Molecular Embryology of Flowering Plants
Molecular Embryology of Flowering Plants V. Raghavan The Ohio State Universi...
354 downloads
2123 Views
18MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Molecular Embryology of Flowering Plants
Molecular Embryology of Flowering Plants V. Raghavan The Ohio State University
CAMBRIDGE UNIVERSITY PRESS
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge CB2 1RP, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, United Kingdom 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Cambridge University Press 1997 This book is in copyright. Subject to statutory exception and to the provision of collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1997 Printed in the United States of America Typeset in Palatino Library of Congress Cataloging-in-Publication Data
Raghavan, V. (Valayamghat), 1931Molecular embryology of flowering plants / V. Raghavan. p. cm. Includes index. ISBN 0-521-55246-X (hardcover) 1. Botany-Embryology. 2. Angiosperms - Embryology I. Title. QK665.R34 1997 96-44410 571.8'62-dc21 A catalog record for this book is available from the British Library
ISBN 0 521 55246 X hardback
CIP
This work is dedicated to the memory of my parents.
Contents Preface / ix Abbreviations I xi List ofcDNA clones, genes, protein products, and mutants / xiii
1 1. 2. 3. 4. 5. 6.
Reproductive biology of angiosperms: retrospect and prospect 1 An overview of angiosperm reproduction / 1 How reproductive biology is studied / 4 Genetic and molecular perspectives / 7 Molecular and genetic engineering techniques / 11 Scope of molecular embryology / 13 Organization of the text / 14
SECTION I GAMETOGENESIS
3. Physiological and biochemical aspects of male sterility / 132 4. Role of the mitochondrial genome in male sterility / 137 5. General comments / 151 6 Megasporogenesis and megagametogenesis 152 1. Carpel determination / 153 2. Megasporogenesis / 154 3. Organization of the megagametophyte / 162 4. Molecular biology of megagametogenesis / 173 5. Nutrition of the megagametophyte / 174 6. Apomixis / 176 7. General comments / 177
15
2 1. 2. 3. 4. 5.
17 Anther developmental biology Anther determination / 17 Anther development / 25 Microsporogenesis /41 Anther culture in the study of meiosis / 57 General comments / 60
3 1. 2. 3. 4.
Pollen development and maturation Morphogenesis of pollen walls / 62 The first haploid mitosis / 75 Vegetative and generative cells / 82 General comments / 89
61
4 Gene expression during pollen development 90 1. Genetic basis for pollen gene expression / 90 2. Nucleic acid and protein metabolism / 94 3. Isolation of genes controlling pollen development / 104 4. Promoter analysis of pollen-specific genes / 112 5. Pollen allergens / 117 6. General comments / 118 5 Pollen abortion and male sterility 120 1. Phenotypic and genotypic aspects of male sterility / 121 2. Structural, developmental, and functional aspects of male sterility / 122
SECTION II POLLINATION AND FERTILIZATION 7 Stigma, style, and pollen-pistil interactions 1. The stigma / 181 2. The style / 192 3. Pollen-pistil interactions / 201 4. General comments / 208 8 1. 2. 3. 4. 5. 9 1. 2. 3. 4.
179 181
In vitro pollen germination and pollen tube growth 209 Pollen germination / 210 Biochemistry of pollen germination / 220 Structure of the pollen tube / 228 Growth of the pollen tube / 240 General comments / 242 Developmental biology of incompatibility 244 Heteromorphic self-incompatibility / 245 Homomorphic self-incompatibility / 251 Cross-incompatibility / 269 General comments / 271
10 Molecular biology of self-incompatibility 1. The Brassica system / 273 2. The Nicotiana system / 284 3. Petunia and other Solanaceae / 288
272
VII
Contents
VIII
4. Papaver rhoeas and other plants / 290 5. General comments / 292 11 Fertilization: the beginning of sporophytic growth
6. Regulation of expression of defenserelated genes / 434 7. General comments / 438 293
1. Cellular nature of the sperm / 294 2. Odyssey of the sperm and double fertilization / 301 3. In vitro approaches to the study of fertilization / 313 4. General comments 317 SECTION III ZYGOTIC EMBRYOGENESIS 12 Developmental biology of the endosperm
1. 2. 3. 4. 5. 6. 7. 8.
319 321
1. Genetic control of embryogenesis / 395 2. Gene expression patterns during embryogenesis / 404 3. Characterization of genes essential for embryogenesis / 413 4. Expression of genes for housekeeping and structural proteins / 422 5. Expression of genes related to desiccation tolerance and dormancy / 426
SECTION IV ADVENTIVE EMBRYOGENESIS
465
16 Somatic embryogenesis
467
1. Historical background / 468 2. Dynamics of somatic embryogenesis / 470 3. Cell biology and physiology of somatic embryogenesis / 475 4. Biochemistry of somatic embryogenesis / 486 5. Regulation of gene expression / 492 6. General comments / 498 17 Embryogenic development of pollen grains
357
The zygote / 358 The multicellular embryo / 361 Cellular aspects of embryogenesis / 371 Tissue and meristem differentiation during embryogenesis / 374 The suspensor / 376 Nutrition of the embryo / 382 Embryo culture / 384 General comments / 393
14 Genetic and molecular analysis of embryogenesis
440
1. Cytology of reserve deposition / 441 2. Regulation of storage protein synthesis / 443 3. Isolation and characterization of storage protein genes / 453 4. General comments / 464
1. Histology and cytology of the endosperm / 322 2. Nutritional role of the endosperm / 327 3. Gene action during endosperm ontogeny / 331 4. Synthesis and accumulation of endosperm storage reserves / 334 5. Analysis of storage protein gene expression / 342 6. Cloning and characterization of endosperm storage protein genes / 350 7. General comments / 356 13 Embryogenesis and physiology of growth of embryos
15 Storage protein synthesis in developing embryos
394
500
1. Historical background / 501 2. Factors controlling pollen sporophytic growth / 505 3. Cytology of pollen embryogenesis / 510 4. Molecular biology of pollen dedifferentiation / 515 5. General comments / 521 SECTION V APPLICATIONS
523
18 Genetic transformation of embryos
525
1. Transformation through pollen grains and ovules / 526 2. Transformation of developing embryos / 527 3. Transformation of embryogenic cells and somatic embryos / 530 4. General comments / 531 References
Index 1669
533
Preface This book is derived from my long-standing interest in the reproductive biology of vascular plants, in particular of the ferns and flowering plants. During the initial planning stages of this venture, the book was intended as a revision of my Experimental
Embryogenesis
in Vascular
Plants
(Academic Press) published in 1976, with the specific aim of describing how our present ideas in molecular and genetic biology apply to embryo development in vascular plants. After I wrote two or three chapters on angiosperm embryogenesis, I began to view the project essentially as a revision of my 1986 book, Embryogenesis in Angiosperms: A Developmental and Experimental Study (Cambridge
University Press), and aborted the idea of covering embryogenesis of ferns and gymnosperms. Reflecting the perspective of current research, it began to dawn on me that molecular and genetic principles underlying embryo development in angiosperms have much in common with principles regulating the whole gamut of reproductive processes in flowering plants. This periodic reshaping and refining of the various drafts during the past five years has resulted in the present volume providing a synthetic review of molecular and cellular aspects of the familiar sequence in the reproductive biology of flowering plants, beginning with the flower and terminating with the embryo. It is thus a wholly new book and has very little to share with its predecessors. My objective in writing this book is to explain the progress achieved toward a molecular understanding of the reproductive processes in angiosperms with particular emphasis on embryology. In reviewing the materials for each chapter, my prime consideration has been to suppress the enthusiasm to cover strictly molecular work, but to provide the necessary structural and developmental framework into which the molecular data are inserted. The view of angiosperm embryology that this book presents, therefore, owes as much to research done during these past few decades as to the research performed during the early part of this century. I have also made an attempt to show how, by a shift in focus and by the application of new and sophisticated techniques, the study of angiosperm embryology
has become a dynamic and exciting discipline. The book is written as a scholarly work of reference for use by researchers interested in the field of angiosperm embryology and as a reference text in graduate level courses on the subject. I have attempted to provide coverage of materials that were published up to mid-1996, when the final draft of the manuscript was made press-ready. I am aware of the fact that some of the topics reviewed in this book have continued to be enlivened by a steady stream of high-quality publications reporting interesting new observations. To incorporate this information in the book at advanced stages of production would have required additions and revisions of a magnitude beyond what could be deemed practical. The book has benefited tremendously from the generosity, cooperation, and help of many professional colleagues and others. I am deeply appreciative to many colleagues who have provided illustrative materials in the form of original pictures or photographs, prints, negatives, or slides from their publications for inclusion in the book. These scientists, listed individually, along with others have provided the necessary stimulus for this book through their devotion to the study of plant embryology. The final draft of the book was read by Professor Abraham D. Krikorian (State University of New York at Stony Brook) and Professor Joseph P. Mascarenhas (State University of New York at Albany). I am thankful to them for their comments and for convincing Dr. Robin Smith, Life Sciences Editor, Cambridge University Press, New York, that the length of the book is one of its strengths. The credit for providing consistency in style and punctuation and for converting my manuscript into the finished product goes to Ms. Dorothy Duncan, the copyeditor at Bookworks. I appreciate her professional expertise and the informed intuitions that she brought to the job. Ms. Elizabeth A. (Bette) Hellinger, administrative secretary of my department, aided me immensely in various ways and spent considerable time to rectify what appeared to be never-ending problems in my word-processing operations. Ms Vicki A. Payne, secretary of the department, provided extensive
IX
Preface
and dependable help in typing and printing. I was really fortunate to have had this kind of assistance in the "front office." At this time, I should also like to acknowledge with gratitude the role played by my teachers, Professor William P. Jacobs (Princeton University) and the late Professor John G. Torrey (Harvard University), in leading me to the study of problems in plant development, of which embryology is just one.
During the gestation period of this book, I was helped by my wife, Lakshmi, who dispensed wisdom with understanding, and by my daughter, Anita, by her sense of humor. Their love and devotion enabled me to complete this work sooner than I expected. V. Raghavan April 1,1997
Abbreviations ABA ACC
= =
ACP ADF ADH ADP AMP ATP ATPase
= = = = = = =
cAMP
=
CaMV CAT
= =
CCC
=
cDNA cms
= =
coA con-A 2,4-D DAPI DCCD DMS DNA DNase ds-DNA dUTP EBN EMS ER ethrel
= = = = = = = = = = = = = =
FeEDTA
=
FITC Gl, G2 GA GTP GTPase GUS H2B histone H3 histone
= = = = = = = =
abscisic acid 1-aminocyclopropane-lcarboxylic acid acyl carrier protein actin-depolymerizing factor alcohol dehydrogenase adenosine diphosphate adenosine monophosphate adenosine triphosphate adenosine triphosphatase or synthase cyclic adenosine monophosphate cauliflower mosaic virus chloramphenicol acetyl transferase 2-chloroethyl-trimethyl ammonium chloride complementary DNA cytoplasmic male sterility; cytoplasmic male steriles coenzyme-A concanavalin-A 2,4-dichlorophenoxyacetic acid 4'6-diamidino-2-phenylindole dicyclohexylcarbodiimide dimethylsulfate deoxyribonucleic acid deoxyrixbonuclease double-stranded DNA deoxyuridine triphosphate endosperm balance number ethyl methanesulfonate endoplasmic reticulum 2-chlorethylphosphonic acid, ethylene-releasing agent ferric ethylenediaminetetraacetic acid fluorescein isothiocyanate growth phases gibberellic acid guanosine triphosphate guanosine triphosphatase (5-glucuronidase lysine-rich histone arginine-rich histone
HSP IAA IBA ICL 2iP
L-PPT L-protein morphactin
=
mRNA MS NAA NAD
NAD-3 NADH NADPH NMR oligo-dT
=
PAS PAT PCR
P-DNA P-grains
=
PHA
PHA-E PHA-L poly(A)-RNA =
heat-shock protein indoleacetic acid indolebutyric acid isocitrate lyase N6(A2-isopentenyl) adenine L-phosphinothricin leptotene protein methyl-2-chloro-9-hydroxy-fluorene-(9)carboxylate messenger RNA malate synthase naphthaleneacetic acid nicotinamide adenine dinucleotide nicotinamide adenine dinucleotide dehydrogenase subunit-3 nicotinamide adenine dinucleotide, reduced nicotinamide adenine dinucleotide phosphate nuclear magnetic resonance oligo-deoxythymidylic acid periodic acid-Schiff reagent phosphinothricin acetyltransferase polymerase chain reaction pachytene DNA premitotic pollen grains phytohaemagglutinin erythrocyte-agglutinating phytohaemagglutinin lymphocyte-agglutinating phytohaemagglutinin polyadenylic acid-containing RNA
poly(U) P-particles PPDK
=
psn-DNA psn-RNA rDNA recA-protein
=
polyuridylic acid polysaccharide-rich particles pyruvate orthophosphate dikinase pachytene small-nuclear DNA sequences pachytene small-nuclear RNA ribosomal DNA DNA-binding or DNA-reassociation protein
Abbreviations
XII
RFLP
=
RH-0007 RH-531
= =
RNA RNase RNase Tl rRNA Rubisco
= =
restriction fragment length polymorphism fenridazon sodium-l(p-chlorophenyl)-l,2dihydro-4,5 dimethyl-2oxonicotinate ribonucleic acid ribonuclease
=
RNase from Aspergillus oryzae
= =
ribosomal RNA ribulose-l,5-bisphosphate carboxylase
SC-1058, SC-1271
=
SDS
=
phenylcinnoline carboxylate compounds sodium dodecyl sulfate
SDS-PAGI S-period (phase) ss-DNA 2,4,6-T T-DNA tRNA UDP U-protein WGA Z-layer zyg-DNA zyg-RNA
sodium dodecyl sulfate polyacrylamide gel electrophoresis DNA synthetic period single-stranded DNA 2,4,6-trichlorophenoxyacetic acid transferred DNA transfer RNA uridine diphosphate DNA-unwinding protein wheat germ agglutinin Zwischenkorper or oncus zygotene DNA zygotene DNA transcript
List ofcDNA Clones, Genes, Protein Products, and Mutants [Abbreviations and names of cDNA clones and genes are presented here and in the text in italicized capital letters; mutants are indicated in italicized lowercase letters. Protein products are given in capital letters. In a few cases, the convention used by the authors is followed. Abbreviations for the same gene, protein product, and mutation are not listed separately.) AS, A-6, A-8, A-9
=
aba
=
abi
=
ABR
ac
=
ACT-1 ae
=
ag
AG-1
agl al AMBA-1 AMBtV ant
ap
=
tapetum-specific cDNA clones from Brassica napus ABA-deficient mutant seed of Arabidopsis thaliana ABA-insensitive mutant seed of Arabidopsis thaliana protein induced by ABA in embryos of Pisum sativum activator, mobile mutator element of Zea mays actin gene of Oryza sativa amylose extender mutant of Zea mays agamous mutant of Arabidopsis thaliana flower protein that binds to ATrich sequences of phaseolin gene promoter agamous-like mutant of Arabidopsis thaliana flower viviparous mutant of Zea mays pollen allergen of Ambrosia artemisiifolia gene for allergen from Ambrosia trifida aintegumenta, ovule mutation of Arabidopsis thaliana flower apetala mutant of Arabidopsis thaliana flower
APG
=
ARK-1
=
AT2S
=
ATEM-1
=
ATP-6
=
ATP-9
=
ATP-A ats
=
AX-92
=
AX-110
=
B11E
=
B-19
=
B22E
=
B-32, B-70
=
BA-112, BA-118
=
BAG-1
=
BAR
=
BARNASE
=
BARNSTAR
=
anther-specific cDNA clone from Arabidopsis thaliana receptor protein kinase gene from Arabidopsis thaliana albumin gene from Arabidopsis thaliana LEA gene of Arabidopsis thaliana mitochondrial subunit-6 of Fj-ATP-synthase mitochondrial subunit-9 of Fj-ATP synthase a-subunit of Fj-ATP synthase aberrant testa shape mutant of Arabidopsis thaliana ovule axis-abundant gene of Brassica napus embryos cDNA clone from somatic embryos of Daucus carota aleurone-specific cDNA clone from Hordeum vulgare LEA gene family of Hordeum vulgare cDNA clone from the embryo and endosperm of Hordeum vulgare proteins from Zea mays endosperm tapetum-specific cDNA clones from Brassica napus MADS-box gene from Brassica napus gene encoding phosphinothricin acetyltransferase gene for RNase from Bacillus amyloliquefaciens gene that counteracts BARNASE
XIV
ListofcDNA Clones, Genes, Protein Products, and Mutants
BCP-1
pollen-specific cDNA clone from Brassica campestris bel-1 bell mutant of Arabidopsis thaliana ovule BETV-I pollen allergen of Betula verrucosa genomic clone of BCP-1 BGP-1 from Brassica campestris basic region/helix-loopbHLH helix protein bio-1 biotin auxotrophic mutant of Arabidopsis thaliana embryo BiP immunoglobulin heavychain-binding protein BmT or HmT = toxin produced by the fungus Bipolaris (Helminthosporium) maydis hno = defective kernel (embryo) mutant of Zea mays BP-4, BP-10, BP-19 = pollen-specific genomicclones from Brassica napus corresponding to cDNA clones BP-401, BP-408, and BP-405, respectively BP-4A, BP-4B, BP-4C = pollen-specific cDNA clones of BP-4 family from Brassica napus BP-401, BP-405, pollen-specific cDNA BP-408 clones from B. napus inbred line of Zea mays BSSS-53 brittle endosperm bt mutant of Zea mays domain of a leucine zipbZIP per DNA-binding protein C-1 gene in the anthocyanin pathway of Zea mays C-1, C-2 callus-specific proteins of somatic cells of Daucus carota pollen-specific cDNA 3C-12 clone from Zea mays cDNA clone from C-98 anthers of Brassica napus DNA-binding protein of CA-1 phaseolin gene chlorophyll a/b-binding CAB protein CAN protein that binds to
CAT CDC-2
cDNA clone 17 cDNA clone 9622
=
cDNA clones 92-b, 108,127-A
CDPK
CELP
CEM-1, CEM-6 CHI CHI-A 4-CL
CLV-3
CMS-3 cms-C
=
cms-S
=
cms-T
=
COB COR-6.6
COX-I, COX-II
motifs in the phaseolin gene CATALASE gene from Zea mays cell cycle control protein kinase gene from Arabidopsis thaliana and Zea mays pollen-specific cDNA clone from Brassica napus cDNA clone expressed in the style of Lycopersicon esculentum tapetum-specific cDNA clones from Lycopersicon esculentum gene for calmodulindependent protein kinase from pollen of Zea mays gene for cysteine-rich, extensinlike proteins from flowers of Nicotiana tabacum cDNA clones from embryogenic cells of Daucus carota gene for chitinase gene encoding chalcone flavanone isomerase gene that encodes parsley 4-coumarate:coenzyme A ligase CLAVATA3, gene for shoot and floral meristem development in Arabidopsis thaliana cytoplasmic male-sterile line of Helianthus annuus cytoplasmic male-sterile Charrua line of Zea mays cytoplasmic male-sterile U.S. Department of Agriculture line of Zea mays cytoplasmic male-sterile Texas line of Zea mays apocytochrome-6 COLD-REGULATED gene induced by cold temperature in Arabidopsis thaliana cytochrome oxidase-c subunits
List of cDNA Clones, Genes, Protein Products, and Mutants
cp
CRU CKY-1A
CVC-A CYCl-At
cyd
=
D-7, D-ll, D-19 DC-2, DC-3, DC-5, DC-7, DC-8, DCS, DC-13, DC-59
de-17 de*-B30
def dek DLEC ds DT-A E-l, E-2
E-2 ECP-31, ECP-40
EF-1
EM
emb EMB-1
=
defective kernel (embryo) mutant of Zea mays gene for the storage protein cruciferin gene encoding a protein derived from Bacillus thuringiensis gene for the storage protein convicilin cell cycle control cyclin gene from Arabidopsis thaliana cytokinesis-defective mutant of Pisum sativum LEA proteins of Gossypium hirsutum
cDNA clones from embryogenic cells of Daucus carota defective seed mutant of Zea mays defective endosperm-B30 mutant of Zea mays deficiens mutant of Antirrhinum majus flower defective kernel mutant of Zea mays gene encoding lectin in Phaseolus vulgaris dissociation transposable element diphtheria toxin chain-A embryonic proteins of somatic cells of Daucus carota cDNA clone from anthers of Brassica napus cDNA clones from embryogenic cells of Daucus carota elongation factor-la EARLY METHIONINE-labeled gene from Triticum aestivum embryos embryo-specific mutant of Zea mays cDNA clone from embryogenic cells of Daucus carota
EMB-30
=
EMB-564
=
eml-1
=
EP-1, EP-2, EP-4
-
ett
=
EXOPG
=
F2S
=
FBP
=
fdh
=
fie
=
fk
=
fl
=
flo
=
FR, FR-2
=
FS
=
fus
=
G-l G-10
= =
G-box
=
G-box/C-box
=
embryo-lethal gene (112A) of Arabidopsis thaliana LEA gene from Zea mays embryos embryoless mutant of Oryza sativa extracellular proteins secreted by embryogenic or nonembryogenic cells of Daucus carota ettin mutation of the gynoecium of Arabidopsis thaliana gene for exopolygalacturonase from Zea mays pollen-specific cDNA clone from Brassica napus FLORAL BINDING PROTEIN gene of Petunia hybrida ovule fiddlehead mutant of Arabidopsis thaliana fertilization-independent endosperm mutant of Arabidopsis thaliana fackel pattern mutant of Arabidopsis thaliana embryo homeotic floral mutant of Arabidopsis thaliana; also floury endosperm or embryo mutant of Zea mays floral mutant of Arabidopsis thaliana flower nuclear restorer genes for restoring fertility in Phaseolus vulgaris FASS gene that affects embryo shape in Arabidopsis thaliana fusca seedling mutant of Arabidopsis thaliana gene for glycinin 1 genomic clone of the pollen-specific cDNA clone TP-10 from Nicotiana tabacum CACGTG motif of the phaseolin gene GCCACGTCAG or CACACGTCAA motifs of the phaseolin gene
List of cDNA Clones, Genes, Protein Products, and Mutants
GCN4
gH2B
gH3
=
= =
gib-1
=
GLB
=
glo
=
GLTP
=
8n
=
GOS-2
=
G-protein GT
= =
Su
-
GY
=
H3-1
=
HA
=
HAG
=
HBP-1
=
HMG-2
=
HmT or BmT
=
(for General Control Nonderepressible); a class of mammalian and yeast regulatory proteins variant lysine-rich histone of Lilium longiflorum pollen nuclei variant arginine-rich histone of Lilium longiflorum pollen nuclei gibberellin-deficient mutant of Lycopersicon esculentum gene for globulin storage protein globosa mutant of Antirrhinum majus flower gene for nonspecific lipid transfer protein from ray florets of Gerbera hybrida gnom pattern mutant of Arabidopsis thaliana embryo Oryza sativa genomic clone involved in the initiation of translation GTP-binding protein glutelin gene from endosperm of Oryza sativa gurke pattern mutant of Arabidopsis thaliana embryo gene for glycinin storage protein of Glycine max embryo histone H3 gene from Medicago sativa parent cell line of Daucus carota gene for helianthinin of Helianthus annuus embryos Triticum aestivum histone DNA-binding protein gene encoding 3hydroxy-3-methylglutaryl coenzyme A reductase toxin produced by the fungus Bipolaris (Helminthosporium) maydis
HSP-17, HSP-18.2, HSP-70, HSP-81, HSP-82, HSP-90 = 1-3
1-49
ig Iwt
JIM-5 JIM-7 JIM-8 JUN KBG
KEU
KIN-1 KN
KN-1 KNF
KT1 L-3
heat-shock protein genes pollen-specific cDNA clone from Brassica napus protein involved in the resistance of pea pods to plant pathogens indeterminate gametophyte mutant of Zea mays conserved region in the promoter of an ABAresponsive gene monoclonal antibody for unesterified pectin monoclonal antibody for esterified pectin monoclonal antibody for arabinogalactans JUNNA, a class of yeast regulatory proteins cDNA clone of allergenic protein of Poa pratensis pollen KEULE gene that affects embryo shape in Arabidopsis thaliana gene induced by cold temperature in Arabidopsis thaliana KNOLLE gene for pattern formation in Arabidopsis thaliana embryo KNOTTED gene of Zea mays KNOPF gene that affects embryo shape in Arabidopsis thaliana gene for Kunitz trypsin inhibitor gene for 16-kDa oleosin isoform of Zea mays
LAT-51, LAT-52, LAT-56, LAT-58, LAT-59
LE LEA LEB
pollen-specific cDNA clones from Lycopersicon esculentum lectin gene from embryos of Glycine max late embryogenesis abundant gene gene for the storage protein legumin, B-type
List ofcDNA Clones, Genes, Protein Products, and Mutants
lee LEG LELB
Ify L1M LIM
LIM-15 LMP-131A, LMP-134 LOLP LTP
lug MA MA-12 MAC-207 MADS-box MB me MCM
MHSP18-1
leafy cotyledon seedling mutant of Arabidopsis thaliana genes for the storage protein legumin, A- and B-types gene for the storage protein legumin, B-type leafy mutant of Arabidopsis thaliana flower meiotic-specific gene from microsporocytes of Lilium longiflorum (for LIN-U, 1SL-1, and MEC-3 genes) a conserved motif of metalbinding, cysteine-rich proteins of animals bacterial reassociation protein cDNA clones expressed in mature pollen grains of Lilium longiflorum cDNA clone of allergenic protein of Lolium perenne pollen for L1PID TRANSFER PROTEIN, aleurone-specific genomic DNA clone from Hordeum vulgare leunig mutant of Arabidopsis thaliana flower a myrosinase gene from embryos of Sinapis alba LEA gene from Zea mays embryos, now referred to as RAB-17 monoclonal antibody for arabinogalactans functional motif shared by MCM-1, AG, DEF, and SRF genes myrosinase gene from embryos of Sinapis alba mucronate endosperm mutant of Zea mays gene that causes defects in the maintenance of minichromosomes of yeast heat-shock protein gene from Zea mays
XVII
MIC MLG-3 mn Mo-17 mp ms-4 ms-H mu MYB
MYC
NAG-1 NAM
NE1F-4A8 NOD
NOS
=
NPG-1 NPT1I
=
NTGLO
=
NTP-303
=
ogu
=
OHP-1
=
MICKEY gene that affects embryo shape in Arabidopsis thaliana LEA protein of Zea mays embryos miniature seed endosperm mutant of Zea mays inbred line of Zea mays monopteros pattern mutant of Arabidopsis thaliana embryo male-sterile mutant of Glycine max male-sterile mutant of Arabidopsis thaliana mutator transposon of Zea mays recognition site in the genome, identified with myeloblastosis-associated viruses recognition site in the genome, associated with myelocytomatosis-associated viruses MADS-box gene from Nicotiana tabacum NO APICAL MERISTEM gene required for shoot meristem formation in Petunia sp. embryo pollen-specific cDNA clone from Nicotiana tabacum nodulation gene of Rhizobium leguminosarum gene for nopaline synthase cDNA clone highly expressed in Nicotiana tabacum pollen gene encoding neomycin phosphotransferase MADS-box gene from Nicotiana tabacum pollen-specific cDNA clone from Nicotiana tabacum ogura cms line of Brassica napus and Raphanus sativus protein product of a Zea mays cDNA clone
XVIII
List of cDNA Clones, Genes, Protein Products, and Mutants
MADS-box gene from the orchid, Aranda sp. op opaque endosperm mutant of Zea mays OP-2 MODIFIER gene that modifies op-2 mutation open reading frames of ORF-13, ORF-25 TURF-13 gene associated with cms-T line of Zea mays ORF-138 open reading frame associated with cms in Brassica napus and Raphanus sativus open reading frame ORF-522 associated with cms in Helianthus annuus 0SG-6B Tapetum-specific genomic clone from Oryza sativa anthers OSMADS MADS-box gene from Oryza sativa ovu = ovulata mutant of Antirrhinum majus flower P-2 pollen-specific cDNA clone from Oenothera organensis p34 c d c 2 yeast cell division cycle phosphoprotein conserved sequence in PA the promoter of an ABAresponsive gene promoters of a gene for PA-1, PA-2 chalcone synthase isomerase from Petunia hybrida PAL-2 gene for phenylalanine ammonia lyase from Phaseolus vulgaris PCBC-3 monoclonal antibody for arabinofuranosyl residues Petunia fusion gene, PCF mitochondrial fusion gene associated with cms phytohaemagglutinin PDLEC gene of a lectin-nonproducing mutant of Phaseolus vulgaris pEMB-4, pEMB-94 = embryoid-specific cDNA clones from cultured anthers of Triticum aestivum OM-1
PEX-1 pi
pin-1 pie Pm
PMA-2005 PMAH-9 po
pop PPE-1 PREM PRK-1 prl
PS-1 PS-IB
PSB-A ptd PVALF
PVS
pZE-40
pollen-specific cDNA from Zea mays pistillata mutant of Arabidopsis thaliana flower pin-formedfloralmutant of Arabidopsis thaliana plena mutant of Antirrhinum majus flower toxin produced by the fungus Phyllosticta maydis (teleomorph Mycosphaerella zeaemaydis) LEA gene from embryos of Triticum aestivum LEA gene from embryos of Zea mays a recessive mutation of Petunia hybrida mutation in Arabidopsis thaliana defective in pollen-pistil interactions pollen-specific cDNA clone from Petunia inflata pollen retroelement of Zea mays pollen-specific cDNA clone from Petunia inflata prolifera mutant of Arabidopsis thaliana affecting the embryo sac gene highly expressed in Oryza sativa pollen cDNA clone encoding S-associated proteins from S-l allele of Petunia hybrida 32-kDa quinone-binding protein associated with photosystem II defective kernel (embryo) mutant of Zea mays gene for Phaseolus vulgaris ABAI-3-like factor from embryos of Phaseolus vulgaris mitochondrial DNA sequence associated with cms in Phaseolus vulgaris embryo-specific cDNA clone from Hordeum vulgare
List ofcDNA Clones, Genes, Protein Products, and Mutants
r,rb
=
RAB
=
RAD-51 RBC-L
= =
RBC-S
=
re
=
RF-1, RF-2, RF-3, RF-4
=
rgh
=
Rho
=
Ri
Rise-56, Ris0-15O8
rugosus, genetic loci that control wrinkled condition in Pisum sativum seeds RESPONSIVE TO ABA protein synthesized in response to ABA application bacterial reassociation protein large subunit of Rubisco small subunit of Rubisco
S-1, S-2, S-3, S-5, S-6, S-7, S-8, S-9, S-12, S-13, S-14, S-22, S-29, S-63
Brassica, Nicotiana, and Petunia Sa, Sf-11, Sz
S-alleles identified in
SBH-1
Nicotiana alata Glycine max homeobox-
reduced endosperm mutant of Zea mays
se, seg
nuclear restorer genes for fertility restoration in
SEC-7
cms Zea mays
SEF
campestris shrunken endosperm mutants of Hordeum vulgare
secretory protein of yeasi soybean embryo factor, a protein that binds to the conglycinin gene of
defective kernel (embryo) mutant of Zea mays protein factor from
sex
=
causes the termination and release of RNA molecules during transcription in vitro root-inducing plasmid
SF-3, SF-16, SF-17
=
=
of Agrobacterium rhizogenes mutants of Hordeum vul-
gare defective in hordein accumulation transcription factor from Oryza saliva endosperm ROOT LOCI genes for the hairy root of
S-gene sh
=
sin-1
=
ROL-B, ROL-C
= =
RRN-26 RTBV
=
RTL
=
=
=
=
gene for small ribosomal subunit protein-12 gene encoding 26S rRNA gene of RICE TUNGRO BACILLIFORM VIRUS ROOTLESS gene of Arabidopsis thaliana
embryo 5'-CATGCAT-3', 5'-CATGCAC-3', and 5'-CATGCATG-3' motifs found in the glycinin gene linear episomal DNAs of cms-S line of Zea mays
=
Glycine max sepaloidea mutant of Antirrhinum majus
flower
sl-2
Agrobacterium rhizogenes
S1,S2
containing gene mutation from selfincompatibility to compatibility in Brassica
scf-1
sep
=
RY-repeats
S-alleles identified in
Escherichia coli that
RITA-1
RPS-12
XIX
SLA SLG SLR sp-1
spm
shrunken endosperm expressing xenia mutant of Zea mays
pollen-specific cDNA clones from Helianthus annuus sterility gene shrunken endosperm mutant of Zea mays short integuments mutant of Arabidopsis thaliana stamenless-2 mutant of Lycopersicon esculentum
flower S-LOCUS ANTHER gene S-LOCUS GLYCOPROTEIN gene S-LOCUS-RELATED gene small pollen mutant of Zea mays suppressor mutator,
mobile mutator of Zea mays
SRF SRK
serum response factor from vertebrates S-LOCUS RECEPTOR KINASE gene
List of cDNA Clones, Genes, Protein Products, and Mutants
XX
S-RNase
=
STYLAR RNase, a product of the S-gene
TRP-E
=
STA39-3, STA39-4
=
ts
=
STA44-4
=
ts-llc, ts-59, ts-85
=
STIG-1
=
STM
=
late pollen genes from Brassica napus gene highly expressed in Brassica napus pollen stigma-specific gene from Nicotiana tabacum pistil SHOOT MERISTEMLESS gene required for shoot meristem formation in Arabidopsis thaliana embryo pistil-specific genes from Solanum tuberosum mutation from selfincompatibility to compatibility in Brassica oleracea; also sugary endosperm mutant of Zea mays superman mutant of Arabidopsis thaliana flower suspensor embryo-lethal mutant of Arabidopsis thaliana SUCROSE SYNTHASE gene of Zea mays male-fertile mutant of Zea mays
tu
=
TUA-1
=
TUBa
=
TUBS, TUB-4, TUBS
=
TURF-13
-
twn
=
STS-14, STS-15
=
su
=
sup
=
sus
=
SUS
=
t-4
=
TA-13, TA-26, TA-29, TA-32, TA-36
=
TA-20, TA-56
=
Ti plasmid
=
TM-6 TP-10
=
TPC-44, TPC-70
=
TR-2'
=
tapetum-specific cDNA clones from Nicotiana tabacum cDNA clones from Nicotiana tabacum anthers tumor-inducing plasmid of Agrobacterium tumefaciens MADS-box gene from Lycopersicon esculentum pollen-specific cDNA clone from Nicotiana tabacum pollen-specific cDNA clones from Tradescantia paludosa nannopine synthase gene from Agrobacterium tumefaciens
ufo URF-13
=
URF-RMC
=
VIR
=
vp w-3
= =
WGA-B
=
WUN
=
WUS
=
wx y-9
= =
YEC-2 ZAG
=
gene for tryptophan biosynthesis from Escherichia coli tassel-seed mutant of Zea mays temperature-sensitive mutant cell lines of Daucus carota impaired in somatic embryogenesis tunicate mutant of Zea mays oc-tubulin gene from Arabidopsis thaliana a-tubulin gene from Zea mays (3-tubulin genes from Zea mays gene associated with male sterility in cms-T line of Zea mays twin embryo-lethal mutant of Arabidopsis thaliana universal floral organs mutant of Arabidopsis thaliana flower 13-kDa protein associated with cms-T line of Zea mays chimeric gene associated with cms in Oryza sativa VIRULENCE gene of Agrobacterium tumefaciens viviparous mutant viviparous mutant of Zea mays gene encoding wheat germ agglutinin isolectin B gene encoding woundinducible protein in potato tuber WUSCHEL gene required for shoot meristem formation in Arabidopsis thaliana embryo waxy mutant viviparous mutant of Zea mays gene from yeast MADS-box gene from Zea mays
List of cDNA Clones, Genes, Protein Products, and Mutants
ZAP-1
=
Z-locus
=
ZM-58.1, ZM-58.2
=
ZMABP-1, ZMABP-2, ZMABP-3
=
XXI
MADS-box gene from Zea mays locus unlinked to S-locus that controls self-incompatibility in members of Poaceae pollen-specific cDNA clones from Zea mays
ZMC-13, ZMC-26, ZMC-30
ZMG-13
genomic clone of pollen-specific cDNA clone ZMC-13 from Zea
eenes from Zea mays pollen that encode actindepolymerizing factors
ZMPK-1
receptor protein kinase gene from Zea mays
ZMEMPR-9
pollen-specific cDNA clones from Zea mays heat-shock protein gene from Zea mays
mays
Chapter 1
Reproductive Biology of Angiosperms: Retrospect and Prospect An Overview of Angiosperm Reproduction (i) Sexual Reproduction (ii) Asexual Reproduction How Reproductive Biology Is Studied (i) Developmental Cytology (ii) Genetics Genetic and Molecular Perspectives (i) Genes and Flowering (ii) Gene Expression during Anther and Pollen Development (iii) Embryo-specific Genes (iv) Gene Activity during Fruit Development and Ripening Molecular and Genetic Engineering Techniques Scope of Molecular Embryology Organization of the Text
Recent advances in molecular biology and genetic engineering have brought new and powerful methodologies to bear upon investigations into the reproductive biology of flowering plants or Angiospermae (angiosperms). These studies, which have been undertaken using a few model systems, have provided novel insights into the role of genomic information during a dynamic phase in the life of higher plants. Naturally, the modern approaches owe their origin to foundations laid in the past; taken together, the old and the new studies have led to the conclusion that the steps in the reproductive biology of plants are integrated systems of changes that occur at levels ranging from morphological to molecular. The purpose of this introductory chapter is to highlight the impact of novel conceptual approaches as well as emerging technologies in unraveling the complex mechanisms that underlie the various phases of sexual and asexual reproduction in angiosperms. Since most of our food and many other natural commodities are generated directly or indirectly from angiosperms, these studies collectively
hold great promise of opening new frontiers to improve our commercial exploitation of flowering plants. Thus, the story to be told in the following chapters is as intellectually exciting as it is economically important, with many ramifications of a practical nature. For purposes of orientation, we will begin with a brief account of the flower and the processes leading to gamete, zygote, seed, and fruit formation. The detailed structural changes that occur during these processes and the underlying mechanisms will be considered later at appropriate places in the book. Because the topics covered in this chapter are treated in a general way, no attempt has been made to key specific statements to books or to journal articles. 1. AN OVERVIEW OF ANGIOSPERM REPRODUCTION Angiosperms constitute the division Anthophyta, which is separated into two large classes, Dicotyledonae (dicotyledons or dicots) and Monocotyledonae (monocotyledons or monocots). As the names imply, dicots have two cotyledons in their embryos, whereas just one cotyledon is characteristic of monocot embryos. Reproductive potential of almost all angiosperms is due to sexual or asexual processes or a mixture of both. The future of research in angiosperm reproduction has never been brighter, and the prospect of introduction of useful foreign genes into flowering plants to subvert the reproductive processes and modify the progeny to our advantage appears a distinct reality. (i) Sexual Reproduction
Angiosperms are by far the most diverse and widespread of all plants and constitute the domi-
Reproductive Biology ofAngiosperms:
nant vegetation of the land. Among the features attributed to the ecological spread of angiosperms are their unique sexual reproductive structure, the flower, and the fruit that is born out of it. The conventional flower is an abbreviated shoot with four whorls of modified leaves constituting the sterile and the fertile parts. The sterile parts are an outer whorl of sepals, which are usually green and enclose the rest of the flower before it opens, and an inner whorl of brightly colored petals, which aid in attracting insects and other pollinators. The fertile organs of the flower directly concerned with sexual reproduction are the stamens, representing the male units, and the carpels (or pistil, consisting of one carpel or of several), representing the female units. These four whorls produced by the floral meristem or the receptacle are precisely determined according to a genetic blueprint characteristic of each species. The stamen consists of a threadlike stalk known as the filament, which bears at its distal end four sporangia held together in a swollen portion, the anther. While the flower is developing in the bud, microspores or pollen grains are formed in the sporangia by meiotic division of diploid microsporocytes (microspore mother cells). A subsequent step in the differentiation of the pollen grain gives rise to a large vegetative cell and a small generative cell. While these changes occur within the pollen grain, a two-layered wall, the outer layer of which (exine) is often finely sculptured, develops around each pollen grain. The final step in the differentiation of the pollen grain is the division of the generative cell to form two sperm cells. The mature pollen grain enclosing the two sperm constitutes the male gametophyte; the complex of the vegetative cell and the two sperm is christened the male germ unit. The essential parts of the carpel are an enlarged basal portion known as the ovary and a distal receptive surface known as the stigma; these two parts are often separated by a robust stalk known as the style. The ovary contains ovules, which enclose a mass of undifferentiated cells or the nucellus. While the male gametophyte is being fabricated, a parallel process leading to the formation of the female gametophyte is orchestrated in the ovule. Within each ovule, the megasporocyte or megaspore mother cell, which is differentiated by the prior division of a nucellar cell, undergoes meiosis to form four haploid megaspores. Not long after they are carved out, three of these cells disintegrate, and the fortunate one that survives enlarges to form the embryo sac. As a result of three sequential mitotic divisions of the nucleus of the surviving
megaspore followed by wall formation, the mature embryo sac comes to enclose six haploid cells: an egg, two synergids, and three antipodals. There are also two free haploid nuclei within the embryo sac that fuse to form a diploid polar fusion nucleus, now confined to the remaining part of the embryo sac known as the central cell. The mature embryo sac, consisting of seven cells, represents the female gametophyte. Of these cells, the egg and the polar fusion nucleus persist in the embryo sac, whereas the synergids and antipodals have a transitory existence. In analogy to the male germ unit, the egg, synergids, and the central cell are collectively designated as the female germ unit, but the term has not been really caught on. Fertilization of the egg is preceded by the transfer of pollen grains from the anthers of one flower to the stigmatic surface of another flower on the same or a different plant and by the subsequent germination of pollen grains. Upon germination, a pollen grain develops a tube that pushes through the style and the tissues of the ovary to reach the vicinity of the embryo sac, where the sperm cells are discharged. "Double fertilization" in angiosperms is a unique process that has no match in other divisions of the plant kingdom, although some genera in the gymnosperms are believed to come close. Double fertilization occurs when one sperm fuses with the egg and a second sperm fuses with the polar fusion nucleus. These two fusion events result in the formation of a diploid zygote and a triploid primary endosperm nucleus, respectively. The zygote is the first cell of the sporophytic generation. It displays a rapid burst of mitotic activity to develop into the embryo, whereas the primary endosperm nucleus differentiates into a nutritive tissue known as the endosperm. Externally, the striking changes in the embryo during this time are caused by the formation of embryonic organs like the root, hypocotyl, cotyledons, and shoot. At the cellular level, during their growth both the endosperm and parts of the embryo begin to store massive amounts of nutrients such as carbohydrates, proteins, and fats. The ovule enclosing the developing embryo and endosperm is transformed into a seed contained within the ovary. A ripened ovary becomes the fruit, which protects the seeds and aids in their dispersal. The mature, full-grown embryo enclosed within the seed is the structural link between the completed gametophytic generation and the future sporophytic generation of the plant. Two fundamentally distinct, but interdependent, phenomena characterize the fate of the embryo in the seed. One
Retrospect and Prospect
is the capacity of the embryo to remain metabolically inactive in a quiescent or dormant state spanning a sizable segment of an unfavorable season or until such time that an essential environmental condition is met. Second is the reactivation of growth during germination leading to sweeping structural changes that result in a seedling plant. The seedling matures into the adult, which eventually flowers and sets seeds for the next generation to continue the life cycle. (ii) Asexual Reproduction
from a rapidly proliferating callus generated from the secondary phloem of Daucus carota (carrot; Apiaceae) were nurtured in a nutrient broth, they gave rise to organized growth through units that resembled various stages of embryogenic development of the zygote. Because the embryolike structures (embryoids) have their origin in somatic cells, as opposed to germ cells, the term "somatic embryogenesis" has been frequently applied to this phenomenon. Normal plants are easily regenerated from embryoids by simple manipulative methods, such as transferring them to a medium of a different composition or nurturing them under high humidity in soil or in vermiculite. A simmering controversy about the precise origin of embryoids namely, whether they are from single cells or from cell clumps - was laid to rest with the unequivocal demonstration that in different plants, single callus cells carefully nurtured in isolation give rise to complete mature plants by organogenesis or embryogenesis. Thus, activation of a single cell to choose the embryoid pathway may significantly resemble the triggering of the egg to embryogenic divisions by fertilization.
Propagation of plants by cuttings of stems and leaves and by underground organs like roots, tubers, and rhizomes is an age-old horticultural practice. When nurtured under appropriate conditions of moisture and humidity, a favorable temperature, and the normal composition of the atmosphere, meristematic regions of the transplanted organs regenerate new plants, which continue the life cycle exactly the same way as did the original plant. A well-known natural phenomenon in angiosperms leading to the generation of new plants is adventive or asexual embryogenesis, that is, the development The challenge is to apply these technologies adeof embryo-like units that are not products of sexual quately and economically to crop improvement. recombination. Obviously, when plants reproduce Methods are at hand for the long-term cryopin the absence of sex, there is no sorting of genes. reservation of nonembryogenic and embryogenic Although many plants frequently reproduce asexu- callus cells and for retrieving them later to induce ally from vegetative organs or, rarely, reproduce plant regeneration. Storage of embryoids under from a variety of cells and tissues like the nucellus, ordinary conditions may be achieved by encapsusynergids, and antipodals of the embryo sac under lating them in synthetic seed coats that allow natural conditions, it was the development of tissue embryoids to be handled more or less like convenculture methods that brought to light the startling tional seeds. Many attempts have been made to regenerative powers of plant cells, tissues, and achieve this, but the results have for the most part organs. Early investigations showed that when tis- been equivocal due to limitations in the developsues such as the pith of Nicotiana tabacum (tobacco; ment of appropriate coating systems and the low Solanaceae)1 stem were cultured aseptically in a quality of embryoids and seedlings produced. defined medium, they grew as disorganized masses Overcoming or bypassing the technical problems is of cells known as callus. By subtle manipulations of central to the further use of this method for rapid, growth hormones, such as auxins and cytokinins, large-scale vegetative propagation of plants from supplied to the medium, it was possible to induce a embryoids. pervasive state of growth of the callus, resulting in By employing a mixture of enzymes that the regeneration of shoot buds. At this stage it was hydrolyze cell wall components, it is possible to necessary only to remove the buds from the original isolate protoplasts from cells; these naked cells remedium and transplant them onto a maintenance form walls and attain respectability, divide to form medium of a different composition or to vermiculite calluses, and eventually regenerate plantlets by or soil to induce rooting and formation of normal organogenesis or by embryogenic episodes. In plants in full multicellularity, sexuality, and strucextensions of this work along several lines, ture. In an extension of this work, it was found that progress has been made that permits production of when single cells and clusters of cells sloughed off triploid hybrids by the fusion of gametic and somatic protoplasts, isolation of generative cells ' When first mentioned in the book, the binomials are and sperm cells from pollen grains, microinjection followed in parentheses by the common name and/or of sperm cells into embryo sac, fusion of isolated the family name of the plant.
Reproductive Biology of Angiosperms:
sperm and egg cells, and isolation of viable embryo sacs and their component cells and protoplasts. Protoplasts are the choice systems for fusion experiments, because cells that retain their walls would not be able to fuse. In yet another unexpected discovery, it was shown that when anthers of certain plants excised at an appropriate stage of development are cultured in a medium supplemented with growth hormones and other additives, a small number of the enclosed pollen grains dedifferentiate and develop into embryoids and plantlets with the haploid or gametic number of chromosomes. Confirmation of the ploidy of the regenerants is an important control in these experiments, since it provides direct support for the origin of plants from pollen grains. Embryogenic development of pollen grains has proved to be a boon for geneticists and plant breeders, because haploid plants can be obtained in quantity by anther culture techniques and transformed into true-breeding isogenic diploid lines by simple procedures. From anther culture has developed the free pollen culture system designed to control the development of pollen grains along either the gametophytic or the sporophytic pathways. Although the pollen grain is normally programmed for terminal differentiation, under culture conditions it follows a sporophytic type of program, demonstrating that the nucleus retains a developmental potential that vastly exceeds its normal fate. In a general sense, these observations and those demonstrating the formation of plants from single somatic cells carry implications far beyond asexual reproduction; they support the concept of totipotency, that is, all plant cells, except those that are committed to irreversible differentiation, possess the ability to reprogram their genome and regenerate a normal plant. Finally, in the context of direct programming of specific cells into new developmental pathways, the induction of flower buds from identifiable cells of cultured explants is of great interest. When surface layer explants consisting of only the epidermal and subepidermal cells derived from the inflorescence stalk of tobacco are cultured, floral meristems and fertile flowers are produced by divisions of the subepidermal cells without intermediate callus formation, and they appear outside on the surface of the explant. It is appropriate to emphasize here that the nonmeristematic cells of the explant reconstitute the shape and the cell, tissue, and organ specificities that we recognize in a flower without the benefit of internal control mechanisms that operate in the shoot apex where flowers normally appear.
These nonmeristematic cells of the explant should obviously contain the genetic information for the initiation and formation of floral meristems and their differentiation into flowers. The diverse ways of regenerating plants without involving sexuality and of regenerating flowers capable of sexuality from differentiated cells indicate that cells of the angiosperm plant body are in an unstable state and do not retain their existing state of differentiation under experimental conditions. The reproductive biology of flowering plants described in this section involves several independent developmental systems suitable for analysis at various levels. As we shall see in the chapters that follow, each phase of angiosperm reproduction is a clearly ordered process whereby the structural and functional organization of the participating cells becomes expressed and integrated. This provides the necessary continuum for the biological processes of reproduction. 2. HOW REPRODUCTIVE BIOLOGY IS STUDIED Much of our basic knowledge of reproduction of angiosperms has been garnered from studies at the morphological, histological, cytological, and physiological levels. For the most part, achievements made thus far in the study of floral morphology and pollination biology have not required an array of sophisticated instruments and equipment or specialized facilities. Especially during the early part of this century, plant scientists acquired new and unprecedented opportunities to explore the flora in the different parts of the world, conduct field observations, collect plants, and even cultivate them in new habitats. These studies, which were directed to other objectives, have proved to be invaluable to the basic science of plant reproductive biology. Investigations of this nature are being continued even to this day. Not only have these endeavors allowed researchers to probe the mysteries of sex in diverse groups of plants, but they have also opened up new approaches for developing economically important strains of some of our cultivated plants and for preserving their wild relatives. (i) Developmental Cytology It is light microscopy, involving examination of stained serial sections of fixed materials, that has contributed the basic framework for our attempts to compile a complete structural description of the remarkable processes of anther and ovule develop-
Retrospect and Prospect
ment, microsporogenesis, and megasporogenesis in angiosperms. This approach has also helped to define the developmental changes that occur at the level of individual cells, an accomplishment that is not possible by biochemical, molecular, or physiological methods. Initiation of floral primordia is reflected in a series of morphological changes not only in the shoot apex of the vegetative plant but also in its individual cells. The chronological sequence of changes in the distribution and positional appearance of the different parts of the flower as seen in sections in the light microscope or in the scanning electron microscope has provided a means of correlating the results of more sophisticated studies. Unfortunately, both superficial and sophisticated analyses frequently involve only a limited number of experimentally amenable species. Comparative studies of the origin of floral parts in a wide range of plants is needed to consider the currently available data in a broad biological context and to focus attention on features that are common to flowers of many species and those that are unique to flowers of just one or two species. Some of the early studies on anther development were bedeviled by controversy as to whether the tissues of the anther arose from a single hypodermal cell (archesporial initial; archesporium) or from a group of cells several layers beneath the epidermis; it is now clearly established that a hypodermal origin of the wall layers and sporogenous cells is the mode in most, if not all plants, examined. Present interest in the tapetum - a layer of cells immediately surrounding the sporogenous cells of the anther - has been stimulated by divergent views on its origin, as deduced from light microscopic observations. Because of the dense cytoplasmic nature of the tapetal cells and the frequent divisions they undergo, some investigators erroneously thought that the tapetum is a modified outer layer of cells of the sporogenous tissue; however, based on later studies, the view on the origin of the tapetum from the descendants of the parietal cell, which itself is born out of a periclinal division of the archesporial initial, is now uncontested. More than ever, the coming of the transmission electron microscope has enriched research in the reproductive biology of angiosperms. During the past three decades, the electron microscope has portrayed the structure of the cells of the anther and of the pollen grain as amazingly complex packets of organelles, differing from cell type to cell type, yet fundamentally the same in all. We also owe to light and electron microscopic analyses our current level of knowledge of the structure and ori-
gin of the exine of the pollen wall. Here the focus has been on the tapetal cells, which apparently synthesize precursors of sporopollenin, the primary component of the polymer that forms the exine. Sophisticated electron microscopic protocols such as rapid fixation and freeze-substitution, immunological methods, and biophysical approaches are beginning to unravel the structure of the pollen tube and the mechanism governing the ebb and flow of its cytoplasm and growth of the tube. The identification of deoxyribonucleic acid (DNA) as the genetic material was preceded by the formulation of an elegant cytochemical procedure, the Feulgen reaction, that provided the basis for the quantitative estimation of this material. A colored complex is generated when DNA interacts with Schiff's reagent under acidic conditions that hydrolyze the purine-deoxyribose linkages and expose the free aldehyde groups. The color produced is proportional to the amount of DNA present and the quantity of DNA can be measured by densitometry. The application of this method combined with microspectrophotometry led to the determination of DNA amounts in pollen nuclei; it was confirmed that doubling of DNA occurs only in the nucleus of the generative cell, which gives rise to the sperm, whereas the vegetative cell nucleus, which does not divide, retains the expected haploid level of DNA. A multidisciplinary approach involving histological, immunochemical, and cytochemical methods and electron microscopy has also been employed to study the female reproductive units of angiosperms. In the majority of angiosperms examined, the basic structure of the embryo sac appears to be conserved, although multiple pathways are followed to evolve this structure. At the ultrastructural level, the angiosperm egg provides a fine example of a cell whose metabolic activity is at a low ebb. In a similar way, the structure of synergids with internal projections known as the filiform apparatus, can be rationalized in terms of their function in the transport of metabolites from the nucellus to the developing egg cell. Histochemical changes are valuable for monitoring the chemical state of individual cells, especially for localizing specific substances or sites of chemical activity in the cells. Relative levels of histochemically detectable proteins, nucleic acids, and carbohydrates have been used to define the component cells of the embryo sac and of the developing anther, embryo, and endosperm. The presence of ribonucleic acid (RNA) and proteins in particular cell types, such as the egg and the synergids, has
Reproductive Biology of Angiosperms:
in a limited way alerted workers to the possibility that gene activation is involved in the differentiation of these cells. This is not surprising, since a cell which is programmed to behave in a certain way in response to an internal signal will inevitably show changes when the signal is received; what is surprising is that such changes can be detected in the light microscope. Although the histochemical approach presents a picture of cells that accumulate macromolecules at a given point in time, it does not demonstrate the unique ability of cells to synthesize any of the substances nor does it identify the sites of their synthesis. Synthesis of macromolecules by specific cells is monitored by autoradiography using radioactive precursors. Autoradiography of incorporation of precursors of DNA, RNA, and proteins into developing anthers has reinforced the image of differential nucleic acid and protein synthetic activity of the vegetative and generative cells of the pollen grain. Because autoradiographic methods are generally applied at the cellular level, they are particularly useful when heterogeneity of the tissues precludes the use of conventional biochemical techniques. Interest in the process of fertilization has been revived by computer simulations of the three-dimensional structure of sperm cells based on electron microscopic analysis of serial sections. In attempts to understand the essentially uncontrolled way in which one sperm is presumed to fuse with the egg and the other with the diploid fusion nucleus in the act of double fertilization, a hopeful note was introduced with the discovery that in some plants the two sperm cells are structurally and functionally different. In the analysis of the beginnings of the sporophytic development of the plant, the embryo and endosperm have figured in several different levels of investigations. One of the important concepts to emerge from light microscopic analysis of embryo development in plants is that the early division sequences of the zygote are carried forward according to a blueprint characteristic of each species. In this scenario, the first division of the zygote is almost invariably asymmetric and transverse, cutting off a large vacuolate basal cell attached to the embryo sac wall (micropylar end) and a small, densely cytoplasmic terminal cell toward the embryo sac cavity (chalazal end). The organogenetic part of the embryo is derived from the terminal cell with little or no contribution from the basal cell; on its way to form the embryo, the terminal cell may divide transversely, longitudinally, or obliquely. Divisions of the basal cell create a small subtending structure of the embryo, known as the suspensor. The advantage of identifying these precise division steps is that it
enables one to trace the adult organs of the embryo back to the cells cut off during early embryogenesis. Embryo development has been studied so exhaustively from a descriptive point of view that it is easily one the best understood areas of angiosperm embryology. Although fusion events that give rise to the zygote and the endosperm occur almost synchronously, examination of postfertilization embryo sacs of a number of plants has clearly shown that the primary endosperm nucleus begins to divide ahead of the zygote. The development of a free-nuclear or cellular endosperm in the embryo sac before the division of the zygote has an adaptive value in ensuring the availability of an adequate food supply to the developing embryo. There is a further type of behavior of the embryo and endosperm, which was uncovered by a combination of light and electron microscopy, microspectrophotometry, and biochemical analysis and which mirrors their distinctive functions. This is the synthesis and accumulation of an acervate complex of food reserves, mainly storage proteins that are used by the embryo during its development and germination. In the embryo itself, storage proteins accumulate almost exclusively in the cells of the cotyledons, which undergo senescence after germination. In contrast, in the endosperm, all of its cells dedicate their synthetic machinery to the production and accumulation of food reserves. The study of storage protein synthesis has done much to stimulate interest in the dynamics of DNA changes in the embryo and endosperm cells when major increases in their fresh and dry weights occur. It has been found that the cells of developing cotyledons and endosperm continue to synthesize DNA after cell divisions cease, leading to the production of cells with disproportionately high DNA values. This abbreviated version of the mitotic cycle, known as endoreduplication, is fairly far removed from normal cellular activity, but it does offer some explanation for the continued accumulation of food reserves in the cells. It can be envisaged that an increase in gene copies might affect the amplification of those sequences concerned with the transcription of storage protein messenger RNAs (mRNAs). This is of some advantage to the plant since it allows rapid utilization of amino acids and other metabolites that are attracted to the developing fruits and seeds.
(ii) Genetics We now consider the role of genetic studies in unraveling the reproductive biology of angiosperms. In many flowering plants, self-fertiliza-
Retrospect and Prospect
tion is prevented by a genetically controlled mechanism known as self-incompatibility. Self-incompatibility is manifest in the failure of the pollen grains to germinate on the stigma or of the pollen tube to grow through the style of the same flower. One of the outstanding contributions of genetic studies to angiosperm reproductive biology is the formulation of the view that incompatibility reactions are controlled by a single gene (sterility or S-gene) with multiple alleles. The result is that a pollen grain bearing an S-allele identical to one of the two alleles carried by the pistil suffers growth inhibition, whereas pollen bearing alleles different from both pistil alleles is able to grow. There does not exist very much published material commenting on genetic control in other aspects of angiosperm reproduction because, until recently, research emphasis was on the comparative and descriptive aspects of the processes and not on their genetics. As alluded to earlier, this approach has resulted in a wealth of information on floral morphology, anther and ovule development, microand megagametogenesis, embryogenesis, and endosperm formation in a large number of plants. At the biochemical level, chemical components of the various organs and tissues concerned with reproduction have been analyzed in a covert attempt to find a chemical basis for their function. As a result, morphologists, cytologists, physiologists, and biochemists working with different aspects of plant reproduction considered genetic control to be a sideshow to the development of the structure or organ modulated by other mechanisms.
3. GENETIC A N D MOLECULAR PERSPECTIVES
The morphological changes that occur during the reproductive development of plants are correlated with modifications in the biochemical properties of the participating cells. The central dogma of developmental biology and genetics is that these modifications are due to differential gene action. Assuming that all cells contain an identical complement of genes, the hypothesis implies that the primary cells involved in each differentiation event utilize different genes from their inherited gene pool. The major question in the analysis of the reproductive biology of plants then becomes, How is the genetic information regulated at different stages of the life cycle of the plant? When we recall that the reproductive life of angiosperms consists of a series of landmark stages, it becomes apparent that different cell types should be expressing different genes. Principles governing the genetic and molecular analysis of reproduction in angiosperms then involve the differential expression of specific genes in specific cells at specific times. (i) Genes and Flowering
Because of the ease of generating mutants, flower development has proved to be simpler and more amenable to genetic analysis than other aspects of plant reproduction. Genetic studies of flower development have concentrated on gene systems that regulate the transformation of the Although Drosophila melanogaster (fruit fly) was vegetative shoot apex into a floral meristem and on mainly worshiped by geneticists on the zoological those that cause the meristem to differentiate into side, and a small core of animal cytologists and cell types characteristic of each floral organ. embryologists took to the liking of the fly to cham- Although the transformation of a vegetative shoot pion the role of genes in embryogenesis, no compa- apex into a floral meristem is clearly distinct from rable organism was available to plant biologists. the determination of floral organs, genes of both This was the story until about 1970. During the classes are considered to be homeotic, as mutations last 30 years there have been several breakthroughs in them cause certain organs to be replaced by that have laid the foundation for the genetic analy- other organs not normally found in that position. sis of flower pattern formation and floral morpho- In Arabidopsis, a broad range of mutants have been genesis, isolation of incompatibility genes, isolation described, including plants with petals instead of and characterization of genes that are expressed stamens in the wild-type flowers, or sepals develduring male gametogenesis and embryogenesis, oping in positions occupied by petals in the wild isolation of seed protein genes, gene transfer, and type, or extra flowers in place of the ovary. Cloning transposon mutagenesis. These investigations, cou- and sequencing of representative organ identity pled with the development of Arabidopsis thaliana genes have led to the conclusion that the wild-type (Brassicaceae; hereafter referred to as Arabidopsis) products of the genes involved in certain mutations as a model system comparable in many respects to act by allowing cells to determine their positions D. melanogaster, promise to provide new insights within the floral meristem and thus to specify the into the genetic control of angiosperm reproductive fate of the organ. Were this not the case, the entire biology. A brief background of this research is pro- structural organization of the flower would be in jeopardy. vided in the following section.
Reproductive Biology of Angiosperms:
Special DNA sequences known as transposons possess the ability to move about in the genome. Because the insertion of a transposon into the coding sequence of a gene will prevent the expression of that gene, transposon insertion has been employed to produce homeotic mutants for structural abnormalities and pigmentation pattern in flowers. A series of mutations similar to those found in Arabidopsis that alter the identities of floral organs have been characterized by this strategy in Antirrhinum majus (snapdragon; Scrophulariaceae), which belongs to a taxonomically distant family. This suggests that the genetic mechanisms controlling floral organ identity are very ancient and are therefore highly conserved in the evolution of the dicotyledons. Transposon mutagenesis, pursued in Petunia hybrida (Solanaceae), promises to provide a useful molecular probe to identify genes involved in flower development. Using a complementary DNA (cDNA) library made to mRNA isolated from thinlayer explants of tobacco induced to flower, it has been possible to isolate recombinants corresponding to messages present during the early stages of differentiation of the floral meristem and to follow their expression at the cell and tissue levels. What is evident from a consideration of these results is that the morphological pattern of the flower and its organs is determined by a multitude of influences, of which genes are paramount. Clearly, if floral differentiation involves the activation of genes, then the stability of the differentiated state within the flower also requires the involvement of genes. Support for this notion comes from the observation that each floral and vegetative organ of the tobacco plant has only a small number of common mRNAs, but has a large number of mRNAs specific for the particular organ. Cells of the floral organs are capable of a degree of organization that is matched by the presence of discernible mRNA subsets. (ii) Gene Expression during Anther and Pollen Development
Several independent gene expression programs that correlate with the differentiated state of the specific cell types have been established in developing anthers. These include mainly mRNAs that are specific for the wall layers and taperum of the anther and for the premeiotic pollen precursor cells and mature pollen grains. A delayed gene expression pattern apparently accounts for the elaboration of the exine of the pollen; here, transcription
that determines the exine pattern is initiated during the premeiotic phase, but the exine itself is not formed until after meiosis. At least two sets of genes have been implicated in the development of the pollen grain. Genes of the first set, which are not specific to the pollen, appear to be active soon after meiosis and reach a peak during pollen development but diminish in the mature pollen. Genes of the second group, comprising pollen-specific genes, are activated at the bicellular pollen stage and their transcripts are found maximally in the mature pollen. Clearly, other gene sets, having different patterns of expression, must also be involved in pollen development; on the whole, data thus far available present compelling evidence for differential gene expression during male gametogenesis in angiosperms. Male sterility resulting in pollen abortion in the anther has been previously shown to be caused by mitochondrial genes. At the molecular level, in comparison to corresponding DNA segments of the male-fertile plant, a highly ordered rearrangement of portions of the mitochondrial genome appears to be characteristic of the male-sterile plant. From this, one can predict that the rearranged gene might cause mitochondrial dysfunction by its inability to code for functional enzymes. Since many male-sterile mutations interfere with tapetal development, genetic engineering methods have opened up the way to introduce into plants chimeric genes that selectively destroy the tapetum during anther development, prevent pollen formation, and lead to the production of male-sterile plants. In the area of self-incompatibility, impressive progress has been made in identifying products of the S-allele in members of Solanaceae and Brassicaceae and in determining how they might cause the inhibition both of pollen germination and of pollen tube growth following an incompatible pollination. This study has been made possible by immunological localization of antigens specific to various S-locus alleles in stigma homogenates, identification of stylar glycoproteins with individual S-alleles, and by the cloning and sequencing of the S-genes and their protein products. Suggestive of a functional involvement of S-proteins in the self-incompatibility interactions are the correlations of the temporal expression of these proteins with the onset of self-incompatibility and the spatial expression at the sites of pollen tube rejection in the stigma and style and the inhibition of selfpollen tube growth in vitro by purified S-proteins. The most interesting current question is how the Sproteins might mediate recognition between the
Retrospect and Prospect
pollen grain and the stigma. There is considerable focus now on arabinogalactan as a protein mediating in pollen recognition on the stigma and in pollen tube growth in the style independent of incompatibility reactions. Gene expression during female gametogenesis in angiosperms is an area that has not lent itself to molecular analysis. This has been attributed to the fact that the formation of the egg cell occurs within the embryo sac, which itself is buried within several layers of somatic cells of the ovule. Protocols for the enzymatic isolation of living embryo sacs from ovules and of egg from the embryo sacs of certain plants have introduced a new and hopeful note that before long it will be possible to identify genes that are required to establish and maintain polarity of the egg and to specify maternal factors for embryogenic development. A strategy most appropriate for this would be to select egg-specific recombinants from a cDNA library made to mRNAs of isolated embryo sacs or egg cells by subtraction hybridization or polymerase chain reaction (PCR) amplification. (iii) Embryo-specific Genes The developing embryo has already been recognized as the major organ of the plant where cellspecific gene expression programs operate. Genetics has been a powerful tool in the early stages of this work. Analysis of embryo- and seedling-defective mutants has led to the identification of genes responsible for pattern formation and morphogenesis in early-stage embryos. Growth arrest at a particular stage by a mutation implies that, due to a defective gene, the embryo is unable to complete a metabolic reaction or to synthesize a specific nutrient substance required for development. But this type of information is no substitute for hard-core molecular studies of the processes that direct polarity and specify cell lineages in the embryo; unfortunately, very little work has been done to address these problems, which first appear in the zygote and in the cells immediately derived from it. Identifying the protein products of some genes affecting divisions of earlystage embryos as ubiquitous regulators of cell wall synthesis has only reinforced the importance of cell wall formation in embryogenesis. To a large extent, genes expressed during later stages of embryogenesis, especially during cotyledon initiation and growth, persist throughout maturation and are stored in the embryo of the mature seed; many are present in both the seedling and the adult plant. Another set of genes, designated as late embryoge-
nesis abundant (LEA), whose mRNAs begin to accumulate at still later stages of embryogenesis, are thought to be involved in seed desiccation when the embryo lapses into a state of quiescence or dormancy. Seed protein genes that encode the abundant storage proteins, however, represent the most striking example of a highly regulated embryo-specific gene set. The general features of a seed protein gene expression program include accumulation and decay at precise stages of embryogenesis, almost exclusive expression in embryos in contrast to other parts of the plant, and spatial expression within specific parts of the embryo. There is also evidence to indicate that seed protein gene expression is regulated by both transcriptional and posttranscriptional events. Finally, the expression of different seed protein genes in transgenic plants suggests that DNA sequences of seed proteins are highly conserved and that the regulatory factors that control seed protein synthesis are common to seeds of different plants. Although a great deal has been learned about the expression of storage protein genes and their structure, our understanding of the mechanisms controlling their temporal, spatial, and cell-specific expression still needs to be fine tuned. The endosperm of cereal grains has also proved to be valuable for basic studies of storage protein gene expression. With the benefit of existing background knowledge on the chemistry of endosperm proteins of our common cereals, research on the isolation, characterization, and expression of endosperm storage protein genes has taken off dramatically. However, analysis of the gene activation program in the endosperm concerned with expressional divergence from the embryo developmental program has proved to be difficult in both monocots and dicots. Progressive embryogenesis depends upon the control of the gene expression program by the environment, both in its hormonal and, to some extent, in its physical aspects also. For example, the expression of some classes of embryo-specific genes, especially those encoding storage proteins and proteins that confer desiccation tolerance in embryos, appears to be under the control of the hormone abscisic acid (ABA). This reliance of embryo gene expression on ABA is based on the fact that mRNAs for storage protein genes accumulate precociously in immature embryos in response to ABA application. Current knowledge indicates that in Triticum aestivum (wheat; Poaceae) embryos, an 8-bp region of the gene is the conserved core in the ABA-responsive element and that a trans-acting
w factor is functionally important in the ABA response of the gene. Gene transfer is a powerful tool for tailoring flowering plants to perform functions far beyond the scope of their intended role and for the demonstration of DNA regions important for seed protein gene expression. This is an area of research that has made rapid strides since the 1980s. The highly conserved nature of storage protein genes of the embryo and endosperm and the similarity of the regulatory factors involved in their expression in different plants require that the input of chimeric seed protein gene promoters into the cellular machinery of heterologous hosts be subject to correct molecular processing. This has been found to be the case when embryo and endosperm storage protein genes of several plants introduced into tobacco or Petunia using Agrobacterium tumefadens-mediated plasmid vectors became incorporated into the genome of the recipient's cells and were found to be expressed in a controlled, spatial, and temporal fashion. Significant as this discovery was, the amount of protein made by the gene in the alien environment was not sufficient to match its action in a homologous milieu. As storage protein transcriptional units are identified, they can be attached to regulatory sequences that allow their expression in plants, and the constructs can be used for transformation experiments. This will also simplify the methods to assess the stability of the genetic material in the engineered plant and to increase the yield of the protein product. For a number of years, monocots have remained highly recalcitrant in their regenerative ability in tissue culture. This roadblock has now been largely cleared, and methods are at hand to regenerate many of the common cereals by organogenesis or by somatic embryogenesis. Unfortunately, until recently, monocots have not been readily infected by A. tumefaciens, although they are amenable to direct gene transfer into protoplasts or by microprojectile bombardment. Because of their versatility and ease of handling, embryogenic cells growing in suspension cultures and embryogenic pollen grains might be exploited for a functional characterization of the gene expression program of cells embarking on the embryogenic pathway. In view of the fact that embryogenesis is induced in carrot cell suspension by the simple expedient of removing auxin from the medium, not surprisingly, only minor qualitative changes occur in the protein profile during transition from a nonembryogenic to an embryogenic state. The benefits to the embryogenic cell of using
Reproductive Biology of Angiosperms:
more or less the same genetic program as its nonembryogenic counterpart for differentiation are not fully understood. Genes identified during somatic embryogenesis are also found to be expressed during zygotic embryogenesis. In addition, analysis of proteins secreted into the medium by embryogenic cells has revealed the presence of a plant lipid transfer protein not found in the medium of nonembryogenic cells; expression of the gene for this protein in the early-stage somatic embryos of carrot has suggested a role for lipid transfer proteins in cutin biosynthesis on the epidermal surface of developing embryoids. The similarity in the gene expression programs of embryogenic cells of carrot and developing zygotic embryos may be more than a coincidence; it might reflect the universality of the two phenomena, resulting in identical end products. The evidence for gene activation during pollen embryogenesis has been hard to come by. Differences have been observed in the regenerative ability of pollen grains of cultured anthers of different genotypes of the same species. This might suggest that the process is largely determined by a genetic component of the individual plant that at some point impinges on the process of pollen dedifferentiation. The clearest model of gene action during pollen dedifferentiation is based on cytological studies. The results suggest that in contrast to normal pollen development, embryogenic development is due to the transcription of new mRNA in the uninucleate pollen grains of cultured anthers. It appears that as much as possible of the gametophytic program is masked and that a sporophytic program is unfolded in but a few pollen grains. At the molecular level, some specific phosphoproteins have been associated with the induction of embryogenic divisions in pollen grains, whereas other phosphoproteins seem to be characteristic of pollen grains that continue the gametophytic program. (iv) Gene Activity during Fruit Development and Ripening
Fruit development and ripening involve a complex series of physiological processes whose molecular biology is still incompletely understood. A brief look at the gene activation program during these events is justified not only because of their biological importance, but also because genes concerned with ripening are those that cause cellular degradation. One group of genes that has been monitored is the photosynthetic genes whose
Retrospect and Prospect
expression patterns vary at different times during fruit development. During ripening, a decline in the transcript levels of a number of plastid-encoded proteins is found to correlate with the dedifferentiation of chloroplasts into chromoplasts. In addition, genes for polygalacturonase, cellulase, and other ripening-related proteins have been cloned and their expression followed. These studies, along with the analysis of extant and in vitro synthesized proteins of fruits of various ripening stages, have shown that genes for both synthesis and degradation of proteins are activated during the ripening process. Since the hormone ethylene controls the ripening of many fruits, antisense RNA strategy has been used to inactivate the synthesis of a rate-limiting enzyme in the biosynthetic pathway of ethylene and to delay fruit ripening. This discovery has raised the prospects of extending the life span of fruits, thereby preventing spoilage. Indeed, genetically engineered tomatoes with a longer shelf-life than normal ones are already being sold in the supermarkets. 4. MOLECULAR AND GENETIC ENGINEERING TECHNIQUES The achievements in the realm of angiosperm reproductive biology described in the previous section have been made possible by refinements in the existing methods in biochemistry and cell and molecular biology and by the introduction of the new disciplines of biotechnology and genetic engineering. These methods are not only numerous, but they are also very complex. Since it is assumed that readers will be familiar with these methods, no attempt is made here to describe them. However, the principles underlying some of these techniques to which references are made in this book without further explanation will be described. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE): This method is used to
fractionate nucleic acids and proteins using polyacrylamide as a support medium. When electrophoresis is performed in the presence of the ionic detergent sodium dodecyl sulfate (SDS), the method can be used to determine the molecular mass of migrating proteins. After electrophoresis, the gel is stained to locate the bands or is autoradiographed against an X-ray film to locate radioactively labeled macromolecules. Two-dimensional electrophoresis involves the fractionation of proteins in two dimensions based on independent properties such as the isoelectric point and the ability to complex with SDS. A stag-
11
gering number of proteins can be separated by this method. Cell-free translation: This method employing preparations from wheat germ or rabbit reticulocytes as the source of cytoplasmic ingredients necessary for protein synthesis is used to confirm the functional competence of mRNA or polyadenylic acid-containing RNA [poly(A)-RNA] isolated from cells. A labeled amino acid is added to the reaction mixture to determine the amount and nature of proteins synthesized. For the latter purpose, the proteins are separated by SDS-PAGE. Molecular hybridization: It is possible to assay for
the immediate products of gene expression by RNA-DNA hybridization and to determine gene sequence complexity by DNA-DNA hybridization. In practice, one population of the single-stranded nucleic acid, present in low concentrations, is radioactively labeled; the other population, present in excess, is unlabeled. Separation of the hybrids is accomplished by fractionation on hydroxyapatite column or by Sl-nuclease digestion. The former is used to bind selectively double-stranded molecules, whereas the latter selectively degrades single-stranded molecules. By either method, the percentage of hybrids formed is obtained by determining the radioactivity in the fractions. In situ hybridization: This method is ideal for localizing gene sequences on chromosomes at the light microscope level by DNA-DNA hybridization using labeled DNA probes and in cells by RNA-RNA hybridization using labeled antisense RNA probes. Total RNA is localized in cells and tissues by annealing with labeled polyuridylic acid [poly(U)] as a probe. Sites of annealing of labeled probes are detected in both cases by autoradiography. Nonradioactive digoxygenin-labeled oligodeoxythymidylic acid (oligo-dT) probes are also widely used to localize specific transcripts and total RNA in cells. To localize proteins, antibodies that are radioactively labeled, or coupled with an enzyme, or linked to an electron-dense heavy metal like gold, are used and the sections are viewed in a light, fluorescence, or electron microscope. cDNA library: A cDNA library or a cDNA clone bank is a population of DNA copies of homogeneous mRNA sequences coding for a specific protein, or DNA copies of representatives of most, if not all, mRNAs of a cell, tissue, organ, or organism, inserted into a plasmid or phage vector and cloned into bacterial cells. Cloning is initiated by the isolation of mRNAs transcribed by genes involved in the synthesis of a particular protein or active in a specific organism or its constituent parts.
12
Differential screening is employed to isolate genes specifically expressed in a cell or organ or genes involved in physiological processes whose protein products are unknown. For example, many genes specifically involved in pollen development as well as others common to the pollen, stem, and leaf of the plant are likely to be present in a cDNA library made to poly(A)-RNA from mature pollen grains. To eliminate recombinants common to all systems and to enrich the library for pollen-specific clones, the cDNA library is replica-plated and probed with labeled cDNAs made to mRNAs from pollen, stem, and leaf. In this method, bacterial colonies with inserts that are pollen specific will give positive signals only to the probe made from pollen mRNA, whereas colonies with inserts containing sequences similar to those present in the pollen, stem, and leaf will react positively to all three cDNA probes. If the protein product of the gene is known, one of the simplest procedures for screening the cDNA library is the use of a protein detection method based on immunological techniques. Alternatively, oligonucleotides synthesized to the amino acid sequence of the protein of interest are used as probes to pick specific clones from the library. Genomic library: In contrast to the cDNA library, which is made from mRNA, the genomic library has its origin in the total DNA of the organism. Genomic clones are useful for revealing the structure of the gene in the chromosome and the structure of the sequences flanking it. They may also provide information about the presence or absence of introns that are not found in the mRNA. Genomic cloning is generally accomplished using bacteriophage lambda as a vector, and hybridization techniques similar to those described for cDNA clones are applied to identify individual genomic clones. Filter hybridization: This constitutes a family of methods in which nucleic acids or proteins are initially adsorbed to a nitrocellulose filter before they are hybridized with labeled probes or reacted with antibodies to assess their size, tissue distribution, and quantitative expression. Southern blot is employed to identify electrophoretically separated DNA fragments transferred from agarose or polyacrylamide gels to membrane filters. When RNA fragments separated on agarose gels under denaturing conditions are transferred to a filter and hybridized with complementary single-stranded DNA (ss-DNA), the method is known as Northern blot. For some experiments it is preferable to spot nucleic acids directly on the filter membrane using a specially designed apparatus to produce uniform
Reproductive Biology of Angiosperms:
spots, rather than transfer them after electrophoretic separation. This is the basis for slot-blot or dot-blot hybridization; after hybridization, the spots are excised and the hybridized radioactivity is counted. Another filter hybridization technique, known as Western blot, is used for the identification and quantification of specific proteins in complex mixtures by transferring them from a SDS-PAGE gel to a membrane and probing with antibodies that react with target proteins. DNA sequencing: Determination of the arrangement of the particular sequences of base pairs in a DNA molecule is the basis of DNA sequencing and constitutes the method of choice to determine the differences between individual molecules of DNA. The Maxam-Gilbert chemical degradation method and the Sanger dideoxy-mediated chain termination method are currently employed to sequence DNA. Polymerase chain reaction: This new technique is
extremely efficient for exponential amplification of DNA sequences. Using two oligonucleotides as primers and repeated cycles of denaturation, annealing, and DNA synthesis, substantial amounts of double-stranded DNA (ds-DNA) terminated by the primers is obtained. The method is especially useful for amplifying small pieces of target sequences from a pool to generate probes to screen a cDNA library. Antisense RNA strategy: This is used to down-regulate the expression of specific genes of interest in cells and tissues. When antisense genes are introduced into plants, they promote the synthesis of noncomplementary RNAs, which subsequently hybridize with target RNAs and prevent their expression. Cells that express antisense genes are thus starved for proteins normally coded by complementary RNAs and eventually they disintegrate. Construction of chimeric genes: Here, the regula-
tory elements from one gene are fused to the coding sequence of a prokaryotic reporter gene whose product can be easily detected by simple staining reactions or which produces discernible effects, such as cytotoxicity, on target cells. Among the reporter genes that can be readily detected by means of sensitive histochemical tests are those encoding for fi-glucuronidase (GUS) and chloramphenicol acetyltransferase (CAT). Diphtheria toxin is a member of a group of protein toxins that inhibit protein synthesis, and expression of diphtheria toxin chain-A gene (DT-A) driven by a promoter causes cell death as a visible phenotypic expression. Gene transfer: Gene transfer methods involve the introduction of DNA into plant cells, integration of the foreign DNA into the host plant's genome,
13
Retrospect and Prospect
regeneration of transgenic plants, and selection of transformants. Techniques used to deliver foreign DNA into plant cells include particle bombardment, microinjection, electroporation, and Agrobacterium-mediated transformation. Particle bombardment capitalizes on the fact that metal microprojectiles (particles) carrying DNA fragments of appropriate size, if propelled by an acceleration device to high velocities, penetrate the cell walls and cell membranes of plants without damaging them. For bombardment, gold and tungsten particles coated with precipitated DNA or frozen micropellets of DNA acting as microprojectiles are used. In the microinjection method, glass micropipettes with openings of less than 1 fim in diameter are used to deliver DNA into targets such as protoplasts, isolated pollen grains, and single cells or cells in small clusters. Electroporation is based on the application of a pulse of high-voltage electric field to targets suspended in a buffer, which results in a reversible permeabilization of the plasma membrane to macromolecules.
used to screen regenerants and identify transformants, perhaps the easiest are assays for GUS or CAT genes. At a more sophisticated level, DNA is isolated from the putative transformants and subjected to Southern blot hybridization to identify the inserted genes, or the inserts are separated from restriction endonuclease-digested DNA by electrophoresis and amplified by PCR. 5. SCOPE OF MOLECULAR EMBRYOLOGY
Against the background of review of the reproductive biology of angiosperms presented in this chapter, it is now necessary to describe the scope of molecular embryology, the subject of this book. Whereas embryogenesis as applied to angiosperms is concerned with the study of zygotic, somatic, and pollen-derived embryos and of the endosperm, most plant biologists include under embryology the study not only of the embryo and endosperm, but also of the ensemble Because of the ease with which different strains of morphological, cytological, and anatomical of A. tumefaciens deliver foreign DNA into the host changes concerned with sporogenesis, gametogengenome, transformations with high efficiencies esis, and fertilization. As conceived in this book, have been achieved in many species by the use of embryology involves both premeiotic and postthis vector. As is well known, A. tumefaciens readily meiotic events in the anther and ovule, encominfects various species of dicots to produce tumor- passing both diploid and haploid phases in the life ous outgrowths known as crown-gall tumors. cycle of the plant; in short, embryology is the Tumor induction is the result of transfer and inte- plant's own strategy to perpetuate the progeny. gration of a specific segment of the tumor-inducing Thus, the topics that are germane for discussion (Ti) plasmid DNA from the bacterium into the under the rubric of embryology include anther and nuclear genome of the susceptible plant cells. The ovule development; microsporogenesis and inserted DNA segment of the Ti plasmid is known megasporogenesis; prefertilization events; fertilas the transferred DNA (T-DNA); the other critical ization; zygotic, somatic and pollen embryogeneregion of this plasmid is a sequence encompassing sis; and endosperm formation. These past studies the virulence (V1R) genes encoding proteins in embryology, described in some classical books and scores of scientific papers, have been carried required for T-DNA transfer. forward by many investigators in laboratories To serve as an effective vehicle for gene transfer, around the world. They provided a useful framethe bacterium is first disarmed by deleting the Twork for an appreciation of the range of variations DNA genes and then is equipped with a marker in the development and organization of the male gene for antibiotic resistance to give a selectable and female reproductive units of angiosperms and advantage to the transformants. The desired forthe process of embryogenesis. Subsequent evaluaeign gene is then ligated between the T-DNA bortion of the data led to a diversification in the outders of the vector, and bacteria with the genetically look in research in plant embryology. The engineered plasmid are used to infect cultured utilization of embryological information in deterplant materials, which typically include leaf discs, mining the phylogenetic affinities of certain hypocotyls, stems, roots, embryos, and embryocontroversial families, genera, and species of flowgenie callus. Depending upon the plant material, ering plants was the immediate outcome of such the type of explant, and the composition of the diversification. Embryological data have also been regeneration medium, regenerants may appear as used to find new affinities for certain species and callus, shoot buds, or somatic embryos. These are genera of plants that were rather uncomfortable in further subjected to appropriate tissue culture protheir old alliances. tocols to obtain normal plants. Among the methods
14 Since the 1930s, advances made in the fields of plant physiology, biochemistry, and genetics, and refinements in the culture of plant organs, tissues, cells, and protoplasts under aseptic conditions have had much influence on the direction of embryological research. The outcome of research in these fields introduced the discipline of experimental embryology, involving the control of pollination and fertilization and manipulations of the anther, pollen grains, ovary, ovule, and embryo by excision and culture, by chemical, hormonal, and surgical treatments, and by exposure to selected day-lengths and temperature conditions, as a way to study the controlling mechanisms that affect the form and structure of plant reproductive organs. Studies in experimental embryology were enshrined in the hope that as we gain familiarity with the laws that control the reproductive processes in plants, we will get new clues about how to control them to our advantage. Thus, experimental embryology transformed an era of descriptive studies into one of experiments and deductions, designed to discover the physiological basis of the reproductive processes in angiosperms. From experimental embryology, it is perhaps one further step of complexity to molecular embryology. In recent years, the most profound impact on the study of angiosperm embryology has come from advances in molecular biology, recombinant DNA technology, and gene cloning. The ability to isolate specific genes that modulate developmental processes in living organisms opened the way for dramatic new approaches in the study of angiosperm embryology. The insights obtained in our understanding of the molecular and genetic basis for the different embryological processes in angiosperms are considered here as molecular embryology. In general, these studies are concerned with the spatial and temporal activation of genes involved in embryological processes, isolation and cloning of genes, sequencing the genes, determination of the amino acid sequence of the proteins encoded by the genes, and identification of the protein products. The use of embryos and embryogenic cells for gene transfer is also considered to be a topic in molecular embryology. The developments that were to herald molecular embryology may be said to have had their origin in the 1970s; although somewhat late blooming, research in this field has taken off dramatically, providing an increasing awareness of the reality of genetic engineering for improvement of our agricultural crops. Therein lies the importance of molecular embryology in the reproductive biology of angiosperms.
Reproductive Biology of Angiosperms:
6. ORGANIZATION OF THE TEXT Following is an abbreviated survey of the contents and organization of this book. Based on the wider comprehension of embryology, the book is divided into four major sections, entitled "Gametogenesis" (five chapters), "Pollination and Fertilization" (five chapters), "Zygotic Embryogenesis" (four chapters), and "Adventive Embryogenesis" (two chapters). Included in the section entitled "Applications" is a concluding chapter on the genetic engineering of embryos. The first section of the book closely examines the morphological, cytological, biochemical, and molecular aspects of the development of the anther and the ovule that result in the formation of functional male and female gametophytes. The background of events thereby established serves as a basis for the second section, dealing with pollination and fertilization. The problems of pollination and fertilization are approached from the points of view of pollen deposition on the stigma and the resulting recognition reactions, in vivo and in vitro growth of pollen tubes, phenomena of self- and cross-incompatibility, and, finally, sexual recombination heralding the sporophytic growth. The third section is concerned with the growth of the embryo and endosperm and with gene expression in cells with different genetic backgrounds. Two chapters on adventive embryogenesis, dealing with somatic embryogenesis and pollen embryogenesis, form the focus of the fourth section. Despite the fact that work on the transformation of embryos has not made a significant dent in practical or theoretical considerations, a short account of the current work in this area rounds off the descriptive part of the book. Although the framework of discussion is provided mainly by angiosperms, some contributions from the gymnosperms are also stressed, especially when they illustrate principles that are generally applicable. According to one viewpoint, the life of the embryo is not considered complete until the seed germinates; unfortunately, the many fascinating and intriguing biochemical and molecular aspects of germination of the seed (which by themselves would have formed another bulky section of the book) have not been included here. Notwithstanding this deliberate omission, it is hoped that the book will be of use in bringing together the diverse and scattered literature on the developmental and molecular aspects of plant embryology and will give a general idea of the current and future perspectives of this exciting field.
SECTION I
Gametogenesis
Chapter 2
Anther Developmental Biology 1. Anther Determination (i) Timing of Anther Determination (ii) Genetics and Molecular Biology of Anther Determination 2. Anther Development (i) The Anther Wall (ii) The Tapetum 3. Microsporogenesis (i) Ultrastructural Changes (ii) Molecular Biology of Meiosis 4. Anther Culture in the Study of Meiosis 5. General Comments
The anther is a morphologically simple organ of the flower concerned with some unique functions, such as microsporogenesis and the production of pollen grains. The size, shape, and orientation of the anther on the stamen and the diversity of the cells and tissues of the anther make it a morphogenetic system of great interest. A typical anther is a two-lobed organ with two locules or microsporangia in each lobe and, so, the anther is functionally a group of four microsporangia. A single vascular strand is embedded in the connective, which is the tissue found in the central region of the anther between the two lobes. In most plants, anthers are readily accessible to observations and manipulations; for these reasons, and in recognition of their role in the sexual reproduction of plants, anther developmental biology is one of the most extensively studied topics in plant embryology. Over the years, the goal of these studies, aptly termed "classical," has been to gain insight into the mechanisms by which the anther reacts to developmental information communicated to it. As a result, there has been an accumulation of considerable data on the histology underlying the differen-
tiation of specialized cells and tissues of the anther and the physiology of anther growth, as well on the isolation and characterization of genes preferentially expressed during anther development. In this chapter, most attention will be devoted to the molecular aspects of anther development in angiosperms leading to meiosis in the microsporocytes and the production of haploid microspores. However, for a proper appreciation of these studies, it is necessary to preface the accounts with brief surveys of a cluster of related problems that include anther determination and the ontogeny and physiology of the anther. These topics have been reviewed periodically during the past 30 years (Vasil 1967; J. Heslop-Harrison 1972; Mascarenhas 1975; Bhandari 1984; Shivanna and Johri 1985). 1. ANTHER DETERMINATION In the analysis of the factors concerned with stamen or anther determination in a flower, various experimental approaches coalesce. As in the case of leaves, the filament is the first-formed part of the stamen and the anther is a later-derived part. The stamens begin their life on the floral meristem, which initially is a mass of homogeneous cells but later becomes a mosaic of determined cell types. Following the initiation of the sepals and petals, the stamens arise as small, radially symmetrical outgrowths of cells with potentially organogenetic properties. These outgrowths presage determination, after which their developmental pathway becomes fixed as stamens. With further growth, the protuberances become cylindrical at the base and bilaterally symmetrical at the top. Once initiated, the stamen primordium completes its development rapidly, within a few days, to satisfy the structural 17
18
and functional requirements of the organ. During development, the position of the primordia on the floral apex apparently determines their identity as stamens and causes them to follow the characteristic genetically determined program. In this scenario, determined cells of each of the incipient stamens must assess or know their relative positions within the primordium and differentiate into the cell types appropriate for their parts, such as the filament, connective, and anther. However, the factors responsible for the way in which cells destined to become the anther determine their position on the stamen primordium, establish priority, and differentiate are not fully known. This is attributed partly to the microscopic size of the floral meristem and partly to the lack of clues about the diversity of causal factors. For example, a cell's knowledge of its position in the floral meristem preparing to form an organ could be imparted by the concentration gradient of a hormone in the meristem.
Gametogenesis
wild-type stamens has yielded further evidence of developmental flexibility of the stamen. Despite the striking phenotypic expression of the mutation, the primordia of normal and mutant stamens resemble each other. Whereas treatment of mutant plants at the stamen primordial stage with GA leads to the development of stamens with wild-type features, later application of the hormone results in the appearance of abnormal mutant stamens (Sawhney and Greyson 1973a, 1979). It is logical to interpret these data as indicating that the primordia that respond to GA are undetermined and are not affected by the mutation. In contrast, older primordia are presumably determined and become sensitive to the genetic lesion.
Insofar as overt morphological differentiation of the stamen is concerned, there is some evidence suggesting that cues from other floral organs on the meristem may influence the process. For example, in cultured floral meristems of tobacco, excision of the territory of the sepals and petals or suppression of growth of the sepals and petals interferes with (i) Timing of Anther Determination the formation of stamens (McHughen 1980). It also In an early study, stamen development was mon- appears that developmental interactions between itored in flower buds of Aquilegia formosa floral organs affect the initiation of stamens. Tepfer (Ranunculaceae), in various stages of development, et al (1963) found that in cultured flower buds of A. cultured in media supplemented with plant hor- formosa, the sepals had to be removed if the remainmones such as indoleacetic acid (IAA), 2,4- der of the floral organs were to faithfully recapitudichlorophenoxyacetic acid (2,4-D), gibberellic acid late their normal mode of differentiation. The (GA), and kinetin. Even the very early stage flower course of development pursued by the stamens is buds cultured in a favorable medium developed sta- also markedly affected by the presence of GA in the mens with normal anthers, indicating that determi- medium, because the hormone stimulates the nation has already occurred in the floral primordium growth of all floral organs except that of the sta(Tepfer et al 1963). Similar results were obtained mens. This observation is of interest in indicating when excised flower primordia of Nicotiana tabacum that the altered potential for organ initiation on the were cultured in a medium with or without kinetin floral meristem might result from its changing hor(Hicks and Sussex 1970; McHughen 1980). These monal status. observations do not necessarily identify a role for It is also possible that interactions between other hormones in fixing the developmental fate of cells in floral organs are involved in the formation of stathe differentiative pathway, nor do they give any mens in unisexual flowers. In Cucumis sativus insight into the process itself. (cucumber; Cucurbitaceae), there are separate Other studies have shown that under certain genetic lines of plants, which produce male, female, conditions it is even possible to alter the develop and bisexual flowers. Although the kind of organ mental pathway embarked upon by cells of the sta- destined to develop may be inherent in the bud men primordium. For example, Hicks (1975) found stage, there is compelling evidence from the culture that potential stamens of N. tabacum are partially of flower buds to show that the fate of the stamen is transformed into carpeloid and stigmatoid organs originally undetermined. Thus, very young potential when the primordia are cultured in a medium con- male buds tend to form carpels when cultured in the taining kinetin. This perplexing finding presumably basal medium; it is also possible to increase the indicates that the stamen is a developmentally flex- prevalence of femaleness in slightly older buds with ible organ of the flower. The reversion of stamens visible stamen primordia by the inclusion of IAA in with twisted, distorted anthers bearing naked exter- the medium. Simultaneous addition of IAA and GA nal ovules found in the stamenless-2 (sl-2) mutant of negates the effect of auxin in promoting carpel develLycopersicon esculentum (tomato; Solanaceae) to opment. Suggestive of an antagonism between the
Anther Developmental Biology
development of stamens and of carpels is the observation that in no case do the flowers become bisexual; no hormone treatment is also found to transform the potential female flowers into males (Galun, Jung and Lang 1963). These experiments provide tangible evidence for the operation of subtle and complex influences even as early as the determinative phase of anther development in the flower. Some observations on the role of kinetin in the selective stimulation of stamen development and of kinetin and GA in the development of the pistil and abortion of stamens in cultured ear shoots of Zea mays (maize; Poaceae) also seem to bear on the undetermined nature of the stamen (Bommineni and Greyson 1987,1990). (ii) Genetics and Molecular Biology of Anther Determination The fundamental influences that determine whether and where a stamen with a normal anther will differentiate on thefloralmeristem are ultimately attributed to genetic, biochemical, and molecular factors. In Nicotiana langsdorfii X N. alata1 hybrids, diver-
19
Maize is a particularly suitable object for genetic studies because many mutations that affect sex determination of the normally unisexual flowers are known (Veit et al 1993). Some mutations affect sex determination of the florets, permitting the development of pistils in the staminate floret. Tassel seed (ts) and tunicate (tu) are two mutants in this category that develop pistillate florets instead of staminate florets in the tassel; genetic analysis has shown that the genes concerned normally function in pistil abortion in the male florets (Irish and Nelson 1993; Irish, Langdale and Nelson 1994; Langdale, Irish and Nelson 1994). The molecular and biochemical basis for these mutations remains largely unknown. However, the sequence similarity between a TS protein and short-chain alcohol dehydrogenases (ADH) has fueled the speculation that the mutation may be brought about by a series of sex determination genes some of which act on gibberellins or steroidlike molecules of the floral meristem (DeLong, CalderonUrrea and Dellaporta, 1993).
Homeotic Mutants in Arabidopsis. A major gence of the stamen primordium from the normal contribution to our understanding of the mechadevelopmental pattern, giving rise to a stigmatoid nism of anther determination has come from an organ from the anther, is reported to be a heritable impressive series of studies on Arabidopsis and trait involving nuclear genes (White 1914). In N. deb- Antirrhinum majus; much of this work has been neyi X N. tabacum hybrids, development of stigmatic reviewed (Coen and Meyerowitz 1991; Meyerowitz characters on the anthers has been traced to cytoplas- et al 1991; Sommer et al 1991; Okamuro, den Boer mic elements of the female parent (Sand and Christoff and Jofuku 1993; Ma 1994; Haughn, Schultz and 1973). In this case, determination of the fate of the Martinez-Zapater 1995). Flowers of Arabidopsis are organ appears to have been made early in ontogeny, favorable for research because of their bilateral since the phenotypic expression is not altered by in symmetry consisting of only one whorl each of the vitro culture of undifferentiated, prestigmatic stamen calyx with four sepals, corolla with four petals, primordia (Hicks and Sand 1977). By observing androecium with six stamens, and gynoecium of changes in the phenotype of the stamen following two fused carpels. Other fortuitous biological feacrosses, these experiments have revealed something tures, such as the small size of the plant, its abbreof the origin of the genetic factors controlling deter- viated life cycle, and the small size of its genome, as mination of this floral organ. In line with the geneti- well as the availability of a good background of cally based biochemical differentiation of floral classical genetics, have also made this species a organs on the receptacle, polypeptides unique to sta- model system for research in plant molecular biolmens and carpels have been detected in one-dimen- ogy and genetics. In the approach used to generate sional and two-dimensional electrophoresis of mutants, seeds are chemically mutagenized and soluble and in vitro-translated proteins of floral pri- phenotypes that interfere with the orderly developmordia of certain plants, but the available data do not ment of floral parts are selected. A general characpermit informed speculation about the nature of the teristic of the mutations is that they cause havoc on proteins triggering sex expression (Sawhney, Chen the floral meristem as a whole, rather than specific and Sussex 1985; Bassett, Mothershed and Galau defects on the individual floral organs. Although 1988; Galli et al 1988; Bhadula and Sawhney 1989a; abnormalities in floral structure are usually held in Bommineni et al 1990; Rembur et al 1992; Thompson- abeyance in the wild type, mutations affect the difCoffe et al 1992). ferentiation of organs in particular positions on the meristem and produce abnormal flowers. Mutant 1 In designating crosses, the first-named is the female flowers therefore reveal the operation of genetic influences and indeed provide what is in many parent.
20
Gametogenesis
respects a convenient system for the molecular terminal portion to the DNA-binding regions of investigation of the determination of floral organs. transcription factors from a yeast gene that causes Especially useful in such studies are genetic lesions defects in the maintenance of minichromosome that produce floral organs in alien locations on the (MCM-1) and to the vertebrate serum response facreceptacle, such as petals in the place of stamens in tor genes (SRF). Based on this relationship, the the third whorl and a new flower primordium apparent uniqueness of the AG gene is that it probinstead of carpels in the fourth whorl (agamous [ag]) ably encodes a transcription factor that regulates (Bowman, Smyth and Meyerowitz 1989,1991), fac- genes determining stamen and carpel function in similes of carpels and stamens in the place of sepals wild-type flowers (Yanofsky et al 1990). The tranand petals, respectively, (apetala [ap-2],fl-40,fl-48, scription factor in all likelihood binds to key eleand leunig [lug]) (Komaki et al 1988; Kunst et al ments of the AG gene, thereby triggering a cascade 1989; Bowman, Smyth and Meyerowitz, 1989,1991; of regulatory events whose ultimate consequence is Liu and Meyerowitz 1995), carpels where stamens the activation of stamen- and carpel-determining are normally present {pistillata [pi], ap-3, and uni- genes (H. Huang et al 1993). The distinctive feature versal floral organs [ufo]) (Hill and Lord 1989; of the mutants defective at the AG locus is that they Bowman, Smyth and Meyerowitz 1989,1991; Levin fail to produce the transcription factor, thereby and Meyerowitz 1995), stamens that lack anther causing biochemical dysfunction of the stamenlocules (fl-54) (Komaki et al 1988), stamens in the and carpel-determining pathways. Additional AGplace of carpels (floral \flo-10], and superman [sup])LIKE genes (AGL-1 to AGL-6, AGL-8) that share this (Schultz, Pickett and Haughn 1991; Bowman et al conserved sequence motif have been isolated from 1992), carpels in the place either of sepals or of the flowers of Arabidopsis (Ma, Yanofsky and entire perianth (flo-2, flo-3, and flo-4) (Haughn and Meyerowitz 1991; Mandel and Yanofsky 1995). Somerville 1988), and secondary flowers in the The similarity between the phenotypes of ap-3 axils of leaflike first whorl organs of the primary mutant of Arabidopsis and deficiens (def) mutant of flower (ap-2) (Irish and Sussex 1990). The phenoAntirrhinum majus, to be described later, offered an types of these mutants, and the many ancillary experimental opportunity to use a DEF cDNA observations about them that are described in the clone to probe an ArabidopsisflowercDNA library references cited, leave little doubt that determinaand isolate AP-3 gene. The surprising observation tion of the floral organs in the wild-type plant is from a comparison of nucleotide and amino acid dictated by the products of the corresponding sequence data is that like DEF and AG, AP-3 progenes. Further analysis of the genetic control of tein contains a conserved motif similar to the tranorgan formation in the flowers of Arabidopsis has scription factors identified in other systems (Jack, involved cloning and sequencing of the genes and Brockman and Meyerowitz 1992). Molecular identification of their protein products. Cloning of cloning of AP-1 (Mandel et al 1992b) and PI (Goto the AG gene was facilitated by using a mutant line and Meyerowitz 1994) genes by other standard to tag genes using the T-DNA of the Ti plasmid of methods and determination of their amino acid Agrobacterium tumefaciens as the insertional mutasequences have also shown the protein products to gen. Transformation was accomplished by infectbelong to the same family of transcription factors ing germinating seeds of Arabidopsis with the as the AG gene product. Agrobacterium construct. One of the transformed Since the small size of the floral organ primordia lines selected by this method was found to display stable phenotypic alterations of the ag mutant. precludes RN A extraction and Northern blot analyBased on this observation and genetic analysis, it sis, gene expression in the floral meristem has been has become clear that T-DNA can be used as a mol- visualized by in situ hybridization of the compleecular probe to isolate the AG gene. This has mentary mRNA with labeled probes (Bowman, opened the way for the use of restriction fragments Drews and Meyerowitz 1991; Drews, Bowman and spanning the T-DNA insertion site to probe a Meyerowitz 1991). When sections of floral meriscDNA library made to poly(A)-RNA of wild-type tems of wild-type Arabidopsis plants of different Arabidopsis flowers, to select sequences comple- stages 35of development are hybridized with antimentary to the AG gene, to analyze their structure sense S-RNA probe derived from sequences and regulatory function, and to identify the protein within an AG cDNA clone, the label is detected first products they encode. The amino acid sequence in the floral meristem as the sepal primordia are inicomparison of the AG gene product based on DNA tiated. Subsequently, AG RNA is expressed almost sequence analysis showed similarities in the amino exclusively in the stamen and carpel primordia with barely detectable label in the sepals and petals.
Anther Developmental Biology
These data are consistent with the view that the AG gene plays a role in the early determination of the stamen and carpel. In contrast, AG RNA is present in the primordia of all four sets of floral organs of an ap-2 mutant that causes carpels and stamens to appear in whorls reserved for sepals and petals in the wild type. The paradox arising from these observations has been resolved on the basis of genetic evidence and data on the expression of AP1 gene in transgenic Arabidopsisflowerscarrying an antisense AG gene construct; the results corroborate the view that the function of the AP-2 gene product is to negatively regulate AG gene activity, rather than to act on its own (Bowman, Smyth and Meyerowitz 1991; Mizukami and Ma 1995). The product of the LUG gene appears to be earmarked for a similar function (Liu and Meyerowitz 1995). That the AG gene interferes with the expression of AP-1 activity has been revealed by in situ hybridization of AP-1 probe to sections of floral meristems of wild type and ag mutants. Although AP-1 transcripts are uniformly distributed during the early stage of development of the floral meristem of wild type Arabidopsis, they disappear from the potential sites of stamen and carpel primordia at later stages. Since this coincides with the expression of AG gene transcripts in the same sites of the floral meristem, a possible interaction between AG and AP-1 genes was followed by monitoring the expression of AP-1 in the floral meristem of an ag mutant. The surprising finding was that AP-1 RNA levels were quite high in the inner whorls of the flower when ag was in the mutant form. This gives strong support to the idea that AP-1 expression is negatively regulated by AG in the domains of the stamen and carpel primordia (Mandel et al 1992b; Gustafson-Brown, Savidge and Yanofsky 1994). Based on these and other results it has been proposed that organ specification in flowers depends upon regulatory interactions between different homeotic genes in the different parts of the meristem; this is no small feat considering all the network of peripheral reactions that cannot be left to chance (Bowman et al 1993).
21
ses, which showed that AP-1 and LUG are expressed in the sites of potential sepal and petal primordia for most of floral development, has led to the addition of ap-1 and lug to the list of mutants lacking A activity. From the functional perspective, the gene products are assumed to act in three overlapping fields such that field A controls the sepals and petals, B is responsible for petals and stamens, and C modulates stamens and carpels. By assigning genes AP-l/AP-2, AP-3/PI, and AG to function in fields A, B, and C, respectively, a picture of a unique pattern of expression of floral organs in each of the four whorls controlled by these gene products, acting singly or in combination, is apparent. Thus, when these conditions are fulfilled, cells in which AP-1/AP-2/LUG and AG alone are expressed might be expected to form sepals and carpels, respectively, whereas the combined effects of AP-1/AP-2/LUG and AP-3/PI will lead to petal formation. Similarly, both AP-3/P1 and AG gene products are necessary to create conditions for the formation of stamens (Bowman et al 1993). Provided the correct gene product is diverted to the floral meristem, any type of organ can indeed regenerate in any whorl of the flower, as shown by the observation that ectopic expression of AP-3 and PI genes gives rise to flowers with petals in the two outer whorls and stamens in the inner. The uniqueness of this ectopic expression is underscored by the fact that the same regions of the functional protein define the specificity of the two genes (Krizek and Meyerowitz 1996a, b).
Some of the results from in situ hybridization studies just considered are explained by yet another feature of the model, which proposes the existence of a mutual antagonism between A and C functions. Because this proposal implies that A function represses the action of C in whorls 1 and 2, and that C function prevents the action of A in whorls 3 and 4, it comes as no surprise that AG RNA is present in all four whorls of an ap-2 mutant. Direct evidence that the product of the AG gene prevents AP-2 function has come from the demonstration that chimeric AG gene introduced into The confluence of genetic and molecular studies transgenic Arabidopsis under the control of caulihas led to a working model to explain the function flower mosaic virus (CaMV) 35S promoter induces of mutant genes in the three extensively character- homeotic transformations of sepals into carpels, ized homeotic mutations, ag, ap-2, and ap-3 or pi and petals into stamens, very much like ap-2 (Figure 2.1). The model envisages that a different mutant phenotype (Mizukami and Ma 1992). genetic activity necessary for normal flower pat- Similar results were obtained when AG gene tern formation is missing in each of these muta- cloned from Brassica napus (rapeseed; Brassicaceae) tions. Designating these activities as A, B, and C, was expressed in the flowers of transgenic tobacco ap-2 mutant is assumed to lack A, ap-3 and pi (Mandel et al 1992a). These results clearly indicate mutants lack B, and C is missing in ag. Later analy- that homologous genes control the development of
22
Gametogenesis
NORMAL FLOWER SEPAL
1
PETAL STAMEN CARPEL
2
3
4
WHORL APETALA2 MUTANT LACKS ACTIVITY A CARPELSTAMEN STAMEN CARPEL
<SHS
STAMEN APETALA3AHO
P1ST1LLATA MUTANTS LACK ACTIVITY B
SEPAL SEPAL CARPEL CARPEL
-
Q
CARPEL
MM 1
2
3
Bftggag 4
WHORL X G 4 M 0 U S MUTANT LACKS ACTIVITY C SEPAL PETAL PETAL
SEPAL
Figure 2.1. A diagram showing the arrangement of floral parts in the wild-type and three groups of homeotic mutants of Arabidopsis. The genetic activities in each whorl are
shown in the block diagrams. (From E. M. Meyerowitz 1994. The genetics of flower development. Sci. Am. 271 (5):40-47. © 1994 by Scientific American Inc. All rights reserved.)
the stamens and carpels even in distantly related plants and that it is feasible to manipulate in a predictable manner the production of these floral organs on the receptacle by altering the expression of a single regulatory gene. AP-2 gene has been cloned and found to encode a putative nuclear protein that bears no significant similarity to any known plant, fungal, or animal regulatory protein. However, the gene is expressed at the RNA level in all organs of wild-type Arabidopsisflowersand is not restricted to sepals and petals as predicted by the model; this has raised some questions as to whether AG and AP-2 genes are mutually antagonistic (Jofuku et al 1994). AGL-1 to AGL-6 and AGL-8 genes isolated from
flowers of Arabidopsis have provided somewhat different expression patterns when sections of the floral meristem are hybridized in situ with 35S-labeled RNA probes complementary to AGL mRNA. For example, in contrast to AG, expression of AGL-1 and AGL-2 begins at a much later stage in floral ontogeny. Whereas AGL-1 is expressed preferentially in the carpels, AGL-2 is detected additionally in the sepals, petals, and stamens. Furthermore, AGL-2 is also expressed during early development of all four whorls of flowers defective in the organ identity genes AG, AP-2, and AP-3; the function of these genes is therefore not essential for AGL-2 expression per se. One possible implication of these results is that AGL-1 and AGL-2 may regulate a bat-
Anther Developmental Biology
23
evidence to account for the failure of AP-3 expression in the carpels formed in place of stamens in the pi mutant, there is compelling reason to believe that to form petals and stamens in the correct orientation on the meristem, interaction between AP-3, AG, and PI gene products is necessary (Jack, Brockman and Meyerowitz 1992). This general picture of interaction of AG, AP-3, and PI genes is further complicated by the spatial and temporal expression patterns of other genes that have been cloned subsequently. For example, although the phenotypic effects of ap-3 and pi mutations are nearly identical, the pattern of accumulation of their respective transcripts in the receptacle of wild-type flowers is strikingly variable. In contrast to AP-3 expression confined to the petal and stamen primordia of wild-type flowers, PI transcripts, at a similar stage of development, accumulate in the three inner whorls of the floral meristem. The mystery is deepened by the fact that this is a transient stage in the expression of PI transcripts because by the time stamens begin to differentiate, these transcripts are no longer expressed in the carpel territory; at this stage, the patterns of expression of PI and AP-3 genes become similar (Goto and Meyerowitz 1994). Interestingly, it has How the different genes interact at the floral been shown that when AP-3 is expressed in the apex is a conceptually puzzling problem, but some fourth whorl of flowers of a transgenic line of pointers have come from studies on the expression Arabidopsis, high levels of transcripts of PI accumupattern of the AP-3 gene in the wild-type and late in all three inner whorls throughout flower mutant flowers of Arabidopsis. Transcripts of AP-3 development. This provides good evidence that the are detected at the same stage of development in disappearance of PI transcripts from the carpel terthe floral meristem of wild-type flowers as those of ritory is due to the absence of simultaneous AP-3 the AG gene, but the subsequent expression of the expression (Jack, Fox and Meyerowitz 1994). gene is restricted to progenitors of the petals and Associated with the homeotic transformation of stamens. The coincident expression of AG and AP- stamens to carpels in the floral meristem of ufo 3 in the floral meristem specifying the petal, sta- mutant, there is a decrease in AP-3 and PI RNA men, and carpel primordia is just what one would expression, suggesting that the UFO gene holds a expect if initiation of the stamen primordium does sway over the activities of B-field genes (Levin and not depend upon the differentiation of other floral Meyerowitz 1995). Based on PI RNA expression organs on the apex. The expression pattern of the levels in the receptacles of sup and ap-2 mutant AP-3 gene in the floral meristem of ag mutant also flowers, SUP and AP-2 genes have also been idenoffers a clue about the relationship between AG tified as controlling PI gene expression (Goto and and AP-3 genes. In ag mutant, AP-3 gene expres- Meyerowitz 1994). The present view of the SUP sion is restricted to the second and third whorl gene is that it is a transcriptional regulator petals. This indicates that the AG gene product nor- involved in regulating the boundary between the mally does not regulate the function of the AP-3 stamen and carpel whorls on the receptacle (Sakai, gene. What is more appealing for speculative pur- Medrano and Meyerowitz 1995). poses about gene interactions in floral organ speciIn contrast to the genes promoting floral organ fication is the observation that although the AP-3 initiation, meristem identity genes control floral gene is normally expressed in the progenitors of meristem initiation and size, but there are indicaboth petals and stamens in the receptacle of pi tions that they also control the initial activation of mutant flowers, its later expression is restricted to the homeotic genes. The candidate genes that conthe petals. Despite the lack of sufficiently detailed trol such activation are LEAFY (LFY), AP-1, AP-2,
tery of genes required for the establishment of floral organs different from organs that are under the influence of AG and other genes (Ma, Yanofsky and Meyerowitz 1991; Flanagan and Ma 1994). Three other AGL genes studied in great detail are AGL-4, AGL-5, and AGL-8. The remarkable interaction of AGL-4 and AGL-5 genes with the AP-2 gene is seen in a significant increase in their transcript levels in ap-2 mutant flowers, suggesting that the AP-2 gene might negatively regulate the expression of AGL-4 and AGL-5 genes (Savidge, Rounsley and Yanofsky 1995). Stamens in the third whorl of wild-type flowers normally lack AGL-8 mRNA, but the gene is expressed in the carpels that replace stamens in ap-3 mutants. This assigns an undetermined role for the AP-3 gene product in blocking AGL-8 expression in the third whorl organs (Mandel and Yanofsky 1995). A general role has been ascribed to the AGL-3 gene because it is somewhat unique; unlike other AGL genes, its expression is global and is confined to all above-ground parts of the plant (Huang et al 1995). On the whole, the choice of whether a floral organ is established under the influence of one or the other set of organ identity genes probably depends upon the interaction of genes active in the meristem at the time of floral induction.
Gametogenesis
and UFO. Like ap-1 mutant, Ify mutation causes the partial transformation of flowers into abnormal structures, shoots, or inflorescences (Schultz and Haughn 1991, 1993; Huala and Sussex 1992). Because the determination of the floral meristem precedes the determination of floral organs, there is justification to consider that LFY and AP-1 genes are expressed before homeotic genes are activated; indeed, transcripts of cloned LFY gene are expressed strongly in wild-type flower primordia (Weigel et al 1992). In situ hybridizations performed with AP-3, PI, and AG gene probes on sections of floral meristems of wild type and single or double Ify and ap-1 mutants have indicated that the initial activation of the homeotic genes is accomplished by synergistic interaction with products of LFY and AP-1 genes (Weigel and Meyerowitz 1993). Apparently, the gap between the developmental pathway from a determined floral meristem to a determined flower primordium is bridged in a specific way by the products of LFY and AP-1 genes.
two outer whorls (Carpenter and Coen 1990). The def, glo, and sep mutants have phenotypes similar to the ap-3 and pi mutants of Arabidopsis and thus affect the B region on the receptacle. The pie mutant is similar to Arabidopsis ag mutant and affects the C region, whereas ovu, like ap-2, controls the A region. The DEF gene, which has been cloned, is also found to encode a protein with the same type of DNA-binding domain as AG and AP-3 genes from Arabidopsis. The functional motif shared by these diverse genes has been given the name MADS-box (for MCM-1, AG, DEF, and SRF) (Sommer et al 1990; Ma, Yanofsky and Meyerowitz 1991). In situ hybridization of sections of flower buds of wildtype A. majus has shown that DEF gene is very strongly expressed in the petal and stamen primordia (Schwarz-Sommer et al 1990); this is not surprising, since these organs are most sensitive to the effects of the mutation. Moreover, a homologue of the DEF gene isolated from wild-type Arabidopsis complements the ap-3 mutation in transgenic plants of the same species and produces flowers with petals and stamens of normal size, shape, and number (Okamoto et al 1994). As DEF and AP-3 have very similar protein products, mutant phenotypes, functions, and spatial expression patterns, we cannot fail to be impressed by the argument that the same molecular mechanisms probably operate in organ specification in the flowers of the two genera.
Homeotic Mutants in Antirrhinum and Other Plants. Following the discovery of homeotic mutants in Antirrhinum majus in the early 1920s, there has been a recent revival of interest in the molecular analysis of these mutants. This has been aided by the isolation and characterization of transposable elements that can be used for transposon tagging and mutagenesis. Additional advantages of A. majus for molecular studies are the large size To gain further insight into the molecular basis for of the flowers, availability of a good genetic map the interaction of the three sets of genes in the floral with many well-characterized mutations affecting meristem of A. majus, both PLE and GLO genes have flower development, and the ease of vegetative been cloned and characterized. The PLE locus was propagation by cuttings. However, unlike identified and isolated using AG from Arabidopsis as Arabidopsis, Antirrhinum flowers are zygomorphic, a probe. The functional similarity of PLE to the AG having only one plane of mirror-image symmetry, gene was shown by the restricted expression of PLE with a pentamerous arrangement of sepals, petals, transcripts in the stamens and carpels of wild-type and stamens, and two united carpels. In the Antirrhinum flowers. The complementary nature of approach followed to generate homeotic mutants, the phenotypes of pie (normally restricted to the two plants known to carry active transposons are self- inner whorls) and ovu (expressed in the two outer pollinated. Ml plants that are raised are also self- whorls) has been attributed to the opposite orientapollinated, allowing the isolation - in the M2 tions of a transposon in the intron of the PLE gene generation - of homozygous recessive mutations (Bradley et al 1993). The spatial and temporal expresdisplaying the same characteristic misrepresenta- sion patterns of DEF and GLO genes in the second tion of positions of the floral organs as in and third whorls of thefloralreceptacles of wild-type Arabidopsis. Among such mutations, those that are and various DEF and GLO alleles have implied that of interest to us at this stage are def, globosa (glo),the two genes are transcribed independently. From and sepaloidea (sep), in which the first three whorls other studies it appears that, whereas GLO or DEF of the flower are altered to yield sepal, sepal, and proteins do not bind DNA, in combination they bind carpel instead of sepal, petal, and stamen; plena to a motif found in the promoters of DEF or GLO (pie), with homeotic conversion of sex organs to genes (Schwarz-Sommer et al 1992; Trobner et al sterile whorls; and ovulata (ovu), in which carpels 1992; Zachgo et al 1995). These results favor a crossreplace sepals, and stamens replace petals in the regulatory role for the two gene products in specify-
25
Anther Developmental Biology
ing organ identity at the floral meristem; if this is correct, mutations could cause alterations in the transcription of genes normally controlled by MADS proteins and lead to homeotic changes in the floral meristem. These pioneering studies on Ambidopsis and Antirrhinum have not only resulted in a common genetic model of flower development but have also led to further research on the principles of organ identity in flowers of several other plants. MADSbox genes associated with floral organ development have now been described in tomato (Pnueli et al 1991, 1994a, b), Brassica napus (Mandel et al 1992a), Petunia hybrida (Angenent et al 1992, 1993, 1995a; Kush et al 1993; Tsuchimoto, van der Krol and Chua 1993; van der Krol et al 1993; Cafias et al 1994), tobacco (Hansen et al 1993; Kempin, Mandel and Yanofsky 1993), Solanum tuberosum (potato; Solanaceae) (Garcia-Maroto, Salamini and Rohde 1993), Aranda sp. (Orchidaceae) (Lu et al 1993), maize (Schmidt et al 1993; Mena et al 1995), and Sinapis alba (Brassicaceae) (Menzel, Apel and Melzer 1995). Close homology has been shown to exist between the homeotic genes BAG-1 of B. napus (Mandel et al 1992a), ZAG-1 of maize (Schmidt et al 1993), OM-1 of Aranda (Lu et al 1993), NAG-1 of tobacco (Kempin, Mandel and Yanofsky 1993), OSMADS-3 of Oryza sativa (rice; Poaceae) (Kang et al 1995) and the AG gene of Ambidopsis; between ZAP-1 of maize and AP-1 of Arabidopsis (Mena et al 1995); between ZAG-3, ZAG-5 of maize and AGL-6 of Arabidopsis (Mena et al 1995); between TM-6 of tomato and DEF of Antirrhinum (Pnueli et al 1991); between NTGLO of tobacco and GLO of Antirrhinum (Hansen et al 1993); and between OSMADS-2, OSMADS-4 of rice, GLO of Antirrhinum and PI of Arabidopsis (Chung et al 1995). MADS-box genes also appear to play a role in sex differentiation in the flowers of dioecious plants, Silene latifolia (Caryophyllaceae) (Hardenack et al 1994) and Rumex acetosa (Polygonaceae) (Ainsworth et al 1995). Using cDNA clones with MADS-box homology isolated from a male-flower cDNA library of S. latifolia, Hardenack et al (1994) have shown that each gene is expressed in the same floral whorl of the male and female flowers as is the homologous gene from snapdragon or Arabidopsis, such as AG, AP-3, DEF, and PI. The M^lDS-box gene is presumed to function here by reducing the size of the fourth whorl in the male flower. A recessive flower mutant described in tomato, which shows transformation of petals to sepals in whorl two, and of stamens to carpels in whorl three, also supports in principle the mechanism underlying
the phenomenon of homeosis in floral architecture (Rasmussen and Green 1993). In summary, tissue culture, genetic, and molecular investigations have led to some generalizations regarding determination of the floral organs, especially of the stamens and carpels. One is that floral organs are largely undetermined as they are initiated on the meristem, but the primordia soon acquire features of the organs they are destined to represent. Mutations that provide access to detailed molecular processes underlying determination of stamens indicate that specific genes direct stamen position on the floral meristem. However, sufficient data are not at hand to understand the role of genes in determining how the stamen primordium attains its mature form, consisting of the anther, filament, and connective, after its identity has been specified. Since the basic mechanisms that specify stamen determination are the same in widely distant species, the genetic organization for the functioning of genes and for their expression appears to have a long evolutionary history. 2. ANTHER DEVELOPMENT
Study of anther development in angiosperms has spanned more than a century and has provided a stable foundation for later work in the areas of physiology and, more recently, in molecular biology. Much of the early work is compiled by Maheshwari (1950); the later studies have been surveyed by Bhandari (1984). For physiological and molecular investigations, a nondestructive method of predicting the stage of microsporogenesis in the anther while it is still enclosed in the flower bud, is necessary. For several plants, this has resulted in the establishment of allometric relationships between the growth of the flower bud, the length of the anther, and the stage of microsporogenesis. Erickson (1948) showed for the first time that in the flower buds of Lilium longiflorum (lily; Liliaceae), from an initial length of 10 mm to a final length of 150 mm, cytological events of microsporogenesis occur at predictable anther lengths. This does not necessarily imply a steady distribution of activity along the entire length of the anther, as marking experiments have shown the existence of a rather basipetal wavelike distribution pattern of growth in length and cell division activity (Gould and Lord 1988). Mature florets and anthers of Oryza sativa attain only about 5% of the length of the flower buds and anthers of lily; however, the correlations observed between the growth of florets and anthers and stages of microsporogenesis in rice are similar
26
Figure 2.2. Anther development in rice, (a) Cross section of an anther primordium showing the mass of meristematic cells, (b) Differentiation of archesporial initials (arrows).
Gametogenesis
Both photographed with Nomarski optics. Scale bars = 10 (jm. (From Raghavan 1988.)
to those found in lily (Raghavan 1988). Significant and a primary sporogenous cell toward the inside. correlations have been established between the Although these two cells are dedicated to produce cytological stages of microsporogenesis and the different cell types, their origin from the same inimorphological features of anther development in tial cell indicates that they are ontogenetically Triticum aestivum (Bennett et al 1973b) and Zea mays related. When the distribution of mRNA [poly(A)(Chang and Neuffer 1989). RNA] in the anther primordium of rice is assayed 3 Initially, the anther consists of a mass of meris- by in situ hybridization using H-poly(U) as a tematic cells that do not display any differentiated probe, poly(A)-RNA is found to be uniformly discharacteristics, bounded on the outside by an epi- tributed in all cells. During the formation of the dermis. Subsequently, cell differentiation proceeds archesporial initial and the primary parietal and in such a manner that the anther becomes a four- primary sporogenous cells, there is no differential lobed organ. At this stage, each lobe becomes pop- accumulation of poly(A)-RNA in the progenitor ulated along its entire length by rows of large cells cells (Raghavan 1989). with dense cytoplasm and deeply staining nuclei. Later divisions of the primary parietal cell and There is virtually universal agreement that these the primary sporogenous cell can be summarized cells, which constitute the archesporial initials, are as follows. The former undergoes repeated anticlihypodermal in origin (Figure 2.2). Although multi- nal (perpendicular to the outer wall of the anther ple rows of archesporial initials are the norm in lobe) and periclinal divisions to form two to five anthers of many plants studied, in others the arch- concentric layers of cells constituting the anther esporium may consist of only a single longitudinal wall and tapetum. The primary sporogenous cell file of cells. Indicative of its morphogenetic poten- functions directly as the microsporocyte or resumes tial, each archesporial initial divides periclinally mitotic divisions to generate a mass of secondary (parallel to the outer wall of the anther lobe) to sporogenous cells that function as microsporocytes yield a primary parietal cell toward the periphery (Figure 2.3). Four microspores, with the haploid or
Anther Developmental Biology
Figure 2.3. Differentiation of wall layers and microsporogenesis in the anther of rice, (a, b) Cross sections of anther lobes showing two successive stages in the formation of the primary parietal cell (pp) and the primary sporogenous cell (ps). (c) An anther lobe showing the primary sporogenous cell (arrow) surrounded by a three-layered wall, (d) An anther lobe show-
27
ing the wall constituted of the epidermis (e), endothecium (en), and middle layers (m) and tapetum (t). The primary sporogenous cell has undergone cross-wise vertical divisions to form secondary sporogenous cells or microsporocytes (arrow). Sections were photographed with Nomarski optics. Scale bars = 10 urn. (From Raghavan 1988.)
28
gametic number of chromosomes, are born out of each microsporocyte by a meiotic division. This development, which in its extreme form may involve a single diploid cell, presents in a microcosm an excellent example of a meiotic pathway ending with the formation of specialized cells. There is a large body of anatomical evidence showing that the anther is derived either exclusively from the subepidermal layer of cells or from both subepidermal and epidermal cell lineages of the floral meristem. Utilizing transposable elements and X-rays to generate clonal sectors in the tassels, Dawe and Freeling (1990) showed that either the epidermal or subepidermal cell layer in maize can independently form the anther wall, but cell lineages from only the subepidermal layer generate the sporogenous cells. The total number of cells in each of the two layers participating in anther development has been estimated to be 12 or less (Dawe and Freeling 1992).
Gametogenesis
two locules of each anther lobe. As regards the middle layers, anthers inherit a fixed number, which is virtually invariant from species to species within a genus. Thus, the genome of each species is already programmed for determination of cell layers in the anther wall in a specific lineage and for the fate of the cells. However, after the middle layers are carved out, they become compressed, crushed, or obliterated by the burgeoning cells of the endothecium to the outside and the tapetum to the inside. The classic question as to how descendants of the same progenitor cell differentiate into structurally and functionally different cell types can be posed in the analysis of the anther wall. Although the role of the anther wall previously alluded to makes understanding of the physiology and molecular biology of its cells clearly important, little is known of the gene activation program in these cells. The problem here is compounded by the limited number of cells present in the anther wall and the difficulty of separating them from the sporogenous cells and pollen grains. Among examples of (i) The Anther Wall gene expression in the wall layers of the anther that It has long been tacitly assumed that the function have been investigated is the localization of in rice anthers by in situ hybridizaof the anther wall is to protect the sporogenous cells poly(A)-RNA 3 H-poly(U). Preparatory to meiosis in the tion with and developing microspores and to facilitate dismicrosporocytes, there is a sharp decrease in the charge of the baggage of enclosed pollen grains. What are the unique features of the anther wall that poly(A)-RNA concentration in the epidermis and serve this function? The wall of the mature anther middle layers of the anther wall, although the label consists of an epidermis, followed to the inside by a is predominant in the endothecium. Poly(A)-RNA layer of cells of the endothecium and cells compris- concentration attains low levels in the persistent ing two or three middle layers. In some plants, cells endothecium of the postmeiotic anther. An interof the anther wall and connective are traversed by pretation consistent with these observations is that plasmodesmata that allow assimilates to pass to the transcriptional activity is required for the differenlocule (Clement and Audran 1995). The most strik- tiation of the endothecium, perhaps in the fabricaing changes in the anther wall are observed in the tion of the radial thickening bands (Raghavan endothecial cells. As perceived in Lens culinare 1989). In anther sections hybridized with a 1.3-kb(Fabaceae), endothecial cells develop wall thicken- long H3 (arginine-rich) histone gene probe isolated ings at the vacuolate pollen grain stage; at the from a genomic library of 10-day-old rice seedlings, mature pollen grain stage, they harbor long strands transcripts are restricted to the endothecium of rough endoplasmic reticulum (ER), polysomes, (Figure 2.4); in line with this observation, the proand plastids with and without starch. Later, the moter of an H3 histone gene from wheat is found to cytoplasm becomes diffuse, accompanied by the confer expression of the GUS gene in the endotheappearance of fibrillar thickenings (Biddle 1979). cium of anthers of transgenic rice (Raghavan 1989; These thickenings constitute the beginnings of Terada et al 1993). In attempts to delimit the fibrous bands of the endothecial cells of mature nucleotide sequences in the 5'-region of wheat anthers. The bands, which arise chiefly along the H3 histone gene responsible for its expression, inner tangential wall, assume a U-shaped pattern, Terada et al (1995) demonstrated that replicationwith the gap directed toward the epidermis. During independent expression of the gene in the anther dehydration of the anther preparatory to dehis- wall of transgenic rice is not affected by mutations cence, the endothecial cells lose water and begin to in the highly conserved hexamer and octamer shrink in an uneven manner, resulting in the open- motifs; in contrast, the mutated gene is expressed at ing of the anther. Opening occurs through longitu- an appreciably reduced intensity in the rapidly dinal slits, known as stomium, located between the dividing cells of the root meristem of the transgenic
Anther Developmental Biology
29
m
Figure 2.4. (a) Autoradiographic localization of H3 histone probe in the longitudinal section of a premeiotic anther of rice, (b) Autoradiograph showing localization of the same probe in the section of a postmeiotic rice anther, e, epider-
mis; en, endothecium; m, microsporocyte; p, uninucleate pollen grains; t, tapetum. Scale bars = 10 urn. (From Raghavan 1989. J. Cell Sci. 92:217-229. © Company of Biologists Ltd.)
plant. Rice histone H3 gene probe also hybridizes Since the enzyme plays a key role in phenylto sections of anthers of a dicot, Hyoso/amus niger propanoid metabolism, presence of transcripts of (henbane; Solanaceae), indicating that the fidelity the gene in the lignified cells of the endothecium is of the gene is highly conserved between the two significant. major classes of angiosperms. In H. niger, autoradiAmong anther-specific cDNA clones identified ographic silver grains generated by in situ in a cDNA library made to poly(A)-RNA of tobacco hybridization are found to be present in more or anthers are those encoding a thiol endopeptidase less the same density in the cells of the anther (TA-56) and an unknown protein (TA-20) primordium and, later, in the epidermis and (Koltunow et al 1990). When RNA extracted from endothecium (Raghavan, Jiang and Bimal 1992). anthers and other floral organs and vegetative Unexpected as these findings are, they serve to parts of tobacco are hybridized with the cDNA indicate that histone transcripts are not necessarily clones in a Northern blot, they produce strong confined to actively dividing cells as one would hybridization signals with anther RNA only; this is expect if their protein products are used to complex considered as good evidence for anther specificity with DNA. Specific hybridization of probes made of the two cDNA clones and for their relative to the gene for 4-coumarate:coenzyme A ligase prevalence in the anther. Dot-blot hybridization of from Petroselinum crispum (parsley; Apiaceae) and RNA isolated from anthers of different stages of Solarium tuberosum to the cells of the endothecium development showed that although mRNA encodis also observed in anthers of tobacco transformed ing thiol endopeptidase accumulates from the with this gene (Reinold, Hauffe and Douglas 1993). stage of microsporocyte meiosis to the stage of the
30
Gametogenesis
STAGE 1 2 3 4 5 6 7 8 9 10 11 Control
# •
•
• i
•
Figure 2.5. Dot-blot hybridization of tobacco anther cDNA clones with anther mRNAs at different stages of development. TA-20 and TA-25 dots contained 0.25 ug mRNA; the other dots contained 0.05 fjg mRNA. Control dots contained
an equivalent amount of soybean embryo polysomal mRNA. Stages of anther development are described in Koltunow et al (1990). (From Koltunow et al 1990. Plant Cell 2:1201-1224. © American Society of Plant Physiologists.)
first haploid mitosis in the pollen grain, mRNA encoded by TA-20 becomes detectable long in advance and is present at the same level in anthers of all stages. Although the concentration of TA-56 mRNA remains relatively constant until the disruption of the connective, it declines to low levels thereafter (Figure 2.5). As predicted from this observation, in situ hybridization shows that at the earliest stage of anther development, transcripts for TA-56 are localized in the anther wall in the region that corresponds to the stomium. As the anther matures, the intensity of hybridization signals in the stomium decreases. The outcome was different in similar experiments using TA-20 mRNA as a probe; here the gene transcripts are expressed initially in all layers of the anther wall, although in older anthers they become concentrated in the connective adjacent to the vascular bundle. From the functional point of view, the early and from-the-outset-localized appearance of TA-56 transcripts at the stomium can be reasonably related to the fact that the gene belongs to a class whose primary function is to encode enzymes involved in the degeneration of cells in this part of the anther. Both TA-56 and TA-20 genes apparently function by responding to regulatory signals from the anther wall and are expressed during anther ontogeny when enzymes necessary for the differentiation of the anther wall are required. A similar function can be ascribed to genes encoding proline- and glycine-rich proteins, which are expressed in the epidermis of Helianthus annuus (sunflower; Asteraceae) anthers before pollen maturation (Domon et al 1990, 1991; Evrard et al 1991; Domon and Steinmetz 1994), and chal-
cone synthase and thionin, which are expressed throughout the wall layers of cells of tobacco anther (Drews et al 1992; Gu et al 1992). The expression of Phaseolus vulgaris (bean; Fabaceae) phenylalanine ammonia lyase (PAL-2)-GUS gene construct and of the gene for 4-coumarate:coenzyme A ligase from parsley and potato in the potential cells of the stomium of transgenic tobacco plants is thought to be critically related to the biosynthesis of lignin (Reinold, Hauffe and Douglas 1993; Shufflebottom et al 1993). Anther specificity has been established for several cDNA clones isolated from Lycopersicon esculentum anthers. Transcripts of these recombinants are also expressed in the somatic cells of the anther, especially in the epidermis and endothecium (Figure 2.6). The predicted protein sequence of the LAT-52 gene showed some homologies to the protein encoded by a pollen-specific gene from maize and to trypsin inhibitors (Twell et al 1989b; Ursin, Yamaguchi and McCormick 1989; McCormick 1991). Collectively interpreted, these data show that there are a few known molecular markers for the tissues of the anther wall. It should be noted that the evidence to date concerns genes that are selected from cDNA libraries made to poly(A)RNA of the entire anther. The appearance of lowabundance mRNAs during the development of the anther wall would escape detection by this method. Such sequences could be crucially significant for the analysis of gene expression in the cells of the anther wall. A series of antibodies with a high degree of specificity to the cells of the stomium have been raised in
31
Anther Developmental Biology
Strengthening the protective function of the anther epidermis in some unknown ways is envisaged as the role for patatin in the cells. (ii) The Tapetum
1 Ic
Figure 2.6. In situ localization of a tomato anther-specific cDNA clone (LAT-58) to longitudinal sections of tomato anther, (a) Drawing of the section of an anther to identify endothecium (en), filament (fl), and locule (Ic). Regions 1, 2, and 3 correspond to tip, middle, and base of the anther as shown by dashed lines, (b) Dark-field autoradiographs of longitudinal sections from regions 1, 2, and 3 hybridized with 35S-labeled RNA probe. (From Ursin, Yamaguchi and McCormick 1989. Plant Cell 1:727-736. © American Society of Plant Physiologists.)
tobacco flower extracts; this observation further supports the view that at the protein level also, the stomium is biochemically different from the rest of the cells of the anther wall (Trull et al 1991; Canas and Malmberg 1992). Immunocytochemical analysis has shown that patatin, the major glycoprotein of the potato tuber, is present in the epidermal cells of anthers of Solatium tubewsum and Capsicum annuum
(pepper; Solanaceae) (Vancanneyt et al 1989).
As the limiting tissue between the somatic cells and germ cells of the anther, the tapetum is in immediate contact with the sporogenous cells to the inside and with the middle layers to the outside. Compared to the cells of the anther wall, the tapetal cells are in a dynamic state during their short life period; they constantly undergo divisions; there is a substantial increase in the total DNA content of the cells, mostly by endoreduplication or polyteny (presence of bundles of interphase chromosomes composed of many parallel fibrils resulting from repeated rounds of DNA replication without separation of the daughter chromatids); cytoplasmic upheavals, such as accumulation of diverse kinds of complex macromolecules, are common in the tapetal cells; and the differentiated state of the tapetum is associated with the expression of unique genes. By virtue of its control of the passage of materials into and out of the anther locule, the tapetum has an important role in the nutrition of the sporogenous cells and microspores. The morphological, structural and cytological features of the tapetum that facilitate this function, as well as the other postulated role of the tapetum in the formation of the pollen exine, will be first considered here. This will be followed by an analysis of the cytology and gene expression patterns in the tapetum. For reviews of the evolution, structure, and function of the tapetum and its changing physiological relationship with the sporogenous cells, see Pacini, Franchi and Hesse (1985), Albertini, Souvr6 and Audran (1987), and Chapman (1987). Papers presented at a symposium on the tapetum collected together in the book edited by Hesse, Pacini and Willemse (1993) are also germane to our discussion. Structure-Function Relationships. Following the genesis of the tapetal cells, their subsequent fate might take one of two routes. In most plants, the tapetum remains as a single layer of cells characterized by a densely staining cytoplasm and prominent nuclei; however, in a few plants it becomes biseriate or multiseriate throughout the anther (Bhandari 1984). The tapetal cells attain their maximum development at the tetrad stage of microsporogenesis, after which degradative changes set in, resulting in the collapse of cells. Beginning with a study by Echlin and Godwin
32
(1968a), there have been several investigations on the ultrastructure of development of the tapetum in both dicotyledons and monocotyledons. These studies have clearly established that the tapetal cell is quite different from the cells of the wall layer to which it is ontogenetically related. However, the differences in the structure of the tapetum among various species are mainly concerned with the relative timing of developmental episodes compared with the stage of microsporogenesis, whereas generalizations mostly involve the type and distribution of the different organelles. In Avena sativa (oat; Poaceae), the newly formed tapetal cell walls are perforated by plasmodesmata that connect the cells to each other, to the pollen mother cells, and to the cells of the middle layer. A mechanism for the possible coordination of various functions of the tapetum emerges from the observation that during microsporocyte meiosis in Zea mays, the plasmodesmata between tapetal cells are replaced by cytoplasmic channels (Figure 2.7) that allow for free exchange of cytoplasm and organelles (Perdue, Loukides and Bedinger 1992). Microtubules, which show preferential orientations during the division of the tapetal cells of A. sativa, are also found in the peripheral cytoplasm of the nondividing cells along the tangential and radial walls (Steer 1977a). The rest of the cytoplasm of the cells is highly organized in spatial terms with a generous supply of mitochondria, plastids, dictyosomes, ribosomes, and smooth and rough ER. An interesting feature of the tapetal cells of A. sativa is the marked association among the smooth ER, plastids, and mitochondria. The majority of these organelles are completely encircled by the distended cisternae of the ER, but a few are ensheathed only partially; in either case, there are no direct connections between the organelles and the ER. It has been suggested that the ER-organelle association represents a route for the transfer of metabolites from the organelles to the anther locule. ER has also been implicated in the movement of dictyosome products to the surface of tapetal cells of A. sativa (Steer 1977b). In many plants, a distinct phase in the tapetal cell ontogeny that can be easily observed in the electron microscope at the time of microsporocyte meiosis is characterized by the amplification of ER, dictyosomes, mitochondria, and ribosomes (Echlin and Godwin 1968a, b; Christensen, Horner and Lersten 1972; Carraro and Lombardo 1976; Lombardo and Carraro 1976a; Pacini and Cresti 1978; Pacini and Juniper 1979b; Stevens and Murray 1981; Misset and Gourret 1984; Tiwari and Gunning 1986b; Keijzer and Willemse 1988;
Gametogenesis
Figure 2.7. Development of the wall of the tapetum in Zea mays, (a) Premeiotic tapetal cells and microsporocyte (m). Inset shows typical plasmodesmata between cells, (b) Postmeiotic tapetal cells and microspore (ms) soon after release from the tetrad. Pro-Ubisch bodies (pu) line the tapetal wall. Inset shows cytoplasmic channel. Scale bars = 1 um; inset scale bars = 0.5 um. (From Perdue, Loukides and Bedinger 1992.)
Audran and Dicko-Zafimahova 1992; Perdue, Loukides and Bedinger 1992; Brighigna and Papini 1993; Fernando and Cass 1994; Hess and Hesse 1994; Owen and Makaroff 1995). In Oka europaea (olive; Oleaceae), cytoplasmic vesicles found to peak in the tapetal cells during the meiotic phase of microsporogenesis are believed to function in the
Anther Developmental Biology
33
Figure 2.8. Electron micrographs of tapetal cells of Lycopersicon esculentum at the binucleate pollen grain stage, (a) A tapetal cell devoid of cytoplasm, but with orbicules lining the cell boundary, (b) Another tapetal cell with degenerated cytoplasm. Organelles are reduced to the nucleus (n) and isolated fragments of endoplasmic reticulum (er); part of the pollen grain (pg) is also seen. Orbicules
are seen within the tapetum as well, lining the outside wall. Scale bars = 1 urn. (From P. L. Polowick and V. K. Sawhney 1993a. Differentiation of the tapetum during microsporogenesis in tomato [Lycopersicon esculentum Mill.] with special reference to the tapetal cell wall. Ann. Bot. 72:595-605, by permission of Academic Press Ltd., London. Photographs supplied by Dr. V. K. Sawhney.)
transport of substrates mobilized from the anther wall (Pacini and Juniper 1979a). The roles attributed to the ER-organelle complex and dictyosomes in the tapetal cells of A. sativa and to the cytoplasmic vesicles in O. europaea appear to be attractive from the structural point of view, but as yet there is no clear evidence in support of the notions. This uncertainty reflects a general difficulty in assigning a role for the tapetum based on static electron micrographs alone. Of interest from the functional perspective is the presence of membrane-lined lipoidal bodies in the tapetal cells (Figure 2.8). These units are currently known as orbicules in preference to older terms such as Ubisch bodies or plaques (Echlin and Godwin 1968a; J. Heslop-Harrison 1968f; J. HeslopHarrison and Dickinson 1969; Risuefto et al 1969; Horner and Lersten 1971; Christensen, Horner and Lersten 1972; Horner and Pearson 1978; Stevens and Murray 1981; El-Ghazaly and Nilsson 1991; Polowick and Sawhney 1993a). As shown in Helleborus foetidus (Ranunculaceae), progenitors of the orbicules (pro-orbicules) appear at the sporoge-
nous cell stage in the tapetal cytoplasm as medium electron-dense vesicles in close association with ribosomes and ER. Coincident with the completion of microsporocyte meiosis, the tapetal cell walls lyse, extruding the pro-orbicules through the cell membrane into the anther locule. After their release into the locule, there is some mechanism, yet to be identified, that causes these bodies to become rapidly impregnated with sporopollenin and transformed into orbicules. From these observations it is possible to consider the orbicules as fragments of tapetal cytoplasm sufficiently organized to contribute to the formation of the pollen exine (Echlin and Godwin 1968a). A wall fraction isolated from the tapetal cells of anthers of Tulipa sp. (Liliaceae) by acetolysis was found to show an infrared spectrum similar to that of acetolyzed pollen grain from the same species, asserting the chemical identity of the two materials (Rittscher and Wiermann 1983). Materials originating from the tapetum of certain members of Asteraceae Q. Heslop-Harrison 1969b) and the gymnosperm Pinus banksiana (Pinaceae) (Dickinson 1970b) form a snugly fitting membrane
34
Gametogenesis
around the microspore tetrads. In Gasteria verrucosa is the glandular or the secretory type, in which the (Liliaceae) and Petunia hybrida, there is an excellent cell wall lyses, but the cell contents remain stationcorrelation between the timing of disappearance of ary and progressively become disorganized and Ca2+ from the taperum and its presence on the undergo autolysis; the other is the amoeboid or the pollen surface, suggesting that the degenerating periplasmodial type, in which the cell walls break tapetum is the source of Ca2+ that mediates in the down and the protoplasts fuse to form a periplaspollen recognition reaction on the stigma (Tirlapur modium. This syncytium later invades the anther and Willemse 1992; Bednarska and Butowt 1994). locule and envelops and nourishes the developing These observations tell us that orbicule formation is pollen grains. Rarely, as in Canna sp. (Cannaceae), not always the final fate of the material released by the tapetal protoplasts invade the locule individually without forming a periplasmodium (Tiwari the tapetal cells. Electron microscopic observations indicate that and Gunning 1986d). At the cellular level, the activthe proteins produced by tapetal cells aggregate in ities that culminate in the glandular mode of tapethe outer cavities of the exine. Perhaps this is not tum are analogous to those described in other cell surprising, given the coincidence between the cyto- types that senesce under natural or experimental chemical reactions of the proteins in the tapetum and conditions. In Helleborus foetidus, mitochondria and on the exine. These proteins, together with pigments, plastids are the first organelles that succumb as the lipids, and monosaccharides, apparently function as tapetum follows the degradative pathway. Later, recognition substances during pollen-stigma interac- similar considerations apply to the dictyosomes tions in plants. Another group of proteins generally and ribosomes, which show a decrease in number. found in the intine is enzymatic and antigenic in Shortly thereafter, the tapetum resembles a ghost nature and is probably produced in the pollen grain. cell consisting of only the plasma membrane In several members of Malvaceae (J. Heslop- together with the nucleus and ER (Echlin and Harrison et al 1973) and in Iberis amara (Brassicaceae) Godwin 1968a). A different mode of cell break(J. Heslop-Harrison, Knox and Heslop-Harrison down is seen in Citrus limon (lemon; Rutaceae), in 1974), the exine proteins are believed to have their which the tapetal cells intrude among the origin in the tapetal cells, specifically in the mem- microspores at the stage of starch accumulation; brane-bound cisternae of the ER, with a large num- later, the plasma membrane ruptures, releasing the ber of attached ribosomes and with inclusions of a protoplasm, which consists of more or less intact fibrillar material. Esterase is a marker enzyme for mitochondria, plastids, and nucleus (Homer and exine proteins that shows a dramatic increase in Lersten 1971). In Pisum sativum (pea; Fabaceae), activity in the tapetal cells of Brassica oleracea (cab-cytoplasmic breakdown of tapetal cells is reflected bage) beginning with the microspore to the stage of in their extreme vacuolation and is accompanied mature pollen grain. In contrast, acid phosphatase, a by the disintegration of the nucleus (Biddle 1979). A major difference between the glandular tapemarker enzyme for the intine proteins, shows barely tum and amoeboid tapetum is that the cells of the detectable activity in the tapetum (Vithanage and latter possess an organized, functional ultrastrucKnox 1976). The difficulty in assigning specific roles ture even as they go into a decline. This is evident to the exine and intine proteins is underscored by the fact that both apparently act together in the recogni- from the detailed account of development of the tapetum in Tradescantia bracteata (Commelinaceae) tion reaction on the stigmatic surface. (Mepham and Lane 1969) and T. virginiana (Tiwari Claims made to the effect that DNA from the and Gunning 1986a, b). In addition to the presence tapetum, especially the free nucleic acid bases and of abundant rough ER, plastids, mitochondria, and nucleorides, is transferred to the microsporocytes dictyosomes, the tapetal cells accumulate raphides during meiosis have proven to be spurious. By at the microsporocyte stage. Another feature of this monitoring the fate of labeled tapetal nuclei during phase of tapetal development is the presence of microsporogenesis in Lilium longiflorum, it has been small vesicles and their coalescence, followed by shown that materials from the disintegrating the convolution of the plasma membrane and lysis tapetal nuclei are not incorporated into either the of the cell wall. During meiosis, the tapetal cytosporogenous cells or the microsporocytes (Takats plasm, now bereft of its plasma membrane, ER, dic1959,1962). tyosomes, and microtubules, streams into the What happens to the protoplasm of the col- anther locule, where it engulfs the microsporolapsed tapetal cells is just as important as how it cytes. The reorganization of the naked protoplasm functions in the intact cell. From this perspective, that occurs at this stage involves the reappearance two principal types of tapetum are recognized. One
Anther Developmental Biology
35
of rough ER, dictyosomes, and microtubules, the (Gentianaceae) (Lombardo and Carraro 1976b), appearance of electron-lucent vesicles, and the syn- Tradescantia virginiana (Tiwari and Gunning 1986b, thesis of enzymes that degrade callose (a polymer c), and Butomus umbellatus (Butomaceae) (Fernando of (l,3)-F3-glucan). The periplasmodium that sur- and Cass 1994) has led to some questions about the rounds the tetrads develops long amoeboid validity of orbicules as a distinguishing character. processes, the callose walls of the tetrad being pro- This should perhaps alert us to focus on the formagressively degraded in front of them. Other signs of tion of the periplasmodium as the most stable critapetal activity evident at this and later stages are a terion to differentiate between glandular and decrease in the number of raphides, accumulation amoeboid types of tapetum. of plastoglobuli in the plastids and their release Among the substances released from the degeninto the cytoplasm, proliferation of both rough and erating tapetum is pollenkitt, a generic term smooth ER, accumulation of granular material in applied to compounds imparting stickiness to the the smooth ER, and increase in polysomes and dic- pollen. Pollenkitt is composed of lipoidal materials, tyosomes. No signs of disintegration are seen in the flavonoids, carotenoids, and degeneration prodtapetal nuclei. At the binucleate pollen grain stage, ucts of tapetal proteins. Based on an assay using the tapetal plastids enter a new phase of polysac- isolated cell layers of Tulipa sp. anthers, the comcharide synthesis. These changes in tapetal cells are plete sequence of enzymes involved in flavonoid analogous to reorganization rather than degenera- metabolism has been localized predominantly in tion; this view is supported by the work of Souvre the tapetal cells (Herdt, Siitfeld and Wiermann and Albertini (1982) on tapetal development in 1978; Rittscher and Wiermann 1983; Kehrel and Rhoeo spathacea (Commelinaceae) and Owens and Wiermann 1985; Beerhues et al 1989). In Lilium Dickinson (1983) in Gibasis karwinskyana and G. longiflorwn 0. Heslop-Harrison 1968f; Dickinson venustula (Commelinaceae). In the tapetal periplas- 1973), L. hybrida (Reznickova and Dickinson 1982; modium of Arum italicum (Araceae), cytoplasmic Keijzer 1987), Oka europaea (Pacini and Casadoro reorganization mainly centers around a cluster of 1981), Raphanus sativus (radish; Brassicaceae) microtubules that acquire various orientations in (Dickinson 1973), Gasteria verrucosa (Keijzer 1987), the vicinity of the microsporocytes and later sur- Anemarrhena asphodeloides (Liliaceae) (Chen, Wang round the microspores (Pacini and Juniper 1983). and Zhou 1988), Brassica oleracea, B. napus (Murgia et al 1991a; Evans et al 1992), and Apium nodiflorum Some of the aforementioned facts suggest that (Apiaceae) (Weber 1992), coalesced lipids or their the amoeboid tapetum is not only the site of the precursors, proteins, and carotenoids of pollenkitt packaging of metabolites, but also of their synthesis. Whether these are exclusive roles performed by are synthesized in the plastids, which degenerate the amoeboid tapetum is not certain. Cytochemical prior to tapetal collapse. At this stage the contents and autoradiographic experiments on the amoe- of plastids are seen as osmiophilic globules freely boid tapetum of R. spathacea have shown increases distributed in the cytoplasm of the tapetal cells. in the amount of polysaccharides and basic and Tapetal activity is terminated by the collapse of acidic proteins of the tapetal cells and in the incor- cells and the release of the globules. It is not poration of 3H-acetate and 3H-choline into the lipids unusual for the tapetal cytoplasm also to partici(Albertini, Grenet-Auberger and Souvre 1981) and pate in the production of pollenkitt (Pacini and of 3H-tryptophan into proteins (Albertini 1975) of Casadoro 1981; Pacini and Keijzer 1989). the cells during the tetrad stage of microsporocyte Osmiophilic bodies harboring pollenkitt precursors meiosis. Demonstration of synthetic activity in the are associated with the ER during tapetal developtapetal cells is an example of the type of evidence ment in Ledebouria socialis (Hyacinthaceae) anthers; that is necessary to establish a correlation between however, it is not established whether these bodies are derived from the ER (Hess and Hesse 1994). A the structure and function of the tapetum. variety of roles, such as insect attractant, agent for As electron microscopic observations show, wall pollen dispersal, and an antidote against the damdissolution and invasion of the locule by the tapetal aging effects of UV rays, have been attributed to contents occur in both glandular and amoeboid the pollenkitt. Lipoidal components of the poltypes of tapetal cells. Presence or absence of lenkitt have an important role in controlling water orbicules was previously considered as a striking loss from the germ pores of the exine following difference between the two types of tapetal cells; pollen dispersal. The pollenkitt is also probably however, the presence of bodies having similar involved in controlling the recognition reactions on electron density and structure as orbicules in plasthe stigmatic domain. Tryphine is often treated as modial tapetal cells of Gentiana acaulis
36
Gametogenesis
different from pollenkitt. It is a mixture of other cases, the chromatids formed in each chrohydrophilic substances derived from the break- mosome by DNA replication remain aligned with down of tapetal cells, whereas pollenkitt is princi- one another, forming polytene chromosomes. In pally made up of hydrophobic lipids containing some species with multinucleate tapetal cells, cytospecies-specific carotenoids. In R. sativus, the main logically complex cells are formed when restitution bulk of the tryphine is synthesized in the tapetal nuclei undergo endoreduplication, resulting in cells after they have been stripped off their walls; nuclei with high chromosome numbers and high this coincides with the stage of the first haploid DNA content (D'Amato 1984). Ploidy levels in the pollen mitosis. Tryphine apparently consists of pro- uninucleate tapetal cells range from 4N in teinaceous materials secreted by the ER and degen- Helleborus foetidus (Carniel 1952) to 16N in Cucurbita erate lipid-rich cytoplasm, along with lipid-filled pepo (Cucurbitaceae) (Turala 1958). In plants with plastids. It is believed that besides pollenkitt, multinucleate tapetum, 4N nuclei formed by restitryphine is the main encrustation of the pollen tutional mitosis are common in Zea mays (Carniel exine and aids in the adhesion of pollen grains to 1961), whereas 32N nuclei by endomitosis and the stigma and also possibly regulates recognition restitutional mitosis are found in Lycopersicon escureactions on the stigma (Dickinson 1973; Dickinson lentum (Brown 1949). and Lewis 1973b). The timing of synthesis and accumulation of DNA in tapetal nuclei was demonstrated in the Cytology and Gene Expression. Few areas of early applications of autoradiography and microinvestigations on the anther demonstrate more densitometry to plant tissues and later by histoforcefully the interplay of old-fashioned cytology chemical methods. The first autoradiographic and modern molecular biology and genetic engi- studies carried out using 32PO4, 3H-cytidine, and 14 C-thymidine showed that tapetal cells of Tulbaghia neering than does the study of the tapetum. Tapetal cells may be uninucleate, binucleate, or multinucle- violacea (Liliaceae) (Taylor 1958), Lilium longiflorum ate. Nuclear divisions are initiated in the cells dur- (Taylor 1959), and L. henryi (J. Heslop-Harrison and ing meiotic division of the microsporocytes and Mackenzie 1967), respectively, begin to synthesize may continue through the active life of the tape- DNA after the microsporocytes enter meiosis. tum. According to Wunderlich (1954), who has Earlier, microdensitometry had shown that the reviewed the cytology of the tapetum of DNA content of tapetal nuclei of L. longiflorum angiosperms, the multinucleate tapetum is more increases during microsporocyte meiosis, followed common than the uninucleate type and has conse- by a decrease and then an increase even up to 8C quently received much experimental attention. level (Taylor and McMaster 1954). In maize, the Since the early 1900s, several reports of abnormal tapetum reveals a unique pattern of DNA accumudivisions of tapetal nuclei resulting in increased lation. Although the early cell divisions are not DNA contents, polyploidy, and polyteny have also accompanied by a modal increase in DNA values, appeared to suggest that nuclear aberrations are an synthesis does ensue during the later times and intrinsic feature of tapetal cytology. Although these attains AC values (Moss and Heslop-Harrison abnormal divisions have been precisely docu- 1967). According to Albertini (1971), tapetal cells of mented, their impact on the function of the tape- Rhoeo spathacea begin to incorporate 3H-thymidine tum, for the most part, is largely unknown, and as the microsporocytes enter the pachytene stage of none have yet been studied at the molecular level. meiosis. In Nigella damascena (Ranunculaceae), the For these reasons, only a summary of the cytologi- intensity of Feulgen staining of the tapetal nuclei cal abnormalities is given here. Because distur- appears to increase during the entire period of bances in the mitotic spindle result in a slow meiosis of the microsporocytes (Bhandari, Kishori movement of individual chromosomes and due to and Natesh 1976). The range of C values observed their stickiness, polyploid restitution nuclei are a in the tapetal nuclei of Petunia hybrida varies from frequent occurrence in both uninucleate and multi- 2C to 4C during meiotic prophase in the nucleate tapetal cells. A variation of this theme, microsporocytes, advancing to 7C to 8C at the early resulting in the same end product, namely, poly- pollen grain stage (Liu, Jones and Dickinson 1987). ploidy, is endomitosis, which causes a doubling of Thus, the timing mechanism responsible for DNA chromosome number without the formation of a synthesis in the tapetal nuclei appears to be linked mitotic spindle. Endoreduplication leads to modal to the entry of the sporogenous cells into the meiincrease in DNA values of the nuclei of tapetal cells otic cycle. in the absence of chromosome duplication. In yet Because of the increase in genome size of the
Anther Developmental Biology
37
tapetal cell nuclei as revealed by observations just 3H-leucine incorporation is detected at later stages described, it is clear that they engage in wide- (Sauter 1969a). As shown by 3H-arginine and 3Hspread and intense transcriptional activity. tryptophan incorporation, tapetal cells of Evidence for transcription has come from the Tradescantia paludosa synthesize histone and nonhiautoradiographic pattern of incorporation of pre- stone nuclear proteins coincident with microsporocursors of RNA synthesis into tapetal cells. As cyte meiosis (De 1961). Moss and Heslop-Harrison shown by 3H-uridine labeling, in R. spathctcea and L. (1967) showed that like RNA, proteins are present longiflorum (Albertini 1965, 1971; Williams and in the premeiotic tapetum of maize anthers, but Heslop-Harrison 1979), marked RNA synthesis increase to attain a peak at the tetrad stage. By occurs in the tapetal cells at all stages of meiosis, immunological methods, two specific proteins with a peak during diplotene. Somewhat similar have been shown to accumulate exclusively in the results were obtained by Sauter (1969a) on the tapetum of lily anthers coincident with the peak of incorporation of 3H-cytidine into the tapetum of tapetal secretory functions (Wang et al 1992a, b, Paeonia tenuifolia (Ranunculaceae). Although L. 1993; Balsamo et al 1995). longiflorum has the glandular type and R. spathacea Although protein synthetic activity of the has the amoeboid type of tapetum, it is interesting tapetal cells might reflect changes in enzyme titer, to note that RNA synthetic activity is related to the there is little experimental support for this belief. stage of microsporogenesis and not to the ultimate The only published study that addresses this quesfate of the tapetum. tion is that of Linskens (1966), who found marked Transcription of mRNA in the tapetal cells has fluctuations in the activity of lactic dehydrogenase, been visualized by in situ hybridization with 3H- thiamine pyrophosphatase, and acid phosphatase poly(U) as a probe. As seen in Figure 2.9, in anther in isolated tapetal cells of L. henryi; these fluctuasections of Hyoscyamus niger, the tapetal cells act as tions were related to the stages of microsporocyte a focal point for binding of the label from the time meiosis. It seems probable that tapetal metabolism of completion of meiosis in the microsporocytes to is the outcome of a stage-specific gene expression the stage when the tapetum begins to disintegrate program. (Raghavan 1981a). A similar pattern of poly(A)RNA distribution has been observed in the anthers Tapetum-specific Genes. Recent investigations of rice (Raghavan 1989). The absence of poly(A)- have correlated the expression of certain unique RNA accumulation in the premeiotic tapetum classes of mRNAs with the differentiated state of the argues against any active role for this tissue in the tapetum (see Schrauwen et al 1996 for review). From nutrition of the microsporocytes or in inducing a cDNA library made to poly(A)-RNA of preanthemeiosis in them. In contrast, the tapetal tissue of sis stamens of tomato, Smith et al (1990) isolated the anthers of Lilium sp. is found to anneal 3H- three tapetum-specific clones that are stringently poly(U) throughout meiosis (Porter, Parry and regulated in this tissue over time. As shown by Dickinson 1983). Intense annealing of a cloned rice Northern blot, clones 92-B, 208, and 127-A hybridize histone H3 gene to tapetal cells of H. niger presum- to RNAs of 600, 700, and 750 bp, respectively, and ably indicates that histones are made in the tapetal are detected only in RNA isolated from stamens cells (Raghavan, Jiang and Bimal 1992). The overall prior to anthesis. In situ hybridization showed that impact of these studies is to confirm that long their expression is limited to the tapetum spanning a before the tapetum begins to disintegrate, it narrow window from late microsporocyte meiosis to acquires a complex set of specifications in the form immature pollen grain stage; this provides a conof mRNAs. vincing correlative interpretation of the data from Tapetal cells also actively synthesize proteins, Northern blot (Figure 2.10). The deduced proteins of although there is some variation between species in tapetum-specific clones have the characteristics of the periods of peak synthetic activity. At the same glycine-rich or cysteine-rich secretory proteins time, there are similarities in the timing of accumu- (Aguirre and Smith 1993; Chen and Smith 1993; lation and synthesis of RNA and proteins in the Chen, Aguirre and Smith 1994). Tapetum-specific tapetum. As noted by Taylor (1959), intense incor- genes that encode small secretory proteins have also poration of 14C-glycine into proteins of tapetal cells been isolated from flower buds of Antirrhinum majus occurs during the later stages of meiosis in the (Nacken et al 1991a, b), flowering apices of Sinapis microsporocytes of L. longiflorum. In contrast, in P. alba (Staiger and Apel 1993; Staiger et al 1994), tassels tenuifolia, compared to tapetal cells at the early of maize (Wright et al 1993), and microsporocytes of stage of microsporocyte meiosis, less than 30% of Lilium henryi (Crossley, Greenland and Dickinson
38
Gametogenesis
.^'''
Figure 2.9. Autoradiographs showing the binding of 3 Hpoly(U) to tapetal cells of anthers of Hyoscyamus niger at different stages of development, (a) Anther with microsporocytes (ms) at an early stage of meiosis. (b) Anther with microspores (m) after release from tetrads, (c) Early unicel-
\
lular pollen grain stage of the anther (p). (d) Unicellular, partially vacuolate pollen grain stage of the anther (p). In all cases, arrows point to tapetal cells. Scale bars = 20 um. (From Raghavan 1981a.)
Anther Developmental Biology
39
250 \xm
Figure 2.10. Localization of 35S-labeled antisense and sense probes of tapetum-specific cDNA clone 108 in sections of tomato anthers, (a) Section hybridized with the antisense probe showing intense localization in the tapetal cells (t). (b) Section hybridized with the sense probe showing lack of hybridization in the tapetum. Little or no hybridization is seen in the microspores (m) of both sections. (From Smith et al 1990. Mol. Gen. Genet. 222:9-16. © Springer-Verlag, Berlin.)
1995). By immuno-electron microscopic methods, the secreted proteins have been localized in the boundary between the tapetum and the middle layers of the anther (peritapetal membrane) and in the exine of early vacuolate microspores of S. alba (Staiger et al 1994). The identification of a protein of tapetal origin in the microspore exine by a high-resolution method is significant as support of the view that some of the exine materials are synthesized in the tapetum. Several cDNA clones (TA-13, TA-26, TA-29, TA-32,
and TA-36) that represent transcripts expressed exclusively in the tapetum have been described from the tobacco anther cDNA library considered earlier (Koltunow et al 1990). These transcripts encode pro-
teins that are functionally important in the physiology of the tapetum. For example, TA-13/TA-29 and TA-32/TA-36 mRNAs encode glytine-rich cell wall proteins and lipid transfer proteins, respectively, that are present in the tapetum and are probably transferred to the exine of the pollen grains. Southern blot and DNA sequencing studies have indicated that each of these mRNAs is encoded by a small gene family (Seurinck, Truettner and Goldberg 1990); mRNAs for these genes accumulate coordinately in the tapetum following microsporocyte meiosis and decay before the first haploid pollen mitosis. Because positive signals are obtained by hybridizing blots containing TA-29 gene sequences with run-off transcripts made by isolated anther nuclei, it has been concluded that tapetal-specific expression of this gene is controlled at the transcriptional level. This is based on the assumption that run-off transcripts reflect in vivo transcription, since the intact nuclei should preserve the native state of chromatin and its regulatory proteins. To localize the tapetum-specific TA-29 transcriptional domain, chimeric TA-29/GUS gene or TA-29/GUS gene containing a minimal CaMV 35S promoter (-52 to +8 bp) and different amounts of the 5'-region of TA-29 gene were constructed. The chimeric genes were next transferred to tobacco plants using the disarmed Agrobacterium tumefaciens Ti plasmid vector in the tobacco leaf disc assay. This assay (Figure 2.11) showed that a 122bp region (-207 to -85 bp) of TA-29 gene contains all the information required to program tapetal-specific expression in anthers of transgenic plants (Koltunow et al 1990). In another work, Mariani et al (1990) demonstrated that, like the hybridization profile of
-279
-207 -178 -150
-85
Figure 2.11. Deletion analysis of tobacco TA-29 gene. Schematic representation of the -279 bp 5'-region of chimeric TA-29/CUS constructs with or without a minimal CaMV promoter. Shaded 5'-regions represent chimeric genes that produced the tapetal-specific GUS expression. NOS, nopaline synthase gene, used as a selectable marker. (From Koltunow et al 1990. Plant Cell 2:1201-1224. © American Society of Plant Physiologists.)
40
Gametogenesis
TA-29 antisense RNA probe in the normal tobacco plant, TA-29IGUS antisense RNA probe produces intense hybridization signals in the tapetum of the transformed tobacco plant. In sum, it appears that sequences within a restricted domain of the TA-29 gene are required to activate transcription factors in the anther tapetum during specific times in the life cycle of the tobacco plant. Several genes (A-3, A-8, A-9, BA-112, and BA118) isolated from a cDNA library made to poly(A)-RNA of anthers of Brassica napus at the sporogenous cell stage have been found to be tapetum specific. Characterization of the A-9 gene has shown that its predicted protein sequence has homology to a family of seed proteins. It is difficult to assign a critical role for seed proteins in the development of the tapetum; in line with this fact, it was found that transgenic B. napus plants containing A-9 antisense gene are phenotypically normal and produce viable pollen grains (Scott et al 1991; Paul et al 1992; Shen and Hsu 1992; Turgut et al 1994). In a similar way, the tapetum specificity of a storage protein gene from the endosperm of maize in transgenic Petunia hybrida is puzzling (Quattrocchio et al 1990). Another cDNA clone from B. napus (A-6) and its corresponding Arabidopsis genomic clone encode a protein with some similarity to callase (l,3)-(3-glucanase. Because the A-6 gene is tapetum specific and is expressed in the tapetum of transgenic plants with a peak in activity coincident with callase activity in the anther, its protein is considered to be a component of the callase enzyme complex (Hird et al 1993). A gene for callase isolated from the anthers of tobacco is expressed in the tapetum just prior to tetrad dissolution (Bucciaglia and Smith 1994). Clones from a cDNA library made from anthers of B. napus at the microspore stage appear to be expressed in the tapetum as well as in the microspores. A protein encoded by a cDNA clone (E-2) from this group is found to be homologous to a lipid transfer protein, possibly functioning in the formation of both membranes and storage lipids; another cDNA clone encodes a novel proline-rich protein with no homology to any other known protein (Scott et al 1991; Foster et al 1992; Roberts et al 1993a). Acetolactate synthase is an enzyme in branched-chain amino acid biosynthesis, and the expression of transcripts of the gene encoding this enzyme in the tapetal cells of tobacco anthers suggests a requirement for branched-chain amino acids for tapetal function (Keeler et al 1993).
source of lipid materials has been brought into sharp focus by the characterization of genes related to oleosins from a flower cDNA library. Oleosins are specialized proteins of the electron-dense proteinaceous membrane enclosing the naked oil droplets of storage triglycerides; they are present in embryos and endosperms of various plants, as well as in pollen grains of B. napus. The transcripts of these genes are specifically expressed in the tapetum during a narrow window in its active life when lipid accumulation is also believed to occur (Robert et al 1994c; Ross and Murphy 1996). In this context, the demonstration that the promoter of a stearoyl-acyl carrier protein (ACP) desaturase gene isolated from embryos of B. napus is active in the tapetum of anthers of transgenic tobacco, is noteworthy (Slocombe et al 1994).
By differential screening of a mature B. campestris pollen cDNA library, Theerakulpisut et al (1991) have identified a clone, BCP-1, whose unique feature is its occurrence in the tapetum and in pollen grains. In situ hybridization with probes labeled with digoxygenin-deoxyuridine triphosphate (dUTP) reporter molecule and an alkaline phosphatase detection system showed that the gene is activated in the tapetal cells beginning at the bicellular pollen grain stage and that its expression continues until tapetal dissolution. The molecular mass of the putative protein encoded by BCP-1 mRNA is approximately 12 kDa. Although the identity of the protein has not been established, it appears to be a major pollen protein. Transcripts of two clones isolated from a cDNA library made from rice anthers at the microspore stage are localized exclusively in the tapetum. Further analysis of the genomic clone of one cDNA revealed that its upstream region can regulate GUS gene expression in the tapetum of transgenic tobacco. A set of 5'-deletions generated from this clone (Figure 2.12) showed that tissue regulatory elements reside between the -1,273 and -1,095 bp regions of the gene for promoter activation in transgenic tobacco (Tsuchiya et al 1994). Mitochondrial replication occurs very actively in the developing tapetum to meet the changing energy demands of the cell. The dispersion of genes in the mitochondria suggests that the developmental control of organelle biogenesis may be regulated at the level of mitochondrial gene expression. In attempts to equate increasing mitochondrial number to changes in mitochondrial gene expression, it has been shown by tissue imprinting that both nuclear-encoded alternative peroxidase protein and The importance of the tapetum of B. napus as a mitochondrially encoded proteins of oc-subunit of
Anther Developmental Biology
-1500
-1000
H
1—
-500
—r—
-1648 [
41 GUS Activity (4MU nmol/mln/mg protein)
+1
H JWWXtttfT
0
xo
4
a i H^
iy>
8 O CDO
3Dcoo co a n o
-1469 [_ -13471
' V
JJ
-1273 1
XEDOO 003300 O HXEOO CO
-1095 [_
_u
-M7[
>o mo
12
03
16 O
oo O O
O
O
O O
as 0
uo -123 | ] _ Osg6B promoter
30)
^ TATA Box
NosT Control
J
Figure 2.12. Deletion constructs of a tapetum-specific rice genomic clone (OSG-6B) and their expression in transgenic tobacco. The deletion constructs are indicated on the left
and GUS activity in individual plants in each group and in an untransformed control plant are plotted on the right. (From Tsuchiya et al 1994.)
Fj-adenosine triphosphate (ATP) synthase (ATP-A) and cytochrome oxidase-c subunit-II (COX-II) accumulate in the tapetal cells of Petunia hybrida at different stages of microsporocyte meiosis (Conley and Hanson 1994). The accumulation of transcripts of ATP-A, mitochondrial subunit-9 of F,-ATP synthase (ATP-9), and mitochondrial 26S ribosomal RNA (rRNA) (RRN-26) genes in the tapetum of sunflower anthers is also indicative of a cell-specific regulation of gene expression (Smart, Moneger and Leaver 1994). The mounting evidence in favor of a role for tapetal cells in sporopollenin and pollenkitt synthesis has led to a renewed interest in their flavonoid metabolism. Tapetal-specific expression of genes encoding two key enzymes of flavonoid pathways in P. hybrida is an interesting case study. When promoter regions of chalcone synthase and chalcone flavanone isomerase genes are fused with GUS reporter gene, expression of the genes in the anthers of transgenic plants is confined mostly to the tapetum (Koes et al 1990; van Tunen et al 1990). Transcripts of a cDNA clone isolated from Brassica napus anthers sharing reasonably good amino acid homology with chalcone synthase are also expressed in the tapetum (Shen and Hsu 1992). The work on the tapetum of the angiosperm anther has now come full circle to provide a conceptual framework for its probable function. The most influential recent approach in regard to the functional aspect of the tapetum is to introduce chimeric ribonuclease (RNase) genes TA-29/RNase-Tl (regulatory fragment of tobacco TA-29 fused to a gene
encoding RNase-Tl from the fungus Aspergillus oryzae) and TA-29/BARNASE (TA-29 fused to RNase gene BARNASE from Bacillus amyloliquefaciens) into tobacco and B. napus plants and show that their expression selectively destroys the tapetum, prevents pollen formation, and leads to male sterility in the transformed plants (Mariani et al 1990; de Block and Debrouwer 1993; Denis et al 1993). Similar results were obtained (Figure 2.13) when promoter fragments from B. napus AS and A-6 genes fused to BARNASE were expressed in transgenic tobacco (Paul et al 1992; Hird et al 1993) or when an antisense construct using the regulatory region of BCP-1 gene from B. campestris was introduced into Arabidopsis (Xu et al 1995a). Premature tapetal degeneration was also observed when Arabidopsis anther-specific gene APG-BARNASE construct was introduced into transgenic tobacco, although tapetal ablation after the late uninucleate stage of microspore development was no longer effective in causing pollen sterility (Roberts, Boyes and Scott 1995). In this respect, the onset of tapetal independence of the microspores offers a clear contrast with results from the other studies. In any event, the tapetum appears to be involved in pollen formation; the challenge now is to provide a complete formal description of how the tapetum accomplishes this function. 3. MICROSPOROGENESIS
Investigation of the structure, physiology, and cytology of the microsporocytes as they are transformed into microspores is considered in this sec-
42
Gametogenesis
Figure 2.13. Transverse sections of wild-type and transgenic tobacco anthers, transformed with A-6-/BARNASE construct, (a) Wild-type anther (3 mm) at the tetrad stage, (b) Transgenic anther (2.5 mm) with tetrads, (c) Wild-type anther (4 mm) at
the microspore stage, (d) Transgenic anther (3.5 mm) showing the complete degeneration of the tapetum. c, callose wall; m, microspore; t, tapetum; td, tetrad. Scale bars = 100 urn. (From Hird et al 1993; photographs supplied by Dr. R. J. Scott.)
tion under microsporogenesis. Research in microsporogenesis in flowering plants has, as its major goals, an understanding of the causes and consequences of the origin of cells with a balanced but reduced number of chromosomes, and a comprehension of the molecular mechanisms involved in meiosis of the microsporocytes. In a sense, by a series of events that flow as a continuum in time, microsporogenesis bridges the gap between the diploid microsporocytes, on the one hand, and the haploid microspores, on the other. Cytological and biochemical discoveries have become an indispensable information package to define landmarks in the study of microsporogenesis in angiosperms.
it. Soon after the microsporocyte is cut off, it is covered by a primary cellulose cell wall. However, shortly before meiosis, the primary wall gives way to another distinctive wall made of callose; in occasional species, the cellulose wall persists around the microsporocyte and callose deposition occurs within the primary wall (Bhandari, Bhargava and Geier 1981). At the same time as callose appears as a sheath around each microsporocyte, massive protoplasmic strands connect the microsporocytes with one another. Attention has been called to the fact that the presence of cytoplasmic channels between microsporocytes makes the contents of an anther locule a single massive coenocyte and thus probably facilitates synchrony of meiosis (J. HeslopHarrison 1966b). At the ultrastructural level, the cytoplasm of the microsporocyte presents the profile of a metabolically active cell enriched with ribosomes, mitochondria, and plastids. Among the distinctive properties of the latter two organelles, one is that they show little internal differentiation.
(i) Ultrastructural Changes
We start our account of microsporogenesis with the microsporocyte. Processes concerned with cell growth and development converge on the microsporocyte and events of meiosis diverge from
Anther Developmental Biology
Dictyosomes and fine strands of rough ER dot the rest of the cytoplasm (van Went and Cresti 1989). The cytoplasmic contour of the microsporocytes of Canna generalis is ascribed to the presence of stacked annulate lamellae, which merge into the rough ER. Pores, which suggest similarity to similar openings in the nuclear membrane, are a regular feature of the architectural design of the lamella (Scheer and Franke 1972). In the microsporocytes of Lycopersicon esculentum, major changes are restricted to the nucleus, nucleolus, and mitochondria. Formation of vacuoles by the invagination of the inner membrane of the nuclear envelope during the early prophase period is the change observed in the nucleus. The nucleolus is transformed from a granular and fibrillar condition during the preprophase stage to a fibrillar state during late prophase; mitochondria are simple, with lightly stained matrix and a few cristae to begin with, but they acquire a densely stained matrix with dilated cristae later (Sheffield et al 1979; Polowick and Sawhney 1992). Transformation of the ER into double-membraned islands enclosing blobs of cytoplasm enmeshed in a stroma of the regular cytoplasm has been found to be an important cytological feature of the microsporocytes of Lilium longiflorum as they enter the meiotic prophase (Dickinson and Andrews 1977). Although these cytoplasmic enclaves are detected readily in the microsporocyte cytoplasm of other plants also, their functions are not clearly defined (Echlin and Godwin 1968a; Biddle 1979). Perhaps they represent a temporary storage facility for the cytoplasm for the duration of meiosis or serve to support and maintain the integrity of the cell during meiosis. More will be said about this later. As is well known, meiosis consists of two successive nuclear divisions, but chromosome replication occurs only once. Most authors recognize five distinct stages of meiosis, namely, leptotene, zygotene, pachytene, diplotene, and diakinesis, and our account of microsporogenesis here will also be keyed to these stages. DNA synthesis is initiated shortly before leptotene and continues into leptotene. Leptotene lapses into zygotene, the visible evidence of which is apposition or pairing of homologous chromosomes (synapsis), leading to the formation of bivalents. What constitutes the basic mechanism that facilitates homology recognition once the homologues are side by side is yet to be determined, but it appears that chromosomes undergo several dramatic morphological changes, such as partial separation of sister chromatids, increase in the surface complexity of the chromatid
43 fibers, and elongation of the heterochromatic knobs that make this possible (Dawe et al 1994). In L. longiflorum and other plants, pairing - and probably the physical basis of it - is the formation of a tripartite structure, known as the synaptonemal complex, between meiotic homologues (Moens 1968). Indicative of the proteinaceous nature of this structure, more than 20 proteins were identified in SDS-PAGE as components of the synaptonemal complex of L. longiflorum (Ohyama et al 1992). Zygotene is followed by pachytene, when a shortening and thickening of the bivalents takes place. In each bivalent, the homologous chromatids break apart and exchange segments, thus re-forming complete chromatids but with a reassortment of genetic units. The region of contact between homologous chromatids where crossing-over occurs is known as the chiasma. The events that highlight pachytene are followed by diplotene, when the bivalents begin to contract. A major feature of diplotene is the separation of the two paired chromosomes by splitting along the synaptonemal complex. However, maximum contraction of the bivalents is attained at diakinesis, which signals the end of prophase. The conventional view of chromosome pairing at zygotene and crossing-over at pachytene has not been universally accepted. Although data on the timing of DNA synthesis and chromosome replication in the premeiotic interphase period challenged the view of pachytene crossing-over, later cytological and biochemical studies seem to reinforce the thesis that synapsis takes place during zygotene, followed by crossing-over at pachytene (Walters 1970). Considerable uncertainty also prevails about the precise time at which crossing-over occurs in relation to the assembly of the synaptonemal complex. It has been suggested that some undefined cytological and biochemical changes occur in the microsporocytes prior to synapsis, entailing an arbitrary selection of potential sites for crossingover (Stern, Westergaard and von Wettstein 1975). Prophase is followed by the other regular stages of mitosis (meiosis I, heterotypic division), resulting in the formation of two daughter nuclei with half the number of chromosomes of the mother cell. These nuclei promptly complete a second mitotic division (meiosis II, homotypic division), generating four cells with the haploid number of chromosomes. If we view meiosis in its totality, the dismantling of the paired chromosomes and their reconstruction stand in sharp contrast to other events. It is necessary to point out here that this bare-bones account of meiosis is intended to guide
44 the reader through the biochemical and molecular events that are considered later and does not do justice to the many processes that together make up this division. With keen foresight, some investigators working at the light microscopic level suggested that meiosis in the microsporocytes is accompanied by some striking changes in cytoplasm and organelle complement. The consequences of these changes, later confirmed in the electron microscope, are most easily seen in the ribosome population, nucleoli, plastids, and mitochondria. In L. longiflorum, for which extensive data are available, by diakinesis the microsporocytes are almost completely denuded of ribosomes; this is followed by a repopulation of the cytoplasm with ribosomes beginning about the metaphase-anaphase stages of meiosis I (see Dickinson and Heslop-Harrison 1977, Porter et al 1984, for reviews). When changes in the RNA content of microsporocytes of L. longiflorum, Trillium erectum (Liliaceae) (Mackenzie, Heslop-Harrison and Dickinson 1967), and Cosmos bipinnatus (Asteraceae) (Knox, Dickinson and Heslop-Harrison 1970) are determined, substantial reductions during the zygotene-pachytene interval and an increase toward the end of the division are observed. This supports connections between the elimination of ribosomes and a decline in the RNA content of the microsporocytes, on the one hand, and the increase in RNA content and the flooding of cells with ribosomes, on the other hand. In C. bipinnatus (Knox, Dickinson and Heslop-Harrison 1970) and Lilium sp. (Bird, Porter and Dickinson 1983), the low RNA content of microsporocytes also follows a period of lytic enzyme activity, mainly of acid phosphatase. According to Dickinson and Heslop-Harrison (1970b), replenishment of the ribosome population in this RNA cycle is related to the formation of nucleolus-like bodies from nucleolus-organizing regions of the chromosomes; these supernumerary nucleoli, termed nucleoloids, are later set free in the cytoplasm of the microsporocyte at anaphase of meiosis I. Additional evidence from cytochemistry, autoradiography of 3H-uridine incorporation, and in situ hybridization using cloned rRNA probes has shown that the increase in RNA content is due to a significant synthesis of new rRNA in the accessory nucleoli and in the nucleolus-organizing regions, as well as in a small number of nucleolus-like inclusions that appear associated with the chromosomes (Williams, Heslop-Harrison and Dickinson 1973; Dickinson and Willson 1985; Sato, Willson and Dickinson 1989; S. Sato et al 1991; Majewska-Sawka and Rodrfguez-Garcfa 1996). Apparently, the riboso-
Gametogenesis mal DNA (rDNA) formed by chromosomal amplification is packaged in the extruded nucleoli. Although not widespread, chromosome-associated nucleolus-like bodies and cytoplasmic nucleoloids have been noted during microsporocyte meiosis in Rhoeo spathacea (Williams and Heslop-Harrison 1975) and Oka europaea, respectively (RodriguezGarcia and Fernandez 1987; Alche, Fernandez and Rodrfguez-Garcfa 1994). Changes seen in the structure of plastids and mitochondria during microsporocyte meiosis in L. longiflorum are the very epitome of new developmental potencies in the organelles. To begin with, normal plastids present in the microsporocyte undergo a cycle of dedifferentiation and redifferentiation, characterized by the loss of starch, changes in shape, and erosion of internal structure. These events, which are completed by late zygotene, relegate plastids to the status of hollow double-membrane organelles consisting solely of osmiophilic droplets, small vesicles, and membranous tubes. The redifferentiation of the plastids is initiated during metaphase of meiosis I with the transient appearance of granule-double membrane associations in the stroma. Following the release of microspores from the tetrad, the membrane-particle associations disappear and the plastids reconstitute their regular structure. Based on enzymatic digestion and cytochemical tests, there is some evidence to show that RNA-, protein-, and carbohydrate-containing structures are involved in the dedifferentiation of plastids (Dickinson and Heslop-Harrison 1970a; Dickinson 1981; Dickinson and Willson 1983). Since these plastids develop into amyloplasts in the microspore, the possibility that the dedifferentiated plastid is the progenitor of the amyloplast deserves serious consideration (Dickinson and Willson 1983). In lily microsporocytes, the mitochondria also show a cycle of changes similar to that of the plastids, such as structural simplification by the zygotene stage and recovery of the normal structure in the microspore. In certain plants, there is a coincident accumulation of transcripts of mitochondrial genes and their proteins during the early stages of microsporocyte meiosis (Conley and Hanson 1994; Smart, Moneger and Leaver 1994). It was mentioned earlier that a considerable portion of the microsporocyte cytoplasm is engulfed by unit membranes. Interestingly enough, the ribosome population in the enclosed cytoplasm undergoes much less degradation than the ribosomes in the rest of the cytoplasm. Coincident with the redifferentiation of plastids and mitochondria in the
Anther Developmental Biology
45
microspores, the investing membranes tear apart, (Orchidaceae) (Brown and Lemmon 1991a), allowing diffusion of the contents into the Magnolia denudata, M. tripetala (Magnoliaceae) microsporocyte cytoplasm, while segments of the (Brown and Lemmon 1992a), and Cypripedium calimembranes themselves move to the vicinity of the fornicum (Orchidaceae) (Brown and Lemmon 1996) plasma membrane. It is believed that these mem- have not revealed a configuration of microtubules brane fragments play a major role in the formation in the form of a preprophase band. In G. verrucosa, of the pollen wall layer (Dickinson and Andrews microtubules are randomly oriented during meio1977). Spherosomes also show some changes corre- sis, suggesting that their main function is to ensure lated with meiosis; these organelles display a cycle the integrity of the cytoplasm during division. As of aggregation and dispersion without undergoing shown in Figure 2.14, following meiosis II, arrays any internal gyrations (J. Heslop-Harrison and of microtubules radiate from the nucleus and thus Dickinson 1967). On the whole, these observations apparently define the cytoplasmic domains of the leave little doubt that transition of the microsporo- future microspores (van Lammeren et al 1985). A cyte to the gametophytic phase involves a radical similar role has been ascribed to a postmeiotic sysreorganization of the cytoplasm, resulting in the tem of microtubules in establishing the division elimination of much of the sporophytic program plane for quadripartitioning of microsporocytes in and in the installation of a program for gameto- most of the other species just listed. Thus, the disphytic functions. Although the ultrastructural position of the four haploid nuclei with their own changes during microsporogenesis have been fol- microtubule cytoskeleton seems to determine the lowed in Tmdescantia paludosa (Bal and De 1961;future cleavage pattern of the tetrad. The radiating Maruyama 1968), Hyacinthoides non-scripta microtubules are found to exclude plastids from (Liliaceae) (Luck and Jordan 1980), Nymphaea alba around the microsporocyte nuclei of L. longiflorum; (Nymphaeaceae) (Rodkiewicz, Duda and Bednara a wildly speculative view is that this paves the way 1989), Triticum aestivum (Mizelle et al 1989), and for the subsequent formation of plastid-free generGossypium hirsutum (cotton; Malvaceae) (Hu, Wang ative cells from microspores (Tanaka 1991). and Yuan 1993), the data are neither complete nor as Microfilaments are best detected by immunoflucomprehensive as those provided for L. longiflorum. orescence using antibodies for actin (the main The basic features of the cytokinetic apparatus chemical constituent of the microfilament) or by and the factors that control the division planes of treating unfixed, permeabilized cells with fluoresthe nuclei during microspore meiosis have hardly cent-dye-labeled phallotoxins (rhodamine-phalbeen investigated. As is well known, the cytoskeleloidin staining). According to Sheldon and Hawes tal network of plant cells is constituted of micro(1988), actin cytoskeleton appears to be indepentubules and microfilaments. Microtubules are dent of the microtubule system both within the easily identified in the electron microscope; they mitotic spindle and in the cytoplasm during are also detected by immunocytochemical methods microspore meiosis in Lilium hybrida; these authors using an antibody for tubulin. During cell division, suggest that here actin may function with the microtubules are thought to play an important role microtubules in maintaining cytoplasmic integrity. in the formation of the preprophase band that preHowever, actin filaments accompany microtubules dicts the division plane and helps to guide the cytoin phragmoplast formation during microspore kinetic apparatus along a predetermined path. meiosis in L. hybrida (Sheldon and Hawes 1988), G. However, electron microscopic and /or immunofluverrucosa (van Lammeren, Bednara and Willemse orescence analyses of microtubular patterns during 1989), Z. mays (Staiger and Cande 1991), and microspore meiosis in Tradescantia paludosa Phlaenopsis sp. (Brown and Lemmon 1991a, b). In S. (Clapham and Ostergren 1984), Gasteria verrucosa melongena, a dense web of randomly oriented actin (van Lammeren et al 1985), Lilium henryi (Sheldon filaments is thought to mark the future division and Dickinson 1986), L. longiflorum (Tanaka 1991), plane after meiosis I, very much like the preLycopersicon esculentum, Ornithogalum virens prophase band of microtubules, but such a config(Liliaceae) (Hogan 1987), lmpatiens sultani uration of actin is not detected during meiosis II (Balsaminaceae), Lonicera japonica (Caprifoliaceae) (Traas, Burgain and de Vaulx 1989). Bands of actin (Brown and Lemmon 1988; van Went and Cresti filaments are also seen at the level of cell division 1988b), Solanum melongena (eggplant) (Traas, planes constituting cleavage furrows following Burgain and de Vaulx 1989), Doritis pulcherrima both meiosis I and II in the microsporocytes of (Orchidaceae) (Brown and Lemmon 1989), Zea Magnolia soulangeana (Dinis and Mesquita 1993). In mays (Staiger and Cande 1990), Phalaenopsis sp. the most detailed studies on G. verrucosa, microfila-
a1
c1
d1
f Figure 2.14. Microsporogenesis in Casteria verrucosa. (A-F) Stages in microsporogenesis as visualized by Nomarski optics, (a-f) Fluorescently labeled microtubules in the corresponding stages, (a'-f) Diagrammatic representations of microtubule configurations in the corresponding stages. (A, a, a1) Microsporocytes at leptotene with microtubules arranged in a crisscross pattern. (B, b, b') Microsporocytes at pachytene; microtubules are arranged in a crisscross pattern and are also concentrated near the nucleus. (C, c, c1) Microsporocytes at diakinesis with
46
microtubules at the polar center (arrow), and the bundles at the equatorial poles at metaphase. (D, d, d') Dyads, showing microtubules radiating from the nuclei and interdigitating between the nuclei. (E, e, e') Tetrads of nuclei, with microtubules radiating from the nuclei. (F, f, f) Tetrads after wall formation, with thick bundles of microtubules (arrow). Ca, callose wall; Ch, chromosome; N, nucleus. Scale bars = 10 urn. (From van Lammeren et al 1985. Planta 165:1-11. © Springer-Verlag, Berlin. Photographs and line drawing supplied by Dr. A. A. M. van Lammeren.)
Anther Developmental Biology
ments and microtubules appear to interact during the different stages of microsporocyte meiosis (van Lammeren, Bednara and Willemse 1989). Since no site-specific actin was localized during meiosis, it is difficult to assign a function for actin in the division process. Perhaps there is sufficient reason to consider a spatial control of meiosis in terms of the dynamics of the entire cytoskeleton, rather than of the microtubules or microfilaments alone. Moreover, it seems that key processes of meiosis involving microtubules and microfilaments are controlled by genes. Recent studies on certain meiotic mutants of maize have shown that they exhibit distinct alterations in cytoskeletal behavior, especially of the microtubules, associated with aberrations in the meiotic pathway (Staiger and Cande 1990, 1991; Liu, Golubovskaya and Cande 1993). Identification of the genes regulating entry into, and exit from, the meiotic processes and of the proteins responsible for the phenotype will provide further insight into cytoskeletal functions during microsporocyte meiosis.
47
this is a series of biochemical and molecular activities concerned with pairing, crossing-over, and a reduction in chromosome number in a small population of cells over a short period of time. Two aspects of the molecular biology of meiosis that will be addressed here are those concerned with the transition of microsporocytes from a state of mitosis to meiosis and the completion of meiosis itself. A question of primary importance concerns the nature of the trigger that prompts sporogenous cells to cease mitotic divisions and enter the meiotic cycle. It has been suggested that the synthesis of some factors in tissues other than the sporogenous cells is involved in triggering the cells to enter meiosis. These substances have been given such trivial names as "meiosis-inducing substance" or "meiosis determinants," but their chemical nature remains obscure (Walters 1985). It would seem that some activities connected with meiosis, such as the synthesis of the spindle material, are initiated during the premeiotic period and might even be a prerequisite to division, but we do not yet have a catalogue of these events. Among other biochemical events of the immediate premeiotic period is the apparent accumulation of nucleic acid precursors in the microsporocytes. According to Foster and Stern (1959), the premeiotic interval in the microsporocytes of Lilium longiflorum is associated with a marked accumulation of soluble deoxyribosidic compounds. Other studies have shown the presence of enzymes associated with the production of deoxynucleosides and their phosphorylation during the interval adjacent to DNA synthesis, with the caveat that some enzymes that mediate in the nucleoside transformations are not directly tied to DNA synthesis (Hotta and Stern 1961a, b). The timing of appearance of the nucleosides may be variable, as in L. henryi, in which nucleosides, nucleotides, and free bases are detected in the midmeiotic prophase; this observation might influence the judgment as to whether they really accumulate before meiosis (Linskens 1958).
Considerable variation exists in the formation of the phragmoplast and cell plate during microsporocyte meiosis. When meiotic cytokinesis occurs successively after each nuclear division, cell plates are laid down typically in association with centrifugally expanding phragmoplasts. An usual feature of cell plate formation in the microsporocytes of Lilium longiflorum is the presence of numerous Golgi-derived, coated vesicles with fuzzy hairlike spikes that contribute to the formation of the phragmoplast (Nakamura and Miki-Hirosige 1982). Absence of phragmoplast after meiosis I is a general feature of simultaneous cytokinesis, in which the four microspores are cleaved from a common cytoplasm. A complex of phragmoplasts is, however, generated after meiosis II, and wall formation occurs by centripetally growing furrows, by cell plate formation, or by a combination of both processes with the help of microtubules that determine spore domains. With simultaneous cytokinesis in Doritis pulcherrima (Brown and Lemmon An important group of compounds that are 1989), Phalaenopsis sp. (Brown and Lemmon 1991b), thought to have a structural involvement in cell Magnolia denudata, and M. tripetala (Brown and division is the -SH group, whose presence is Lemmon 1992a), typical phragmoplasts and associ- related to the formation and functioning of the ated cell plates are formed after meiosis I, but they mitotic spindle. Although the role of -SH groups in are short-lived and are routinely resorbed before the bonding of macromolecules of the mitotic meiosis II. apparatus is not well understood, fluctuations have been noted in the relative levels of free and bound -SH groups associated with microsporocyte meio(ii) Molecular Biology of Meiosis sis in L. longiflorum and Trillium erectum. In both The sense and purpose of microsporogenesis attain their climax at meiosis. Tangible evidence of
species, particularly striking is a rise in the concentration of soluble sulfhydryls preceding meiosis
48 and a decline before the fabrication of the mitotic apparatus (Stern 1958). This observation does not vitiate the suggestion that sulfhydryl compounds are used prior to prophase; however, it appears that we have a long way to go before the preparative molecular changes as they relate to discrete events in the premeiotic microsporocytes involving the utilization of -SH groups are understood. A promotive effect of glutathione on meiosis in cultured anthers of L. henryi has been linked to a requirement for -SH compounds in cell wall formation during meiosis (Pereira and Linskens 1963).
Gametogenesis
that there is no further synthesis until after the completion of meiosis. The premeiotic DNA synthetic period was referred to as the preleptotene period (Plaut 1953; Taylor and McMaster 1954). Although label incorporation in these investigations is found almost exclusively in the nucleus, Holm (1977) has described 3H-thymidine label in the cytoplasm of premeiotic microsporocytes of L. longiflorum; more than half of the incorporation was resistant to digestion by deoxyribonuclease (DNase). Two early reports indicated that a reduced level of DNA synthesis occurred during prophase after completion of the main S-period. This might appear necessary for chromosome pairing at zygotene and crossing-over at pachytene. According to Sparrow, Moses and Steele (1952), in the microsporocytes of Trillium erectum a small increase in DNA content occurs between pachytene and diplotene, but it is insufficient to account for an expected doubling if the entire premeiotic DNA is synthesized during this interval. In anthers of L. henryi, there is a burst of DNA increase in the microsporocytes during prophase after the Speriod (Linskens and Schrauwen 1968b). The reality of a midprophase DNA synthesis during microsporocyte meiosis in the liliaceous plants Lilium and Trillium has now been established by the detailed investigations of Hotta and Stern (see Stern and Hotta 1977, 1984, for reviews). The microsporocytes of several hybrids of L. longiflo-
The Timing of DNA Synthesis. To understand the concepts of pairing of homologous chromosomes and recombination during microsporocyte meiosis, we must first examine whether there are any unique patterns of DNA synthesis associated with the meiotic cell cycle. The problem here is infinitely more complex than DNA synthesis during mitosis, since it is conceivable that the freedom of the homologous chromosomes for recombination depends upon a round of DNA synthesis for the repair of breaks formed during crossing-over. During the past several decades, the study of DNA synthesis during microsporocyte meiosis in angiosperms has proceeded from microspectrophotometry through autoradiography to biochemical and molecular approaches. A pioneering investigation of Swift (1950), which showed an increase in the DNA content of microsporocytes of Tmdescantia paludosa during the early period of rum, L. henryi, L. speciosum, L. tigrinum, and T. erecmeiosis, suggested that the conventional synthetic tum have proved to be favorable for this work period (S-period) is at the onset of prophase. In a because of their abundance, natural synchrony, and later work, Taylor and McMaster (1954) substanti- ability to undergo part of the meiotic division ated this finding from their work with anthers of L under in vitro conditions. Moreover, compared to longiflorum. Particularly striking was their observa- other plants, meiotic division in the microsporotion that as the microsporocytes underwent syn- cytes of liliaceous plants occurs at a leisurely pace chronous DNA replication in anticipation of and is spread over a period of 9 to 12 days (Ito and meiosis, they settled at a DNA content correspond- Stern 1967). In much of the work on liliaceous ing to the AC amount. Subsequently, they attained a plants, meiotic cells at different stages were labeled transient phase with 2C DNA and ended up as by exposing flower buds to solutions containing microspores containing the haploid amount of the isotope or by suspending extruded microsporo32 3 DNA. Following this study, the pattern of autora- cytes directly in the isotope; both PO4 and Hdiographically or microspectrophotometrically thymidine were used to label the cells. DNA measurable DNA synthesis during microsporocyte isolated by a standard method was next subjected meiosis has been followed in other species of Lilium to centrifugation to equilibrium in cesium chloride, (Plaut 1953; Taylor 1953, 1959), T. paludosa (Taylor and the positions of the labeled peaks in the gradi1953; Moses and Taylor 1955; De 1961), Tulbaghia ent were determined by analyzing fractions (Figure violacea (Taylor 1958), and Secale cereale (rye; 2.15). The most interesting outcome of these studies Poaceae) (Bhaskaran and Swaminathan 1959). is the identification and characterization of two These studies have generally established that DNA groups of DNA sequences that replicate during the synthesis in the microsporocytes occurs at late pre- prophase interval. This was accomplished by meiotic interphase or at the onset of prophase and DNA-DNA hybridization involving DNA synthe-
Anther Developmental Biology
20
30
40
50
49
30
40
50
TUBE NUMBER Figure 2.15. DNA labeling during microsporocyte meiosis in lily anthers. Microsporocytes cultured for 24 hours were exposed to 32PO4 for 20 hours. DNA extracted was centrifuged in cesium chloride, fractions (indicated by tube number) were collected, and radioactivity (broken lines) and absorbance (solid lines) were determined. The stages during which microsporocytes were exposed to the isotope are (a) zygotene, (b) late zygotene, early pachytene, (c) late pachytene, and (d) late pachytene, early diplotene. (From Hotta, Ito and Stern 1966.)
sized exclusively during the premeiotic S-phase and zygotene and pachytene intervals and DNA prepared from somatic cells. The synthesis of the first group of DNA sequences coincides with the beginning of chromosome pairing at zygotene and continues to the end of pachytene. This sequence is designated as zygotene DNA (zyg-DNA) and constitutes about 0.1% to 0.2% of the genome whose replication is delayed until halfway through the meiotic prophase (Hotta, Ito and Stern 1966; Hotta and Stern 1971b; Hotta et al 1985b). Despite its presence in limited amounts, zyg-DNA is remarkable for the complex it forms with proteins and phospholipids (Hecht and Stern 1971). Another interesting aspect of this DNA synthesis is that the newly replicated strands are not immediately ligated to the body of the nuclear DNA, thus causing gaps between zyg-DNA and adjacent nuclear DNA. This was demonstrated in sedimentation profiles of 14C-labeled S-phase DNA and 3H-labeled zyg-DNA in sucrose or glycerol gradients, where the zyg-DNA sediments toward the top of the gradient while the S-phase DNA remains at the bottom (Hotta and Stern 1976). The characteristic buoyant
density of zyg-DNA is probably due to its protein-phospholipid association. In an attempt to account for the selective suppression of zyg-DNA synthesis until zygotene and the delayed completion of its replication, Hotta, Tabata and Stern (1984) have invoked a specific protein named leptotene protein (L-protein). This protein has been purified from the nuclear membranes of meiotic cells of L. speciosum and has a molecular mass of 73 kDa. It has been found to be highly specific in inhibiting zyg-DNA synthesis in isolated prezygotene nuclei, whereas nuclei from premeiotic anthers, from postzygotene microsporocytes, and from microsporocytes before or during S-phase are insensitive to the protein inhibitor. The rationale for a protein inhibitor in suppressing zygDNA synthesis during the premeiotic S-phase and through the prezygotene interval is the need for zyg-DNA synthesis to occur coincident with the pairing of homologous chromosomes. Synthesis of zyg-DNA has proved to be a useful molecular marker for the onset of meiosis in liliaceous plants. Hotta et al (1985b) utilized poly(A)RNA prepared from microsporocytes of L. speciosum to test for its ability to hybridize with fractions of lily DNA centrifuged to equilibrium in a cesium chloride gradient. A major peak of hybridization coinciding with zyg-DNA has pointed to the poly(A)-RNAs, being a zyg-DNA transcript (zyg-RNA). The accumulation of this mRNA occurs only after leptotene and does not extend beyond midpachytene (Figure 2.16). Since zyg-RNA was not detected in nonmeiotic tissues such as the anther wall, young anthers, and callus, it is believed that its function is essentially meiotic. A speculative view is that zyg-RNA is translated into proteins that are involved in the formation of the synaptonemal complex. It has been argued that a critical role must be envisaged for zyg-DNA in meiosis, perhaps in the actual process of chromosome pairing itself. The effects of inhibitors of DNA synthesis (for example, deoxyadenosine) on meiosis in cultured microsporocytes of L. longiflorum have added substance and support to this argument (Ito, Hotta and Stern 1967; Sakaguchi et al 1980). If the inhibitors are applied at late leptotene or very early zygotene stages, microsporocytes are totally arrested in their development and end up with fragmented chromosomes. Anarchic chromosome shattering is reduced when microsporocytes are exposed to the inhibitors at the midzygotene stage; these cells also go through an abortive meiosis I. Effects of inhibition of DNA synthesis during late zygotene or early
50
Gametogenesis
so
10
40
f 20
10
8
//
SOM II
ON
T
Figure 2.16. Changes in the level of zygotene DNA transcript (zyg-RNA) during microsporocyte meiosis in lily anthers. The amount of zyg-RNA was determined from the amount of RNA adsorbed from poly(A)-RNA in the zygDNA column. Closed circles are values expressed relative to DNA; open circles are values expressed as percentage of total poly(A)-RNA. Values for somatic tissues are also shown. SOM, somatic tissues; I, interphase; L, leptotene; Z, zygotene; P, pachytene; DIV, end of meiosis; T, tetrads. (From Hotta et al 1985b. © Cell Press.)
pachytene stages are generally reflected in chromosome fragmentation at meiosis II. The best pointer to the effect of inhibitors of DNA synthesis is the electron microscopic demonstration that deoxyadenosine apparently interferes with the initiation and formation of the synaptonemal complex with subsidiary effects on the disjunction of the paired homologues at diplotene stage and the separation of sister chromatids at anaphase (Roth and Ito 1967). Mitomycin-C is another inhibitor of DNA synthesis; it interferes with the formation of the synaptonemal complex if it is applied to microspores at the leptotene-zygotene stage (Sen 1969). It is difficult to judge from the electron micrographs alone whether inhibitors prevent the organization of the axial core or the interdigitation of the transverse filaments of the synaptonemal complex. In an attempt to define the relationship of zyg-DNA synthesis to chromosome pairing, it has been proposed that synthesis of DNA during the zygotene-pachytene stage requires the simultaneous synthesis of certain nuclear proteins. Selective application of cycloheximide, an inhibitor of protein synthesis, not only arrests protein synthesis during the zygotene interval, but also results in the inhibition of DNA synthesis and in the failure of
chromosome pairing and fabrication of the chiasma (Hotta, Parchman and Stern 1968; Parchman and Stern 1969). This has fueled some speculation that the protein is the component of an axial element involved in chromosome pairing and chiasma formation and that its function is related to simultaneous DNA replication. Various types of evidence have been adduced to support the view that a second group of DNA sequences that replicate during the preprophase interval of microsporocyte meiosis afford an efficient mechanism for genetic recombination, especially in the limited repair activities at the break points of DNA molecules involved in crossing-over at pachytene. Synthesis of this DNA is of a much smaller magnitude than that occurring at zygotene and, since pachytene replication does not cause a net increase in DNA, it has been concluded that the regions of the DNA strand undergoing replication are simply replacing preexisting ones (Hotta and Stern 1971b). The first major indication that breakage and restoration of continuity by repair are largely confined to pachytene came from the demonstration that activities of a microsporocytespecific endonuclease and of polynucleotide kinase and polynucleotide ligase, which oversee chromosomal DNA breakage and reunion events, peak at the end of the zygotene phase and decline sharply during late pachytene (Howell and Stern 1971). Evidence for a programmed breakage of chromosomal DNA during pachytene has been provided by zonal sedimentation profile of denatured, ss-DNA prepared from microsporocytes of Lilium sp. at different stages of the meiotic cycle. The striking result of this experiment reflecting the introduction of nicks at pachytene is that the denatured DNA displayed a bimodal profile during pachytene, with two principal peaks corresponding to sedimentation values of 104S and 62S, and a more or less unimodal profile at all other stages of meiosis. No change was seen in the sedimentation profile of native DNA, which showed a major peak in the 250S region of the gradient (Figure 2.17). In addition, it was found that sedimentation profiles of DNA prepared from microsporocytes of hybrids of L. speciosum in which chiasma formation was suppressed or very much reduced by colchicine treatment were distinguished by the absence of a 62S peak in the chiasmatic forms (Hotta and Shepard 1973; Hotta and Stern 1974). In other studies, the seemingly elaborate organization of pachytene DNA (P-DNA) has been unraveled by reannealing kinetics and thermal stability analysis. It has been deduced from these experiments that families of
Anther Developmental Biology
300S
ZOOS
I
1
51
MS-Z IOOS 78S
\
\
200 •
FRACTION NUMBER Figure 2.17. Sedimentation profiles of native and alkalidenatured DNA extracted from microsporocytes of lily at meiotic stages, (a) Sedimentation of native DNA. (b) Sedimentation of denatured DNA, both in a 10%-30% glycerol gradient. O, premeiotic interphase to early leptotene; A, zygotene; • , early to midpachytene; X, diplotene to anaphase I. Zygotene DNA was not measured in (b). Sedimentation values are given at the top of each figure. MS-2 and T7 are marker DNAs. (From Hotta and Stern 1974. Chromosoma 46:279-296. © Springer-Verlag, Berlin.)
moderately repeated sequences, of modal length of about 1,500-2,000 bp, scattered throughout the chromosomes may be the sites of specific nicks and repair synthesis. Moreover, there is a strong preference for nicking and rejoining functions to a particular subset of repeats as opposed to a weak preference for repetitive sequences in general (Smyth and Stern 1973; Bouchard and Stern 1980). Perhaps the most remarkable complexity associated with the functioning of P-DNA results from factors that render these DNA sequences susceptible to endonuclease nicking. Hotta and Stern (1981) found that when pachytene cells of L. speciosum are cultured in the presence of 3 H-uridine,
P-DNA appears to be labeled. By RNase digestion, the label was identified in RNA bound to P-DNA. These RNA molecules, referred to as pachytene-small-nuclear RNA (psn-RNA), revealed a single size class, with a sedimentation value in the 5S to 6S range, and were found to be synthesized during meiotic prophase coincident with chromosome pairing. Since P-DNA sequences rendered accessible to endonuclease nicking are partially complementary to psn-RNA, the specificity of chromatin sites selected by psn-RNA is attributed to sequence complementarity. By modifying chromatin organization in the regions housing P-DNA sequences, it has been claimed that psn-RNA is a major factor in rendering P-DNA sequences open to endonuclease nicking. Based on the work just reviewed, a tentative model of the organization of pachytene nick-repair sites in the genome of L. speciosum has assigned each P-DNA three different regions. According to this model, each end of the P-DNA consists of pachytene-small-nuclear DNA sequences (psnDNA) ranging in length from about 150 to 300 bp and a middle region that does not share homology with psn-DNA (Hotta and Stern 1984). A discordant note in this otherwise convincing account of P-DNA is the concern raised by the work of Smyth and Shaw (1979) showing that a lot more DNA synthesis, monitored by 3H-thymidine incorporation, occurs in the cytoplasm than in the pachytene nuclei of microsporocytes of L. henryi. A later work involving electron microscope autoradiography showed that about 75% of pachytene synthesis might represent plastid rather than nuclear DNA synthesis, the latter accounting for only about 10% (Smyth 1982; see also Bird, Porter and Dickinson 1983). In addition to nuclear DNA synthesis, varying degrees of cytoplasmic DNA synthesis have also been described during microsporocyte meiosis in Agapcmthus umbellatus (Liliaceae) (Lima-de-Faria 1965). In conformity with biochemical studies, DNA synthesis during early meiotic prophase has been demonstrated by autoradiographic methods, although an unambiguous separation of the synthesis of zyg-DNA and P-DNA was not possible. In the microsporocytes of cultured anthers of Triticum aestivum, where meiosis is completed in about 24 hours, incorporation of 3H-thymidine is found to occur at all stages of meiosis, although a possible effect of wounding on the observed response is not eliminated (Riley and Bennett 1971). Ito and Hotta (1973) provided evidence to show that in the microsporocytes of L. longiflorum, zyg-DNA is synthesized by all the chromosomes without any obvi-
52
Gametogenesis
8000 ous localization, although in long-term 3H-thymidine labeling experiments, the sites of DNA synthesis are obscured by nonspecific labeling by products 6000 of thymidine catabolism. The reality of DNA synthesis during zygotene and pachytene episodes in S the microsporocytes of L. longiflorum has been con•S! 4000 . firmed by high-speed scintillation autoradiography of DNA fibers (Sen, Kundu and Gaddipati 1977). In do the microsporocytes of this same species, electron 2000 microscope autoradiography has revealed preferential association of label with the synaptonemal complex (Kurata and Ito 1978) and with the condensed P DD MAT DY LZ or condensing chromatin (Porter, Bird and Meiotic stage Dickinson 1982), supporting the speculation that DNA synthesis is involved in some aspect of chro- Figure 2.18. Changes in the specific activity of poly(A)RNA isolated from microsporocytes of lily at different mosome pairing.
stages of meiosis. The three lines indicate three separate experiments. L, leptotene; LZ, leptotene-zygotene; Z,
RNA Synthesis and Metabolism. The character- zygotene; P, pachytene; DD, diplotene-diakinesis; MAT, istics of RNA synthesis and metabolism during metaphase, anaphase, telophase; DY, dyad; T, tetrad. (From microsporocyte meiosis correlate in most general Porter, Parry and Dickinson 1983.7. Cell Sci. 62:177-186. respects with what has already been described at © Company of Biologists Ltd.) the ultrastructural level in Lilium longiflorum (Mackenzie, Heslop-Harrison and Dickinson 1967). played by specific RNA species during meiosis is At the heart of these changes is the simultaneous not apparent from these studies, although the posdecrease in the RNA content and in the ribosome sibility of a selective inhibition of rRNA and mRNA population of microsporocytes during meiosis; this synthesis has been implied. For example, Das (1965) result was foreshadowed by previous works from showed that in the microsporocytes of Zea mays, the other laboratories. Using 32PO4 as an indicator of nucleolus rapidly loses its ability to incorporate preautoradiographically detectable RNA synthesis, cursors of RNA synthesis beginning with leptotene Taylor (1958) identified a period of active synthesis and extending into diakinesis. In the microsporoduring premeiotic interphase and one of declining cytes of L. longiflorum, there is not only a decrease in synthesis during zygotene pairing in the the rate of rRNA transcription during meiosis, but microsporocytes of Tulbaghia violacea. Somewhatthe processing of nucleolar RNA into mature ribosimilar results were obtained in L. longiflorum using somes is delayed as well (Parchman and Lin 1972). 14 C-orotic acid as a precursor of RNA synthesis; the Indicative of the inhibition of mRNA synthesis, dracomparatively high resolution of the autoradi- matic decreases have been found in the levels of ographs afforded by this isotope compared to 32PO4 poly(A)-RNA during prophase of meiosis in the also made it possible to follow cytoplasmic and microsporocytes of Lilium sp. (Figure 2.18) and nuclear incorporation of the label. Generally, no Nicotiana tabacum, with a low level being registered cytoplasmic labeling of microsporocytes is at the pachytene stage (Porter, Parry and Dickinson observed during the period of decreasing RNA syn- 1983; Chandra Sekhar and Williams 1992). thesis beginning at leptotene (Taylor 1959). Rhoeo Cytochemical analyses of the changes in RNA discolor is another monocot in which a decreasing content of microsporocytes of Ribes rubrum pattern of RNA synthesis is noted in the (Saxifragaceae) (Geneves 1966), Z. mays (Moss and microsporocytes beginning with pachytene and Heslop-Harrison 1967), P. tenuifolia (Sauter and extending into metaphase of meiosis I (Albertini Marquardt 1967a, b), and L. longiflorum (Linskens 1965,1971). The strategy utilized for RNA synthesis and Schrauwen 1968a) at different stages of meiosis during meiosis by microsporocytes of the dicot are consistent with the autoradiographic data. Paeonia tenuifolia is strikingly similar to that of the These studies have also shown that later stages of monocots (Sauter and Marquardt 1967b; Sauter meiosis of the microsporocytes are invariably 1969a). Evidently, even in taxonomically distinct accompanied by a definitive increase in RNA conclasses of plants, there is a regulated pattern of RNA tents. According to Linskens and Schrauwen synthesis during meiotic intervals when chromo(1968b), there is a considerable increase in the somes undergo extensive alterations. The role polysome content of the microsporocytes of L.
Anther Developmental Biology
longiflorum during metaphase, indicating the engagement of mRNAs. In contrast to the autoradiographic and cytochemical data, biochemical analyses of RNA synthesis during microsporocyte meiosis have been bedeviled by a set of conflicting data. A decrease in RNA synthesis monitored by 14C-uracil incorporation is noted between leptotene and zygotene stages in the microsporocytes of Tulipa gesneriana (tulip), but this is not associated with corresponding changes in RNA polymerase activity (Hotta and Stern 1965b). However, a precipitous decrease in RNA synthesis coincident with the zygotene stage is not always the case. For example, Hotta and Stern (1963a) determined the synthesis of RNA by continuous labeling of microsporocytes of intact anthers of Trillium erectum with 14C-cytosine and found two short peaks at leptotene-zygotene and a substantial peak at the tetrad stage. In this experiment, in contrast to a short pulse of radioactivity, the net incorporation reflects the difference between synthesis and turnover. Similar results were also obtained using isolated microsporocytes and a short pulse with 14C-uracil. Based on the data relating to inhibition of RNA synthesis by actinomycin-D and nucleotide analysis of pulse-labeled RNA, it has been deduced that part of the RNA synthesized during meiotic prophase in T. erectum has messenger properties (Hotta and Stern 1963b; Kemp 1964).
53
and Heslop-Harrison 1967). Based on SDS-PAGE of proteins of anthers of L. longiflorum, there is some indication that specific polypeptides may be involved in the events of meiosis (Wang et al 1992a). A uniform pattern is also encountered in considering the synthesis of histones and nonhistone nuclear proteins during microsporocyte meiosis. By using 3H-arginine as a marker for histone synthesis and 3H-tryptophan for nonhistone proteins, combined with 3H-thymidine incorporation, De (1961) showed that microsporocytes of Tradescantia paludosa synthesize histones at the early leptotene stage synchronously with DNA synthesis. Considerable nonhistone nuclear protein synthesis occurs transiently at late leptotene when no further DNA is synthesized. Similar results have been reported for the incorporation of 3H-arginine and 3 H-tryptophan during meiosis in the microsporocytes of R. spathacea (Albertini 1971, 1975). According to Sauter and Marquardt (1967b), there is an inverse correlation between the nucleohistone content of the microsporocytes of P. tenuifolia and their RNA and protein contents: Meiotic cells characterized by high RNA and protein contents have little or no nucleohistones, whereas those with reduced RNA and proteins have invariably high nucleohistone content. In Hippeastrum belladonna (Amaryllidaceae), the lysine-rich histone profile of the sporogenous cells appears to be complex and different from that of the tapetum, suggesting that it may be associated with changes in the metabolic activity of the cells (Pipkin and Larson 1972).
Protein Synthesis and Metabolism. Patterns of An interesting observation about the microprotein synthesis and metabolism during microsporocyte meiosis are in many respects simi- sporocytes of L. longiflorum and Tulipa gesneriana is lar to those described for RNA. Perhaps this is not that they contain a unique histone that is easily sepsurprising, because variations in RNA metabolism arated by acrylamide gel electrophoresis. This hismight be expected to reflect variations in enzyme tone, designated as the meiotic histone, is virtually activity. Several investigators have exploited the absent in the somatic tissues of the plants but is possibilities offered by autoradiographic localiza- prevalent throughout microsporocyte meiosis tion of labeled amino acids to follow protein syn- (Sheridan and Stern 1967). A meiotic-specific histhesis during microsporocyte meiosis, with the tone has also been detected in microsporocytes of same end result in the diverse examples. This is L. candidum (Bogdanov, Strokov and Reznickova illustrated by the results from Taylor's (1959) work 1973; Strokov, Bogdanov and Reznickova 1973) and on Lilium longiflorum, which showed intense L. speciosum (Sasaki et al 1990; Sasaki and Harada nuclear and cytoplasmic incorporation of 14C- 1991). The meiotic histone is probably identical, or glycine into the microsporocytes through leptotene related, to a chromatin-associated protein desigand a reduced labeling during the rest of the meiotic nated as meiotin-1 (Figure 2.19), found in prophase. In the microsporocytes of Rhoeo discolor microsporocytes of L. longiflorum immediately pre(Albertini 1967, 1971) and Paeonia tenuifolia (Sauter ceding and during meiosis (Riggs and Hasenkampf 1968), there is very little protein synthesis from lep- 1991; Hasenkampf et al 1992). The origin, nature, totene until pachytene; these data are also consis- and molecular mode of action of this protein tent with a decline in the accumulation of remain elusive; one does not even know whether it cytophotometrically detectable proteins during affects directly or indirectly the meiotic cycle of the meiosis in the microsporocytes of Zea mays (Moss microsporocyte.
54
Figure 2.19. Immunostaining of meiotic microsporocytes of Lilium longiflorum. (a) Microsporocytes at pachytene stage immunostained with lily histone H1 immune serum. (b) Same, immunostained with lily meiotin-1 immune serum. Scale bar = 10 urn. (From C. Hasenkampf et al
Current thinking about the role of protein synthesis during microsporocyte meiosis has been largely influenced by a series of studies by Hotta, Stern, and associates. This has led to the view that one aspect of the molecular regulation of meiosis in liliaceous plants lies in the controlled synthesis of structural proteins required for the formation of the synaptonemal complex and of catalytic proteins involved in the recombination events. An early work (Hotta and Stern 1963b) established that in isolated microsporocytes of Trillium erectum, there is a surge in protein synthesis at the pachytene stage. Although the four labeled precursors used, namely, 14C-arginine, 14C-glycine, 14C-leucine, and 3 H-tryptophan, showed some fluctuations in the rate of incorporation, in other respects the four patterns were similar (Figure 2.20). Part of the proteins synthesized during pachytene could be inhibited by actinomycin-D; seen in the light of the claimed specificity of this drug as an inhibitor mRNA synthesis, this result lends itself to the conclusion that some of the proteins are encoded by newly transcribed mRNA. Reference was made earlier to the relationship between the synthesis of these proteins, synthesis of DNA, and the key events of meiosis like synapsis and crossing-over (Hotta, Parchman and Stern 1968; Parchman and Stern 1969). Other proteins that might play a role in initiating meiotic recombination are the endonucleases (Howell and Stern 1971).
Gametogenesis
1992. Temporal and spatial distribution of meiotin-1 in anthers of Lilium longiflorum. Devel. Genet. 13:425-434. © 1992. Reprinted by permission of Wiley-Liss Inc., a subsidiary of John Wiley & Sons Inc. Photographs supplied by Dr. C. Hasenkampf.)
pieces for heteroduplex formation (Hotta and Stern 1978). U-protein from lily microsporocytes has not yet been examined in sufficient detail to specify its role. There is nonetheless the intriguing possibility that it promotes recombination between DNA duplexes by producing single-strand tails of DNA. The finding of a specific DNA-binding or reassociation protein ("recA-protein") in the nuclei of microsporocytes of L. speciosum has made it highly probable that this protein has a special meiotic function, namely, that of catalyzing the reassociation of ss-DNA. Its transient appearance at prophase and absence at all other intervals of meiosis suggest a role for this protein in chromosome pairing and/or recombination (Hotta and Stern 1971a). Two homologues of the bacterial recA-protein (RAD-51 and LIM-15) are found abundantly in the leptotene and zygotene chromosomes and decrease significantly at pachytene (Terasawa et al 1995). Moreover, certain primary properties of recA-protein are modified by its state of phosphorylation. This is illustrated by the observation that dephosphorylation with phosphatase increases the affinity of the protein for ss-DNA and bestows the ability to bind ds-DNA also. Although cyclic adenosine monophosphate (cAMP)-dependent protein kinase cannot reverse the effects of dephosphorylation, a protein kinase present in the microsporocytes is able to reactivate the dephosphorylated form (Hotta and Stern 1979). According to Hotta et al (1979), the activity of this protein is regulated in Among the catalytic proteins, a DNA-unwind- some manner by homologous chromosome pairing ing protein ("U-protein") has been found to at zygotene. This is based on the evidence that increase in the microsporocytes of Lilium sp. coinci- microsporocytes of a variety of lily, "Black Beauty," dent with the appearance of chromosomes in which is characterized by the absence or near homologous alignment and subsequent crossing- absence of synapsis, shows a reduced level of recAover during meiosis. This protein, whose proper- protein during zygotene and pachytene. Induction ties are similar to those of the prokaryotic of tetraploidy in the microspore mother cells by U-proteins, apparently unwinds DNA duplexes colchicine has, however, allowed extensive chromofrom ends or nicks and thus provides single-strand
55
Anther Developmental Biology
L e p t o t e n e - Zygotene
120
100
x—X
C-Leucine C^GIycinetXIO''.) H 3 -Tryptophan (X10' 2 ) C^-Arginine
80
^
60
CL
U 40
10
20
30
40
50
60
70
80
90
Days of Cold Exposure Figure 2.20. Incorporation of 14C-arginine, 14C-glycine, 14Cleucine, and 3 H-tryptophan into proteins of isolated microsporocytes of lily at different stages of meiosis. The dotted line represents the pattern of 14C-leucine incorporation
by microsporocytes in situ. Plants were kept in storage at 1 °C to induce microsporocyte meiosis. (Reproduced from Hotta and Stern 1963a. / Cell Biol. 19:45-58 by copyright permission of the Rockefeller University Press.)
some pairing as well as restoration of the characteris- lated from lily microsporocytes have revealed its tically high level of recA-protein activity during the great similarity to the protein from E. coli. From the zygotene-pachytene stages. Another clue to the rela- standpoint of meiotic physiology, an important tionship of this protein to chromosome pairing is the observation made in this work is that there is a observation that disruption of homologous pairing major increase in recA-protein activity in the affects the levels only of recA-protein, but not of oth- microsporocytes at meiosis with a peak activity at ers. How this protein might function in the process of early pachytene (Figure 2.21). In another experirecombination is speculative, although it is likely that ment, recombination between two mutant plasit facilitates pairing between complementary DNA mids (recA~, HB101) that restores tetracycline strands at the zygotene and pachytene stages. resistance in transformed E. coli was used in an Whether phosphorylation is a mechanism for modi- assay to measure the activity of recA-protein in fying the DNA-binding properties of the protein extracts of meiotic cells. This work showed that, paralleling the increase in protein titer, the recomremains to be determined. Hotta et al (1985a) have also purified a recA-like bination frequency begins to increase as the protein from meiotic microsporocytes of a Lilium microsporocytes enter zygotene, reaching a peak at hybrid. Central to the concept of homologous pachytene. These observations point strongly to the recombination, recA-protein from Escherichia coli is conclusion that the enzymatic steps leading to known to rapidly promote the pairing of homolo- recombination are consummated at the pachytene gous DNA strands and to more slowly modulate interval (Hotta et al 1985a). the exchange of one strand in a linear duplex DNA for a strand from a closed circular ss-DNA moleMeiosis-specific Genes. In an attempt to clone cule, thereby creating heteroduplex DNA (Radding the genes that transcribe meiosis-related mRNAs, a 1982). The tests carried out with recA-protein iso- cDNA library was constructed to poly(A)-RNA
56
Gametogenesis
15
° 10 X
H
z >
5 5
I
L
' D' T
Figure 2.21. Changes in the levels of recA-like proteins during microsporocyte meiosis in lily. The protein was purified from meiotic cells at different stages and the DNAdependent adenosine triphosphatase (ATPase) activity was assayed as an indicator of recA-like protein activity. The open squares (bottom right) probably do not represent recA-like protein because of its very low DNA-dependent ATPase activity. I, premeiotic interphase; L, leptotene; Z, zygotene; P, pachytene; D, diplotene-diakinesis; T, tetrads. (From Hotta et al 1985a. Chromosoma 93:140-151. © Springer-Verlag, Berlin.)
clones that are expressed at low levels before microsporocytes enter the meiotic cycle (L1M-17, LlM-18) and decrease subsequently; one clone (LIM-15) appears transiently during early prophase. Heat-shock proteins (HSP), serine proteases, and proteins involved in DNA repair are some of the translated amino acid sequences of these gene products (Kobayashi et al 1994). Although it seems that further analysis of these recombinants will help to elucidate the molecular biology of meiosis, it is possible that most of the proteins identified will be more homologous to those that are generally found in other meiotic cells than to those specific to lily. An early meiosis cDNA clone from wheat florets identified by indirect screening using a lily meiosisspecific clone encodes a leucine-rich protein with a structure that resembles leucine zipper proteins. This DNA-binding motif generally found in eukaryotic genomes consists of a heptameric repeat of leucines that serves as the "zipper" in dimer formation. The dimers are located adjacent to a cluster of positively charged amino acids ("basic motif") that recognize sequence-specific target DNA. The gene is expressed only after leptotene, predominantly at the zygotene and pachytene stages, and the protein probably plays a structural role in the fabrication of the synaptonemal complex 0i and Langridge 1994). Enhanced GUS gene expression during the leptotene, zygotene, and pachytene stages of meiosis in lily microsporocytes by activation of yeast-derived meiotic promoter elements has been demonstrated; this is a strong indication of the ubiquitous nature of the cis-acting sequences required for transcription of meiosis-specific genes (Tabata et al 1993).
In summary, studies reviewed here, elaborate though they may seem, are only the beginnings for a from Lilium sp. microsporocytes spanning the general molecular analysis of meiosis in the entire premeiotic interphase through pachytene. Of microsporocytes of angiosperms. The impression the several meiotic-specific clones identified by dif- that one gains at first glance is that the entry of the ferential screening of the library with cDNA from microsporocyte into meiosis is a specialty far mitotic cells, one has been inferred from its DNA removed from the events of mitosis. This is probably sequence to code for a protein homologous to the true because of the unique process of generic recomsoybean (Glycine max; Fabaceae) heat-shock protein bination that occurs during the meiotic episode. The HSP-17 (Appels, Bouchard and Stern 1982; molecular studies have extended our knowledge of recombinational activities to a point where we can Bouchard 1990). Subtraction screening of a cDNA library made to talk in terms of a vulnerable period when the activzygotene-specific mRNAs of L. longiflorum ity of a vast array of enzymes for initiating recombimicrosporocytes with a probe made to premeiotic nation is controlled in a sensitive way. The key to microsporocytes has led to the identification of a this activity involves not only the availability of large number of meiotic-specific genes. As seen in genes, but also the frequency with which available Figure 2.22, most of the recombinants continue to be genes are transcribed. Obviously, any insight into expressed throughout meiosis beginning at the the mechanisms whereby the activities of meiotic zygotene stage. Included in this group are also genes are regulated would be valuable.
57
Anther Developmental Biology
B Y I
Z
P D T
LIM 1 3 4 5 6 7
8
10 **•+••
"'
9 10 11 12 14 15 16
rbc tub
Figure 2.22. Characterization of meiotic-specific genes from microsporocytes of Lilium longiflorum. Poly(A)-RNA from microsporocytes (A) and total RNA from other tissues (B) were subjected to Northern blot hybridization using probes made to premeiotic microsporocytes. Controls consisted of probes containing wheat histone H3 gene (H3), lily elongation factor 1 (EF1), tobacco large subunit of ribu-
lose-1,5-bisphosphate carboxylase (rbc), and lily a-tubulin (tub). Y, young anthers; I, interphase and leptotene stages; Z, zygotene; P, pachytene; D, diplotene-diakinesis; T, tetrads; M, total microsporocytes; Po, mature pollen grains; S, stem; L, leaves. The cDNA clones are numbered from LIM-1 to LIM-18. (From Kobayashi et al 1994; photograph supplied by Dr. S. Tabata.)
4. ANTHER CULTURE IN THE STUDY OF MEIOSIS Some of our modern understanding of the molecular biology of meiosis is derived from experiments in which a part, or the whole, of the
syndrome of meiotic events is followed in microsporocytes grown under controlled conditions. This technique involves excision of flower buds and anthers or extrusion of the microsporocytes and their subsequent culture in a defined medium. The conditions of an ideal culture system
58
Gametogenesis
would be such that the microsporocytes undergo the time of culture of the anthers, collapse of this synchronous meiotic development in culture when tissue being almost complete in anthers cultured at they are grown in the absence of the somatic tissues the leptotene stage and less so at later stages of the anther, but this aim has not been achieved. (Reznickova and Bogdanov 1972). Early experiments demonstrated that microA variation of the anther culture method was sporocytes need to go through part of the meiotic developed by Hotta and Stern (1963a). In this division cycle on the mother plant before anthers method, the leaves and perianth from flowers of can be cultured and microsporocytes allowed to Trillium erectum were excised, and the intact group advance through the rest of the division. Generally, of anthers attached to the receptacle along with a success has been sporadic if cultures are initiated at segment of the pedicel was transferred to the stages earlier than pachytene. Moreover, other con- medium. This protocol served as the cornerstone ditions being equal, anthers attached to the flower for several investigations on mitotic and meiotic bud are more prone to complete meiosis in culture regulation in the anthers of liliaceous plants (Hotta than anthers separated from the bud and cultured. and Stern 1963b, c, d; 1974, 1976; Shepard, For a period of time, these observations led to a Boothroyd and Stern 1974). In all of these different lively debate on the possible deficiencies of the methods of anther culture, the problem of explainnutrient medium and on the absence of certain ing the role of the somatic tissues of the anther, essential substances required for completion of especially of the tapetum, in meiotic development meiosis in the anther at the time of culture. See of the microsporocytes remains difficult. From this Vasil (1967) for a detailed discussion of these inves- perspective, a method developed for the in vitro tigations; a few representative cases will be culture of microsporocytes of liliaceous plants is described. significant (Ito and Stern 1967). The basic protocol Tradescantia paludosa illustrates a case in whichhas centered around the feasibility of extruding the sporogenous cells of anthers excised and cul- microsporocytes from the cut end of an anther into tured 5-8 days before leptotene continue mitotic a nutrient medium as a coherent filament by gentle divisions and end up as abnormal cells showing no pressure from the opposite end. Interestingly, even parallel to meiotic stages. Microsporocytes of in the most complex medium tested, only a small anthers excised during pachytene and later stages percentage of the microsporocytes cultured at the nearly always complete the meiotic division in cul- leptotene stage completes meiosis without causing ture (Taylor 1950). In an attempt to induce aberrant segregation and failure of the spindle microsporogenesis in excised anthers of Trillium organization at meiosis I. As the extruded erectum, Sparrow, Pond and Kojan (1955) used a microsporocytes approach a higher stage limit, variety of media and found that the best develop- such as late zygotene or pachytene, meiotic diviment was achieved by the incorporation of sion is nearly normal. Microsporocyte culture has 25%-50% coconut milk in the medium. A small per- been employed in several cytological and molecucentage of anthers excised even as early as the lar studies of meiosis in liliaceous plants (Hotta, Ito pachytene stage completed meiosis in culture, and Stern 1966; Ito, Hotta and Stern 1967; Hotta, although the survival rate increased in anthers cul- Parchman and Stern 1968; Parchman and Stern tured at later stages of the meiotic prophase. Vasil 1969; Hotta and Hecht 1971; Hotta and Shepard (1957, 1959) has reported experiments that show 1973; Ito and Hotta 1973; Sakaguchi et al 1980; that with the addition of kinetin, GA, and auto- Appels, Bouchard and Stern 1982). What has claved nucleic acids to the medium, meiosis pro- emerged from subsequent investigations is that the ceeds to completion in the microsporocytes of commitment of isolated microsporocytes to enter anthers of Allium cepa (onion; Liliaceae) cultured as the meiotic cycle is a function of the stage of the early as the leptotene stage. In cultured anthers of premeiotic cell cycle (Parchman and Roth 1971; Secale cereale, naphthaleneacetic acid (NAA) and Ninnemann and Epel 1973; Takegami et al 1981; Ito kinetin trigger microsporocyte meiosis without and Takegami 1982). It has been shown that necessarily completing the process and thus have a microsporocytes transplanted at growth phase 1 rather limited role (Rueda and Vazquez 1985). (Gl), S, or early G2 phases divide mitotically, Survival and limits of meiotic development of whereas those explanted at the late G2 phase promicrosporocytes of Lilium candidum in culture have ceed to divide meiotically (Figure 2.23). This has much in common with anthers of other species. led to the view that commitment of microsporoCompletion of tapetal development is also corre- cytes to meiosis is made during the G2 phase of lated with the stage of microsporocyte meiosis at premeiosis (Takegami et al 1981; Ito and Takegami
59
Anther Developmental Biology
%§1
#-v Figure 2.23. Meiosis in cultured microsporocytes of Trillium sp. (a, c) and Lilium longiflorum (b, d). Microsporocytes cultured at the completion of premeiotic DNA synthesis
reverted to mitotic division (a, b), whereas those cultured at the late G2 phase completed normal meiosis (c, d). (From Takegami et al 1981.)
1982). Although high yields of protoplasts of improve the development of anthers in culture. microsporocytes at meiotic stages can be obtained Meiotic division appeared to proceed with minimal routinely by enzyme treatments, success in induc- disorder in isolated microsporocytes of Secale ing them to form walls and complete the cell divi- cereale cultured in a mineral-salt medium containsion cycle has been sporadic (Ito 1973; Ito and ing amino acids and sucrose, although microspore Maeda 1973; Bajaj 1974a, b; Takegami and Ito 1975; development was arrested at the uninucleate stage (de la Pefia 1986). Maeda, Yoshioka and Ito 1979). Zea mays is the only example in which A simplification of the culture medium and completion of meiosis in the microsporocytes of microsporocyte meiosis has been induced in infloanthers cultured at the primordial stage of devel- rescence cultures. The explants were generally opment have been achieved when flower buds or 1-cm-long pieces of the tassel bearing approxiinflorescences are cultured. Using this approach, it mately six hundred spikelet primordia; the most was possible to follow by 3H-thymidine labeling mature spikelets in the explant consisted of undifthe entire meiotic cycle of the microsporocytes in ferentiated anther primordia enclosing essentially hermaphrodite flower buds of Cucumis melo nonvacuolate meristematic cells. Although (melon). The only difference between in situ and in microsporogenesis is found to proceed to the stage vitro development is the lack of synchrony of uninucleate microspores in anthers of tassels culbetween individual anthers of a flower bud in the tured in a mineral-salt-sucrose medium, addition latter condition (Porath and Galun 1967). Izhar and of kinetin is found to enhance the frequency of Frankel (1973b) found that culture of flower buds responding florets (Polowick and Greyson 1982, of Petunia hybrida and P. axillaris in a simple min- 1984; Pareddy and Greyson 1985). When spikelets eral-salt-sucrose medium supports the complete are cultured, microsporocytes proceed through cycle of microsporocyte meiosis in a small number meiosis and produce normal pollen grains even in of anthers. The spartan requirement for anthers of the absence of kinetin in the medium (Stapleton cultured flower buds to complete normal develop- and Bedinger 1992). From these results one can ment is underscored by the fact that the addition of appreciate why in most cases excised anthers are casein hydrolyzate or coconut milk does not unable to complete meiotic development in vitro. It
60 is not because the effects of the culture medium are trivial; it is because the culture medium lacks some essential ingredients that are provided by the flower or by the inflorescence. 5. GENERAL COMMENTS This chapter has assembled a considerable body of information on the development of the anther from its primordial beginnings and on the fate of each of the cell types formed. Although somewhat outside the scope of this account, the genetic and epigenetic factors concerned with the determination of the anther on the floral meristem have also been considered. The results show the power of mutational approaches in unraveling the complex mechanisms involved in the determination and differentiation of the floral parts. The investigations reviewed in the first part of this chapter underlie the dominant theme that will be evident throughout this book, namely, the study of plant reproductive biology has undergone a quiet revolution from a descriptive science to one emphasizing molecular and genetic concepts. The account of anther development has been built upon the traditional view that the anther as a whole consists of two distinct cell types programmed for different functions. The anther wall and the tapetum have taken on added significance with the demonstration that specific genes are expressed in these cells to define their functions. In this context the endothecium and stomium no longer appear to be passive cell types, but their
Cametogenesis
assigned role in the dehiscence of the anther seems to be modulated by the expression of specific genes. Although the tapetum remains an enigmatic tissue, evading a thorough examination, the ability to manipulate this tissue offers new opportunities to genetically engineer male-sterile plants. The picture of meiosis in the microsporocytes appears increasingly complex as more and more data have become available. Considering the fact that microsporocyte meiosis, like its counterpart in the megasporocyte, is a critical phase in the life cycle of the plant, it is surprising that what we know at the molecular level is restricted to studies on a model system. Moreover, in comparison to the work done on meiosis in animal cells and yeast, our knowledge of the molecular biology of meiosis in the microsporocytes is very limited. No meiotic genes from plants have yet been cloned and their protein products identified. The premeiotic anther enclosing diploid microsporocytes proceeds through an orderly series of changes to produce haploid microspores. Refinements in tissue culture have made it possible to study meiosis under controlled conditions in cultured microsporocytes, flower buds, and inflorescences or tassels. Future advances in the study of anther development will certainly involve the isolation and characterization of additional antherspecific genes, especially those activated during meiosis. The following chapter will return to the products of meiotic division of the microsporocyte to complete the discussion of the development of the angiosperm male gametophyte.
Chapter 3
Pollen Development and Maturation 1. Morphogenesis of Pollen Walls (i) The Callose Wall (ii) Assembly of Pollen Walls (iii)Pollen Wall Proteins (iv) Adaptive Significance of Pollen Walls 2. The First Haploid Mitosis (i) Biochemical Changes (ii) Ultrastructural Changes 3. Vegetative and Generative Cells (i) The Vegetative Cell (ii) The Generative Cell (iii) Biochemical Cytology of Vegetative and Generative Cells (iv) Division of the Generative cell (v) Isolation of Pollen Cells and Their Nuclei 4. General Comments
The work reviewed in the previous chapter has established that meiotic division of the microsporocyte results in the formation of four haploid microspore nuclei. They remain encased in the original callose wall of the microsporocyte to form a tetrad - the four-celled stage at the end of meiosis. There are two basic patterns of wall formation followed by these nuclei before they attain the status of cells. In most monocotyledons, immediately after each meiotic division of the microsporocyte, cell plate formation occurs in concert with a centrifugally expanding phragmoplast to produce the tetrad (successive cytokinesis). Alternatively, the norm in dicots is a type of division in which the four nuclei are walled off at the end of meiosis II (simultaneous cytokinesis). In either case, the first wall delimiting the microspore nuclei from each other is constituted of callose and not of cellulose. Later, after its release from the tetrad, each microspore forms its own wall comprising the
exine and intine. According to an informal morphological concept, the microspore represents the beginning of the male gametophytic generation, with the term "pollen grain" being reserved for the older microspore, particularly after its release from the tetrad. In some accounts, the first haploid mitosis is considered to terminate the life of the microspore and usher in the reign of the pollen grain. However, the terms "microspore" and "pollen grain" continue to be used interchangeably and synonymously in standard embryology literature to refer to the first cell of the male gametophytic generation of angiosperms. At a gross level, the pollen grain represents the structural unit that is programmed for terminal differentiation into sperm cells. For descriptive purposes, it is convenient to consider pollen development in terms of a series of changes beginning with the microspore after its release from the tetrad. These changes are very diverse and are geared to the ultimate function of the pollen grain as the transport vector for the sperm cells or their progenitor cell. Outside the microspore wall, sporopollenin is deposited in a diagnostic way to form the characteristic architectural pattern of the exine. Within the cytoplasm, vacuolation occurs, accompanied by a rearrangement of the organelle population of the microspore and leading to the first haploid pollen mitosis. The result is the formation of a large vegetative cell and a small generative cell. In some plants, the generative cell divides to form two sperm cells while enclosed in the pollen grain, but in the large majority of angiosperms, this division is postponed until the pollen tube is formed (Figure 3.1). It is believed that the bicellular condition is primitive and that the tricellular state originated indepen67
62
Cametogenesis
1. MORPHOGENESIS OF POLLEN WALLS archesporial cell
uninucleate microspore
bicellular pollen
(tricellular type)
(bicellular type)
Figure 3.1. Diagrammatic representation of microspore development and male gametogenesis beginning with the archesporial cell. G, generative cell; S, sperm cell; V, vegetative cell. (From Tanaka 1993.)
dently many times during angiosperm evolution (Brewbaker 1967). In order to cover in a logical manner the many diverse and seemingly unrelated events of pollen development, this chapter will begin with a review of the structure and morphogenesis of the pollen wall. Working from the outside to inside, the chapter will then examine the cytology of pollen development, which culminates in the formation of the vegetative cell and generative cell. Integration of the biochemical and molecular work into a review of the state of our present knowledge on gene expression during pollen development will be taken up in the next chapter. For an appraisal of the early literature on pollen development, the book by Maheshwari (1950) is still unsurpassed; the later work has been reviewed by Steffen (1963), Mascarenhas (1975), and Knox (1984a).
There are several important questions concerning pollen wall morphogenesis that reflect, on the whole, the process of maturation of the microspore from the tetrad stage to the point when sperm cells are generated. What is the significance of the impervious callose wall around the microspores? What is the source of the exine materials and what are the level and limits of morphogenetic control of exine fabrication? What specific processes concerned with wall differentiation are orchestrated in the pollen cytoplasm and in the tapetal cytoplasm? As will be seen, what emerges from a study of the diversity of processes is that pollen wall assembly is the result of a series of coordinated changes both outside and inside the cell. (i) The Callose Wall
The deposition of the callose wall begins centripetally on either side of the newly formed cell plate immediately after completion of microsporocyte meiosis. The daughter cells born out of meiotic division are devoid of plasmodesmatal connections, so that each cell carved out of the microsporocyte floats in the incipient tetrad, isolated from its neighbors but ensheathed in its own callose wall 0. Heslop-Harrison 1966a). Because of the nature of callose as a sealing agent, it is apparent that the walls of the tetrad and of the microspore present effective barriers to the penetration of large molecules from the anther locule and to the traffic of materials between the microspores. From this perspective, the reason why 14 C-thymidine and 3H-phenylalanine do not penetrate the callose-invested microspores of cultured anthers or flower buds is understandable (J. Heslop-Harrison and Mackenzie 1967; Southworth 1971). These same precursors are freely incorporated into microsporocytes or into microspores that are at later stages of development and are denuded of the callose sheath. At the cytochemical level, fluorescein diacetate is easily detected by fluorescence microscopy in microspores of various angiosperms mechanically released from their callose walls, but intact dyads and tetrads fail to show fluorochromasia. This could not happen if the callose wall had allowed free access to fluorogenic substrate into the microspore (Knox and Heslop-Harrison 1970b). J. Heslop-Harrison (1964,1968b) has made a case that the significance of insulating the newly formed microspores from the external milieu is to allow them to assert their genetic independence and
Pollen Development and Maturation
63
an outer exine and an inner intine is now widely accepted. In members of Poaceae, a granular, nonstructured middle layer termed Zwischenkorper (also known as Z-layer or oncus) is recognized; this layer is particularly prominent beneath the germination pores on the exine (Rowley 1964; Christensen and Horner 1974; J. Heslop-Harrison 1979a; El-Ghazaly and Jensen 1986a). However, other investigations have tended to view the Zlayer, which overlies the main intine layer, essentially as a part of the intine. It has a slightly higher electron density than the main intine material, but lacks the cytoplasmic tubules that traverse the latter 0. Heslop-Harrison and Heslop-Harrison 1980a; Kress and Stone 1982). In this chapter, pollen wall morphogenesis will be broadly categorized as the deposition of the outer patterned layers of the exine and the inner nonpatterned layers of the intine. One might say that morphological, developmental, chemical, and genetic differences between the exine and intine Despite its postulated importance, the callose could not be more distinct. It is now known that the wall is only transiently associated with the tetrad as pollen walls function not only as protective it is degraded at the termination of meiosis. This envelopes for the cytoplasm but also as storage results in the liberation of microspores with par- sites for a variety of proteins and other physiologitially constituted walls from the sectored cally active substances concerned in the interacmicrosporocyte; the timing of callose breakdown is tions of the pollen grain with the stigma and in correlated with the activity of callase (Eschrich 1961; some critical early functions in pollen germination. Echlin and Godwin 1968b; Mepham and Lane 1970; Electron microscopy has played a major role in Frankel, Izhar and Nitsan 1969; Izhar and Frankel uncovering the timing and pattern of the formation 1971; Stieglitz and Stern 1973; Stieglitz 1977). of the pollen walls. The development of the exine Analysis of callase activity separately in the will be considered first. microsporocytes and anther tissues of Lilium sp. has disclosed the interesting fact that the enzyme is conThe Exine. As alluded to in the previous chapcentrated mostly in the somatic cells of the anther ter, the exine consists of sporopollenin, a wall suband that the tapetum is probably the source stance known for its resistance to chemicals and (Stieglitz and Stern 1973; Stieglitz 1977). Exclusive biodegradation. The exact chemical composition of localization of transcripts of an anther cDNA clone sporopollenin is not known; one view is that it is a from Nicotiana tabacum encoding callase in the polymer of carotenoids and carotenoid esters tapetal cells of tobacco anthers provides strong sup- (Brooks and Shaw 1978). Another view of more port for this view (Bucciaglia and Smith 1994). recent vintage, based on solid state nuclear magThese observations underscore the fact that all com- netic resonance (NMR) spectroscopy, is that it conponent tissues of the developing anther are respon- sists mainly of aliphatic domains in addition to sive to one another through mechanisms whereby carbohydrates and aromatic compounds (Guilford individual cells in a tissue can assess the develop- et al 1988; Espelie et al 1989). Negative evidence mental state of cells in a neighboring tissue. against the presence of carotenoids as a major constituent of sporopollenin has come from the observation that norflurazon, an inhibitor of carotenoid (ii) Assembly of Pollen Walls synthesis, does not significantly inhibit the accuWall development involves the synthesis of mulation of sporopollenin in the anthers of materials in which the pollen grain is wrapped and Cucurbita pepo (Prahl et al 1985). Of potential value is part of the preparation toward the final journey of for gaining a complete understanding of the chemthe pollen to the stigmatic surface of another flower. ical composition of the exine is the development of The familiar differentiation of the pollen wall into methods for the isolation of purified exine fractions
establish cellular autonomy as the first cells of a new generation. As one can see, meiosis ensures the production of genetically different cells; the callose wall eliminates any possible effects of the informational pool of the sibs on the cytoplasm bequeathed by the microsporocyte. Another view of the significance of the callose wall is that it acts as a template for the development of the exine under the control of the microspore cytoplasm (Waterkeyn and Bienfait 1970). This remains a moot point; although the absence of callose in the tetrad is associated with poor exine deposition on the microspores (Vijayaraghavan and Shukla 1977; Pettitt 1981; Takahashi 1987; Worrall et al 1992; Fitzgerald et al 1994), tetrads without callose sheath also generate microspores with perfect exines (Periasamy and Amalathas 1991). In mutant Ambidopsis plants that release pollen as permanent tetrads, a callose wall may be absent but the exine development is no less complete as compared to wild-type pollen (Preuss, Rhee and Davis 1994).
64
Cametogenesis
caput -sexine — baculum
-exine
nexine 1 -nexine nexine 2 •intine
Figure 3.2. Section of the wall of an untreated pollen grain of Lilium sp. showing the stratification, m, fibrillar matrix material. (Reprinted with permission from J. Heslop-
from pollen grains and the production of antibodies to exine proteins (Southworth 1988; Southworth et al 1988; Chay et al 1992). Based on early light microscopic images, a complex terminology is now in use to describe exine stratigraphy. Briefly, this subdivides the exine into an outer, intricately sculptured portion, the sexine, and an inner, more or less continuous floor layer, the nexine. Rarely, as described in the pollen grains
Harrison 1968b. Pollen wall development. Science 161:230-237. ©1963. American Association for the Advancement of Science.)
that the signal for formation of the sexine is the completion of microsporocyte meiosis. At this stage, when the microspores are still in the tetrad configuration, the first elements of the sexine, referred to as primexine, appear as a layer of cellulose microfibrils interspersed between the plasma membrane of the microspore and its callose wall. The basic outline of the final sexine pattern is determined shortly thereafter by the formation of radiof Brassica oleracea and B. napus, the exine is sur- ally oriented columns of electron-opaque material rounded by a coating with properties of a biologi- in the cellulose matrix of the primexine. These, cal membrane for possible involvement in the referred to as probaculae, represent the progenitors recognition reaction on the stigma (Gaude and of the baculae. Observations on microspore Dumas 1984; Elleman and Dickinson 1986). The ontogeny in Lilium longiflorum have shown a sculpturing of the sexine is attributed to the pres- dynamic picture of pleated and folded lamellae ence of radially directed free-standing rods, or bac- that eventually compact into the baculae. The ulae, with enlarged heads (caputs). It should also lamellae are probably of plasmalemmal origin, be noted that heads of the baculae are linked at the organized in an assembly of membranelike structop to produce a reticulate pattern or are sealed to tures to form the baculae (Dickinson and Heslopcreate the facsimile of a roof or tectum. The tectum Harrison 1968). As described in S. pendula and S. and the nexine are, however, perforated by microp- officinalis, in the final phase of microspore wall ores to maintain communication with the outside. morphogenesis within the tetrad, the radial The sexine architecture dominates the entire pollen columns extend and link together over the outer grain surface except at the site of the germination face of the primexine to assemble the tectum; at the apertures, where it is absent or attains a low point. same time, they become confluent below the radial The nexine is subdivided into an outer nexine-1 columns to form the beginnings of nexine-1. With and an inner nexine-2, separated by an electron- some variations, the essential details of the prestransparent discontinuous region (Figure 3.2). One ence of isolating callose walls around microspores recurrent feature of exine development in the and of the primexine template of fibrillar material microspores of various angiosperms is the cyto- have been confirmed as common denominators in chemical variability between species and in the the formation of the pollen wall sexine in several same species during a developmental time span (J. other plants (Banerjee, Rowley and Alessio 1965; Skvarla and Larson 1966; Angold 1967; Godwin, Heslop-Harrison 1968b, 1975a). Echlin and Chapman 1967; Echlin and Godwin Pioneering work by J. Heslop-Harrison (1963a, 1968b; J. Heslop-Harrison 1969a; Dickinson 1970a; b) on pollen wall ontogeny in Silene pendula and Horner and Lersten 1971; Christensen, Horner and Saponaria officinalis (both Caryophyllaceae) showed
Pollen Development and Maturation
65
Lersten, 1972; Pettitt, 1976; Horner and Pearson Caesalpinia japonica (Fabaceae) and Bougainvillea 1978; Nakamura 1979; Shoup, Overton and Ruddat spectabilis (Nyctaginaceae) have shown that prior to 1980; Owens and Dickinson 1983; Southworth exine formation, the plasma membrane invaginates 1983; El-Ghazaly and Jensen 1986b; Takahashi and and forms a reticulate pattern corresponding to the Kouchi 1988; Takahashi 1989a, 1994; Jordaan and mature exine pattern. The plasma membrane, Kruger 1993; P£rez-Munoz, Jernstedt and Webster rather than the primexine, it is argued, is the prime 1993a; F. Xu et al 1993; Hess and Frosch 1994; mover in the determination of the basic exine patSimpson and Levin 1994; Nepi, Ciampolini and tern (Takahashi 1989b, 1995; Takahashi and Skvarla Pacini 1995). An extreme case, where a virtually 1991). According to this view, a protectum, which mature sexine is assembled around microspores appears as irregularly oriented threads, is the first while they are still protected within the callose exine layer to be deposited on the reticulate plasma sheath, has been described in Tradescantia bracteatamembrane. The threads probably serve as the scaffolding upon which the exine is built (Takahashi (Mepham and Lane 1970). The chemical nature of primexine remains 1993). An examination of primexine formation in unknown, but from the work of J. Heslop-Harrison pollen grains of Brassica campestris by rapid freeze(1968g) on microspore development in L. longiflo- substitution technology has added another dimenrum, we have some pointers about its closeness to sion to the controversy. Although it has confirmed sporopollenin. The strategy used in this investiga- the appearance of the primexine matrix as the first tion was to follow by a combination of ultrastruc- event followed by an undulation of the plasma tural and cytochemical methods the characteristics membrane, it has also shown that fibrous materials of the wall during the tetrad period and also after inserted into the invaginations of the plasma memthe release of microspores from the tetrad. These brane contribute to the beginning of probacula forexperiments showed that at the time of its first mation (Fitzgerald and Knox 1995). That there is no appearance, the material of the probaculae is not universal exine initiation template is not surprising resistant to acetolysis and apparently does not con- in view of the bewildering array of mature exine tain sporopollenin; however, like sporopollenin, patterns in angiosperm pollen grains. the material later deposited within the tetrad wall A foreshadowing of the germination pores on displays some physical and chemical properties the exine is first evidenced by the discontinuity of that confer a capacity to survive acetolysis. In other primexine over those areas of the microspore wall words, the microspore protoplast holds the key to destined to become pores. An important question the synthesis of the real sporopolleninlike materi- concerning pores is what cytoplasmic or genetic als; what form this key might take, and whether it factors determine their location. If a single cytois present in the microsporocyte as a long-lived plasmic process does exist in the control of sexine molecule or is a transient product of meiosis, morphogenesis, this process is the association of requires further work. elements of ER with certain areas of the plasma From a developmental perspective, two other membrane, marking out the sites of the pores. This views on the pattern determination of the exine are relationship, first described in a study of pollen worthy of consideration. One invokes a role for a wall ontogeny in Silene pendula (J. Heslop-Harrison plasma membrane-associated compound rich in 1963a), has since then been visualized in polysaccharides and proteins, known as glycoca- microspores as diverse as those of Zea mays lyx, in the process. According to Rowley and (Skvarla and Larson 1966), Helleborus foetidus Skvarla (1975), spinules of the exine of Canna gener- (Echlin and Godwin 1968b), Lilium longiflorum alis pollen are initiated within a three-dimensional (Dickinson 1970a), Sorghum bicolor (sorghum; network constituted of a compound resembling Poaceae) (Christensen and Horner 1974), Cosmos glycocalyx. On this basis, it has been suggested that bipinnatus (Dickinson and Potter 1976), Helianthus genetic expression of the form of the exine prior to annuus (Horner and Pearson 1978), Gibasis venusprimexine formation is mediated by glycocalyx, tula (Owens and Dickinson 1983), Triticum aestivum perhaps in association with the plasma membrane. (El-Ghazaly and Jensen 1986a), and Vigna vexillata The other view, with a different perspective, is that (Fabaceae) (P6rez-Munoz, Jernstedt and Webster the initial exine pattern is determined by the 1993b). A possible explanation for a role for ER in plasma membrane of the microspores while they defining the location of the germination aperture are still enclosed within the callose sheath of the on the exine is that it acts as a shield, preventing tetrad. Attempts to identify explicit early ultra- access to precursors of cellulose biosynthesis outstructural indications of exine development in side the plasma membrane (J. Heslop-Harrison
66
1971b); unfortunately, we have very little experimental evidence to support this view. Other observations indicate that the location of pores is programmed into microspores by virtue of their geometrical relationship in the formative tetrad stage. Attention was first drawn to this relationship by Wodehouse (1935), whose extensive light microscopic studies revealed that pollen grains have an inherent polarity because of the constant position of germination pores; in later years, it was shown that the siting of pores in the microspores is determined by the polarization imposed by the mitotic spindle and the plane of division during meiosis II (El-Ghazaly and Jensen 1986a). A causal relationship between the orientation of the mitotic spindle and the location of pores on the exine has also been established in experiments using centrifugation to displace cytoplasmic components, or colchicine to disturb spindle formation during microsporocyte meiosis. For example, if microsporocytes of L. henryi are centrifuged before the assembly of the mitotic spindle, pores are found to be wholly lacking in the microspores formed. In contrast, displacement of the spindle by centrifugation at later stages of microsporogenesis results in microspores with pores mostly in the normal configuration (J. Heslop-Harrison 1971b; Sheldon and Dickinson 1983). When microsporocytes are treated with colchicine, primary effects of the drug are evident in the failure of the surviving microspores to become polarized due to spindle disassembly; in such cells, pore sites assume a random distribution or appear ill defined (J. Heslop-Harrison 1971b). Colchicine treatment has also been shown to induce anarchic divisions in the microsporocytes of T. aestivum, with the location of pores remaining constant in relation to the spindle axes of the surviving microspores (Dover 1972). Taken together, these observations show that the spindle axes specifying the cleavage planes of the microsporocyte tetrad also identify the pore sites in the sexine of the microspores. Whether centrifugation and colchicine effects on spindle formation during meiotic division of the microsporocytes are reflected in the assembly of cytoplasmic components in the region of the microspores destined to form germination pores, remains to be seen. A regular participation of membranous structures is possibly the most distinctive feature of the disposition of baculae on the sexine. Just as there is a correlation between the position of the ER and sites of the germination pores on the exine, so too there is an association between the formation of primexine and a random orientation of elements of the ER close to the plasma membrane. It has been
Gametogenesis
observed that in Silene pendula, although there is no increase in the elements of ER in the microspore cytoplasm immediately after meiosis, this situation is quickly followed by the presence of ER at intervals in ordered arrays closely apposed to the plasma membrane. This precedes the appearance of the cellulose matrix outside the plasma membrane, as described earlier (J. Heslop-Harrison 1963a). The involvement of ER as a membranous template for primexine formation has also been observed in the microspores of Zea mays (Skvarla and Larson 1966), Berberis vulgaris (Berberidaceae) (Gabara 1974), Cosmos bipinnatus (Dickinson and Potter 1976), and Vigna vexillata (Pe'rez-Munoz, Jernstedt and Webster 1993b), whereas in other plants, the primexine-forming potential is associated with cytoplasmic organelles resembling vesicles or mitochondria. In Pinus banksiana, the vesicles appear to be derived from stacks of membranous cisternae somewhat similar to dictyosomes in their structure, and they have as their destination the periphery of the microspore cytoplasm beneath the plasma membrane (Dickinson 1971a). Although the disposition of dictyosomes and membrane proliferations seems to presage the arrangement of the baculae on the sexine of Lilium longiflorum (Dickinson 1970a), Gasteria verrucosa
(Willemse 1972), and Helianthus annuus (Horner and Pearson 1978), mitochondria are reported to be associated with the organization of the sexine in Linum usitatissimum (flax; Linaceae) (Vazart 1970). The presence of microtubules in the peripheral cytoplasm in the vicinity of the plasma membrane of microspores of P. banksiana, C. bipinnatus, and L.
longiflorum (Dickinson 1976), and as elements radiating from the nucleus to the plasma membrane in L. henryi (Dickinson and Sheldon 1984), V. vexillata (Mufloz, Webster and Jernstedt 1995), and V. unguiculata (Southworth and Jernsted 1995), reinforces the view that the cytoskeleton is responsible for the movement of the vesicles and membrane elements involved in sexine pattern formation on the cell surface. However, based on experiments involving colchicine treatment of microsporocytes of L. henryi, microtubules do not appear to be responsible for determining the sexine pattern. These experiments showed that despite the obliteration of microtubules by the drug and the chaotic nature of meiosis that ensued, pattern generation on the sexine was unaffected (Sheldon and Dickinson 1986). Somewhat similar results were obtained by treating microspores of Tradescantia virginiana with colchicine during the period of wall development (Tiwari 1989). These observations relegate the cytoskeleton to a secondary role in sexine
Pollen Development and Maturation
formation, perhaps molding the materials into the pattern already established by some other agents. A particularly fascinating aspect of these observations is the attention they have directed to the genetic control of sexine pattern formation. To what extent does the activity of the haploid genome determine the species-specific sexine architecture on the microspores? Since microscopically detectable elements of the sexine first appear on microspores encased in the tetrad, it is logical to assume that the process is under control of the haploid genome. In a thoughtful essay, Godwin (1968) has discussed some of the older evidence from the literature in support of this view, largely borne out by electron microscopy. Later work has, however, shown that the deposition of the basic patterned component of the sexine is associated with cytoplasmic control extending back to the diploid microsporocyte. Particularly instructive are the results of experiments using colchicine or centrifugation stress to disrupt meiosis in the microsporocytes of L. henryi. An extreme situation is seen where anucleate cell fragments produced by these treatments bear the typical sexine pattern approaching that of unstressed cells (J. HeslopHarrison 1971b; Sheldon and Dickinson 1983). Similarly, the development of the normal sexine pattern in microspores with incomplete chromosome complements in sterile hybrids arising out of crosses between the tetraploid Linum cateri and various related diploids also indicates a cytoplasmic, rather than a nuclear, control of sexine pattern determination (Rogers and Harris 1969). These results excited quite a lot of interest at that time, for they seemed to argue against the possibility that the control of sexine differentiation depends on gene action in the microspore. For how could the sexine pattern be formed so precisely in such genetically deficient, enucleate cytoplasmic enclaves if the microspore nucleus was the controlling agent? The mechanism by which the enucleate microspore fragments assemble the sexine is not understood, but it is hard to imagine how this could happen otherwise than by transcription of the genes in the microsporocyte and transmission of information from the diploid parent to haploid daughter cells by "delayed action" messages.
67. essential for the structural reinforcement of this layer (Dickinson and Heslop-Harrison 1968). Once the microspore has thus assembled the primexine and nexine-1, dissolution of the callose wall of the tetrad is the next order of business. This ensures the release of microspores, growth of nexine-2, maturation of the sexine, and assembly of the intine, all drawing upon the metabolites of the anther locule. Rowley and Southworth (1967) first showed that, during the formation of the inner exine layer (presumably nexine-2) around germination pores of the microspore of Anthurium sp. (Araceae), layers of lamellae consisting of unit membranes are routinely formed and that sporopollenin is subsequently deposited on them to form a homogeneous layer. Similar observations have been made on the mode of origin of nexine-2 in other plants, such as Hippuris vulgaris (Hippuridaceae), Populus tremula (Salicaceae) (Rowley and Dunbar 1967), Ipomoea purpurea (Convolvulaceae) (Godwin, Echlin and Chapman 1967), Lilium longiflorum, Lilium sp. (Dickinson and Heslop-Harrison 1971; Willemse and Reznickova 1980), and Silene alba (Shoup, Overton and Ruddat 1980, 1981). Of interest in terms of the origin of the lamella, in L. longiflorum the initiating structure has been designated as the parent lamella, which is presumed to be formed on the outer surface of the plasma membrane (Dickinson and Heslop-Harrison 1971). Structures strongly suggestive of lamellar organization of the lower layer of the sexine have also been described in earlier studies on the ontogeny of nexine-2 (Larson, Skvarla and Lewis 1962; Larson and Lewis 1963), so that the basis for the origin of this layer can be considered as a secure generalization.
Sporopollenin Synthesis. It is clear from the foregoing that the actual patterning of the exine is imprinted while the young microspores are protected within the callose sheath. Subsequent work has left little doubt that the molding of the structure of the mature exine by sporopollenin accumulation occurs after the release of microspores from the tetrad. Research on the origin and accumulation of sporopollenin on the pollen exine has a long history dating back to the mid-19th century. Early work was, necessarily, a domain in which the light microAs described in Lilium longiflorum, nexine-1 is scope was the preferred tool. By the 1930s, several invariably organized after the early definition of cytologists had documented the accumulation of the probaculae, but the precise events are subtle bodies with the chemical properties of sporopoland often difficult to detect. Accompanying the lenin on the locular faces of tapetal cells. The speadvance of the matrix material at the base of the cific location of these bodies or orbicules probaculae and fully integrated with them are immediately suggested a connection between the lamellae of plasma membrane origin. Nexine-1 is tapetum and sporopollenin synthesis. A developformed by the apposition of the lamellae, a process ment that followed from this in later years has been
68
Gametogenesis
the electron microscopic examination of orbicule sors are laid down upon the scaffolding of special development in pollen grains of a number of plants, initiating sites outside the plasma membrane such most notably, Poa annua (Poaceae) (Rowley 1963), as the probaculae or the lamellae. Considered in its Cannabis sativa (hemp; Moraceae), Silene pendula (J. entirety, the dramatic aspect of exine pattern forHeslop-Harrison 1962, 1963b), Zea mays (Skvarla mation on the pollen wall is the precise timing of its and Larson 1966), certain species of Oxalis initiation, the apparently simple beginning, the (Oxalidaceae) (Carniel 1967b), Salix humilis speed with which it goes forward, the complexity (Salicaceae), Populus tremula (Rowley and Erdtman of the final product, and the gaps in our knowledge 1967), Helleborus foetidus (Echlin and Godwin 1968a), of the critical steps. Lilium longiflorum, Lilium sp. (J. Heslop-Harrison and Dickinson 1969; Reznickova and Willemse Exine Inclusions. Intuitively, one might assume 1980), Citrus limon (Horner and Lersten 1971), that the function of the exine with its load of Sorghum bicolor (Christensen, Horner and Lersten sporopollenin is to provide mechanical protection 1972), Pinus banksiana (Dickinson and Bell 1972a, to the developing male gametophyte. Though this 1976), Helianthus annuus (Horner and Pearson 1978), is undoubtedly true, it also goes without saying Triticum aestivum (El-Ghazaly and Jensen 1986b), that several structural modifications of the exine and Ledebouria socialis (Hess and Frosch 1994). In H.have adaptive significance for the role of the pollen foetidus (Echlin and Godwin 1968a), L. longiflorum (J. in gamete transport. For example, it is a matter of Heslop-Harrison and Dickinson 1969), and T. aes- common observation that exine elaboration attains tivum (El-Ghazaly and Jensen 1986b) in which the its highest level of complexity in the pollen grains ontogeny of the tapetal cell has been closely fol- of insect-pollinated flowers, whereas pollen of lowed, progenitors of orbicules that have a different wind-pollinated species have relatively smooth electron density than the rest of the tapetal cyto- exine. In recent years considerable attention has plasm have been identified. Subsequently, these been focused on another facet of exine function, cytoplasmic territories accumulate in the vicinity of namely, in the storage and release of lipids, the plasma membrane, from which they are carotenoids, and proteins during pollen-stigma extruded, followed by their acquisition of a interaction. Reference was made in the previous sporopollenin coating in the anther locule. In the chapter to the fact that the presence of precursors of final lap of their journey from the tapetum, the pollenkitt and tryphine is a typical feature of orbicules align themselves on the microspore wall degenerating tapetal cells of certain plants. As to and are integrated into the developing exine (see the origin of pollenkitt, it is now known that Figure 2.8). These studies have enormously aug- osmiophilic globules of precursor materials dismented the light microscopic observations and charged from the tapetal cells migrate between reasserted the chemical identity of tapetal orbicules microspores and are eventually trapped in the cavwith sporopollenin of the exine. At the same time, ities of the exine, where they polymerize into polsome conflicting suggestions regarding the origin of lenkitt. Some globules are even lodged in the substrates for sporopollenin synthesis have interprobacular cavities (Pankow 1957; J. Heslopappeared in the literature. Among these suggestions Harrison 1968e, f; J. Heslop-Harrison et al 1973). are the possible formation of substrates in the mitoThe only detailed treatment of tryphine deposichondrial population of tapetal cells Q. Hesloption on the exine is that described in Raphanus Harrison 1962), their elaboration in both the sativus (Dickinson 1973; Dickinson and Lewis tapetum and the anther locule (Rowley, Miihlethaler and Frey-Wyssling 1959; Rowley 1964), and their 1973b). Of the two components of tryphine in this elaboration in the microspore cytoplasm through species, the proteinaceous material is applied first, dictyosome vesicles (Larson and Lewis 1963). These followed by a thick layer of the lipoidal mass. ideas either have been withdrawn or remain uncon- During the final stages of pollen maturation, these firmed, and so the origin of sporopollenin substrates components dry down, forming a thick encrustaremains a mystery. What is now established is that tion on the exine. The possible role of tryphine in following the early period of sexine assembly within pollen-stigma interaction will be considered later the tetrad, continued thickening of the wall occurs in this section. A significant component of the exine inclusion through the polymerization of sporopollenin precursors derived mostly from the tapetum and of pollen grains of Brassica napus consists of lipids. developing microspores. The goal of a diagnosti- These are presumed to be the products of the tapecally patterned exine is attained when the precur- tum and apparently transferred to the pollen surface late in development (Evans et al 1987). Various
Pollen Development and Maturation
69
mination pores is somewhat more complex than in the nonapertural areas and involves the incorporation of additional cellulosic strata, as in Linum grandiflorum (Dulberger 1989) and Cobaea scandens (Polemoniaceae) Q. Heslop-Harrison and HeslopHarrison 1991b), the formation of transfer-cell type The Intine. Developmentally, structurally, and wall ingrowths (Charzyriska, Murgia and Cresti physiologically, the intine is quite different from 1990), or the incorporation of pectocellulosic thickthe exine. This layer is hidden within the exine, enings encrusted with proteins (Nepi, Ciampolini although it begins to develop only after the disso- and Pacini 1995). In terms of function, the intine lution of the tetrad. Thus, in contrast to the exine, layers aid in the interaction of the pollen grains intine organization is programmed entirely by the with the stigma by serving as a repository of enzyhaploid genome of the microspore. Unlike the matic and other proteins and by facilitating pollen exine, the intine is nonsculptured and is therefore germination and pollen tube growth (J. Heslopstructurally a simple layer of fibrillar material Harrison and Heslop-Harrison 1991b). whose components merge with those of the exine. Since intine assembly is a rapid, multistep The presence of cellulose microfibrils embedded in process, it is not easy to identify a single, truly inia matrix of pectin and hemicellulose also defines tial step. However, in Olea europaea there is little the intine and gives it an identity different from doubt that the early reactions occur at the plasma that of the exine, but redolent of the primary wall membrane, which separates irregularly from the of a somatic cell. By progressive chemical extrac- inner layer of the exine, leaving an undulating area tion of the exine and the enclosed cytoplasmic con- beneath. Soon this area becomes filled with fibrillar stituents, the microfibrillar skeleton of the intine precursors of pectocellulosic wall materials; attaincan be isolated as an intact component (intine ment of uniform thickness follows (Pacini and "ghost") for analysis of its structural features. As Juniper 1979a). Dictyosomes are the prime movers shown in representative species, the orientation of in the formation of the intine and apparently supmicrofibril aggregates in the intine necessarily ply wall precursors for its microfibrillar organizatends to conform to the major topographical tion. Suggestive of the contribution of dictyosomes features of the exine and to the shape of the to the growing intine layer is the finding that pollen grain as a whole (J. Heslop-Harrison coated vesicles are dispersed near the plasma 1979a; Y. Heslop-Harrison and Heslop-Harrison membrane of freshly released microspores of 1982b). In some plants, especially in members Lilium longiflorum (J. Heslop-Harrison 1968c; of Cannaceae (Skvarla and Rowley 1970), Dickinson and Heslop-Harrison 1971; Nakamura Heliconiaceae (Kress, Stone and Sellers 1978), 1979) and O. europaea (Pacini and Juniper 1979a). Cymodoceaceae (McConchie, Knox and Ducker Attempts to correlate these observations with a 1982), and Asclepiadaceae (Vijayaraghavan and general mechanism for the transfer of GolgiShukla 1977), the pollen grains display extreme derived materials to the intine layer have not yet reduction in wall structure, with the exine very produced unequivocal results; one idea is that much reduced and the intine considerably thick- intine matrix substances leave the Golgi vesicles by ened. In these plants and in members of Poaceae, it exocytosis (Hess 1993). According to Mepham and is not difficult to identify by cytochemical methods Lane (1970), in addition to dictyosomes, ER also an outer layer of the intine, presumably pectic appears to be involved in intine secretion in the polysaccharide in nature, and an inner layer pri- microspores of Tradescantia bracteata. The basis for marily of cellulose (Kress and Stone 1982); rarely, as this assertion is the frequent occurrence of profiles in Lilium longiflorum, a third, fibrous layer of intine of ER-producing terminal vesicles that fuse with has been identified (Nakamura 1979). Even in other the plasma membrane, and the correspondence plants, in which a definitive separation of two between the materials of the tubular elements of ER intine layers is not possible, the outer stratum is and the developing intine, with electron microuniversally thought to be pectic, and the inner cel- scopically detected microtubule disposition lulosic, in nature. Although the intine appears beneath the intine layer at this time. It has been rapidly as a continuous layer at the inner edge of suggested that ER plays a role by absorbing soluble the incipient exine of the microspore, often it does precursors and transferring them to the dicnot attain its final thickness until late in pollen tyosomes for incorporation into the growing intine ontogeny. Intine specialization subtending the ger- skeleton. A similar role has been envisaged for ER analyses have demonstrated interesting biochemical differences between the internal cytoplasmic lipids of pollen grains, presumed to be the products of the haploid genome and the external exine-associated lipids (Evans et al 1987,1990,1991).
Gametogenesis
70
in intine function in the microspores of L. longiflorum (Nakamura 1979). This is not surprising in view of the fact that dictyosomes are an obligate intermediary between ER, on the one hand, and secretory granules and plasma membrane, on the other. The ability of isolated pollen protoplasts to regenerate a wall in culture does not seem to result in the formation of an intine; we can assume that besides the cytoplasmic organelles, there must be some other precursors in limited supply necessary for intine assembly (Miki-Hirosige, Nakamura and Tanaka 1988; Y. Wu and Zhou 1990). Summarizing, we see that formation of the pollen wall is a complex process for which information is derived from both sporophytic and gametophytic genomes. What are the forces that drive this dynamic morphogenetic process? Although this question will remain unanswered for some time, the temporal and spatial separation of the actions of the two genomes suggests that the primary events rely on information transfer and the synthesis of specific proteins. (iii) Pollen Wall Proteins
There is general agreement that pollen wall development is a period of accumulation of great many proteins of sporophytic origin. The proteins accumulate in both the exine and intine layers; deposits in the exine are confined to the outer sexine, especially to the interbacular spaces, where they are found in association with pollenkitt and tryphine. Although proteins are initially deposited in the pectic part of the intine, in the mature pollen grain they form a conspicuous zone between the pectic and cellulosic strata that can be considered as a separate layer by itself (J. Heslop-Harrison and Heslop-Harrison 1991b). Tsinger and PetrovskayaBaranova (1961) are generally credited with the first demonstration of the presence of included proteins in the pollen walls. By staining tests, these investigators noted high concentrations of proteins in the walls of pollen grains of Paeonia sp. and Amaryllis sp. (Amaryllidaceae) and showed that some of the proteins possess enzymatic properties. As later research showed, their conclusion that the pollen "wall. .. appears to be a living, physiologically active structure, playing a very responsible role in the processes of interchange between the pollen grain and its substrate," has proved to be prophetic. This conclusion also provided the starting point for a chain of interactive investigations, utilizing electron microscopy, cytochemistry, and immunology to gain a deep insight into the nature
and function of the wall proteins. Outstanding among the many contributions of more recent years has been the work of Knox and HeslopHarrison (1969, 1970a), who showed by high resolution cytochemistry the presence of hydrolytic enzymes like acid phosphatase, esterase, RNase, amylase, and protease in the inner cellulosic intine layer of pollen walls of about 50 angiosperms. In this survey, which included all major pollen types, it is interesting that the enzyme activity is strikingly localized in the vicinity of the germination pores, without entirely bypassing the rest of the intine. Upon hydration of the pollen grain, the proteins diffuse through the exine, particularly at the germination pores, and are released into the medium. SDS-PAGE electrophoresis of diffusates from pollen grains of Cosmos bipinnatus and Ambrosia elatior (Asteraceae) has shown that the wall proteins are heterogeneous in nature. The diffusate from C. bipinnatus revealed more than 15 bands, including two glycoproteins and three or four major fractions (Howlett et al 1975). Using antisera prepared in rabbits against diffusates from pollen walls of Gladiolus gandavensis (Iridaceae), Phalaris tuberosa (canary grass; Poaceae), Populus alba, P. deltoides (poplars), A. trifida, and C. bipinna-
tus, the intine was shown by immunofluorescence to be the principal site of protein accumulation (Knox, Heslop-Harrison and Reed 1970; Knox and Heslop-Harrison 1971c; Knox, Willing and Ashford 1972; Howlett, Knox and Heslop-Harrison 1973). The antiserum from Ambrosia trifida pollen was found to cross-react with pollen from four other species of the genus, namely, A. artemisiifolia (ragweed), A. elatior, A. psilostachya, and A. tenuifolia,
and in each case the association was restricted to the intine and, to some extent, to the exine; however, there was no evidence for the presence of antigen in the microspore cytoplasm or in the vegetative cell of the mature pollen grain (Knox and Heslop-Harrison 1971a; Howlett, Knox and Heslop-Harrison 1973). The characteristic parts of the intine where enzymes predominate are the ribbonlike tubules within the cellulosic layer (Figure 3.3). This was established by ultracytochemical localization of acid phosphatase activity in the intine of Crocus vernus (Iridaceae) (Knox and Heslop-Harrison 1971b). Similar tubules, presumed to harbor RNase, have been described in the intine of Malvaviscus arboreus (Malvaceae) (J. Heslop-Harrison et al 1973). Another fine-structural study has called attention to the presence of stratified ER encrusted with ribosomes and of cytochemically detectable proteins encased in the cellu-
Pollen Development and Maturation
71
Figure 3.3. (a) Naphthol AS-B1 phosphate pararosaniline reaction for localization of acid phosphatase in the intine of pollen grains of Crocus vernus. (b) Control: reaction without substrate, (c) Electron micrograph of the pollen wall showing localization of acid phosphatase in the ribbonlike
tubules of the intine. e, exine; i, intine; pm, plasma membrane. Scale bars = 10 um (a, b) and 1 um (c). (From Knox and Heslop-Harrison 1971b. J. Cell Sci. 8:727-733. © Company of Biologists Ltd.)
lose matrix of the microspores of Cosmos bipinnatus during the period of intine deposition (Knox and Heslop-Harrison 1970a). As shown in Figure 3.4, immunolocalization of cutinase (the enzyme that breaks down cutin) in the pollen grains of Brassica napus by a polyclonal antiserum to Fusarium cutinase has revealed that the enzyme is confined predominantly to the intine, with a faint presence in discrete regions of the pollen cytoplasm (Hiscock et al 1994). In extending this work in a developmental context, both cytochemical and immunofluorescence methods have shown that although microspores freshly released from the tetrad lack wall-associated enzyme activity, they acquire it soon after,
during the period of expansion and vacuolarion, when the intine growth is most active (Knox and Heslop-Harrison 1970a; Knox 1971). When the activity of acid phosphatase as a marker enzyme for intine proteins of pollen grains of B. oleracea was followed by quantitative cytochemical methods, two periods of enzyme accumulation were observed. The first coincided with the late vacuolation phase of the microspore, and the second with the period of pollen maturation (Vithanage and Knox 1976). A similar pattern of acid phosphatase activity has been verified by qualitative observations and quantitative analyses in the intine of Lolium perenne (ryegrass; Poaceae) pollen (Vithanage and Knox 1980). In Silene alba, acid
72
Figure 3.4. Immunolocalization of cutinase in the intine of pollen grains of Brassica napus. (a, b) Fluorescence micrographs of sections of pollen grains treated with cutinase polyclonal antiserum followed by goat antimouse fluorescein isothiocyanate (FITQ-conjugated secondary antibody showing fluorescence in the intine and in discrete regions of the cytoplasm (arrows), (c, d) Controls showing light micrograph (c) and fluorescence micrograph (d) of sections of pollen grains treated with mouse pre-immune serum followed by FITC-conjugated secondary antibody. Arrow in (c) indicates the intine. Scale bar = 8 urn. (From Hiscock et al 1994. Planta 193:377-384. © Springer-Verlag, Berlin. Photographs supplied by Dr. H. C. Dickinson.)
Gametogenesis
Harrison 1974). In contrast to the intine proteins, the exine materials are derived from the tapetum, which during dissolution releases its melange of materials, including membrane-bound cisternae, into the anther locule, where they lodge in the exine cavities. The cisternae containing granular materials are thought to be precursors of the proteins. In the tapetal cells of B. oleracea, esterase, a marker enzyme of the exine, attains its peak activity before the cells begin to disintegrate. Indicative of the transfer of the enzyme from the tapetum to the exine, an increase in the activity of esterase in the pollen wall exine is found to coincide with the dissolution of the tapetum (Vithanage and Knox 1976). The pattern is comparable to that seen for esterase activity in anthers of Helianthus annuus (Vithanage and Knox 1979) and Lolium perenne (Vithanage and Knox 1980). Although acid phosphatase is of gametophytic origin and is present mainly in the intine in the pollen of B. oleracea and L. perenne, in H. annuus it is present in both exine and intine sites. Moreover, coincident with tapetal dissolution, enzyme activity in H. annuus is localized in the sexine; this is indicative of a sporophytic origin of the exine-bound enzyme (Vithanage and Knox 1979). It should be pointed out that pollen grains of the species just listed also harbor intine proteins, which are found mainly, but not exclusively, in the vicinity of the germination pores.
Among other proteins accumulating in the pollen walls are the allergens that cause the seasonal disease known as hayfever. In Ambrosia, a phosphatase is initially detected in the peripheral genus that includes the ragweeds, the most active pollen cytoplasm prior to its exclusive localization allergens are two globular proteins, antigen-E and in the domain of the intine (Shoup, Overton and antigen-K, with the former being somewhat more Ruddat 1981). It would seem from these and the active than the latter. By immunofluorescence foregoing observations that intine proteins are syn- method, Knox and Heslop-Harrison (1971a) thesized by the microspore cytoplasm and are showed that antigen-E is localized in the intine of therefore gametophytic in origin. This makes sense the pollen grain of A. artemisiifolia and that its since, unlike the exine proteins, the intine proteins appearance coincides with the earliest indication of do not show homology to proteins released by the antigen reaction with the whole pollen extract antiserum. The ragweed pollen, with the structural tapetum. specialization of its intine as a repository for allerBecause of its morphological nature with freegens, provides a striking contrast to the ryegrass standing baculae or the reticulate pattern with pollen, where the active allergens are found in the open cavities, the exine provides large volumes of more or less enclosed space for the storage of pro- exine. The major allergen of ryegrass pollen is an teins. An unequivocal demonstration of the deposi- acidic glycoprotein, and immunocytochemical tion of proteins in the exine was possible using the methods have permitted detection of the allergen large pollen grains of Malvaviscus arboreus, Hibiscus not only in the exine but also in the pollen cytorosa-sinensis, Abutilon megapotamicum, and Anoda plasm (Vithanage et al 1982; Staff et al 1990). The crispa (all of Malvaceae) (J. Heslop-Harrison et al allergic protein of Cryptomeria japonica (Japanese 1973), Cosmos bipinnatus (Howlett, Knox and cedar; Taxodiaceae) is localized all over, and Heslop-Harrison 1973), Iberis semperjlorens, and J. between, the exine and intine layers, as well as in sempervirens (J.Heslop-Harrison, Knox and Heslop- the generative cell wall and in organelles like the
Pollen Development and Maturation
Golgi vesicles (Miki-Hirosige et al 1994). Recent developments in the characterization of the allergens of pollen grains have included isolation of cDNA clones encoding the allergens and determination of their nucleotide sequences and the deduced amino acid sequences; these investigations are described in Chapter 4. The discovery of allergens in the exine and intine layers of pollen grains and their characterization hold considerable promise for the future in terms of a better understanding of the allergic response, which could help alleviate the suffering due to hayfever. (iv) Adaptive Significance of Pollen Walls
An appropriate way to end the discussion on pollen walls is to consider their functional significance in a broad biological context. Pollen grains of angiosperms are astonishingly diverse in the details of their exine ornamentation. Palynology literature is replete with examples of the species-specific architecture of the pollen exine, which has permitted investigation of the systematic affinities of certain groups of plants; in some groups, additional features related to pollination ecology are associated with the exine. Although the pollen wall structure is now recognized as an adaptation of great importance in the evolution of angiosperms, the concern here is with the significance of pollen walls in the overall reproductive strategy of this group of plants. From this perspective, interest in pollen walls arises due to their roles such as protection against excessive desiccation, facilitation of efficient germination of the pollen grain, and promotion of interaction of the pollen grain with the stigma. In periodic reviews, J. Heslop-Harrison (1971a, 1975a, 1976), J. Heslop-Harrison and Heslop-Harrison (1991b), and }. Heslop-Harrison et al (1975) have extensively commented on these aspects of the function of pollen walls and the following account has been drawn largely from these articles. For a long time, researchers have been preoccupied with the protective function of the pollen walls. The presence of a thick sporoderm material is clearly not the only mechanism employed in nature to prevent the loss of water from cells, for both single-celled and multicellular plants achieve the same result with the cuticle. The development of the concept of pollen walls nevertheless provides an interesting example of the upper limit of structural specialization that is achieved by plant cells in response to selective pressures. The reason why pollen grains need to have a complex wall for
73
protection is clearly related to the unpredictable conditions of their life on the dry land after they leave the anther. There are difficulties and uncertainties associated with this interpretation, but it provides a framework for understanding how different cell types respond to the same environmental stress. A problem of enduring interest relates to the role of pollen walls during the critical period when a dehydrated pollen grain lands on the receptive stigma and begins to germinate. In recent years it has been possible to correlate the development of the germination pores of the pollen grain with physiological changes that follow upon hydration and germination. This has made it possible to demonstrate that besides its traditional role as a preferred path of emergence of the pollen tube, the aperture controls the water balance during germination. The ingress of water into the dehydrated pollen due to exposure of the intine surface at the pore site, in contrast to the exclusion of water due to occlusion of the pores by sporopollenin deposit, has revealed a positive feedback system during hydration; this results in an increase in the surface of the intine available for water uptake with each increase in volume. The exine also functions in a similar way in the control of water uptake in nonaperturate pollen grains (J. Heslop-Harrison 1971a). Examples of specialization of the intine for a critical early function in pollen germination are described by J. Heslop-Harrison and HeslopHarrison (1991b). Investigations of the movement of included proteins from the exine and intine domains are also changing our concepts of their possible role in the physiology of the pollen grain. Indications are that following pollination, proteins leach out of the pollen walls and take up temporary residence on the surface of stigmatic cells. Some of these proteins, especially those held in the exine, include enzymes that play important roles in the pollen-stigma interaction, possibly in the germination of the pollen grain and early growth of the pollen tube on the stigma. Intine-borne proteins, on the other hand, appear responsible for the penetration of the stigma by the emerging pollen tube, as shown in studies in certain Brassicaceae. In Raphanus sativus and other members of Brassicaceae, and in members of Caryophyllaceae, Asteraceae, and Malvaceae, the surface of the stigma papillae bears an external proteinaceous coating overlying the cutinized outer layer of the cell wall (Mattsson et al 1974; J. Heslop-Harrison, Heslop-Harrison and Barber 1975). This layer,
74
known as the pellicle, is the first barrier encountered by pollen grains and serves to attach them to the stigma. Dickinson and Lewis (1973a) showed that penetration of the pollen tube through the cuticle proceeds without hindrance in both compatible and incompatible pollinations and that incompatibility reactions set in as the pollen tube begins to grow. It appears that pollen tube penetration of the cuticle depends on the presence of a precursor of cutinase that is carried in the intine and is activated when the pollen tube comes in contact with the pellicle. The basis for this interpretation is the observation that although enzymatic degradation of the pellicle in Agrostemma githago (Caryophyllaceae) does not impair pollen germination on the stigma, it prevents the penetration of the pollen tube, possibly due to the failure of cutinase activation (J. Heslop-Harrison and Heslop-Harrison 1975). The interpretation also receives support from similar experiments done in Crocus chrysanthus. In this species, in which the stigmatic papillae do not bear a pellicle, the germinating pollen grain directly encounters the cuticle, whose enzymatic erosion permits the penetration of the pollen tube. Protein secretions from the underlying cell, stored in the lacunae or cavities of the cuticle, apparently contain the factor that activates intine-held cutinase; here also, as in A. githago, pollen tube fails to penetrate the stigmatic papillae from which the surface secretions have been removed by enzyme treatment (Y. HeslopHarrison 1977). A conceptual reorientation in the field of plant breeding was initiated by studies that showed that exine-held proteins released by pollen grains during their early period of attachment to the stigma play a role in determining compatibility relationships. Although this remarkable fact was suspected toward the end of the last century, the impetus for much of the later experimental studies was the postulation of J. Heslop-Harrison (1968a) that implicated exine-based tapetum-derived proteins in sporophytic incompatibility systems. A series of experiments involving germination of pollen grains and growth of pollen tubes in compatible and incompatible pollinations in a wide range of plants, and the effects of proteins isolated from pollen walls of compatible genotypes in overcoming the inhibition of germination of incompatible pollen, have demonstrated the interference by proteinaceous factors of the pollen wall as a basic physiological characteristic of the incompatibility system. These proteins are designated as "recognition substances" that control pollen germination
Cametogenesis
and pollen tube growth during both compatibility and incompatibility interactions. The role of exine proteins in compatibility control was first clarified in experiments involving members of Brassicaceae and Asteraceae. In /bens sempervirens, which has an intraspecific incompatibility system, the early events following self- and cross-pollinations are more or less similar. These include hydration of the pollen grain and release of exine-bound proteins onto the stigmatic surface. In selfed plants, the rejection reaction develops soon after pollen tube formation and is accompanied by callose occlusion of the stigmatic papilla adjacent to the pollen grain. The rejection reaction is also induced in excised stigmas by agar blocks enriched with exine proteins, as well as by isolated tapetal fragments as a substitute for the exine proteins (J. Heslop-Harrison, Knox and Heslop-Harrison 1974). A similar syndrome of changes at the stigmatic surface accompanies compatible and incompatible pollinations in Raphanus sativus. A chemical fraction isolated from the exine of R. sativus, consisting mainly of proteins and tryphine, elicited the production of a callosic reaction on the stigmatic papillae in the appropriate genotype in a way similar to an incompatible pollination. This is what one would expect if the incompatibility reaction is due to exine-held materials (Dickinson and Lewis 1973a, b). Callose occlusion of the stigmatic papillae following incompatible pollination has also been described in Cosmos bipinnatus. Genetic evidence for the role of exine-borne proteins in intraspecific incompatibility in this species was provided by offsetting self-incompatibility by pollination with pollen mixes of killed compatible pollen and fresh self-pollen or by self-pollination following application of compatible pollen wall diffusate to the stigma (Knox 1973; Howlett et al 1975). Although the percentage of successful illegitimate crosses was not overwhelming, these observations indicate that exine proteins are factors to be reckoned with in controlling the breeding behavior of plants. A function for the intine proteins in the control of interspecific incompatibility has been demonstrated in poplars. Failure of pollen grains to germinate on the foreign stigma is the usual cause of hybrid failure in Populus deltoides X P. alba crosses. However, this incompatibility barrier can be breached by disguising pollen grains from the foreign donor by mixing them with acceptable pollen or by substituting for the recognition pollen a protein preparation leached from the pollen intine. By this method, P. deltoides X P. alba hybrids were
Pollen Development and Maturation
obtained, using P. deltoides as recognition pollen (Knox, Willing and Ashford 1972; Knox, Willing and Pryor 1972). Studies reviewed in this section are perhaps only the beginnings of future research geared to apply the new knowledge about pollen wall structure toward ensuring efficient pollination and fertilization in flowering plants. There are several questions at the molecular level that remain to be answered before our understanding of the functions of the pollen wall can be considered complete. They range from the reactions triggered by the recognition materials to the secondary signals elicited in the participating cells. Research addressed to these questions as they apply to the genetics of self-incompatibility will be considered in Chapter 9 of this book. 2. THE FIRST HAPLOID MITOSIS In the ontogeny of the pollen grain, the most studied and the most impressive division is the first haploid mitosis, heralding the formation of the vegetative and generative cells. The elaborate network of preparatory steps that are completed by the microspore before this division suggests that they may represent a set of common denominators for an asymmetric division. Although microspores of different stages of development collected from intact anthers continue to provide excellent materials for cytological and biochemical studies, successful attempts have been made to induce division in microspores of cultured immature inflorescences (Barnabas and Kovacs 1992; Pareddy and Petolino 1992) and cultured anthers (Sparrow, Pond and Kojan 1955), in microspores cultured at the uninucleate stage of development (Tanaka and Ito 1980, 1981b; Tanaka, Taguchi and Ito 1980; Benito Moreno et al 1988; Stauffer, Benito Moreno and Heberle-Bors 1991; Tupy, Rihova and Zarsky',1991; Touraev et al 1995b) and in microspore and pollen protoplasts divested of their walls by enzymatic digestion (Bhojwani and Cocking 1972; Rajasekhar 1973; Bajaj 1974a; Pirrie and Power 1986; Baldi, Franceschi and Loewus 1987; Tanaka, Kitazume and Ito 1987; Lee and Power 1988; Miki-Hirosige, Nakamura and Tanaka 1988; Pental et al 1988; Zhou 1989a, b; Y. Wu and Zhou 1990, 1992). Apparently, the microspore genome contains all the information needed to complete its maturation, which thus appears to be independent of the somatic tissues of the anther. Results of investigations on the maturation of microspores under in vitro conditions have opened up questions relating
75
to the need for the mitotic division of the microspore nucleus and its asymmetrical disposition for continuation of the gametophytic program. The work of Touraev et al (1995b) showed that the gametophytic program resulting in pollen tube growth is faithfully continued by cultured microspores of Nicotiana tabacum even when they divide symmetrically into two cells. Pollen tubes are also formed in microspores of a transgenic tobacco cultured in the presence of high concentrations of colchicine, which effectively blocks the first haploid mitosis (Eady, Lindsey and Twell 1995). The in vitro methods are obviously of great advantage in the manipulation of pollen cells and in the analysis of cell cycle parameters of microspores unimpeded by the influence of the parent plant and in the absence of the tapetum. In addition, a high probability of success in the gametophytic selection of desirable genes for transmission to the progeny has been indicated using a transgenic approach with in vitro cultured microspores (Touraev et al 1995a). A striking and consistent feature of the newly released microspore is the presence of a dense, granular, nonvacuolate cytoplasm poised for resumption of metabolic activity. This cytoplasm is in part originally bequeathed from the microsporocyte and is in part synthesized by the microspore immediately after its liberation from the tetrad. It is, of course, not unreasonable to expect nucleo-cytoplasmic interactions to be important determinants of gametophytic expression in the microspore at this stage. Ultrastructural features suggestive of such interactions are nuclear invaginations in the microspores of Podocarpus macrophyllus (Podocarpaceae) (Aldrich and Vasil 1970) and Pinus banksiana, invaginations in the latter probably enclosing RNA (Dickinson and Bell 1970, 1972b; Dickinson and Potter 1975), and the association of the nuclear envelope with membrane-bound inclusions in the microspores of Lilium longiflorum (Dickinson 1971b) and Cosmos bipinnatus (Dickinson and Potter 1979). Unfortunately, the control exerted by these interactions in the subsequent reprogramming of the microspore remains a mystery. Increase in the microspore cytoplasm inevitably leads to an increase in the volume of the microspore, which is accentuated by vacuolation. The process of vacuolation is truly remarkable, because it presents an entirely new structural profile of the microspore and is accomplished by the coalescence of a large number of tiny vacuoles or by the formation of a single large vacuole. In Phalaris tuberosa, vacuolation occurs in a pre-
76
Gametogenesis
dictable time interval and occupies a major part of the developmental time span of the microspore, when its diameter more than doubles to about three-quarters of the final size. Toward the end of maturation of the microspore, the vacuole is resorbed and the resulting void is filled with new cytoplasm (Vithanage and Knox 1980). A phenomenal increase in size, up to about 12 times the volume of the postmeiotic products, occurs during vacuolation of microspores of Triticum aestivum (Mizelle et al 1989). The other event in the microspore that is almost invariably linked to vacuolation is the displacement of the nucleus from its central position to the periphery of the cell, where mitosis occurs. It is remarkable that in some plants nuclear migration is not a random event, but is site specific as the organelle is propelled to a site opposite the germination pore (Sanger and Jackson 1971a; Christensen and Homer 1974). The location of the pore thus appears to dictate where the nucleus will migrate and where the wall partitioning the cell will be formed.
1200
800-
400
, pp. 131-174. Berlin: Springer-Verlag. © Springer-Verlag, compatible; —>, incompatible. (From Shivanna 1982. In Berlin.) Pin flower (b) and thrum flower (c) of Linum pubesExperimental Embryology of Vascular Plants, ed. B. M. Johri, cens. Scale bar = 10 mm. (From Dulberger 1973.)
lous species. In the typical distylous species, one prevent self-pollination in cleistogamous flowers, morph is characterized by long style and short sta- cleistogamy will not be taken up in this chapter mens and the other possesses short style and long again. stamens (Figure 9.1). The preferred terms used to By far, one of the most extensively investigated designate the long-styled and short-styled morphs distylous genera is Linum (Ockendon 1968; are pin and thrum, respectively. The salient feature Dulberger 1973, 1974, 1981, 1987; Rogers 1979; of the breeding behavior of the two morphs is that Ghosh and Shivanna 1980a, b). In several species of successful pollination ensues between flowers of Linum, the reciprocal arrangement of the stamens different morphs (intermorph), whereas incompat- and style is associated with differences in the suribility syndromes are unrestrained after pollination face features of pollen grains and stigma. Generally, between flowers of the same morph (intramorph). the pollen exine of thrum flowers is ornamented Although most of the claims of distyly are based with excrescences of uniform size, whereas in pin largely on stamen and style length, in many species morphs the outgrowths are of two or more types the incompatibility package includes differences in (Dulberger 1973, 1981; Rogers 1979). Similarly, the pollen characters such as size, shape, and exine stigmatic papillae of the two morphs are of differsculpturing, and in the morphology of the stigma, ent sizes and exhibit differences bordering on the for example, the density and shape of the papillae. structure of the walls of the papillae. For example, Another situation in which two morphologi- in L. grandiflorum and L. pubsecens, the stigmatic cally different flower types are borne on the same papillae of pin morphs are nearly twice as long as plant, goes by the name "cleistogamy." The hall- those of thrum morphs; in the former, the cuticle mark of cleistogamy is the production of closed covering the papillae is wavy, continuous, and (cleistogamous) flowers, which self-pollinate in the thicker than in the thrum papillae, where it is bud and chasmogamous flowers, which open nor- highly irregular and nonuniform. A secondary mally and outcross. Dimorphic flowers of plants effect of cuticular modification in the thrum morph like Collomia grandiflora (Polemoniaceae) show sig- is the accumulation of secretion products on the nificant differences in pollen exine characteristics, surface of the papillae, making the stigma a wet structure of stigma papillae, and style lengths, very type (Dulberger 1974, 1987; Ghosh and Shivanna much as distylous flowers do. However, unlike in 1980a, b). Intermorph differences between exine distylous plants, intermorph crosses between cleis- ornamentation on pollen grains and the thickness togamous and chasmogamous flowers are incom- of the papillar cuticle have also been described in patible (Lord and Eckard 1984, 1986). Since no several members of Plumbaginaceae (Baker 1966; genetic or other control mechanisms are known to Dulberger 1975). No differences are seen between
Developmental Biology of Incompatibility
the cellular organization of the stigmas of the pin and thrum morphs of Primula vulgaris (Y. HeslopHarrison, Heslop-Harrison and Shivanna 1981). In P. obconica, however, intermorph differences are seen in the stigmatic papillae, which are small, smooth, and dry with a thin layer of exudate in the thrum morph, and large, uneven, and wet with copious exudate in the pin morph; aspects of dimorphism even extend to the nature of the enzymes localized in the exudates of the two types of stigmas (Schou 1984; Schou and Mattsson 1985). Besides obvious differences in the structure of the stigmatic papillae, morph differences in the stigma length have been reported in certain distylous genera of Rubiaceae (Murray 1990). Physiological differences between pollen grains and stigmatic surfaces of floral morphs are important in controlling incompatibility responses of distylous plants. Since such differences are less obvious than the readily detected morphological and structural differences, they have not been subjected to critical investigations. Most of the efforts to identify physiological differences have taken place since the 1940s. Lewis (1943) called attention to the existence of a difference in osmolarity between the stigmatic cells and pollen grains of L. grandiflorum and suggested that this apparently accounts for the incompatibility response. It is generally observed that pollen grains of the pin morph do not absorb water from the papillae nor swell on the pin stigma due to low osmolarity of the pollen grain relative to that of the stigmatic cells. In contrast, thrum pollen grains imbibe water readily on the pin stigma due to their lower water potential compared to pin pollen. Pollen water economy appears to be a factor in intermorph differences in P. vulgaris also. This is indicated by a requirement for controlled hydration for normal germination and pollen tube growth of thrum pollen, whereas pin pollen is little affected by atmospheric conditions. The two types of pollen grains also display marked differences in their capacity to imbibe atmospheric moisture and swell, pin pollen being slower than thrum pollen in this respect. In a related way, the ability of thrum pollen to imbibe water is lost when surface materials of the exine are leached by organic solvents; modifying the surface of pollen grains of the pin morph does not interfere with the hydration process. Since absorption of water is the initial preparatory process for crossing the incompatibility barrier, membrane functions appear to be at the hub of the physiological processes controlling pollen acceptance or rejection reactions (Shivanna, Heslop-Harrison and Heslop-
247 Harrison 1983). In L. grandiflorum, the stigma of the thrum morph is characterized by an activity of nonspecific esterases and acid phosphatases higher than that of pin morph. Both morphs also display differences in the stigmatic protein profile (Ghosh and Shivanna 1980a, b). However, differences in lipid components are nonexistent between the two types of flowers in Forsythia intermedia (Dumas 1977a). The most obvious physiological manifestation of dimorphism is in the external character of the stigma itself, but this has not lent itself to unequivocal conclusions. For example, in P. obconica, the thrum stigma is dry and the pin stigma is wet (Schou 1984), but the opposite is the condition in several species of Linum (Dulberger 1987). (ii) Tristyly
Analysis of tristyly as an outbreeding mechanism has been pursued in a few members of Oxalidaceae, Pontederiaceae, and Lythraceae. The distinctive feature of tristyly is the presence of three floral morphs, namely, long-styled, mid-styled, and short-styled. The attachment of stamens of two different lengths in each morph, corresponding to the lengths of the styles in the other two morphs, adds to the complexity of the incompatibility system in tristylous plants. Thus, long-styled morphs have mid-length and short stamens, mid-styled have long and short stamens, and short-styled have long and mid-length stamens. In terms of compatibility relations, the general principle is that pollinations are successful between anthers and stigmas situated at an equivalent level in the different floral morphs (Figure 9.2). For example, in a long-styled morph, successful pollinations are feasible between the pollen of short stamens and the style of a shortstyled morph, and between the pollen of midlength stamens and the style of a mid-style morph (Devi 1964; Ornduff 1964, 1966; Dulberger 1970; Barrett 1977; Price and Barrett 1982; Glover and Barrett 1983; Richards and Barrett 1987; Scribailo and Barrett 1991a). Tristyly in Lythrum junceum (Lythraceae)shows certain similarities to distyly, such as heteromorphism in the size of pollen grains and stigmatic papillae. A correlation is found between long-level anthers and large pollen grains, short-level anthers and small pollen grains, and mid-level anthers and pollen grains of intermediate size. Similarly, papillae of long-level stigmas are significantly longer than those of the mid-level and short-level stigmas (Dulberger 1970). Although long-level anthers of Pontederia cordata, P. rotundifolia, and P. sagittata
248
Pollination and Fertilization
and Barrett 1882; Glover and Barrett 1983). It would seem that differences in pollen and stigma morphology also function to prevent self-pollination in tristylous plants. (iii) Genetics of Heteromorphic Self-Incompatibility
Mid
Because heteromorphic characters of stamens and stigmas are involved in the incompatibility syndrome, heterostyly is clearly a sporophytically controlled phenomenon. Compared to homomorphic systems, the mechanism of pollen rejection at the genetic level is in heteromorphic systems coupled to differences in floral morphology to accentuate the outbreeding potential of the plant. De Nettancourt (1977) and Lewis and Jones (1992) have described in some detail the genetics and inheritance of self-incompatibility in heteromorphic systems, and the following is a summary of this information. The genetic basis of distyly is governed by the S-gene complex with two alleles, S and s. The dominant allele is found in the shortstyled plants in a heterozygous state (Ss), whereas the s-allele is present in a homozygous state (ss) in long-styled plants. From this it follows that a compatible Ss X ss cross will break down into almost equal numbers of long-styled and short-styled morphs. There is also convincing evidence from genetic analysis of rare crosses showing that the Slocus harbors a complex gene with at least five, and probably more, linked genes, one each for stylar incompatibility, pollen incompatibility, length of the style, pollen size, and anther level. It has been suggested that the action of the S-gene is controlled and modified by a number of unrelated genes in addition to the linked genes. This represents the best information that we have on the genetics of distyly, but the precise way these genes are integrated in controlling the incompatibility phenomenon remains to be elucidated.
Our knowledge of the genetic control of tristyly is based on the work done on Lythrum and Oxalis. It has been reasoned that the action of a single gene Figure 9.2. (a) Diagrammatic representation of tristyly. locus cannot account for the inheritance of three >, compatible; — > , incompatible. (From Shivanna 1982. In Experimental Embryology of Vascular Plants, ed. B. variable sets of characters, and this has been found M. Johri, pp. 131-174. Berlin: Springer-Verlag. © Springerto be the case. Two loci, S and M, with a dominant Verlag, Berlin), (b-d) Long-styled, mid-styled and shortand recessive allele in each locus have been implistyled morphs, respectively, of Lythrum junceum. (From cated in the inheritance of tristyly. To explain the Dulberger 1970.) segregation patterns obtained in crosses between (Pontederiaceae) produce larger pollen grains than compatible and incompatible partners, it is mid- or short-level anthers, pollen production assumed that the dominant S-allele is epistatic to M tends to be inversely correlated with anther level, and is found in all short-styled morphs. Longwith short-level anthers yielding more pollen than styled morphs are homozygous recessive for both anthers located at higher levels (Barrett 1977; Price genes, whereas mid-style morphs carry homozyShort
Developmental Biology of Incompatibility
249
gous recessive s and either homozygous or het- ferences in the aperture wall or in its functioning at erozygous dominant M. Claims have been made hydration are observed between pin and thrum that either the S or the M gene is composed of sev- pollen grains (Dulberger 1989). In the search for an eral subunits, but these claims have not been sub- explanation for this anomaly, attention has been stantiated. A study of the developmental basis of called to the fact that whereas the thrum stigma of tristyly in Eichhornia paniculata (Pontederiaceae) L. grandiflorum is of the wet type, with a disrupted has indicated that gene action does not control cuticle-pellicle layer and viscous stigmatic exudate, flower initiation or the development of floral the pin stigma is close to the dry type (Ghosh and organs in the different morphs, but first becomes Shivanna 1980b). Thus, compared to the pin stigma, evident in the premeiotic anthers (Richards and the thrum stigma could provide just the right subBarrett 1984); in Pontederia cordata, the genes appearstratum to support adhesion of both thrum and pin to regulate the relative growth rates of the stamens pollen grains regardless of their surface properties. and styles by differential effects on cell division To test for a possible surface tension effect, pin stigand cell elongation (Richards and Barrett 1987). mas were wetted and, not surprisingly, a good From these observations, it appears that the genes number of self-pollen were found to adhere to the controlling tristyly do not act through common stigma and to germinate; pollen tubes, however, developmental pathways to regulate the mature succumbed to inhibition within the stigmatic tissue itself (Ghosh and Shivanna 1982a). Even if pollen floral structures. adhesion depends on the state of the stigma, it may be just one factor in the causal complex - essential, (iv) Physiology of Heteromorphic but not sufficient, for overcoming the incompatibilSelf-Incompatibility ity barrier. Other possible explanations for pollen adhesion are also worth exploring: age of the An implicit assumption in most investigations of stigma, electrostatic and electrodynamic forces, heteromorphic incompatibility is that the rejection gravity and receptor molecules (Woittiez and reactions are initiated on the stigma or in the style. Willemse 1979; Dumas and Gaude 1981). It was A related factor that seems to have been ignored is seen in Chapter 7 that surface components of the the nature of the chemical system that underlies the pollen grains clearly have much to do with pollen rejection reaction. According to our current knowladhesion on the stigma during compatible pollinaedge, incompatibility may arise due to the failure of tion. Apparently, these compounds are also imporpollen grains to adhere to the stigma, the failure of tant in the adhesion of pollen in incompatible pollen germination, or the failure of pollen tube pollinations. As shown by Ghosh and Shivanna growth in the style. It is difficult to make a realistic (1982a), treatment of pollen grains of the pin morph assessment of the effectiveness of each of these of L. grandiflorum with nonpolar, anhydrous solimpediments to self-pollination when they do not vents such as hexane or acetone increases their abilfunction in the same species. ity to adhere to the same morph stigma. It is not Since pollen grains must adhere to the stigma in known whether removal of surface components order to germinate under natural conditions, it is causes the pollen grain to interact with recognition no wonder that the phenomenon of pollen adhe- substances of the stigma or whether the pollen sion following inter- and intramorph pollinations adheres passively due to other forces. has been intensively studied. Evidence that pollen grains do not adhere to the stigma after intramorph Whatever may be the basis of the pollen adhepollinations has been provided in Linum grandiflo- sion process for the failure of intramorph pollinarum (Dickinson and Lewis 1974), some species of tions, it is generally accepted that the impairment Limonium, and in Armeria maritima (both of of pollen germination and pollen tube growth is a Plumbaginaceae) (Baker 1966; Dulberger 1975). The causal factor in incompatibility. This comes as no importance of the stigmatic surface in controlling surprise because, compared to the physical forces pollen adhesion is dramatically demonstrated in L. that determine pollen adhesion, both pollen grandiflorum, in which pin pollen grains do not germination and pollen tube growth involve easily adhere easily to pin stigmas, but both pin and identifiable morphogenetic changes. Linum grandithrum pollen court thrum stigma. Even the physi- florum, in which thrum pollen lodges very easily on ology of a few pin pollen grains that accidentally the thrum stigma, will again serve as the first adhere to the pin stigma remains unchanged, with example. Although pollen grains germinate and neither swelling nor degradation of the exine produce tubes, the latter drastically change their occurring. This is surprising, since, except for the course in the incompatible style as they swell and presence of morph-specific exine processes, no dif- burst at the tip. The compatible tubes from the
250 alternate morph, in contrast, grow straight down in the style to their destination in the ovary (Lewis 1943). In an experiment wherein massive germination of self-pollen easily occurred on the thrum morph stigma of Luculia gratissima (Rubiaceae), it was possible to localize the incompatibility reaction zone in the stigma, because amputation of the stigma and upper part of the style eliminated the incompatibility reaction (Murray 1990). In contrast, in Limonium vulgare, incompatibility relies on the failure of pollen grains to germinate or of the pollen tubes to enter the stigma (Baker 1966). This rules out any possibility of the effect of stylar tissues in promoting incompatibility. The perennial herb Primula has served as a superb material for study of the efficacy of pollen germination and the fate of pollen tubes during intra- and intermorph pollinations because of the transparency of the style and the small size, hardiness, and availability of the popular garden plant in many parts of the world. Although there are some conflicting reports on the fate of the pollen tube following intramorph pollinations, it is clear that the behavior of the pollen grain varies between the morphs. According to Lewis (1942), differential growth of pollen tubes characterizes intramorph pin and thrum pollinations in P. obconica and P. sinensis; the growth of the thrum pollen tube in the style of the same morph is poorer than in the pin X pin cross. In P. obconica, other reports have claimed the existence of rigorous incompatibility reactions characterized by the lack of pollen tube penetration of the stigma after pin and thrum selfpollinations, despite respectable levels of germination (Dowrick 1956; Stevens and Murray 1982). This result leaves no doubt that the expression of incompatibility occurs after germination. The breeding behavior of P. veris illustrates yet another aspect of the incompatibility phenomenon in this genus. Production of short, broad pollen tubes, coupled with their inability to make inroads into the style, accounts for incompatibility in pin X pin pollinations; in thrum X thrum pollinations, pollen germination is very low, and the pollen tubes are so unusually short and sinuous that it is doubtful that they could ever cross the stigmatic papillae and grow (Richards and Ibrahim 1982). From this it appears that, on the generic level, the progress of the incompatible pollen in Primula may be checked at several stages, such as during germination, during pollen tube penetration of the stigma, or during pollen tube growth in the style. Similar conclusions have been derived from a study of P. vulgaris; here a persuasive case has been made that incompatibil-
Pollination and Fertilization
ity results from the additive effects of inhibition at different levels rather than at a single barrier (Shivanna, Heslop-Harrison and Heslop-Harrison 1981). In an investigation of within-morph crosses involving 52 species of Primula, pollen tube growth inhibition has been identified to occur in the stigma, style, or ovary, although most often failure of pollen germination at the stigmatic surface results in efficient incompatibility (Wedderburn and Richards 1990). Following pin X pin pollination in Pulmonaria afftnis (Boraginaceae), a few pollen tubes make their way to the ovary wall but most end up in the style, as do pollen tubes from a thrum X thrum cross (Richards and Mitchell 1990). Inhibition of pollen tube growth in different regions of the stigma and style also occurs following incompatible pollinations of thrum and pin stigma with self-pollen in certain distylous species of Rubiaceae (Bawa and Beach 1983) and in Averrhoa carambola (Oxalidaceae) (Wong, Watanabe and Hinata 1994b). Impairment of pollen tube growth in the stigma or in the upper part of the style probably accounts for incompatibility in the tristylous species, as seen from the analysis of intra- and intermorph pollinations in Lythrum salicaria and L. junceum (Esser 1953; Dulberger 1970). In Pontederia sagittata and P. cor-
data, pollen tubes make their way into a considerable length of the style, even into the ovary, before rejection occurs (Glover and Barrett 1983; Anderson and Barrett 1986; Scribailo and Barrett 1991b). It is not possible to offer a general mechanism to explain the incompatibility behavior in heterostylous plants. A few ideas current in the literature are totally inadequate for more than a superficial insight into the factors involved in the rejection reactions. Unfortunately, there seems to have been no rigorous molecular studies on these intriguing systems. Since water uptake is an early event in pollen germination, some perception of incompatibility may be gained from systems in which the rejection reaction is due to the failure of germination. A good example is Linum grandiflorum, in which there is a large-scale failure of pollen adhesion following pin X pin pollination, or failure of swelling of the few pollen grains that adhere. As indicated earlier, the failure of pollen grains to absorb water has been attributed to their low osmolarity compared to the stigmatic cells. Following thrum X thrum pollination in this species, the pollen tube bursts in the tissues of the style, presumably due to its higher osmotic potential relative to that of the cells of the style (Lewis 1943). Although determina-
Developmental Biology of Incompatibility
251
tion of the relative osmotic potential of the pollen involvement of protein synthesis in the incompatigrain and stylar cell in L. grandiflorum supports a bility reactions. Suggesting a link between peroxirole for osmotic changes in pollen tube demise in dase activity and incompatibility response, an the style, the work has not been extended to other ultracytochemical study has revealed a marked accumulation of the enzyme in pin X pin pollinated species. styles of Primula acaulis and its absence in the crossAnother mechanism that has been proposed for incompatibility in heterostylous species is based on pollinated styles (Carraro, Lombardo and Gerola the differences in the distribution of the flavonols 1985). Differences reported in the soluble protein quercetin and rutin in the pollen grains of short- profiles of the styles of thrum and pin morphs of styled and long-styled plants of Forsythia intermedia Averrhoa carambola are too insignificant to account (Moewus 1950). Quercetin is found almost exclu- for inhibition of pollen tube growth following polsively in the pollen grains of pin morphs and rutin lination with the same-morph pollen (Wong, is found exclusively in the pollen grains of thrum Watanabe and Hinata 1994a). morphs; paradoxically, the specific enzymes that Based on a study of pollen-pistil interactions in degrade quercetin and rutin occur only in the styles Pontederia sagittata, it has been suggested that incomof the opposite morphs. Because flavonols inhibit patibility in heterostylous systems is governed more pollen germination, a model in which rutin of the by the interaction of heteromorphic characters than thrum pollen is destroyed by the enzyme present in by an active rejection mechanism based on molecuthe pin style and quercetin of the pin pollen is lar specificities. Compelling evidence comes from destroyed by the enzyme found in the thrum style, the observation that incompatibility in certain would assuredly explain successful compatible pollen-pistil combinations involving pollen grains intermorph pollinations. On the other hand, follow- smaller than the legitimate size class is accompanied ing incompatible pollinations, the flavonol by an attrition of pollen tubes that occurs due to inhibitors are not destroyed, due to lack of appro- their inability to grow the required distance to the priate enzymes, and the result is the failure of ovary. This, in turn, is attributed to the presence of pollen germination. Obviously, these effects of storage reserves in the pollen grain insufficient for flavonols could serve as effective biochemical the tube to reach the ovule or due to impairment of mechanisms to promote or impede pollen germina- an essential metabolic function (Scribailo and Barrett tion in vivo, but the hypothesis did not really take 1991b). off, so to speak, and has not been useful for studyThe problems relating to the physiology of ing the mechanism of heteromorphic incompatibil- incompatibility in heterostylous systems rest here ity for two reasons. One is the failure to identify the at the present time. Admittedly, we do not have two specific enzymes that inactivate quercetin and very much to say on the physiology of incompatirutin; second, and more importantly, is the reported bility in distylous species, and the basis of incomfailure to reproduce the results in populations of the patibility in tristylous species stands out as a species (Lewis 1954). conspicuous unknown in this whole area of limited Because of their enzymatic nature, proteins have understanding. Not only is heterostyly an imporemerged as natural candidates to stop the growth tant genetic adaptation, but an understanding of it of pollen tubes following intramorph pollinations. will no doubt inform us about aspects of the physA recurrent view is that inhibition of growth of the iology of the pollen grains, stigma, and style that incompatible pollen tube is somehow dependent are crucial in the incompatibility response. upon continued protein synthesis, for when protein synthesis in the style is inhibited by cycloheximide, inhibition of pollen tube growth is partially 2. HOMOMORPHIC SELF-INCOMPATIBILITY reversed (Ghosh and Shivanna 1982a). Proteins Homomorphic self-incompatibility is relatively present in the stylar extracts have also been shown more widespread than the heteromorphic type, to differentially inhibit in vitro growth of pollen and representative genera exhibiting homomortubes of pin and thrum morphs; pollen tubes of the phic systems are found in more than half of the same morph are more sensitive than those of the families of angiosperms, with prospects that many other morph to a stylar extract (Golynskaya, more will be identified (Brewbaker 1959). Whereas Bashkirova and Tomchuk 1976; Shivanna, Heslop- heteromorphic systems possess morphological Harrison and Heslop-Harrison 1981). In the mechanisms in the flower to discourage self-polliabsence of refined biochemical studies, these obser- nation, homomorphic plants are able to identify vations fall short of a completely critical test for the and reject self-pollen by genetic and physiological
252
Pollination and Fertilization
Sporophytic incompatibility Genotype expressed is that of pollen parent
S,S2 S,S2
Genotype of individual pollen grains (haploid)
Genotype of pistil (diploid)
S,S2 S,S2
s,s 2
if t? s,s ,s 2
Gametophytic incompatibility Genotype expressed is that of individual pollen grains
S, S2
Genotype of pistil Figure 9.3. Genetic control of self-incompatibility in gametophytic and sporophytic systems. In both systems, pollination is incompatible when the S-alleles (5,, S2, etc.) expressed in pollen grains and pistil match. In the gametophytic system, the phenotype of the pollen grain is determined by its haploid genotype, whereas in the sporophytic system, the pollen phenotype is determined by the diploid genotype of its parent. In the gametophytic system, both 5, and S2 pollen are inhibited in an S, S2 style; when there is no match of alleles (for example, 5, S2 pollen grains on an S3 S4 pistil), pollen tubes of both genotypes grow through the style to the embryo sac. The S2 pollen will also grow successfully through the S, S3 style. In the sporophytic system, a match of the alleles of the pollen parent with those
of the pistil (for example, S, S2 or S, S3) causes arrest of pollen germination on the stigma surface. When there is no match, as in an S3 S4 style with S, S2 pollen grain, the latter germinates and grows through the style to the embryo sac. The central panel applies only if the allele 5, is dominant to, or codominant with, 52 in the pollen and S, is dominant to, or codominant with, S3 in the style. If S3 is dominant to S, in the style, or S2 dominant to 5, in the pollen, the pollen from the S, S2 parent will be compatible. (From R. Bernatzky, M. A. Anderson and A. E. Clarke 1988. Molecular genetics of self-incompatibility in flowering plants. Devel. Genet. 9:1-12. © 1988. Reprinted by permission of Wiley-Liss Inc., a subsidiary of John Wiley & Sons Inc.)
means. In practical terms, it turns out that flowers produced by different plants of a species in the homomorphic series do not show any morphological variations and are of the same type. Perhaps manifestations of the incompatibility syndrome in flowers of the same morphology has made the homomorphic system more amenable to genetic and molecular analyses than the heteromorphic system. Because different mechanisms are at work, homomorphic self-incompatibility has been wisely divided into gametophytic and sporophytic types. In both types, the behavior of pollen grains that land on the stigma is plainly crucial, for the degree of incompatibility is determined by the extent of pollen germination and pollen tube growth. In gametophytic self-incompatibility, the haploid genotype of the pollen grain determines the pheno-
type for incompatibility, and fertilization is prevented if the S-allele of the pollen is identical to either S-allele present in the recipient female sporophytic tissue. In contrast, in sporophytic selfincompatibility, the outcome is determined by the alleles carried by the diploid, pollen-producing parent plant, rather than by the S-allele of the pollen itself; when there is no match between an allele of the pollen parent with that of the female sporophytic tissues, fertilization occurs (Figure 9.3). Although in-depth genetic studies of selfincompatibility in the majority of homomorphic systems are lacking, most species have been assigned to the gametophytic type by virtue of the cytological symptoms they display; in general, sporophytic self-incompatibility has been described only in three families: Brassicaceae, Asteraceae, and Convolvulaceae.
Developmental Biology of Incompatibility
It has been known for some time that gametophyric self-incompatibility is characteristic of species that shed pollen at the two-celled stage and in which pollen rejection mechanisms operate in the style. On the other hand, species that display sporophy tic selfincompatibility have three-celled pollen grains and operate pollen rejection at the stigmatic surface (Brewbaker 1957). An apparent correlation also exists between the genetics of self-incompatibility and the nature of the stigma; gametophytic selfincompatibility is invariably associated with species having wet stigmas, whereas dry stigmas are the norm for species with sporophytic self-incompatibility (Y. Heslop-Harrison and Shivanna 1977). With a few exceptions, these correlations have withstood the test of time and thus can be considered as solid generalizations to identify the two types of homomorphic self-incompatibility systems. (i) Structural Basis of Self-Incompatibility
253 Gorrez 1967), in which pollen tubes grow through the style to the ovary before growth is arrested, and Theobroma cacao (cacao; Sterculiaceae) (Cope 1962), in which the rejection reaction is activated only after the pollen tube discharges the gametes in the embryo sac. In the neotropical trees Chorisia chodatii, C. speciosa (Bombacaceae), Tabebuia caraiba, and T.
ochracea (Bignoniaceae), the late-acting genes act so late that both zygote and endosperm are formed in the selfed ovules before they abscise (Gibbs and Bianchi 1993). In the woody mimosoid legume Acacia retinodes, compatible and incompatible pollen tubes travel at similar rates through the style and the ovary to the ovules. However, the incompatible pollen tube does not grow into the ovule beyond the nucellus (Kenrick and Knox 1985; Kenrick, Kaul and Williams 1986). The suggestion from these works that nucellar arrest of the pollen tube is a more primitive form of gametophytic selfincompatibility than stylar arrest provides a good starting point for painstaking microscopical investigations of incompatibility rejection in a wide range of primitive plants.
There is considerable diversity of S-gene expressions in the gametophytic incompatibility systems, so it is impossible to describe the cytology of indiA unifying feature of flowers of many cultivars vidual species, or indeed the diversity that occurs in of Lilium is the presence of a hollow style rather some closely investigated families such as than the solid type. The cultivars are also predisSolanaceae. It is unequivocally accepted that in posed to several forms of self-incompatibility many species, following both compatible and which can result in the inhibition of pollen tube incompatible pollinations, the pollen grains germi- growth near the stigma, or halfway down the style, nate normally. Whereas pollen tubes grow swiftly or even near the ovary (Brock 1954; Ascher and through the stigma into the style after compatible Peloquin 1968; Ascher and Drewlow 1975). In an pollinations, tubes resulting from incompatible pol- assay designed to follow pollen tube growth in linations travel only a very short distance in the segments of the style of L. longiflorum, it was found style before they collapse (de Nettancourt 1977). that incompatible pollen tubes grow as efficiently Extreme cases of inhibition are seen in Oenothera as the compatible ones from the stigmatic end to organensis (Dickinson and Lawson 1975), Papaver the ovary or in the opposite direction. This was rhoeas (Lawrence 1975), Citrus varieties (Kahn and taken to indicate that the incompatibility reactions DeMason 1986), and in some members of Poaceae are not localized in any particular portion of the (Hayman 1956; Lundqvist 1956; Knox and Heslop- style, but are due to interactions between pollen Harrison 1971c) and Commelinaceae (Owens 1981). tubes and the stylar tissue (Ascher 1977). In L. longiIn these plants, pollen grains germinate normally, florum, growth of pollen tubes in incompatible but pollen tube growth is arrested after the tube crosses is also related to the age of the style and, penetrates the surface layers of the stigma; in a generally, incompatibility is strongest following sense, these can be considered as hybrids between anthesis (Ascher and Peloquin 1966a). gametophytic and sporophytic systems, since the A fundamental feature of sporophytic selfsite of pollen growth inhibition is in the stigma, but incompatibility response as exemplified by memrecognition is under gametophytic control. At the bers of Brassicaceae and Asteraceae is the profound other end of the spectrum are the so-called "late-act- effect of the stigmatic cells on self-pollen. This is ing self-incompatibility" systems, in which rejection manifest either in the failure of pollen germination of the selfed flowers occurs after pollen tubes have on the stigma or in the failure of the short pollen entered the ovary (Seavey and Bawa 1986). tubes produced to grow beyond the barriers of the Included in this list are a few unrelated genera such stigma (Christ 1959; Kroh 1966; Kanno and Hinata as Hemerocallis (Stout and Chandler 1933), Ananas 1969; Knox 1973; Dickinson and Lewis 1973a). In (Bromeliaceae), and Gasteria (Brewbaker and Brassica oleracea, events prior to germination, such
254
Pollination and Fertilization
Figure 9.4. Fluorescence micrographs of aniline with dispersed granules in a papilla, (d) Same as in (c), blue-stained stigmatic papillae of Raphanus sativus show- focused to show callose in the plasmalemma. Scales bars = ing callose formation following an incompatible pollina100 urn (a) and 10 urn (b, c, d). (From Shivanna, Hesloption, (a) Section of a portion of the stigma showing callose Harrison and Heslop-Harrison 1978; photographs supplied formed at the plasmalemma of each papilla, (b) Formation by Dr. K. R. Shivanna.) of annular callose in a papilla, (c) Diffuse callose reaction
as adhesion and hydration of pollen grains and Callose Deposition. One aspect of S-gene erosion of pollen cell wall components, seem to be expression peculiar to sporophytic self-incompatiimportant in the rejection of self-pollen. Since self- bility systems is the reaction of the stigmatic papilpollen grains initially bind to the stigma less firmly lae following self-pollination (Figure 9.4). When than cross-pollen, there is reason to believe that self-pollen grains make contact with the stigma, the components that bind the pollen to the stigma are cuticle of the papilla is eroded at the same time that differentially altered during self- and cross-pollina- the external layer of the cell is digested. This is a tions (Stead, Roberts and Dickinson 1979). An early prelude to the deposition of callose at the tip of the scanning electron microscopic study claimed that papilla between the cell wall and the plasma memin B. oleracea, the incompatibility barrier is a waxy brane (Dickinson and Lewis 1973a; Knox 1973; J. layer on the stigmatic surface on which pollen Heslop-Harrison, Knox and Heslop-Harrison 1974; grains fail to adhere and germinate. Adhesion and Vithanage and Knox 1977; Shivanna, Heslopgermination are feats easily accomplished by the Harrison and Heslop-Harrison 1978). Callose compatible pollen grains whose tubes penetrate the deposition appears to be a Ca2+-dependent process cuticle (Roggen 1972). However, once self-pollen in self-pollinated Brassica oleracea because it is abolgrains stick to the stigmatic cells, they germinate, ished by deprivation of the ion by Ca2+ channel and pollen tubes that penetrate the base of the antagonists and induced by Ca2+ ionophore. stigma have an equal chance as cross-pollen in However, germination of self-pollen was preeffecting fertilization (Ockendon and Gates 1975). vented, and incompatibility prevailed when callose Irrespective of the extent of pollen adhesion on formation was abolished by treatment of stigmatic the stigma, some rejection signal is transmitted papillae with 2-deoxy-2-glucose. This observation from the stigmatic cells to the pollen grains. seems to fit rather well with the view that callose Symptoms of self-rejection are seen in the physical formation in the stigma is not a requirement for the appearance of the pollen grains on the stigma and demise of self-pollen or for incompatibility funcin the loss of their cellular integrity. Pollen grains of tions (Singh and Paolillo 1990; Elleman and Brassica oleracea subject to a self-incompatibilityDickinson 1996). episode at the stigmatic surface have higher levels It has been claimed that the callose response is of hydrolases (Dhaliwal and Malik 1982) and mem- not pollen specific (Sood, Prabha and Gupta 1982); brane-bound Ca2+ (Singh, Perdue and Paolillo 1989) however, a number of observations made on B. olerthan cross-pollen. Rejection signals evoke permis- acea, such as the induction of callose response on the sive conditions within the confining membrane of stigmatic papillae by a diffusate of self-pollen but the pollen grain that sequester enzyme molecules not by that of compatible pollen, activity of the difand Ca2+ ions, although this can hardly be consid- fusate at low concentrations, and the rapidity of the ered as the primary cause of pollen rejection. response, among others, seem to indicate that the
Developmental Biology of Incompatibility
callose plug formation on the stigmatic papillae is a dependable phenotypic expression of incompatible pollination (Kerhoas, Knox and Dumas 1983). The importance of callose deposition in incompatibility mechanisms is dramatically demonstrated in the growth of pollen tubes following self-pollination in gametophytic systems (Figure 9.5). In many plants, callose deposition at regular intervals is a normal way of life for pollen tubes as they grow following a compatible mating. The suspected specific function of callose in the pollen tube is to isolate the cytoplasm harboring the sperm cells and the vegetative cell nucleus from the empty pollen grain. The distinctiveness of the incompatible pollen tube is the irregular nature of the callose deposits, often leading to the plugging of the entire tip of the pollen tube. Some pollen tubes accumulate so much callose that the tips swell until they burst to death within the style (Tupy 1959; Schlosser 1961; de Nettancourt et al 1974; J. HeslopHarrison, Knox and Heslop-Harrison 1974; Sastri and Shivanna 1979; Vithanage, Gleeson and Clarke 1980); in others, callose is deposited even inside the pollen grain, beginning at what appears to be the site of pollen tube emergence (FranklinTong and Franklin 1992). Ultrastructural Cytology. Our understanding of the structural basis of incompatibility reactions is linked almost exclusively to electron microscopic investigations of pollen tubes following self- and cross-pollinations. The foundation for fine structural study was laid by van der Pluijm and Linskens (1966), who noted that incompatible pollen tubes of Petunia hybrida have thickened walls and their organelles and cytoplasmic contents suffer a loss of identifiable form. A follow-up work showed that there is a scarcity of smooth tubular ER and polysomes in the incompatible pollen tubes; in addition, fibrous masses found in the nuclear zone of the compatible pollen tubes are absent in the corresponding region of the incompatible tubes. These observations led to the suggestion that incompatibility is probably due to the secretion and production of membrane components, synthesis of growth hormones, decreased protein synthesis, and callose accumulation (Cresti et al 1979). The anarchic ultrastructural changes seen in the incompatible pollen tubes are not surprising in view of the rapid rejection reactions that set in following self-pollination. Herrero and Dickinson (1981) confirmed many of these findings and showed that although both compatible and incompatible tubes of P. hybrida begin to grow
255
rapidly upon entering the style, the growth of incompatible tubes slow downs. This might signify a metabolic imbalance in the incompatible pollen tube that leads to a decreased turgor pressure, rather than reflect a deposition of callose, as is generally assumed. There are no major ultrastructural differences during the activation and germination of compatible and incompatible pollen grains of Lycopersicon peruvianum (Cresti, Ciampolini and Sarfatti 1980). However, modifications of the pollen tube tip remain an essential baseline for interpreting the demise of germinating pollen grains following incompatible pollination. Crucial changes in the tip of the incompatible pollen tube during its journey through the style are the appearance of numerous granules in the cytoplasm, configuration of the rough ER in the form of whorls of concentric parallel membranes, disappearance of the inner wall, thickening of the outer wall, and lysis of the pollen tube. Incompatibility reactions that have been proposed as causal for the growth arrest of the pollen tubes are cessation of protein synthesis and the binding of an incompatibility protein to some constituents of the inner wall to produce particles that accumulate in the cytoplasm (de Nettancourt et al 1973a, 1974). Two brief investigations on Lilium longiflorum seem to reflect the differences in metabolism and synthesis of nucleic acids and proteins by pollen tubes following self-compatible and selfincompatible pollinations and suggest that these changes may possibly influence pollen tube wall formation as well as pollen tube growth (Campbell and Ascher 1975; Li and Linskens 1983a). The results clearly show that even though light microscopic work led to the belief that callose formation is a prelude to the commencement of incompatibility reaction, it is not the primary event that causes the arrest of pollen tube growth. Be that as it may, callose formation may well be necessary for the continuation of the rejection reaction. In grasses, it has been found that a thickening of the pollen tube tip by the apposition of microfibrillar pectic materials is the earliest morphological abnormality in the incompatible pollen tube and is initiated after the contact of the pollen grain with the stigma. This may be due to a dislocation of the normal pattern of wall growth, which involves the precise cross-linking of microfibrils in the emerging pollen tube. Callose is initially deposited behind the pollen tube and later extends even to its tip (Shivanna, Heslop-Harrison and Heslop-Harrison 1982). According to Geitmann et al (1995), during the incompatibility reaction in Brugmansia suave-
256
Figure 9.5. Callose formation during self-incompatibility in Caudinia fragilis (Poaceae). (a) Stigma surface 2 hours after a compatible pollination, showing pollen tubes with only occasional callose. (b) Stigma surface 2 hours after self-pollination, showing short pollen tubes occluded with callose deposition. Both (a) and (b) are fluorescence micrographs of sections stained with aniline blue, (c) Photograph of an inhibited pollen grain and pollen tube following an incom-
Pollination and Fertilization
patible pollination. Numbers 6 and 7 refer to planes of sectioning made in another study, (d) Fluorescence micrograph of the same field as in (c), stained with aniline blue, showing massive callose (c) in the pollen grain and pollen tube. The apical 12 urn of the tube (Pe) was identified as pectic in nature. Scale bars = 100 urn (a, b) and 10 u (c, d). (From Shivanna, Heslop-Harrison and Heslop-Harrison 1982; photographs supplied by Dr. K. R. Shivanna.)
Developmental Biology of Incompatibility
olens it is possible to disentangle effects of the Sgene on the processing of cell wall materials from those on the physical properties of the pollen tube wall. In this work, pectin synthesis detected by monoclonal antibodies was not diminished in pollen tubes after incompatible pollination; on this count, arrest of pollen tube growth might be attributed to the inhibition of both the transport and the incorporation of pectic vesicles into the wall. (ii) Genetics of Self-Incompatibility Current concepts on the inheritance of the selfrejection process in homomorphic systems have evolved mainly from the studies of Bateman (1952, 1954, 1955) and Lundqvist (1975). The popular genetic view of self-incompatibility is that in most plants it is governed by a single S-locus (monofactorial), whereas in a minority of species various allelic or intergeneric interactions occur that result in the control by two or more loci (bi-, tri-, or tetrafactorial). As alluded to earlier, the genetic basis of self-incompatibility is that any pollen that has an S-allele that does not match with the one present in the female tissue may cause successful fertilization. The monofactorial gametophytic self-incompatibility system is common in species belonging to Solanaceae and Fabaceae. One fascinating aspect of plants exhibiting this type of incompatibility is the high number of different alleles that operate at the S-locus. For example, as many as 150 to 250 different alleles have been reported in Trifolium sp. Studies in grasses have shown that self-incompatibility is controlled by two independent loci, S and Z, each with allelic series of considerable size. The two loci behave independently, so that full selfincompatibility is expressed when alleles matching both loci are present in the pollen and the pistil. Complex bifactorial incompatibility relations have been described in certain members of Solanaceae, for example, Physalis ixocarpa, Solarium pinnatisectum, S. phureja, and S. stentomum (de Nettancourt
1977). There is also evidence favoring the involvement of three or even four S-loci in self-incompatibility in certain Ranunculaceae (0sterbye 1975), Chenopodiaceae (Larsen 1977), and Fabaceae (Lundqvist 1993). These examples are potentially of great evolutionary significance, because their confirmation in a great number of species will provide evidence of early ancestry of polyfactorial gametophytic self-incompatibility. The genetics of sporophytic self-incompatibility
257
has been subject to less intensive study, but it also seems to follow the same principle as the gametophytic system. Sporophytic self-incompatibility governed by a single genetic locus with two alleles is found in Capsella grandiflora and Cardamine pratensis (Brassicaceae); polyallelic control at a single locus is found in Parthenium argentatum (Asteraceae), Crepis foetida, Cosmos bipinnatus, and
several members of Brassicaceae (de Nettancourt 1977). The incompatibility system in Eruca sativa appears to be controlled by at least three loci, of which two are analogous to the S and Z loci of grasses (Verma, Malik and Dhir 1977). An unusual genetic mechanism apparently operates in the self-incompatibility of Theobroma cacao, which displays characteristics of both gametophytic and sporophytic systems (Cope 1962). Despite the discharge of male gametes in the embryo sac, genetic recombination is hampered in varying percentages of ovules in an ovary. This is attributed to the action of the S-gene both before and after meiosis in the anther and ovule. An additional complexity is due to the presence of accessory loci, A and B, which provide nonspecific precursors to which Salleles impart their specificities. Observations on the frequency of distribution of ovules in which fertilization has not occurred (nonfusion ovules) in nature and in various crosses have provided genetic evidence for sporophytic and gametophytic control of self-pollen rejection in this system. A prevalent hypothesis to explain the evolutionary switches between sporophytic and gametophytic incompatibility systems makes the not unreasonable assumption that a difference in the timing of gene action is at the crux of the differences between the two systems. This hypothesis, advocated by Pandey (1958, 1970), argues that in the sporophytic systems, the S-gene is activated before or during microsporocyte meiosis. This would ensure that the incompatibility substances synthesized during the diplophase are distributed equally among the products of meiosis. By the same token, if gene action in the gametophytic system is assured to occur immediately after meiosis, it will allow an accumulation of products of one Sallele in one pair of microspores and products of the other S-allele in the other pair. However, it is well known that in certain species showing sporophytic incompatibility, synthesis of incompatibility substances such as pollenkitt and tryphine takes place in the tapetum and the substances are subsequently transferred to the pollen exine; these facts are certainly damaging to the timing hypothesis of
258 S-gene action in so far as it relates to sporophytic incompatibility. A few comments about the organization of the Slocus are also relevant to understanding the functioning of the self-incompatibility gene. A model that can claim some validity is attributed to Lewis (1949, 1960), who proposed that the S-gene is composed of three closely linked parts. It has been reasoned that in order for the alleles of the same gene to be effective in both the pollen and the style for the incompatibility reaction, allelic specificity of both is determined by a common segment of the gene; accordingly, the first segment of the gene, referred to as the specificity part, is assumed to carry genetic information affecting pollen tube growth in the style following an incompatible cross. The second and third segments of the gene apparently determine the allelic specificity of the pollen and style, respectively. Although there are a number of lines of evidence to support this model, the most direct one is based on mutants that affect the pollen part without any effect on stylar behavior and that affect the style without interfering with the pollen activity part. Several questions about the molecular structure of the S-locus beg resolution, the most important one being the number of functional genes in the S-locus. Interpretations favoring the involvement of one, two, or even three genes to effect S-allele-specific recognition following incompatible crosses have been proposed, but hard evidence in support of them is scanty (Singh and Kao 1992).
(iii) Physiology and Biochemistry of SelfIncompatibility Under this heading, several aspects need to be considered. First, what are the specific physiological changes triggered by self-compatible and incompatible pollinations? What is the nature of the S-allele products and where are they synthesized? What is the biochemical basis for self-incompatibility recognition? The events alluded to in these questions mark the early signs of change in the stigma and style, and represent a basic feature of the homomorphic self-incompatibility system. Many investigators studying the physiology of cross- and self-pollination have looked at the postpollination reactions of the flower, especially those of the pistil. One view of the physiological changes in the flower following pollination is based on the principle of translocation of organic substances from the site of their synthesis (source) to the site of their utilization (sink). In an unpollinated flower,
Pollination and Fertilization
stamens and carpels serve as the sink for attracting metabolites derived from the vegetative parts of the plant as well as from the calyx and corolla. According to Linskens (1975), in the unpollinated flower of Petunia hybrida, the anthers function as the primary sink for 14C-protein hydrolyzate, with some label accumulating in the style also. Among pollination-induced changes is a strong influx of the label into the style and the ovary about 10 hours after self-pollination and a weak influx into the same organs of cross-pollinated flowers. After about 18 hours, the ovary continues to function as the sink in the cross-pollinated flower, whereas the self-pollinated flower suffers a major efflux of nutrients. Changes in phytic acid levels that occur in the flower can be considered in this context, since incompatible pollination leads to a more rapid decrease of phytic acid in the pistil than does compatible pollination (Jackson, Kamboj and Linskens 1983). Because the growth of the pollen tube in the self-pollinated flower stops about 12 hours after pollination, the efflux of nutrients after this time is not unexpected. Although significant metabolic changes are set in motion in the pistil following pollination, it has been difficult to identify specific differences between self- and cross-pollinated pistils in all instances. In general, the pistil of P. hybrida registers a dramatic increase in respiratory activity as the first of a cascade of changes following both compatible and incompatible matings, although the incompatibly pollinated pistil is unable to carry out normal respiration after about 12 hours (Linskens 1955). Some investigators have shown that an increase in CO2 concentration in the air effectively eliminates the incompatibility reaction in Brassica campestris by overcoming the barrier to pollen germination (Nakanishi, Esashi and Hinata 1969; Nakanishi and Hinata 1973; Dhaliwal, Malik and Singh 1981; O'Neill, Singh and Knox 1988). Although no clues are at hand to help us understand the effect of CO2 in this system in overcoming self-incompatibility, the possibility that the gas modifies the surface properties of the stigmatic papillae should not be discounted. There are some reports of changes in the protein and nucleic acid contents of the style following selfand cross-pollinations. In P. hybrida, following crosspollination, a single new glycoprotein fraction appears in the style. After selfing, however, two glycoprotein fractions with different mobilities are identified. Apparently, these glycoproteins are synthesized by contributions from both the pollen grains and the style (Linskens 1959). In this same
Developmental Biology of Incompatibility
259
species, there is a fall in the protein content of the Ultracytochemical analysis has shown that in P. style in the first 24 hours after both self- and cross- hybrida, there is a substantial increase in the number pollinations, but the fall is more marked in the selfed of peroxidase-containing cells in the transmitting tisstyles. The decrease in protein content does not sue of self-pollinated styles compared to that of result in a corresponding rise in the free amino acid unpollinated and cross-pollinated styles (Carraro, pool, probably because the amino acids released Lombardo and Gerola, 1986). from protein breakdown are used in the respiratory We have also some information on the relationpathway (Linskens and Tupy 1966). Somewhat sim- ship of isozymes of esterase to incompatibility. This ilar changes occur in the free amino acid pools of the is based on the observation that heat treatment of pistils of Nicotiana alata following self- and cross-pol- styles of self-incompatible genotypes of several linations (Tupy 1961). Differences noted in the elec- species of Nicotiana and Lilium longiflorum leads to trophoretically separated protein banding patterns the disappearance of certain isozyme bands while between prefertilization ovaries of self- and cross- others remain stable; the compatible genotypes pollinated Lotus corniculataflowerssuggest that bio- show no changes in the isozyme pattern (Pandey chemical self-recognition occurs in the ovary before 1973). Since heat treatment is known to eliminate fertilization (Dobrofsky and Grant 1980). incompatibility reactions in L. longiflorum, these One of the enzymes implicated in incompatibility observations support the concept of differential in a causal way is peroxidase. This assessment of the genetic control of the synthesis of a specific enzyme role of peroxidase was originally based on the work (Ascher and Peloquin 1966b; Dickinson, Moriarty of Pandey (1967), who showed that a specific perox- and Lawson 1982). Studies on P. hybrida have shown idase isozyme is associated with the S-allele of N. a correlation between the appearance of isozymes of alata. Since peroxidase is known to destroy IAA, it glutamate dehydrogenase and S-genotypes; activiwas logically thought that the enzyme might cause ties of several glucan hydrolases register an increase incompatibility by destroying IAA necessary for that is greater in cross-pollinated styles than in pollen tube growth. Although subsequent workers selfed ones (Roggen 1967; Linskens et al 1969). were unable to observe any apparent association Other investigations on P. hybrida have revealed between self-incompatibility and peroxidase activ- distinct differences in the synthesis of RNA and ity, persistent differences in enzyme activity in some proteins and in the size of the free nucleotide pool plants nonetheless occur between self- and cross- in the styles after cross- and self-pollinations. pollinated styles (Desborough and Peloquin 1968; Particularly striking is the observation that RNA Nasrallah, Barber and Wallace 1970; Nishio and synthesis is appreciably enhanced as early as 3 Hinata 1977; Bredemeijer and Blaas 1980). For exam- hours in cross-pollinated compared to self-polliple, after one day, self- and cross-pollinated pistils of nated styles; protein synthesis in the compatibly N. alata do not register a change in the total peroxi- pollinated styles also peaks earlier than in their dase activity compared to that of unpollinated pis- selfed counterparts (van der Donk 1974a, b; tils. In the next three days, there is a steady increase Deurenberg 1976). Electrophoretic analysis of proin the total enzyme activity, which is higher in cross- teins translated by polysomes isolated from selfpollinated styles than in self-pollinated ones. After and cross-pollinated pistils and injected into eggs four days, enzyme activity remains more or less con- of Xenopus laevis revealed some minor differences stant in self-pollinated styles but increases markedly in the protein bands as early as 6 hours after polliin cross-pollinated ones (Bredemeijer 1974). A posi- nation; the differences were accentuated by 24 tive correlation also exists between peroxidase hours (Deurenberg 1977). These synthetic activities isozyme-10 and the intensity of the incompatible are probably triggered by pollen recognition events reaction, indicating that peroxidase-10 is in some and therefore reflect differential gene expression way involved in pollen tube rejection. It has been concerned with pollen acceptance or rejection. proposed that after a compatible pollination, the Observations such as these are of course inadepollen tube grows so fast that its tip marauds those quate as anything more than a first step in impliparts of the style where peroxidase-10 is not present; cating nucleic acid and protein synthesis in the on the other hand, after an incompatible pollination, development of incompatibility reactions. because of the slow growth of the pollen tube, the tip Another useful first step in ascertaining the role is inextricably caught in the part of the style with a of nucleic acid and protein synthesis in self-incomhigh peroxidase-10 activity and, as a result, growth patibility reactions is the use of metabolic is inhibited (Bredemeijer and Blaas 1975). inhibitors. One general approach employed con-
260
sists of the injection of the inhibitor into the style during the period of development of incompatibility and the monitoring of pollen tube growth or seed set as a sign of overcoming incompatibility. Ascher (1971) found that injection of inhibitors of RNA synthesis such as acridine orange, actinomycin-D, or 6-methylpurine into the style of L. longiflorum before pollination appreciably stimulates the growth of incompatible pollen tubes. In P. hybrida, incompatibility is overcome by treating the buds with metabolic inhibitors, the most effective ones being oligomycin for RNA and cycloheximide for protein synthesis (Kovaleva, Milyaeva and Chailakhyan 1978). It has also been shown that in 8. oleracea, which exhibits sporophytic selfincompatibility, cycloheximide treatment of the stigma is in itself sufficient to suppress the selfincompatibility mechanism. The reversal of selfincompatibility by protein synthesis inhibition can be traced back to an effect of the stigmatic papillae on pollen hydration, which is probably the earliest physiological event after the self-pollen grains establish contact with the stigma (Roberts, Harrod and Dickinson 1984b; Sarker, Elleman and Dickinson 1988). These results suggest that the synthesis of a new RNA or protein is triggered in the stigma during the development of self-incompatibility and that these macromolecules are necessary for the incompatibility reactions to ensue. But synthesis of new macromolecules may also be simply the result of physiological changes induced in the stigma and may not have anything to do with self-incompatibility.
Pollination and Fertilization Table 9.1 Growth of self-incompatible and selfcompatible pollen tubes of Papaver rhoeas challenged with stigmatic extracts in the presence of metabolic inhibitors. (From Franklin-Tong, Lawrence and Franklin 1990.) MEAN POLLEN TUBE LENGTH*
Self (S,S2) pollen
Compatible pollen
S,S2 stigmatic extract alone
22.50
59.58
S ^ stigmatic extract + cycloheximide
5.58
5.75
S ^ stigmatic extract + tunicamycin
57.08
61.67
S,S2 stigmatic extract + actinomycin-D
48.33
62.50
Control, water only
82.08
68.33
Treatments
*Mean pollen tube length is expressed in units (1 unit = 32 urn).
for gene expression for the progress of incompatibility reactions, indicated by these results, was confirmed by in vitro translation of RNA isolated from pollen grains participating in an incompatible reaction; the results revealed the presence of novel proteins associated with the incompatibility response A second approach has utilized in vitro germina- (Franklin-Tong, Lawrence and Franklin 1990). A physiological observation that has some beartion of pollen grains and growth of pollen tubes in the presence of metabolic inhibitors and stigma ing on the molecular studies to be considered in extracts to monitor the involvement of nucleic acid Chapter 10 is that there is a pronounced increase in and protein synthesis in the incompatibility reaction. phosphorylation activity when pollen grains of rye Ferrari and Wallace (1976) found that a delayed are treated with self-stigma proteins. A correspondaddition of cycloheximide inhibits the germination ingly decreased phosphorylation activity is of pollen grains of B. oleracea with the same kinetics observed in self-fertile mutants; the self-incompatias the self-stigma extract. The inference from these bility reaction is also eliminated by protein kinase results, which suggest that incompatibility sub- inhibitors and Ca2+ antagonists applied to isolated stances in the self-stigma extract might interfere stigmas (Wehling, Hackauf and Wricke 1994). Since with protein synthesis, is also supported by the genes that encode for putative protein kinase observation that the stigma extract selectively receptors have been cloned from a sporophytic selfinhibits 14C-leucine incorporation into proteins. incompatibility system, it is of interest that the When pollen grains of Papaver rhoeas are grown in intracellular signal transduction pathway in a the presence of stigma extract from plants with the gametophytic system might also be mediated by a same S-genotype, inhibition of pollen tube growth protein kinase. characteristic of the incompatibility reaction is We now have a rough picture of the major physobserved; however, if pollen grains are challenged iological changes accompanying cross- and selfwith actinomycin-D and stigma extract, inhibition is pollinations in certain plants. Obviously these considerably alleviated (Table 9.1). A requirement changes occur subsequent to a fundamental recog-
Developmental Biology of Incompatibility
nition reaction involving pollen-generated molecules and those present in the pistil. The location and nature of these incompatibility factors have become topics of great interest in the context of cell-to-cell recognition, because an essential feature of such interactions is the involvement of specific proteins. Flowers show a range of incompatibility responses to self-pollen. Functions of the stigmatic cells change after anthesis due to extracellular and intracellular factors, and these changes are reflected in the flowers becoming self-compatible toward the end of the flowering season (Stout 1931; East 1934; Pandey 1960). The nature of the regulatory circuits that are activated to restore compatibility is not determined, although some evidence indicates that reversion to self-compatibility might be traced to cytological aberrations. This is the case in tetraploid clones of Petunia axillaris in which pollen grains are able to grow normally following self-pollination and to cause seed set, whereas diploid branches on the same plant remain strictly self-incompatible (Stout and Chandler 1941). Some mutants of P. inflata obtained by irradiation of a highly selfincompatible seed stock show an S-gene bearing a centric fragment in the genome that marks their transition to self-compatibility (Brewbaker and Natarajan 1960).
261 flowers (Clark et al 1990). The onset of self-incompatibility in flowers of Nicotiana alata is associated with the development of a pink color in the petals; buds with green petals remain self-compatible (Bernatzky, Anderson and Clarke 1988). In B. oleracea, Raphanus sativus,
and Cheiranthus
cheiri
(Brassicaceae), young buds are initially self-compatible; self-incompatibility is acquired late in floral development, generally about one to two days prior to anthesis (Shivanna, Heslop-Harrison and Heslop-Harrison 1978; Sood, Prabha and Gupta 1983). Several biochemical studies support a correlation between a shift to incompatibility and an increase in the amount and rate of synthesis of some proteins directly associated with a given Sallele specificity (Figure 9.6). This was first shown in B. oleracea, in which the onset of incompatibility in progressively older flowers was correlated with increasing concentrations of an S-allele antigen in the stigma (Nasrallah 1974). Somewhat similar results were obtained when glycoproteins from stigmas were analyzed by isoelectric focusing and SDS-PAGE combined with sensitive staining techniques (Roberts et al 1979; Nasrallah and Nasrallah 1984). Although excised stigmas of B. oleracea begin to synthesize S-glycoproteins before the selfincompatibility syndrome becomes functional, the It is now established that the acquisition of com- rate of synthesis dramatically peaks before anthepetence for self-pollen rejection is developmentally sis. The peak synthesis coincides with the switch of controlled, being strong in mature flowers and rel- the receptive surface of the stigma from one favoratively weak in immature flowers and unopened ing self-pollen to one programmed for complete buds. Evidence for this has come from immature incompatibility. Based on the continued synthesis bud pollinations, from biochemical analysis of of these glycoproteins, albeit at a low rate, and on changes of specific proteins, and by molecular the low rate of their turnover in stigmas of open hybridization. Studies on the development of the flowers, it has been suggested that the shift from self-pollen rejection mechanism were first con- compatibility to incompatibility occurs when a cerducted by bud pollination of Brassica oleracea (Attia tain threshold level of S-specific glycoproteins 1950) and P. axillaris (Shivanna and Rangaswamy accumulates (Nasrallah, Doney and Nasrallah 1969). As seen in the latter species, bud pollination 1985). S-genotype-specific proteins have been with self-pollen one or two days before anthesis identified in mature stigmas of self-incompatible significantly increases seed set but is less effective lines of B. campestris but are absent in the self-comat still earlier periods. Although the results with B. patible stigmas (O'Neill, Singh and Knox 1989). A oleracea were less clear-cut, elimination of incom- developmental regulation of S-allele-associated patibility due to bud pollination was attributed to proteins also prevails in the flowers of P. hybrida the absence of an inhibitor; later biochemical stud- (Herrero and Dickinson 1980) and Lycopersicon ies have shown that this was a prophetic interpre- peruvianum (Mau et al 1986; Rivers and Bernatzky tation. In P. hybrida, bud pollination is effective in 1994). Several proteins mobilized in the pistils of overcoming incompatibility even at earlier periods mature flowers after self-pollination are absent in in the ontogeny than in P. axillaris, because buds the pistils of immature buds of P. hybrida; in L, perupollinated three to four days before anthesis fail to vianum, extracts of the style of incompatible mature show the incompatibility reaction. Self-incompati- flowers and yellow buds contain higher concentrability is, however, unmistakably seen in buds pol- tions of S-glycoproteins than extracts of compatible linated one day prior to anthesis and in open immature buds (Figure 9.7).
262
Pollination and Fertilization
S.C.
•0- p
\
Oo 3.1
days ..
before anfhesis
Su Figure 9.6. Immunodiffusion (central panel) showing precipitation patterns of stigmatic antigen along the inflorescences of incompatible (left) and compatible (right) homozygous lines of Brassica oleracea. The diagram on the
left shows an incompatible inflorescence with self-compatible (S. C.) and self-incompatible (S. I.) zones; open circles, flowers; solid circles, buds 1, 2, and 3 days before anthesis. (From Nasrallah 1974.)
t B
— 30K
Figure 9.7. Development of S-allele-associated proteins in the style of Lycopersicon peruvianum (genotype S2 S3). The panel shows a segment of the SDS-PACE of extracts of styles of flowers at the different stages of development indicated at the top. Lanes 1 and 4, green bud (A); lanes 2 and 5, yellow
bud one day before opening (B); lanes 3 and 6, mature bud one day after opening (C). Lanes 1-3, extracts prepared from six styles; lanes 4-6, extracts with equal amounts of protein. (From Mau et al 1986. Planta 169:184-191. © SpringerVerlag, Berlin. Photographs supplied by Dr A. E. Clarke.)
The isolation and characterization of genes encoding components of the self-incompatibility response (Chapter 10) have led to the establishment of correlations between the accumulation of S-locus RNA and the acquisition of self-incompatibility.
This was first shown in B. oleracea by Northern blot hybridization of RNA from stigmas of different stages of development with a cDNA clone containing sequences specifying an S-locus-specific glycoprotein. The level of the transcript was found to
Developmental Biology of Incompatibility
263 Bud length
60 50
Style length
»- S1 mRNA
0.70
B
A
• ! ^ l 81
40 30
a s % |
S-2
0.25% l l f
^ ^ ^ —3
•
0.50 /
. .
_5 /
-6
0.031%;
0.40
/1
0.125%^ O.O62%| X
20
0.60 /
-7
-
•
^
6.20
10
0.10
,—-—— a—
-10
-9
—;
-8
0.30
1
1
-7
-6
V"***^ -5
r
i
i
- 4 - 3 - 2 - 1
i
0.00 0
1
Days Before Anthesis Figure 9.8. Developmental expression of S-locus mRNA in a self-incompatible homozygous line of Petunia hybrida. Inset: Total RNA isolated from floral buds of different stages or RNA transcribed in vitro from a cDNA clone (PS-IB) encoding the incompatibility proteins was hybridized with the same cDNA clone in a slot-blot apparatus. Lane A,
hybridization of serially diluted RNA transcribed in vitro by PS-IB; lane B, total RNA isolated from styles of different ages as indicated by days prior to anthesis on the right. (Clark et al 1990. Plant Cell 2:815-826. © American Society of Plant Physiologists.)
increase as a function of maturity of the stigma and to reach maximum level with the onset of the selfincompatibility response (Nasrallah et al 1985). This was confirmed in a subsequent work, which involved in situ hybridization of stigma sections with a labeled probe (Nasrallah, Yu and Nasrallah 1988). Both Northern blot and in situ hybridization methods have shown that the expression of a cDNA clone encoding S-allele protein of Nicotiana data correlates well with the development of incompatibility, transcripts for the gene being generally absent in the immature style, but appearing in the style of mature flowers (Anderson et al 1986; Cornish et al 1987). In P. hybrida (Figure 9.8), quantitative slot-blot assay revealed that the accumulation of S-locus mRNA parallels the acquisition of incompatibility, with a major accumulation of transcripts occurring during the transition from compatibility to incompatibility (Clark et al 1990). In Solarium tuberosum, the expression pattern of cloned mRNA transcripts fits with the functional requirement of providing the mature pistil with high concentrations of S-proteins (Kaufmann, Salamini and Thompson 1991). It has not been determined whether the relevant mRNAs that code for recognition molecules in the style and stigma are produced at an early stage in
floral ontogeny but remain masked until a late stage. It was mentioned earlier that in plants expressing self-incompatibility late in floral development, bud pollination results in the production of a full complement of seeds. Evidently, the components necessary for normal seed set are synthesized early in floral development and remain stable until anthesis, but the expression of incompatibility factors is confined to a narrow developmental window during floral ontogeny. Why do incompatibility factors remain unexpressed until late in the development of the flower? This is an intriguing question that cannot be answered at present. (iv) Pollen Recognition Factors
Rejection of self-pollen in an incompatible reaction is initiated following contact of pollen grains with the stigmatic cells of a flower carrying the same S-allele. This interaction tacitly assumes a role for recognition factors of pollen grain origin in the rejection process. In Brassica oleracea and other plants exhibiting sporophytic self-incompatibility, substances responsible for the rejection reaction are presumably synthesized in the tapetum and transferred to the cavities of the sculptured layer of the pollen exine. Because these substances can be eas-
264
Pollination and Fertilization
ily eluted from the exine, they have become prime et al 1995). These observations make it certain that candidates for the kind of rapid rejection reaction a pollen-stigma dialogue involving proteins from in the stigma following an incompatible pollina- both cell types is central to any hypothesis to tion. It was mentioned earlier that in most cases of explain self-incompatibility or self-compatibility in sporophytic incompatibility, self-pollinations lead Brassica sp. to the failure of pollen germination and to the formation of callose at the point of pollen contact with Gametophytic Systems. Considering the fact the stigmatic papillae. Using callose formation as a that the most common self-incompatibility system bioassay, J. Heslop-Harrison, Knox and Heslop- is of the gametophytic type, amazingly little is Harrison (1974) induced a rejection reaction on the known about the location of pollen recognition facstigma of Iberis sempervirens by deposition of an tors in this group of plants. The problem of identiagar film in which the exine proteins are collected. fying factors of pollen grain origin in the It was possible to duplicate the rejection reaction by gametophytic system is technically difficult apposing pieces of the intact tapetum isolated from because the rejection reaction results in the inhibiincompatible anthers to the stigma. The presence of tion of growth of the pollen tube in the style. A sugrecognition substances in the pollen exine and the gestion that the outflow of proteins into the role of the substances in the incompatibility reac- stigmatic papillae must be mainly from the intine tion have also been neatly demonstrated in another sites has come from the observation that pollen work. When anthers of Raphanus sativus were grains of grasses are generally impoverished of minced and centrifuged, a fraction enriched for the substances of tapetal origin in the exine (J. Hesloptryphine coating of the pollen was collected; Harrison 1979b). Moreover, inhibition of pollen tryphine thus extracted from incompatible pollen tube growth takes place at the stigmatic surface in and adsorbed on a membrane was able to induce a grasses. In Saccharum bengalense, the timing of the callose reaction on the stigma in a manner analo- rejection reaction is found to depend upon the way gous to that of an incompatible pollen tube. In con- pollen grains establish contact with the stigmatic trast, tryphine from the cross-pollen failed to elicit papillae. If pollen grains land with their germ pore the same reaction (Dickinson and Lewis 1973b). in contact with the stigmatic papillae, germination is inhibited and the germ pore is occluded with calEvidence for the role of exine proteins in the incompatibility reaction in Cosmos bipinnatus has lose. Pollen grains whose germ pore is away from come by another way. Here, self-incompatibility is the stigmatic papillae germinate and produce short partially overcome when flowers are pollinated fol- tubes whose growth is inhibited as soon as they lowing the application on the stigma of a diffusate come in contact with the papillae. Assuming that from the compatible pollen. Apparently the dif- the intine proteins are released when the germ pore fusate from the compatible pollen acts as a "men- is in contact with the papillae, these observations tor" in stimulating germination and growth of considerably bolster the role of intine proteins in self-pollen (Howlett et al 1975). As stated previ- the rejection reaction (Sastri and Shivanna 1979). In Oenothera organensis also, in which pollen tube ously, after pollen grains are deposited on the growth is stymied in the surface cell zone of the stigma, responses of the stigmatic cells and style stigma, an early release of the reactants lodged in are also crucial in the recognition reaction. That the the intine at the apertural site appears likely (J. cells of the stigma of B. oleracea store incompatibilHeslop-Harrison 1978). ity factors has been shown by the observation that a leachate of the stigma selectively inhibits germination of self-pollen without affecting cross-pollen; some inhibitors of pollen germination are not detected in unpollinated stigmas or in stigmas after a compatible pollination (Ferrari and Wallace 1975, 1976; Hodgkin and Lyon 1984). According to Doughty et al (1993), mixing crude stigmatic extract of B. oleracea with polypeptides extracted from pollen coats elicits the formation of an interaction product involving a 7-kDa pollen coat protein. Similar fusion proteins are also formed by interaction of pollen coat-borne peptides from B. napus with two stigmatic S-family glycoproteins from a compatible line of the same species (Hiscock
Information on the location of incompatibility factors in gametophytic systems in which the inhibition occurs in the style is no more solid than in those cases in which stigmatic surface inhibition is the rule. The incompatibility response in Lilium longiflorum is eliminated by treating the whole, or parts of, the style at 50 °C in a water bath for 5-6 minutes and pollinating the stigma by self-pollen (Hopper, Ascher and Peloquin 1967; Fett, Paxton and Dickinson 1976). Heating the half of the style with the stigma has little effect, indicating that the recognition substances are confined to the lower half of the style and that the stigma does not have any profound effects on the incompatibility reac-
Developmental Biology of Incompatibility
tions (Fett, Paxton and Dickinson 1976). A critical test of the biological role of the recognition factors is that they should inhibit pollen tube growth with S-genotype specificity. Consistent with the postulated location of incompatibility factors in the style of L. longiflorum is the observation that an extract from the incompatible style inhibits in vitro growth of pollen tubes and that heat treatment shown to inactivate self-incompatibility eliminates two glycoprotein bands from the electrophoretic profile of the stigmatic extract (Dickinson, Moriarty and Lawson 1982). Compared to pollen germination and pollen tube growth of compatible pollen grains of Petunia hybrida in an in vitro bioassay containing a homogenate of unpollinated pistils, germination and tube growth were markedly inhibited in a similar bioassay using incompatible pollen (Sharma and Shivanna 1982). In an extension of this work, it was found that incorporation of lectins such as conA into the medium containing the pistil homogenate, or treatment of the stigma with lectin, effectively overcomes the inhibition of growth of self-pollen. A biochemical explanation of these results assumes that since sugar-containing molecules constitute the bulk of the recognition factors in the pistil, their binding with lectin makes them ineffective in recognizing self-pollen. Another experiment showed that when pollen grains are coated with either sugars or lectin and are cultured in a medium containing the pistil homogenate, sugar-coated pollen grains overcome the growth inhibition, but pollen treated with lectin do not. If lectinlike molecules of the pollen are involved in self-recognition, these results indicate that blocking lectin with sugars makes the molecules ineffective in an ensuing recognition reaction (Sharma and Shivanna 1983; Sharma, Bajaj and Shivanna 1985). Results of a later study showed that application of an extract from compatible pistils to the stigma before pollination promotes pollen germination and tube growth and overcomes incompatibility. It has been argued that treatment of the stigma with a compatible pistil extract apparently masks the recognition molecules of the pistil and thus overcomes incompatibility (Sharma and Shivanna 1986). The production and utilization of Sallele-specific proteins in P. hybrida have also been examined by monitoring the effects on pollen tube growth of proteins translated in the eggs of Xenopus laevis by RNA isolated from cross- and self-pollinated styles. The results, which showed an inhibition of pollen tube growth when the S-alleles of the style contributing the proteins matched Salleles of the pollen, suggest that S-specific proteins
265
are synthesized in the postpollination style (van der Donk 1975). In other experiments, specific inhibition of pollen tube growth has been demonstrated using crude pistil extracts in some incompatible genotypes of Petunia inflata (Brewbaker and Majumder 1961), Papaver rhoeas (Franklin-Tong, Lawrence and Franklin 1988), and Malus domestica (Speranza and Calzoni 1988). Purification of the stylar extract from M. domestica showed that inhibition is due to two glycoprotein fractions (Speranza and Calzoni 1990). In the gametophytic self-incompatibility systems, the role of proteins bound to the pollen wall is not fully defined and it may be that such proteins are not responsible for triggering the incompatibility response. This was shown by experiments in which leaching out of substances loosely bound to the pollen walls of Lilium longiflorum (Fett, Paxton and Dickinson 1976) and Petunia hybrida (Gillissen and Brantjes 1978) was not found to alter the incompatibility reaction or seed set. Use of mentor pollen has shown that in some gametophytic selfincompatibility systems, it is possible to mask the effects of the self-pollen by mixing with nonviable cross-pollen (Pandey 1977; Sree Ramulu et al 1979; van Tuyl, Marcucci and Visser 1982). Whether the mentor pollen grains "fool" the incompatible pollen by providing recognition substances or stimulate germination by the release of nonspecific growth hormones is not known. Characterization of S-Allele Products. Once the idea took hold that the signal involved in selfincompatibility is a chemical substance, identification of the substance and the unambiguous demonstration of its association with the expression of the S-allele loomed as important questions. The pioneering studies in this area began with immunolocalization of unique antigenic components in the pollen and pistil extracts of self-incompatible plants. Lewis (1952) found that pollen antiserum of incompatible genotypes of Oenothera organensis gave precipitation reactions with proteins from macerated pollen grains of the same genotype. In a later work, even single pollen grains, when placed on a thin layer of agar mixed with pollen antiserum, were shown to respond to the specific immunological reaction related to the incompatibility gene (Lewis, Burrage and Walls 1967). This points to a reaction produced by antigens diffusing from the pollen. The only other report of the presence of S-related molecules in the male gametophyte is from immunological studies of macerated pollen grains of Petunia hybrida that show a correlation between self-incompatibility
266
Pollination and Fertilization
reactions of the pollen and the presence of S-spePH cific proteins (Linskens 1960). Pistil proteins correlated with specific S-alleles •4 were first identified in the sporophytic system of — — — Brassica oleracea. In this work, antipistil antiserum i i i — •5 from known genotypes was used to localize S-specific antibodies in diffusates of the intact pistil. That h incompatibility is correlated with the appearance g q g g g — 6 t i i of particular antigens is also suggested from the e e e » fact that stigmas of immature buds that are genetie •7 .... cally self-compatible lack the antigens (Nasrallah d c d 8 b and Wallace 1967b). In other studies, these antigens a 9 were shown by electrophoresis and immunodiffusion to cosegregate with their respective alleles in S39S39 S22S22 S45S45 S13S13 S7S7 S2S2 genetic crosses. For example, an electrophoretic separation of stigmatic homogenates of several S- Figure 9.9. Diagram showing the separation of stigmatic allele genotypes showed that each has a specific S- protein bands in different homozygous lines of Brassica oleracea. The separation of bands labeled a to / differs in allele protein band. More to the point is the fact that the different lanes. (From Nishio and Hinata 1977.) in heterozygous plants, protein bands associated with both S-alleles could be resolved (Nasrallah, Barber and Wallace 1970). Analysis of stigma linked to it. The evidence from PAS staining of the extracts of Fl and F2 populations of homozygous S-allele protein bands led to their identification as self-incompatible genotypes showed that certain glycoproteins. The molecular mass of the proteins antigens present in Fl and F2 progenies are pre- purified from stigma extracts ranged from 51 kDa cisely correlated with the segregation of S-alleles to 65 kDa. The glycoproteins had basic Pis (pH and with a functional incompatibility response 10.3-11.1) and showed immunological specificity (Nasrallah, Wallace and Savo 1972). Because of like that of the crude stigma homogenates (Hinata, their agreement with immunodiffusion and elec- Nishio and Kimura 1982; Nishio and Hinata 1982). trophoretic data, these results suggest that S-anti- The glycoproteins were precipitated by con-A and gens are specific protein products. In another work, were not detected in pollen grains or other plant Nasrallah (1974) showed that the degree of self- tissues. In accordance with the established incomincompatibility in B. oleracea is dependent upon the patibility reactions, low quantities of S-specific glyrelative concentrations of the S-protein, because coproteins occur in the stigmas of self-compatible self-incompatibility, partial self-incompatibility immature flowers and their concentration increases and self-compatibility displayed by parental geno- with flower maturity and attainment of self-incomtypes and mutant phenotypes are accompanied by patibility. As expected, stigmas of flowers homozycorresponding changes in the level of this protein. gous for an S-allele have twice the amount of However, an apparent lack of cross-reaction glycoproteins as heterozygotes (Sedgley 1974; between S-antigens also led to the view that they Nishio and Hinata 1977). These studies have made are not polypeptide variants of a protein molecule B. oleracea a classic system for incompatibility studies, and the criteria used to establish the relation(Nasrallah 1979). Results (Figure 9.9) that are completely in ship between the S-allele and the protein product accord with the aforementioned have come from have gained general acceptance in molecular invesinvestigations of different genotypes of B. oleracea tigations of self-incompatibility in other plants. If the glycoproteins are indeed S-gene products (Kucera and Polak 1975; Nishio and Hinata 1977). In a follow-up study involving isoelectric focusing that interfere with the germination of incompatible of stigmatic proteins of 13 S-alleles in diallelic pollen grains on the stigma, it should be possible to crosses of B. campestris and B. oleracea, Hinata and demonstrate the potency of the compounds in situ Nishio (1978) showed a provocative correlation on the stigma; this has been done. Pretreatment of among seven allelic families between S-genotypes pollen grains of B. oleracea with a highly purified Sdetermined by breeding analysis and specific pro- allele-specific glycoprotein isolated from stigmas teins. Since the protein that corresponds to a partic- induces an incompatible reaction leading to the ular S-allele cosegregates with that allele, the inhibition of pollen germination on otherwise comlogical conclusion is that the protein is the product patible stigmas bearing the same allele (Ferrari, either of the S-allele or of a gene sequence closely Bruns and Wallace 1981). Since germination of sim-
Developmental Biology of Incompatibility
267
+
ilarly pretreated pollen grains of other S-allele genotypes is not prevented, the incompatibility reaction induced by the glycoprotein is considered $1 to be specific to pollen containing the particular genotype. Following these studies, using techniques similar to those described for Bmssica, S-allele protein relationships have now been established in several species showing the gametophytic type of incompatibility (Figure 9.10). As will be seen in Chapter 2 10, a generous portion of the current work on the molecular biology of self-incompatibility has been done on Nicotiana data. In this species, S-glycoproteins corresponding to three different S-alleles were first identified in both style and stigma (Bredemeijer and Blaas 1981). The proteins are generally present in the highest concentration in the zone of the pistil where inhibition of pollen tube 3 growth occurs (Anderson et al 1986; Jahnen et al 1989; Kheyr-Pour et al 1990). Since in vitro growth of the pollen tube is inhibited by S-glycoproteins with S-genotype specificity, it is clear that they are gene products of S-alleles (Harris, Weinhandl and Clarke 1989; Jahnen, Lush and Clarke 1989). Another genus in which considerable advance has been made in the analysis of S-proteins of selfincompatibility is Petunia. The most abundant protein of the pistil of self-incompatible P. hybrida is a glycoprotein with molecular mass in the range of 27-30 kDa and PI in the range of 8.3-9.3. A striking correlation of distribution of the glycoprotein with a given S-allele has been demonstrated by both one- and two-dimensional electrophoresis using homozygous and heterozygous lines segregating for four different S-alleles, their Fl progeny, and backcrosses. These proteins are generally localized in the stigma and in the upper part of the style, and their distribution pattern in flower buds of different ages is consistent with the development of selfincompatibility. They also display biological activity as indicated by the allele-specific inhibition Figure 9.10. Two-dimensional SDS-PACE of stylar proteins mature flowers of different homozygous lines of of pollen tube growth (Kamboj and Jackson 1986; from Nicotiana alata. The part of each gel corresponding to the Broothaerts et al 1989, 1991). In P. inflata the pro- molecular mass range of about 25-40 kDa and to the neuteins that cosegregate with their S-alleles have mol- tral to basic pH range is shown. Arrows indicate proteins ecular masses of 24 and 25 kDa (Ai et al 1990). common to all genotypes; arrowheads indicate 5Stylar proteins identified from other solanaceous allele-specific proteins. (From Jahnen et al 1989. Plant Cell 1:493-499. © American Society of Plant Physiologists. species, such as Lycopersicon peruvianum (Mau et al Photograph supplied by Dr. A. E. Clarke.) 1986), Solanum tuberosum (Kirch et al 1989), and S. chacoense (Xu et al 1990a), are also associated with the S-gene product. It is restricted to the mature S-alleles. style, whereas the other (antigen-P) is present in S-allele proteins have been identified with some styles of all stages of development in three other success only in a few other plants. Of the two anti- species of the genus: P. cerasus, P. armeniaca, and P. gens detected in stylar components of Primus persica (Raff, Knox and Clarke 1981). The antigen is avium, one (antigen-S) typically corresponds to the also secreted into the medium by callus cells product of the S-allele and is thus a candidate for derived from both leaf and stem of P. avium; rather
30 S
30 S
30
30
30
t
268
Pollination and Fertilization
than showing the ubiquitous nature of the antigen, to the 1970s was in identifying and characterizing this probably reflects on the totipotency of the cal- the S-allele recognition factors of the style. There lus cells and their ability to express the total are grounds for believing that in most self-incomgenome of the plant (Raff and Clarke 1981). The S- patibility systems, the interaction between recogniantigen has been identified as a glycoprotein hav- tion factors contributes to physiological changes in ing a 16.3% carbohydrate content and consisting of the stigma or style that results in the demise of selftwo closely related components with an apparent pollen. Models to explain the mechanism of selfmolecular mass of 37-39 kDa. It is also a potent incompatibility not only differ in the signaling inhibitor of in vitro pollen tube growth of the same substances thought to participate in inhibiting genotype of P. avium (Mau, Raff and Clarke 1982; pollen germination and pollen tube growth, but Williams et al 1982). Two S-associated glycoprotein also in the emphasis placed on the reactions themfractions of the style of a self-incompatible cultivar selves. However, without any data on the nature of of Malus domestica selectively inhibit self-pollen the S-allele products of the pollen grain, it is pregermination and tube growth without any effect on mature to generalize the models proposed. cross-pollen. In some incompatible cultivars, variThe tenet of the theory on pollen-pistil interacetal differences have been observed in the S-glyco- tion is that through chemical signals, a basic comproteins (Speranza and Calzoni 1990; Sassa et al patibility between the two partners has evolved, 1994). In Pyrus serotina, sugar-containing basic pro- and interactions controlled by S-genes are superteins of molecular mass of about 30 kDa are associ- imposed upon this compatibility. As a working ated with self-incompatibility. The glycoproteins hypothesis, Lewis (1965) proposed that following are predominantly expressed in the style, have self-pollination, a specific interaction of gene prodhigh RNase activity, and are considered homolo- ucts of each S-allele present in the pollen and in the gous to those identified in Solanaceae (Sassa, style causes the synthesis of a dimer repressor. This Hirano and Ikehashi 1992,1993). Demonstration of compound, presumably a tetramer, functions as the S-gene activity of a component isolated from the incompatibility complex and serves to inhibit the stigma of Papaver rhoeas stems from the work of growth of the pollen tube. In an interesting variant Franklin-Tong et al (1989). When extracts from stig- of this hypothesis, Ascher (1966, 1975) proposed matic tissues were fractionated on immobilized that a regulatory system involving a repressor molcon-A, a fraction in the molecular mass range of ecule generated from the interaction between iden20-25 kDa bound to con-A was found to function tical S-locus products of the pollen grain and the as an S-allele-specific inhibitor of pollen tube style prevents the growth of the pollen tube in the growth. The uniqueness of this substance is that it style to an appreciable extent; the result is selfappears as the component of a 22-kDa protein in incompatibility. When there is a mismatch between SDS-PAGE of stigma extracts but is absent from S-allele products of the pollen grain and the style, a similar preparations of buds, ovary, petal, leaf, and rapid-growth operon, which switches the pollen root. In the only published report of a link between tube from a slow-rate metabolism to a fast-rate incompatibility and stylar proteins in a member of metabolism, is activated; this gives rise to a comPoaceae, Tan and Jackson (1988) showed that patible reaction. A feature of the model, which is stigma extracts of various incompatible genotypes based on a long series of experiments on Lilium of Phalaris coerulescens exhibit differences in the longiflorum, is the requirement for a special stylar pattern of proteins in the molecular mass range of substance to support compatible pollen tube 43-97 kDa. Since self-incompatibility in grasses is growth; based on heat-treatment experiments, it is controlled by two unlinked loci, S and Z, it is also believed that this substance is synthesized beginclear that their alleles control the expression of the ning at anthesis and continuing until senescence of stigma proteins; however, from the available data the flower. it is difficult to assign specific proteins to each Mulcahy and Mulcahy (1983) have proposed a allele. heterosis model to explain gametophytic incompatibility as a passive mechanism. The basic premise of this model is that there are many incompatibility (v) Models of S-Allele Action genes (supergenes), each containing one dominant The defining genetic feature of the self-incom- and several deleterious recessive alleles. Pollen tube patibility system enforcing outbreeding in growth resulting in self-incompatibility or crossangiosperms centers around the gene products of compatibility is a manifestation of pollen-style the stigma and style. Perhaps the most significant interactions on a quantitative level. Accordingly, advance in our understanding of this system prior after cross-pollination, a large number of incompat-
269
Developmental Biology of Incompatibility
ibility loci would be heterozygous in the pollen and style; this raises the number of heterotic pollen-style interactions and leads to an increase in the rate of pollen tube growth. Following self-pollination, a high proportion of the supergenes would be homozygous and deleterious and would thus cause a reduction in the extent of pollen tube growth. The major thrust of this model is in asserting that self-incompatibility is not due to the action of S-allele-specific inhibitory molecules, but results from a slow growth of pollen tubes to effect fertilization before floral abscission. This model has been subject to some criticism (Lawrence et al 1985). Extrapolating from the features of self-incompatibility system in grasses, J. Heslop-Harrison and Heslop-Harrison (1982b) proposed a model of pollen acceptance and rejection based on carbohydrate-protein interaction. Proteins with lectinlike properties seem to be involved on the pistil side, either binding to the complementary sugar sequences of the wall carbohydrates of the incompatible pollen tube or ignoring similar sequences on the compatible tube due to noncomplementarity. S-alleles in the pollen grain specify the sugar sequences that become ordered into distinct configurations in the polysaccharide matrix synthesized and deposited at the tip of the pollen tube. Binding of pistil S-factors to the specific sugars of the pectic wall components of the pollen tube tip may set off a train of determinative events that culminate in the disruption of pollen tube growth. For sporophytic systems, a function for cutinase has been invoked to explain the self-incompatibility phenomenon. This might appear reasonable since the presence of a cuticularized dry stigma is one of the features that distinguish sporophytic systems from gametophytic systems. Although both direct and indirect lines of evidence exist for the presence of cutinase in germinating pollen grains, it is not understood how S-proteins affect the enzyme present in the responding self- and nonself-pollen grains (Linskens and Heinen 1962; Shayk, Kolattukudy and Davis 1977). S-proteins on an incompatible stigma might inactivate the cutinase present in the pollen or they might furnish permissive conditions on a compatible stigma for pollen cutinase or its precursors to be activated (Christ 1959). The assumptions of the models described have been based on evidence primarily from genetic analysis and pollen behavior rather than from the chemical nature of the reactants or the reaction products. It is therefore not surprising that contradictory views have been presented against these models. The models are described here to stimulate
the kind of research that will generate new information about the basic mechanism of self-incompatibility in the future. On a general level, it is safe to say that self-incompatibility covers a multiplicity of physiological processes in the stigma and style and that it would be unwise to focus exclusively on a single hypothesis. 3. CROSS-INCOMPATIBILITY
Gametic fusions between plants belonging to different species, and even between different genera, account for the introduction of beneficial alien genes despite natural reproductive barriers. Be that as it may, pressures to maintain the identity of species are so great that such fusions are impeded by cross-incompatibility mechanisms. This apparent contradiction between wide hybridization and cross-incompatibility is one of the reasons why there are so few reports of successful interspecific or intergeneric hybridizations in plants. Unlike self-incompatibility, which is a prefertilization barrier, cross-incompatibility can cause both pre- and postfertilization events to suffer. The magnitude of failure depends upon the stage at which the rejection reaction occurs; although prefertilization rejection generally results in the inhibition of pollen tube growth, postfertilization lesions cause more drastic effects, such as the arrest of embryo and endosperm development. This section will consider only examples of cross-incompatibility due to prefertilization blocks; an account of postfertilization blocks leading to cross-incompatibility is given in Chapter 14. (i) Physiological Background
Many interspecific crosses are doomed to fail when self-incompatible pistillate parents are crossed with self-compatible staminate parents, although success is often assured in the reciprocal cross. This one-way isolation, known as unilateral incompatibility, is widespread in many cultivated plants and appears as an extension of self-incompatibility. In addition to occurring in interspecific crosses, unilateral incompatibility has been described in intergeneric matings between members of Solanaceae and even in crosses between members of two families (Lewis and Crowe 1958). An analysis of unilateral incompatibility in Solatium illustrates some structural and developmental aspects of the phenomenon. Clones showing unilateral incompatibility exist among different species of Solarium, although some incompatible clones of a species could be successfully crossed
270
reciprocally with self-compatible clones of another species (Grun and Radlow 1961). In crosses
Pollination and Fertilization
cells during the period of their active growth (Gaget et al 1984; Rougier et al 1992). On the basis between clones of S. chacoense, S. soukupii, and S. of experiments using recognition pollen and treatverrucosum and other genotypically differing ment of pollen grains and stigma with lipid soldiploid and dihaploid potato, unilateral incom- vents, pollen wall and stigmatic proteins have been patibility is traced to the inhibition of pollen tube implicated in interspecific incompatibility reactions growth in the upper half of the style or to a block in Populus (Knox, Willing and Ashford 1972; that operates after the pollen tubes reach the Willing and Pryor 1976). External surface features ovary (Grun and Aubertin 1966; Trognitz and of the stigma, such as a dry-type stigma covered by Schmiediche 1993). Consistent pollen tube inhibi- a pellicle, or a wet-type stigma with copious exution just below the stigma, accompanied by tip date, also seem to determine the success of interswelling, and the bursting of pollen tubes and specific crosses in the genus (Villar et al 1987). extensive callose deposition, is typical of most An important unifying concept that has emerged interspecific tuber-bearing X non-tuber-bearing from studies of cell-to-cell recognition in animals is species crosses, whereas in interspecific tuber-bear- that binding of signal molecules to most target cells ing species crosses, pollen tubes make considerable will activate adenylate cyclase to produce cAMP as headway in the style before callose deposition a second messenger. This molecule has attained stymies growth. Surprisingly, electrophoresis of great significance in diverse biochemical reactions, stylar proteins does not reveal any S-allele prod- such as activation of protein kinases, Ca2+ flux, comucts involved in the incompatibility reaction (Fritz partmentalization, phosphorylation, and gene tranand Hanneman 1989). scription. In an attempt to link cellular recognition events during interspecific incompatibility in In most interspecific crosses in Lycopersicon, pollen tubes end up below the stigma with swollen Populus to enzyme-mediated signal transduction, or burst tip, which means that only short tubes are one study has shown a strong correlation between produced (McGuire and Rick 1954). According to de adenylate cyclase activity in the stigmatic cells and Nettancourt et al (1973b), the incompatibility bar- the adhesion, hydration, and germination of pollen rier between L. esculentum and L. peruvianum at the grains on the stigmatic surface following compatiultrastructural level has many aspects in common ble pollinations in P. deltoides. In contrast, not only is with changes following self-pollination and some the activity reduced in the stigma cells following aspects that are specific to unilateral incompatibil- interaction between the P. deltoides stigma and P. ity. In both types of incompatibility, the rejection alba pollen grains, it completely disappears from the reaction initially results in a progressive disappear- stigmatic cells after pollen germination (Rougier, ance of the callosic inner wall and in an accumula- Jnoud and Dumas 1988). On the basis of these tion of particles in the pollen tube cytoplasm. Later, results, it is possible, although by no means proven, the interspecific cross elicits additional reactions, that growth of the incompatible pollen tube is inhibsuch as breakdown of the outer pollen tube wall, ited due to biochemical ramifications connected folding of the inner wall, and disappearance of the with the disappearance of adenylate cyclase. callose plug at the tip of the pollen tube. Structural and physiological studies on crossVarious studies on interspecific crosses in incompatibility in other systems have not been carPopulus have established the importance of pollen ried out to the same detailed level as in Solarium, tube growth inhibition in the stylar tissues. In the Lycopersicon, and Populus. A two-step incompatibilreciprocal cross combinations of P. tremuloides and ity system, one present in the stigma and the other P. trichocarpa, growth inhibition is expressed by the in the style, seems to be operating in Sesamum presence of short, twisted, or extended tubes, but in mulayanum X S. indicum (Pedaliaceae) cross. no case do pollen tubes penetrate the stigmatic sur- Although the stigmatic barrier is overcome by the face (Stettler, Koster and Steenackers 1980); in other presence of recognition pollen, the pollen tube interspecific crosses, incompatible pollen tubes are meets its nemesis in the style, where its growth is able to grow through the stigmatic tissues but are inhibited (Sastri and Shivanna 1976). In interspecific arrested in the style by a swelling of the tips and by and intergeneric crosses in Poaceae, pollen tubes are callose plug formation (Li et al 1991; Villar et al most commonly arrested in the transmitting tract of 1993). Peculiarities of the P. deltoides X P. alba crossthe style or in the upper realm of the ovary wall. are reflected in the presence of callose in the trans- These late-acting barriers to fertilization are not mitting tissue of the style in the vicinity of the dead accompanied by the formation of massive callose pollen tubes, which, however, have functional male plugs in the pollen tubes (J. Heslop-Harrison 1982).
Developmental Biology of Incompatibility
271
From the point of view of diversity of sites of pollen is not rigidly fixed, because the stigma of a species tube arrest, of variations in pollen tube abnormali- may take pollen grains of other species of the same ties, and of anomalies in the pattern of callose depo- genus through the usual round of rejection reacsition, interspecific crosses in Rhododendron are tions. The question arises as to whether, during unmatched (Williams, Knox and Rouse 1982). In interspecific pollination, the stigma relinquishes its controlled germination of pollen grains of various role in discriminating between compatible and genera on the stigma of Crocus chrysanthus, Y. incompatible pollen grains. Once the incompatible Heslop-Harrison (1977) found that the stigma is pollen grains escape the stigmatic barrier, the pistil hospitable not only to pollen grains of certain other must experience strong selective pressure to evolve genera of Iridaceae, but also to pollen grains of tax- an effective rejection response. This forms the basis onomically remote members such as Arabis cypria for a hypothesis that interspecific incompatibility and Erysimum capitulum (both of Brassicaceae). The (designated as incongruity) differs from self-incompollen grains, which are freely accepted on the patibility and does not involve the participation of stigma, germinate and breach the cuticle of the stig- the S-locus (Hogenboom 1973). Rather, the fate of matic papillae, but pollen tubes are arrested on their the pollen grain may hinge on the relationship between the matching gene sets present in the way to the style. pollen and the pistil. Thus, nonmatching of the genetic program resulting from a lack of informa(ii) Genetic Models tion on one or more pistil characters in the pollen The most critical issue bearing on the genetics of will result in incompatibility. unilateral incompatibility relates to the role of the S-locus acting either alone or in association with other loci. The involvement of the S-locus in self- 4. GENERAL COMMENTS incompatibility rules out the possibility that a difOur current understanding of the biology of ferent gene controls unilateral incompatibility incompatibility has involved morphological, because it is unlikely that plants have specific and genetic, physiological, biochemical, and molecular different genes to control the recognition of each of investigations. Most morphological and genetic the incompatibility traits. It seems more probable research has favored the heteromorphic self-incomthat plants have retained the ability to discriminate patibility system, whereas homomorphic selfbetween pollen grains that land on the stigma incompatibility has figured in most physiological, through common surface molecules before S-allele biochemical, and molecular investigations. products become involved in the acceptance or Although the applied aspects of incompatibility rejection reaction. An interpretation resulting from research are not emphasized in this chapter, conthis analysis is that self-incompatibility and inter- siderable work on cross-incompatibility has been specific incompatibility are expressions of the same concerned with overcoming the pre- and postfertilgene and that the S-locus displays a dual function ization barriers to embryo development and seed (Lewis and Crowe 1958). Pandey (1964) has corrob- set in plants. orated this hypothesis of direct control by the SAlthough there is much interest in understanding locus of incompatibility in self-incompatible the mechanism of heteromorphic self-incompatibilspecies when it overlaps with a related self-com- ity and cross-incompatibility, this has been dwarfed patible species. by the spectacular progress made in the molecular In a second hypothesis, interactions between the analysis of homomorphic self-incompatibility. The S-gene locus and other loci are thought to play a initial strategy that ushered in this advance was to role in the interspecific incompatibility response. identify stylar proteins that, by their temporal and Abdalla and Hermsen (1972) proposed a two- spatial localization, possess the expected properties power competition hypothesis in which unilateral of an incompatibility protein. Although traditional incompatibility genes prevent fertilization by any models of self-incompatibility have proposed the pollen grains bearing a self-compatibility allele. existence of a pollen-specific component as well, the According to this view, plants develop unilateral nature of the pollen incompatibility molecules has so incompatibility in response to the emergence of far remained elusive. Despite this major gap, the conself-compatible alleles within self-incompatible cepts formulated have proved to be of great value in providing a framework for our interpretation of the species. The specificity of rejection in interspecific incom- developmental biology of homomorphic self-incompatibility, compared to that in self-incompatibility, patibility in plants.
Chapter 10
Molecular Biology of Self-Incompatibility 1. The Brassica System (i) Structure of the Gene and Its Protein (ii) The S-Multigene Family (iii) A Gene for Pollen Recognition and Signal Transduction (iv) S-Locus Genes in Anthers and Pollen Grains (v) Analysis of the Breakdown of SelfIncompatibility (vi) S-Gene Expression in Normal and Transgenic Plants 2. The Nicotiana System (i) Ribonucleases as S-Glycoproteins (ii) Expression of S-Locus Glycoprotein Genes 3. Petunia and Other Solanaceae 4. Papaver rhoeas and Other Plants 5. General Comments
The stigma and style are the two differentiated organs of the flower that carry out the unique function of gamete screening and selection during pollination. These floral organs undergo subtle morphological and physiological changes from the time they are carved out on the receptacle, and by the end of their growth phase they develop a set of unusual features for the specified functions. It was pointed out at some length in Chapter 9 that the current models for pollen-stigma-style interactions during self-incompatibility predict the existence of a precise control system involving expression of the S-locus both in the pollen grains and in the stigmatic papillae or the style. Supporting this prediction, S-gene products from the pistil were identified as specific glycoproteins, although thus far no pollen component has been found. Based on this model, the widely accepted view is that self-incompatibility is due to the composite reaction of two 272
somewhat similar, if not identical, molecules. The past two decades have witnessed major advances in the isolation and characterization of S-locus genes of the stigma-style complex and of their protein products which mediate in self-incompatibility reactions in some model systems. Included in the list of model systems are plants that display either gametophytic and sporophytic types of self-incompatibility. This chapter will deal with the fundamental components of S-locus genes and their protein products. The background of understanding from this research has generated new thoughts on how pollen grains respond to a natural, hostile environment of the self-stigma as they mark their time toward premature demise. There is no paucity of reviews on the molecular and genetic biology of self-incompatibility, and one can even say that this subject has been overreviewed. In chronological order, a sampling of general reviews includes those by Nasrallah and Nasrallah (1986), Gaude and Dumas (1987), Cornish, Anderson and Clarke (1988), Bernatzky, Anderson and Clarke (1988), Cornish, Pettitt and Clarke (1988), Ebert et al (1989), Dickinson (1990), Haring et al (1990), Dzelzkalns, Nasrallah and Nasrallah (1992), Singh and Kao (1992), Thompson and Kirch (1992), Hinata et al (1993), Sims (1993), Franklin, Lawrence and Franklin-Tong (1995), Golz, Clarke and Newbigin (1995), and Dodds, Clarke and Newbigin (1996); in addition, reviews summarizing the work done on specific incompatibility systems such as Brassica (Nasrallah and Nasrallah 1989, 1993; Nasrallah, Nishio and Nasrallah 1991; Trick and Heizmann 1992; Nasrallah et al 1994), Nicotiana (Newbigin, Anderson and Clarke 1993), and Papaver (FranklinTong and Franklin 1992,1993) have also been pub-
Molecular Biology of Self-Incompatibility
273
lowed the pattern of synthesis of SLG in the stigma. Two additional, well-supported experiments have confirmed the inference that the SLG-6 gene in fact encodes an S-allele-related protein or that it represents a gene sequence tightly linked to the S-locus. The first experiment used Western blot analysis, which showed that a fusion protein produced by a construct containing the open reading frame derived from the SLG-6 gene in an expression vector cross-reacts with antibodies made against SLG purified specifically from styles of plants bearing the S-6 allele. In the second experiment, Southern blot analysis of genomic DNA from homozygotes, heterozygotes, and segregating F2 populations of B. oleracea revealed S-allele-associated restriction site polymorphism and cosegregation of specific restriction fragments with particular genotypes in controlled genetic crosses (Nasrallah et al 1985; Brace, Nicotiana, Petunia, and Papaver; very little molecular Ockendon and King 1993). work has been done on other genera.
lished. The book edited by Williams, Clarke and Knox (1994) is another addition to the list. Despite the fundamental differences between gametophytic and sporophytic incompatibility systems, there is no reason to believe that these systems do not operate in basically the same ways at the cellular and molecular levels. However, in order to highlight the progress and problems intrinsic to each species studied, the considerable accumulated data on the molecular biology of selfincompatibility will be presented as case histories. This account is arranged in the order of decreasing volume of information. In this way one is prone to view the unfolding of the problems from various angles, analyze the answers to significant questions in concrete terms in each case, and plan strategies for future research. The self-incompatibility systems considered in detail are those of Brassica,
(i) Structure of the Gene and Its Protein 1. THE BRASSICA SYSTEM
The identification of S-allele products of Brassica oleracea as glycoproteins in the early 1970s (Nasrallah, Barber and Wallace 1970) and the simultaneous development of recombinant DNA technology provided the main impetus to clone and sequence the S-gene and its translation products, termed S-locus-specific glycoproteins (SLGs), from this species. The isolation of cDNA clones encoding the protein backbone of style glycoproteins that segregate with particular S-alleles was perhaps the most important initial result of the application of molecular approaches to the study of self-incompatibility in B. oleracea. In the strategy followed in this work, cDNA libraries made to poly(A)-RNA of mature incompatible stigmas were initially screened with labeled cDNA made to stigma mRNA and counterscreened with cDNA made to seedling mRNA. A key issue in considering the significance of cDNA clones isolated by differential screening is their relation to SLG mRNA. This was established by Northern blot hybridization of RNA prepared from stigmas of different stages of development and from vegetative organs of the plant with one of the cloned cDNA probes. This clone, derived from a S-6 homozygote of B. oleracea and designated as SLG-6, was found to hybridize to an mRNA of approximately 2.0 kb. The transcripts of this clone accumulated to high levels in maturing stigmas coincident with the acquisition of incompatibility, but were absent in leaves or seedlings. Moreover, the transcriptional activity always fol-
Further evidence that SLG-6 and similar cDNA clones encode SLGs is provided by nucleotide sequence analysis of the cDNA inserts and a comparison of the properties of their predicted polypeptides. By screening cDNA libraries, Nasrallah et al (1987) generated cDNA clones homologous to the SLG-6 gene from three lines of B. oleracea homozygous for S-6, S-13, and S-14 alleles. A new perspective on the close relationship between these S-alleles is provided by the observation that nucleotide sequences of representative cDNA clones of each allele predict closely related polypeptides. Based on this information, a model that has been proposed for the SLG-6 gene is that it has a single exon that contains a continuous open reading frame of 436 amino acids and includes a hydrophobic domain of about 31 amino acids in the NH2-terminus (Figure 10.1). To account for the presence of the conserved NH2-terminal region, a primary assumption is made that there is a need for a signal polypeptide to move the SLG across the plasma membrane in a relatively short time. Variations, such as substitution, insertion, and deletion, as well as changes in the number and position of potential sites of N-glycosylation at which carbohydrate chains are attached to the asparagine residues of SLGs, are found in the amino acid sequences of alleles (or haplotypes as they are currently known). Focusing on the variations in the amino acid sequences is the observation that differences are confined to two highly conserved domains, A and C, which alternate with two
274
Pollination and Fertilization
Residue:15 33 +
8390
•
i *
t
•
•
•
119
214
230 251
+
•
•
•
+
+
•
i.s
-2
I . JB
es.
iff
er K S3 ft
"S
+
T
L -1
S i r
359
Sl3
-0 —
296
•
+ • t
284
IK
sg
"5
Mr
? I
33 ^ S
1*8
g as
I r
B
-3
-31
80%
182
42%
275
81%
Figure 10.1. Hydropathy profile of 5-locus glycoprotein predicted from an S£G-6-derived sequence showing the hydrophobic signal peptide at the N-terminus (residues -31 to 1). Hydrophobic region, above the horizontal line; hydrophilic region, below the horizontal line; arrowheads, potential N-glycosylation sites from an SZ.G-6-derived sequence; +, potential N-glycosylation sites derived from S-
6, S-13, and 5-74 alleles. A, B, C, and D refer to regions showing about 80%, 42%, 81%, and 46% conservation, respectively. The highly variable B region (residues 182 to 274) is relatively hydrophilic. Ten conserved cysteine residues are in region C (residues 275 to 378). (From Nasrallah et al 1987. Reprinted with permission from Nature 326:617-619. © Macmillan Magazines Ltd.)
variable domains, B and D, within the molecule. Interesting, although still uninterpreted in terms of functional significance, is the fact that conserved region C is characterized by the presence of a backbone of several invariant cysteine-rich clusters. The initial phase of investigation into the structure of SLG came full circle with the work of Takayama et al (1987) and Isogai et al (1987), who isolated S-glycoproteins from three homozygotes (S-8, S-9, and S-12) of B. campestris and sequenced the peptides generated from them. Not surprisingly, amino acid sequences of the two species showed slightly more than 80% identity (Figure 10.2). However, lack of major differences in the numbers and positions of potential glycosylation sites or in cysteine residues and other molecular properties of SLGs from the different S-haplotypes has left open the question of allelic specificity of the glycoproteins. Following these initial investigations into SLG structure, cDNA clones for other S-alleles were isolated from stigma tissues of different cultivars of B. oleracea. These included lines homozygous for S-22 (Nasrallah, Yu and Nasrallah 1988), S-2 (Chen and Nasrallah 1990), and S-5 (Scutt et al 1990). Plants homozygous for the S-29 allele express in the stigma two distinct, but related, abundant transcripts (Trick and Flavell 1989). Additionally,
nucleic acid sequences have been reported for genomic clones derived from S-13 and S-63 homozygotes of B. oleracea and from S-8 homozygote of B. campestris, (Trick 1990; Dwyer et al 1991) and for cDNAs cloned from self-incompatible B. napus (Goring et al 1992a, b). Some comparisons of the nucleotide and amino acid sequences of SLGs from the different homozygotes are germane here. For example, at the nucleotide level, the S-5 and S6 sequences show 71% homology, whereas at the amino acid level, homology between the two is reduced to 64% (Scutt et al 1990). S-2, which is a pollen-recessive self-incompatibility allele, exhibits only a weak DNA homology to S-6, S-13, S-14, and S-22 genes (Chen and Nasrallah 1990). Although nucleotide sequences of the two clones from the S29 line are clearly homologous, their sequence conservation is limited to only 70% (Trick and Flavell 1989). Sequences of the S-8 gene from B. campestris and the S-13 gene from B. oleracea display 92% homology and differ by only 105 nucleotides. Interestingly enough, the deduced amino acid sequence of glycoprotein of S-8 homozygote from B. campestris differs by only two residues from the amino acid sequences obtained by direct protein sequence analysis of S-8, S-9, and S-22 alleles of this species. The conservation of an equal number of cys-
275
Molecular Biology of Self-Incompatibility 1
S6 I N ' S8 VN T S9 1 S T SI 2 I N I
L L L L
SSTBSt S ST B S L SSTESl SSTESL
R T T T
I I I I
S S S S
P P P P
G G G G
N N F B L G F F R Tl D V F E L G F FJLTJ N I F E LGF N VF E L G F
60 RTYVWVANRDNPLSN RTYVWVANRDNPLSN
S6
RTYVWVANRDNPL 100 P VV P V V PV V P V L
80 NLVLL N L V L L
s N M
S-N
120 A A A A
E E E E
L L L L
L L L L
S A A A
NGN NGN NGN NGN
R D S R D S R P S
N R F N R F L
N N
W H X X
R R X X
S 9 N
160 S D Y D S 5 X
200
CSGDGFTRMKNMKLPET CSGDGFTRMKNMKLPET
S6 S8 S9 S12
C C C C
T T T T
A A A A
F F F F
A N A D A N A D A N A D A N A P
I R N I RN I I
401 S6 S8 S9 S12
A D L Vl
Figure 10.2. Amino acid sequences of S-glycoproteins of Brassica species. S6: S-6 glycoprotein of B. oleracea; S8, S9, and S12: S-8, S-9, and S-12 glycoproteins of B. campestris. Potential glycosylation sites are circled.
Common sequences among S-glycoproteins are boxed. The underlined part shows relatively different sequences. Asterisks indicate the 12 cysteine residues. (From Isogai et al 1987.)
napus (Goring et al 1992a, b). It would appear from the amino acid sequence comparisons that the pattern of glycosylation and the sequence of the more 12 haplotypes of B. campestris, of S-6, S-13, S-14, S- variable parts of the polypeptide may be significant 22, and S-29 haplotypes of B. oleracea (Dwyer et al in determining allelic specificity in the self-incom1991; Yamakawa et al 1994), and of a clone from B. patibility reaction. teine residues, the presence of common and unique glycosylation sites, and a variable central region characterize the amino acid sequences of S-8 and S-
276
Pollination and Fertilization
self-incompatibility. One conclusion that seems justified from the presence of a single band on Southern A particularly interesting and totally new blot of genomic DNA from B. oleracea strains probed insight into the organization of the S-gene and its with SLR-1 gene and from a comparison of SLR-1 mode of function in the discrimination of self- gene sequences isolated from three different S-allele pollen was obtained in Southern blot hybridization homozygotes is that this gene is highly conserved of Brassica oleracea genomic DNA with SLG-cDNA and, despite allelic differences, encodes identical proprobes. This work revealed multiple bands of teins in different strains of B. oleracea. Another conhybridization, indicating the presence of several clusion, that SLR-1 is not genetically linked to the sequences related to the SLG structural gene. In a S-locus gene battery and therefore is not likely to typical blot, as many as 12 genomic regions with determine the specificity of self-incompatibility reacvarying degrees of signal intensity were observed tions except in an ancillary role, is reinforced by the (Nasrallah et al 1985; Nasrallah, Yu and Nasrallah observation that segregation of SLR-1 polymorphism 1988). This was an unexpected finding in view of in F2 plants is independent of the S-genotype. An the genetic prediction that a single gene is involved alternative role that has been speculated for SLR-1 in determining the specificity of incompatibility, protein is that it may have some general function in and it suggested that the SLG gene is a member of pollination, perhaps to ensure the adhesion of the a multigene family. Dwyer et al (1989) cloned sev- pollen grains to the stigma or in some way to sort out eral genomic regions containing S-sequences hav- the self-incompatibility reactions at the stigmatic suring different degrees of homology to the SLG gene face. The fact that the SLR-1 gene is not restricted and confirmed the presence of S-sequences as mul- to self-incompatible lines of B. oleracea, but is tiple related copies in the B. oleracea genome. These expressed in self-compatible (Lalonde et al 1989) data, considered along with the genetic analysis of and self-incompatible (Isogai et al 1988, 1991; F2 populations showing cosegregation of polymor- Yamakawa et al 1993) lines of B. campestris, in phisms for different S-haplotypes, referred to ear- amphidiploid self-compatible and incompatible B. lier, leave little doubt that although most of the napus (Trick and Flavell 1989; Oldknow et al 1995), S-related gene copies are at the S-locus or tightly and in Arabidopsis (Dwyer et al 1992,1994), supports linked to it, other copies of the same multigene this interpretation. family are found elsewhere in the genome; the data In addition to extensive regions of amino acid also provide a preview of the immense molecular identity, the predicted SLR-1 protein has in comcomplexity of the genetic apparatus that functions during self-incompatibility in Brassica. Perhaps these mon with SLG the putative signal sequence at the two groups of genes affect each other's expression amino terminus, N-glycosylation sites, and a cysteine-rich domain at the carboxy terminus. These during self-incompatibility reaction. features suggest that SLR-1 protein, like SLG, is a However, the key issue in considering the ulti- secreted glycoprotein (Lalonde et al 1989; Umbach mate function of genes that are genetically unlinked et al 1990). The generality of the occurrence of SLR to the S-locus is the extent of similarities and differ- genes in Brassica was demonstrated by the characences between their nucleotide sequences and the terization of a second, stigma-expressed SLR gene proteins they encode. To highlight the individuality (SLR-2) from S-2 and S-5 genotypes of B. oleracea of the genes that are not linked to the S-locus and (Boyes et al 1991) and from a genotype of B. yet to emphasize their sequence homology to campestris (Tantikanjana, Nasrallah and Nasrallah SLG genes, the former are designated as S- 1996). The import of these investigations is that LOCUS-RELATED (SLR) genes. Lalonde et al (1989) since the SLR-2 gene segregates independently of found that among the clones in a mature B. oleracea the S-locus, like SLR-1, it cannot be implicated in stigma cDNA library that hybridized with stigma- determining the specificity of self-incompatibility. derived cDNA probes were several that showed a Moreover, DNA sequences of the SLR-2 gene from reduced homology to SLG clones. Molecular analysis the two lines of B. oleracea are homologous to the of one of these sequences revealed that it corre- sequence of an S-5 homozygote (Scutt et al 1990). sponds to an expressed gene, SLR-1, with a 68% Scutt and Croy (1992) have described another nucleotide sequence homology to the consensus SLG expressed S-gene family closely related to the SLRgene and a 58% amino acid sequence homology. As 2 gene from the S-5 homozygote; a puzzling findin the SLG gene, transcription of the gene encoding ing from this work is that, regardless of the genetic SLR-1 message occurs only in the stigma, and accu- background of the plants, this sequence is specific mulation of the message coincides with the onset of to lines containing the S-5 allele. (ii) The S-Multigene Family
Molecular Biology of Self-Incompatibility
277
from S-8, S-9, and S-12 haplotypes of B. campestris and from S-29 haplotype of B. oleracea has revealed something of the complexity of the genes. The Of the two groups of genes considered so far, apparent copy number is estimated to be around SLG genes appear to confer allelic specificity in 20 and the SRK gene is thus a member of a multiself-incompatibility, whereas SLR genes have pos- gene family showing sequence divergence among sibly a more general role in pollination events. its members (Kumar and Trick 1993; G. Suzuki et al There is another gene residing in the S-locus that 1995; Yamakawa et al 1995). might offer a link between recognition at the The molecular structure of the SRK-6 gene has stigma surface and the signal transduction path- been visualized by comparison of genomic and way that culminates in pollen rejection. The gene cDNA sequences and from the predicted amino concerned is S-LOCUS RECEPTOR KINASE (SRK); acid sequences. At least three features in the struca hint of the presence of this gene came from an ture of the SRK-6 gene that set it apart from SLG analysis of overlapping clones from an S-6 gene can be distinguished. One is that the predicted genomic library that showed homology to the SLG SRK-6 protein has a potentially glycosylated extragene but had a different restriction map (Nasrallah, cellular domain (S-domain) that displays about Yu and Nasrallah 1988). Subsequently, SRK-cDNA 90% homology to SLG-6, including the hallmark clones were isolated by screening an S-6-line alignment of cysteine residues. Secondly, inherent stigma cDNA library with a kinase domain-spe- in the structure of SRK-6 is a cytoplasmically oricific probe or by PCR amplification from cDNA ented, putative kinase catalytic domain consisting using appropriate primers. A hypothetical relation- of blocks of conserved protein sequences harboring ship between SRK-6 and SLG genes that has the key amino acid residues; thirdly, there is a remarkadvantage of testability is that both are genetically able, small stretch of about 20 amino acids constilinked together in the S-locus. This was indeed tuting a transmembrane bridge joining the found to be true from an analysis of restriction S-domain with the kinase segment (Stein et al 1991). fragment length polymorphism (RFLP) of popula- The amino acid sequence of SRK-6 protein is tions of Bmssica oleracea lines segregating for two entirely consistent with its role as a plasma memdifferent S-alleles (Stein et al 1991). In view of the brane-localized protein kinase mediating cell-tofailure to identify any overlapping regions of DNA cell recognition on the stigmatic surface. The flanking the two genes, there has been some spec- identification of this protein in the stigma memulation that a physical linkage may exist between brane fraction of B. oleracea and the specific targetthem and permit their coexistence in the same ing of the protein to the plasma membrane of locus (Figure 10.3). Confirmation of this came from stigmatic cells of transgenic plants have essentially pulse-field gel electrophoresis combined with confirmed this role (Stein et al 1996). DNA blotting analysis, which showed that in two In its gross features, the molecular structure of S-genotypes studied, SLG and SRK genes are physproteins of the SRK-910 and SRK-AU genes isoically linked in a single genomic fragment and seplated from self-incompatible B. napus lines is simarated by a distance of 220-350 kb (Boyes and ilar to that of the protein of the SRK-6 gene, with Nasrallah 1993). Characterization of SRK genes an extracellular receptor domain homologous to S-locus glycoproteins, a cytoplasmic kinase domain, and a putative transmembrane bridge. SRK SLG SRK-6, SRK-910, and SRK-AU genes also display remarkable specificity in their expression; their transcripts are not expressed in the vegetative parts such as leaves, roots, and cotyledons, but are present predominantly in the pistil and to some extent in the anthers (Stein et al 1991; Goring and Rothstein 1992; Goring et al 1993; Glavin et al Figure 10.3. A diagram showing the evolution of the SLG-SRKgene pair within 5-2 and S-6 genotypes and their 1994). On the other hand, exclusive stigma-specific expression is characteristic of SRK-29 gene (Kumar divergence between genotypes. The predicted SLC-6 and SRK-2 proteins are represented schematically. Numbers and Trick 1994), SRK-3 gene isolated from an S-3 indicate percentage identity between putative domains. C, haplotype of B. oleracea (Delorme et al 1995), and C-terminus domain; JM, juxtamembrane domain; PK, proSRK-9 gene isolated from B. campestris (Watanabe tein kinase domain; S, S-domain; SP, signal peptide; TM, et al 1994). SRK-like genes detected in the S-63 haptransmembrane domain. (From Stein et al 1991.) (iii) A Gene for Pollen Recognition and Signal Transduction
278
Pollination and Fertilization 1.6 kb mRNA
tm
SLG2
X E
B
1.8kbmRNA
2 Xb
\B
HXb
SRK2 1363
2785
tm
3.0 kb mRNA 500 bp
Figure 10.4. Generation of two forms of transcripts by the SLG-2 gene and the relationship between the SLG-2 and SRK-2 genes. Partial restriction map of the two genes and their flanking regions are shown. B, Bamhfl; E, EcoR\; H, H/ndlll; P, Psfl; X, Xho\; Xb, Xba\. The regions of sequence similarity shared by the two genes are indicated by boxes; stippled box, protein coding region; white boxes, shared
sequences located 5'-of the first exon and in the first intron. The black boxes are the mRNAs encoded by the genes; the putative signal sequences are indicated by stripes. The regions corresponding to the two mRNA probes (a) and (b) are shown by bold lines above the SLG-2 gene, tm, transmembrane domain. (From Tantikanjana et al 1993. Plant Ce//5:657-666. ©American Society of Plant Physiologists.)
lotype of B. oleracea display a different expression ancestor (Tantikanjana et al 1993). Characterization pattern, because homologous transcripts are found of SRK proteins of the S-3 haplotype of B. oleracea in both stigma and leaf mRNAs (Oldknow and has revealed the presence of a soluble, truncated Trick 1995). Not surprisingly, sequence data for form of a protein receptor kinase; a clue to the SRK genes from both B. oleracea and B. napus reveal importance of this extracellular protein may lie in a high degree of similarity of the encoded proteins its similarity to animal receptor kinases (Giranton to receptor protein kinase genes ZMPK-1 from et al 1995). maize (Walker and Zhang 1990) and ARK-1 from Part of the current interest in the protein kinase Arabidopsis (Tobias, Howlett and Nasrallah 1992). family of enzymes stems from their potential role With heterologous proteins, homology not only is in mediating the response of cells to inter- and observed in the S- and kinase-domains, but also intracellular signals. These enzymes are either serextends to the transmembrane domain that links ine-threonine specific or tyrosine specific, phosthe two. phorylating serine-threonine or tyrosine, Proceeding from the structure of SLG and SRK respectively. Although most transmembrane progenes, the next problem is to determine the origin tein kinases described so far are of the tyrosine of this gene pair. Often it turns out that the exons of class, there is growing evidence for the view that one gene are related to those of the other gene in a receptor kinases possessing serine-threonine phospair. This might indicate that a common ancestral phorylation function are active in plants. In line gene duplicated to give rise to the two genes, after with this thought, it is significant that SRK genes which differences accumulated between the copies. from B. oleracea and B. napus show sequence homolA prediction that the SLG gene has evolved from an ogy at the terminal end to serine-threonine kinases ancestral SRK-like gene by a duplication event was (Stein et al 1991; Goring and Rothstein 1992). tested by a multifaceted molecular analysis of the Compelling evidence that the products of SRK-6 SLG-2 gene derived from the S-2 line of B. oleracea. and SRK-910 genes are functional serine-threonine A critical observation made in this work was that kinases has been provided by the demonstration the SLG-2 gene generates two forms of transcripts. that protein kinase domains of these polypeptides One transcript, similar to the transcripts from other expressed as fusion proteins in Escherichia coli are SLG-alleles, encodes a secreted glycoprotein, able to autophosphorylate on serine and threonine, whereas the second one is predicted to encode a but not on tyrosine (Goring and Rothstein 1992; membrane-anchored form of SLG (Figure 10.4). Stein and Nasrallah 1993). Sequence similarity between SLG-2 and SRK genes It has become increasingly clear that the degree in the extracellular S-domain as well as, to some of phosphorylation of a particular protein is depenextent, in the cytoplasmic domain is considered dent not only upon the activity of the protein strong evidence for their origin from a common kinases but also upon that of phosphatases. The lat-
Molecular Biology of Self-Incompatibility
ter enzymes exhibit actions antagonistic to the kinases and they dephosphorylate substrates involved in the phosphorylation reaction. As we are now in the era of gene isolation and sequencing, it suffices to state here that a serine-threonine phosphatase sequence showing homology to the rat inhibitor-1 protein has been isolated from a B. oleracea stigma cDNA library (Rundle and Nasrallah 1992). Evidence for the involvement of protein phosphorylation in self-incompatibility is also provided by the observation that okadaic acid, an inhibitor of protein phosphatases, modifies selfincompatibility in B. oleracea (Rundle, Nasrallah and Nasrallah 1993; Scutt, Fordham-Skelton and Croy 1993). This is opposite to what would be expected if a protein kinase were involved in signal transduction, since inhibition of this enzyme would result in increased levels of the phosphorylated substrate. Clearly, much more work is needed before we have a clear understanding of the interaction of phosphatases and SRK in the self-incompatibility syndrome. Binding of a protein to its receptor is a complex phenomenon, and the product of this interaction has important consequences for the cell. Unfortunately, in the realm between the molecular and the cellular, the biology of the transitions from the latter to the former has been little explored, and so the following comments are speculative. By analogy to other receptor kinases, the protein domain encoded by the S-locus would be the receptor that recognizes self-pollen. The receptor protein functions by interaction of a ligand with the extracellular domain of SRK protein, and this triggers the protein kinase activity of the cytoplasmic domain. On the basis of the biochemical findings, it seems entirely possible that the protein kinase then phosphorylates specific substrates; this phosphorylation leads to a chain of events, such as the production of second messengers, amplification of the signal, and alteration or activation of cellular processes. A very important question for our understanding of the dynamics of the recognition reaction is the molecular interaction between SLG and SRK protein. In addition, the plausibility of any hypothesis involving signal transduction depends upon hard evidence that identical elements are active on the stigma surface and on the pollen grain. (iv) 5-Locus Genes in Anthers and Pollen Grains
The closest we have come to demonstrating the presence of proteins with homology to SLG epitopes in anthers and pollen grains of self-incom-
279 patible plants is by the use of transgenic plants (see pp. 281-284) and by immunolocalization and PCR amplification techniques. Watanabe et al (1991) used a polyclonal antibody for SLG from the S-8 line of B. campestris in blots of stigma and anther proteins and showed that the antibody cross-reacts with proteins from stigmas and anthers of S-8, S-9, and S-12 homozygotes. Based on electrophoretic mobility and lack of cross-reactivity with peroxidaselabeled con-A, the protein displays properties different from the stigma SLG described earlier. Its pattern of accumulation - starting in anthers more than 10 days before anthesis and reaching a maximum in anthers at the bicellular pollen grain stage also seems to set it apart from the stigma SLG. In mature anthers, it is present in the anther walls rather than in the pollen grains. All told, it would appear that this protein is a different kind of recognition molecule, neither identical to SLG nor very different from it. Amplification of SLG gene sequences from anther cDNAs of the S-9 homozygous line of B. oleracea has been made possible by PCR reactions with primers synthesized to the S-6 cDNA sequence. The presence of mRNA sequences homologous to the DNA sequence of the SLG clone in anthers during the unicellular and bicellular pollen grain stages seems to fit the prediction that the DNA sequence of the SLG gene may be transcribed in the anthers (Guilluy et al 1991). However, because of the complex hybridization pattern seen in Northern blots and the heterogeneous nature of S-related genes in B. oleracea, it does not necessarily follow that the PCR products obtained are specific to anther SLG genes. Further evidence for the presence of an Slocus-encoded signal emanating from self-pollen is derived from the analysis of a gene cloned from the S-2 haplotype of B. oleracea. This gene, designated as S-LOCUS ANTHER (SLA), is physically linked to the SLG gene and is transcribed from two promoters to produce a spliced and an unspliced transcript. The transcripts of the SLA gene are specific for the pollen grains, where they are expressed differentially, those of the spliced one appearing early in pollen development and the unspliced one restricted to mature pollen grains. Not surprisingly, the transcripts are absent in a self-compatible B. napus carrying a haplotype like S-2 (Boyes and Nasrallah 1995). As a result of these and other findings, attention might now turn from the pistil proteins to the pollen proteins to determine whether the latter molecules enable the stigmatic papillae to distinguish between self- and nonself-pollen grains.
280
Pollination and Fertilization
stigmatic transcription of these genes (Nasrallah, Kandasamy and Nasrallah 1992). If this hypothesis proves to be correct, a pistil-specific trans-acting It is not uncommon to find self-compatible factor may hold the clue to the functioning of the strains of Brassica oleracea in commercial stocks or promoters of these genes. in well-defined genetic populations; in these In many ways the genetics and molecular biolplants, self-pollen grains escape from inhibition ogy of self-compatibility in a line of B. oleracea and produce normal pollen tubes on the stigma appear very different from these aspects of the lines (Thompson and Taylor 1971; Nasrallah 1974). considered earlier. The main difference is in the Except that the breakdown of self-incompatibility presence of SLG not only in the stigmas of selfis ascribed to a single suppressor or modifier gene compatible plants, in a spatial and temporal patunlinked to the S-locus, for the most part the pri- tern similar to that found in the stigmas of mary genetic or molecular effects of the mutation self-incompatible ones, but also in the stylar transare not known. A discussion of the disruption of mitting tissue. Moreover, the SLGs segregate with incompatibility is included here as a reminder that the self-compatibility phenotype in the F2 progeny; there is more to the molecular biology of this phe- the significance of this observation is the implicanomenon than what is presented by the analysis of tion that self-compatibility may be due to changes plants habitually displaying self-pollen rejection. at the S-locus. These data are also intelligible in terms of the generalization that the self-compatible The interesting example analyzed by Nasrallah (1974) concerns a self-compatible line of B. oleracea phenotype is in some ways linked to the S-locus. A presumably derived from an incompatible line by a central feature of the work on the self-compatible spontaneous mutation referred to as su. The genetic line of B. oleracea is the demonstration that the basis of this mutation was demonstrated in crosses amino acid sequence deduced from a cDNA clone involving an inbred strain and the self-compatible encoding the self-compatibility gene is highly mutant combined with the assay of antigens in homologous to the SLG. Collectively, the results homogenates of the stigmas of the different geno- would argue that the breakdown of self-incompattypes (Nasrallah and Wallace 1967a). The biochem- ibility is due to changes at the S-locus or due to the ical effect of the mutation is a reduction in the action of a modifier gene closely linked to this locus amount of immunologically detected S-proteins in (Gaude et al 1993,1995; Kleman-Mariac et al 1995). the stigmas of the compatible plants as compared Evidence for a passive role for SLGs in self-comto the amount of proteins in incompatible ones patibility has come from the observation that the (Nasrallah 1974). Descriptively, the hospitality of rate of transcription and translation of the SLG the stigma to all pollen grains, including self- gene in two additional self-compatible strains of B. pollen, is also the characteristic feature of another oleracea is comparable to that observed in the selfspontaneous mutation (scf-1) identified in a rapidly incompatible strain. Here, a mutation that impairs cycling B. campestris. So far as the molecular effects SRK function has been found to result in the breakof the mutation are concerned, not only does this down of the self-pollen rejection syndrome mutant have a drastically reduced level of SLG, but (Nasrallah, Rundle and Nasrallah 1994). SLG, SLR-1, and SLR-2 genes are coordinately Other investigators have obtained results that down-regulated at the RNA level. At the same indicate that self-compatible lines of B. napus carry time, expression of the SRK gene very much resem- a normal SLG gene and that its transcripts are bles that of the wild type. At the cytological level, expressed at levels comparable to those of SLG correspondingly low levels of SLG and SLR pro- genes from self-incompatible lines (Goring et al teins are detected on the walls of the stigmatic 1993; Robert et al 1994a). One self-compatible line papillae. On the whole, the hypothesis that all even contains disproportionately large amounts of members of the S-multigene family participate in SLR glycoprotein relative to SLG (Hiscock et al the operation of self-incompatibility in B. campestris 1995). The problem of determining the possible finds strong support in the effects of the mutation causes for self-compatibility in one of the lines was on the down-regulation of the three genes and on unexpectedly simplified by the discovery that the the breakdown of self-incompatibility. Quite apart coding region of the SRK gene from the self-comfrom genetic considerations, it has been hypothe- patible lines suffers a 1-bp deletion in the SLG gene sized that the mutation interferes with the func- homology domain when compared to other SRK tioning of a regulatory gene that possibly encodes a genes. Due to this frame-shift mutation, in the nortrans-acting factor necessary for the high level of mal sequence of synthesis and assembly of the gene (v) Analysis of the Breakdown of SelfIncompatibility
Molecular Biology of Self-Incompatibility
281
clonal antibodies specific for a protein epitope of the SLG gene has revealed accumulation of proteins in the walls of the papillar cells of the mature stigma of B. oleracea. The pattern of accumulation of SLG in the stigmatic papillae is diagnostic of the development of incompatibility in the stigma, because papillar cells of young self-compatible stigmas are devoid of the protein. At the subcellular level, labeling in the (vi) S-Cene Expression in Normal and cells of the stigmatic papillae is associated with the Transgenic Plants rough ER and Golgi vesicles, suggesting that SLG is Since the S-gene plays a critical role in pollen synthesized in the ER and secreted into the papillar rejection at the stigmatic surface during self-incom- walls by dictyosomes (Kandasamy et al 1989; patibility, it is logical to inquire whether its pattern Kandasamy, Parthasarathy and Nasrallah 1991). The of expression reflects the functional activities of the question as to whether the other related glycoprostigmatic cells. A coincidence of messenger local- teins encoded by SLR-1, SLR-2, and SRK genes are ization in the stigmatic papillae in preference to also present in the stigmatic cells is obviously other cells would confirm a role for the gene and its important in the consideration of specificity of the protein product in pollen rejection at the papillae. incompatibility reaction. Immunolocalization with In a similar way, transformation with SLG genes polyclonal antibodies raised against a fusion protein from Brassica of plants that display gametophytic between SLR-1 coding sequences and the Escherichia self-incompatibility or that are self-compatible coli tryptophan biosynthesis gene, TRP-E, has shown would allow a comparison of the sites of expres- that SLR-1 protein is expressed in the cell walls of sion of the incompatibility genes in the two sys- the stigmatic papillae exactly in the same spatial and tems. On the basis of molecular data discussed temporal configuration as is SLG (Umbach et al earlier and the cytological findings to be consid- 1990); the similarity between the expression patterns ered now, it has been unambiguously established of these glycoproteins is provocative in view of the that the stigma bears S-locus-associated determi- corresponding homologies in their protein sequences. The cytochemical distribution of SLR-2 nants of recognition. and SRK proteins remains to be studied. In situ hybridization of SLG-6 probe to longitudinal sections of the stigma and style of B. oleracea A series of investigations into SLG gene regulahas demonstrated an exclusive localization of tran- tion in transgenic tobacco and Brassica sp. has proscripts in the stigmatic papillae. Indeed, the vided further evidence of the correlation of cellular intensely labeled cells of the stigmatic papillae, and developmental expression of the gene with the which appear as a crown over the unlabeled basal site and timing of pollen inhibition in the recipicells and the cells of the transmitting tissue of the ents. In one version of the experiment, tobacco was style, provide a striking visual representation of transformed by the SLG gene of the S-13 haplotype this experiment. Hybridization of sections of the of B. oleracea, which comprises the coding region stigma and style at different stages of development flanked by approximately 4.2 kb of upstream, and has shown that the intensity of signals in the papil- 5.0 kb of downstream, genomic sequence. The most lae is correlated with the development of incom- general change observed in the transformants was patibility, because no hybridization is detected in expression of the SLG gene at the RNA and protein the cells of the immature stigma and a nearly 10- levels in the stigma and style extracts (Moore and fold increase occurs in the cells of the mature Nasrallah 1990). In some ways the developmental stigma one day prior to anthesis (Nasrallah, Yu and expression of SLG and its gene transcripts in the Nasrallah 1988). The spatial and temporal expres- transformed plants was similar to that seen in the sion of SLR-1 gene transcripts harmonizes with self-incompatibility phenotype of B. oleracea, and in that of SLG transcripts and, like the latter, SLR-1 other ways it was close to that of self-incompatible messages are concentrated in the papillar cells, rel- Nicotiana alata, but broad correlation was noted ative to the remainder of the stigma, prior to and between the maximum levels of SLG and its tranimmediately after anthesis (Lalonde et al 1989). scripts and the onset of incompatibility (Nasrallah Part of the incompatibility program of the pistil et al 1985; Anderson et al 1986). Analogy to N. alata preparatory to receiving the self-pollen is the accu- was seen in immunolocalization, which revealed mulation of receptor proteins in the appropriate that the protein product of the SLG gene was cells. Indirect immunogold labeling with mono- expressed in the stylar transmitting tissue of trans-
product, a termination of translation and the production of a truncated SRK protein that may have no relevance to signal transduction or self-incompatibility can be expected (Goring et al 1993). Alternative explanations are necessary to account for self-compatibility in other lines of B. napus.
282
Pollination and Fertilization
transcripts in the mature stigma of N. alata and traces the route taken by compatible pollen tubes during their growth to the ovary (Cornish et al 1987). Nishio et al (1992) have shown that self-compatible B. napus plants transformed with several SLG genes isolated from B. oleracea and B. campestris
exhibit, at a reduced level, the same pattern of developmentally regulated and cell-specific gene expression as they do in the self-incompatible phenotype. Despite the clear evidence from these experiments for the similarity of the ris-acting elements of the donor and recipient plants, the expression of SLG genes, interestingly enough, failed to impart self-incompatibility in transgenic plants in the experiments just described. An instance in which introduction of an SLG gene perturbed selfincompatibility and induced full self-compatibility in the transgenic recipient was described by Toriyama et al (1991a) following transformation of an incompatible line of B. oleracea with the S-8 allele of B. campestris SLG gene. Introduction of antisense RNA of the same gene led to the synthesis of SLG and the breakdown of self-incompatibility in transgenic B. campestris (Shiba et al 1995). In these experiments, the effect of the regular gene is less easy to explain than the effect of the antisense gene; the breakdown of the self-incompatibility system has been attributed to an unexpected interference by the antisense gene in the expression of the homologous host gene, which leads to a decrease in the levels of endogenous SLG. There is little doubt that other side effects of the antisense gene will be described in the future. Figure 10.5. An immunostained, longitudinal section of the stigma and style of transgenic Nicotiana tabacum showing the distribution of an antigen encoded by Brassica oleracea SLC-22 gene. Scale bar = 100 urn. (From Kandasamy et al 1990. Plant Cell 2:39-49. © American Society of Plant Physiologists. Photograph supplied by Dr. M. Kandasamy.)
genie plants and not in the stigmatic papillae (Figure 10.5). Similar results were obtained in tobacco transformed with the SLG-22 gene in another version of this experiment, which also defined the spatial pattern of expression of the gene. The transgene-encoded protein product was localized in high concentrations in the style in the intercellular matrix of the transmitting tissue, with some label spilling over into the secretory matrix in the basal region of the stigma and into the placental epidermal cells adjoining the ovules (Kandasamy et al 1990). Localization of the SLG22-encoded protein in transgenic tobacco coincides precisely with the expression of tobacco S-allele
Although models of self-incompatibility predict the presence of S-locus components in the pollen, stigma, and style, it has not been established from studies reviewed here that pollen grains express Slocus transcripts or their protein products. On the surface, these results suggest that S-locus genes are very tightly regulated or are present in very low abundance in pollen grains. One strategy employed to compensate for the low abundance of putative SLG transcripts is to use the SLG promoter to activate very sensitive toxic genes or reporter genes in transgenic plants. In experiments with the heterologous host tobacco, activity of a promoter fragment isolated from the SLG-13 gene was followed by fusing it to the DT-A gene; in parallel experiments, the promoter fused to the GUS gene was used to characterize the sites of gene expression by a standard histochemical assay. Since the DT-A gene is a potent inhibitor of protein synthesis and causes autonomous cell death, the effect of gene action was a genetical wipe-out of cells
Molecular Biology of Self-Incompatibility
expressing the fusion gene. This catastrophic though specific effect of the DT-A gene caused seemingly obvious ablation of stylar tissues in most transformed plants and led to stunted growth of the style with a concomitant disruption of the transmitting tissue. The effect of the toxic gene was also targeted at the normally bilobed stigma, which became multilobed with abnormal papillae. Generally, the fusion gene was not fatal to the cells of young buds, but the abnormalities became severe in open flowers, so that the pistils were incapable of supporting pollination and pollen germination. Consistent with the gametophytic expression of the fusion gene, 50%-75% of pollen grains in the transgenic plant were empty and nonviable (Thorsness et al 1991). Altogether, it is not surprising that DT-A gene activity primed by SLG gene promoter affects those very cells and tissues that favor pollen capture, pollen germination, and pollen tube growth. Similar phenotypic effects, confined to ablation of the stigmatic papillae and abnormal pollen development, are observed with impressive regularity in Ambidopsis and in self-fertile B. napus plants transformed with the same fusion gene. However, the effects of the fusion gene on the functional competence of the stigmatic papillae to support pollen germination and pollen tube growth are completely different in the two transformants. For example, a consequence of the toxic gene on Arabidopsis is the failure of self-pollen to germinate on the ablated papillar cells, whereas these same cells support germination and growth of pollen from untransformed plants; in contrast, ablation of the stigmatic papillae of B. napus almost completely blocks the ability of pollen grains of any genotype to germinate (Kandasamy et al 1993; Thorsness et al 1993). Apart from the obvious implication of these results that B. napus requires active papillar cell metabolism for successful pollination and Arabidopsis does not, the results underscore the complexity of the pollination system, even in self-fertile B. napus, and indicate an important function for the papillae in the initial recognition event. GUS gene expression in tobacco plants transformed with SLG-13/DT-A constructs has resulted in the cytological detection of promoter activity at a high level in the stigmatic papillae and transmitting tissue of the style and at a reduced level in the placental epidermis of the ovary (Figure 10.6). This is just what one would expect in gametophytic selfincompatibility systems in which stylar inhibition of pollen tube growth is the norm. In the male
283
gametophyte, GUS activity is expressed at the uninucleate microspore stage and it increases up to the binucleate stage. The percentage of pollen grains in the anther locule showing GUS activity is consistent with the gametophytic, rather than with the sporophytic, mode of self-incompatibility (Thorsness et al 1991). The spatial and temporal pattern of expression of GUS gene directed by a B. oleracea SLR-1 gene promoter in transgenic tobacco is similar to that observed for SLG-GUS fusion gene. Unlike SLG-GUS fusion gene, SLR-1-GUS construct was not expressed in the placental epidermis of the ovary; despite this minor difference, the combined results suggest a close interaction between the different S-locus genes in pollination (Hackett, Lawrence and Franklin 1992). When SLG-13-GUS construct was introduced into Arabidopsis, the pattern of expression of the GUS gene was somewhat like that of the SLG-13 gene in untransformed B. oleracea (Figure 10.6). Activity was observed in the stigmatic papillae, but not in the underlying stylar tissue. Some transformants also showed activity in the tapetum during a narrow developmental window confined to the Brasska
Micrtxporo
SligmiPipilbe
TYansgenic Nicotiana
TYansgenic Arabidopsis
Figure 10.6. Diagrams showing the tissue specificity of SLG-13 promoter activity in transgenic Brassica napus, Nicotiana tabacum, and Arabidopsis. Shaded areas indicate regions of expression. Patterns of expression seen are, in transgenic 6. napus: stigmatic papillae, tapetum, pollen grains, and, to a lesser degree, the stylar transmitting tissue and placental epidermis; in transgenic N. tabacum: stigmatic papillae, transmitting tissue, placental epidermis, and 50% of pollen grains; and in transgenic Arabidopsis: stigmatic papillae and tapetum. (From Dzelzkalns, Nasrallah and Nasrallah 1992.)
284
Pollination and Fertilization
-415 -339 -291 •206 -163 -143 -117 -79 ATO , period immediately after microsporocyte meiosis, H H 1 1— 1 r— but pollen grains did not express the reporter gene H (Toriyama et al 1991b). This observation is in harmony with the view that in plants with the sporophytic type of self-incompatibility, S-locus expression occurs postmeiotically in the tapetum. Proximal Pollen Element Although the main outlines of expression of the SLG-GUS fusion gene in transformed self-compatible B. napus and self-incompatible B. oleracea plants Figure 10.7. A diagram based on deletion analysis showing the arrangement of the functional elements of the SLC-13 are similar to those described in transformed promoter. The deletion end points of the constructs are Arabidopsis, two unexpected findings have put the shown at the top. The positions of the stigma-style domain, relationship between gametophytic and sporo- the two pollen elements, and the pollen silencer element phytic types of self-incompatibility in a new per- are shown relative to the positions of the sequence motifs spective. One is that the SLG-GUS fusion gene is found to be highly conserved in SLC and SLR genes (boxes 1 to 5). (From Dzelzkalns et al 1993. Plant Cell 5:855-863. expressed not only in the stigmatic papillae, but © American Society of Plant Physiologists.) also to a reduced extent in the stylar transmitting tissue and ovary of the transformed plants. The other is that the SLG promoter is active not only in the tapetum of the anther, but also in the haploid 2. THE NICOTIANA SYSTEM microspores (T. Sato et al 1991). Although the hisThe discovery of an abundant glycoprotein with tory of contention about the gametophytic and a molecular mass of 32 kDa in the style of Nicotiana sporophytic types of self-incompatibility actually alata carrying S-2 alleles was among the had to do with the timing of gene action, comments antecedents of attempts to isolate the S-2 gene have also appeared periodically about the origin of product and to clone the gene encoding the protein. the two systems. Results from transgenic Brassica The natural assumption that the protein segregates plants showing features of a pollen recognition sys- with S-2 breeding behavior was confirmed by simtem based on gametophytic control in sporophyti- ple one-dimensional SDS-PAGE of stylar extracts of cally self-incompatible plants might support the S-2 genotypes. In addition, expression of this proview that the gametophytic self-incompatibility tein with the development of self-incompatibility system is the ancestral type and that the sporo- in maturing styles and its concentration in the part phytic system is derived from it. of the style where inhibition of pollen tube growth
The ability to break sequential parts of the promoter and to join them to a reporter gene, or to place promoter modules under the appropriate control so that they can be transcribed from a new promoter, allows one to determine the minimum sequence elements of the SLG promoter necessary for gene expression (Figure 10.7). Expression patterns of truncated versions of SLG-13 promoter fused to GUS have shown that a 411-bp promoter fragment adjoining the translation initiation codon directs a normal level of GUS activity in the stigma and style and a reduced activity in the pollen grains of transgenic tobacco plants. A 196-bp region (-339 to -143 bp) is also sufficient to confer independent stigma- and style-specific expression at a reduced level. Although the 196-bp fragment does not direct expression in the pollen grains, two distinct but functionally redundant flanking domains (-415 to -291 bp, and -117 to +8 bp) allow expression of the gene in the pollen (Dzelzkalns et al 1993). It is clear that even when large fragments of the truncated promoter are expressed, selective mechanisms could operate to reduce the expression to unimpressively low levels in some tissues.
occurs also seem to support the notion that the 32kDa protein is a product of either the S-2 allele or a gene tightly linked to it. In the approach followed to clone the gene encoding the self-incompatibility glycoprotein from the stigma-style complex of N. data, the protein was isolated, purified, deglycosylated and its N-terminal sequence determined. From this sequence, a partial 30-bp DNA probe for the encoding gene was synthesized to screen a cDNA library prepared to poly(A)-RNA of mature styles of an S-2/S-3 genotype of N. alata. Following differential screening with labeled cDNA made to the mRNA of mature styles, of immature styles and of ovaries, cDNA clones highly expressed in mature styles were further screened with the 30-bp synthetic oligonucleotide. A positive cDNA clone obtained by this method was used to screen another cDNA library to isolate a full-length cDNA clone of the glycoprotein. Sequencing of this cDNA clone revealed an open reading frame of 642 bp that encodes a 204-residue polypeptide with a classic signal sequence of 22 amino acids and a hydrophobic N-terminus of 15 amino acids. Other features of the predicted amino acid sequence of
Molecular Biology of Self-Incompatibility
the protein are consistent with the view that it is a secreted glycoprotein whose transfer is directed by a signal sequence. In Northern blot hybridization, the cDNA clone was found to cross-hybridize intensely with RNA from mature styles bearing the S-2 allele and weakly with RNA from nonhomologous alleles. No hybridization was detected with RNA of immature styles of any genotypes (Anderson et al 1986). Taken together, these data provide persuasive evidence that the cDNA characterized from N. alata encodes a protein product of the S-2 allele. Different approaches to screening cDNA libraries have been employed to isolate clones corresponding to the glycoproteins from other S-locus alleles of N. alata. Following differential screening of cDNA libraries with probes prepared from mRNA of S-3 and S-6 genotypes, potential S-3 and S-6 cDNA clones were identified by their ability to hybridize strongly to S-3 and S-6 probes and weakly to S-6 and S-3 probes, respectively (M. A. Anderson et al 1989). To isolate cDNAs from Sz, Sf11, S-2, and Sa alleles, a cDNA library made to Sz pistil mRNA was differentially screened with pistil and leaf mRNA probes. A single clone representing the most abundant class of pistil-specific cDNA was used as a hybridization probe to screen cDNA libraries made to mRNA from S/-22, S-2, and Sa genotypes (Kheyr-Pour et al 1990). A critical question about the deduced amino acid sequences of the alleles is the extent of variation and homology. Although the incompleteness of some of the cDNA clones has rendered estimates of amino acid sequences of limited accuracy, comparison between S-2, S-3, and S-6 sequences has revealed about 56% overall homology at the amino acid level. The average homology was reduced to 54% when amino acid sequences of all seven S-locus alleles were compared. Variations are mainly due to amino acid substitutions, deletions, and insertions found throughout the sequence, but many of these are seemingly bunched together in four hydrophilic regions confined to the amino half of the sequence. The amino terminus, the cysteine residues, the multiple glycosylation sites, and the carboxy terminus are generally considered to be conserved regions and they account for the other half of the sequence. S-l, S-3, and S-6 proteins have 10 cysteine residues, whereas others, except Sa, have the traditional 9 residues (Figure 10.8). The Sa protein has only 7 of the 9 conserved cysteine clusters. When glycosylation sites are compared, 4 sites are found to be conserved among proteins of S-2, S-3, and S-6 alleles, and the S-3 allele has gained a new site (M. A. Anderson et al 1989; Woodward et al
285 -22 -1 l KFKSOLTSVLFIVLSVLSPTT& MLNSDtTSAVLFLLFALSPIYV MNKSrvTSVLLIFLFTLSPISG MSK50LTSVFFILLCALSPIYG KFNLFLTSVFVIFLrALSPIYG
s
• A A A
8 SF S, Sa S, S2
TNX.GW r P.
f ; L I U G I C F >SKVKNVIDCPHPKTCI< 'HCNRCIK "P.
CELttMSICFDSTAimVIDCPMPICTCKffASWtlGlft-p. BIO C5« O OOOOO* COB*O HV5 K »
Figure 10.8. Alignment of the amino acid sequences of seven S-proteins of Nicotiana alata. The five conserved (C1 to C5) and five hypervariable (HV1 to HV5) regions are enclosed in solid lines. Small boxes within the conserved regions are residues that are different from the conserved residues at those sites. Open circles enclose residues that are different between the two homology classes but are conserved within one or two homology classes. The nine cysteine residues are indicated by asterisks. The circled asparagine residues are potential N-glycosylation sites. (From Kheyr-Pour et al 1990. Sex. Plant Reprod. 3:88-97. © Springer-Verlag, Berlin.)
1989; Kheyr-Pour et al 1990)). The occurrence of conserved cysteine residues and glycosylation sites in the alleles suggests specific structural roles for these molecules in the stability, conformation, and biological activity of the glycoproteins. Is the SLG gene of N. alata a member of a multigene family? When Southern blots of endonuclease-digested genomic DNA from S-2, S-2, S-3, S-6, and S-7 allelic plants were probed with cDNAs corresponding to the alleles, the fragments hybridizing from different alleles were found to be dissimilar in size, and, in most cases, single diagnostic restriction fragments hybridized intensely with the probe. Each cDNA probe bound strongly to a DNA fragment of that allele and weakly or not at all to DNA fragments of other S-alleles. These
286
observations are compatible with the view that the isolated cDNA clones are allelic products of the same S-locus and that this is a single-copy genetic system. Additional data confirming that the cDNA sequences are allelic products of the same S-locus have come from breeding experiments combined with DNA blot analysis, two-dimensional electrophoresis, and cation exchange liquid chromatography of style glycoproteins. It was found that the S-genotypes of individual plants determined from conventional breeding tests correlate precisely with their DNA blot analysis using S-allele-specific DNA probes and with the presence of S-allele-specific glycoprotein spots and protein elution profiles (Bernatzky, Anderson and Clarke 1988; M. A. Anderson et al 1989; Jahnen et al 1989). Of further interest is a report that a small fragment located at the 5'-end of the coding sequence of S-2 DNA shows homology with mitochondrial DNA in Southern blots (Bernatzky, Mau and Clarke 1989). How the transfer of sequences between nuclear and mitochondrial DNA might have occurred, and what the role of this sequence is in the function and expression of self-incompatibility, remain to be explored. (i) Ribonucleases as S-glycoproteins A feature of the amino acid sequences of S-glycoproteins isolated from N. alata was their homology to extracellular fungal RNase. The homology itself was not particularly striking, as only 30 of the 122 amino acids conserved among S-2, S-3, and S-6 allelic glycoproteins were aligned with identical molecules in the fungal enzyme. Another 22 amino acids in the glycoproteins aligned with closely related amino acids in the RNase, and 5 out of the 9 or 10 cysteine residues and 2 catalytic histidine residues were conserved between the two proteins. Nonetheless, this finding led to intensive research on the RNase activity of glycoproteins and eventually to the identification of glycoproteins as RNases. That RNase activity is an inherent component of the S-glycoproteins became clear from the fact that the enzyme is found to co-elute perfectly with the individual S-glycoproteins by ion-exchange chromatography and that it has a specific activity close to that of purified RNase T2. In physiological terms, the specific activity of the enzyme from stylar extracts of the self-incompatible N. alata is found to be 100-1,000-fold higher than that of the self-compatible N. tabacum. This correlation implies a possible role for the stylar product of the self-incompatibility gene, designated as S-RNase, in self-pollen rejection reactions (McClure et al 1989). S-RNase also
Pollination and Fertilization
appears to have a bearing in pollen rejection associated with unilateral incompatibility in certain species combinations in Nicotiana. This was shown by up-regulating RNase expression in transgenic plants that acquire the capacity to reject normally compatible pollen grains (Murfett et al 1996). A general mechanism by which RNase might function in the self-incompatibility response is by the degrading of pollen RNA, which results in pollen tube death in an incompatible pollination. Support for this view has come from the work of McClure et al (1990), who followed the integrity of RNA extracted from the style of N. alata of a defined genotype after compatible and incompatible pollinations with 32P-labeled pollen grains. The results were straightforward in showing extensive RNA breakdown in incompatible styles but not in compatible ones (Figure 10.9). Furthermore, the results suggest that pollen tube rRNA could have suffered degradation in the incompatible style due to RNase cytotoxicity, because RNA appeared as a Style Fraction RNAs
Stigma Fraction RNAs
op op op cp op op op cp op op op dp epopofcffcpopofof *op*op*ofvof * * * * * * * * op op op op op op op op op op op op of op op of of-op op op of op op of -28S -18S
#
Equal Counts Loaded
Equal RNA Loaded
Equal Counts Loaded
Figure 10.9. Agarose gel electrophoresis of 32P-labeled RNA from compatible and incompatible crosses in Nicotiana alata. RNAs from style and stigma parts of the flower were treated separately. Compatible crosses represented are 5,53 X S2S2 and S2S2 X 5,S3; incompatible crosses represented are S2S2 X S2S2 and S^S^ X S.Sy Mostly intact RNAs are seen in the styles of compatible crosses and in the stigmas of all crosses, whereas RNAs of the styles in the incompatible crosses are degraded, which results in the accumulation of low-molecular-mass species. (From McClure et al 1990. Reprinted with permission from Nature 347:757-760. © Macmillan Magazines Ltd. Photograph supplied by Dr. A. E. Clarke.)
Molecular Biology of Self-Incompatibility
smear of low-molecular-mass species in an agarose gel run under denaturing conditions. Results from other experiments have been marshalled to show that the inhibition of pollen tube growth in the style results from the entry of S-RNase and the consequent degradation of pollen tube RNA. For example, following autoradiography of 3H-labeled S2-RNAse incorporation, the label is present in the pollen grain cytoplasm and, to a lesser extent, in the pollen tube. Immunogold electron microscopy using an antibody raised to a synthetic peptide corresponding to an allele-specific region of S2-RNase showed labeling in the walls of the germinated pollen grain and in the pollen tube cytoplasm. Treatment of pollen grains with S2-RNase inhibited pollen tube growth and, concomitantly, protein synthesis. S2-RNase also inhibited cell-free translation of proteins by self-pollen RNA to about 70% of the control rate (Gray et al 1991). In the absence of data on the comparative effects of S2-RNase on compatible pollen grains, the significance of these results leaves much to be desired. Despite this criticism, the results lead to the inference that RNA degradation is instituted soon after the incompatible pollen tubes reach the style, as a result of which their growth is stymied. Analyses of S-RNases recovered separately from the stigma and style have shown that the two enzymes are identical, although their respective transcripts differ in the length of the polyadenylate tail. This has raised the obvious question of why pollen tube growth is arrested in the style rather than in the stigma after self-pollination. Different mechanisms of enzyme uptake by the pollen tube from the stigma and stylar regions might underlie their differential sensitivity, but this remains to be further investigated (McClure et al 1993). In one working model proposed to explain the operation of the system, it is believed that in an incompatible pollination, pollen tubes carry S-locus-specific receptors that bind SRNases. How these molecules find their way into the pollen tube is a mystery, but once inside, they decimate the endogenous RNA of the pollen tubes and inhibit their growth. Compatible pollen tubes survive the RNase onslaught since they have some way of preventing the uptake of the enzyme. Alternatively, it has been postulated that RNases are taken up by pollen tubes regardless of their specificity but, inside the pollen tube, specificity of action prevails due to an intracellular inactivation of the enzyme or a compartmentalization of the enzyme that makes the RNA pool inaccessible (Franklin-Tong and Franklin 1993; Dodds, Clarke and Newbigin 1996).
287
Attractive though these models are, the mechanism by which allelic specificity is controlled during the incompatibility episode remains unclear. Since the gametophytic self-incompatibility system is based on the expression of the S-gene in the pollen as well as in the style, the problem is compounded by the fact that S-gene products of the pollen have not been identified. The dilemma has been heightened with the discovery that the SRNase gene is also expressed at a low level in the pollen grains of N. alata; it is difficult to visualize an incompatible reaction as a result of the interaction of two identical molecules (Dodds et al 1993). Based on the fine structural analysis of S-glycoproteins, N-linked carbohydrate chains are thought to be involved through posttranslational modification in the function of the S-alleles (Woodward et al 1992). This implies that the choice between the fates of S-alleles in the pollen grain and style is predetermined rather than stochastic. From observations showing that S-RNases are effective substrates for protein kinases of pollen tubes, a role for phosphorylation of S-RNase involving Ca2+-dependent protein kinases in the signal transduction pathway of self-incompatibility appears as a distinct possibility (Kunz et al 1996). (ii) Expression of 5-Locus Glycoprotein Genes
An important aspect of the cloning of S-locus glycoprotein genes from N. alata concerns their temporal and spatial appearance. At what period are the transcripts of the gene expressed relative to the stage of floral development at which overt selfincompatibility syndromes appear? Does the accumulation of transcripts coincide with the site of pollen tube inhibition? Fortunately, these questions have generated complementary sets of data that indicate that the pattern of gene expression reflects the age of the flower when it becomes self-incompatible and the age of the tissues where the reactions occur. From Northern blot analysis of stylar RNA from flower buds of different ages probed with S-2 cDNA, it would appear that gene expression is most intense in mature styles, weak in the ovary, and absent in immature styles. At the cellular level, the S-2 gene is expressed in the stigmatic papillae and stylar transmitting tissue of mature flowers as a pathfinder for the pollen tube before it succumbs, but the gene is not prevalent in the immature stigma-style complex before it becomes self-incompatible. The weak expression of the gene detected in the ovary in Northern blot is seen by in
288 situ hybridization as labeling in the epidermis of the placenta. The epidermal cells are probably engaged in arresting the growth of self-pollen tubes that escape inhibition in the style (Anderson et al 1986; Cornish et al 1987). Immunocytochemical studies have shown that S-glycoproteins accumulate in the same tissues in which the gene transcripts are found (M. A. Anderson et al 1989). In an approach using transgenic plants to localize the sites of S-gene expression, self-compatible N. tabacum plants were transformed with constructs encoding the S2-glycoproteins/ and gene expression was followed by protein gel blot and Northern blot analyses and by immunolocalization. The aspects of gene expression in transgenic tobacco plants that can be correlated with self-incompatibility in N. alata are the localization of S2-glycoproteins in the stylar transmitting tissues, an increase in the level of the glycoprotein with maturity of the style, and RNase activity of S2-glycoproteins. However, no change in the self-incompatibility phenotype of transformants was observed as they spontaneously self-pollinated (Murfett et al 1992). In another work, prompt rejection of self-pollen was seen when a high level of RNase expression was engineered in transgenic self-compatible hybrids between N. alata and N. langsdorfii. Conversely, hybrids of N.
Pollination and Fertilization
and P. inflata. In studies of P. hybrida, cDNA libraries made to mRNA of mature styles of different S-alleles were screened with an oligonucleotide homologous to the conserved N-terminal region of N. alata glycoprotein to generate style-specific clones. These clones reacted specifically with RNA from the style, where the levels of mRNA homologous to the cloned cDNAs increased dramatically with the transition of the flower from self-compatibility to self-incompatibility. The clones were of low copy number and, consistent with their allelic nature, hybridized to polymorphic DNA restriction fragments (Clark et al 1990). A cDNA clone encoding S-locus glycoprotein in a S-2 homozygote of P. inflata was selected from a cDNA library using N. alata cDNA as a probe; subsequently, cDNA clones corresponding to the glycoproteins of other S-locus alleles of P. inflata were obtained by probing with S2 cDNA clones (Ai et al 1990). Antisense technology has been successfully used to demonstrate a correlation between inhibition of S-protein synthesis in the style and enforced breakdown of self-incompatibility in transgenic P. inflata (Lee, Huang and Kao 1994).
Combining the predicted amino acid sequences of SLGs of P. hybrida and P. inflata allows the following conclusions to be drawn about their plumbaginifolia X N. alata transformed with anti- sequence conservation and variability. Like N. alata sense construct targeted to the S2-RNase gene of N. glycoproteins, Petunia proteins are about 220 amino alata had low RNase activity and were compatible acids in length and include a 20-amino acid signal with N. alata pollen. These ingenious crosses were peptide characteristic of secreted proteins. The necessary to overcome the problems of the low sequences reveal a general pattern of several highly expression of S-RNase genes in transgenic N. alata conserved domains, interspersed by regions of and the poor regeneration of this species in tissue sequence variability. Among the individual culture (Murfett et al 1994, 1995; Murfett, Bourque residues that are highly conserved are eight cysand McClure 1995). Such problems will no doubt teine clusters. Although SLGs of N. alata have sevbe examined critically as more S-RNase expressions eral N-glycosylation sites, S-glycoproteins of are monitored in transgenic plants with the insight Petunia have only one or, at best, two conserved gained from the knowledge of the structure of residues that constitute potential N-glycosylation RNase genes and their promoter sequences sites. The importance of RNase to the self-incom(Matton et al 1995). For the present, these experi- patibility mechanism in Petunia is underscored by ments come close to providing a direct demonstra- the presence of conserved domains in the S-glycotion that the S-RNase gene is sufficient for proteins with strong homology to active site self-pollen rejection. regions of fungal RNases (Broothaerts et al 1989, 1991; Ai et al 1990; Clark et al 1990; Ioerger et al 1991; Lee, Singh and Kao 1992). Additionally, SLGs 3. PETUNIA AND OTHER SOLANACEAE from self-incompatible genotypes of P. hybrida and Petunia enjoys the distinction as one of the first P. inflata have been purified and shown to be genera to be used in self-incompatibility research, RNases, as are the SLGs of N. alata (Broothaerts et yet our knowledge of the molecular biology of self- al 1991; Singh, Ai and Kao 1991). A further point of incompatibility in this genus lags behind that of interest is that the carbohydrate moiety of the SBrassica and Nicotiana. cDNA clones encoding S- RNase is not required for the self-incompatibility allele-associated proteins were isolated and char- reaction in P. inflata, suggesting that the S-allele acterized almost simultaneously from P. hybrida specificity of the protein is determined by its amino
Molecular Biology of Self-Incompatibility
acid sequence and not by the glycan chain (Karunanandaa, Huang and Kao 1994). Direct proof implicating RNase in self-incompatibility in P. inflata has come from an experiment in which a mutant s-gene that lacked RNase activity due to a site-directed mutagenesis of one of the catalytic histidines was introduced into transgenic P. inflata and failed to reject self-pollen bearing the same Sallele as the transgene (Huang et al 1994). However, the extent to which stylar RNase activity in the self-incompatible lines of Petunia inhibits self-pollen tube growth is at present uncertain, and this gap in our knowledge has thwarted attempts to formulate a unitary concept of S-gene action in the gametophytic systems. The uncertainty stems from an examination of the levels of Sproteins and RNases in pseudo-self-compatible and self-compatible lines of P. hybrida. When RNase activity of these lines is compared, equal levels of activity are found in both (Clark et al 1990). Similarly, the S-proteins of self-compatible lines of P. hybrida display the same spatial and temporal pattern of expression as the SLGs of self-incompatible P. inflata lines. Proteins of both lines also have comparable levels of RNase activity (Ai, Kron and Kao 1991). This evidence begs the question of how S-RNase distinguishes between self- and nonselfpollen and how the recognition reaction leads to growth inhibition of the pollen tube. Conflicting data exist about the expression of SRNase transcripts in pollen grains of Petunia. Microprojectile bombardment showed that chimeric S-RNase-GUS gene accumulates to low levels in developing anthers of P. hybrida prior to anthesis (Clark and Sims 1994). However, the selfincompatibility trait of pollen grains of transgenic P. inflata is unaffected in transgenic plants whose phenotype is transformed from self-incompatibility to self-fertility (Lee, Huang and Kao 1994). Some insight into the structure of the S-gene of P. inflata is provided by an analysis of genomic clones of two S-locus alleles. The structure associates both alleles with a single, short intron that is embedded in the region identified with high interallelic variability in SLGs and that shows small variations in size and position between alleles. The clones of both alleles share the same TATA-box structure and have short, untranslated leader sequences. In both, there is a lack of interallelic homology in the flanks of the coding regions, which themselves contain repeated sequences. Despite these and other general similarities showing that the clones are alleles of the same S-gene, the clones appear to be basically heterogeneous,
289
except for conserved sites within the coding regions (Coleman and Kao 1992). The sequence divergence in the flanking regions might reflect an inevitable effect of an independent evolution of the two alleles. Attempts to clone self-incompatibility genes from other solanaceous genera have used oligonucleotide probes or heterologous probes to screen cDNA libraries. Since conserved regions of SLGs thus far characterized display some homology, it might be reasonable to expect heterologous probes to identify sequences similar to the primary structural features of S-locus proteins from cDNA libraries of related genera. In one round of experiments, cDNA or genomic clones encoding S-locus proteins were isolated from Solanum chacoense (Xu et al 1990b; Despres et al 1994; Saba-El-Leil et al 1994), S. tuberosum (Kaufmann, Salamini and Thompson 1991), and Lycopersicon peruvianum (Tsai et al 1992; Rivers et al 1993; Chung et al 1994). A rare, variant self-compatible line of L. peruvianum was used to illustrate the essentiality of RNases in preserving self-incompatibility in this species. Crosses between the variant line and a self-incompatible line showed that the mutation to self-compatibility is located at the S-locus and is associated with the absence of stylar RNase activity. The derived amino acid sequence for the glycoprotein from the compatible line was found to differ from that of the incompatible line by the loss of a histidine residue; as histidines are essential for the catalytic activity of the enzyme, this leads to the production of a protein that lacks RNase activity (Kowyama et al 1994; Royo et al 1994). Comparison of the deduced amino acid sequences of S-locus proteins of S. chacoense, S. tuberosum, and L. peruvianum with SLGs of Nicotiana alata, Petunia
hybrida, and P. inflata has shown a set of five conserved regions, including domains similar to those in the active site of fungal RNases. Equally significant is the presence of eight conserved cysteine residues in all S-locus proteins. Extreme sequence diversity between alleles of the same, as well as of different species, is also a fundamental property of the proteins characterized from the four genera of Solanaceae (Tsai et al 1992). Comparison of the amino acid sequences of S-locus proteins of 11 alleles from N. alata, P. inflata, and S. chacoense has
revealed the interesting fact that sequence similarity is relatively low within species, whereas some interspecific homologies are higher than intraspecific homologies (Ioerger, Clark and Kao 1990). That extreme sequence diversity can be pervasive between populations of even a single species is
290 apparent from a study in which the S-allele from L. peruvianum was used to probe DNA from several natural populations of the species (Rivers et al 1993). These and other observations have been put together in the form of a hypothesis about the phylogenetic relationships of the S-alleles that states that S-allele polymorphism has a common ancestry and that it arose prior to speciation in the four genera of Solanaceae (Ioerger, Clark and Kao 1990; Rivers et al 1993). In common with the structure of the S-gene of P. inflata, genomic clones corresponding to two S-alleles of S. tuberosum have a single intron and a short, untranslated leader sequence (Kaufmann, Salamini and Thompson 1991). From the data available, no major differences are discernible between the gene structures of the alleles. Expression of an S-RNase gene from S. tuberosum under the control of the tomato pollen-specific LAT-52 gene is developmentally regulated in the pollen grains of transgenic tobacco, but that alone does not appear sufficient to influence the compatibility phenotype of the transformants (Kirch et al 1995).
Pollination and Fertilization
In seeking a common denominator with other gametophytic self-incompatibility systems, it was found that the focus on RNase as a candidate for stigmatic S-locus protein in P. rhoeas was misplaced. This summation is based on several lines of evidence, such as the presence of negligible amounts of RNase in the stigma of P. rhoeas, the lack of correlation of enzyme activity with the development of incompatibility, the lack of RNase activity in purified S-locus protein, and the insensitivity of pollen grains to RNase in in vitro bioassays (Franklin-Tong et al 1991). On the other hand, the rapidity of the self-incompatibility reaction in P. rhoeas has led to the view that this is mediated by cytosolic Ca2+ acting as a second messenger. In support of this view, using laser scanning confocal microscopy and fluorescence ratio imaging of pollen tubes stained with the fluorescent dye Calcium Green-1 to localize Ca2+ fluxes, researchers have visualized transient increases in cytosolic Ca2+ levels in growing pollen tubes after the addition of partially purified, incompatible stigmatic protein extract or the recombinant S-protein. Evidently, changes in Ca2+ levels are farreaching, for the growth of pollen tubes that generate a rise in Ca2+ when challenged with the stigmatic extract or S-protein is inhibited (Figure 10.10). There 4. PAPAVER RHOEAS A N D OTHER PLANTS is no rise in the Ca2+ levels of pollen tubes after Molecular biology of self-incompatibility in addition of either compatible or heat-denatured, Papaver rhoeas may provide a testing ground for incompatible stigmatic extract. Even when the level concepts applied to gametophytic and sporophytic of endogenous Ca2+ is raised artificially by UV actisystems of self-incompatibility. Although the selfvation of caged Ca2+, an inhibition of pollen tube incompatibility system in this species is of the growth is noted in the absence of incompatible stiggametophytic type, inhibition of self-pollen occurs matic extract (Franklin-Tong et al 1993; Franklinin the stigma, as in the sporophytic type. What Tong, Ride and Franklin 1995). Conceivably, the makes this system interesting is that it is possible to transcriptional activation of genes necessary for the reproduce the self-incompatibility reaction in vitro initiation of self-incompatibility reactions in P. and that identification of the putative S-gene prod- rhoeas pollen might be the consequence of a rise in ucts has been made through their biological activ- Ca2+ level that is induced by incompatible S-locus ity of the inhibition of pollen tube growth proteins and is accompanied by an increased phos(Franklin-Tong, Lawrence and Franklin 1988; phorylation of proteins, but we have a long way to Franklin-Tong et al 1989). Transcript expression of go before the signal transduction pathway and the the S-gene cloned from P. rhoeas correlates well spa- mode of action of gene products are defined tially and temporally with the appearance of (Franklin-Tong, Lawrence and Franklin 1990; Rudd S-protein in the stigma and the onset of self-incom- et al 1996). patibility. The functional significance of this gene is underscored by the fact that the recombinant proBrassicaceae, Solanaceae and Papaveraceae thus tein isolated from Escherichia coli expressing the represent the three families in which S-genes and gene inhibits pollen tube growth in culture, as does their products have been intensively studied at the the stigmatic extract (Foote et al 1994). Finally, molecular level; very little information is available some preliminary evidence has been obtained for on other plants. A beginning has been made toward the presence in mature P. rhoeas pollen grains of a the molecular analysis of self-incompatibility in plasma membrane-associated glycoprotein that members of Rosaceae by characterizing cDNA binds to the stigma S-proteins in a nonspecific clones representing S-locus alleles of apple and manner (Hearn, Franklin and Ride 1996). As is evi- Pyrus serotina. As is the case in the S-locus genes of dent by now, these observations hold the key to the other plants, specificity of expression of transcripts basic reactions in self-incompatibility in all plants. of apple cDNA clones in the stylar tissues is inher-
Molecular Biology of Self-Incompatibility
291
50000
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
400
600
800
1000 1200 time (s)
1400
1600
1800
2000
40 2200
Figure 10.10. Changes in Ca2+ and growth of pollen tubes of Papaver rhoeas microinjected with Calcium Green-1. The filled triangles represent fluorescence (pixel values) from the first 100 urn of the pollen tube; the open squares indicate the relative length of the pollen tube at the corresponding time point, (a) Addition of compatible stigmatic S-glycoproteins (first arrow) followed by addition of incompatible stigmatic S-glycoproteins (second arrow). No
increase in Ca2+ and no significant inhibition of pollen tube growth are observed after treatment with the compatible protein; after addition of the incompatible protein, there is a transient rise in Ca2+, indicated by the two peaks, and a cessation of pollen tube growth around 2,000 seconds, (b) Photoactivation of caged Ca2+ (arrow) and cessation of pollen tube growth almost immediately. (From Franklin-Tong et al 1993.)
ent to their function, and, similar to S-proteins of members of Solanaceae, purified S-proteins from apple and P. serotina correspond to RNases (Broothaerts et al 1995; Sassa et al 1996). It appears that RNases are likely to mediate the self-incompatibility syndrome in certain lines of Antirrhinum majus also (Xue et al 1996). So, if a single regulatory molecule is required to activate the stylar self-pollen rejection mechanism, it would probably be the same molecule in members of Solanaceae, Rosaceae, and Scrophulariaceae. There is preliminary evidence showing that SLG and SRK-like genes modulate the sporophytic type of self-incompatibility in Ipomoea trifida (Kowyama et al 1995). What appears to be the pollen component of the self-incompatibility allele has been cloned from the
grass Phalaris coerulescens (X. Li et al 1994). Unlike the other gametophytic systems studied, selfincompatibility in P. coerulescens is under the control of two unlinked loci, S and Z, and self-pollen rejection is the rule when both S and Z alleles are present in the pollen as well as in the style. A cDNA clone identified by differential screening was found to be expressed only in mature pollen grains and was absent from all other tissues tested, including the stigma. The deduced amino acid sequence showed that the protein has a variable N-terminus that has no homology to known proteins and a conserved Cterminus with a thioredoxin domain. The absence of a signal sequence and the presence of thioredoxin sequences probably indicate that, unlike other Slocus proteins, this is a cytoplasmic protein. How
292 this protein controls both recognition specificity • and inhibition of pollen tube growth in the selfincompatibility reaction remains to be elucidated. Considering the self-incompatibility systems for which good molecular data are available, it is clear the pollen self-rejection system evolved independently several times in the evolution of angiosperms. Millions of years have probably separated the different incompatibility systems. The molecular data from Brassicaceae, Solanaceae, and Papaveraceae show that there are some obvious, basic similarities, as well as differences, in the operation of signaling systems in the three families. A viewpoint that has gained momentum in recent times is that cellular responses unleashed during sporophytic self-incompatibility may share some components of host-pathogen responses in plants and that self-incompatibility is evolutionarily related to the system that conditions resistance to invading pathogens. Besides superficial similarities between the genetics, physiology, and biochemistry of the self-incompatibility and host-pathogen responses, the possible involvement of a receptor kinase in both strengthens the view that there is a connection between the two responses (Hodgkin, Lyon and Dickinson 1988; Lamb 1994). 5. GENERAL COMMENTS
Molecular analysis shows that specific characteristics of the self-incompatibility genes are varied in the four systems in which they have been studied
Pollination and Fertilization
intensively, and this makes it difficult to interpret the mechanism of action of the encoded proteins in the rejection of self-pollen. In a general sense, we know that in Brassica the structure of the proteins suggests a role in the signal transduction mechanism, whereas in Nicotiana and Petunia the protein that shows similarity to RNases functions by disabling incompatible pollen tubes. Particularly intriguing is that in Papaver the protein elicits a cellular response that results in a transient increase in Ca2+ concentration. Dramatic progress is currently being made in understanding the way the different proteins work in the model incompatibility systems. Identification of the recognition sequences expressed by the pollen grains assumes an obvious significance in the expected flow of information. At this stage, we do not know whether the recognition sequences expressed by pollen grains are different from those expressed by the female part of the flower. In analyzing the conditions that lead to the incompatibility reactions, delineation of the parts of the incompatibility molecules important for recognition begs resolution. Equally significant is the role of other parts of the incompatibility protein, perhaps in the initiation of events that subsequently result in pollen tube collapse. It is encouraging to note that more than one model system of self-incompatibility has been investigated to a level insightful enough that realistic progress in our understanding of the molecular basis of the recognition reaction and its consequences can be expected in the foreseeable future.
Chapter 11
Fertilization: The Beginning of Spomphytic Growth Cellular Nature of the Sperm (i) The Sperm Cytoskeleton (ii) The Male Germ Unit (iii) Isolation and Characterization of the Sperm Odyssey of the Sperm and Double Fertilization (i) Growth of the Pollen Tube through the Style (ii) Passage of Sperm into the Embryo Sac (iii) Fusion of Nuclei (iv) Cytoplasmic Inheritance In Vitro Approaches to the Study of Fertilization (i) Intraovarian Pollination (ii) Test-Tube Fertilization (iii) In Vitro Fertilization (iv) Gynogenesis 4. General Comments
In the life cycle of plants, fertilization is invariably associated with sexual reproduction and therefore represents the genetic switch that initiates the diploid, or the sporophytic, phase of development. The history of research into fertilization in angiosperms may be said to have begun with the discovery of syngamy, or the actual fusion of the male and female gametes, by Strasburger (1884) and of double fertilization by Nawaschin (1898) and Guignard (1899). The latter two investigators called attention to the streamlining of the male gametophyte's economy by providing light microscopic evidence for the fusion of one sperm with the egg cell to form the diploid zygote and of the other sperm with the polar fusion nucleus of the central cell to generate the triploid endosperm. In spite of an abundance of studies on the development of the male and female gametophytes and on
postfertilization events in angiosperms in the years following these discoveries, relatively few investigators have ventured into the area of fertilization research. This has led to the unenviable situation that, as a biological process, fertilization in angiosperms has remained as an unknown black box. The paucity of information on fertilization is undoubtedly related to the fact that events leading to the egg-sperm encounter take place in the privileged interior of the multicellular ovule, where, hidden from the eye, they are not prone to experimental assault and to the technical difficulties that have so far prevented the development of an ex ovulo in vitro fertilization system. However, during the past three decades, there has been considerable progress in our understanding of many facets of fertilization, particularly the ultrastructure of the male and female gametes and the sequence of events leading to gametic fusion. Extensive use of the electron microscope, experiments involving immunological and molecular approaches, and nondestructive isolation techniques have all contributed to this progress. Chapter 6 described in some detail the structure of the egg at the time of fertilization, and so this material will not be considered here again except as it relates to the male gametophyte preparatory to gametic fusion. The primary focus of the present chapter will be on the structure of the sperm cells, their fate in the pollen tube, and their eventual fusion with the cells of the embryo sac. The final section of this chapter is devoted to a review of the work done on the manipulation of the fertilization process under controlled conditions. Until recently, the subject of fertilization has been reviewed almost always within the larger context of male and female gametogenesis (Linskens 1969; Jensen 293
294 1973; Kapil and Bhatnagar 1975; van Went and Willemse 1984), but more focused reviews (Knox and Singh 1987; Knox, Southworth and Singh 1988; Russell, Cresti and Dumas 1990; Russell 1992) are beginning to appear. These should serve as excellent sources of information on various aspects of the biology of fertilization.
Pollination and Fertilization
appears as densely staining granular chromatin (Knox and Singh 1987; Knox, Southworth and Singh 1988). The nucleus is enclosed in a nuclear envelope in which the usual pores appear to be lacking or are few in number (Mogensen and Wagner 1987; Southworth, Platt-Aloia and Thomson 1988; Taylor et al 1989; Southworth 1990). Not much should be read into this last observation except that the nucleus of the mature sperm is tran1. CELLULAR NATURE OF THE SPERM scriptionally inactive and negotiates a limited trafOne point of special concern in the early studies fic with the surrounding cytoplasm. on the structure of the sperm was whether the sperm It is within the framework of the subcellular is a true cell with all the trappings or just an structure that factors responsible for the differentiorphaned nucleus. Although some light microscopic ation of cells reside; however, in all cases examined, investigations on this question remained inconclu- the cytoplasm of the male gamete appears truly sive, the cellular nature of the sperm was clearly undistinguished and does not give any hints about established by the early 1940s. Subsequently, elec- the specialized character of the cell (Figure 11.1). In tron microscopy, particularly the pioneering investi- classical descriptions, limited amounts of ER, ribogations of Jensen and Fisher (1968b) on bicellular somes, polysomes, lipid and protein bodies, mitopollen grains of Gossypium hirsutum, and of Hoefert chondria, dictyosomes, microtubules, and vesicles (1969b) and Cass (1973) on tricellular pollen grains of appear to make up the bulk of the sperm cytoBeta vulgaris and Hordeum vulgare, respectively, pro- plasm, with the mitochondrion varying with vided the first direct evidence that the male gamete respect to its morphology in certain species in angiosperms is not a naked nucleus but is a rela- (Dumas, Knox and Gaude 1985; McConchie, tively simple cell, with its own cytoplasmic sheath Hough and Knox 1987; Charzynska, Ciampolini containing a reduced complement of organelles. and Cresti 1988; Hu and Yu 1988; Southworth and Exhibiting great simplicity, the sperm in G. hirsutum is Knox 1989; Taylor et al 1989; Yu, Hu and Zhu 1989; reduced to little more than a bag for a prominent Cresti, Murgia and Theunis 1990; Theunis 1990; P. nucleus bathed in a thin layer of cytoplasm contain- E. Taylor et al 1991; Theunis, Cresti and Milanesi ing dictyosomes, ER, polysomes, vesicles, and 1991; van der Maas et al 1994). Mitochondria are organelles resembling mitochondria (Jensen and the most prominent organelles of the sperm of Zea Fisher 1968b). Besides these cytoplasmic organelles, mays and, as seen in three-dimensional reconstrucsperm cells of B. vulgaris and H. vulgare also harbor tion, they form a filamentous complex, each indifiles of microtubules along their length, suggesting vidual filament partially wrapping around itself that the cytoskeletal elements might play a role in (McConchie, Hough and Knox 1987; Mogensen, maintaining the shape of the sperm or in affecting Wagner and Dumas 1990). At the other extreme are sperm motility (Hoefert 1969b; Cass 1973). the conventional, small, ellipsoidal mitochondria found in the sperm cells of Brassica campestris and Further information about the more intimate B. oleracea (McConchie et al 1987). An important details of the structure of sperm cells has come question raised in these studies relates to the inherfrom electron microscopy combined with threeitance of plastids - whether plastids are paternally dimensional reconstruction, freeze-fracture analyinherited or not - and hence to their presence in the sis, confocal laser scanning fluorescent microscopy, cytoplasm of the sperm cells. The answer to this and immunochemistry. From these investigations question appears equivocal, because sperm cells of emerges the following summary picture. Sperm some plants, like Plumbago zeylanica (Russell and cells generally extend in length to about 35 fim and 1981; Russell 1983), Rhododendron laetum, and Cass 3 are approximately 1-3 fjm in width and 7-133 ^m R. macgregoriae (Taylor et al 1989), come equipped in volume. The characteristic shape of sperm cells, with plastids in their cytoplasm, whereas those of as seen in mature pollen grains or in pollen tubes, Spinacia oleracea (Wilms, Leferink-ten Klooster and is elongate or spindle shaped. The most conspicuvan Aelst 1986), Z. mays (McConchie, Hough and ous organelle of the sperm is the nucleus, which Knox 1987), B. campestris, B. oleracea (McConchie occupies an enlarged region at one end of the cell; et al 1987), B. napus (Murgia et al 1991b), and one or more tapering cytoplasmic extensions are Nicotiana tabacum (Yu, Hu and Zhu 1989) lack found at the other end. The nucleus of the mature plastids. sperm contains paternal genetic material, which
Fertilization: The Beginning of Sporophytic Growth
295 (i) The Sperm Cytoskeleton
Analysis of the cytoskeletal machinery of the sperm has followed a two-tier course, analogous to that employed in the study of the cytoskeleton of the vegetative and generative cells (Chapter 3). Early studies were conducted at the ultrastructural level, whereas later studies enlisted immunofluorescence methods. These investigations have provided a surprisingly uniform picture of the organization of sperm microtubules, which attain conformations compatible with the function of the sperm cells. Commonly, microtubules, singly or in bundles, are aligned longitudinally or helically in relation to the axis of the sperm (Hoefert 1969b; Cass 1973; Karas and Cass 1976; Zhu et al 1980; Dumas, Knox and Gaude 1985; Pierson, Derksen and Traas 1986; Charzynska, Ciampolini and Cresti 1988; Taylor et al 1989; Cresti, Murgia and Theunis 1990; Del Casino et al 1992; Palevitz and Liu 1992). A striking feature of the sperm of Tradescantia virginiana is the presence of numerous interconnected helical to longitudinal branches of microtubules that appear to arise from the main microtubule bundle and invest the entire cell (Palevitz and Cresti 1988). Whether the microtubule elements of the sperm are inherited from the interphase cytoskeletal network of the generative cell at the time of division or are assembled anew from stable precursors at the end of division is not clear. According to Palevitz and Cresti (1989), the entire array of microtubules of the sperm of T. virginiana is inherited from those seen in the late anaphase of the generative cell mitosis; these microtubule bundles are derived from the kinetochore fibers, which remain enmeshed with each other and with the surrounding microtubules during anaphase. Consistent with this view is the observation that the precise, stepwise aggregation of the kinetochore fibers at the poles of the generative cell is also reflected in the cytoskeleton of the young sperm (Palevitz 1990). However, circumstantial evidence such as the presence of fewer and lessprominent microtubule bundles in sperm cells compared to their progenitor cell, as seen in Rhododendron laetum and R. macgregoriae (Taylor et al
Figure 11.1. Structure of the sperm cell in the pollen tube of Populus deltoides. Two elongate sperm cells seen in the longitudinal section of the pollen tube are connected by narrow cytoplasmic projections. PT, pollen tube; SC, sperm cell. Scale bar = 1 urn. (From Rougier et al 1991; photograph supplied by Dr. C. Dumas.)
1989), argues against the inheritance of microtubules from the previous generation. On the whole, the situation presents an invitation for more experimentation about the origin of microtubules in the sperm cells of angiosperms. In ultrastructural investigations of sperm cells, not much attention has been paid to the presence of actin-containing microfilaments, and, if microfila-
296 ments are present, they have clearly remained elusive. Experiments using rhodamine-phalloidin as a fluorescent detection probe have also failed to reveal the presence of microfilaments in sperm cells
Pollination and Fertilization
Plumbago zeylanica (Russell and Cass 1981), and barley (Mogensen and Wagner 1987). The contacts and physical associations of the two sperm cells born out of the division of a generative of Lilium longiflorum, Nicotiana tabacum, Petunia cell were occasionally commented upon in some of hybrida, and Tradescantia virginiana (Pierson, the early light and electron microscopic studies. Are Derksen and Traas 1986; Palevitz and Cresti 1988, the two sperm cells closely linked together as a pair, 1989). As indicated in Chapter 3, ultrastructural or are they aligned with the vegetative cell nucleus, and immunofluorescence studies have failed to or do they float aimlessly in the pollen grain or in detect microfilaments in the generative cell; in this the pollen tube cytoplasm? Although the occurrespect, sperm cells appear to share a unique prop- rence of paired sperm cells and their closeness to erty with their progenitor cell. The only reports of the vegetative cell nucleus in the pollen tube of cotthe presence of microfilaments are in sperm cells of ton was noted by Jensen and Fisher (1968b, 1970), it wheat (Zhu et al 1980) and Plumbago zeylanica was the work of Russell and Cass (1981) that intro(Russell and Cass 1981), but these observations duced a new level of understanding of the charachave not been corroborated by other researchers. teristics of the sperm cell association and the One study using antiactin immunofluorescence as sperm-vegetative cell package. This work showed a specific probe has reported an impressive local- that in the tricellular pollen grain of P. zeylanica, the ization of actin in the generative cells and sperm two sperm are directly linked to each other by a cells of Rhododendron laetum and R. macgregoriae common lateral wall transgressed by plasmodes(Taylor et al 1989), but a subsequent work using the mata. It was also found that while both sperm are same and other materials and the same antiactin enclosed in the inner plasma membrane of the vegprobe failed to substantiate this observation etative cell, one of them is interlocked with the (Palevitz and Liu 1992). On the weight of the evi- nucleus of the vegetative cell in a complex way. The dence, the latter investigators have concluded that conformational relationship of the sperm cell to the vegetative cell nucleus is traced to a cytoplasmic microfilaments do not form part of the cytoskeleton extension that wraps around the periphery of the of the sperm cells of angiosperms. vegetative cell nucleus and is partially enclosed by convolutions of this nucleus (Figure 11.2). The sperm cell remains associated with the nucleus of (ii) The Male Germ Unit the vegetative cell during the period of active The electron microscopic image of the outer growth of the pollen tube, but the connection is sevboundary of the male gametes in the pollen grain is ered following the discharge of the pollen tube conthat of protoplasm covered by a plasma membrane, tents. Although it has been suggested that the with two sperm cells enclosed together in the inner sperm-vegetative cell nucleus assembly might facilplasma membrane of the vegetative cell. The mem- itate gamete delivery, its importance is not fully branes may be tightly appressed or may form a understood because the association of the naked loose association with intramembrane pockets that nucleus and the adhering cells cannot be easily enclose vesicles and osmiophilic droplets (Zhu et al manipulated or examined in situ. 1980; Dumas, Knox and Gaude 1985; McConchie, Hough and Knox 1987; McConchie et al 1987; An exciting concept demonstrated in a later Southworth and Knox 1989). The paired nature of work (Russell 1984) is that the two sperm cells in the outer membrane system of the sperm has been the generative cell, rather than being isomorphic, confirmed by visualizing cells by freeze-fracture, in display varying degrees of dimorphism in size and which the frozen membranes are cleaved along organelle content. This conclusion was made possitheir hydrophobic interior and the resulting surface ble by an examination of serial, ultrathin sections of is observed in the electron microscope; this sperm cells, by computer-aided three-dimensional approach did not reveal any connections or bridges reconstruction of their structure, and by the actual between the membranes (Southworth, Platt-Aloia counting of organelles in sections. The dimorphism and Thomson 1988; Southworth 1990; van Aelst, in shape is due to a membranous extension from Theunis and van Went 1990). This observation the one sperm of the pair that superficially undoubtedly relegates the male gamete to the sta- embraces the nucleus of the vegetative cell. This tus of a natural protoplast; however, a rudimentary sperm appears to be the larger of the two; the wall has been reported to be present in the sperm smaller one, which lacks a cellular projection, is of wheat (Zhu et al 1980; Hu, Zhu and Xu 1981), unassociated with the vegetative cell nucleus.
Fertilization: The Beginning of Sporophytic Growth
297
Figure 11.2. Reconstruction of the sperm cells and the male germ unit of Plumbago zeylanica. (a) The two sperm cells; the sperm cell associated with the vegetative cell nucleus is Svn; the other sperm cell, not associated with the vegetative cell nucleus, is Sua. (b) The physical association between Svn and the nucleus of the vegetative cell (VN). The cellular
protrusion of the sperm cell (arrowheads) is found within the clasping region of the lobed vegetative cell nucleus and extends through a shallow hole in the vegetative cell nucleus at one location (arrow). (From Russell 1984. Planta 162:385-391. © Springer-Verlag, Berlin.)
Perhaps the most dramatic difference between the two sperm lies in their relative content of heritable organelles. The large sperm is rich in mitochondria and is typically impoverished of plastids, whereas the small one is poor in mitochondria but has numerous plastids. Since P. zeylanica is unusual in lacking synergids in the embryo sac, sperm-vegetative cell nucleus association and sperm dimorphism seen in this species raised objections that they may not be representative of the wide range of angiosperms characterized by the presence of two synergids flanking the egg cell. This objection was soon dispelled when similar features were found in sperm cells of
lacks a cytoplasmic outgrowth, is linked to the larger one by a common cell junction. The small sperm cells in both species of Brassica also appear unique in their paucity of mitochondria, with 4 in B. campestris and 6 in B. oleracea, compared to a maximum of 15 and 34 in the large sperm cells of B. oleracea and B. campestris, respectively. Neither
Brassica campestris and B. oleracea (Dumas, Knox
and Gaude 1985; McConchie, Jobson and Knox 1985; McConchie et al 1987), both of which possess synergids. In both, dimorphism in shape is due to the presence of an evagination on the sperm that is attached to the vegetative cell nucleus, which thus makes it the larger of the two sperm cells. The distinction between the two species lies in the fact that in B. oleracea the evagination on the sperm surrounds the surface of the vegetative cell nucleus superficially, whereas in B. campestris it penetrates the enclaves of the vegetative cell nucleus and emerges at the other end as a protuberance (Figure 11.3). In both species, the smaller sperm, which
MT MT
Fig. 11.3. A schematic diagram of the male germ unit of Brassica campestris. M, mitochondria; MT, microtubules; N, nucleus; VN, nucleus of the vegetative cell; SC1, SC2, sperm cells 1 and 2. (From McConchie, Jobson and Knox 1985.)
298
Pollination and Fertilization
sperm cell contains any plastids. The published This attention to detail seems justified because other accounts of the structure of the sperm of Spinacia electron microscopic studies, with or without threeoleracea are much less detailed than those of Brassica dimensional reconstruction of serial sections, have sp., although differences in the sperms' size and demonstrated a physical relationship between the association with the nucleus of the vegetative cell sperm and the vegetative cell nucleus; included in are obvious in the electron micrographs (Wilms, these studies are male germ units identified in pollen Leferink-ten Klooster and van Aelst 1986; Theunis, tubes of bicellular pollen grains of Hippeastrum vittaCresti and Milanesi 1991). tum (Mogensen 1986), Petunia hybrida (Wagner and Thanks to the skills of the electron microscopist Mogensen 1988), Rhododendron laetum, R. macgregriae and the availability of techniques for computer (Taylor et al 1989), Cyphomandra betacea (Solanaceae) simulation of the data, these observations add a (Hu and Yu 1988), Nicotiana tabacum (Yu, Hu and new dimension to our knowledge of sperm struc- Zhu 1989; Yu, Hu and Russell 1992; Yu and Russell ture and organelle inheritance in angiosperms. 1994b), and Populus deltoides (Rougier et al 1991), Despite the minimal structural attributes displayed and in tricellular pollen grains of Spinacia oleracea by the sperm, this cell appears to be highly differ- (Wilms, Leferink-ten Klooster and van Aelst 1986), entiated in relation to its developmental tasks, Catananche caerulea (Asteraceae) (Barnes and especially to respond with preestablished speci- Blackmore 1987), and B. napus (Murgia et al 1991b). ficity and to discriminate between the egg nucleus In C. betacea, N. tabacum, and P. hybrida, the two and the polar fusion nucleus during fertilization. sperm are linked together by a common transverse The presence of significantly different numbers of wall as in Plumbago zeylanica. In N. tabacum, the wall heritable organelles in the two sperm cells of a pair appears to be further specialized by the presence of suggests the possibility of sophisticated mecha- fibrils, small tubules, and callose (Yu, Hu and nisms of organelle transmission at fertilization. The Russell 1992). differences in the sizes of the two sperm and in The importance of three-dimensional reconstructheir organelle contents pose some important ques- tion of the sperm-vegetative cell nucleus assembly tions on differential gene activity during sperm cell lies in the demonstration that complex association differentiation. between cells can be understood in simple graphic The novel features of the structure of the sperm pictures. The impressive precision of these studies provided the impetus for further work, which caused many to believe that it was only a matter of brings the observations in line with the current time before the concept of the male germ unit could thinking on the nature of the male gamete. Dumas be considered as a secure generalization applicable et al (1984) designated the functional complex of the universally to angiosperms. This did not prove to two sperm cells of common origin plus the nucleus be the case. From light and electron microscopic of the vegetative cell in the pollen grain, or in the observations of pollen grains and pollen tubes of pollen tube, as the male germ unit; the rationale for wheat (Chandra and Bhatnagar 1974), Helleborus this concept is the view that all the nuclear and foetidus (J. Heslop-Harrison, Heslop-Harrison and cytoplasmic DNA of male heredity is linked Heslop-Harrison 1986), Euphorbia dulcis (Murgia together, forming a single functional unit for the and Wilms 1988), and Alopecurus pratensis (J. recognition of target female cells, the transmission Heslop-Harrison and Heslop-Harrison 1984), it of hereditary materials, and fusion during double appears that there is a spatial or temporal separafertilization. Some investigators (Kaul et al 1987; Yu tion of the sperm pair from the nucleus of the vegeand Russell 1992) have referred to a complex asso- tative cell. In live pollen tubes of Secale cereale, not ciation formed between the vegetative cell nucleus only is there no clear-cut evidence of linkage and the generative cell as the male germ unit; in between the sperm cells, but the cells have been other cases, the transient union noted between the observed to shoot past each other and move in two sperm after they are formed from the genera- opposite directions (J. Heslop-Harrison and tive cell has been identified as the male germ unit Heslop-Harrison 1987). According to Weber (1988), (Levieil 1986; Charzynska et al 1989; Charzynska who followed the division of the generative cell in and Lewandowska 1990). It is a moot point whether tricellular pollen grains of Galium mollugo, such associations, however complex they are, Trichodiadema setuliferum, and Avena sativa, there is should be included within the concept of the male no linkage between sperm cells that separate from germ unit, which was coined to convey a precise one another soon after they are formed. These functional union involving the two sperm and the observations, though drawing attention to the indevegetative cell nucleus prior to double fertilization. pendent existence of sperm cells and the nucleus of
Fertilization: The Beginning of Sporophytic Growth
299
the vegetative cell, do not take into account the fact try. Further work has shown that the high degree of that an association between them may be formed at cytoplasmic difference displayed by the sperm of a later stage, and so further work is called for before Plumbago zeylanica, characterized by the presence of the issue can be considered settled. There is strong significantly different numbers of plastids and support for the view that in some plants the physi- mitochondria, is somewhat unique and that far cal connection between sperm cells of a pair and more widespread are cases in which organelle difbetween one sperm and the vegetative cell nucleus ferences are restricted to mitochondrial number or is ephemeral and would have been completely nuclear size (McConchie, Hough and Knox 1987). missed in routine electron microscopic observa- Moreover, pollen grains of some plants with pertions. For example, by three-dimensional analysis of fectly standard male germ units have revealed no ultrathin sections, it was found that in the pollen appreciable differences in organelle contents by grains of barley (Mogensen and Rusche 1985) and serial thin-sectioning, computer reconstruction, maize (McConchie, Hough and Knox 1987), the and quantification. Such isomorphic sperm cells sperm cells are discrete and do not exist in physical have been found in barley (Mogensen and Rusche association with the vegetative cell nucleus as the 1985), Petunia hybrida (Wagner and Mogensen male germ unit. The major effort in a subsequent 1988), Rhododendron sp. (Taylor et al 1989), Medicago study on barley was to look for evidence of a phys- sativa (Zhu, Mogensen and Smith 1992), and ical association between sperm cells and the vegeta- Nicotiana tabacum (Yu, Hu and Russell 1992). In the tive cell nucleus during hydration and germination most detailed quantitative analysis of the sperm in of the pollen grain (Mogensen and Wagner 1987). N. tabacum, no statistically significant differences This work showed that following controlled polli- were found between the two sperm cells in several nation, before sperm cells move into the emerging cellular parameters (Yu, Hu and Russell 1992; Yu pollen tube, there is a close association between and Russell 1994b). sperm cells and between them and the nucleus of In summary, it appears that in many angiothe vegetative cell. Although the vegetative cell sperms, part of the preprogramming of sperm cells nucleus is not seen to migrate into the pollen tube, involves the formation of the male germ unit with sperm cells remain attached in the pollen tube; as or without a concurrent asymmetry in the distribusperm cells approach the embryo sac, there is a furtion of heritable organelles. How this specialized ther intimacy in their association and they appear to packaging of the sperm influences the fertilization be connected at both ends. In maize pollen also, the process is not known at present. Dimorphism vegetative cell nucleus is found to be separated raises questions about the identity of the sperm cell from sperm cells, but it remains closely associated that fuses with the egg and of the one that fuses with sperm cells as a male germ unit in the course with the polar fusion nucleus, with perhaps the of postpollination journey (Rusche and Mogensen hardest question to answer being, what is the 1988; Mogensen 1992). In Medicago sativa, the assonature of the factors that endow the specificity? We ciation of sperm cells with the nucleus of the vegealso need to know the mechanisms that cause the tative cell is tenuous in the pollen grain, but is well unequal partitioning of organelles and the trigger defined in the pollen tube throughout its growth that induces it. (Zhu, Mogensen and Smith 1992). It has been argued from these observations that the male germ unit is not the developmental consequence of divi(iii) Isolation and Characterization of the sion of the generative cell; rather, it is induced as a Sperm result of pollen germination (Mogensen 1992). Since To gain insight into the molecular biology of male germ unit formation is a morphogenetic process, it is also possible that, in some plants, the double fertilization, it is necessary to know somestructure formed in the pollen grains is not stabi- thing about the dynamics of the participating lized before germination; germination is the trigger gametes. Although it is relatively easy to monitor that stabilizes the male germ unit in the pollen sperm cells microscopically, it is hard to investigate their metabolic activities experimentally. This objectubes. tive is best achieved by isolating sperm from pollen It was mentioned earlier that in certain plants grains or from pollen tubes. As the following survey that possess a male germ unit, the sperm cells show shows, from a modest beginning using sperm cells dimorphism in size and in the content of heritable extruded from pollen grains, the stage is set now for organelles. But the presence of a male germ unit is the large-scale isolation of viable sperm cells from not always associated with cytoplasmic asymme- pollen grains and pollen tubes of various plants. A
300
brief review of the present state of the art is provided by Theunis, Pierson and Cresti (1991). The first attempts to isolate sperm used tricellular pollen grains of barley (Cass 1973) and Plumbago zeylanica (Russell and Cass 1981). In recognition of the fact that the milieu of the pollen grain is in a state of high osmotic pressure, pollen contents were released into a medium of high osmolarity secured by the addition of 30% sucrose. Although the number of sperm cells isolated by this method was not large, some basic observations, such as the change in the shape of extruded sperm from spindle shaped to round or spherical, suggested a role for the cytoskeletal framework in the process. Similar methods, with variations in the concentrations of sucrose, were used subsequently for the small-scale isolation of sperm from tricellu-
Pollination and Fertilization
is necessary when pollen cannot be burst osmotically; however, germinating the pollen grains for a short period of time to generate tubes was sufficient to allow sperm release by osmotic shock (P. E. Taylor et al 1991). All these methods for the largescale isolation of sperm have used tricellular pollen grains; an osmotic shock combined with treatment with pectinase and cellulase has been devised to obtain high sperm yields from pollen tubes of several bicellular species (Mo and Yang 1992). Viability of isolated sperm cells is routinely assessed by using Evans-blue as an exclusion dye (Russell 1986; Southworth and Knox 1989) or by fluorochrome reaction (Dupuis et al 1987; Yang and Zhou 1989; Mo and Yang 1991). Zhang et al (1992a, b) have advocated the use of flow cytometry as a highly sensitive method to evaluate the viability of lar pollen grains of Brassica oleracea, Triticum aes- isolated sperm cells. Not unexpectedly, viability was tivum (Matthys-Rochon et al 1987), and Zea mays found to decrease with time of storage of isolated (Cass and Fabi 1988). The first successful isolation sperm cells (Russell 1986; Dupuis et al 1987; Theunis of sperm from bicellular types of pollen grains was and van Went 1989; Yang and Zhou 1989; Mo and accomplished by enzymatic hydrolysis of the Yang 1991). Among the measures recommended to pollen tube wall of Rhododendron macgregoriae andenhance the longevity and viability of isolated Gladiolus gandavensis (Shivanna et al 1988). For this sperm of maize are the use of a buffered galactose purpose, pollen tubes allowed to grow through the solution or the use of a dilute mineral-salt solution style into a block of agar were treated with a mix- especially low in calcium and magnesium (Zhang et ture of macerozyme and cellulase, and sperm cells al 1992b; Roeckel and Dumas 1993; G. Zhang et al were separated by microfiltration. 1995). A procedure for increasing the yield and viaGiven the differences in density between the bility of sperm cells isolated from Lolium perenne has sperm and the rest of the pollen grain or pollen used vitamin-C and fetal calf serum during isolation tube contents, one might expect to enrich a pollen and storage (van der Maas et al 1993b). Maintaining fraction for sperm by differential centrifugation. isolated sperm in a prime condition is an issue to This was achieved by subjecting osmotically which attention needs be focused for sperm to fulfill shocked pollen grains of P. zeylanica to centrifuga- their potential use in in vitro fertilization and as tion in a sucrose gradient (Russell 1986). gene vectors in genetic manipulation. In other Moderately high yields of sperm essentially free of respects, such as in the absence of the common contamination by vegetative cell nuclei and starch plasma membrane of the vegetative cell, isolated and suitable for physiological manipulations were sperm cells resemble true protoplasts (Russell 1986; obtained by this method. Dupuis et al (1987) and Dupuis et al 1987; Cass and Fabi 1988; Theunis Theunis and van Went (1989) used Percoll gradi- 1990). Morphometric analysis and three-dimenents to obtain high sperm yields from pollen grains sional reconstruction of serial sections of isolated of maize and spinach, respectively. Sperm cells at a sperm of Z. mays have revealed that there is considconcentration of 20,000/ml were obtained from erable heterogeneity in the volume and number of pollen grains of Gerbera jamesonii by gentle homog- the complement of organelles in a given population enization followed by microfiltration (Southworth of sperm (Wagner, Dumas and Mogensen 1989; and Knox 1989). Both grinding and a two-step Mogensen, Wagner and Dumas 1990). From a osmotic shock procedure have been claimed to give heuristic standpoint, this shows that strengthening high yields of sperm from pollen grains of B. the differences between two groups of sperm in a campestris, B. napus, Z. mays, and Secale cereale (Yang population is not an easy task. and Zhou 1989; Mo and Yang 1991). In the latter Except in two instances (see Section 3), it has not procedure, pollen grains hydrated in a high con- been determined whether isolated sperm cells are centration of sucrose were shocked to burst and functionally competent to fuse with an egg cell or release the sperm by dilution of the medium. This with a central cell in an in vitro environment.
Fertilization: The Beginning of Sporophytic Growth
However, a few reported cases of fusion between isolated sperm cells of the same species and between sperm of different species and between isolated sperm and isolated microspore protoplasts have provided some hopeful hints of a wider application of the technique. This curious behavior has been described among sperm cells of Gladiolus gan-
301
both transcription and translation (Zhang, Gifford and Cass 1993; Matthys-Rochon et al 1994). But the evidence is not really conclusive as to whether this is a reflection of the synthetic activities of sperm cells or is a reaction to isolation stress. It is clear from this account that in recent years a major effort in the study of male gametes of davensis, Zephyranthes Candida, Hemerocallis minor, angiosperms has been directed toward isolation proH. fulva, and Hippeastrum vittatum; between sperm cedures. Despite the fact that well-established procecells of G. gandavensis and H. vittatum; between dures are already at hand for the biochemical sperm and microspore protoplasts of H. minor; and characterization of sperm, such studies are limited. between sperm of H. vittatum and microspore pro- The special appeal of male gametes derives from the toplasts of H. fulva (Mo and Yang 1992). fact that each sperm of a pair in the pollen grain The metabolic and biochemical activities of iso- is programmed for different fusion events. lated sperm cells have hardly been investigated. Characterization of the recognition factors in the Preliminary attempts have been made to character- sperm cells is an essential prerequisite to underize the surface antigens of sperm cells by employ- standing the physiological and molecular basis of ing monoclonal antibodies. Hough et al (1986) double fertilization. raised hybridoma cell lines against isolated male gametes of Brasska campestris and, by screening them for antibody specificity by a simplified 2. ODYSSEY OF THE SPERM AND DOUBLE FERTILIZATION immunofluorescence technique, identified a few sperm surface-specific antigens. In another investiThe fate of male gametes in the pollen tube gation, Pennell et al (1987) screened hybridoma growing in the style is certainly the most critical antibodies elicited by Plumbago zeylanica sperm issue to be considered now, as it lies at the heart of against both sperm and pollen fractions. Whereas a much of our understanding of double fertilization. number of antibodies were found to bind to The pollen tube makes its way into the embryo sac, contaminating fractions from the vegetative cell where it discharges its cargo for fertilization events. cytoplasm and soluble macromolecules, a high per- The tasks involved here are far too prodigious to be centage of the reactants were also sperm specific. accomplished by sperm cells without some outside Although much remains to be done to characterize help. For convenience, the events leading up to the antibodies in both species, this is a task that pollen tube discharge are designated as the might eventually lead to insights into the basic progamic phase; this is followed by the syngamic problems of sperm recognition. phase, in which the predominant event is a set of In isolated sperm cells of P. zeylanica, a beginning nuclear fusions. has also been made to study the protein profile of the sperm by SDS-PAGE with a view toward identi(i) Growth of the Pollen Tube through fying recognition molecules during egg-sperm the Style interaction (Geltz and Russell 1988). Obtaining evidence for the existence of sperm-specific proteins In species with tricellular pollen grains, the was difficult in this work because of contamination sperm cells and the vegetative cell nucleus migrate of the preparation with vegetative cell nuclei, into the emerging pollen tube, whereas in plants organelles, and water-soluble proteins. Nonetheless, with bicellular pollen grains the generative cell about 14% of the most conspicuous polypeptides divides in the pollen tube to give rise to sperm were specific to sperm cells; whether these include cells. The stigmatic tissue on which pollen grains cell surface determinants of male gametes remains land is composed of glandular cells that facilitate to be investigated. In similar studies with isolated germination of pollen grains and growth of pollen sperm cells of maize, the protein synthetic profile tubes (see Chapter 7). These cells lead to the transwas found to change with time after isolation. The mitting tissue of the style, which is also made up of increased incorporation of radioactive precursors of cells with glandular properties. What is the mechaRNA and protein synthesis into isolated sperm cells nism by which the pollen tube is guided toward the that was observed in these investigations evidently ovary? It is now established that pollen tubes grow shows that changes in protein synthesis involve through the intercellular matrix of the style, and
302
indications are that a variety of control mechanisms operate within the style to maintain the downward directionality of the pollen rube. It has been argued quite convincingly that in some plants the guidance of the pollen tube is mechanical or electrical and that such cues interact with the intercellular growth machinery of the stylar matrix. Mechanical guidance begins to take hold when the tip of the pollen tube is properly oriented in the stylar matrix. To examine this proposition closely, Iwanami (1953) dusted the stigma of Lilium auratum with pollen and followed the course of the pollen tube. The finding in this work that the normal site of pollen tube extension was along the longitudinally oriented cells surrounding the stylar canal led to the view that the pollen tube is directed mechanically by these cells. That the pollen tube is not guided in this situation by some polarized gradient of a chemical substance was subsequently established in an extensive range of delicate surgical experiments on the styles of L. auratum and L. longiflorum. In one experiment, when cut ends of the style were grafted to its morphologically inverted pieces, the artificial style was found to serve admirably as a medium for the growth of pollen rubes (Iwanami 1959). This could not have happened if chemical gradients were at operation in directing pollen tube growth. Some investigations on pollen-stigma interaction in maize and other Poaceae are also consistent with the view that pollen rubes are oriented in the direction of the ovary by mechanical constraints (Y. HeslopHarrison, Reger and Heslop-Harrison 1984b). Following germination of the pollen grain on the receptive stigmatic trichome in the species of Poaceae studied, the orientation of the pollen tube tip in the stylar tract seems to determine its unhindered growth toward the ovary. This occurs even in the absence of any visible structural adaptations directing the pollen tube toward the transmitting tracts. A long-distance influence, probably of a mechanical nature, on the direction of growth of the pollen tube through the style has also been indicated in Nicotiana alata (Mulcahy and Mulcahy 1987). Reports of the responses of pollen tubes to applied electric fields have nurtured the view that endogenous currents may direct pollen tube growth. Since directional growth of pollen tubes in vitro toward both the anode (Wulff 1935; Wang, Rathore and Robinson 1989; Nozue and Wada 1993) and the cathode (Marsh and Beams 1945; Zeijlemaker 1956; Nakamura et al 1991) has been observed, it is difficult to evaluate the significance
Pollination and Fertilization of the data. In Narcissus pseudonarcissus, the ovary is
electrically negative by about 100 mV with respect to the stigma; the potential voltage difference may be the guiding force of the pollen tube to the ovary (Zeijlemaker 1956). Given the delicate nature of the pollen tube and the need for micromanipulative procedures, the mechanism directing electrical field-induced pollen tube growth is likely to remain mysterious for some time. The attraction of pollen rubes to a source of Ca2+, to be described later, coupled with the evidence that calcium ions constitute one component of the endogenous current in the pollen tube (Weisenseel, Nuccitelli and Jaffe 1975; Weisenseel and Jaffe 1976) and that Ca2+ accumulates at the tip of the pollen tube (Jaffe, Weisenseel and Jaffe 1975; Reiss, Herth and Nobiling 1985), has engendered much speculation about the possible role of redistribution of these ions in influencing the direction of pollen rube growth through applied electrical fields. The most commonly mentioned mechanism guiding the direction of growth of pollen tube through the style is the attraction toward unspecified chemical substances in the style itself. The glandular cells of the style have drawn attention as a possible source of chemotropic substances. Following upon the work of several earlier investigators going back to the end of the 19th century, Tsao (1949) demonstrated chemotropism of pollen tubes to pistil parts of 9 out of 36 species tested on artificial media; it was further shown that the active factor from the pistil of L. superbum was diffusible into agar blocks, which then attracted pollen tubes. A later study by Miki (1964) confirmed and extended these results using L. longiflorum and L. japonicum and showed that the chemotropic factor is a heat-labile, diffusible, small molecule. Linck and Blaydes (1960) demonstrated chemotropism in four additional species, of which Clivia nobilis was particularly noteworthy for the attraction of pollen tubes not only by slices of the carpel but also by those of sepal, petal, and leaf. Attempts to elucidate the mechanism regulating pollen tube pathway in other species of Lilium have also been made. For example, pollen tubes of several hybrids of L. longiflorum are found to respond positively to the upper part of the style, although the activity is weak or absent in the mid and lower segments; the ovary displays no activity at all (Rosen 1961). In L. leucanthum and L. regale, the glandular cells lining the stylar canal have been identified as the source of the putative chemotropic factor, which is apparently distributed in a basipetal gradient (Welk, Millington and Rosen 1965). An influence of the ovary on
Fertilization: The Beginning of Sporophytic Growth
303
lowing the early studies and, until recently, other investigations on this topic have not been directed at chemotropic agents in the style. Pollen tubes of Pennisetum glaucum have been shown to respond positively to diffusates of the pistil tissue as well as to glucose; a totally different understanding of the in vivo metabolism of the style has emerged from the discovery that the pistil diffusate response is due to glucose produced by tissue-bound invertase action rather than to free glucose (Reger, Chaubal and Pressey 1992; Reger, Pressey and Chaubal 1992). Instead of positing the growth of pollen tubes to be under a chemotropic influence, some investigations have led to a model of pollen tube growth in the style as a directional cell migration along a gradient of a factor in the secretory matrix of the style; this model is analogous to a biochemical recognition-adhesion system. The initial experimental support for this model came from the observation Progress in our understanding of the interaction that latex beads approximating the size of pollen between the stylar matrix and the growing pollen tube tips, when introduced into the stylar transmittube has been bolstered by attempts to identify the ting tracts of certain plants, traveled to the site of chemotropic substance. A key piece of information the ovary at rates similar to those of pollen tubes in this context came from the work of Mascarenhas regardless of the stylar gravitational orientation and Machlis (1962b), who showed that in (Sanders and Lord 1989). Since substrate adhesion Antirrhinum majus the positive chemotropic factor is molecules probably confer this ability on the beads a small, heat-stable, water-soluble molecule associ- to travel in the stylar matrix, attempts have been ated with larger molecules. This finding led to the made to identify these substrate molecules and search for inorganic ions and to the discovery that their receptors in the style. A candidate molecule Ca2+ is chemotropically active for pollen tubes of A. identified in the substrate adhesion role is vitmajus, Narcissus pseudonarcissus, and Clivia miniata ronectin. Immunolocalization of vitronectin-like (Mascarenhas and Machlis 1962a, 1964). It was also proteins in cross sections of the pistil of Vicia faba found that the levels of Ca2+ present in chemotropi- showed strong reaction in the stylar transmitting cally active parts of the pistil of A. majus were well tract, suggesting a possible role for these proteins within the physiological range of the ion active in in guiding pollen tube growth in a manner similar bioassays of chemotropism. In the pistil of Petunia to cell migration in animals (Sanders et al 1991). hybrida, membrane-bound Ca2+ is found in a Jauh and Lord (1995) have further shown that by decreasing gradient from the stigma to the base of excising different lengths of the profusely pollithe style before pollination, and in the reverse direc- nated style of lily, it is possible to cause the tip of tion after pollination (Bednarska 1995). However, the pollen tube, in the absence of the remains of the there is some skepticism as to whether Ca2+ is a uni- pollen grain and the spent pollen tube, to convey versal chemoattractant for pollen tubes. In this con- the sperm cells to the ovules and effect fertilization. text, in vitro bioassays showing that pollen tubes of The tip of the pollen tube can be considered analoOenothera longiflora and other species grow in a ran- gous to a population of migrating cells that adhere dom manner in the presence of calcium compounds to one another and to the cells of the transmitting have had a particularly powerful impact on the tissue of the style during its journey to the ovule. skeptics (Glenk, Schimmer and Wagner 1970). It is The role of adhesive substrates in pollen tube also relevant to point out that there is a lack of a gra- growth has unfolded further with the demonstradient of ionic calcium in the style of A. majus, which tion that glycoproteins isolated from the stylar seems to argue for the presence of an additional fac- transmitting tissue of tobacco adhere to the surface tor as a chemotropic agent (Mascarenhas 1966a). of pollen tubes and are incorporated into the wall There has not been much research output on pollen of the pollen tubes growing in the style. It appears tube chemotropism in the style in the period fol- that within the style, pollen tube-bound enzymes pollen tube growth in Petunia hybrida has been demonstrated using a "semi-vivo" technique in which the ovary is removed at different times after pollination. Whereas the growth of pollen tubes in the style detached from the ovary was inhibited irrespective of the time after pollination, their growth in the intact pistil occurred normally toward the ovary (Mulcahy and Mulcahy 1985). Despite the fact that most of these observations are based on in vitro experiments that may have little or no parallel to events in vivo, they nonetheless suggest a complex chemical interaction between the growing pollen tube, the stylar matrix, and an ovule-derived guidance mechanism controlling pollen tube growth (Hiilskamp, Schneitz and Pruitt 1995). It is worth noting that use of the same methods described here has not led to the demonstration of pollen tube chemotropism in certain other plants (Cook and Walden 1967; Hepher and Boulter 1987; Kandasamy and Kristen 1987b).
304
Pollination and Fertilization
deglycosylate the proteins and thus release sugar (Tilton 1980). The filiform apparatus is believed to molecules that are presumably utilized by the produce most of the protein- and carbohydrate-rich growing pollen tubes (Wu, Wang and Cheung exudate in G. verrucosa, with minor contributions 1995). These observations suggest that stylar glyco- from the micropylar cells of the nucellus (Franssenproteins of tobacco serve an important role in Verheijen and Willemse 1993). The characteristics matrix adhesion and in cell metabolism. of the chemotropic systems guiding the pollen tube The general path of pollen tube entry into the to the ovule that do not involve Ca2+ provide interovule has been described convincingly in numer- esting comparisons with the growth of the pollen ous classical studies. These works showed that tube in the style, considered earlier, and its final after reaching the ovary, the pollen tube almost push into the embryo sac to be considered next. always grows downward over the outer integument, then between the inner ovary wall and the (ii) Passage of Sperm into the Embryo Sac inner integument of the ovule, toward the placenta. When the placental region is reached, the pollen This part begins with a brief account of the tube executes a sharp turn and continues its jour- ovary. As is well known, the ovary not only proney between the placenta and the outer integument vides the physical protection for the enclosed to the micropyle. At the micropyle, the tube grows ovules, but also harbors the tissues through which intercellularly through the nucellus to the embryo the pollen tube grows during the final phase of its sac. Observations of this path reported from con- existence. As described in Chapter 6, the embryo ventional histological studies were later confirmed sac is formed within the ovule, which itself is born in research using fluorescent microscopy to trace on the placenta. Before the pollen tube reaches the the pollen tube in the ovary (Y. Heslop-Harrison, embryo sac, it encounters the placenta, funiculus, Heslop-Harrison and Reger 1985; Reger, Chaubal and integuments of the ovule. Light microscopic and Pressey 1992). How does the pollen tube find studies have revealed three general pathways its way to the micropyle from the base of the style? through which the pollen tube enters the ovule and From indirect lines of evidence, the basic plan races to the embryo sac. In the most common route, appears to be one guided by a chemotropic attrac- known as porogamy, the pollen tube enters the tant produced at a site close to the micropyle. The ovule through the micropyle. In some plants, presence of such substances has been demon- mucilaginous canals and ancillary structures such strated in the micropylar region of ovules of as obturator, embellum, ponticulus, and papillae Paspalum orbiculare (Poaceae) (Chao 1971), that aid in the entry of the pollen tube into the Ornithogalurn caudatum (Tilton 1980), maize (Y. micropyle have evolved (Kapil and Vasil 1963; Heslop-Harrison, Heslop-Harrison and Reger Kapil and Bhatnagar 1975; Busri, Chapman and 1985), beet root (Bruun and Olesen 1989), Gasteria Greenham 1993; Martinez-Palle and Herrero 1995). verrucosa (Franssen-Verheijen and Willemse 1993), Two other preferred sites of entry of the pollen tube and Asclepias exaltata (Asclepiadaceae) (Sage and are at the chalazal end (chalazogamy) and on one Williams 1995), although the origin of the sub- side of the ovule (mesogamy), but they are used stances has not been clearly determined in all cases. only sporadically (Maheshwari 1950). These obserThe substance found in P. orbiculare is PAS positive vations suggest that pollen tubes may be directed and proteinaceous; it consists of secretion products to specific parts of the ovule by a gradient of chemof the nucellus and integuments, as well as the ical substances. Irrespective of the pathway taken product of dissolution of cells of the outer and by the pollen tube to the ovule, the tip of the pollen inner integuments close to the micropyle (Chao tube invariably makes contact with the embryo sac 1971, 1977). The substance in maize stains for pro- at its micropylar end. teins and pectic polysaccharides, although in the This leads to the consideration of the entry of light microscope its origin is not easily resolved (Y. the pollen tube into the embryo sac and the disHeslop-Harrison, Heslop-Harrison and Reger charge of sperm. It is an interesting commentary on 1985). In beet root, the middle lamella of the nucellar cells in the micropylar region comes apart dur- the state of uncertainty in this field that up to the ing the maturation process and these cells probably early 1960s, based on light microscopy, as many as secrete PAS-positive substances into the micropyle six different modes of entry of the pollen tube into (Bruun and Olesen 1989). Proteins and RNA are the the embryo sac involving one or both synergids main components of the exudate secreted by the were described. As summarized by Kapil and nucellar cap and inner integument of O. caudatum Bhatnagar (1975), the pollen tube might (a) enter between the two intact synergids, (b) enter between
Fertilization: The Beginning of Sporophytic Growth
the egg and the synergid, (c) enter between the embryo sac wall and one or both synergids, (d) penetrate one of the synergids, destroying it in the process, (e) destroy the synergid on contact, or (f) enter the cytoplasm of the synergid, which has degenerated before the arrival of the pollen tube. Route (d), by which the pollen tube enters the embryo sac and destroys one of the synergids in the process, has been described most frequently, whereas the other modes of entry have been documented in only a few cases. It should be pointed out that in the light microscope it is often difficult to distinguish between a healthy and a degenerating synergid in sections of the embryo sac clouded by pollen tube discharge. The era of electron microscopy led to a new level of understanding of the passage of the pollen tube into the embryo sac. What appeared as a basic pattern of pollen tube entry at the electron microscope level was described by Jensen and Fisher (1968a) in Gossypium hirsutum, which to date remains as the one species in which events associated with fertilization are best known. In G. hirsutum, as in many other species, the two synergids are identical before pollination. However, one of the synergids begins to disintegrate prior to the entry of the pollen tube into the embryo sac. In fact, pollination triggers the degeneration process, which is most frequently initiated when the pollen tube is in the style. The first signs of synergid collapse are irregular thickening and darkening of the plasma membrane and of the outer membranes of adjacent plastids and mitochondria, accompanied by a general increase in the electron density of the cytoplasm. When the pollen tube reaches the embryo sac, the tip of the tube grows through the filiform apparatus into the degenerating synergid. The degeneration of the synergid initiated before pollen tube entry into the embryo sac is accelerated after entry of the pollen tube. Signs of accelerated degeneration of the synergid are transformation of the plastids and mitochondria into ghosts of their original structure, collapse of vacuoles, and breakdown and disappearance of the plasma membrane. As to the fate of the pollen tube, after passage through the cytoplasm of the degenerating synergid, pollen tube growth comes to a standstill and then the pollen tube discharges its contents of sperm and cytoplasm. The observations on G. hirsutum imply that rather than being the result of mechanical intrusion of the pollen tube, synergid degeneration is caused by pollination and is actually required for normal pollen tube entry into the embryo sac. Fascinating
305
though this thesis is, its possible significance in the fertilization process was not widely appreciated until there was a long list of plants in which it was established that synergid degeneration begins before the arrival of the pollen tube and that the pollen tube consistently favors the degenerating synergid. In subsequent years, other species, such as barley (Cass and Jensen 1970), Zea mays (Diboll 1968; van Lammeren 1986), Epidendrum scutella (Cocucci and Jensen 1969c), Linum usitatissimum (Vazart 1969; Russell and Mao 1990), Quercus gambelii (Mogensen 1972), Proboscidea louisianica (Mogensen 1978b), Nicotiana tabacum (Mogensen and Suthar 1979; Huang and Russell 1992), Spinacia oleracea (Wilms 1981b), Crepis capillaris (Kuroiwa
1989), Glydne max (Dute, Peterson and Rushing 1989), Populus deltoides (Russell, Rougier and Dumas 1990; Rougier et al 1991), and Arabidopsis (Murgia et al 1993), have gained general acceptance as examples in which synergid degeneration occurs in apparent response to pollination. The pattern of cytoplasmic reorganization observed in one of the synergids following pollination varies widely, from a series of disordered changes in the organelles, as in G. hirsutum (Jensen and Fisher 1968a), to only a difference in the density of the cytoplasm due to an increase in the number of ribosomes, as in N. tabacum (Mogensen and Suthar 1979). That the loss of synergid integrity is not an artifact of the chemical fixatives used in these investigations has been confirmed by observing similar changes in synergids of cryofixed ovules of N. tabacum, although the full extent of cellular degeneration is delayed until pollen tube penetration (Huang, Strout and Russell 1993). Considerable variation in the condition of the synergid in relation to the timing of pollination and pollen tube growth has been observed in other plants. For example, in wheat, degeneration of both synergids is initiated in embryo sacs collected just prior to pollination; the degeneration cycle is completed in one synergid after entry of the pollen tube, whereas the other appears unaffected (You and Jensen 1985). In Brassica campestris, both synergids show degenerative changes in pollinated as well as in unpollinated flowers; this implies that degeneration of the synergid does not depend upon pollination (van Went and Cresti 1988a; Sumner 1992). In Beta vulgaris (Bruun 1987) and Oryza sativa (Dong and Yang 1989), one of the synergids readies to receive the pollen tube and begins to display signs of degeneration before anthesis of the flower. There are also reports of synergids remaining healthy even after pollination and
306
Pollination and Fertilization
pollen tube growth; apparently, when the pollen been employed to localize gross Ca2+ and loosely tube makes contact with the synergid, the degener- sequestered exchangeable Ca2+, respectively, in the ative program of the cell is set in motion. Examples embryo sac cells of cereal grains. The synergids in which this situation is found are Capsella bursa- were found by both methods to display high conpastoris (Schulz and Jensen 1968b), Petunia hybrida centrations of Ca2+, suggesting an active role for the (van Went 1970c), Ornithogalum caudatum (van ion as an erotropic substance for the attraction of Rensburg and Robbertse 1988), and Lilium longiflo- pollen tubes (Chaubal and Reger 1990,1992a). Since rum (Janson and Willemse 1995). Although the the pollen tube moves directionally to the synergid, 2+ paired synergids of Helianthus annuus do not show the existence of a continuous Ca gradient from the placenta to the synergid has been postulated any structural differences before pollination, degen(Chaubal and Reger 1990). In other cases, PAS-stainerative changes begin to surface in one of them gradually after pollination, whereas the other one ing substances secreted by the synergid probably remains perfectly healthy until after fertilization function as part of the pollen tube guidance system (Newcomb 1973a; Yan, Yang and Jensen 1991a). (Chao 1971; Bruun 1987; Willemse, Plyushch and Despite the fact that ultrastructural analysis has not Reinders 1995). However, as is true of most cells been carried out to the same depth in all of the growing in isolation within a tissue of different oriaforementioned cases, sufficient details are at hand gin, the movement of the pollen tube may be condito establish the essential validity of the hypothesis tioned by the physical state of the neighboring cells. that irrespective of the prior response of the syn- For example, compared to the intact cells of the ergid to pollination stimulus, the pollen tube grows female gametophyte, the degenerating synergid into one of the two synergids and then, as the pollen might offer a lower resistance to the growing pollen tube discharge occurs, degenerative changes set in. tube (Russell 1992). Moreover, when penetrating It is not, however, easy to fit into this model genera the synergid, the pollen tube conforms in a regithat lack synergids, such as Plumbago and mented way to the filiform apparatus. Since microtubules are associated with the filiform apparatus, Plumbagella. According to Russell (1982), arrival of the possibility that they may somehow be involved the pollen tube at the micropylar end of the embryo in giving directionality to the pollen tube also sac in Plumbago zeylanica is not heralded by the disdeserves attention (Cass and Karas 1974; B.-Q. array of a receptor cell; rather, the pollen tube peneHuang et al 1990). trates the wall ingrowths of the filiform apparatus at the base of the egg and grows in the intercellular The degeneration of the synergid before or after space between the egg and the central cell until it entry of the pollen tube is obviously an important reaches the vicinity of the nucleus near the chalazal step in the fertilization process. We do not yet end of the egg. A similar sequence of events occurs understand how this happens, even in plants in in Plumbagella micrantha (Russell and Cass 1988). It which pollination triggers the degeneration prois clear from this that, whatever else is going on, cel- gram. One common view of the nature of the stimlular degeneration is not an indispensable part of ulus is that it is a hormonal substance. In support of this, it has been shown that in cotton ovules grown successful pollen tube penetration. By and large, the signal for pollen tube penetra- in culture, synergid degeneration occurs when a tion into the embryo sac is provided by a synergid. mixture of IAA and GA is added to the medium. As The question arises, what is the nature of this sig- in ovules of pollinated flowers, only one synergid nal? The most persistent hypothesis to account for in cultured, unpollinated ovules responds to the the directional growth of the pollen tube is that it is hormonal signal Qensen, Schulz and Ashton 1977). attracted to a chemical substance. In the absence of It has also been shown that supplementation of the direct evidence for the presence of a chemotactic medium with GA alone triggers the degeneration substance in the synergids, a speculative view is of the synergid in ovules of unpollinated flowers that the release of Ca2+, especially from the degener- (Jensen, Ashton and Beasley 1983). These experiating synergid, is associated with pollen tube pene- ments by themselves do not prove that the hortration. Jensen (1965b) visualized inorganic ash in mone directly mediates the synergid response, but the vacuoles and cytoplasm of microincinerated they alert us to such a possibility in any hypothesis sections of cotton ovules. The ash, which is largely to explain synergid2+degeneration. Another suggesabsent from the egg, is assumed to contain Ca2+. tion involving Ca and related elements is that their loss after pollen tube discharge reduces the More recently, high-energy X-ray dispersion analyosmotic pressure of the synergid, leaving the way sis and antimonate precipitation methods have
Fertilization: The Beginning of Sporophytic Growth
open for an efflux of water from the cell and for subsequent shrinkage of the cell (Chaubal and Reger 1992b). Since there are two synergids in most species investigated, the next question is whether degeneration is a random event or whether one of the synergids is more predisposed to degenerate than the other. Based on examination of a limited number of species, indeed it appears that synergid degeneration is a preprogrammed, nonrandom event. In Helianthus annuus, the synergids are isomorphic; however, one is found closer to the placenta than the other. Using this feature for orientation purposes, it was found that in 81% of the ovules examined, the degenerating synergid was the one aligned with the placenta (Yan, Yang and Jensen 1991a). In barley, Mogensen (1984) found that in one sampling there was a greater tendency for the synergid closer to the placental attachment of the ovule (proximal) to degenerate than for the distal one. In these instances, therefore, the force for degeneration is derived from the sporophytic tissues due to physiological differences between the two synergids, such as the transport of nutrients or erotropic substances through the placental vasculature to one of the synergids. On the other hand, in Linum usitatissimum, some kind of gametophytic control seems to be operating in determining which one of a pair of synergids degenerates (Russell and Mao 1990). Based on the orientation of synergids relative to the egg, there is a greater preference for the left than the right synergid to degenerate. Since the volume and the surface area of the filiform apparatus of the left synergid are significantly lower than those of the right, the model predicts a degree of control by intrinsic factors that affect synergid function. A particularly fascinating and important aspect of the race of the pollen tube into the synergid is the fact that sperm cells keep pace with the distal end of the growing pollen tube. The mechanism by which sperm cells move down the pollen tube continues to attract the attention of investigators even to this day, but no satisfactory hypothesis has emerged. Because sperm cells share a number of properties with the generative cell, including shape and the cytoskeletal framework, comments made in Chapter 8 about the movement of the generative cell in the pollen tube are germane here. Although sperm cells have their own endowment of microtubules, it is doubtful whether they function as a motility system. Cass (1973) found that isolated sperm cells of barley show little or no directional movement, which should not be the case if their
307
microtubules are called into service. In view of the absence of microfilaments in the sperm, it comes as no surprise that no one has implicated them in gamete movement. Because of the negative evidence, it has been speculated that, like the generative cell, the sperm cells are propelled along the longitudinally disposed actin cytoskeletal elements of the pollen tube cytoplasm (J. Heslop-Harrison and Heslop-Harrison 1987). It is hard to justify the presence of microtubules in the short-lived sperm cells if they do not have a role in locomotion; it might be that sperm movement in the pollen tube is due to the combined efforts of the sperm microtubules and the actin system of the pollen tube. Once the pollen tube has made contact with the synergid, it grows some distance into this cell and readies itself to release sperm. Based on an examination of the cytoskeletal organization of the embryo sac of Plumbago zeylanica at various stages of pollen tube penetration, it has been proposed that positions of both microtubules and actin bundles change significantly to guide the pollen tube to the site of sperm delivery (B.-Q. Huang et al 1993). In other plants, the electron microscope has revealed several features of the final hours of the pollen tube and the life of the sperm as free cells. It is no accident that the first detailed description of the fate of the pollen tube contents in the synergid was provided in cotton, for this plant has proved to be remarkable for the highly regulated way in which fertilizationrelated events occur. In cotton, the arrest of pollen tube growth in the degenerating synergid is followed by the formation of a thick cap at the tip. The most dramatic event at this juncture is the appearance of a pore at the back of the tip through which the male gametes and the vegetative cell nucleus are released along with a generous amount of the pollen tube cytoplasm. Adding to the chaotic environment of the synergid is the discharge of polysaccharide spheres presumably constituting unused cell wall material (Jensen and Fisher 1968a). The force of the discharge, combined with changes in the physical properties of the discharged cytoplasm, apparently propels the vegetative cell nucleus and sperm cells to the chalazal end of the synergid, close to the egg nucleus. The end of the discharge is signaled by the formation of a callose plug over the pore, effectively preventing any cytoplasmicflowbetween the pollen tube and the synergid. In other plants, such as Capsella bursa-pastoris (Schulz and Jensen 1968b), Epidendrum scutella (Cocucci and Jensen 1969c), Petunia hybrida (van Went 1970c), Epilobium palustre (Bednara 1977),
308
Pollination and Fertilization
Quercus gambelii (Mogensen 1972), Plumbago zeylan-accomplished between the plasma membranes of the ica (Russell 1982), Plumbagella micrantha (Russell sperm and of the egg or the central cell, leading to the and Cass 1988), and Brassica campestris (Sumner formation of a bridge through which the sperm 1992), the discharge of the pollen tube occurs slides. The chalazal part of the egg, where the cell through a terminal aperture. In P. zeylanica, growth wall is incomplete, serves as the window where the of the pollen tube to the summit of the egg is often sperm fuses (Jensen and Fisher 1968a). In Petunia the prelude to pollen tube rupture. Sperm cells, hybrida, contact between the sperm and the female vegetative cell nucleus, and a small amount of cyto- reproductive cells is followed by cell fusion someplasm, which are released from the pollen tube, what analogous to the fusion of protoplasts (van find a place between the egg and the central cell Went 1970c). Regarding the fate of the male nucleus, (Russell 1982). Various mechanisms have been pro- some observations suggest that the chromatin of the posed to account for the rupture of the pollen tube, male nucleus becomes dispersed and indistinguishbut they seem incomplete in their explanations. able from the chromatin of the egg nucleus (Hu and One view is that as the pollen tube stops growing, Zhu 1979). A specific role for the degenerating synosmotic pressure built within it causes rupture at ergid in transporting sperm to the site of the egg and the weakest point. In P. zeylanica the pollen tube central cell has been described in Triticum aestivum. wall is relatively thin at the tip near the point of Transport is accomplished by a tubelike extension of rupture. Moreover, fibrillar elements near the aper- the degenerating synergid containing the pollen tube ture are less organized than in the proximal region, discharge and the sperm nuclei, which penetrates indicating both an apparent constraint in the syn- into the space between the egg, central cell, and the thesis of wall material and the onset of stress persistent synergid (Gao et al 1992a). (Russell 1982). Another view is that changes in oxySeveral events are played out in a compressed gen tension stop the growth of the pollen tube and time frame during gametic fusion in the synergidcause its rupture (Stanley and Linskens 1967). This less angiosperm Plumbago zeylanica. Immediately is based on the observation that low levels of oxy- after their release from the pollen tube, sperm cells gen promote bursting of pollen tubes in vitro and shed their investment of the inner plasma memin the style; whether pollen tubes experience a sim- brane of the vegetative cell and the incipient cell ilar stress in the embryo sac is not known. It has wall materials. At the same time, cell wall compoalso been suggested that the cessation of pollen nents of the egg are dispersed, thus exposing the tube growth and the rupture of its tip may be egg cell membrane. The stage is now set for caused by supraoptimal concentrations of Ca2+ in gametic fusion in which plasma membranes of the sperm cells become directly appressed to the the synergid (Chaubal and Reger 1990). plasma membranes of the egg and central cell. Fusion initiated at one location between the partic(iii) Fusion of Nuclei ipating cells spreads in a wavelike fashion through The observations terminating in sperm discharge the membrane and results in the transmission of lead to the consideration of the mechanism by which both nucleus and cytoplasm. Aggregates of actin gametic fusion is accomplished. The range of the found at the sperm delivery site between the egg search for this mechanism is wide and involves such and the central cell, designated as a corona, have aspects of cell behavior as movement, recognition, been assumed to cause the two sperm nuclei to and adhesion. The first issue concerns the actual align with their respective female nuclei (Russell transfer of sperm to the egg or the central cell. The 1983; B.-Q. Huang et al 1993). In the final act of two sperm cells are presumed to move separately, fusion, the sperm cell is incorporated into the cytoone to the egg and the other to the central cell. plasm of the egg or the central cell, and the plasma Electron microscopy has provided some pointers on membrane remaining from the fusion events dissithis, although it is not possible to determine the pates as fragments in the cytoplasm of the female mechanism that underlies the movement by inspec- cells involved. The events described here occur tion of electron micrographs alone. In Torenia fournieri within a time span of about 30 minutes. The actin (van der Pluijm 1964), the receptive synergid opens corona is also present between the synergids, egg, up a passage at a specific site at its chalazal end, pro- and central cell, and near the chalazal end of the viding direct access of the sperm to the egg and the degenerating synergid, during fertilization in IV. central cell. In terms of the end result, the transfer of tabacum (Huang and Russell 1994). The corona has the sperm cells to the egg and the central cell in cot- no counterpart in the embryo sac before one of the ton falls into the same type as T. fournieri. Fusion is synergids begins to degenerate.
Fertilization: The Beginning of Sporophytic Growth
309
tribute to the nuclear envelope. During nuclear fusion, pockets of cytoplasm containing fragments of organelles are trapped in the nuclear material, but they are probably forced out prior to the complete fusion of the nuclei (Jensen 1964; Jensen and Fisher 1967). Membrane fusion events similar to those observed during sexual karyogamy have been described during the fusion of the male nucleus with the polar fusion nucleus in Capsella bursa-pastoris (Schulz and Jensen 1973), Spinacia oleracea (Wilms 1981b), and Triticale (Hause and Schroder 1987). As shown in maize, karyogamy in the central cell occurs between the sperm nucleus and the diploid polar fusion nucleus or with one of the haploid polar nuclei (M61, Matthys-Rochon and Dumas 1994). Tangible evidence for the completion of double fertilization in conventional embryo sacs is the presence of dark-staining oval structures, referred to as X-bodies, in the degenerate synergid. Although these structures were variously interpreted by early embryologists, present opinion favors the view that they are nuclei or fragments of nuclei, presumably of the vegetative cell and of the degenerate synergid (Fisher and Jensen 1969; Cass and Jensen 1970; van Went 1970c; Huang, Strout and Russell 1993). A vexing question in angiosperm embryology is whether cell fusions during double fertilization are random or preferential events. In other words, are there traceable differences between sperm cells that fuse with the egg, on the one hand, and with the polar fusion nucleus, on the other? Based on light microscopic studies, the idea that both sperm cells Impatiens glandulifera, and Tradescantia paludosa, the are structurally identical and, hence, there is no sperm nucleus is already in the interphase stage at preferential fusion, held center stage for several the time of fusion, and the first zygotic division fol- decades. The first dent in this model was adminislows without delay (Kapil and Bhatnagar 1975). In tered by Roman (1948), who provided genetic eviGnetum gnemon (Gnetales), the gametic nuclei pass dence for the existence of nuclear dimorphism in through the S phase and double their DNA content the sperm cells of maize. It was found that when Bbefore they begin to court each other and fuse chromosomes do not separate during sperm cell (Carmichael and Friedman 1995). formation, the sperm containing an extra set of BElectron microscopy has added a wealth of new chromosomes fuses in the majority of cases with details about karyogamy in a selected list of plants. the egg. Although not striking at first glance as a In cotton, the sperm nucleus may enter the egg and revolutionary insight, the significance of this obsertravel the circumference of the egg before align- vation in the context of double fertilization is that ment with its nucleus preparatory to karyogamy. in maize, the presence of supernumerary B-chroKaryogamy is initiated when the two nuclei come mosomes seems to confer a competitive edge on together and the bordering ER fuse at several certain sperm cells as true male gametes. A fluorespoints and join their outer nuclear membranes. cence in situ hybridization method developed by This results in the formation of bridges between the Shi et al (1996) using a B-chromosome-specific nuclei. Next, the inner layers of the nuclear DNA sequence to identify and track maize sperm envelopes also fuse and the nucleoplasm becomes cells containing B-chromosomes now makes it poscontinuous at several points. When a fusion sible to follow up this question. nucleus is formed, membranes of both nuclei conOne of the most provocative pieces of evidence Nuclear fusions, or karyogamy, between the sperm and egg nuclei and between the sperm and central cell nuclei are at the heart of double fertilization and constitute a major reason for sustained interest in the process itself. GerassimovaNavashina (1960) used the cyclic state of the nuclei as an elegantly simple criterion to identify two types of karyogamy involving the egg and sperm. In the most common type, known as the premitotic type, the sperm nucleus fuses immediately with the egg nucleus and, after a short hiatus, the zygote nucleus passes through S phase and continues through G2 and mitosis to complete the first zygotic division. Illustrative examples of this type of karyogamy are found in members of Asteraceae and Poaceae. Premitotic karyogamy has been nicely demonstrated recently (Figure 11.4) in isolated egg and zygote protoplasts of barley (Mogensen and Holm 1995). In the postmitotic type, frequently observed in members of Liliaceae, the male and female nuclei approach one another and remain in close proximity; then follows an increase in the size of the nucleus and nucleolus of the sperm. Although the nuclei enter mitosis independently, a common metaphase plate is formed, leading to the formation of a zygote nucleus. The underlying basis of postmitotic karyogamy in the gymnosperm Ephedra trifurca (Gnetales) entails only the completion of S phase of the cell cycle by the gametic nuclei during the period of courtship before fusion occurs (Friedman 1991). Intermediate situations also prevail, depending upon the duration of karyogamy. For example, in Mirabilis jalapa,
Pollination and Fertilization
Figure 11.4. Premitotic karyogamy in isolated egg and zygote protoplasts of barley, (a) Nomarski optics photograph of an unfertilized egg. (b) Unfertilized egg, showing the fluorescence of DAPI-stained nucleus, (c) Egg showing the sperm nucleus beginning to fuse, (d) Fertilized egg showing the brightly fluorescing male chromatin. (e) Zygote showing the dispersion of the male and female chromatin.
in support of preferential fusion was provided in Plumbago zeylanica (Russell 1985). It was pointed out earlier that the paired sperm cells of P. zeylanica display heteromorphism in size, shape, and organelle contents. Because plastids are virtually absent from the cytoplasm of the sperm that is in
(f) Zygote before beginning of mitosis, (g) Zygote at the early prophase stage, (h) Zygote at the metaphase stage. (i) Late telophase stage of zygote mitosis, en, egg nucleus; sn, sperm nucleus. Scale bar = 5 |jm. (From Mogensen and Holm 1995. Plant Cell 7:487-494. © American Society of Plant Physiologists. Photographs supplied by Dr. H. L. Mogensen.)
physical association with the nucleus of the vegetative cell, it is possible to use paternal plastids as markers to follow the fate of the two male gametes in the embryo sac. By painstaking identification and counting in electron microscopic profiles of the number of plastids of paternal origin in the egg after
Fertilization: The Beginning of Sporophytic Growth
311
fertilization, it was established that the plastid-rich, (iv) Cytoplasmic Inheritance mitochondrion-poor sperm fuses with the egg cell. Because fusion of the plastid-impoverished sperm The transfer of male cytoplasm along with the with the central cell is as frequent as fusion of the nucleus during fusion of the sperm with the egg plastid-rich sperm with the egg, this brings us tan- raises an important question: Is the inheritance of talizingly close to considering the occurrence of a plastids and mitochondria uniparental or putative recognition event at the gametic level. In biparental? Chapter 3 described several instances of analogy to animal systems, this might involve the the first haploid pollen mitosis, during which the presence of different cell surface molecules on the generative cell is cut off with few or no plastids or two sperm cells that will recognize the egg and the mitochondria. This led to the view that in the majorpolar fusion nucleus, respectively. Although this ity of plants, cytoplasmic male DNA is not inherappears to be an attractive hypothesis, we should ited, which results in uniparental maternal not lose sight of the fact that it is based on the study organelle inheritance. In a few instances in which of one unusual species and that a great many prob- sperm cells are endowed with plastids and mitolems have to be overcome before gamete surface chondria, mechanisms also operate to exclude them recognition substances are identified. P. zeylanica from the egg prior to fertilization. Rarely, plastids has made an interesting contribution to our under- and mitochondria are transferred from the sperm to standing of the structural basis of double fertiliza- the egg during fertilization; however, as the fertiltion, one that is far richer than has been apparent ized egg begins to divide, it displays characteristic from studies of other angiosperms. uniparental inheritance. How does this happen? Mogensen (1996) has reviewed the elaborate Little is known about the existence of any cryptic pattern of recognition system on the female side. ways by which cytoplasmic organelle transmission Pennell et al (1991) have demonstrated the presence is restricted or enhanced during and following of a plasma membrane-bound arabinogalactan pro- genetic recombination in angiosperms. The classitein epitope in the paired sperm cells and in the egg cal method of excluding sperm cytoplasm from the cell, but not in the central cell, of Brassica napus; the egg at fertilization is by allowing passage of only differential accumulation of the marker protein in the nucleus into the egg cell, leaving the cytoplasm the egg and central cell may suggest a role for the outside. This is probably the case in cotton, where arabinogalactan protein in female gametic cell no trace of the sperm cytoplasm in any form or shape is detected in the egg (Jensen and Fisher recognition. To round out this account, brief mention should 1967,1968a). Ultrastructural evidence for the exclube made of reports of fusion of a second sperm sion of sperm cytoplasm and organelles during nucleus with egglike counterparts in the female karyogamy has also been reported in barley gametophytes of Ephedra and Gnetum. These fusion (Mogensen 1982, 1988). In the barley embryo sac, events have been billed as double fertilization; in E. an enucleate cytoplasmic body, approximating the trifurca, in which it has been examined in detail, the size of a prefusion sperm cell and containing a genproduct of second fertilization yields an embryo erous number of mitochondria and a limited numafter a brief free-nuclear interlude (Friedman 1990, ber of plastids, is frequently observed within the 1992; Carmichael and Friedman 1995). The direct degenerate synergid adjacent to a recently fertilanalogy of the situation in the gymnosperms to ized egg cell. This body has been interpreted as the that in angiosperms may be too facile a compari- entire male cytoplasm left out after fusion of the son, and the view of a likely origin of the egg and sperm. Since only one cytoplasmic body is endosperm in angiosperms from an embryolike observed in each postfertilization embryo sac structure in gymnosperms may be far-fetched. It examined, it is believed that uniparental inheridoes not take into account the fact that the product tance is restricted to the zygote and that the of the second fertilization event does not produce endosperm receives the paternal cytoplasm. What the nutritive tissue that is the hallmark of the causes the separation of the intact cytoplasm from angiosperms. In the absence of evidence to refute the nucleus is difficult to comprehend, although that the second fusion event is another manifesta- cytoplasmic degeneration may certainly contribute tion of polyembryony, for which gymnosperms are to this effect. In Populus deltoides, the sperm cells notorious, the thesis that double fertilization, in develop large vacuoles during their descent into the sense we understand the process in the pollen tube, and it is believed that this is the sigangiosperms, originated in gymnosperms cannot nal that leads to the separation of the cytoplasm from the nucleus at the time of fertilization be supported.
312
(Russell, Rougier and Dumas 1990). An unusual way of eliminating the male cytoplasm from the egg probably occurs in spinach in which plasma membranes of the two sperm cells initially fuse together to form a binucleate cell. It has been claimed that fusion of the male nuclei with the egg nucleus and with the polar fusion nucleus of the central cell is followed by sealing of the plasma membranes, thus excluding any remaining cytoplasm from the female cells (Wilms 1981b). Just as the complete elimination of the sperm cytoplasm at the time of fertilization offers an explanation for maternal organelle inheritance, so also might a process of progressive cytoplasmic diminution during sperm cell maturation explain maternal inheritance. Mogensen and Rusche (1985) first reported that during maturation of barley sperm cells, lateral projections containing mitochondria are frequently pinched off, resulting in a conspicuous reduction in the surface area, volume, and mitochondrial number of the sperm. In Nicotiana tabacum, cytoplasmic diminution, which occurs in the sperm cells as they descend into pollen tubes, is accomplished by the formation of enucleate cytoplasmic bodies, the spontaneous vesiculation of the tip of cytoplasmic extensions, and the accumulation of vesiculate bodies in the periplasm of the sperm cells (Yu, Hu and Russell 1992). It is difficult to reconcile these observations with those of a subsequent work, which showed that during fertilization in N. tabacum, male cytoplasm including heritable organelles is transmitted into the female target cells (Yu, Huang and Russell 1994). Since sperm cells do not produce nuclei denuded completely of their enveloping cytoplasm, it is likely that processes such as those described in barley and tobacco, combined with cytoplasmic elimination at the time of fertilization, might offer an explanation for maternal organelle inheritance in the majority of plants in both species, with the caveat that occasional biparental inheritance also occurs. Transmission of a part or all of the cytoplasm of the sperm cell in the act of fertilization certainly is the basis of biparental or paternal inheritance of organelles. In Chapter 3 it was reported that in some plants the generative cell is cut off with an appreciable number of plastids, which suggests that the plastids are transmitted to the sperm and subsequently into the egg cell during gametic fusion. This seems to be true where the fate of the sperm cell has been followed electron microscopically. Very clearcut, indeed, is the evidence for the transmission of sperm cytoplasm into the egg and central cell in
Pollination and Fertilization
Plumbago zeylanica, in which, based on size and fluorescent-staining differences, it was possible to discriminate between plastids and mitochondria of sperm cell origin from those present in the egg at the time of fertilization (Russell 1980, 1987; Sodmergen et al 1995). Electron microscopic evidence has augmented genetic data to help explain certain cases of uniparental, paternal inheritance of organelles. One such case is Medicago sativa, in which RFLP for plastid DN A has indicated a predominant paternal inheritance of this organelle in reciprocal crosses between two subspecies (Schumann and Hancock 1989). Paternal inheritance is also correlated with a high plastid content in the generative cell, although there are no significant differences in plastid abundance between genotypes known to differ in their male plastid transmission pattern (Zhu, Mogensen and Smith 1990, 1991). Likewise, there is no consistent correlation among genotypes between the mean number of plastid DNA aggregates (nucleoids) in the generative cell and the strength of male plastid transmission (Shi et al 1991). A quantitative, threedimensional analysis of the plastid distribution patterns in egg cells of genotypes designated as weak plastid transmitters or strong plastid transmitters has provided an explanation of this paradox. Apparently, eggs of weak plastid transmitters are characterized by a largely micropylar plastid distribution, whereas plastids of the strong females are positioned in the chalazal part of the egg. This would ensure that following the first division of the zygote, the plastid population of the weak females is segregated to the basal micropylar cell, which becomes the transient suspensor, and the apical organogenetic part of the embryo will be blessed with the plastids delivered by the sperm. On the other hand, in strong plastid transmitters, the apical cell, which inherits the plastids at the time of the first division of the zygote, will perpetuate maternal or biparental plastid transmission (Zhu, Mogensen and Smith 1993; Rusche et al 1995). This viewpoint assumes that fertilization is a passive event and that organelles stay put in whichever locations they find themselves at the time of nuclear fusion. As described in Chapter 12, electron microscopic studies have revealed wholesale rearrangement in the distribution of organelles in the egg cells of several plants following fertilization. A strong bias in favor of paternal inheritance of plastids also prevails in the interspecific cross Daucus muricatus X D. carota. The basis for this is
the presence of an exact replica of restriction fragments of D. carota plastid DNA in self-pollinated Fl
Fertilization: The Beginning of Sporophytic Growth
plants (Boblenz, Nothnagel and Metzlaff 1990). This is attributed to the presence of relatively few intact plastids in the egg of D. muricatus compared to the other species and to the presence of degenerate plastids (Hause 1991). However few in number the plastids may be, how these plastids are excluded from the zygote is an unsolved problem. Among the ideas suggested are the domination of paternal plastids and eventual elimination of the maternal plastid genome, differential plastid replication with the paternal plastids' holding an advantage, and the wholesale degeneration of maternal plastids. Investigations that take these ideas into account seem to be necessary to provide complete structural evidence for the genetic data for paternal plastid inheritance. In recognition of the fact that some male plastids invariably find their way into the egg cytoplasm at the time of fertilization, efforts have been made to explain plastid behavior in the zygote that results in uniparental inheritance. According to Zhu, Mogensen and Smith (1991), who have collated this information, elimination of male plastids from the zygote occurs by (a) differential degradation of male plastid DNA within the zygote; (b) recombination between different plastomes in the zygote, with subsequent DNA repair using the strong plastome of the female parent as the template; (c) differential multiplication of plastids of the two parents within the zygote and proembryo, with the maternal plastids' holding an edge; and (d) segregation of maternal plastids to the apical cell and paternal plastids to the basal cell in embryo types in which the apical cell of the two-celled embryo becomes the organogenetic part and the basal cell becomes a short-lived suspensor. Perhaps symptomatic of the primitive state of our understanding of the fate of the paternal plastids following fertilization is the fact that there are no markers to follow them in the chaotic environment of the zygote. 3. IN VITRO APPROACHES TO THE STUDY OF FERTILIZATION In the reproductive phase of angiosperms, fertilization and seed set are modulated by several external environmental and internal genetic factors. Among the former are female receptivity, the age and quality of pollen grains, and properties of the ambient atmosphere. In traditional plant breeding programs involving incompatible pollinations, prefertilization blocks such as the failure of pollen grains to germinate on the stigma, physiological disturbances in the directional growth of pollen
313
tubes through the style toward the embryo sac, and failure of fertilization itself often thwart seed set. When transfer of beneficial alien genes across interspecific and intergeneric barriers is attempted, such as in wide hybridizations, it is seen that in spite of normal pollination and fertilization, deleterious genes carried forward into the zygote cause embryo lethality and seed collapse. Since success in agricultural practices would depend upon devising techniques to control reproductive processes in plants to our advantage, methods to overcome these natural bottlenecks to sexual recombination have attracted a great deal of attention. As described in Chapter 10, advances made in our understanding of the molecular biology of incompatibility offer potential for the development of methods for overcoming this barrier to crossability in plants. Methods to overcome embryo abortion in wide hybrids are briefly considered in Chapter 12. The purpose of this section is to provide an overview of the in vitro methods that have been developed to study fertilization in angiosperms under controlled conditions. Detailed protocols for most of the in vitro approaches currently in use are given by Yeung, Thorpe and Jensen (1981). (i) Intraovarian Pollination The primary objective of in vitro approaches to study fertilization is to secure the encounter of the egg and sperm in a relatively uncomplicated way that bypasses pollination and pollen tube growth in the style. An extensive discussion would be required to recount all of the early attempts in the experimental manipulation of the fertilization process. Only the highlights can be described here; refer to the brief review by Tilton and Russell (1984) for a historical perspective of this work. In one of the first published full investigations of this kind, Bosio (1940) was able to induce fertilization in ovules of three species of Paeonia by combining excision of the style with actual deposition of pollen grains on the upper part of the ovary. Because it effectively bypasses stigmatic pollination and stylar growth of pollen tubes in a distinctive way, the technique is christened as "intraovarian pollination." Unfortunately, neither the technique nor the observations sparked much interest until Kanta (1960) used a modified method to induce fertilization and seed set in Papaver rhoeas. The experimental design involved emasculating receptor flowers before anthesis and injecting into the ovary small quantities of a suspension of pollen grains prepared in boric acid. This
314
Pollination and Fertilization
resulted in the germination of pollen grains, in the the filiform apparatus of the synergid as it does growth of pollen tubes and their entry into the during in vivo fertilization. embryo sac, and eventually, in fertilization. Results Test-tube fertilization also occurs by crossing acquired in terms of seed set demonstrated the suc- interspecific and intergeneric barriers. One of the cess of intraovarian pollination in P. rhoeas, as well most convincing examples is fertilization of culas in other members of Papaveraceae such as P. tured ovules of Melandrium album with pollen from somniferum, Argemone mexicana, A. ochroleuca, and M. rubrum, Silene schafta, and S. tatarica, all of Eschscholzia californica. Although reciprocal crossesCaryophyllaceae (Zenkteler 1967). Indeed, this between the diploid A. mexicana and the tetraploid appears to be a choice method for raising interspeA. ochroleuca failed due to inhibition of pollen tube cific hybrids because hybrid embryos complete growth in the style, intraovarian pollination offered their full-term development. Cultured ovules of M. the potential for eliminating this roadblock to fer- album are also found to favor pollen grains of other tilization when either species was used as the genera of Caryophyllaceae such as Cerastium female parent (Kanta and Maheshwari 1963a). An arvense and Dianthus serotinus, and fertilization extension of this method is to inject sperm directly ensues; however, the competence of zygotes to into the embryo sac; this approach holds promise form full-term embryos is found to fade after a few in plants where the sperm cells do not reach the divisions (Zenkteler, Misiura and Guzowska 1975). embryo sac after intraovarian pollination. Beginning with the works of Sladky and Havel (1976) and Gengenbach (1977a), a great effort has been made to identify the conditions for maxi(ii) Test-Tube Fertilization mizing fertilization and kernel development A step toward a more controlled encounter of under in vitro conditions in maize. These studies, the egg and sperm was achieved by culturing which basically have involved culturing pieces of unpollinated receptive ovules of P. somniferum in a the cob and pollinating the exposed silk or dustnutrient medium and dusting them with pollen ing the nucellar tissue with pollen grains, have grains. During the ensuing period of two to three examined the influence of donor plant genotype weeks, pollen germination, pollen tube growth, fer- (Gengenbach 1977b; Bajaj 1979; Higgins and tilization, embryogenesis, and endosperm develop- Petolino 1988), the ratio of ovary to cob tissue ment occur, as attested by the transformation of (Sladky and Havel 1976; Higgins and Petolino cultured ovules into mature seeds that survive on 1988), pollen quantity (Raman, Walden and the original medium. This was confirmed by histo- Greyson 1980), explant size (Higgins and Petolino logical preparations of ovules that established 1988), silk length (Dupuis and Dumas 1989), tissue beyond doubt the occurrence of double fertilization continuity of the silk (Booy, Krens and Bino 1992), in the test tube; hence, in analogy to animal sys- stage of spikelet development (Dupuis and Dumas tems, the technique is designated as "test-tube fer- 1990a), and temperature stress (Mitchell and tilization" (Kanta, Ranga Swamy and Maheshwari Petolino 1988; Dupuis and Dumas 1990b) on fertil1962). Using these general guidelines, the protocol ization frequency and seed set. Using this system, it was also successfully extended to other members has been shown that compared to the mass pollinaof Papaveraceae, as well as to members of other tion that occurs in nature, no more than a single families (Kanta and Maheshwari 1963b; Kameya, pollen grain is required to affect fertilization of an Hinata and Mizushima 1966; Zenkteler 1967; ovule, although the frequency of successful fertilZubkova and Sladky 1975). In a modified version of ization events varies greatly in different trials the protocol, whole pistils were cultured and polli- (Raman, Walden and Greyson 1980; Hauptli and nated in vitro with results similar to those obtained Williams 1988; Kranz and Lorz 1990). Since pollen with cultured ovules; this is analogous to conducttubes are able to travel a shorter distance within the ing stigmatic pollination in vitro. The technique style under in vitro conditions to affect fertilization was successful with, among others, Petunia violacea than in vivo, it appears that sperm cells do not (Shivanna 1965), Nicotiana tabacum (Dulieu 1966), undergo an extended period of maturation in the N. rustica (Rao and Rangaswamy 1972), Antirrhinum majus (Usha 1965), Dicranostigma pollen tube to become fusion competent (Dupuis franchetianum (Papaveraceae) (Rangaswamy and and Dumas 1989). Havel and Nova~k (1981) used Shivanna 1969), and Papaver nudicaule (Olson and genetic color characters of embryos and kernels to Cass 1981). In the last-mentioned example, the demonstrate double fertilization in vitro. The pollen tube enters the embryo sac by growing into development of kernels after in vitro pollination is not proof that the process of double fertilization
Fertilization: The Beginning of Sporophytic Growth
315
has occurred or that subsequent developmental (iii) In Vitro Fertilization programs have been successful, because occasional absence of embryos and abnormalities in From the developmental point of view, in vitro endosperm development are noted in apparently fertilization with isolated, single gametes presents normal-looking kernels (Schel and Kieft 1986). In an ideal opportunity to analyze karyogamy at the this case, the lesion is not in the fertilization level of the single cell, uninfluenced by the presprocesses but in the development of the embryo and ence of other cells. From the long-range point of endosperm. view, in vitro fertilization will provide a route for One aspect of test-tube fertilization technique to direct transformation of the egg using external be emphasized here concerns its potential use in sources of DNA and for transmission of alien cytoeliminating self-incompatibility barriers. In the pro- plasm to study organelle inheritance. Using the tocol used to overcome self-incompatibility in methods available for isolation of embryo sacs and Petunia axillaris, a placenta with attached ovules, sperm cells, a beginning has been made in the together with a short length of the pedicel, was cul- study of the conditions necessary for single gamete tured and the ovules were pollinated by dusting fusion and culture of artificial zygotes in maize. them with pollen. The result was prompt germina- The initial strategy in this work is to use cleverly tion of pollen grains, rapid growth of pollen tubes, engineered microcapillary techniques to select fertilization of the egg within about two days after sperm cells individually from ruptured pollen pollination, and eventual seed set (Rangaswamy grains and to select eggs and other component cells and Shivanna 1967). Additional manipulations from enzymatically treated embryo sacs. Selection have established that cultured ovules of self-incom- is followed by the transfer of gametes to a micropatible plants do not show any preference to the drop of a fusion medium (0.55 M mannitol) placed pollen grains they encounter, as long as they belong on a coverslip; pairs of gametes are fused electrito the same species. For example, when ovules on cally. By this method, fusion has been accomone placenta in the ovary are left unpollinated and plished not only between isolated, single egg and those on another placenta in the same ovary are pol- sperm cells, but also between single sperm cells linated with self- or cross-pollen grains, the unpolli- and single synergids and central cells. The nated ovules wilt, whereas fertilization and seed set egg-sperm fusion products, when nurtured in a occur in both selfed ovules and crossed ovules. The microculture surrounded by a suspension of feeder efficacy of self- and cross-pollinations for fertiliza- cells of maize embryo origin, develop into multition in vitro was demonstrated by the absence of cellular structures (Kranz, Bautor and Lorz 1991a, differences in seed set when ovules on one placenta b). Electrofusion between isolated male and female are selfed and those on another placenta in the same gametes of wheat resulting in multicellular strucovary are crossed (Rangaswamy and Shivanna tures has also been reported (Kovacs, Barnabas and 1971a). In yet another approach, when stigmatic Kranz 1995). In an alternative method, developed pollination and placental pollination are performed without the application of electrical pulses (Figure on the same pistil in culture, the former is found to 11.5), the essential environment for the egg and lead to seed set only with cross-pollination, whereas sperm cell of maize to fuse was found to be a with the latter, both self- and cross-pollinations are medium containing a high concentration of CaCl2 found to be equally effective (Rangaswamy and at pH of 11.0 or a medium containing mannitol and Shivanna 1971b). These protocols are collectively CaCl2 (Faure, Digonnet and Dumas 1994; Kranz designated as "placental pollination." Placental pol- and Lorz 1994). Later studies showed that the first lination has also been successfully used to over- few rounds of division of the artificially created come incompatibility barriers in P. hybrida (Niimi zygote to form the multicellular mass and the 1970; Wagner and Hess 1973), Brassica campestris transformation of the mass into an embryo bear (Zenkteler, Maheswaran and Williams 1987), and obvious similarity to in vivo development of the zygote. Formation of plants occurred by germinaLilium longiflorum 0anson 1993). tion of the bipolar embryos and, occasionally, by Zenkteler (1990) has compiled a list of plants in polyembryony and by organogenesis from the which fertilization and seed development have multicellular mass (Kranz and Lorz 1993; Kranz, been achieved by the aforementioned methods. It von Wiegen and Lorz 1995). should be noted that since fertilization events occur in eggs enclosed within ovules, these experiments What is the possibility that the plants regenerfall short of a true in vitro fertilization system as ated are the products of spontaneous division of understood in animal embryology. the egg in the absence of fusion with a sperm? This
316
Pollination and Fertilization
Figure 11.5. Phase contrast micrographs of in vitro adhesion and fusion between isolated egg (EC) and sperm cells (SC) of Zea mays. Adhesion between gametes occurred 1 minute 50 seconds after transfer to a fusion medium containing 5 mM CaCI2 (a). Progress of gametic fusion is shown in sequences (b) to (i) taken at intervals over a period of 9 seconds, as indicated on the lower right corner. After fusion
(i), the sperm nucleus (SN) is still visible in the egg cell protoplast. Scale bar = 15 urn. (Reprinted with permission from J.-E. Faure, C. Digonnet and C. Dumas 1994. An in vitro system for adhesion and fusion of maize gametes. Science 263:1598-1600. © 1994. American Association for the Advancement of Science. Photograph supplied by Dr. C. Dumas.)
does not appear to be the case, as karyological, phenotypic, and segregation analysis have established that the regenerated plants are indeed hybrids of both egg and sperm parents. Electron microscopic analysis has shown that the fusion of the egg and sperm culminating in karyogamy involves diverse activities that occur in a regimented way at their own times after cell contact is established (Faure et al 1993; Kranz, von Wiegen and Lorz 1995). Of further interest are the increases in membrane-bound Ca2+, calmodulin, ER, and mitochondria in the zygote and in the deposition of cellulose around it (Tirlapur, Kranz and Cresti 1995). Suggestive of the existence of a block to polyspermy at the gamete level is the finding that the addition of extra sperm cells in the vicinity of fertilized eggs does not lead to their penetration of, or fusion with, the egg (Faure, Digonnet and Dumas 1994). A small num-
ber of isolated egg cells were found to divide in the absence of fertilization when activated by a pulse of a high concentration of 2,4-D, although the microcalluses produced did not regenerate plants. Divisions were also induced in isolated maize egg cells after fusion with sperm of other members of Poaceae, such as Coix lacryma-jobi, Hordeum vulgare, Sorghum bicolor, and Triticum aestivum (Kranz, von
Wiegen and Lorz 1995). These observations not only show that it is possible to break down species barriers in the natural system of cross-incompatibility but also pave the way for interesting experiments for the artificial control of fertilization. Alternative methods to study the intimate details of fertilization under in vitro conditions are observations of living embryo sacs isolated enzymatically from pollinated ovules, and microinjection of active sperm nuclei into the egg and central cell
Fertilization: The Beginning of Sporophytic Growth
317
of isolated embryo sacs (Zhou 1987a; MatthysRochon et al 1994). We have a long way to go before in vitro fertilization in angiosperms will become a routine procedure that can be manipulated without the aid of artificial fusion methods such as electrofusion. Considering the state of the art only a few years ago, the stepwise progress made to facilitate the encounter of egg and sperm for observation and analysis is impressive. As fertilization in angiosperms is most certainly mediated by complementary receptor sites on the sperm and the egg plasma membranes, the stage is now set for the exploration of the nature of these receptor molecules. Streamlining of the in vitro fertilization system will also lead to a generalized picture of the mechanisms that block polyspermy, an area in which the evidence is too flimsy for any firm conclusions.
that divided in the embryogenic pathway was not identified with any degree of certainty. Although it seems clear from later investigations that in some cases embryos originate by the division of the unfertilized egg, evidence for their origin from the synergids and antipodals has been circumstantial (Ferrant and Bouharmont 1994). Compared to pollen embryogenesis, gynogenesis of limited value in developmental and molecular studies due to the unpredictable nature of the response of cultured explants and the difficulty of tracing the origin of embryos to specific cells of the embryo sac. 4. GENERAL COMMENTS
It is apparent from the foregoing discussion that a multifaceted approach can be of fundamental importance in the analysis of fertilization in angiosperms. The establishment of the role of the degenerating synergid, formulation of the concept (iv) Gynogenesis of the male germ unit, isolation of relatively pure Gynogenesis refers to the activation of cells of populations of male and female gametes, and the embryo sac to form embryos in the absence of demonstration of in vitro fertilization stand as landfertilization and is a process with dual conse- marks in the progress made in this field during the quences, namely, the substitution of other stimuli last three decades. In a way that is unique to for fertilization and the construction of a complex angiosperm reproduction, the phenomenon of douembryo and plant. Gynogenesis is generally facili- ble fertilization is beginning to make sense at the tated by the in vitro culture of the flower, ovary, and molecular level when we consider the evidence, ovule before pollination. Thus, in contrast to the in albeit limited, in favor of sperm dimorphism. That vitro methods described earlier, which promote an one sperm distinguished by certain ultrastructural egg-sperm encounter, gynogenetic methods essen- features almost always seeks out the egg cell for tially seek to prevent fertilization. Embryos that are recombination indicates that cell surface molecules formed in the absence of fertilization are haploids may determine this preferential union. at the cellular level, and for this reason, gynogeneThe successful demonstration of in vitro fertilsis is often considered, along with pollen embryo- ization does not necessarily make the task of genesis or androgenesis (Chapter 17), as a means of explaining the phenomenon of gamete recognition obtaining haploid plants in quantity for breeding easy. There are few, if any, available facts concernpurposes. ing cell surface ligands and receptors that must be Culture of unpollinated ovaries of barley in a involved in gamete recognition and fusion. The mineral-salt medium led to the first reported next logical step in the development of this imporinduction of gynogenetic embryos and plantlets tant field of plant embryology is to establish in (San Noeum 1976). This work was followed by vitro conditions involving minimum experimental reports of plant regeneration from cultured unpol- manipulations for the fusion of isolated egg and linated flowers, ovaries, or ovules of wheat, rice, sperm cells in a few model systems. It is clear that barley, maize, tobacco, lily, sunflower, beet root, our ability to manipulate gametes of angiosperms Gerbera jamesonii, Lactuca saliva (lettuce; Asteraceae), has improved so rapidly that it is unrealistic to preCucurbita pepo, and Populus sp. (San and Gelebart dict the rate of progress in the biochemical and 1986; Yang et al 1986; Zhou et al 1986b). In these molecular characterization of gametes. It is likely to investigations, the origin of embryos and plantlets be determined as much by the availability of new in the absence of fertilization was attributed to the and simple techniques for manipulation of gametes haploid cells of the embryo sac, but the specific cell as by the interest and expertise of the researchers.
SECTION III
Zygotic Embryogenesis
Chapter 12
Developmental Biology of the Endosperm 1. Histology and Cytology of the Endosperm (i) Structure and Development of the Endosperm (ii) Cellularization of the Endosperm (iii) DNA Amplification 2. Nutritional Role of the Endosperm (i) Endosperm Haustoria (ii) Chemical Composition of the Endosperm (iii) Embryo Abortion and Endosperm Development 3. Gene Action during Endosperm Ontogeny (i) Endosperm Mutants (ii) Synthesis of Housekeeping Proteins 4. Synthesis and Accumulation of Endosperm Storage Reserves (i) Mutants Defective in Starch Synthesis (ii) Storage Protein Synthesis 5. Analysis of Storage Protein Gene Expression (i) Maize Storage Protein Gene Expression (ii) Barley Storage Protein Gene Expression (iii) Wheat Storage Protein Gene Expression (iv) Storage Protein Gene Expression in Rice and Oat 6. Cloning and Characterization of Endosperm Storage Protein Genes (i) Zein Genes (ii) Hordeins and Other Genes from Barley (iii) Gliadin and Glutenin Genes (iv) Rice Glutelin and Other Genes 7. General Comments
The primary endosperm nucleus, born out of fusion of the second sperm cell with the polar fusion nucleus during double fertilization, is the starting point for the development of the endosperm. Repeated divisions of this nucleus within the confines of the central cell generate a
cluster of free nuclei or a cellular tissue known as the endosperm. Although the endosperm is reckoned as a triploid tissue in the vast majority of angiosperms, the ploidy level of the fusion nucleus obviously determines the final nuclear constitution of this tissue. Theoretically, the ploidy level of the endosperm might be expected to vary from diploid in the Oenothera type of embryo sac, to pentaploid in the Penaea, Fritillaria, and Plumbago types, and as high as 9N in the Peperomia type. Although both diploid and pentaploid endosperms have been described, it is, however, doubtful whether a 9N endosperm is formed as a permanent tissue in any seed types investigated. The endosperm has been studied from a number of viewpoints using light and electron microscopy, tissue culture, and biochemical and molecular techniques. For many years, the patterns of division of the primary endosperm nucleus, the morphological nature of the tissue formed, and the interaction of the tissue with the developing embryo were of prime concern to investigators whose studies have been an important factor in revealing that the developing embryo is nourished by the food materials of the endosperm. During the last two decades, a large part of the research on the endosperm has been devoted to the molecular biology of this tissue, and significant information about the regulation of synthesis and the accumulation of storage materials in the endosperm has been obtained. All through these periods of activity, the endosperm of cereal grains, as a repository of reserve food materials used in human and animal nutrition, has been subject to extensive biochemical investigations. It is perhaps not surprising, therefore, that the multifaceted research to improve the quality of our cereal grains that is conducted in laboratories all over the world is 321
322 in reality concerned with the control of endosperm development. Apart from the essential interest in it as a byproduct of fertilization, the endosperm presents basic problems relevant to our understanding of progressive embryogenesis. The objective of this chapter is to consider the endosperm from such aspects that are germane to its function in the developing seed following double fertilization. First, attention will be paid to the period of division and differentiation of the primary endosperm nucleus. Second, there is a considerable amount of evidence showing that genetic characters play a major role in determining the final form of the endosperm in certain plants. The various genetic factors controlling endosperm development will be briefly considered. Third, information will be presented concerning gene expression pattern in the endosperm, with emphasis on the synthesis and accumulation of storage proteins. Much of the early work on the morphology, cytology, and genetics of the endosperm has been considered by Brink and Cooper (1947b) in an exhaustive article. Reviews of relatively recent vintage are by Bhatnagar and Sawhney (1981) and Vijayaraghavan and Prabhakar (1984) on endosperm developmental morphology; Olsen, Potter and Kalla (1992) on the cytology and molecular biology of cereal endosperms; Lopes and Larkins (1993) on gene regulation; and Shewry (1995) on biochemistry. These are immensely challenging topics addressed by a considerable accumulation of literature, and so some degree of subjective judgment has gone into the selection of material to be discussed here.
1. HISTOLOGY AND CYTOLOGY OF THE ENDOSPERM The ontogeny of the endosperm has been followed in many different species to provide a descriptive basis for future biochemical and molecular investigations. These studies have reinforced the view that the endosperm is a relatively simple and amorphous tissue marked by the presence of only a few differentiated cell types. In spite of the fact that fusion events during double fertilization are almost simultaneous, it is generally acknowledged that the division of the primary endosperm nucleus takes precedence over that of the zygote. Little is known about the signal transduction system that potentiates the division of the primary endosperm nucleus ahead of the zygote or of the factors that induce the development of the primary endosperm nucleus into a compact tissue.
Zygotic Embryogenesis
Estimates of the development of the endosperm prior to division of the zygote range from the production of a few isolated nuclei to the formation of a fully developed tissue. A quantitative analysis of mitotic activity in the primary endosperm nucleus and the zygote in Triticum aestivum is particularly revealing of the shift in the balance of competing genomes. In plants reared at 20 °C, the primary endosperm nucleus begins to divide as early as 6 hours after pollination, whereas the division of the zygote is delayed until about 22 hours. The primary endosperm nucleus completes at least four rounds of nuclear divisions by the time the zygote divides, so that by 24 hours after pollination, at least 16 free endosperm nuclei are drifting randomly in the central cell around a two-celled proembryo. By the fifth day after pollination, the endosperm has already amassed more than 5,000 cells, in contrast to little more than 96 cells in the proembryo (Bennett et al 1973b). The same general trend in the rates of nuclear or cell doubling in the developing endosperm and embryo is maintained in other species of Triticum, in Hordeum vulgare, H. bulbosum, and Secale cereale, and in several genotypes of Triticale, with differences of similar magnitude in their cell numbers resulting (Bennett, Smith and Barclay 1975). To the extent that the endosperm possesses some nutrient value even at this early stage in its ontogeny, the interval between the division of the primary endosperm nucleus and the division of the zygote has significance in the nutrition of the zygote. Although the diploid polar fusion nucleus divides spontaneously in apomictic species, autonomous division of this nucleus may also occur in cultured, unpollinated ovules of certain amphimictic angiosperms (M61, Betka and Wojciechowicz 1995). Ohad et al (1996) have isolated a female gametophyte mutation in Arabidopsis termed fertilization-independent
endosperm (fie) that allows for replication of the diploid polar fusion nucleus and endosperm development without fertilization. Continued nuclear divisions in the central cell without the trigger of fertilization argue for a general molecular explanation based on cell cycle control of the fusion nucleus.
(i) Structure and Development of the Endosperm
The patterns of division of the primary endosperm nucleus have served as the basis for a simple classification of endosperms into nuclear, cellular, and helobial types. In the nuclear type, the
Developmental Biology of the Endosperm
323
Figure 12.1. Endosperm development in Ranunculus sceleratus. (a) Longitudinal section of the ovule showing the embryo sac surrounded by two to three layers of nucellar cells and a single layer of cells of the integument. The vacuole is lined by a thin layer of endosperm cytoplasm containing free nuclei, (b) A whole mount of the embryo sac containing 128 endosperm nuclei. Arrows point to walls in
the micropylar region, (c) In this longitudinal section of the ovule, the endosperm has been completely partitioned into uninucleate, open compartments. AC, antipodal complex; AW, anticlinal wall; CV, central vacuole; Em, embryo; En, endosperm; EW, embryo sac wall; I, integument; N, nucleus; NC, nucellus; S, starch. Scale bars = 100 urn (From Chitralekha and Bhandari 1993.)
first several rounds of division of the primary endosperm nucleus are unaccompanied by cytokinesis and thus result in a few thousand free nuclei in the central cell. During the early endosperm ontogeny, the nuclei are scattered throughout the central cell, or are pushed to its periphery by a burgeoning vacuole, or stream into incipient haustorial processes. Ultrastructural observations indicate that the cytoplasm of the central cell housing the free nuclei is undistinguished and is not enriched for any special organelles, although it is considered to be metabolically active (Marinos 1970a; Schulz and Jensen 1974, 1977; Ashley 1975a; Pacini, Simoncioli and Cresti 1975; Gori 1987; Dute and Peterson 1992; Chitralekha and Bhandari 1993). One might assume that as part of the distinction that characterizes the nuclear type of endosperm, the nuclei remain free throughout the ontogeny of the tissue, but this is not always the case. Indeed, close to the final stage of endosperm ontogeny in the majority of species examined, free nuclei and the incipient cytoplasmic domain around them are sequestered with walls and become cellular (Figure 12.1). Following the initial phase of cellularization, invaginations of the plasma membrane similar to transfer cells are seen on the wall of the endosperm cells surrounding the embryo in Vigna sinensis (Hu, Zhu and Zee 1983), Zea mays (Schel, Kieft and van Lammeren 1984; Davis, Smith and Cobb 1990;
Charlton et al 1995), and Vicia faba (Johansson and Walles 1993a, 1994). To emphasize the structural specialization of the basal endosperm transfer cells of maize, it has been pointed out that transcripts and the protein product of a cDNA clone isolated from a 10-day-old endosperm are restricted to these transfer cells (Hueros et al 1995). Whereas the endosperm in the seeds of most legumes is absorbed during seed development, soybean and certain other legumes have a partial or fully developed cellular endosperm. In soybean, the outermost layer of the cellular endosperm is incorporated into the seed as a single layer of cells around the embryo (Yaklich and Herman 1995). The flexibility of the nuclear type of endosperm is illustrated in cereals in which a coenocytic stage with free nuclei precedes a cellular configuration. Later, cells of the outermost layer of the endosperm differentiate into protein-rich aleurone cells, and the inner cells become filled with starch and storage proteins (Figure 12.2). This basic histology of endosperm development has been described repeatedly in wheat (Buttrose 1963a; Evers 1970; Mares, Norstog and Stone 1975; Morrison and O'Brien 1976; Mares et al 1977; Morrison, O'Brien and Kuo 1978; Briarty, Hughes and Evers 1979; Fineran, Wild and Ingerfeld 1982; van Lammeren 1988a), barley (Forster and Dale 1983; Bosnes et al 1987; Engell 1989; Bosnes, Weideman and Olsen
324
Figure 12.2. A section of the endosperm of Hordeum vulgare showing the peripheral aleurone layers and the inner starchy cells. AL, aleurone layer; M, mitosis in the aleurone cells; NUC, nucellus; PB, protein body; S, starch granule; SAL, subaleurone layer; SE, starchy endosperm. Scale bar = 5 um. (From Bosnes, Weideman and Olsen 1992.)
1992), maize (Randolph 1936; Kiesselbach 1949; Reddy and Daynard 1983; Kowles and Phillips 1985), and rice (Terada 1928; Juliano and Aldama 1937; Hoshikawa 1976; Ellis, Gates and Boulter 1987; Ramachandran and Raghavan 1989). Although cell volume in wheat endosperm continues to increase up to 35 days after anthesis, cell divisions plateau off at 16-20 days. In this time course, cellularization is accomplished within 2 - 4 days after anthesis during which period a marked decrease in cell volume is registered (Briarty, Hughes and Evers 1979; Gao et al 1992b). Whereas the aleurone layer in wheat and other cereals is only one cell layer thick, in barley it is at least three cell layers thick; development of the barley endosperm has been claimed to show a pattern in which cells of the starchy endosperm have a shorter proliferative period than those of the aleurone cells (Cochrane and Duffus 1981; Kvaale and Olsen 1986). However, it is doubtful whether there is a relationship between the number of layers of cells in the aleurone and a prolonged mitotic activity because cell divisions have been shown to continue in the single aleurone cells of maize (Randolph 1936), wheat (Evers 1970), and rice (Ellis, Gates and Boulter 1987; Ramachandran and Raghavan 1989) even after cessation of divisions in the starchy cells. Cell division data in conjunction with anatomical studies have led to the separation of endosperm development in barley into syncytial, cellularization, differentiation, and maturation phases. Most of the interesting changes occur during the differentiation phase, when the peripheral endosperm cells become electron dense and highly vacuolate
Zygotic Embryogenesis
and form the first aleurone cells at the placental region known as the crease or groove. At the same time, the large cells in the central endosperm are partitioned into smaller cells. In the maturing grain, aleurone cells adjacent to the crease area become thick walled in comparison to the rest of the aleurone tissue. In the mature endosperm are recognized, according to their disposition, three types of starchy cells, namely, irregular cells at the wings, prismatic cells that fan out from the crease area, and subaleurone cells (Bosnes, Weideman and Olsen 1992). During grain maturation, the contents of the crease aleurone cells are transformed into membrane-enclosed blobs of cytoplasm containing ribosomes, plastids, mitochondria, vacuoles, and a generous supply of dictyosomes. There seems little doubt that if the membrane surrounding the islands of cytoplasm is a plasma membrane, the crease aleurone cells might function as transfer cells (Cochrane and Duffus 1980). Aleurone cells in the crease region of the endosperms of Setaria lutescens (Rost and Lersten 1970), Echinochloa utilis (Zee and O'Brien 1971), and Sorghum bicolor (Giles, Bassett and Eastin 1975) (all of Poaceae) have structural features of typical transfer cells with wall invaginations. In developing wheat grain, wall ingrowths are confined to the crease aleurone and subaleurone cells that lie immediately below the aleurone layer. Moreover, the development of the crease aleurone cells is accompanied by a massive proliferation of the nuclear envelope that results in the accumulation of large amounts of rough ER (Mares, Jeffery and Norstog 1976; H. L. Wang et al 1995). In many cereals there are, besides the aleurone transfer cells, and often bordering the central cell containing the developing endosperm, elongate nucellar cells known as nucellar projections, also decorated with wall ingrowths (Cochrane and Duffus 1980; Smart and O'Brien 1983; Wang, Offler and Patrick 1995). A realistic evaluation of the crease aleurone cells and nucellar projections is that they amplify severalfold the plasma membrane surface area available for solute transport from the mother plant to the endosperm. In cereals, the interface between the embryo and the endosperm is a single layer of aleurone cells known as the germ aleurone. The cells of the germ aleurone layer are free of the aleurone grains that fill the normal aleurone cells of the rest of the endosperm (Bechtel and Pomeranz 1977; Cochrane 1994). Besides in Poaceae, a nuclear endosperm is found in Cucurbitaceae, Fabaceae, Arecaceae, and Proteaceae. Maheshwari (1950) has listed isolated
Developmental Biology of the Endosperm
genera showing the nuclear type of endosperm in additional families of monocots and dicots. In members of certain families, such as Acanthaceae, Cyrillaceae, Gesneriaceae, Icacinaceae, Loasaceae, Loranthaceae, Santalaceae, and Scrophulariaceae, in which the endosperm is of the cellular type, the first and subsequent divisions of the primary endosperm nucleus are followed by cytokinesis. This results in the formation of a population of cells that fills the entire central cell and may even extend beyond its boundary. The helobial type is a less common form of endosperm development and is found only in the monocots. The definitive feature of the helobial type of endosperm is the asymmetrical partitioning of the central cell into a large micropylar cell and a small chalazal cell following the first division of the primary endosperm nucleus. Subsequently, the micropylar cell executes a brief free nuclear phase followed by cellularization, whereas the chalazal cell remains undivided, or is partitioned into small cells, or, more often, divides in a limited way unaccompanied by cytokinesis to become coenocytic (Vijayaraghavan and Prabhakar 1984). As part of overall accounts of embryogenesis and seed formation, descriptions of endosperm development in an increasing number of plants continue to appear in the botanical literature, but they have not introduced any new concepts on the structure and morphology of this tissue. (ii) Cellularization of the Endosperm
As the endosperm is the major storage tissue of developing seeds of many plants, the importance of the cellularization lies in the fact that the accumulation of storage products and their precursors occurs in whole cells and not in isolated nuclei. Apart from this interest, the formation of cell walls in the endosperm presents basic problems relevant to the cytological basis underlying delayed formation of walls between isolated nuclei dispersed in an amorphous cytoplasm. Three questions relating to wall formation in the endosperm have attracted a great deal of interest and discussion: Does cytokinesis take place by the formation of a phragmoplast? What is the role of microtubules in the assembly of walls? What influence does the embryo sac wall exert in the cellularization of the endosperm? Observations based on light and electron microscopy form the basis for most descriptions of wall formation during endosperm development. Whether a phragmoplast is formed during cytokinesis between free nuclei of the endosperm is an issue
325
that has vexed some cytologists. Based on studies of free nuclear stages of endosperms of Helianthus annuus (Newcomb 1973b; Yan, Yang and Jensen 1991b), Stellaria media (Newcomb and Fowke 1973), Triticum aestivum (Morrison and O'Brien 1976; Mares et al 1977), Quercus gambelii (Singh and Mogensen 1976), Gossypium hirsutum (Jensen, Schulz and Ashton 1977), Papaver somniferum (Bhandari, Bhargava and Chitralekha 1986), Phaseolus vulgaris (Yeung and Cavey 1988), Glycine max (Dute and Peterson 1992), and Ranunculus sceleratus (Chitralekha and Bhandari 1993), and of helobial endosperms of Haemanthus katherinae (Newcomb 1978) and Asphodelus tenuifolius (Liliaceae) (Chitralekha and Bhandari 1992), the likely sequence seems to be the centripetal growth of the embryo sac wall toward the central vacuole, dividing the peripheral layer of cytoplasm into uninucleate open compartments independent of phragmoplasts and cell plates; the fusion of the walls with each other laterally at their origin; and nuclear divisions accompanied by the formation of periclinal phragmoplasts, cell plates and walls that close the open ends of the cells formed. It has also been frequently reported that wall formation begins with free nuclei at the micropylar end of the embryo sac and extends to the chalazal end. Dictyosomes and their vesicles, as well as ER and occasional arrays of microtubules, are seen in conjunction with the freely growing anticlinal walls during endosperm cellularization (Newcomb and Fowke 1973; Singh and Mogensen 1976; Jensen, Schulz and Ashton 1977; Newcomb 1978; Fineran, Wild and Ingerfeld 1982; Gori 1987; van Lammeren 1988a; Webb and Gunning 1991; Yan, Yang and Jensen 1991b; Dute and Peterson 1992; Chitralekha and Bhandari 1993; XuHan and van Lammeren 1993). It is important to note that even when it has been claimed that phragmoplasts form between free nuclei, their resemblance to typical phragmoplasts leaves much to be desired (Ponzi and Pizzolongo 1984). A recent study has shown that the placement of initial walls during cellularization of the barley endosperm syncytium is controlled by microtubules. Although the cytokinetic apparatus lacks a preprophase band of microtubules, a radial system of microtubules emanates from the interphase nuclei and organizes the cytoplasm into discrete units surrounding each nucleus. The walls grow at the boundaries of the nucleo-cytoplasmic domains defined by the microtubules and partition the peripheral cytoplasm into an alveolate system (Brown, Lemmon and Olsen 1994). A somewhat similar configuration of microtubules appears dur-
326
Zygotic Embryogenesis
ing cellularization of the endosperm in Phaseolus vul- the number of nuclear divisions completed after garis (XuHan and van Lammeren 1994). The consis- the formation of the primary endosperm nucleus. tent occurrence of microtubules delimiting the Nuclear activity is reflected in the extent to which nucleo-cytoplasmic units has inspired the notion endosperm haustoria with a limited number of that rather than being a passive support system, the nuclei, as compared to the multicellular tissue, microtubular cytoskeleton undergoes a striking undergo repeated endoreduplication. A striking reorganization during the initial cellularization of example is the uninucleate chalazal endosperm haustorium of Arum maculatum, which exhibits a the endosperm. After the inception of the anticlinal wall from ploidy level of 24,576N, corresponding to 13 the embryo sac wall, the pattern of subsequent wall endoreduplication cycles, whereas cells of the formation involves the building up of an incipient endosperm proper remain triploid (Erbrich 1965). phragmoplast. Indeed, the formation of phragmoEndosperms of various cereals offer considerplasts and cell plates is easier to observe during this able advantage for studying nuclear cytology due phase of endosperm development than during cel- to the prolonged cellular phase when progressive lularization of the free nuclei. Some of the early changes in DNA content occur. The extent of DNA studies using the wall-free liquid endosperm cells changes in the developing endosperm of maize is of of Haemanthus katherinae have provided a dynamic special interest since a number of reports have charpicture of mitosis in the endosperm; a beautiful cin- acterized the nuclei extensively. The increase in emicrographic record of the process, produced by nuclear volume that is often seen during polyBajer (1965), continues to be used even to this day ploidization and endoreduplication has been cited to illustrate mitosis in living plant cells. From the as indicative of an increase in the DNA content of model of cell wall formation that incorporates data the endosperm ranging from 6C to 384C (Punnett from polarized light, phase contrast, differential 1953; Tschermak-Woess and Enzenberg-Kunz 1965; interference contrast, immunofluorescence, and Cavallini et al 1995). Following upon an earlier electron microscopy, there seems little doubt that the observation that the DNA content of the endosperm phragmoplast is initially formed at late anaphase by of young maize grains detected by Feulgen the consolidation of small droplets rich in elements microspectrophotometry increased up to 24C (Swift of the dictyosomes or ER, and that it coexists with a 1950), later studies showed that coincident with the massive accumulation of microtubules. Not unlike decline in mitotic activity, there was a variable the situation described in other plant cells, these increase in the DNA content of the endosperm droplets coalesce to form the cell plate (Hepler and nucleus, followed by a decrease. In studies using Jackson 1968; Lambert and Bajer 1972; de Mey et al flow cytometry of nuclear preparations of the entire 1982). Schmit and Lambert (1987,1990) have shown endosperm, only a modest increase of about 12C in that endosperm cells of H. katherinae have a perma- the average DNA content per nucleus was observed nent system of microfilaments, and it is possible, at the peak period (Kowles and Phillips 1985; and even likely, that actin may participate in the Kowles, Srienc and Phillips 1990; Lur and Setter microtubule-guided transport of vesicles during 1993). A tentative conclusion from these data is that phragmoplast formation. preferential amplification or underreplication of part of the genome during endoreduplication is not a way of life for maize endosperm nuclei.
(iii) DNA Amplification Odd as it may seem, cellularization of the endosperm is not generally associated with a normal cycle of DNA synthesis and mitosis. Two conventional routes followed by the cells of the endosperm that escape a normal cell cycle are endoreduplication and polyploidy. D'Amato (1984) has reviewed the literature on the occurrence of these phenomena in the endosperms of diverse angiosperms; as tabulated therein, the range of endoreduplication and polyploidy displayed by endosperms may be considerable, depending upon the location of cells, the presence of haustoria, and
Studies of developing endosperms of hexaploid triticales, oat (Herz and Brunori 1985), wheat (Herz and Brunori 1985; Chojecki, Bayliss and Gale 1986; Brunori, Forino and Frediani 1989), rice (Ramachandran and Raghavan 1989), and barley (Giese 1992) by microspectrophotometry or autoradiography have also shown that some form of endoreduplication occurs in cell populations and causes a gradual rise in DNA content from 3C to 24C in triticales, oat, and wheat, and from 3C to more than 30C in rice (Figure 12.3). In the context of a general increase in DNA content of endosperm cells with time, there is a marked decrease in the
Developmental Biology of the Endosperm
327
however, that aleurone cells belong to the same league as starchy cells, because only a simple doubling of DNA content, with some nongeometric fluctuations, has been observed in the examples studied (Maherchandani and Naylor 1971; Keown, Taiz and Jones 1977). The mechanism of endoreduplication in the endosperm poses a fundamental problem, since it brings into focus the control systems for induction of the S phase and inactivation of the M phase of the cell cycle. Two key observations from a work on maize endosperm deserve to be mentioned in this connection. One is that DNA synthesis during the early stages of endosperm development is maintained by an increase in the amount and activity of S phase related protein kinases, which are implicated in the onset of the S phase in animal cells. Secondly, endoreduplicated cells contain an inhibitor that suppresses the activity of an M phase promoting factor (Grafi and Larkins 1995). The biological and functional significance of endoreduplication in the endosperm cells is still a matter of debate; according to a popular view, as in the cells of legume cotyledons, an increase in the DNA content of endosperm cells may be important for the synthesis and accumulation of storage products. Unfortunately, amplification of specific DNA sequences related to the function of the developing endosperm has not been demonstrated, because the evidence has turned out to be negative (Giese 1992). 0 6
12 IS 24 30 36 42 48 54 60 66 72 78 84 ONA per nucleus (C units)
Figure 12.3. Distribution pattern of DNA content of nuclei in the developing endosperm of Oryza sativa. DA, days after anthesis. (From C. Ramachandran and V. Raghavan 1989. Changes in nuclear DNA content of endosperm cells during grain development in rice [Oryza sativa]. Ann. Bot. 64:459-468, by permission of Academic Press Ltd., London.)
number of nuclei with lower DNA values and a proportionate increase in nuclei with higher DNA values. A similar relationship exists during endosperm development in some noncereal monocotyledons (Marciniak 1993). The basic changes in DNA values just described were elucidated mostly for the starchy cells or for both starchy and aleurone cells of cereal endosperms. Although assumed to remain unaltered during cytodifferentiation, the nuclear DNA content of developing aleurone cells reportedly varies when the cells are analyzed in the absence of contaminating starchy cells. It seems improbable,
2. NUTRITIONAL ROLE OF THE ENDOSPERM Within the confines of the embryo sac, the free nuclei of the developing endosperm surround the embryo like a loose string of beads and eventually form a compact tissue around it. It is therefore obvious that some kind of interaction must take place between the endosperm and embryo. The best documented claims of interaction invoke a role for the endosperm in providing nurture and nutrients to the developing embryo. These claims, along with the structural modifications of developing embryos for nutrient uptake, will be discussed in Chapter 13. This section considers the outgrowth of haustorial structures, the chemical composition of the endosperm, and the relationship between embryo abortion and endosperm activity in wide crosses all of which constitute additional circumstantial evidence that relates the endosperm with embryo nutrition.
328
Zygotic Embryogenesis
(i) Endosperm Haustoria
A general feature of endosperm organization in certain angiosperms is the production of outgrowths known as haustoria. These have been described in exquisite detail by whole mount preparations of the entire endosperm, which facilitate isolation of the parts intact. The haustoria arise at either the micropylar or the chalazal end or at both ends of the developing endosperm and appear as branched tubular processes that invade adjacent parts of the ovule such as the nucellus, integuments, and chalazal tissues. They are found mostly in the nuclear and cellular types of endosperms and occasionally in the helobial type. One arbitrarily selected representative of each of the three categories will be described here to emphasize the infinite complexity of the endosperm haustoria. It might appear unnecessary to obtain evidence of the actual transfer of metabolites to the haustorium in order to recognize that at least in some species, haustorial action might supplement the nutritional resources of the endosperm. An example of the micropylar haustorium is provided by Impatiens roylei. In the most common pattern noted, the first division of the primary endosperm nucleus separates a micropylar and a chalazal chamber. Subsequently, the micropylar chamber is partitioned into three small cells, the uppermost of which is transformed into a tubular haustorium. The haustorium retains its original form for a variable period of time, then it branches extensively in the micropylar region of the embryo sac, extending even into the funiculus (Dahlgren 1934). A view of the chalazal haustorium comes from the account of endosperm development in Nothapodytes foetida (Icacinaceae). The haustorium has its origin in the chalazal endosperm cell formed by the division of the primary endosperm nucleus. This cell enlarges and sends out narrow tubes that extend toward the chalaza. The continued ramification of the haustorium is accomplished by contact between the haustorium initial and an adjacent cell in the chalazal tissue of the ovule. The outcome of this union is the transfer of the cytoplasm and nucleus of the chalazal cell into the haustorium. At the same time as these events occur, the nucleus of the haustorium initial becomes distorted in shape as it streams into the haustorial branches (Swamy and Ganapathy 1957). These observations show that at some point in its development, the haustorium is enriched with protoplasm from the adjacent chalazal cell and thus its function is deeply entwined with the integration of materials from neighboring cells.
Figure 12.4. Development of the chalazal and micropylar haustoria in the endosperm of Rhinanthus minor. A section of the fully developed seed is shown. CH, chalazal haustorium; E, endosperm; EM, embryo; I, integument; M, micropyle; MH, micropylar haustorium; P, placenta; V, vascular connection; W, winglike part of the seed. Scale bar = 200 urn. (From Nagl 1992.)
The endosperm of Rhinanthus minor (Scrophulariaceae) is described as developing both chalazal and micropylar haustoria. These are formed when the products of the first division of the primary endosperm nucleus become polytenic and penetrate the chalazal and micropylar cells of the ovule (Figure 12.4). From the functional perspective, the most significant observation is the appearance of numerous wall labyrinths in the outer region of the chalazal haustorium where it interdigitates with adjacent cells (Nagl 1992). The convergence of numerous mitochondria alongside the wall projections and the lysis of chalazal cells in contact with the haustorium strongly suggest an active transport of the lysed material through the plasma membrane. Histochemical studies of the haustorium of Linaria bipartita (Scrophulariaceae) have shown that the cytoplasm is enriched for RNA, proteins, histones, and starch (Bhatnagar and Kallarackal 1980). These are hallmarks of a metabolically active cell, but whether they are related to its presumed function in the absorption of metabolites is not known. It is within the range of current cytochemical and cytological techniques to provide additional information to place on a sound footing such an important assumption as the role of the endosperm haustorium in nutrient absorption.
Developmental Biology of the Endosperm (ii) Chemical Composition of the Endosperm
The nutrient function of the endosperm is determined not only by the tissue's morphology and its intimacy to the embryo, but also by its chemical composition. Moreover, several observations point to the view that a knowledge of the chemical composition of the nuclear and cellular types of endosperms is necessary to obtain a general picture of the effect of this tissue on the nutrition of the embryo beginning with the zygote stage. In view of the difficulty of obtaining large quantities of nuclear endosperm, biochemical analysis has been limited to the liquid endosperm of Cocos nudfera (coconut; Arecaceae). Popularly known as coconut milk or coconut water, the liquid endosperm of coconut is the morphological equivalent of a nuclear endosperm and is widely used as a growth adjuvant in plant tissue culture media. As described in the next chapter, the first successful culture of proembryos was accomplished by supplementing the mineral-salt medium with coconut milk. What are the chemical factors present in coconut milk that dramatically induce the growth of embryos and other explanted organs in culture? Results from chemical fractionation have shown that coconut milk is a conglomeration of several inorganic ions, amino acids, organic acids, vitamins, plant hormones, sugars, sugar alcohols, and other substances in concentrations that are used in tissue culture media. Indeed, some of the substances may not be of any real value in the promotion of growth of cultured plant organs and, thus far, no single compound or group of compounds has come to be accepted as being able to duplicate the major growth promoting activity of coconut milk in its entirety (Raghavan 1976b). Limited information is available on the chemical composition of the cellular endosperms of some plants and of their effects on the growth of cultured organs. Because this work has been summarized previously, only a few general comments will be made here (Raghavan 1976b). Maize endosperm at the milky stage contains several amino acids and hormones, including the cytokinin zeatin. Among other cereals, wheat, rye, barley, and rice have endosperms that are known to contain auxins. By far the most potent source of endosperm as a growth promoter of cultured tissues is the liquid content of the vesicular embryo sac of Aesculus woerlitzensis (horse chestnut; Hippocastanaceae), which contains both free nuclei and cellular endosperm. Chlorogenic acid, myo-inositol, and auxin are some of the compounds with potential
329 growth promoting properties identified in the endosperm extract of horse chestnut. Based on the information from chemical analyses in conjunction with tissue culture data, it is believed that the innate in vivo function of endosperm tissues of Datura stramonium, Juglans regia (walnut; Juglandaceae), Allanblackia parviflora (Guttiferae), Cucumis sativus, Prunus persica and related species, Hyphaene natalensis (Arecaceae), and Lupinus luteus
is concerned with the supply of nutrients to the growing embryo. Similarly, the identification of substances such as GA and sugars in the endosperm tissues of Aesculus californicus, Echinocystis macrocarpa (Cucurbitaceae), and apple, among others, reflects the potential utilization of these tissues by the embryo. Monitoring the fate of GA during embryo growth might permit alternative explanations, and so these examples need further study. (iii) Embryo Abortion and Endosperm Development
In horticultural and breeding practices, embryo abortion is commonly encountered in unsuccessful crosses that fail to set seeds. Although fertilization occurs normally in such crosses and embryos begin to grow in a relatively healthy way, lethality is rampant during the subsequent growth of embryos. Hybrid embryos display a wide range of irregularities, some of which are described in Chapter 14; the concern here is to examine evidence suggesting the relationship of embryo abortion to a lack of endosperm activity. Renner (1914) first called attention to the fact that the collapse of embryos in unsuccessful crosses between Oenothera biennis X O. muricata and O. biennis X O. lamarckiana was pre-
ceded by the disintegration of the endosperm beginning soon after fertilization, which thus deprives the embryo of an immediate source of food. As shown by detailed cytological studies of endosperm development in aborted seeds in an extensive range of interspecific hybrids, endosperm failure is more frequent than is generally believed and might take place in discrete ways during the division of the zygote. Comparison of ovule development in reciprocal crosses involving diploid and tetraploid races of Lycopersicon pimpinellifolium and in crosses of these races with L. peruvianum provides a fine example of the diversion of the primary endosperm nucleus into an alternative pathway during its early ontogeny. In terms of cell number, the initial rate of development of the endosperm in inviable crosses does not differ appreciably from that observed in successful
330
crosses, but afterward, the endosperm of the latter outstrips that of the former in its growth. The structural changes seen in the disintegrating endosperm cells are more dramatic than the decrease in cell number. In the hybrid ovule as early as 144 hours after pollination, cells of the endosperm at the chalazal end become vacuolate and less dense than those at the micropylar end. Although the endosperm survives for some time more, it undergoes no further divisions. Instead, as a result of the dissolution of cell walls and the fusion of the protoplasm and of the nuclei with those of contiguous cells, the endosperm is reduced to a few giant cells surrounding the embryo (Cooper and Brink 1945).
Zygotic Embryogenesis
An important factor governing the continued growth of the embryo and endosperm in a fertilized ovule is the nature of their interaction with the neighboring cells; therefore, some attention has been paid to the behavior of the nucellus and integuments in ovules of wide crosses. Several investigators, beginning with Renner (1914), have noted that the development of a weak endosperm and embryo abortion in inviable crosses are associated with the proliferative growth of a tissue from the nucellus. In a comparative histological examination of ovules from crosses between Nicotiana species, seed collapse was found to be associated with a retarded growth of the endosperm, which In Trifolium ambiguum X T. repens cross, division of attained only a fraction of the growth attained in the primary endosperm nucleus ceases at different the selfed parent, and with a pronounced hyperstages of development of the hybrid ovules, but in plastic growth of the nucellus. The term "somatono case does the endosperm proceed beyond the plastic sterility" is used to describe the hyperplasia 128-nucleate stage; in many hybrids, endosperm of the maternal tissue. It has been found that development is stymied at even earlier stages. In embryo failure is also associated with a deficiency contrast, selfed T. ambiguum has a complete in the development of the conducting elements endosperm including a well-developed chalazal within the seed, foreshadowing a suppression of haustorium (Williams and White 1976). The ultra- nutrient transport to the endosperm; hyperplastic structural changes in the hybrid endosperm cells growth of the nucellus has been ascribed to the are also impressive and seem to impinge on the syn- consequent abnormal distribution of nutrients thetic activities of the cells and on nutrient transport (Brink and Cooper 1941). into them. For example, in the endosperm cells of A gene dosage effect leading to a change in the Hibiscus costatus X H. aculeatus and H. costatus X H. ratios of ploidy levels among the embryo, furcellatus hybrids, the rough ER appears as short endosperm, and maternal tissues is seen in the irregular segments lacking ribosomes, whereas the ovules of several infertile crosses that show tumor nuclei contain multiple nucleoli (Ashley 1975a, b). formation from the inner epidermis of the integuIndubitable proof that endosperm abnormalities are ment (endothelium). In reciprocal crosses between related to the transport of nutrients is the demon- 4N L. pimpinellifolium and 2N L. peruvianum, the stration that wall ingrowths characteristic of the endothelium assumes a densely cytoplasmic connormal endosperm are almost completely sup- tour and surrounds the endosperm except for a pressed in the hybrid endosperm generated in small gap at the chalazal end. Tumor formation is crosses between normal diploid and autotetraploid initiated from the dorsal region of the endothelium maize (Charlton et al 1995). In the various examples near the chalazal end. As the tumor assumes form, studied, cells of the unhealthy endosperm have it begins to grow without restraint and extends into been found to show uninformative types of mitotic the embryo sac as a burgeoning mass of cells. The abnormalities such as lagging chromosomes, fused cells of the endosperm surrounding the embryo nuclei, and nuclei of high or low chromosome num- deteriorate progressively to the point that the bers. However, an understanding of the extent to embryo is engulfed by the overgrown endothelial which disturbances in the endosperm in inviable cells (Cooper and Brink 1945). It is no wonder that crosses may cause starvation of the embryo is cer- abortion of the already undernourished embryo is tainly lacking at the present time. It must be con- hastened by these changes. cluded on logical grounds that the hybrid embryos An imbalance in the chromosome number are starved to death due to the inability of the phys- between embryo and endosperm is postulated to iologically disturbed endosperm to manufacture cause abnormalities in the endosperm and consethe exacting nutrients required for embryo growth quent embryo abortion in crosses between diploid or to absorb them from the surrounding maternal and tetraploid varieties of Citrus (Esen and Soost tissues and transmit them to the embryo. This 1973). Considerations of the requirement for a balmeans that the deleterious genes exert their primary ance between maternal and paternal genomes in effect on the endosperm rather than on the embryo. the endosperm for seed set have led to a number of
Developmental Biology of the Endosperm
hypotheses to account for endosperm failure and embryo abortion in wide crosses (Kowles and Phillips 1988). A most useful observation based on data from a wide variety of intraspecific- and interspecific-intraploidy crosses is that the endosperm develops abnormally when the maternal-paternal ratio in the tissue deviates from 2:1. This has been incorporated into the notion of endosperm balance number (EBN) to explain endosperm development in crosses with different levels of ploidy. According to this hypothesis, the genome of each species is assigned an EBN on the basis of its crossing behavior with a standard species and, for crosses to be successful, the EBN has to conform to a 2:1 maternal-paternal ratio (Johnston et al 1980; Johnston and Hanneman 1982; Camadro and Masuelli 1995; Ehlenfeldt and Ortiz 1995). From the practical side, the attractiveness of this hypothesis is that the EBN can be manipulated to break the crossing barriers and to create new hybrids; from the theoretical side, it is unclear how the ratio of the maternal-paternal genomes is detected and how a deviant ratio causes endosperm malfunction. Using a mutant indeterminate gametophyte (ig) to cre-
331
methods have been widely used to obtain transplantable seedlings from aborted seeds of interspecific and intergeneric crosses that are traditionally condemned as being incapable of further growth. In certain cases, continued growth of the hybrid embryo is secured by implanting it on a normal endosperm, which is then cultured, thereby initiating a nurse culture. In a limited number of cases, efforts directed toward rescuing hybrid embryos by culturing ovules and ovaries have proved rewarding. Examples of hybrid embryo rescue by these different tissue culture protocols have been described or tabulated in published reviews (Raghavan 1977b, 1985, 1986b, 1994; Collins and Grosser 1984; Williams 1987) or books (Raghavan 1976b, 1986c), and the reader is referred to these sources for further information. The appearance of haustorial structures on the endosperm, the presence in this tissue of chemical substances with potential growth promoting functions, and the rescue of hybrid embryos from ovules lacking a normal endosperm, present evidence of the potential role of the endosperm in the nutrition of the embryo. These observations, considered in conjunction with the various ultrastructural modifications of the embryo and suspensor cells for the absorption and translocation of metabolites from the surrounding cells (Chapter 13), make a compelling case for the nutrition of the embryo by endosperm metabolites.
ate various ploidy level combinations in crosses in maize, Lin (1984) concluded that development of the endosperm is affected by the parental source of its chromosomes and that the 2:1 maternal-paternal chromosome number is critical in the development of a healthy endosperm. Although this hypothesis can readily explain why 2x X 4x or 4x X 2x crosses with 2:2 and 4:1 maternal-paternal chromosome ratios, respectively, in the endosperms will fail, it is 3. GENE ACTION DURING ENDOSPERM not applicable to crosses in which endosperm ONTOGENY development is not governed by a ratio of chromoIt is natural to think of the development of the some numbers. endosperm in terms of the synthesis of storage The histology of the failure of endosperm devel- compounds and proteins necessary for the fundaopment and its side effects on the maternal tissues mental processes of cellular metabolism and houseof ovules of unsuccessful crosses are obviously in keeping functions. Three major experimental line with the view that gene action deprives hybrid approaches have been pursued to gain an underembryos of a continuing supply of nutrients to standing of how genes control the complex process complete development. The postulated nutritional of endosperm development in cereals. Historically, relationship between the embryo and endosperm spontaneous kernel mutations in maize and barley tends to suggest that the cryptic potentiality of the have provided a tractable genetic system for monihybrid embryo to complete development may be toring the effects of marker genes on the developrealized if it is allowed to grow in an artificial ment and differentiation of the endosperm. A large medium supplied with nutrient substances that are body of work has centered on the correlations of normally identified with the endosperm. Early in RNA and protein synthesis with endosperm develthe development of embryo culture as a research opment, and on the isolation of cDNA clones for tool, Laibach (1925) demonstrated that progenies specific proteins and the analysis of their exprescan be obtained from nonviable seeds of Linum sion in the developing endosperm. In recent years, perenne X L. austriacum hybrid by excision and cul- extensive efforts have gone into the sequencing of ture of embryos before the embryos begin to disin- genes regulating endosperm storage product syntegrate. Since this pioneering work, embryo culture thesis and into the search for conserved nucleotide
332
sequences of developmentally regulated genes and for the expression of promoter sequences of genes in transgenic plants. This section will deal with the genetic control of endosperm development, the regulation of RNA and protein synthesis, and the expression of genes for housekeeping proteins in the endosperms of cereals, in which they have been studied intensively. Due to the sheer volume of information, synthesis and accumulation of storage products in the endosperm will be considered in a separate section.
Zygotic Embryogenesis
mutant kernels have an extremely low level of invertase, impairment of the movement of photosynthates between the pedicel and endosperm, and the consequent osmotic imbalance, offer a plausible explanation of the effects of the mutant gene on the endosperm (Lowe and Nelson 1946; Miller and Chourey 1992; Cheng, Taliercio and Chourey 1996). The basic histological manifestation of mutation in the de-17 locus is the failure of endosperm cells facing the placenta to form a specialized absorbing tissue for transmitting nutrients from the maternal plant to the endosperm (Brink and Cooper 1947a). (i) Endosperm Mutants A large number of dek mutants have been The singular strength of maize as a genetic sys- induced by pollinating maize ears with pollen tem among higher eukaryotes is the occurrence of grains mutagenized with ethyl methanesulfonate morphologically identifiable mutants that affect (EMS). The Ml plants raised from mature grains virtually all stages of its life cycle. In considering were self-pollinated and the harvested ears generthe regulatory intervention of genes in kernel ated the stock of dek mutants for subsequent genetic development in maize, most analyses have been analysis. The endosperms of the mutants varied conperformed to date with mutants that affect both the siderably in their phenotypic appearance, with endosperm and the embryo and with those that are many having a collapsed or shrunken appearance; specific for endosperm development. Those in others were fully formed but were distinguished which growth of both embryo and endosperm is from the wild type by a floury or opaque appearimpaired are known as defective kernel (dek) mutants; ance and a soft or fluid consistency. It was not included in the group that affects only endosperm unusual to find mutant kernels that were lighter in development are shrunken (sh), wx, brittle (bt), sugarycolor than the normal due to blockage of carotenoid (su),floury(fl), and opaque (op) mutants. Since the lat- synthesis. Generally, the mutant phenotype is ter group of mutations impinges on the nature of revealed within two weeks after pollination, at the storage reserves in the endosperm, it is appro- which time the growth of the defective endosperm priately considered in the next section. slows down. Effects of the mutant gene on the embryo are seen even earlier than on the endosperm The first dek mutants described turned out to be (Neuffer and Sheridan 1980; Sheridan and Neuffer those in which the endosperm was normal but the grain did not have a viable embryo (Jones 1920). 1980). Compared to the wild type, mutations impose True dek mutants with defective endosperm were serious restraints on the accumulation of dry matter, described by Mangelsdorf (1923), who found total proteins, and storage proteins in the grains in which ovule development was arrested endosperm. From these analyses, an informal recogalmost immediately after fertilization or after a nition of three broad classes of mutants has small amount of endosperm had been formed. emerged: (a) those showing a reduced accumulation Additional defc-like mutants are designated as of dry matter and proteins during the entire reduced endosperm (re-1, re-2) (Eyster 1931), minia- endosperm ontogeny; (b) those showing an initial ture seed (mn) (Lowe and Nelson 1946), and defec- normal rate of accumulation, followed by an early tive seed (de-17) (Brink and Cooper 1947a). In decline; and (c) those in which an initial lag in the general, mutant grains attain only a fraction of the rate of accumulation is coupled with an early termiweight of normal grains while maintaining a regu- nation of accumulation (Manzocchi, Daminati and lar initial pattern of growth of the endosperm and Gentinetta 1980; Manzocchi et al 1980). In almost all embryo. Closer studies of mn mutant kernels have cases examined, the average DNA content of the shown that subsequent disintegration of the vas- nucleus of mutant endosperm cells was found to be cular elements at the chalazal end of the ovary lower than that of the wild type; two idealized posseemingly interrupts nutrient supply from the sibilities - a low proportion of nuclei in the higher maternal tissues through the pedicel. Provision of DNA classes or the occurrence of fewer rounds of the MN gene in the wild-type grain serves to endoreduplication - can obviously tilt the balance in encode an endosperm-specific cell wall invertase favor of a low DNA value (Kowles et al 1992). necessary for the mobilization of sucrose. Since the An array of mutants equally diverse as those
Developmental Biology of the Endosperm
just described has been obtained by the insertion of mutator (mu) transposons into the maize genome. This is accomplished by crossing standard lines with pollen from mu stock; the Fl plants grown from the resulting caryopses are then self-pollinated. Defective kernels are identified on the ears and are used for subsequent genetic dissection. Among the range of phenotypes observed, the most common are those containing a reduced endosperm and a small embryo, and others that have an empty pericarp and are almost always embryoless. Several mutants identified by transposon tagging have been found to be allelic to the dek mutants obtained by EMS mutagenesis (Scanlon et al 1994). These observations lead to the inference that following fusion of the sperm cell with the diploid polar fusion nucleus, many different genes must function in an orderly manner to effect the development of the endosperm into a functional nutritive tissue. The potential to gain deep insight into the genetic program governing endosperm development in maize by molecular cloning has been enhanced by the isolation of mutants from mu stocks.
333
(ii) Synthesis of Housekeeping Proteins
As a rapidly growing tissue, the endosperm must synthesize an enormous range of proteins necessary for cellular metabolism. However, in contrast to the extensive literature on storage protein synthesis, little information is available on the synthesis of metabolic proteins during endosperm development. The main reason for this is that synthetic activities of the widely investigated endosperm tissues are inextricably linked to the accumulation of storage products. Most importantly, the analysis of housekeeping proteins includes changes in the activities of enzymes for the metabolism of starch, proteins, and lipids, and in the levels of amino acids and nucleotides. Of necessity, much of the evidence from such investigations that relates to the synthesis of housekeeping proteins is correlative, and hence its significance is not fully understood. A few studies that have analyzed gene expression patterns in the endosperm dissociated from the synthesis of storage proteins will be described later.
Further evidence for the action of a multiplicity Following fertilization, a readily available supply of genes during endosperm development has of substrates for the division of the endosperm come from analyses of barley mutants with a nucleus must be present. This may be provided by defective endosperm. Deserving recognition for transcription of the genome of the maternal tissues of their importance are spontaneous endosperm the ovule or of the endosperm nucleus. Evidence mutants designated as shrunken endosperm (se and based on autoradiography of 3H-uridine incorporaseg), in which shrunken kernels are caused by the tion shows that active transcription is initiated in the expression of the mutant genes in the maternal tis- endosperm of barley as early as the free nuclear stage sues of the plant. Mutant seg endosperm types are (Bosnes and Olsen 1992). The question of how soon distinguished from the wild type by a premature after fertilization the genome of the endosperm termination of grain filling caused by necrosis of nuclei begins transcription has been investigated in the chalaza, or by upsets in the endosperm growth maize by crossing two strains with different elecpattern that lead to characteristic abnormalities trophoretic isozymes for a particular locus and by but permit normal development of the chalazal tis- screening the endosperm at different stages of develsues (Jarvi and Eslick 1975; Felker, Peterson and opment for expression of the variant enzymes. It was Nelson 1984, 1985, 1987). Another group of found that several enzymes of the maternal allele mutants with a defective endosperm includes the become active two or three days earlier than the gene chemically induced shrunken endosperm expressing products of the paternal allele. The earlier expression xenia (sex) mutants, in which the mutant genes of enzymes from the maternal side has been attribexert their effect on tissues of the entire grain. As uted to their origin in the nucellus. However, other in the case of maize, the ramifications of sex muta- experiments showed that transcription of some tion on the endosperm of barley are extensive, isozyme genes in the endosperm nucleus begins relranging from the total lack of aleurone cells to the atively soon after fertilization but translation of the complete loss of starchy endosperm traits, and message is delayed until the endosperm becomes celeven to the presence of new cell types not found in lular (Chandlee and Scandalios 1983). the wild-type endosperm; the embryo may remain A study of catalase activity profiles has shown viable or nonviable. These effects are so profound that continued development of the maize as to suggest that the mutations cover genes par- endosperm necessitates fundamental changes in ticipating in all phases of development of the the control of gene expression for the three endosperm (Bosnes et al 1987; Bosnes, Weideman unlinked genes of the enzyme (CAT-1, CAT-2, and and Olsen 1992). CAT-3) in the different parts of the kernel. CAT-1 is
334
present throughout the 30-day developmental period analyzed, whereas CAT-2 is expressed toward the period of kernel maturity and CAT-3 appears only during the initial period of development. The appearance of isozymes correlates well with the expression of transcripts for the respective genes in RN A blots, indicating that the synthesis of the enzyme is controlled at the transcriptional level. The presence of CAT-1 mRNA in the pericarp as well as in the starchy and aleurone cells of the endosperm, presence of CAT-2 mRNA only in the aleurone layer, and presence of CAT-3 mRNA only in the pericarp are evidence of strong tissue-specific regulation of the synthesis of these isozymes (Tsaftaris and Scandalios 1986; Wadsworth and Scandalios 1989). The reason for the unique, developmentally controlled pattern of expression of catalase genes is not at present understood, and investigations on the endosperms of other cereals will be valuable for comparative analysis. Like catalase, pyruvate orthophosphate dikinase, a key enzyme in photosynthetic carbon dioxide fixation in C4 plants, illustrates a tissue-specific and developmentally regulated pattern of gene expression during endosperm development in wheat. The changes in the enzyme titer in the endosperm, with a peak at mid-maturation followed by a decrease, are correlated with changes in mRNA activity. At the tissue level, activity of the enzyme confined to the cells of the aleurone layer is paralleled by the appearance of mRNA in these cells (Aoyagi, Bassham and Greene 1984; Aoyagi and Chua 1988). Genes for glyceraldehyde-3-phosphate dehydrogenase and alcohol dehydrogenase (enzymes of the glucose-phosphate metabolism) show an increased expression correlated with progressive endosperm development in maize up to about 15 days after pollination, followed by a decrease (Russell and Sachs 1991). Trans-acting factors required for developmental expression of the maize ADH-1 gene appear to be present in the endosperm of rice (Kyozuka et al 1994). As will become evident in the following section, the pattern of accumulation of storage proteins and their mRNAs in developing endosperms is not much different from that displayed by the photosynthetic and cytosolic enzymes. 4. SYNTHESIS A N D ACCUMULATION OF ENDOSPERM STORAGE RESERVES
Because of the role of the endosperm as a major dietary source, the importance of the analysis of gene expression concerned with the synthesis of storage reserves in this tissue cannot be overemphasized. An
Zygotic Embryogenesis
understanding of the mechanism by which genes for the synthesis of the three principal storage products, namely, starch, proteins, and lipids, are regulated in the endosperm would be of great importance to the genetic engineering of cereal grains with improved nutritional qualities. A good deal of work, aided mostly by mutants in which the coordination and differential expression of endosperm-specific genes are disturbed, has been done to characterize the synthesis and deposition of storage products during endosperm development in our major cereal crops. This section will review the basic data on the nature of the endosperm storage products, mutants that have led to the identification of key steps in the biosynthetic pathway of storage products in the endosperm tissues of various plants, and the cell biology and biochemistry of the accumulation of proteins in the endosperm. A nutritional role for the storage proteins is supported by their presence in large quantities in the seed and by their rapid degradation during the early stages of germination as the embryo is transformed into a seedling. In seeds of some members of Fabaceae and Arecaceae, extensive deposition of polysaccharides like galactomannans and xyloglucans occurs on the inner cell walls of the endosperm and virtually occludes the cell contents. In addition, the cells contain small amounts of storage lipids and proteins (Meier and Grant Reid 1977; DeMason, Sexton and Grant Reid 1983; DeMason 1986; DeMason, Chandra Sekhar and Harris 1989; Edwards et al 1992). Although it is believed that these cell wall polymers serve as food materials, they are not further considered here.
(i) Mutants Defective in Starch Synthesis
This account will be prefaced by a brief reference to conventional biochemical studies that have identified the enzymes that link the synthesis of starch to its main substrate, sucrose. An important initial reaction in the sucrose-starch conversion is catalyzed by sucrose synthase, which splits sucrose to fructose and UDP-glucose. Another set of reactions involving hexokinase and hexose phosphate isomerase results in the transformation of fructose to glucose-6-phosphate. Formation of glucoses-phosphate is also provoked from UDP-glucose by the action of pyrophosphorylase followed by a reaction mediated by a mutase. These changes, which occur in the cytoplasm, are followed by the conversion of glucose-6-phosphate to starch inside the amyloplast. This step is also catalyzed by a pyrophosphorylase and involves the formation of ADP-glucose, which donates its glucose to an increasing chain of
Developmental Biology of the Endosperm
glucose molecules. Starch exists as straight-chain amylose and as the branched-chain amylopectin, the latter being formed by the action of a branching enzyme. Although different biosynthetic pathways that lead to the same end products occur in maize and in some other plants, the basic steps described here could be rightfully considered as the most prevalent. Several genes affect the type and quantity of carbohydrates in the mature endosperm of maize and, consequently, genetic approaches have been used extensively to determine the biochemical lesions caused by mutations. The sh-1 mutation leads to a shrunken or collapsed endosperm and is attributed to the presence of a reduced amount of starch. At the protein level, the mutant endosperm is highly deficient in sucrose synthase; this in turn results in the accumulation of elevated levels of sucrose in the endosperm (Creech 1965; Chourey and Schwartz 1971; Chourey and Nelson 1976; Cobb and Hannah 1983). A special aspect of all sh-1 mutants showing a substantial loss of sucrose synthase is that they retain a residual level of enzyme activity. The arguments about the normal enzyme and the residual enzyme have hinged around the question of whether the two enzymes are coded by the same or different genes; it is now resolved that the residual activity is due to a minor enzyme, similar in catalytic properties to the bulk enzyme, but encoded by a second gene, SUCROSE SYNTHASE (SUS) (Chourey 1981). A model of different in vivo functions for the two isozymes has gained currency with the demonstration that the isozymes are located in different parts of the endosperm. The enzyme encoded by SH-1 generally accumulates in the bulk of starch-containing cells, whereas the SLZS-encoded enzyme is restricted to the aleurone layer and the basal endosperm transfer cells (Chen and Chourey 1989; Heinlein and Starlinger 1989). Finally, transcripts of SH-1 and SUS genes have been localized in the starchy endosperm and the aleurone layer, respectively (Springer et al 1986; Rowland and Chourey 1990). This finding indicates that the rate of transcription or message stability may be a controlling factor in the accumulation of the two isozymes. At the molecular level, both the SH-1 and SUS genes have been cloned and characterized (Werr et al 1985; McCarty, Shaw and Hannah 1986; Gupta et al 1988). Another mutant resulting from upsets in the carbohydrate metabolism of the endosperm is vox, which forms exclusively amylopectin, instead of a mixture of amylose and amylopectin as in the normal endosperm. The enzymatic basis for this differ-
335
ence has been attributed to the absence in the mutant endosperm of starch-granule-bound nucleoside diphosphate sugar transferase enzymes, which transfer glucose from ADP-glucose or UDP-glucose to starch (Nelson and Rines 1962; Nelson and Tsai 1964). As in the case of the sh mutant, the wx endosperm maintains some residual enzyme activity possibly coded by a different gene (Nelson, Chourey and Chang 1978). Cloning of the WX locus and characterization of its gene product have shown that some mutant phenotypes result from the production of a structurally altered protein; several other stable mutant phenotypes are caused by insertions or deletions within the WX gene transcription unit (Echt and Schwartz 1981; Shure, Wessler and Federoff 1983; Wessler and Varagona 1985). Greatly reduced starch synthetic capability is associated with sh-2 and bt-2 mutants. The mutant block has been identified with a reduced activity of the enzyme ADP-glucose pyrophosphorylase in the endosperm; this enzyme catalyzes the formation of ADP-glucose from glucose-1-phosphate and ATP (Tsai and Nelson 1966). However, the presence of detectable levels of enzyme activity in several alleles of both mutants suggested that they may possess a pyrophosphorylase independent of SH-2 and BT-2 gene functions; this proved to be the case when it was found that 95% of the activity in the wild type exists in the heat-labile form and 5% in the heat-stable form. Both mutations reduce the heat-labile form appreciably and the heat-stable activity only marginally (Hannah and Nelson 1976; Hannah, Tuschall and Mans 1980). Although these observations point to the genes' preference for functioning in a nonstrucrural role in some posttranslational steps, two later studies have strengthened the argument that SH-2 is a structural gene. In one strategy, using the transposable element dissociation {ds), which has the ability to inhibit the function of SH-2, it was found that upon removal of the ds element, gene function is restored in the endosperm. Moreover, revertants that restore wildtype enzyme levels possess enzyme molecules with altered allosteric and kinetic parameters (Tuschall and Hannah 1982). Confirming that SH-2 is indeed a structural gene for the maize endosperm ADPglucose pyrophosphorylase, the deduced amino acid sequence of a SH-2 cDNA clone has revealed much similarity to that of the bacterial enzyme (Bhave et al 1990). Using rice ADP-glucose pyrophosphorylase cDNA as a probe, a BT-2 cDNA clone was isolated from maize endosperm and partially sequenced. When RNA isolated from various bt-2 mutants was probed with the maize cDNA
336
clone, some mutants had wild-type levels of transcripts and others had transcripts of altered size, suggesting that BT-2 is a structural gene for the transcript (Bae, Giroux and Hannah 1990). Not much is known about the bt-1 mutant, which also contains extremely reduced amounts of starch in the endosperm. The deduced amino acid sequence of a cDNA clone of the BT-1 gene and the copurification of the protein with plastid membranes have indicated a possible role for the gene product in amyloplast membrane transport, especially in the uptake of adenylate molecules (Sullivan et al 1991; Sullivan and Kaneko 1995). It thus appears that significant differences exist in the molecular mechanism of the inhibition of starch deposition in the mutant endosperms, despite their phenotypic similarities. The su mutant is interesting because the principal carbohydrate storage product in the endosperm is the high-molecular-mass compound phytoglycogen. Since accumulation of phytoglycogen occurs at the expense of amylose and amylopectin, it is not surprising that the su endosperm is deficient in starch-debranching enzyme. Chromatography on hydroxyapatite column showed that su endosperm extract lacks one of the three peaks of enzyme activity associated with the nonsugary endosperm and has much reduced activity in the other two peaks. In the multistep pathway of sucrose-to-starch conversion, it is not clear how a deficiency in the debranching enzyme would increase sucrose levels. A related mutant is ae, whose wild-type gene encodes a major starch-branching enzyme. Effects of the mutant gene are manifest in a reduction in amylopectin, a small increase in amylose, and a minor reduction in starch content (Pan and Nelson 1984). Consistent with their endosperm-specific nature, the mutations that abolish the activity of enzymes in the endosperm do not interfere with enzyme synthesis in the embryo. This is true for sucrose synthase at the SH-1 locus (Chourey and Nelson 1976), starch-granule-bound transferase at the WX locus (Akatsuka and Nelson 1966), ADPglucose pyrophosphorylase at the SH-2 locus (Preiss, Lammel and Sabraw 1971), and debranching enzyme at the SU locus (Pan and Nelson 1984). Some of the genes involved in starch metabolism of the endosperm of maize that have been cloned are characterized by a highly variable number of introns. The number of introns ranges from 2 in the BT-1 gene (Sullivan et al 1991), to 4 in the SU gene (James, Robertson and Myers 1995), 13 in the WX gene (Klosgen et al 1986) and 15 in the SH gene (Werr et al 1985).
Zygotic Embryogenesis
Mutants regulating WX gene expression in the endosperm have been well documented in rice (Sano 1984). As in the case of maize, as many as 13 introns have been reported in the WX gene of rice (Z.-Y. Wang et al 1990; Hirano and Sano 1991). In the grains of normal and transgenic rice, the WX gene is expressed in a tissue-specific manner exclusively in the endosperm. The expression of this gene is also developmentally regulated, transcripts being present at maximum levels during the early to middle stages of endosperm development (Hirano and Sano 1991; Hirano et al 1995; Z.-Y. Wang et al 1995). Spontaneous mutants affecting carbohydrate metabolism of the endosperm are rare in other cereals, although various kinds of endosperm mutants similar to the su and sh have been obtained by chemical mutagenesis in rice (Satoh and Omura 1981; Yano et al 1984, 1985; Kumamaru et al 1988). (ii) Storage Protein Synthesis
The section on this topic will begin with reference to some terminology currently used to describe seed storage proteins. On the basis of their solubilities, seed storage proteins are classified into albumins, globulins, prolamins, and glutelins. The albumins are soluble in water and in dilute aqueous buffer at a neutral pH, whereas the globulins are characterized by their solubility in salt solutions and insolubility in water. Prolamins are soluble in alcohol, and glutelins belong to a class of proteins that are soluble in dilute acid or alkali solutions. What is known about the distribution of the different kinds of endosperm storage proteins indicates that some of the major cereal grains can be identified by the type of reserve proteins of their endosperms. The major storage proteins in the endosperms of maize, barley, and sorghum are of the prolamin type. Glutelins account for nearly half of the storage proteins of the wheat endosperm, whereas about 40% is constituted of prolamins. Glutelins were thought to be present in the largest amount in rice, accounting for at least 80% of the total protein of the starchy endosperm. However, it now appears that the relative contribution of glutelins to the total proteins of the rice endosperm has been overestimated and that prolamins and globulins also contribute significantly to the total protein content (Krishnan and White 1995). Although prolamins of maize continue to receive a great deal of attention, it is to be noted that glutelins, which constitute about 35%-45% of the storage proteins of the endosperm, have hardly
Developmental Biology of the Endosperm
337
been studied. Globulins, which are the predominant the rough ER in the synthesis of wheat prolamins, storage proteins in the embryos of angiosperms and but the exact mode of transport of the protein from gymnosperms, are the major reserve proteins in the the ER to the protein body in the vacuole has been endosperm of oat. Albumin is also of wide-spread interpreted in different ways (Campbell et al 1981; occurrence in embryos; there are only a few reports Miflin, Burgess and Shewry 1981; Bechtel, Gaines of their occurrence in endosperms. Endosperm stor- and Pomeranz 1982; Parker 1982). One telling age proteins are often designated by a common piece of evidence was the identification of the name derived mostly from the generic name of the wheat homologue of an ER-resident protein, plant, such as zein (prolamin from Zea mays), immunoglobulin heavy-chain-binding protein hordein and hordenin (prolamin and glutelin, (BiP), in the ER and protein bodies of the developrespectively, from Hordeum vulgare), avenin (pro- ing endosperm of T. aestivum; the label was scarcely lamin from Avena sativa), gliadin, glutenin, and trit- present in the Golgi apparatus. The most generalicin (prolamin, glutelin, and globulin, respectively, ized views of these observations are that storage from Triticum aestivum and other species of culti- protein molecules aggregate within the ER along vated wheat), secalin (prolamin from Secale cereale), with the resident protein to form the protein body kafirins (prolamin from Sorghum bicolor), coixin (pro- and that intact protein bodies are transported to the lamin from Coix lacryma-jobi), and oryzin and vacuoles by a novel route that does not utilize the oryzenin (prolamin and glutelin, respectively, from Golgi apparatus (Levanony et al 1992; Rubin, Oryza sativa). Levanony and Galili 1992). An apparent absence of the Golgi apparatus during the active phase of proOrigin of Protein Bodies. The usual concomi- tein accumulation in the cells of the endosperm of tants of endosperm development in cereal grains T. aestivum suggested that this organelle is not are the synthesis of storage proteins and their involved in the transport of the protein product to deposition in the vacuoles as protein bodies. A the vacuole (Briarty, Hughes and Evers 1979). The well-defined aspect of storage protein accumula- picture that began to emerge from later investigation is the appearance of protein bodies in the vac- tions is that Golgi bodies are present abundantly in uoles shortly after the endosperm becomes cellular the developing endosperm adjacent to the proteins and an increase in the size and the number of pro- concentrating within the ER. At this critical period, tein bodies thereafter. Synthesis of storage proteins, an enlargement of the cloud of Golgi vesicles, contheir transformation into protein bodies in a taining material resembling storage proteins, and translocation-competent conformation, and, finally, connections between sheets of rough ER and eletheir transport to the site of storage occur in com- ments of the Golgi vesicles are also frequently plicated ways, and even in the best studied systems observed in suitable preparations (Bechtel, Gaines and Pomeranz 1982; Parker 1982; Parker and some uncertainty remains about the processes. Hawes 1982; Bechtel and Barnett 1986). The weight Some of the most influential generalizations on of evidence is thus suggestive, but not complete the mechanism of protein body formation and its enough to say, that the wheat storage proteins are transport have come from investigations on the transported to protein bodies by a mechanism endosperms of different species of Triticum. The involving the Golgi bodies. It is also conceivable usual parameter used to classify gliadins and that as a first step in protein body formation, the glutenins of wheat endosperm is electrophoretic proteins might begin to condense within the Golgi mobility. The subunits of gliadin thus identified are vesicles, for subsequent transport to the vacuole. a-, (3-, y-, andft>-gliadins,whereas the glutenins are Support for this view has come from immunocytoconstituted of a low- and a high-molecular-mass chemical observations showing that in developing class. Early ultrastructural studies of T. vulgare and endosperm cells of T. aestivum, the ER lumen is T. durum showed the presence of electron-dense bypassed in favor of Golgi vesicles for the packagprotein deposits surrounded by tubular mem- ing and transport of gliadin (Kim et al 1988). branes probably originating from the ER in the Another immunocytochemical study, though developing endosperm. Despite the low resolution implicating the Golgi complex as the site of proof the electron micrographs, these observations led cessing and transport of the proteins, does not, to the suggestion that storage proteins are synthe- however, rule out the lumen of the ER as the site of sized on the rough ER and are deposited in protein their synthesis (Stenram, Heneen and Skerritt bodies by a process of internal secretion (Graham et 1991). From these differences in interpretation, one al 1962; Morton, Palk and Raison 1964). Subsequent is left with a less-than-satisfactory view that the investigators have persistently argued for a role for
338
synthesis and transport of storage proteins in the wheat endosperm revolve around the ER and Golgi vesicles, at least transiently, in a complex way. It is unlikely that electron microscopy and immunolocalization can truly reflect the dynamics of the synthesis and transport of storage proteins in the endosperm, especially when the proteins are constituted of different subunits. Some new information about the role of the Golgi apparatus in protein transport has come from the ability of Xenopus laevis oocytes to sponsor the formation of -a and ysubunits of gliadin when injected with the respective mRNAs. Labeling with radioactive precursor combined with electrophoresis was used to follow the accumulation of proteins in the injected oocytes and in the outside medium. The striking observation made in this experiment was that whereas agliadin accumulates inside the oocyte, y-gliadin is secreted into the medium. Since oocytes of X. laevis do not have vacuoles, y-gliadin is believed to be exported from the ER to the Golgi apparatus and secreted into the medium. This view was also supported by the fact that disruption of the Golgi apparatus by monensin halts the secretion of ygliadin from the oocyte (Simon et al 1990). Evidently, some activation event determined by the structural differences between a- and y-gliadins must allow the gliadins to be distinguished for their assignment to markedly different subcellular transport routes. One final point in this context is that the expression of deletion mutants of y-gliadin in the oocytes showed that the C-terminal region is responsible for secretion of the protein (Altschuler et al 1993). Immunocytochemical evidence shows that during the transport of y-secalins, which are synthesized in the ER, to the site of their accumulation, protein body formation in rye endosperm cells is analogous to that in wheat and involves the Golgi apparatus (Krishnan, White and Pueppke 1990). The biochemical similarity of endosperm prolamins of the wild grass Haynaldia villosa (Poaceae) to wheat gliadins has made it of interest to localize the protein using appropriate antibodies. Absence of label in the inclusions of the ER and the specific labeling of protein bodies found in the vacuoles and vesicles associated with the Golgi complexes provide strong support for a role for the Golgi apparatus in the transport of storage proteins, if not in their synthesis (Krishnan, White and Pueppke 1988). The major mobilization of zein protein bodies in the maize endosperm commences when the cells
Zygotic Embryogenesis
are still undergoing division. The first change in the endosperm cell auguring its altered function is the appearance of small, membrane-enclosed, spherical deposits. The deposits begin to accumulate into dense protein bodies in vesicles produced by localized dilations of the ER that occur along the cisternae or at their enlarged ends. It is difficult to decide from the electron micrographs alone whether the protein is synthesized outside or inside the membranes (Khoo and Wolf 1970). Additional understanding of the role of ER in zein synthesis was provided by studies showing that membrane-bound polysomes and the mRNA population present in the rough ER of the endosperm direct the synthesis of virtually the same zein protein in vitro as they do in vivo (Larkins and Dalby 1975; Larkins, Bracker and Tsai 1976; Larkins, Jones and Tsai 1976; Larkins and Hurkman 1978; Melcher, 1979; Torrent et al 1986). Consistent with the view that fully formed protein bodies are surrounded by ER, a cell-free system that synthesizes zein using polysomes and mRNA derived from protein bodies as templates has also been described (Burr and Burr 1976; Burr et al 1978; Viotti et al 1978). A finding with implications for the role of ER in zein synthesis is that injection of zein mRNA into X. laevis oocytes potentiates the synthesis of proteins compartmentalized within ER vesicles (Larkins et al 1979; Hurkman et al 1981). Any disadvantage in the use of endosperm from intact grains of maize for biochemical studies has been partially ameliorated by the successful culture of the endosperm in liquid suspension or on agar medium. The cultured cells maintain their endosperm phenotype and synthesize zeins that accumulate in protein bodies as in the endosperm cells of developing kernels (Shimamoto, Ackermann and Dierks-Vending 1983; Felker 1987). Although the classification and nomenclature of zeins have remained a contentious issue for a long time, a consensus seems to have emerged in recent years. Accordingly, these proteins are classified into five individual zeins of molecular masses of about 10,15,19, 22, and 28 kDa. These are also known as a-zeins (19 and 22 kDa), fi-zein (15 kDa), y-zein (28 kDa), and 5-zein (10 kDa). Immunolocalization has shown that the different classes of zeins are distributed heterogeneously in the protein bodies, with further variation dependent upon the location of the organelles within the endosperm cells. Generally, the protein bodies of the subaleurone cells are small and contain almost exclusively |3and y-zeins, whereas those confined to the more central cells of the endosperm contain a-zeins also.
Developmental Biology of the Endosperm
An interesting feature of the protein bodies containing, a-, (3-, and y-zeins is that the oc-zein is confined to the central region of the organelles and the p"- and y-zeins become peripheral (Lending et al 1988; Lending and Larkins 1989). Cells of the rice endosperm are unusual in harboring different types of protein bodies that differ in their morphology, the nature of accumulated proteins, and their subcellular origin. A model that has gained some acceptance recognizes two types of predominantly prolamin-containing spherical protein bodies and one type of glutelinenriched, irregularly shaped crystalline body. Electron microscopic observations have supported the view that the three types of protein bodies share a common site of synthesis in the rough ER (Harris and Juliano 1977; Bechtel and Juliano 1980; Oparka and Harris 1982; Yamagata and Tanaka 1986). It has been claimed that different domains of the rough ER, designated as the protein body ER and cisternal type ER, are the site of the synthesis of prolamins and glutelins, respectively. The bases for this view are the clear identification of prolamin-derived protein bodies delimited by the rough ER and of glutelin-derived bodies delimited by single-lamellar membranes, and the differences in the content of prolamin and glutelin mRNAs in the two domains of the rough ER. Whereas the cisternal type of rough ER is enriched for glutelin mRNA, the protein body ER harbors mostly prolamin mRNA (Li, Franceschi and Okita 1993). Evidence has also appeared indicating that a special protein located in the ER lumen might bind to the nascent prolamins, promoting their folding and assembly into protein bodies and thus retaining them in the ER (X. Li et al 1993). Some studies have implicated the Golgi vesicles in a formative way in the final configuration of the prolamins and glutelins (Bechtel and Juliano 1980; Oparka and Harris 1982; Yamagata and Tanaka 1986). The definitive view of the role of the Golgi vesicles now appears to be that they mediate the transport of only the glutelin protein bodies synthesized in the cisternal type ER to the site of their deposition (Krishnan, Franceschi and Okita 1986; Li, Franceschi and Okita 1993). There is little evidence that can be marked out as suggestive of a mechanism that preferentially segregates glutelin mRNA and the protein product to the cisternal type of ER and the Golgi vesicles, respectively. Much of the interest in protein body formation in endosperms of other plants hinges on the involvement of ER in the synthesis of the proteins
339
and of Golgi bodies in their transport. In the developing endosperm of oat, globulins and avenin are deposited in spatially distinct subcellular domains. The main aggregation of globulin takes place within the vacuole, whereas most of the avenin appears to assemble directly within the rough ER, whence it migrates to the vacuole (Saigo, Peterson and Holy 1983; Lending et al 1989). Additional support for the synthesis of storage proteins within the rough ER comes from observations on endosperm development in barley (Cameron-Mills and von Wettstein 1980) and Sorghum bicolor (Taylor, Schiissler and Liebenberg 1985; Krishnan, White and Pueppke 1989). In barley, the proteins, following their synthesis in the ER, are transported to the vacuoles, where they form protein bodies; in sorghum, the proteins aggregate within the site of their synthesis. There is no evidence for the involvement of Golgi vesicles in the synthesis or transport of proteins in oat, barley, and sorghum endosperms. Other abundant proteins such as thionins, which are synthesized in the endosperm of wheat and barley, are detected around storage protein bodies as electron-dense spheroids; a role for ER in their synthesis has not been demonstrated (Carmona et al 1993). Although the endosperm of Ricinus communis (castor bean; Euphorbiaceae) stores abundant oil, there seems to be no constraints on the accumulation of proteins in the same cells. The protein body of the castor bean endosperm has a relatively complex structure, containing phytin globoids and protein crystalloids embedded in an amorphous proteinaceous matrix surrounded by a single membrane. The origin of protein bodies has been traced to small vacuoles that become filled with the protein (Tully and Beevers 1976; Gifford, Greenwood and Bewley 1982). Vacuolelike inclusions containing a particulate phase are also believed to be the site of oil body synthesis in the castor bean endosperm (Harwood et al 1971). Protein Bodies of the Aleurone Cells. Unlike the cells of the starchy endosperm, the aleurone layer is made up of living cells. Although the aleurone layer is not a storage tissue like the starchy endosperm, the cells are filled with aleurone grains - spherosomes and protein granules of a chemical composition different from that of the storage proteins already considered. In wheat (Morrison, Kuo and O'Brien 1975), rice (Bechtel and Pomeranz 1977), barley (Jacobsen, Knox and Pyliotis 1971; CameronMills and von Wettstein 1980; Olsen, Potter and Kalla 1992), and oat (Bechtel and Pomeranz 1981;
Zygotic Embryogenesis
340
15 DAYS
20 DAYS
25 DAYS
35 DAYS
Figure 12.5. Diagrammatic representation of the formation of protein bodies in aleurone cells of the endosperm of Zea mays. Protein precipitates are first detected (15 days) in the vacuole, where they continue to accumulate (20 days). Later the proteins condense, forming a large, electron-
transparent core (25 days). Spherosomes are apposed to the surface of the fully developed protein body (35 days). (From Kyle and Styles 1977. Planta 137:185-193. © Springer-Verlag, Berlin.)
Peterson, Saigo and Holy 1985), the aleurone grains are packaged as discrete particles of phytin, protein, and carbohydrate enclosed within a unit membrane. In electron micrographs, the phytincontaining area of the aleurone grain is recognized as one or more electron-dense globoids; another electron-dense area is constituted of a protein-carbohydrate complex that exists as a matrix or as inclusions. Lipid droplets or spherosomes surrounding the enclosing membrane like a ring complete the profile of the aleurone grain (Jones 1969; Jacobsen, Knox and Pyliotis 1971; Morrison, Kuo and O'Brien 1975). Buttrose (1963b) showed that aleurone grains of wheat originate in vacuoles as deposits of moderate electron density, with protein accumulation progressing with the maturity of the caryopsis. The assembly of aleurone grain in the oat endosperm differs significantly from that described in wheat. Although the vacuole remains as the site of its assembly, the aleurone is initiated by a sequential and progressive deposition of phytin and protein into the vacuole, with phytin deposition preceding that of protein. Compelling evidence in support of the view that the two types of deposits are targeted to different compartments is the early subdivision of the vacuole by internal membranes (Peterson, Saigo and Holy 1985). A similar vacuolar origin of the aleurone grain in the form of dense deposits (Figure 12.5) is the rule in endosperms of barley (Cameron-Mills and von Wettstein 1980) and maize (Kyle and Styles 1977).
the work done with barley has shown that cc-amylase concerned with the digestion of starch in the endosperm is synthesized in the aleurone cells in response to GA coming from the embryo during the early hours of imbibition of the grain. Although the level of regulation of oc-amylase by GA is attributable to changes in gene expression, it is not known what keeps the aleurone cells from expressing the genes in the endosperm of developing and mature grains. Synthesis of carotenes and anthocyanins in the aleurone cells of maize endosperm has come under intense investigation from a developmental perspective, and several genes encoding enzymes for the production of these pigments have been cloned. The anthocyanin biosynthetic pathway in the aleurone cells has served as a model for studies in the genetic interactions between regulatory and structural genes. A discovery of great interest for investigations into the genetic and tissue-specific regulation of expression of genes involved in anthocyanin production is that such genes, introduced by microprojectile bombardment into aleurone cells of a mutant endosperm unable to synthesize the pigment, can faithfully program synthesis of the correct pigment exclusively in the target cells (Klein, Roth and Fromm 1989).
The aleurone cells have provoked a great deal of research in connection with the mobilization of starch during germination of cereal grains. Much of
Metabolism of Storage Proteins. The changes in the pattern of accumulation of total storage proteins and their fractions during endosperm development have been followed in wheat, maize, barley, rice, rye, sorghum, and castor bean. In wheat, the production of most storage proteins
Developmental Biology of the Endosperm
occurs throughout the period of grain maturation; toward the latter phase of maturation, however, accumulation of low-molecular-mass proteins overtakes that of high-molecular-mass proteins (Flint, Ayers and Ries 1975; Pavlov et al 1975; Mecham, Fullington and Greene 1981; Singh and Shepherd 1987). Grain filling in maize is due to an increasing accumulation of globulins and zein from two weeks after pollination up to the mature stage, whereas albumins show a two-phase pattern, reaching a peak midway through endosperm development and declining further with time (Jones, Larkins and Tsai 1977; Dierks-Ventling 1981). In all essential respects, the pattern of accumulation of glutelin in the developing maize endosperm parallels that for zein (Ludevid et al 1984). In terms of their increasing molecular mass and other properties, the prolamins of barley endosperm are designated as B-, C-, and Dhordeins; B-hordeins are sulfur rich; C-hordeins are poor in sulfur; and high-molecular-mass proteins are D-hordeins. Based on genetic studies and protein characterization, three subfamilies, B1-, B3-, and y-hordeins, have been recognized in the Bhordein group. During development, the relative amount of C-hordein is greatest in the young endosperm; it then decreases, remaining constant through grain maturity. In the B-hordein group, Blhordein is found to increase throughout endosperm development whereas B3- and yhordeins behave like C-hordein. D-hordein appears to accumulate toward the final phase of endosperm development (Shewry et al 1979; Rahman, Shewry and Miflin 1982). Another protein of potential nutritional value found in the barley endosperm is thionin; accumulation of this protein follows a course almost parallel to that of Dhordein (Ponz et al 1983). The changes in the proportions of a 75-kDa y-secalin and &>-secalin as the major and minor components, respectively, of the developing rye endosperm mirror to a large extent those for Bl- and C-hordeins of barley (Shewry et al 1983). In rice, two groups of polypeptides (22-23 kDa and 37-39 kDa) composed of glutelin continue to increase coordinately during development, whereas the levels of the 57-kDa and 76-kDa polypeptides remain fairly constant throughout endosperm development (Figure 12.6). The relative proportion of glutelin to prolamins accumulated is high in 10-day-old grains but decreases in 25-dayold grains due to an increased accumulation of prolamins (Yamagata et al 1982; Li and Okita 1993). Total proteins of castor bean endosperm (Gifford, Greenwood and Bewley 1982) and prolamins of
341
7-39k(
26k2-23k( 18k13k-
4 5 6 7 8 9 10 11 12 45 9 11 13 15 17 19 21 23 25 28 I
1
1
1
1
1
1
1
DAF mg
"
150-
^ 37 kd" _,> 23 kd g 13 kd
C
#100-
/
.-•57 kd
r
I 50
/ ^ —
/ 1
1
1
1
1
1
1
1
4 5 6 7 8 9 10 11 12 9 11 13 15 17 19 21 23 25
i
45 DAF 28 mg
Figure 12.6. Protein composition of the endosperm of Oryza sativa. (a) SDS-PAGE of the extant proteins of the developing endosperm. Numbers at the bottom refer to the fresh weight of a grain (mg) at the indicated stage, days after flowering (DAF). The apparent molecular masses of the proteins are given in kDa on the left, (b) Changes in the relative amounts of the proteins given as peak heights of densitometric scanning of the gel. The peak height of the 76-kDa protein is designated as 100%. (From Yamagata et al 1982. PlantPhysiol. 70:1094-1100. ©American Society of Plant Physiologists.)
sorghum endosperm (Taylor, Schiissler and Liebenberg 1985) show the same patterns of accumulation characterized by an almost linear increase throughout development. Assay of the major storage albumins and oleosins of the endosperm of Coricmdrum sativum (coriander; Apiaceae) using antibodies raised against similar proteins from carrot seeds showed that the albumins accumulate at a slightly earlier stage of development than the oleosins (Ross and Murphy 1992).
342 The dependence of continued accumulation of storage materials on protein synthesis has been demonstrated by the incorporation of labeled precursors in developing endosperms of rice (Luthe 1983; Zhu, Shen and Tang 1989; Li and Okita 1993), barley (Shewry et al 1979; Giese, Andersen and Doll 1983; Giese and Hejgaard 1984), wheat (Greene 1983), and oat (Luthe 1987). Because of the difficulty of harvesting endosperms at very early stages of development, these experiments have not generated a set of well-documented data on the pattern of synthesis of individual proteins during the developmental time span of the endosperm. From the foregoing account it is not clear whether the inability of endosperm cells at early stages of development to accumulate storage reserves is due to a high turnover or due to lack of the components necessary for their synthesis. Along the same lines, it is also not determined whether differences exist in the turnover rates of the different storage proteins in the species studied.
Zygotic Embryogenesis
products a few days before their detection in the endosperm cells in vivo (Figure. 12.7). Results of these studies in principle correlate well with those on protein metabolism in showing a coincidence between the periods of increased in vitro translational capacity of mRNA and increased protein synthesis or accumulation. (i) Maize Storage Protein Gene Expression
As the major class of storage proteins of the maize endosperm, zein has been studied intensively with the focus on gene expression. Following the first detection of zein mRNAs 12 days after pollination, the accumulation of these mRNAs in the developing maize endosperm reaches a maximum between 18 and 22 days, and thereafter it declines somewhat until maturation. Using groups of cDNA clones corresponding to 10-, 15-, 19-, and 22-kDa zeins as probes, it was estimated that the efficiency of expression of members of different zein families varies somewhat; generally, transcription of the gene encoding 15-kDa zein is initiated earlier and is 5. ANALYSIS OF STORAGE PROTEIN GENE more active than that of 19- and 22-kDa zeins EXPRESSION (Marks, Lindell and Larkins 1985a; Kirihara et al 1988). Patterns of accumulation of transcripts of Studies on the expression of storage protein individual zein genes also depend upon the growth genes during endosperm ontogeny have considered in vitro protein synthesis by mRNA and conditions and genotypes of the plant (Langridge, polysomes, changes in mRNA accumulation and Pintor-Toro and Feix 1982a, b; Kriz, Boston and synthesis, and the expression of cloned genes. Since Larkins 1987). At the cellular level, the noticeable these topics are of vital importance in understand- characteristic of zein genes is their exclusive expresing the structure of endosperm storage protein sion in the nucleus and cytoplasm of the endosperm genes and their expression in transgenic systems, cells, with the exception of the cells of the aleurone layer. As shown in Figure 12.8, a meaningful comthey will be briefly reviewed here. It was mentioned earlier that in vitro synthesis parison of transcript localization allows one to recof zein is programmed by polysomes and mRNA ognize specific territories of expression for each zein isolated from the developing endosperm of maize; mRNA within the developing endosperm (Dolfini et similar studies have reported the cell-free synthesis al 1992). Zein gene transcripts are also expressed in of globulins primed by polysomes of oat suspension cultures derived from the endosperm of endosperm (Luthe and Peterson 1977; Fabijanski maize, but expression occurs at a reduced level as and Altosaar 1985; Fabijanski, Matlashewski and compared to that of the endosperm of developing Altosaar 1985), hordeins and thionins primed by grains (Manzocchi, Bianchi and Viotti 1989; Ueda polysomes of barley endosperm (Brandt and and Messing 1991). Transcripts of the much less Ingversen 1976; Matthews and Miflin 1980; Ponz et investigated glutelin genes show a rate of accumulaal 1983; Rahman et al 1984), prolamins primed by tion in the developing endosperm similar to that of polysomes of sorghum endosperm 0ohari, Mehta 19- and 22-kDa zeins (Perez-Grau et al 1986). and Naik 1977), gliadin primed by polysomes and The apparent lack of coordination in the expresmRNA of wheat endosperm (Greene 1981; Okita sion of the individual zein genes raises the question and Greene 1982; Reeves, Krishnan and Okita of their control during endosperm development. 1986), secalin primed by mRNA of rye endosperm Determination of the levels of the different zein (Shewry et al 1983), and prolamins and glutelins mRNAs in run-off transcription assays from isoprimed by mRNA of rice endosperm (Yamagata lated endosperm nuclei of different ages has proand Tanaka 1986; Yamagata et al 1986). The storage vided evidence that the levels of the two oc-zein proteins are generally detected in the translation transcripts are generally higher than those of genes
343
Developmental Biology of the Endosperm
Time after anthesis (days) 18 22 26 34
14
18
22 26 30
(ft)
B1 60
f / / /
40
»
83 —»
N
20
C1
0 n.
"14 18 22 26 30
Time after anthesis (days) Figure 12.7. In vitro synthesis of hordeins by RNA isolated from the endosperm of Hordeum vulgare. Top panel: SDS-PACE of proteins translated from membrane-bound polysomes (a), polysomal poly(A)-RNA (b), polysomal RNA (c), and total RNA (d) at different days after anthesis.
Bottom panel: Densitometric determination of the relative incorporation of the labeled amino acid into proteins; (e-h) correspond, respectively, to (a-d). (From Rahman et al 1984.)
encoding p*-, y-, and 8-zeins during all developmental periods studied. Among the two a-zeins, the one encoding the 22-kDa a-zein was transcribed more actively than that encoding the 19kDa protein. Since the high transcriptional activity of the 22-kDa a-zein gene is not predicted by the low levels of its mRNA accumulation on polysomes, both transcriptional and posttranscriptional regulation of the zein genes appear to be
operating during endosperm development (Kodrzycki, Boston and Larkins 1989). Preferential accumulation of 10-kDa zein associated with the presence of a high level of its transcripts occurs in the endosperm of an inbred line of maize, BSSS-53, whereas an underexpression of this protein and its transcripts is seen in another inbred line, Mo-17 (Kirihara et al 1988; Schickler, Benner and Messing 1993). Evidence from analyses
344
Zygotic Embryogenesis
cation that they function to modulate transcript stability in some ways. PROBES In considering the relationship between the transcriptional activity of zein genes and the accuM1 mulation of protein products in the endosperm, the occurrence of mutations credited with reducing zein synthesis needs to be taken into account. Two of these mutations characterized initially were op-2 and fl-2, both of which cause a profound reduction in the synthesis of zein polypeptides while permitE19 ting an increase in the high-lysine-containing, nutritionally superior nonzein proteins (Mertz, Bates and Nelson 1964; Nelson, Mertz and Bates 1965). In the op-2 mutant, depending upon the particular M6 maize strain background, there is a partial to an almost complete inhibition of synthesis of 22-kDa zein at the same time that the levels of the lysinerich protein and elongation factor-la (EF-loc) are significantly increased (Lee et al 1976; Habben, G1 Kirleis and Larkins 1993; Habben et al 1995). The endosperm of this mutant begins to synthesize zein several days later and ceases activity earlier than the wild-type endosperm (Murphy and Dalby Figure 12.8. A diagrammatic representation of the pattern of 1971). The impact of EF-lct is that, as a lysine-rich accumulation of transcripts of zein and glutelin genes durprotein that binds aminoacyl-tRNAs to ribosomes, ing the development of the endosperm in Zea mays. The it could be expected to contribute significantly to diagram is based on in situ hybridization using probes for the increased lysine content. Kernels of the defective zein (Ml, M6, and E19) and glutelin (G1) sequences. DAP, days after pollination. (From S. F. Dolfini et al 1992. Spatial endosperm-B30 (de*-B30) mutant also show a preferregulation in the expression of structural and regulatory ential reduction in the accumulation of 22-kDa zein, storage-protein genes in Zea mays endosperm. Devel. whereas another mutation, op-7, causes a greater Genet. 13:264-276. © 1992. Reprinted by permission of reduction in the synthesis of 19-kDa than in the synWiley-Liss Inc., a subsidiary of John Wiley & Sons Inc.) thesis of 22-kDa zein. Other mutations, such as op-6, fl-2, and mucronate (me), suppress the synthesis of all of linkage relationships between the high mRNA zeins to the same extent (Lee et al 1976; Jones 1978; levels and the BSSS-53 allele for the 10-kDa zein di Fonzo et al 1979; Salamini et al 1983; Soave and structural gene locus seems to indicate that a postSalamini 1984). A newly characterized mutant, optranscriptional regulation of the mRNA level by a 15, shows a large reduction in the accumulation of trans-acting factor accounts for the increased accu27-kDa zein, but contains high concentrations of 19mulation of 10-kDa zein in the inbred line BSSS-53; and 22-kDa zeins (Dannenhoffer et al 1995). Effects a mechanism imposed by the lack of one or more of the mutation extend to the level of the structure frans-acting factors might account for the underexof the protein bodies and the distribution of the varpression phenotype in Mo-17 (Cruz-Alvarez, ious classes of zeins in them. For example, the disKirihara and Messing 1991; Schickler, Benner and organized protein bodies with clumps of P- and Messing 1993; Chaudhuri and Messing 1994). y-zeins interspersed within a mass of a-zein found Tissue-specific frans-acting factors also seem to in the endosperm cells of the/Z-2 mutant provide a determine the transcription of the 19-kDa and 15striking visual contrast to the organized organelle kDa zein genes and the 28-kDa glutelin gene by structure in the cells of the wild-type endosperm binding to their 5'-flanking regions. Five sites in the (Lending and Larkins 1992). promoter region of the 19-kDa zein gene capable of binding to a crude protein fraction of maize The complete mechanism by which the mutaendosperm nuclei include a 22-bp region contain- tions cause a reduction in zein synthesis is yet to be ing a 15-bp-long consensus sequence conserved in determined. Convincing evidence that the mutavirtually all zein genes investigated (Maier et al tions affect the regulatory functions of genes rather 1987, 1988; So and Larkins 1991; Ponte et al 1994). than their organization was provided by restriction The action of frans-acting factors carries the impli- analysis of genomic DNA of op-2, op-7, and fl-2 DAP
10
13
17
20
Developmental Biology of the Endosperm
mutants with cloned cDNA probes for zein genes; this work showed that the coding sequences of the mutant genes had no deletions or structural alterations that might affect zein accumulation (Burr and Burr 1982). The patterns of accumulation of specific zein mRNAs in the mutant endosperms are relatively well described in several studies that show that the reduction of zein proteins is correlated with reduced amounts of the corresponding mRNAs (Pedersen et al 1980; Burr and Burr 1982; Langridge, Pintor-Toro and Feix 1982a; Marks, Lindell and Larkins 1985a; Kodrzycki, Boston and Larkins 1989; Lopes et al 1994; Dannenhoffer et al 1995; Neto et al 1995). In particular, during development of the op-2 endosperm, none of the zein mRNAs can be detected at 12 days, as they would be in the normal endosperm. The mRNAs for 19kDa zeins attain relatively normal levels by 28 days. Since the mRNAs arrive at that level slowly, it is clear that the mutation delays the expression of the 19-kDa zein gene (Figure 12.9). The appearance of 15-kDa zein mRNAs is also delayed and the transcript level is greatly reduced in the mature endosperm. Although the level of 22-kDa zein polypeptides is substantially reduced by the op-2 mutation, deficiency in certain subgroups of 22kDa zein transcripts, rather than in the whole range of transcripts that encode this protein, appears to cause the reduction (Burr and Burr 1982; Marks, Lindell and Larkins 1985a). This explanation of the effects of the op-2 mutation on zein protein mRNA synthesis does not, however, preclude additional levels of regulation. In vitro culture studies have shown that the accumulation of 22-kDa zein and its mRNAs in the endosperm of the op-2 mutant is favored when an excess of nitrogen is supplied in the medium (Balconi et al 1993). Thus, there appears to be a correlation between the level of nitrogen supply and the phenotypic effect of the mutation. Two groups of investigators have cloned the OP-2 gene by transposon tagging with the help of the mobile elements suppressor mutator (spm) and activator (ac). As expected, the gene is expressed in the wild-type endosperm but not in nonendospermic tissues or in the endosperm of the op-2 mutant (Schmidt, Burr and Burr 1987; Motto et al 1988; Gallusci, Salamini and Thompson 1994). The theoretical translation of cDNA clones suggests that the gene encodes a transcription factor that contains a leucine zipper DNA-binding protein known as the bZIP domain (Hartings et al 1989; Schmidt et al 1990). The endosperm-specific expression of the OP-2 gene correlates well with the expression of
345
DAP 16 27.000
16.000 14.000 10.000
18
20 27.000
10,000
Figure 12.9. SDS-PACE of zeins extracted from wild-type (a) and op-2 mutant (b) of Zea mays to show the developmental regulation of zein gene expression. Developmental stages are indicated at the top as days after pollination (DAP). Molecular masses of the zein fractions are shown on the right. (From Kodrzycki, Boston and Larkins 1989. Plant Cell 1:105-114. © American Society of Plant Physiologists. Photograph supplied by Dr. B. A. Larkins.)
the OP-2 protein in the nuclei of maize endosperm cells (Varagona, Schmidt and Raikhel 1991). A view advanced to explain the molecular basis for zein gene expression is that OP-2 protein activates zein gene transcription by binding to promoters. Some support for this idea has come from the demonstration that the protein product of the OP-2 gene expressed in Escherichia coli can bind in vitro to upstream regions of a zein gene. Similarly, the OP-2 protein transiently expressed in the protoplasts of cultured maize endosperm cells or in the cells of the intact maize endosperm can transactivate the 22-kDa zein gene promoter (Schmidt et al 1990; Ueda et al 1992; Unger et al 1993). Conversely, the protein of an op-2 mutant causing a disruption in zein gene expression is unable to bind to the same zein gene promoter (Aukerman et al 1991). To account for these complementary results, a 20nucleotide target site has been identified in the promoter of the 22-kDa zein gene (Schmidt et al 1992). Because this sequence is not present in the promoters of other zein genes, this is prima facie evidence to explain the differential effect of the op-2 mutation on the 22-kDa zein. Unraveling the ramifications of the mutation on zein gene expression is likely to assume complex proportions as other tran-
346
scriptional regulatory proteins are identified (Yunes et al 1994; Holdsworth et al 1995). Another bZIP protein that might play a role in the regulation of the 22-kDa zein gene is the product of a maize endosperm cDNA clone (OHP-1) that binds to the OP-2 target site in the promoter region of the 22-kDa zein gene (Pysh, Aukerman and Schmidt 1993). Considering the fact that the OP-2 gene affects the expression of all subunits of zein, a finding with regulatory implications for the role of OP2 protein is that this protein transactivates the promoters of P-zein of maize and a P-zein homologue from the related member of Poaceae, Coix lacryma-jobi (Neto et al 1995). Another relevant observation is that mutation in the OP-2 gene down-regulates the accumulation of cytoplasmic pyruvate orthophosphate dikinase (PPDK) in the endosperm. Since PPDK is an enzyme involved in C4-photosynthesis, this finding might provide a meeting point for the roles of the OP-2 gene in coordinating the expression of zein genes and genes for carbon partitioning (Maddaloni et al 1996). The tide of reports of the isolation of transcription factors from the endosperm has continued to rise. Characterization of a salt-soluble 32-kDa protein (B-32) from the endosperm of wild-type maize and its absence in the op-2 and op-6 endosperms have led to the highly interesting idea that B-32 protein plays a role in the control of zein synthesis, perhaps as a transcription factor. The protein is coordinately expressed with zein during the development of the endosperm but is localized in the soluble cytoplasm, rather than in protein bodies. It has been speculated that B-32 protein is responsible for regulating zein gene expression in an indirect way. This view assumes that B-32 protein is encoded by the OP-6 gene, which in turn regulates zein synthesis, and that the OP-2 gene activates the OP-6 gene (Soave et al 1981; di Fonzo et al 1986). It is, however, doubtful that B-32 protein is a product of the OP-6 locus, and its abundance and location in the cytoplasm would argue against a direct interaction with zein or zein genes as a regulatory protein. The structure of the gene encoding B-32 protein and the deduced amino acid sequence have not given any indication of the presumed relationship of the protein to the OP-6 gene or its translation product (di Fonzo et al 1988; Hartings et al 1990; Bass et al 1992). According to Lohmer et al (1991), the B-32 gene is transactivated in the presence of the OP-2 gene product, indicating that the interaction between the OP-2 and B-32 genes is at the transcriptional level and is not indirect through the activation of the OP-6 gene. Admittedly, B-32
Zygotic Embryogenesis
protein has received a good deal of attention as a regulator of zein gene expression, but its precise function remains unresolved. Another contribution to our understanding of the action of fl-2, de*-B30, and me mutations in maize showed that endosperms carrying these mutations overproduce a 70-kDa water-soluble protein (B-70) associated with both protein bodies and the rough ER (Galante et al 1983). Dramatic increases in the levels of both B-70 protein and its corresponding mRNA occur in the endosperms of these mutants between 14 and 20 days after pollination. This is followed by a decline in mRNA levels to the vanishing point, whereas B-70 protein levels remain high in the morphologically perturbed protein bodies (Boston et al 1991; Zhang and Boston, 1992). It appears that B-70 is a cytosolic chaperonelike protein with affinities to BiP and HSP-70 (Fontes et al 1991; Marocco et al 1991). Another candidate for a regulatory function in zein gene expression is an unusual 24-kDa protein associated with the zein fraction of the endosperm of the fl-2 mutant (Lopes et al 1994). Results from the cloning of the gene for the 24-kDa protein and from its deduced amino acid sequence indicate that the gene is expressed as a 22-kDa ct-zein with a defective signal peptide (Coleman et al 1995). Irrespective of the nature of the B-70 and 24-kDa proteins, there is circumstantial evidence to suggest they are involved in some ways in the disruption of zein processing and deposition in the endosperms of the mutants. The phenotypic effect of the op-2 mutation typically results in the formation of a soft, floury endosperm. A significant development affecting the phenotype of the op-2 mutant is the modification of all or parts of the endosperm to a vitreous or glassy consistency. Genes that condition this phenotype are known as OP-2 MODIFIERS; the same genes are also active in changing the phenotype of the endosperm in other mutants such as fl-2. Wallace et al (1990) showed that the principal change in the endosperm storage protein of the modified op-2 mutant is a two- to fourfold increase in the y-zein content without any appreciable effect on the other classes of zeins. A central question in analyzing the effects of the various modifier genes is whether there are differences in the fundamental mechanisms of gene transcription or posttranslational events affecting protein synthesis. Experiments designed to answer this question have shown that opaque modifiers confer an increase in the level of y-zein mRNAs; since this occurs in the absence of any increase in the number of gene copies, it appears that the increased gene expression is due to
Developmental Biology of the Endosperm
enhanced mRNA transcription or stability (Geetha et al 1991; Lopes and Larkins 1991). However, there is only a marginal difference between the transcription rates of the "y-zein gene in the op-2 mutant and in the modified op-2 genotypes. This has ruled out a direct role for the modifier genes in the transcription of the y-zein gene and suggests a posttranscriptional role for them (Or, Boyer and Larkins 1993). (ii) Barley Storage Protein Gene Expression
Some observations support the view that the amounts of different hordeins synthesized by the developing barley endosperm are determined by the abundance of corresponding mRNAs present, with little room for translational and posttranslational controls. Quantification of the relative amounts of B1-, B3-, and C-hordeins synthesized in vitro by mRNA preparations from developing endosperm showed that the amount of Bl-hordein increased, and that of B3- and C-hordeins decreased, with age. The population of the corresponding mRNA species determined by dot-blot hybridization using cDNA clones also showed similar changes during endosperm development (Rahman et al 1984). It must be kept in mind that this kind of correlative evidence can never be considered complete because it is difficult to ascertain whether the amount of mRNA present at any given time is due to gene transcription or mRNA turnover. However, in vitro transcription rates of all hordein genes are more or less the same despite the fact that the steady-state levels of some hordein mRNAs are low; this indicates that the high levels of hordein mRNAs are due to posttranscriptional regulations (Sorensen, CameronMills and Brandt 1989).
347
Aalen et al 1994; Kalla et al 1994; Aalen 1995). Transcripts of another aleurone-specific recombinant (LTP-1) also accumulate characteristically in the aleurone cells during late grain development (Skriver et al 1992). Steady-state levels of mRNAs that encode a protein with some sequence homology to LEA proteins begin to accumulate in the aleurone and in the embryo midway through development and remain high in the aleurone cells of the mature endosperm (Hong, Barg and Ho 1992). Genes for chitinase isolated from barley endosperm display either aleurone-specific or both aleuroneand starchy endosperm-specific patterns of expression. This protein is generally reckoned as part of the grain's defense arsenal against invading pathogens (Swegle, Kramer and Muthukrishnan 1992; Leah et al 1994). Mutants impaired in the synthesis and deposition of hordeins have been identified in cultivated varieties of barley. The mutant Rise-56 suffers a 30% reduction in overall hordein accumulation and, in particular, the amount of B-hordein is reduced by about 75%. This reduction has been traced to the absence of B-hordein mRNA in the mutant endosperm due to a deletion covering most of the hordein-2 locus (Kreis et al 1983). In another mutant, Ris0-15O8, the amounts of B- and Chordeins are reduced to about 20% and 7%, respectively, of those present in the parent line at the same time as the amount of D-hordein is increased fourfold. The changed amounts of B-, C-, and Dhordeins in this mutant are accompanied by reduced levels of B- and C-hordein mRNAs and a twofold increase in the amount of D-hordein mRNA, and by a dramatic reduction in the in vitro transcription activity of the genes encoding B- and C-hordeins (Thompson and Bartels 1983; Kreis et al 1984; Cameron-Mills and Brandt 1988; S0rensen, Cameron-Mills and Brandt 1989). Thus, the primary lesion caused by the mutations appears to be at the transcriptional level. The catastrophic effects of the mutation in Rise-1508 have been further traced to transcriptional inactivation of the genes by methylation, whereas hypomethylation of the genes is correlated with their specific expression in the endosperm of the parent (Sarensen 1992).
The use of cDNA clones isolated from aleurone cells of the barley endosperm has yielded vital information on the time course and pattern of accumulation of aleurone mRNAs. Transcripts of an aleurone-specific cDNA clone (B11E) and its genomic clone (LTP-2) show a nearly 200-fold increase in abundance exclusively in the aleurone cells beginning in the early-stage endosperm and extending to the stage of peak accumulation. During desiccation of the grain, transcript levels decrease; they disappear completely from the aleurone cells of the mature grain (Figure 12.10). In con(iii) Wheat Storage Protein Gene trast, mRNAs complementary to other cDNA clones Expression expressed in both the aleurone and the embryo are present more or less uniformly in the aleurone Storage protein mRNAs of the developing throughout grain development and decrease coinci- wheat endosperm are attached to both membranedentally with those of B11E (Jakobsen et al 1989; bound and free polysomes, with the former makOlsen, Jakobsen and Schmelzer 1990; Liu et al 1992; ing up the major component of the storage protein
348
Zygotic Embryogenesis
DPA
0
2 4
OPI
6 8 10 12 16 20 25 30 35 40 45 50 2
kb
Figure 12.10. Temporal and spatial patterns of accumulation of transcripts of aleurone layer-specific genes during the development of the endosperm in Hordeum vulgare. (a) Northern blot hybridization showing steady-state levels of B11E mRNA in unfertilized ovules (0 days postanthesis, DPA), deembryonated grains (2-10 DPA), isolated aleurone-pericarp (12-45 DPA), mature aleurone (50 DPA), and mature aleurone imbibed for two days (2 days postimbibition, DPI). (From Olsen, jakobsen and Schmelzer
1990. Planta 181:462-466. © Springer-Verlag, Berlin.) (b, c) In situ hybridization of a section of the grain with labeled LTP-2 antisense strand. Dark-field micrographs showing hybridization of the probe to the peripheral aleurone cells, but not to the starchy cells of the endosperm. A portion boxed in (b) is magnified in (c). AL, aleurone layer; CC, cross-cell layer; MA, modified aleurone cells; NE, nucellar epidermis; PE, pericarp; SE, starchy endosperm. Scale bars = 100 urn (b) and 250 urn (c). (From Kalla et al 1994.)
messages. Based on hybridization kinetics, it has been estimated that both types of polysomes are constituted only of superabundant and intermediate abundant classes of mRNAs. Although there is no evidence for selective amplification of gliadin and glutenin genes in the developing wheat endosperm, it appears that more than half of the translated mRNAs direct the synthesis of the gliadins (Pernollet and Vaillant 1984). The results of changing patterns in RNA synthesis and in levels of the transcripts of different subunits of gliadins and glutenins have been taken to indicate that the subunit genes are expressed coordinately and that the amounts of mRNAs for the proteins increase sharply until the mid-maturation phase of endosperm development and then decline. The major RNA transcripts that accumulate during the peak period of endosperm development encode the a/P- and y-gliadins. The pattern of accumulation of mRNAs of triticin, a globulinlike storage protein that constitutes only about 5% of the total wheat endosperm protein, is very similar to that of
gliadins. As no protein is detected in the developing endosperm before its homologous mRNA is synthesized, a transcriptional control mechanism appears to coordinate storage protein synthesis in the wheat endosperm (Terce-Laforgue and Pernollet 1982; Okita 1984; Bartels and Thompson 1986; Reeves, Krishnan and Okita 1986; Singh et al 1993). These data are in general accord with those described for maize and barley endosperms. In attempts to characterize the protein factors that regulate transcription of gliadin genes, Vellanoweth and Okita (1993) have identified six sequences in the promoter of a oc/p-gliadin gene that interact with different nuclear proteins from the developing endosperm (Figure 12.11). When nuclear extracts from endosperms of different stages of development were assayed, activities of four of the nuclear factors were found to appear at early stage to midstage, reach a maximum at midstage, and then decline later during endosperm ontogeny. The temporal changes in the activities of nuclear factors are coincident with the periods of
Developmental Biology of the Endosperm
349 -163 Spel
-265 Xbal
-46 Ncol
I
CHD
CAAT
xs
TATA
SN
SN
c5 c4 c3 c2 d
1 2
3
4
7
8
9
10
11 12 13 14
Figure 12.11. Interaction of nuclear proteins of Triticum aestivum with the cx/p-gliadin promoter, (a) A schematic representation of the promoter region of the o/P-gliadin gene. Numbers refer to positions upstream of the transcriptional start site. TATA and CAAT elements are shown in boxes. The arrow indicates the direction of transcription. The probe and competitor fragments used in the experiment shown in (b) are indicated, (b) Electrophoretic mobility shift assay using restriction fragments SN (-163 to -46 bp) in lanes 1-7 and XS (-265 to -163 bp) in lanes 8-14.
The amounts of crude nuclear proteins used were (ug): 0.5 (lanes 1, 8); 1 (lanes 2, 9); 2 (lanes 3, 10); 4 (lanes 4, 6, 7, 11, 13, 14); and 8 (lanes 5, 12). Lanes 6 and 13 and lanes 7 and 14 contained a 100-fold molar excess of unlabeled XS and SN DNA fragments, respectively. The five complexes formed with SN are indicated on the left as d to c5; those formed with XS are marked on the right as d ' to c4'. A sixth, low-mobility complex formed with SN is not seen in the figure. (From Vellanoweth and Okita 1993; photograph supplied by Dr. T. W. Okita.)
maximum accumulation and decrease of gliadin mRNA in the endosperm. Activity of a fifth nuclear factor was found at high levels only at mid-development; a sixth nuclear factor displayed a constant level of activity throughout endosperm development. Because the last-mentioned nuclear factor remains active even after transcription for gliadins has stopped, it has been surmised that some binding proteins may regulate gliadin expression by a negative frans-acting effect. An interaction between the putative trans-acting factors and sequences of the endosperm box (see p. 350) and GCN4-like (for General Control Nonderepressible) motifs
that bind yeast regulatory proteins found in the low-molecular-mass glutenin has also been demonstrated by in vivo dimethylsulfate (DMS) footprinting. Not only is the binding highly specific to endosperm nuclear proteins, but the two types of proteins are found to bind sequentially, whereby the endosperm box motif becomes occupied during the early stages of development of the endosperm and the GCN4 sequence is coupled during the later stages (Hammond-Kosack, Holdsworth and Bevan 1993). From these results, a good case can be made that specific sequences in the promoter regions of the gliadin and glutenin
350 genes combine with endosperm nuclear factors to regulate their temporal and spatial expression during development. (iv) Storage Protein Gene Expression in Rice and Oat
Most of the evidence linking storage protein gene expression with endosperm development in rice comes from studies of glutelin and prolamin genes, the former representing a significantly higher proportion of the total proteins than prolamins. What is the nature of the control imposed on the relative expression of these proteins? In the absence of significant differences in the steady-state levels and transcriptional rates of glutelin and prolamin genes in the developing endosperm, it has been argued that differential recruitment of the respective mRNAs into polysomes might explain the differences in protein accumulation (Kim, Li and Okita 1993). In a similar way, there is a predominance of globulin over avenin in the endosperm of oat; yet, the amount of avenin mRNA is equal to or greater than the amount of globulin mRNA present during most periods of endosperm development (Chesnut et al 1989). Given the uncertainties about the involvement of nontranscriptional processes in the synthesis of proteins, the reasons for the translational efficiencies of these genes deserve further study. The possibility that expression of glutelin and prolamin genes during rice endosperm development is noncoordinately regulated arises out of experiments on the transcriptional activities and transcript levels of several cDNA clones of these gene families. Comparative patterns of expression of 15.2-kDa prolamin messages belonging to two different homology classes showed that although both mRNAs increase during the 10 to 25 days following anthesis, there is a three- to fourfold increase in the abundance of one message during the early phase of endosperm development (Kim and Okita 1988b; Kim, Li and Okita 1993). Like the 15.2-kDa prolamin genes, the gene for a 13-kDa prolamin continues to be expressed at a high level up to the late stage of endosperm development (Masumura et al 1990). The expression of glutelin genes belonging to three subfamilies (GT-1, GT-2, and GT-3) shows that although transcripts of the homologous GT-1/GT-2 genes steadily accumulate during endosperm development beginning 5 days after anthesis, those of GT-3 accumulate maximally during the 5 to 10 days after anthesis, followed by a steady decline to imperceptible levels (Okita
Zygotic Embryogenesis
et al 1989; Takaiwa and Oono 1991; Takaiwa et al 1991; Kim, Li and Okita 1993). At different times during rice grain development, the glutelin gene is expressed not only in the endosperm but also in the embryo and pericarp (Ramachandran and Raghavan 1990). Izawa et al (1994) have characterized a bZIP transcription factor from the rice endosperm (RITA1) that shows an affinity to ACGT motifs. The expression of the RITA-1 gene in the aleurone and starchy endosperm cells of transgenic rice grain implies that this transcription factor might play a major role in the control of storage protein gene expression in the endosperm. 6. CLONING AND CHARACTERIZATION OF ENDOSPERM STORAGE PROTEIN GENES
Isolation and characterization of the relevant genes are necessary to gain insight into the molecular mechanisms of storage protein synthesis in the endosperm. The structure of the genes encoding many endosperm storage proteins is now known. Analyses have defined and delineated the differences among the different genes and have led to the recognition of a 7-nucleotide sequence (TGTAAAG), termed the -300 bp element, in the endosperm storage protein genes of several cereals. This sequence is frequently referred to as the "endosperm box" and is believed to program the endosperm-specific expression of the genes (Figure 12.12). Later progress in this area came with the dissection of promoter sequences of the genes, construction of deletion mutants, and introduction of chimeric genes into transgenic plants to follow the genes' spatial and temporal expression. This section presents an overview of recent findings pertaining to the various levels of characterization of endosperm storage protein genes. (i) Zein Genes
Based on classical genetic approaches, nucleic acid hybridization analysis, and Southern blot hybridization, there have been several revisions of the estimated number of genes encoding zein. The consensus of the prevailing opinion is that a-zein is encoded by a large multigene family of 50-100 gene members, whereas (3-, y-, and 8-zeins are encoded by genes present in a few copies. Several cDNA and genomic clones encoding zeins have been characterized to provide a clear picture of the molecular biology of the gene. Genes characterized during the early period of zein gene cloning
Developmental Biology of the Endosperm
PBHR184:
•563
351 -302 WCAT6TAAAGTGMTAAGGT
pW8233:
GTAAAGTGAMTAAGATGAGlrCATGCAT
pMLl: ZG99: -333 Figure 12.12. Conserved distal sequences of the -300 bp endosperm box in hordein (pBHR184), a-gliadin (pW8233), 21-kDa zein (pML1), and 19-kDa zein (ZG99) genes. The sequence of the repeat of the -300 bp element in the B1 hordein gene is incomplete because the repeat contains one of the two EcoR1 sites that define the boundaries of the
sequenced fragment. Asterisks indicate identical residues in each pair of aligned sequences; those parts of the element that are common to the prolamin and zein genes are boxed. Numbering is relative to the ATC codon. (From Forde et al 1985a. Nucl. Acids Res. 13:7327-7339, by permission of Oxford University Press.)
belonged to the two major classes of oc-zeins (19 kDa and 22 kDa). A list of cDNA or genomic clones of the two a-zeins includes those isolated by Weinand, Briischke and Feix (1979), Geraghty et al (1981), Wienand, Langridge and Feix (1981), Burr et al (1982), Geraghty, Messing and Rubenstein (1982), Marks and Larkins (1982), Hu et al (1982), Spena, Viotti and Pirrotta (1982), Pedersen et al (1982), Heidecker and Messing (1983), Marks, Lindell and Larkins (1985b), Kriz, Boston and Larkins (1987), Wandelt and Feix (1989), and Liu and Rubenstein (1993). Clones encoding 15-kDa zein (Marks and Larkins 1982; Marks, Lindell and Larkins 1985b; Pedersen et al 1986), 27-kDa zein (Wang and Esen 1986; Das and Messing 1987), and 10-kDa zein (Kirihara, Petri and Messing 1988; Kirihara et al 1988), were isolated subsequently, completing the characterization of clones representing all zein subclasses. cDNA clones corresponding to some storage proteins of maize endosperm previously considered as zeins have been assigned to 28-kDa and 16-kDa glutelins because of their close homology in both nucleotide and amino acid sequences to gliadin and hordein (Prat et al 1985; Boronat et al 1986; Prat, Perez-Grau and Puigdomenech 1987; Gallardo et al 1988; Reina et al 1990a, b).
genes have consensus CAAT and TATA sequences in their 5'-noncoding region and one or more polyadenylation sites in the 3'-noncoding part. Differences between the different zein genes, which probably account for their operation, can be traced to the duplication of an original gene, the presence of repeated sequences, the insertion of sequences by duplication of small segments, and nucleotide changes. Based on the structure of individual genes, there is some interesting speculation that members of the zein multigene family have evolved from a common ancestral sequence. Protein structures of zeins deduced from cDNA clones show no homology to major storage proteins of other cereals. Among the different zeins, there is close similarity between the 19-kDa and 22-kDa zeins, as both contain a sequence of 20 amino acids repeated nine times. The amino acid repeats are not part of the sequence of the 15-kDa zein, which thus does not share homology with the 19-kDa and 22kDa zeins. The mature 10-kDa has no homology to the other zeins characterized (Wilson and Larkins 1984; Heidecker and Messing 1986). Although the sequences and structure of the 28-kDa glutelins are different from those of zeins, the 5'- and 3'-flanking regions of the two groups of genes show considerable homology (Boronat et al 1986).
Comments about the structure of the zein genes will be limited to general features that are important in their functioning. The zein genes belong to a rare class of eukaryotic genes that do not contain any introns, thus making their coding regions very long. All zein genes have repeated sequences, transcriptional regulatory sequences characterized by the presence of two or more promoters, as well as the —300 bp element and transcription initiation sites. The unusual promoter system may account for the high level of zein expression in the endosperm. Like typical eukaryotic genes, zene
The transcriptional regulation of zein genes has been monitored by introducing individual genomic clones into heterologous tissues and examining their expression in the absence of other members of the multigene families. Transformed yeast cells used as a transcription system are able to recognize the promoter regions of 19- and 22-kDa zein genes and synthesize the appropriate transcripts (Langridge et al 1984). By integrating the gene into Agrobacterium tumefaciens, it was shown that the 5'flanking regions of 15-kDa, 19-kDa, and 22-kDa zein genes contain sufficient information to accu-
352
Zygotic Embryogenesis
rately direct transcription in transformed sun(ii) Hordeins and Other Genes flower tissues or in transgenic Petunia hybrida from Barley (Matzke et al 1984; Goldsbrough, Gelvin and Larkins 1986; Ueng et al 1988). In addition, the proThe first characterized hordein genes included moters of some zein genes have been shown to those for B1-, B3-, and C-hordeins (Forde et al direct expression of fused CAT, GUS, or luciferase 1985a, b). Comparison of the 5'-flanking regions of genes in protoplasts of carrot somatic cells, tobacco barley, wheat, and maize prolamin genes showed mesophyll cells, and maize endosperm trans- the presence of short, highly conserved sequences formed by electroporation or by polyethylene gly- within 600 bp of the translation initiation site. col (Boston et al 1987; Schwall and Feix 1988; Besides a CAAT sequence and a TATA-box, a Quayle, Hetz and Feix 1991; Giovinazzo et al 1992). sequence with potential significance in the control The effect of deletion of DNA sequences flanking of gene expression was conserved around -300 bp the gene encoding a 19-kDa zein on zein mRNA in all three classes of prolamin genes. Sequence expression in sunflower tissues and CAT expres- homologies were also found between B- and Csion in carrot protoplasts has helped to identify hordeins. Other investigators have subsequently functional upstream noncoding sequences that are cloned and characterized additional Bl- (Brandt et required for optimum transcription. This region is al 1985; Chernyshev et al 1989), y- (Cameron-Mills unusual in sharing minor homology with a virus and Brandt 1988), and C-hordein (Rasmussen and enhancer core sequence (Roussell et al 1988; Brandt 1986; Entwistle 1988) genes. The aleuroneThompson et al 1990). specific genes LTP-1 and LTP-2 from barley encode Initial experiments using zein genes under the a putative nonspecific lipid transfer protein; both control of homologous regulatory elements did not genes contain regulatory motifs such as the MYB lead to gene expression in a tissue-specific manner (identified with myeloblastosis-associated viruses) and did not always reveal the synthesis of and MYC (identified with myelocytomatosis-assodetectable protein in the cells. In later experiments, ciated viruses) recognition sites (Linnestad et al transcriptional fusions composed of 15-, 19-, and 1991; Skriver et al 1992; Kalla et al 1994). These con28-kDa zein genes placed under the control of the sensus binding sites might interact in the regulaflanking regions of the gene for (3-phaseolin (the tion of the LTP-1 and LTP-2 genes, whose protein globulin storage protein of the embryos of products may have a protective function in the Phaseolus vulgaris), CaMV 35S promoter, or the GUS aleurone layer. A gene designated as B22E, which is reporter gene were found to be expressed in the expressed in the endosperm and in the embryo of protein bodies of the endosperm and embryo of barley, encodes two unidentified proteins charactransformed Petunia hybrida, tobacco, and terized by high cysteine contents (Klemsdal et al Arabidopsis (Hoffman et al 1987; Williamson et al 1991). 1988; Quattrocchio et al 1990; Geli, Torrent and In the first report of a regulated expression of a Ludevid 1994; Bagga et al 1995). However, reports barley hordein gene in transgenic plants, Marris et of endosperm-specific activity in transgenic al (1988) fused the 5'-flanking region of Bl-hordein tobacco plants of a 22-kDa zein gene promoter gene to the CAT reporter gene. Transfer of the placed upstream of the GUS reporter gene have chimeric gene into tobacco plants showed that the proved to be unreliable (Schernthaner, Matzke and hordein gene regulatory sequences direct the synMatzke 1988; Matzke et al 1990; Matzke, Stoger and thesis of CAT only in the endosperm in a developMatzke 1993). Both 19-kDa and 22-kDa zeins are mentally regulated manner. However, attempts to synthesized in the cytoplasm of unicellular, uninu- transform plants with other constructs of the cleate green alga Acetabularia mediterranea follow- hordein gene have given variable results. For ing microinjection of the genes into the nuclei and example, a chimeric B22E promoter-GlTS gene the transplantation of the nuclei into enucleate introduced into barley grains by particle bombardalgal cells (Langridge et al 1985; Brown, Wandelt ment is functional in directing transcription in the and Feix 1986). In this system, and in the Hela cell aleurone cells of the endosperm, but it is not active and Xenopus germinal vesicle transcription sys- in the seeds of transformed tobacco (Klemsdal et al tems, the 19- and 22-kDa zein genes are transcribed 1991). The expression of the promoter of the Blwith unequal efficiency from two widely placed hordein gene linked to the GUS gene delivered promoters (Langridge, Pintor-Toro and Feix 1982b; into barley endosperm is confined to mature aleuLangridge and Feix 1983; Brown, Wandelt and Feix rone cells and excludes the developing aleurone 1986). cells (Knudsen and Miiller 1991). Use of a similar
Developmental Biology of the Endosperm
type of transient assay for deletion analysis of the promoter of chitinase gene CHI-26 has made it possible to identify a 20-bp sequence that contains two hexameric direct repeats and confers developmental and aleurone-specific expression (Leah et al 1994). In the transient assay, the promoters of the LTP-1 and LTP-2 genes also display activity in the aleurone cells (Skriver et al 1992; Kalla et al 1994). It appears that barley endosperm storage protein genes contain sequences specific to conferring expression separately in monocot and dicot tissues and in the aleurone and starchy cells of the endosperm. (iii) Gliadin and Glutenin Genes A large multigene family encodes the gliadins, which have probably evolved by repeated divergence from an ancestral gene. The various gliadins are related and are coded by three gene subfamilies, namely, ct/p-, y-, and o-gliadins, the oc/p-type being the largest subfamily. The basic structural features of the coding regions and promoter elements of the oc/ P-gliadins have been deduced from cDNA and genomic clones (Bartels and Thompson 1983; Forde et al 1983; Anderson et al 1984; Kasarda et al 1984; Rafalski et al 1984; Okita, Cheesbrough and Reeves 1985; Sumner-Smith et al 1985; Reeves and Okita 1987). Deletions, duplications, and single base substitutions are believed to generate the sequence differences between the genes in each subfamily. Genomic sequences have revealed typical 5'-promoter elements, two polyadenylation signals, and a lack of introns. At the protein level, the deduced amino acid sequences share a primary structure of six domains consisting of a signal peptide, a region of repeated sequences arranged in tandem, and two polyglutamine stretches interrupted by two regions having distinctive amino acid compositions. There is a close correspondence between the coding sequences of oc/p-gliadin and y-gliadin genes, especially in the lack of introns. The amino acid sequences derived from the ygliadin gene reveal a signal peptide, a region high in glutamine and proline, and a C-terminus (Okita 1984; Okita, Cheesbrough and Reeves 1985; Bartels et al 1986; Rafalski 1986; Sugiyama, Rafalski and Soil 1986; Scheets and Hedgcoth 1988). The main focus of nucleotide sequence analysis of the glutenin genes has been on the high-molecular-mass subunits. They are encoded by 6-18 genes and belong to a small multigene family. The coding sequences of the genes are uninterrupted by introns, and the promoter regions contain a
353
sequence that somewhat resembles the conserved -300 bp element of the endosperm box. The proteins encoded by the glutenin subunit genes possess a structure composed of three linked domains, namely, an N-terminal domain of about 100 amino acids, a repetitive region of over 500 amino acids, and a C-terminus of about 40 residues (Thompson et al 1983; J. Forde et al 1985; Sugiyama et al 1985; Thompson, Bartels and Harberd 1985; Halford et al 1987; O. D. Anderson et al 1989). Several ris-acting motifs that include the endosperm box and the DNA sequences that bind the GCN4-JUN (for JUNNA) class of mammalian and yeast regulatory proteins form a conspicuous part of the structure of a low-molecular-mass glutenin gene. At the protein level, the sequence has a signal peptide, a central repetitive region enriched for glutamine and proline residues, and N- and C-terminals (Colot et al 1989). Although the nucleotide sequence of a triticin cDNA clone shares some features of gliadin and glutenin genes, its predicted amino acid sequence shows strong homology to US proteins of legume embryos (Singh et al 1993). The sequences of the gliadin and glutenin genes that regulate their expression have been sought by transient assays and by introduction into heterologous hosts. Promoters of a/P-gliadins were weak in transient expression assays in tobacco protoplasts and in protoplasts prepared from maize endosperm suspension culture (Aryan, An and Okita 1991; Blechl et al 1994). However, parts of the gliadin promoter placed upstream of either a nonfunctional or a functional version of the NOS promoter can restore the transcriptional activity of the nonfunctional element or enhance the activity of the functional sequence in tobacco protoplasts (Aryan, An and Okita 1991). Approaches to identify cfs-acting elements important in the spatial and tissue-specific transcription of glutenin genes have been more rewarding than with gliadin genes. Chimeric genes consisting of low-molecular- and high-molecularmass glutenin gene promoters linked to the CAT gene display endosperm-specific expression in the seeds of transgenic tobacco. The results of deletion series of low- and high-molecular-mass glutenin genes showed that sufficient information to confer endosperm-specific expression lies between -360 to -160 bp and -433 bp upstream of the respective promoters (Colot et al 1987; Robert, Thompson and Flavell 1989). When transcription of truncated high-molecular-mass glutenin gene fragments was enhanced by CaMV 35S promoter, only sequences spanning -375 to -45 bp in the upstream region
354
Zygotic Embryogenesis Endosperm-specific expression
pGUS45T
3.35 ± 2.05 (13) •45
POUS45R
-375
-*—
1
pGUS45
-45
1
•148
I
- * • - » -
1 -375
pousise
1
pQUSZM
1 "I-
1.51 ± 1.00(10)
-ise |
—^-
-375
3.56 ± 2.25 (14)
1 -1S3 |
•378 POUS163
2.74 ± 1.59 (14)
I
- • « •
-375 pOUS148
2.36 ± 1.76 (10)
I
-37S
0.51 ± 0.31 (15)
-234
0.26 ± 0.22 (18)
I
POUS283
0.34 ± 0.16(9)
pen 21 .s
0.32 ± 0.12 (7)
POUS45O
J
0.66 ± 0.58 (9)
Figure 12.13. Deletion analysis of the promoter of a highmolecular-mass wheat glutenin gene in transgenic tobacco. The acceptor plasmid contained CaMV 35S promoter, the CUS gene, and the 3'-end of the NOS gene. The coordinates on the glutenin gene indicate the distance in bp of the end points from the initiation site for glutenin transcription. Polarity of the promoter fragments is shown by arrows.
Data on the right indicate GUS activity (picomoles ± standard error) in the seeds of transgenic tobacco; numbers in parentheses are the number of plants used for CUS assay. Endosperm-specific expression is indicated by + or - . (From Thomas and Flavell 1990. Plant Cell 2:1171 -1180. © American Society of Plant Physiologists.)
were found essential to potentiate endosperm-specific GUS gene expression in transgenic tobacco seeds (Figure 12.13) (Thomas and Flavell 1990). A functional analysis of the 5'-flanking region of a silent high-molecular-mass glutenin gene in transgenic tobacco has led to the proposal that single base changes in the -280 bp sequence immediately upstream of the transcription start site might result in the inactivation of the promoter. Sequences required for tissue and temporal specificity of the expressed high-molecular-mass glutenin gene in transgenic tobacco are also incorporated in the -280 bp region; since this includes only two core sequences of the conserved -300 bp element, there is some question about the essentiality of the latter for tissue-specific expression of the gene (Halford et al 1989). In transient expression assays using protoplasts
the high-molecular-mass glutenin gene. This is apparently mediated by sequences found in the endosperm box of the glutenin promoter and suggests that such binding sites could act as cis-acting elements in regulating gene expression (Hilson et al 1990). The ability of Escherichia coli and yeast cells to express biologically active proteins has made possible an important advance in analyzing wheat storage protein gene expression (Bartels, Thompson and Rothstein 1985; Neill et al 1987; Galili 1989). Galili (1989) showed that the DNA sequence encoding the glutenin high-molecular-mass subunit cloned into a bacterial expression vector directs the synthesis of large amounts of the corresponding protein. Confirmation that the protein synthesized by the bacteria is identical to the native wheat protein subunit was provided by solubility characteristics. A practical consequence of this discovery is that it makes possible the analysis of the structure of the protein and the identification of its sequences that affect the quality of the dough.
of Nicotiana plumbaginifolia, mammalian transcrip-
tion factors of the leucine zipper class are found to stimulate transcriptional activity of promoters of
Developmental Biology of the Endosperm
355
moters effective in directing transcription in isolated tobacco protoplasts and in transgenic tobacco. As indicated in the previous section, three sub- Three regions of the GT-3 promoter, -945 to -726 families comprise the rice glutelin genes; each sub- bp, -346 to -263 bp, and -181 to +7 bp from the family contains five to eight copies. Several cDNA transcription start site, are required for the clones and genomic clones of glutelin genes have endosperm-specific expression and temporal regubeen isolated and characterized (Takaiwa, Kikuchi lation of chimeric genes in transgenic plants (Leisy and Oono 1986, 1987, 1989; Takaiwa et al 1987; et al 1989; Okita et al 1989; Zhao, Leisy and Okita Masumura et al 1989a; Okita et al 1989; Takaiwa 1994; Croissant-Sych and Okita 1996). A major difand Oono 1991). Genes of all three classes contain ference between glutelin genes of the GT-3 and GTthree short introns precisely at the same positions 1 classes is that in the 5'-flanking region of the in the coding regions. A comparison of genomic latter, information required for both tissue speciclones showed that the 5'-flanking regions of GT-1 ficity and temporal regulation is located in a single and GT-2 are closely related, whereas GT-3 has lit- domain (between -441 and -237 bp) (Takaiwa, tle or no homology upstream of -267 bp to either Oono and Kato 1991). A difference between the GT-1 or GT-2. The derived sequence of the protein location of the promoter elements of the GT-1 gene shows an amino terminus with features typical of a in heterologous and homologous hosts was also signal peptide. From molecular analyses of glutelin revealed when the gene was expressed in transcDNA clones, sequence homology of the primary genic rice, suggesting limitations in the recognition sequence of the gene to the US globulin gene of of critical sequences by transcription factors in the heterologous host (Zheng et al 1993). The legume embryos has become apparent. Analysis of rice prolamin cDNA clones encod- endosperm-specific expression of a rice prolamin ing a 15.2-kDa protein revealed that they contain gene in transgenic tobacco appears to be controlled relatively long 5'-untranslated sequences of about by the upstream region from -680 to -18 bp (Zhou 150-200 bp followed by coding regions of about and Fan 1993). As detected by transient assay, 450 bp. The encoded proteins have a signal peptide endosperm-specific expression of the y-kafirin gene but conspicuously lack repeated sequences (Kim is conserved in the grains of maize, sorghum, and and Okita 1988a, b). Prolamin cDNAs that encode C. lacryma-jobi and is caused by a -285 bp proximal 10-kDa and 13-kDa proteins show close similarity region of the promoter (de Freitas et al 1994). to the structure of cDNA clones that code for the An aleurone-specific gene, LTP-2, from barley 15.2-kDa protein (Masumura et al 1989b, 1990). endosperm that encodes a nonspecific lipid transfer The oat globulin gene is a member of a small protein has been discussed earlier. Kalla et al (1994) multigene family with a nucleotide sequence char- have shown that expression of this gene is under acterized by the presence of three introns and the strict temporal and spatial control in transgenic rice, lack of a -300 bp element. The protein product directing transcription exclusively in the aleurone shares structural features with the 11S globulin of layer from shortly after the onset of aleurone cell legume embryos (Walburg and Larkins 1986; differentiation to mid-maturity of the grain. At preShotwell et al 1988, 1990; Schubert et al 1990). The sent, to account for the differential expression of this nucleotide sequence of avenin, a member of a gene in the developing endosperm, it is assumed multigene family of about 25 genes, shows signifi- that the active promoter sequences contain cis-regucant homology to the gliadins of wheat and the B- latory elements that bind to nuclear proteins. hordein of barley; the coding sequence lacks Although transcripts of the 12S oat globulin gene introns, and the promoter region has a -300 bp ele- are found in both the embryo and the endosperm ment (Chesnut et al 1989; Shotwell et al 1990). of transgenic tobacco, the protein product is localGenes for prolamins of rye (y- and a)-secalin) (Kreis ized exclusively in the endosperm. This suggests et al 1985; Hull et al 1991), sorghum (y-kafirin) that posttranscriptional regulatory processes are (DeRose et al 1989; de Freitas et al 1994), and Coix involved in the expression of this gene in the hetlacryma-jobi (coixin) (Ottoboni et al 1993) are also erologous host (Manteuffel and Panitz 1993; characterized by the lack of introns but they con- Schubert et al 1994). tain the -300 bp element. Both y-kafirin and coixin As can be discerned from this account, there is a show varying degrees of homology to different considerable industry for the isolation of various subunits of zein (de Barros et al 1991; Leite et al endosperm storage protein genes and their intro1992; Ottoboni et al 1993). duction into heterologous hosts. This work is Glutelin genes of rice contain functional pro- fueled by the potential practical benefits that will (iv) Rice Glutelin and Other Genes
356
accrue from the ability of the genes to code for useful proteins in seeds of plants of otherwise limited use. As of now, the main stumbling block is a level of expression of these proteins in the heterologous hosts that is too low to be economically useful. 7. GENERAL COMMENTS
As a product of double fertilization, the endosperm has attracted the attention of plant embryologists for the variety of ways this tissue has evolved from the primary endosperm nucleus. The first few rounds of division of the endosperm nucleus follow divergent patterns in different species, but eventually the mature tissue ends up as a formless mass of cells surrounding the embryo. The conspicuous events in the formation of the endosperm are marked by stages in which cell divisions predominate first and then cell enlargement takes over. In the endosperms of many species, the cells compensate for the lack of divisions in the later part of their ontogeny by abnormal nuclear phenomena such as polyploidy and endoreduplication. In the early period of investigations of the endosperm, the importance attached to the ontogeny and cytology of the tissue tended to be overshadowed by views about the tissue's possible functions. The endosperm is considered a potent source of chemical substances that promote the growth of the early stage embryos. Although direct evidence for
Zygotic Embryogenesis
the absorption of endosperm nutrients by the developing embryo is lacking, in attempts to relate embryo growth to endosperm nutrients, the endosperm has been analyzed morphologically, biochemically, and genetically. Persuasive though the results of these analyses are, they fall short of evidence with definitive functional implications. Throughout the history of the study of endosperms, most investigators have focused on the cereal grains. From the physiological point of view, the endosperm of cereal grains supplies the nurture and nutrients for the embryo during its transformation into a seedling plant. Although the topic is outside the purview of this chapter, the biochemical ramifications of the endosperm in germinating seeds and grains have provided fundamental information about the metabolic and catabolic pathways of the major seed storage products, namely, starch, proteins, and lipids. The modern era of molecular biology and genetic engineering is beginning to bridge the gap in our knowledge of the gene expression program leading to the accumulation of storage products during the development of the endosperm. Information currently available about the structure and expression of the major storage protein genes of the endosperms of some of the economically important cereal crops will continue to have a profound influence on our future attempts to improve the quality and quantity of the food crops and to genetically engineer the endosperm storage proteins of other plants.
Chapter 13
Embryogenesis and Physiology of Growth of Embryos The Zygote (i) Volume Changes and Cell Wall Synthesis (ii) Infrastructure of the Zygote The Multicellular Embryo (i) Classification of Embryo Types (ii) Comparative Embryogenesis in Monocots and Dicots (iii) Reduced Embryo Types (iv) Cell Lineage and Cell Fate during Embryogenesis Cellular Aspects of Embryogenesis (i) Ultrastructural and Histochemical Characterization (ii) Cell Division and Cell Elongation (iii) DNA Synthesis Tissue and Meristem Differentiation during Embryogenesis (i) Shoot and Root Apical Meristems (ii) Origin of the Quiescent Center The Suspensor (i) Ultrastructure of the Suspensor Cells (ii) Nuclear Cytology of the Suspensor Cells (iii) Physiology and Biochemistry of the Suspensor Nutrition of the Embryo (i) Nutrient Supply to the Zygote (ii) Embryo-Endosperm Relations (iii) Nutrient Supply from the.Mother Plant Embryo Culture (i) Seed Embryo Culture (ii) Culture of Immature Embryos and Precocious Germination (iii) Culture of Proembryos 8. General Comments
After fertilization, the egg is transformed into the zygote, which embarks upon one of the most critical periods in its development as it is partitioned
into cells that ultimately make up the body of the embryo. What makes the zygote so unique a cell is that it is the product of fusion of two gametes, the sperm contributing the paternal genome and the egg providing the maternal counterpart. The point in time when the egg and sperm fuse together to be woven into a new sporophyte marks the beginning of the ontogeny of the species. The phase of ontogeny concerned with progressive division of the zygote to form the embryo is known as embryogenesis, or as zygotic embryogenesis (to avoid semantic confusion with somatic embryogenesis and pollen embryogenesis). Classical histological analyses of embryo development in a large number of plants have generated an invaluable guide to the division patterns of the zygote and its immediate derivatives, and contemporary studies are beginning to provide some much needed insight into the associated cellular and molecular changes. The history of the study of the development and physiology of embryos can be considered to have evolved in three distinct phases. Much of the early work was focused on the first few rounds of division of the zygote and on the subsequent morphogenesis of cells to give rise to the embryo. This early work, published in scores of scientific papers over the years, gave rise to the thesis that each division plane of the fertilized egg and its descendants is accomplished according to a genetic blueprint characteristic of the species. In later years, a large number of studies were undertaken on the structural characteristics of embryos and their relationship to morphological development and function. One aspect of these studies was a concern with the ultrastructural changes that accompany the transformation of the zygote into a full-term embryo. A third focus has been on the much debated but little 357
358
Zygotic Embryogenesis
understood area of nutrition of the embryo. also play a role in the evolution of the zygote. For Experimental approaches to gaining insight into example, in Gossypium hirsutum (Jensen 1968), the nutritional interactions during embryogenesis Hibiscus costatus, H. furcellatus (Ashley 1972), have relied on perturbations of the growth of Lagerstroemia speciosa (Lythraceae) (Raghavan and embryos that involve their isolation and culture at Philip 1982), and Brassica campestris (Sumner 1992), there is a dramatic decrease in the size of the fertilvarious stages of development. The account in this chapter will begin with ized egg to about half the size it was before fertiloverviews of the zygote and various aspects of its ization; the zygote of Nicotiana tabacum also division that generates the embryo and the subse- experiences a shrinkage, but it is less striking quent tissue and meristem differentiation patterns in (Mogensen and Suthar 1979). In some plants the the embryo. Part of the account will also be con- egg actually enlarges after fertilization as it is cerned with the cell biology and cytology of transformed into the zygote. This is the case in embryos as they complete their genetically permissi- Datura stramonium (Satina and Rietsema 1959), ble development. Later sections will deal with the Cypripedium insigne (Poddubnaya-Arnoldi 1967), physiology of embryogenesis, emphasizing the Jasione montana (Erdelska and Klasova 1978), Oryza nutrition of embryos of different ages. Three reviews sativa (Jones and Rost 1989b), Hordeum vulgare (Raghavan 1991, 1993; Raghavan and Sharma 1995) (Engell 1989), Monotropa uniflora (Olson 1991), and have discussed the topics covered here from slightly Arabidopsis (Mansfield and Briarty 1991). In different perspectives and the reader is referred to Arabidopsis, a change in the distribution of the microtubules, from a random orientation to a prethem for additional information. dominantly transverse orientation in a subapical band, coincides with the growth in length of the zygote (Webb and Gunning 1991). 1. THE ZYGOTE Within the zygote lies the ability to form an entire plant, a feat that the zygote accomplishes by extensive changes in form in defined and dramatic ways and by a progressive change from the undifferentiated to the differentiated state. Thus, the zygote can truly be considered as the unicellular stage of a multicellular organism, possessing all essential properties of interaction of the adult. Common to the changes that occur during the progressive transformation of the zygote into an embryo are such processes as cell division, cell expansion, cell maturation, cell differentiation, and formation of meristems. In a very young embryo, all cells are involved in producing a new generation of daughter cells. As development continues, cell divisions, which lead to the formation of specialized tissues and organs, are segregated to certain parts of the embryo predictable by their position in the cell lineage. Very little is known about the mechanisms that bring about the restriction of functional activities in the developing embryo. (i) Volume Changes and Cell Wall Synthesis
Once fertilization has been accomplished, many properties of the egg change during the first few hours of its life as the zygote. Even though the postfertilization changes occur in the cytoplasm, physical changes occurring outside the cytoplasm
Special attention has been focused on the chalazal part of the egg, which in many species, as seen in Chapter 6, lacks a cell wall in the unfertilized state. The laying down and the consolidation of a new wall at the chalazal end are important milestones in the evolution of the zygote in G. hirsutum (Jensen 1968), Capsella bursa-pastoris (Schulz and Jensen, 1968c), Quercus gambelii (Singh and Mogensen 1975), N. tabacum (Mogensen and Suthar 1979), Glycine max (Folsom and Cass 1992), Helianthus annuus (Yan, Yang and Jensen 1991), and B. campestris (Sumner 1992); increased activity of dictyosome vesicles and cortical microtubules coincident with cell wall regeneration seems to validate these observations (Jensen 1968; Schulz and Jensen 1968c; Yan, Yang and Jensen 1991b; Sumner 1992). In several species of Rhododendron and in Ledum groenlandicum (Ericaceae), a callose wall is laid down around the zygote during the first two days after fertilization (Williams et al 1984). The fact that the wall essentially insulates the newly formed zygote from the influence of cells of a different genotype in the immediate neighborhood probably has some significance in the subsequent induction of the sporophytic divisions. (ii) Ultrastructure of the Zygote
No generalizations can be made about the ultrastructural upheavals that occur in the egg after fertilization except that the simplicity of the zygote as
Embryogenesis and Physiology of Growth of Embryos
359
revealed in the electron microscope almost betrays size and number of vacuoles increase markedly its potential to form a whole new plant. In Gossypium toward the micropylar end of the zygote (Figure hirsutum, polarity is already established in the egg 13.1). With minor variations, these changes are conby the presence of a large micropylar vacuole. The sistently found to occur in the fertilized egg cells of position of the vacuole remains unchanged as the Capsella bursa-pastoris (Schulz and Jensen 1968c), egg is transformed into the zygote. However, cyto- Linum usitatissimum (D'Alascio-Deschamps 1972), plasmic organelles like plastids, ER, mitochondria, and jasione montana (Erdelska and Klasova 1978). In and ribosomes begin to float in the egg within hours Papaver nudicaule and Zea mays, whose unfertilized after fertilization and take up new positions at the egg cells have a micropylar nucleus and a large chachalazal end, where they form an investment lazal vacuole, fertilization is accompanied by a around the nucleus. This shift in organelle positions reversal of cytoplasmic polarity as the zygote accentuates the polarity already present in the unfer- acquires a chalazal nucleus and a micropylar vactilized egg. Observations of the egg at different times uole (Olson and Cass 1981; van Lammeren 1986). after fertilization leave no doubt that the increase in Other ultrastructural observations, such as the total amount of ER as well as in the tube-con- increase in the number and complexity of the taining ER, the increase in size and number of starch organelles, indicate that the zygote is a metaboligrains in the plastids, the elaboration of additional cally active cell. The standard changes suggestive of cristae in the mitochondria, the generation of new altered cell functions observed include alterations ribosomes, and the formation of polysomes from the in the form of the ER, a marked increase in the numexisting ribosomes, are all part of the basic response ber of starch grains in the plastids, the elaboration of the egg to fertilization (Jensen 1968). In of additional cristae in the mitochondria, the Arabidopsis, the large micropylar vacuole of the egg appearance of a new generation of ribosomes and undergoes a transient reorganization upon fertiliza- their transformation into polysomes, and increased tion as it is replaced by numerous, small, vesiclelike dictyosome activity; these changes are described in vacuoles. This is followed by the elongation of the great detail in the zygotes of Z. mays (Diboll 1968; zygote, a decrease in the number of vacuoles, and van Lammeren 1986), C. bursa-pastoris (Schulz and the eventual formation of an extremely large vac- Jensen 1968c), Epidendrum scutella (Cocucci and uole occupying approximately two-thirds of the Jensen 1969c), Linum usitatissimum (Deschamps total length of the cell at the micropylar end. Nuclear 1969), L. catharticum (D'Alascio-Deschamps 1981), migration from the basal part of the zygote to the Q. gambelii (Singh and Mogensen 1975), H. annuus chalazal end completes the program of organelle (Yan, Yang and Jensen 1991b), Arabidopsis redistribution that results in a distinct apico-basal (Mansfield and Briarty 1991), and Brassica campestris polarity (Mansfield and Briarty 1991). The target of (Sumner 1992). In the eggs of Petunia hybrida fertilization in Glycine max also appears to be the (van Went 1970c), L. usitatissimum (D'Alasciolarge micropylar vacuole of the egg, which is frag- Deschamps 1972), and Q. gambelii (Mogensen 1972), mented into a bunch of small vacuoles distributed an increase in the number of lipid bodies after ferthroughout the zygote; additionally, there is an tilization is a feature not observed in other plants. increase in the number and complexity of the amy- The presence of starch-containing plastids and lipid loplasts in the zygote (Folsom and Cass 1992). The bodies is considered an adaptive adjustment of the polar distribution of cytoplasmic organelles of zygote to budget energy-related materials needed mature egg cells is also maintained unchanged in during its early division. The abundance of increasthe zygotes of Petunia hybrida (van Went 1970c; ingly complex mitochondria and the large quantity Vallade 1980), Quercus gambelii (Singh and of ribosomes might indicate a high metabolic activMogensen 1975), Monotropa uniflora (Olson 1991), M.ity of the egg upon fertilization. That fertilization hypopitys (Olson 1993), and Helianthus annuus (Yan, signals the production of mRNAs and their engageYang and Jensen 1991b). In plants in which the egg ment by ribosomes, is reflected in the abundance of cell does not possess an incipient polarity, present polysomes in the zygote. Consistent with this obserindications are that the zygote displays a polarized vation, it has been found that in C. bursa-pastoris distribution of organelles, most strikingly of the ER, (Schulz and Jensen 1968c) and Nicotiana rustica ribosomes, plastids, and mitochondria, which clus- (Sehgal and Gifford 1979), fertilization is accompater around the nucleus at the chalazal end. nied by an increase in the stainable RNA content of Accentuating the polarity of the zygote is a gradient the egg; in situ hybridization of ovules of C. bursaof decreasing density of plastids and mitochondria pastoris with 3H-poly(U) has revealed not only an toward the micropylar end. Conversely, both the increase in the mRNA content of the zygote but
Figure 13.1. Ultrastructural profile of zygotes. (a) The zygote (z) of Quercus gambelii. ds, degenerating synergid; f, filiform apparatus; I, lipid bodies; n, nucleus; ps, persistent synergid. Scale bar = 10 urn. (From Mogensen 1972; photograph supplied by Dr. H. L. Mogensen.) (b) The zygote of Zea mays, en, endosperm; m, mitochondria; n, nucleus; s, starch granule. Scale bar = 10 um (From van Lammeren 1986; photograph supplied by Dr. A. A. M. van Lammeren.)
360
Embryogenesis and Physiology of Growth of Embryos
also a gradient in its accumulation, with a high concentration in the chalazal part (Raghavan 1990b). Eggs of other plants, such as N. tabacum (Mogensen and Suthar 1979) and Triticum aestivum (You and Jensen 1985), show variable ultrastructural details but display little change after fertilization. In Z. mays, postfertilization increases in the activity of oxidative enzymes and the accumulation of reserve substances bypass the zygote and are generally confined to the nucellus and integuments (Singh and Malik 1976). It is clear that several concomitant changes coordinated by many assembly lines must be taking place during the formation of the zygote, but they differ among different species. One consequence of these changes is that the zygote essentially becomes a repository of developmental information that can play a role in morphogenesis. To a large extent, this is reflected in the subsequent division of the zygote to form the embryo. 2. THE MULTICELLULAR EMBRYO
Changes in the ultrastrucrure of the zygote are preparatory to its first division. However, this division might not occur immediately after fertilization or immediately after the reorganization of the cytoplasm, since the zygote remains quiescent for a period of time, ranging from a few hours to several months, before it divides. At the time of its division, the zygote appears as a highly polarized structure. Polarity is attributed to two factors. First, the orientation of the zygote is polar in the sense that its one end is cemented to the embryo sac wall at the micropylar end, and the other end projects into the central cell facing the chalazal end. Second, polarity is also bestowed on the zygote by its ultrastructural profile previously described. The division of the zygote is the next milestone in the establishment of polarity. This division is almost invariably asymmetric and transverse, cutting off a large, vacuolate, basal cell toward the micropylar end and a small, densely cytoplasmic, organellerich terminal cell. In Arabidopsis, the occurrence of zygote division is presaged by the presence of a preprophase band of microtubules corresponding to the position of the future cell plate (Webb and Gunning 1991). In mutants that are unable to form cortical microtubules, cell divisions are random due to irregular cell expansion and the failure of division plane alignment. A line of reasoning based on the observation that mutant embryos differentiate normally has led to the suggestion that the basic differentiation pattern in the developing embryo
361
does not depend upon the ordered arrangement of the cytoskeletal elements (Traas et al 1995). A mathematical analysis of the relative areas of the apical and basal cells seems to contradict the conventional view of the universality of the size relationships of these cells, since in a significant number of cases, the apical cell may be large and the basal cell small, or both cells may be of the same size (Sivaramakrishna 1978). A consistent, unequal division is morphogenetically significant, because it points to some underlying cytoplasmic program that determines the position of the mitotic spindle and the orientation of the cell wall. Ultrastructural and histochemical studies have shown that the marked degree of asymmetry displayed by the two-celled embryo is complemented by subtle variations in the distribution of organelles and in the concentration of macromolecules. For instance, in Gossypium hirsutum (Jensen 1963), Capsella bursa-pastoris (Schulz and Jensen 1968c), and Quercus gambelii (Singh and Mogensen 1975), the terminal cell of the two-celled embryo is similar to the chalazal part of the zygote in that it has a dense cytoplasm enriched with organelles, whereas the basal cell is relatively impoverished of organelles. In the two-celled embryo of G. hirsutum, the terminal cell is also characterized by the presence of a large number of plastids and large and variably shaped mitochondria, whereas the basal cell has a preponderance of vacuoles of various sizes and shapes (Jensen 1963). In G. hirsutum (Jensen 1963) and Vanda sp. (Alvarez and Sagawa 1965b), a functional differentiation between the terminal and basal cells is signaled by a higher concentration of RNA in the former. By contrast, in Stellaria media (Pritchard 1964b), C. bursa-pastoris
(Schulz and Jensen 1968c), and Limnophyton obtusifolium (Alismaceae) (Shah and Pandey 1978), RNA concentration is higher in the basal cell than in the terminal cell. Apparently an increased concentration of macromolecules does not infallibly predict in all cases the fate of the cell in a particular morphogenetic pathway. (i) Classification of Embryo Types
Partitioning of the terminal and basal cells leads to the formation of the mature embryo and, for this reason, the plane of division of one or both these cells and the fate of the daughter cells formed have an important bearing on the classification of embryo types in angiosperms. Generally, the organogenetic part of the embryo is derived from the terminal cell with little or no contribution from
362 the basal cell; during embryo formation, the first division of the terminal cell may be transverse, longitudinal, or, rarely, oblique. These facts have served as a basis for classifying different types of embryos in angiosperms. A widely used system of classification has identified two major groups, one in which the division of the terminal cell of the twocelled embryo is longitudinal, and the other in which the division wall is oriented in a transverse plane (Johansen 1950). Within these major groups, different embryo segmentation types are recognized and identified by the name of the family in which many examples of the type are found. However, it should be noted that in some families more than one type of embryo development, represented by one or two genera in each case, has also been observed. Accounts of the ontogeny of representatives of each segmentation type are given by Johansen (1950) and Maheshwari (1950), and so in this account ontogenetic information is passed over. In describing embryogenesis, the term "proembryo" is used to designate the early stages of development of the embryo; the initiation of cotyledons is considered as a good cut-off point for the end of the proembryo stage.
Zygotic Embryogenesis cotyledons and plumule, and the latter to the root meristem, root cap, and hypocotyl; a small suspensor is also derived from the basal cell (Jones 1927). The Asterad type of embryo development or its variations have been described in members of Asteraceae, Geraniaceae, Lamiaceae, Oxalidaceae, Polygonaceae, Rosaceae, and Urticaceae (dicots), and Liliaceae and Poaceae (monocots) (Natesh and Rau 1984).
Transverse Division of the Terminal Cell. Three embryo developmental patterns (Solanad, Caryophyllad, and Chenopodiad) are recognized on the basis of the transverse division of the terminal cell and the participation of the division products of the terminal and basal cells of the two-celled proembryo in the formation of the mature embryo. The Solanad type, found in several species of Nicotiana, is characterized by the formation of the embryo from the terminal cell and by division of the basal cell to form a suspensor. Besides members of Solanaceae, members of Hydnoraceae, Linaceae, Papaveraceae, and Rubiaceae display the Solanad type of embryo formation. In the Caryophyllad type, the basal cell does not divide after it is cut off, and the organogenetic part of the embryo is derived exclusively from the division of the terminal cell. These characLongitudinal Division of the Terminal Cell. teristics are clearly evident during the embryogeneDepending upon the extent of the contribution of sis of Sagina procumbens (Caryophyllaceae) (Soueges the basal cell of the two-celled proembryo to the 1924). A suspensor is not a regular feature of the formation of the organogenetic part of the mature Caryophyllad type of embryos; if one is present, it is embryo, Crucifer (or Onagrad) and Asterad types also derived from the terminal cell with contribuof embryo development have been recognized in tions from the basal cell, as seen in Saxifraga granulata the group exhibiting longitudinal division of the (Saxifragaceae) (Soueges 1936). The Caryophyllad terminal cell. In the Crucifer type, exemplified by type of embryo development has been described in Capsella bursa-pastoris and Arabidopsis, the basal cell plants included in Crassulaceae, Droseraceae, contributes very few derivatives to the formation Fumariaceae, Holoragaceae, Portulacaceae, and of the organogenetic part of the embryo, which is Pyrolaceae (dicots), and Alismataceae, Araceae, thus generated almost entirely by the terminal cell. Ruppiaceae, and Zannuchelliaceae (monocots). The cells derived from the basal cell form a shortAnother addition to the list is Monotropa uniflora, in lived, filamentous structure known as the suspenwhich the basal cell collapses after it is cut off (Olson sor; more will be said about the suspensor in a later 1991). The Chenopodiad type has the distinction of section of this chapter. This type of embryogeny having the division products of both the terminal has also been described in plants belonging to and basal cells integrated into the organogenetic part Onagraceae, Bignoniaceae, Lamiaceae, Lythraceae, of the embryo. Generally, the hypocotyl of the Papilionaceae, Ranunculaceae, Rutaceae, and mature embryo has its origin in the basal cell, which Scrophulariaceae among the dicotyledons, and in also forms a short suspensor. In addition to being Juncaceae and Liliaceae among the monocotylefound in members of Chenopodiaceae, this type of dons. In contrast, the embryo in the Asterad type is embryo development is typical of plants included in generated by the division of the terminal and basal Amaranthaceae and Polymoniaceae (Natesh and cells of the two-celled proembryo. The description Rau 1984). of embryogenesis in Lactuca sativa reveals a nice division of labor between the derivatives of the terA type of embryo development in which the minal cell and those of the basal cell in the forma- first division of the zygote itself is oblique or lontion of the embryo, the former giving rise to the gitudinal (Piperad type) was slow to be recog-
Embryogenesis and Physiology of Growth of Embryos
363
nized and was thought to be found only in a few generally transcend the boundaries of dicots and families of dicotyledons, such as Piperaceae, monocots. Balanophoraceae, Dipsacaceae, and Loranthaceae. To understand how the zygote progresses from a The genera and species assigned to this type are not single cell to a complex multicellular embryo in many, and only occasionally has it been identified which fundamental tissues and organs are demarin monocots (Tohda 1974; Natesh and Rau 1984). cated, embryogenesis of representatives of dicots We see that descendants of the basal cell of the and monocots will be considered. Any one of the two-celled proembryo defy a neat packaging, many dicot and monocot embryos that have been because in some cases, part of this cell population studied over the years will illustrate the general becomes enmeshed in the organogenetic part of the principles of embryogenesis easily, but some plants embryo, the rest forming a suspensor, whereas in have been favorites for this purpose. Embryogenesis other cases, the basal cell population wholly forms in Capsella bursa-pastoris is almost a textbook examthe suspensor or simply wilts away. The general ple of the Crucifer type of embryogenesis in which conclusion from these observations is that although the first division of the terminal cell of the two-celled the terminal cell is already determined as the prog- proembryo is longitudinal (Figure 13.2). Each of the enitor of the embryo, the determinative state of the two cells thus formed again divides by longitudinal walls to yield a quadrant, followed by an octant basal cell is at best described as being labile. Embryo development in Paeonia anomala, P. stage by division of the four cells by transverse moutan, and P. wittmanniana, first described by walls. Although the embryo is populated at this Yakovlev and Yoffe (1957), does not fit into any of stage by a relatively small number of homogeneous the types just described, because the failure of wall cells, there are functional differences between the formation during the first few rounds of division of upper and lower tiers of cells. For example, the stem the zygote results in embryogenesis that starts with tip and cotyledons in the adult embryo are believed numerous free nuclei suspended in a cellular bag. to be formed from the upper tier of cells, and the Later, the nuclei migrate toward the periphery of hypocotyl generated from the lower tier. In situ the cell, where wall formation is initiated to form a hybridization of embryos of Brassica napus with multicellular proembryo. A further remarkable cloned genes for the storage proteins, cruciferin and change occurs as the marginal cells of the proem- napin, has shown that the determinative event bryo differentiate as embryo primordia. Although established by the transverse division of the quadseveral primordia are born out of a proembryonal rant is reflected in the gene expression patterns in cluster, only one outlives the others and differenti- later stage embryos. In these embryos, cells derived ates into a viable embryo. Independent investiga- from the lower tier accumulate high levels of trantions by other workers (Cave, Arnott and Cook scripts whereas, except in the cotyledons, cells 1961; Carniel 1967a) on additional species have derived from the upper tier are devoid of transcripts essentially confirmed the unusual features of (Fernandez, Turner and Crouch 1991). The point is embryogenesis in this genus. A free nuclear phase important because the pattern of transcript accumuis typical of embryogenesis in gymnosperms, and lation does not appear to reflect tissue or cell types its occurrence in Paeonia appears to be the but follows the boundary set up by an earlier division. However, the developmental fate of a group of angiosperm equivalent of a gymnosperm feature. three or four layers of cells derived from the upper tier of the quadrant (designated as the epiphysis) and of the root apical meristem derived from the ter(ii) Comparative Embryogenesis in minal cell of the suspensor (designated as the Monocots and Dicots hypophysis) is expressed by their failure to accumuThe essence of embryogenesis is the establish- late storage protein transcripts. It appears that gene ment of tissue systems and primordial organs in expression patterns in these groups of cells are set, the embryo. Obviously, a precise sequence of divi- not by the division boundaries, but by their biosions of the zygote and proembryo, as well as a chemical and physiological characteristics. high degree of inductive interactions between the growing parts, bears heavily on the final form The next round of division of cells of the octant attained by the embryo. It is important to note that embryo is periclinal and results in the formation of although mature embryos of dicots and monocots a globular embryo of 16 cells. That some covert appear morphologically different, their early divi- changes may foreshadow the differential potencies sion sequences are basically similar. It is, therefore, of cells of the embryo at this early stage is shown by expected that the modes of embryo development localization of transcripts of a Kunitz trypsin
364
Zygotic Embryogenesis
f Figure 13.2. Stages in the division of the zygote and early development of the embryo and suspensor in Capsella bursa-pastoris. (a) Zygote. (b) Asymmetric division of the zygote. (c) Transverse division of the basal cell and longitudinal division of the terminal cell, (d) Quadrant stage
embryo, (e) Octant stage embryo, (f) Early globular embryo, (g-i) Three stages of the globular embryo, showing the division of the hypophysis (h). Suspensor is outlined in the square bracket. The micropylar end is toward the bottom of the page. (From Soueges 1919.)
inhibitor gene {KTI-3) in the cells at the micropylar of Triticum aestivum in the presence of auxin and pole of globular embryos of soybean (Perez-Grau inhibitors of auxin transport have supported a role and Goldberg 1989). The expression of the KTI-3 for auxin in the establishment of bilateral symmetry gene is one of the earliest indicators of polarity in in monocot embryos (Fischer and Neuhaus 1996). the globular embryo. The globular embryo is The role of polar auxin transport in the maintephased into the heart-shaped stage when it attains nance of embryonic form was the focus of reports bilateral symmetry due to a lateral expansion of the using the more easily accessible somatic embryos distal pole that gives rise to a pair of cotyledons. of carrot. Not only do polar auxin transport The elements of the future shoot apex, which inhibitors block transitions in embryo development remains essentially in a dormant state throughout beginning as early as the globular stage, they also the life of the embryo, are carved out from a few cause the formation of giant globular and oblong cells lying between the cotyledons. Based on the embryos (Schiavone and Cooke 1987). action of inhibitors of auxin transport on cultured Microsurgical experiments on heart-shaped and embryos of B. juncea resulting in the formation of torpedo-shaped somatic embryos have shed furcotyledons as a fused collarlike structure around ther light on the role of polar auxin supply and disthe apex, rather than as two discrete lateral out- tribution by showing that the basal root half of the growths, a model proposed to account for the embryo grows more rapidly when cultured alone attainment of bilateral symmetry in embryos grow- than when it is grafted back to the apical shoot half ing in vivo has invoked a requirement for polar and cultured. However, when grafted somatic auxin transport (Liu, Xu and Chua 1993b). Indeed, a embryos are treated with a polar auxin transport small percentage of embryos of the pin-formed inhibitor, the basal halves behave as if they have flower mutant (pin-1.1) of Arabidopsis, which has a been cultured separately. Not surprisingly, treatdefect in polar auxin transport, strikingly resemble ment of isolated basal sections with IAA forced treated embryos of B. juncea (Okada et al 1991). The them to grow as if they were part of a graft union patterns of growth displayed by isolated embryos with the apical part (Schiavone 1988). These obser-
Embryogenesis and Physiology of Growth of Embryos
Figure T3.3. Late stages in the development of the embryo and suspensor of Capsella bursa-pastoris. (a) Heart-shaped embryo, (b) Torpedo-shaped embryo, (c) Walkingstick-shaped embryo, (d) Mature embryo, c, cotyledon; ra, root apex; s, shoot apex. Suspensor is outlined in the square bracket. The micropylar end is toward the bottom of the page. (From Schaffner 1906.)
vations also attest to a remarkable dependence of the embryonic root, rather than the shoot, on polar auxin transport. After the cotyledons are initiated, division and differentiation of cells in the basal tier of the embryo give rise to the hypocotyl. There is considerable elongation of the cotyledons and hypocotyl at this stage, which gives rise to the torpedoshaped embryo. A root apex delimited in the torpedo-shaped embryo is the progenitor of the embryonic radicle (Figure 13.3). During progressive embryogenesis, continued growth of the embryonic organs, compounded by spatial restrictions within the ovule, causes the embryo to curve at the tip to assume the walking-stick-shaped stage
365
and finally to take on the shape of a horseshoe (Schaffner 1906; Soueges 1919). Although organs characteristic of the adult embryo differentiate as early as the heart-shaped or torpedo-shaped stage, subtle physiological and biochemical changes continue in the cells of the embryo and in the surrounding tissues before the maturation program sets in. It was discussed earlier that a hallmark of the Crucifer type of embryo development is the ambiguity about the role of the descendants of the basal cell of the two-celled proembryo. That some of these cells are ultimately incorporated into the embryo of C. bursa-pastoris has been established by careful cell lineage studies (Schaffner 1906; Soueges 1919). Initially, the basal cell of the two-celled proembryo is partitioned transversely. Further divisions in this newly formed pair of cells are,
366
Figure 13.4. Stages in the development of the embryo of Triticum aestivum. (a) Division of the zygote. (b) Division of the terminal and basal cells, (c) Globular embryo (d) Club-shaped embryo, (e) Formation of the scutellum and coleoptile. (f) Diagrammatic section of the mature embryo. The micropylar end is toward the bottom of the page, cl, coleoptile; co, coleorhiza; ep, epiblast; r, root; s, suspensor; sa, shoot apex; sc, scutellum; v, first leaf of the plumule. (From Batygina 1969.)
however, restricted to the terminal cell, which produces, by transverse walls, 8-10 cells held together as a filament constituting the suspensor. The suspensor is terminated at the chalazal end by the organogenetic part of the embryo, and at the micropylar end by the basal cell, which has now become disproportionately large (Figures 13.2, 13.3). All the cells of the suspensor, except one, are eventually crushed and obliterated by the growing embryo; this fortunate cell (hypophysis) is in contact with the embryo and plays an important role in the formation of the root apex. The hypophysis divides transversely to form two cells, and the descendants of these cells contribute to the formation of the root cortex, the root cap, and the root epidermis of the mature embryo. It is as if a single cell from a group of about 10 cells born out of simple mitotic divisions of a mother cell is directed into a specialized channel of continued differentiation, while others perish along the way. Practically nothing is known about the mechanisms that cause changes in the developmental options of the cells of the suspensor. In their morphological appearance, mature
Zygotic Embryogenesis
embryos of monocots are strikingly different from those of dicots because of the presence of only one cotyledon. Studies on the comparative ontogeny of monocot and dicot embryos were bedeviled by controversy for some time, but there is now general agreement that the development of the embryo up to the octant stage is almost identical in dicots and monocots. The main point of contention is with regard to the origin of the shoot apex and cotyledon in the monocots. For a period of time, the dominant view was that in monocots the shoot apex is lateral, and the cotyledon is terminal, in origin. Extensive ontogenetic studies (Swamy and Lakshmanan 1962a, b; Ba et al 1978; Swamy 1979; Guignard 1984), however, seem to support the view that both the shoot apex and the cotyledon share a common origin from the terminal cell of a three-celled proembryo. Accordingly, what is perceived as a lateral shoot apex is one that is displaced from its terminal position by the aggressive growth of the single cotyledon. Sagittaria sagittaefolia (Alismaceae) exhibits a pattern of embryogenesis that is typical of many monocots. Here the differentiative potencies of the cells are established after the first transverse division of the zygote. The terminal cell gives rise to the cotyledon and shoot apex, whereas the hypocotyl-shoot axis and the entire suspensor are derived from the
Embryogenesis and Physiology of Growth of Embryos
basal cell. The cells destined to form the shoot apex are initially sluggish, so that the rapidly growing cotyledon dwarfs the shoot apex. As the hypocotyl is organized in the lower part of the proembryo, there is a concomitant differentiation of the procambium and formation of root initials (Swamy 1980). In their mature structure and ontogeny, embryos of Poaceae do not have much in common with other monocots. The main features of the grass embryo are the development of an absorptive organ known as the scutellum; the presence of caplike structures covering the radicle and plumule, known respectively as the coleorhiza and the coleoptile; and a flaplike outgrowth, called the epiblast, on one side of the coleorhiza. Embryo development in Triticum aestivum will be considered as an example (Figure 13.4). Here, following a transverse-oblique division of the zygote, both cells divide by oblique walls to form a four-celled proembryo. Repeated divisions of these cells yield the mature embryo. Following a globular stage of 16-32 cells, the embryo becomes club shaped. It is at this stage that the scutellum appears as a vague elevation in the apical-lateral region of the embryo. At the same time, the opposite side of the embryo begins to enlarge, marking the differentiation of the shoot apex and leaf primordia. The final major event of embryogenesis is the differentiation of the radicle, which has an endogenous origin in the central zone of the embryo (Batygina 1969). In some other members of Poaceae, such as Zea mays, the pattern of cell divisions in the proembryo becomes unpredictable even as early as the second round (Randolph 1936). The use of the transcription factor-encoding homeobox gene KNOTTED-1 (KN-1) has provided a promising experimental design to show that the first histologically recognizable sign of the shoot meristem in developing Z. mays embryos is foreshadowed by the expression of this marker gene. The scutellum, which is formed almost simultaneously with the shoot apical meristem, does not express the KN-1 gene. Since this gene is known to be active throughout the postembryonic shoot ontogeny of maize, its transcription probably connotes a consistent feature of the shoot meristem from the time it is cut off in the embryo (Smith, Jackson and Hake 1995). A different result is obtained when the expression of a rice homeobox gene is followed in developing rice embryos. The gene is expressed in wild-type globular embryos well before organs are initiated, as well as in mutant embryos that lack detectable embryonic organs. On the face of it, this observation does not
367
reveal that homeodomain proteins directly determine shoot development during embryogenesis (Sato et al 1996). (iii) Reduced Embryo Types It is clear that embryogenesis in angiosperms is a closely orchestrated process involving a great many cells whose activities are interlaced. As a result, the mature embryo attains the basic organization consisting of a bipolar axis with one or two cotyledons. The embryo is terminated at each pole by an apical meristem, and the point of attachment of the cotyledons separates the embryonic axis into a hypocoryl-root region and an epicotyl-plumule region. The former region has, at its lower end, the primordial root (radicle), whereas the primordial shoot (plumule) is attached to the stem part called the epicotyl. From the earliest investigations of embryogenesis, it has become apparent that embryo development does not proceed to completion in certain angiosperms. These plants thus shed seeds that harbor underdeveloped or reduced embryos (Figure 13.5). Members of Scrophulariaceae, Balanophoraceae, Ranunculaceae, Orobanchaceae, and Pyrolaceae among dicots, and Orchidaceae, Burmanniaceae, and Pandanaceae among monocots, are noteworthy in this respect, as are additional species belonging to at least 13 more families (Rangaswamy 1967; Natesh and Rau 1984). A paradigm of a seed with a reduced embryo is Monotropa uniflora. Here the embryo is generally embedded in the endosperm and consists of no more than two cells separated by a transverse wall (Olson 1980, 1991). Equally striking for its reduced embryo is the seed of Burmannia pusilla (Burmanniaceae), in which embryogenesis terminates at the quadrant stage of the apical cell (Arekal and Ramaswamy 1973). A higher level of embryo differentiation is found in the seeds of various orchids (Arditti 1979) and of root parasites like Orobanche aegyptiaca and Cistanche tubulosa (both of Orobanchaceae) (Rangaswamy 1967); at the time of shedding, seeds harbor globular embryos that do not show any differentiation into radicle, hypocotyl, cotyledons, epicotyl, and plumule. Mature seeds of AlectA36S ra
vogelii
and
Striga
gesnerioides
(both
of
Scrophulariaceae) contain heart-shaped embryos with rudimentary cotyledons and a radicular pole (Okonkwo and Raghavan 1982). Despite the lack of morphological differentiation, it has become apparent that undifferentiated embryos of certain plants display subtle differences between shoot and root poles. For example, in both O. aegyptiaca and C.
368
Zygotic Embryogenesis
sc
Figure 13.5. Seeds with reduced embryos, (a) Longitudinal section of the seed of Monotropa uniflora. E, embryo; En, endosperm; H, haustorium. Scale bar = 10 urn. (From Olson 1980.) (b) Longitudinal section of the seed of Burmannia pusilla. e, embryo; em, endosperm; np, nucellar pad. Scale bar = 10 um. (From Arekal and Ramaswamy
1973.) (c) Longitudinal section of the embryo of Spathoglottis plicata. d, distal end; p, proximal end. Scale bar = 50 |jm. (From Raghavan and Goh 1994.) (d) Longitudinal section of the seed of Striga gesnerioides. em, embryo; en, endosperm; re, root cap; sc, seed coat. Scale bar = 10 um. (From Okonkwo and Raghavan 1982.)
tubulosa, cells at one end of the embryo, identified as tigation; by identifying when and where embryogethe radicular pole, are small, whereas those at the nesis is blocked, one could determine the causal facopposite, plumular pole are large and vacuolate tors involved in the process. (Rangaswamy 1967). Although embryos of Eriocaulon robusto-brownianum and E. xeranthemum
(iv) Cell Lineage and Cell Fate during (Eriocaulaceae) are typically considered as reduced Embryogenesis types, they show some differential cell division activity resulting in a relatively quiescent region Because of the precise regularity of the cleavage and an actively dividing sector. The former region patterns during embryogenesis that lead to the represents the locus of differentiation of the epi- establishment of histologically recognizable tiers of cotyl; the cotyledons arise from the latter region. cells, there are persuasive arguments for the view The embryos, however, do not differentiate a radi- that cell lineages have highly specific developmental cle or a hypocotyl (Ramaswamy, Swamy and fates. This has no doubt formed the basis of the clasGovindappa 1981; Ramaswamy and Arekal 1982). sification of embryo types wherein embryogenesis is In the embryos of the vast majority of orchids, the considered to adhere to certain laws concerned with proximal, chalazal region is occupied by a group of the origin of cells at a given stage, the number of cells slightly smaller in size than those at the distal, cells produced in each generation, the role of cells micropylar region. As shown in Vanda sp. (Alvarez formed, and the determinative fate of the early and Sagawa 1965b) and Spathoglottis plicata (both formed cells (Crete 1963). However, because of the Orchidaceae) (Raghavan and Goh 1994), total RNA difficulty of establishing the precise location of cells and poly(A)-RNA, respectively, are concentrated in descended from the terminal or basal cells of the equal density in all cells of the embryo; however, proembryo beyond a few rounds of divisions, it has the lack of biochemical differentiation between the not been determined whether the early regularity of chalazal and micropylar parts does not preclude the the cell division pattern is continued late in embryocells at the chalazal region from initiating divisions genesis. Consequently, conclusions about cell linto form the shoot apical meristem, whereas the cells eages during embryogenesis based on studies thus of the micropylar part continue to enlarge. These far reviewed reflect more what can be followed than examples serve to illustrate the occurrence of devel- what should be followed. opmental blocks at programmed stages of embryoA way out of this dilemma is to screen chimeric genesis. What causes the arrest of embryo seedlings endowed with genetic markers such as development in these plants requires further invesanthocyanin or chlorophyll deficiency and to follow
Embryogenesis and Physiology of Growth of Embryos
the distribution of groups of marker cells. From an analysis of sectorial chimeras of Gossypium hirsutum, Christianson (1986) concluded that the cotyledons and the first and second seedling leaves, rather than arising from the embryonic apex, are formed from cellular compartments earmarked in the globular stage embryo. Compartments thus identified are two sets of eight cells for the cotyledons, two cells for the first leaf, and a single cell for the second leaf. In another study on the role of cell lineage during embryogenesis, embryos of Zea mays were screened for the production of somatic sectors from genetic mosaics generated by acute X-irradiation of ears heterozygous for cell marker mutations at different times after pollination and fertilization. Several conclusions from this work on the developmental fate of cells of the embryo and on the origin of the embryonic shoot apex deserve mention here. The occurrence of sectors in more than one-half of the scutellum following irradiations carried out after the first longitudinal division of the terminal cell of the two-celled proembryo provided evidence that the products of the first division do not contribute equally to the growth of the embryo. Based on the observation that irradiation after the first division produces mainly whole plant sectors, it may be valid to argue that the shoot meristem is produced from the progeny of only one of the cells of this division. In many mutants, sectors were observed in both the embryo axis and the scutellum, indicating that cell lineages of these two organs do not become separate until some time after the first or second longitudinal division of the two-celled proembryo. The distribution of sectors following germination led to the view that primordial cells of the shoot meristem delineated peripherally during the early stage of embryogenesis are destined to produce the first few nodes of the adult plant, whereas the later nodes are descended from the more internal cells. That the determination of the shoot meristem is a gradual process that proceeds acropetally is evident from the fact that the number of bifurcated shoots (twin shoots) produced by irradiation declines from the multicellular proembryo stage (8-10 days after pollination) and, as later-stage embryos are irradiated, there is an apical progression in the level at which splitting occurs. Later-stage embryos also become progressively insensitive to irradiation, indicating that the shoot system is fully established in the proembryo. The orientation of the sectors from one side of the shoot to the other (saddle sectors) has demonstrated that the shoot meristem arises
369
from the region of the embryo containing longitudinally oriented files of cells (Poethig, Coe and Johri 1986). To determine cell lineages giving rise to the root and hypocotyl, Scheres et al (1994) analyzed bluestaining sectors in transgenic plants of Arabidopsis containing the maize transposable element ac interrupting a GUS gene-CaMV 35S promoter construct. The ac element randomly excises from the construct to activate GUS gene expression in the cells in which the event occurs. Assuming that blue sectors in the seedling represent descendants of cells in which transposon excision restores GUS expression, it is possible to relate specific cells in the heart-shaped embryo to these embryonic organs from an analysis of sectors spanning a part or all of the root and hypocotyl regions. Without going into the details of the complex analysis involved in this work, the results have led to the conclusion that distinct cell lineages are not specified for the root and hypocotyl, although the derivatives of the hypophysis are restricted to the columella and the central region of the root. However, it was possible to trace the hypocotyl, an intermediate zone near the upper boundary of the root, the root and root meristem back to compartments in the four tiers of cells derived from the lower tier of cells of the globular embryo (Figure 13.6). Since the sector boundaries are not as sharp as the morphological boundaries, on the whole it appears that cell lineages play a relatively minor role during embryogenesis and that there is no overt, determinate pattern of cell divisions giving rise to the major embryonic organs. Clonal analysis of embryogenesis in additional species using critical markers should make it possible to reevaluate the different types of embryo development based on histological analysis. Whether the cell fate is specified in the young embryo or not, cells of the postglobular embryo initiate a program of continued divisions and morphogenesis to form the apical meristems. Mutations that target specific meristems in the seed embryo are useful for determining the fate of the cells present in the meristem during postembryonic stages of ontogeny. Some of the investigations undertaken with meristem mutants relate to the clonal analysis of the development of the tassel and ear shoot in maize (Johri and Coe 1983), the cell lineage patterns in the shoot meristem of maize and sunflower embryos (McDaniel and Poethig 1988; Jegla and Sussex 1989), the alteration in the shoot apical meristem associated with reproductive transition (Sung et al 1992; Bai and Sung 1995), and the production of abnormal
370
Zygotic Embryogenesis
Figure 13.6. Mapping of the end points of colored sectors formed on the seedling of transgenic Arabidopsis. The end points of the large sectors demarcate the earliest divisions, and those of progressively smaller sectors mark later divisions. The sectors are indicated in relation to the seedling diagrammed at left. Black bars indicate the exact location of the sector terminus as determined; bars with white ends indicate estimated sector ends. (A) Sector demarcating the division of the upper and lower tiers of the globular embryo. (B) Root and hypocotyl sectors. (C) Cotyledon shoulder sec-
tors. (D) Hypocotyl sectors. (E) Root sectors with apical ends overlapping basal ends of hypocotyl sectors. (F) Root sectors with apical ends disconnected from basal ends of hypocotyl sectors. (G) Intermediate zone sectors. On the vertical axis, the bars represent cell walls of adjacent epidermal cells. 0, lower wall of the first epidermal cell with root hair; 3, third epidermal cell below the junction of cotyledon vascular strands. (From Scheres et al 1994. Development 120:2475-2487. © Company of Biologists Ltd.)
roots in Arabidopsis (Benfey et al 1993). Findings from these investigations have to some extent altered previously held concepts about the relationship between meristems and contiguous regions; however, since these studies do not materially contribute to our understanding of the developmental fate of early division phase embryos, they will not be further considered. Results from the culture of seed embryo segments have shown that the ability to produce working parts of the missing structures is generally restricted to segments containing the shoot apical meristem. Thus, whereas shoot segments develop into complete seedlings, root and hypocotyl segments retain their typical organization without regenerating shoots (Raghavan 1976b). As shown in an investigation using torpedo-shaped somatic embryos of carrot, region specification seems to operate even in the early stages of embryogenesis in a fashion similar to that seen in seed embryos. It appears that, following extirpation of the lower half of the embryo, virtually any part of the shoot consisting of portions of the hypocotyl or of the hypocotyl and root tissues was able to regenerate a new root, but remnants of the shoot with just the cotyledons and the apical notch did not possess root-forming potential. In contrast, restoration of a normal shoot from a root portion was a rare event and occurred only from root segments consisting of the extreme end of the root pole. During root regeneration from the shoot portion, re-formation of the
missing regions of the embryo begins in the same order and polarity as in an uncut embryo, with the root meristem following the hypocotyl-root tissues (Schiavaone and Racusen 1990, 1991). From this it seems that genetic information required for directing the patterning and timing of events of later stages of embryogenesis is present in the cells of the shoot pole fragment. The extent to which the undifferentiated cells of proembryos and the differentiated regions of slightly more advanced embryos are assigned to a particular fate probably depends upon a succession of events and upon changes in a variety of components. In the proembryo, the events are sustained until the cells are determined as the shoot or root regions. The position developed here with respect to the fate of the territories of early torpedo-shaped embryos is that the events continue until the meristems are fully established. As more cells are produced in the developing embryo, cellular interactions become increasingly complex due to unknown changes in the component processes and their effects on the differentiated cells. For example, based on the capacity of shoot and root segments to form the missing parts, one can conclude that different regulatory states exist in the fragments destined to regenerate. Since global statements such as these do not explain the mechanism of cell lineage and cell fate in developing embryos, one may hope that they will stimulate the formulation of cogent questions and experiments about the systems.
Embryogenesis and Physiology of Growth of Embryos
371
developmental fate of the hypophysis is apparently determined after it is cut off, because ribosome density and concentrations of nucleic acids and The foregoing account has been entirely con- proteins in the cells of the root cap and root epidercerned with embryo growth resulting in recogniz- mis cut off by the hypophysis appear intermediate able changes in embryo size and shape. However, between those of the embryo cells and suspensor size and shape changes, as indicators of the growth cells. The earliest stage of embryogenesis in which of a multicellular structure, leave much to be cytological differences between cells are noted in desired, since the tacit assumption remains that the anticipation of tissue and organ differentiation is embryo attains its form by spatial and temporal the heart-shaped embryo. Here the cells of the promodulations in cell division, cell elongation, and cambium and ground meristem appear more cell differentiation. Histological diversity attained highly vacuolate than those of the protoderm in the embryo might be presaged by detectable (Schulz and Jensen 1968a). ultrastructural and cytochemical changes in critical Ultrastructural studies of embryos of other cells and tissues. What patterns might these cellu- angiosperms are less complete than those lar activities follow in a growing embryo? described in C. bursa-pastoris and contribute little to our understanding of cell specialization during embryogenesis. Early embryogenesis in Hordeum (i) Ultrastructural and Histochemical vulgare involves an increase in the population of Characterization ribosomes and mitochondria in all cells. As the It is hard to imagine that, following the distinct embryo completes two or three rounds of divisions, ultrastructural and histochemical differences that there is a progressive decrease in the distribution of characterize the terminal and basal cells of the two- vacuoles in the cells of the suspensor (Norstog celled proembryo, further differentiation of the 1972a). Accumulation of vacuoles and multivesicuproembryo can occur without similar accompany- lar bodies in the cells of the apical part of the ing changes. There is no question that there are proembryo of Secale cereale is considered indicative some subcellular differences in the distribution of of early stages of cell differentiation to form the organelles as the proembryo is repeatedly parti- scutellum (Stefaniak 1984). In Helianthus annuus, tioned, but whether these changes are symptomatic cells of the embryo have more ribosomes than of cell differentiation in the embryo is a debatable those of the suspensor, giving these cells a more issue. In Capsella bursa-pastoris, as the basal cell ofelectron-dense appearance (Newcomb 1973b). the two-celled proembryo divides transversely to According to Singh and Mogensen (1975), cells of form the suspensor, the terminal cell acquires more the embryo and suspensor of Quercus gambelii seem ribosomes and it stains more intensely for proteins to amplify the ultrastructural features of the termiand nucleic acids than do the suspensor cells. On nal and basal cells, respectively, of the two-celled the other hand, ER and dictyosomes are more proembryo. As the embryo is phased out into the abundant in the basal cell than in the terminal cell globular stage, suggestive of increased metabolic (Schulz and Jensen 1968c). As the terminal cell activity, the organelle population undergoes a comforms the globular embryo and the suspensor cell plete reorganization. Increases in plastid number is partitioned transversely into a filament, ribo- and complexity, mitochondrial number, dictyososome density is distinctly greater in the cells of the mal activity, ribosomal aggregation, and the embryo than in those of the suspensor. The pres- amount of ER are a conspicuous part of the ence of starch and lipid bodies in the terminal cell processes that peak in the late globular stage of the three-celled proembryo and their disappear- embryo. By contrast, the number and distribution ance during the division of this cell have suggested of these organelles in the suspensor cells remain that these storage bodies probably function as a unchanged. Besides the conventional subcellular source of energy for mitosis (Schulz and Jensen structure, some unique changes that have been 1968a). When the hypophysis is cut off from the ter- described are the disappearance of starch and lipid minal cell of the suspensor in the globular embryo droplets and an increase in ribosome density and in of C. bursa-pastoris, it acquires a basic subcellular the number of autophagic vacuoles in globular profile different from that of cells of the embryo; embryos of Petunia hybrida (Vallade 1980); the presthis is evident in the highly vacuolate cytoplasm, ence of extensive nuclear evaginations and vesicles the low ribosome density, and the generous in five- to eight-celled embryos of Oryza sativa amounts of ER, starch, and lipids in the cell. The (Jones and Rost 1989b); the appearance of large 3. CELLULAR ASPECTS OF EMBRYOGENESIS
372
mitochondria in the cells of the globular embryo of Arabidopsis (Mansfield and Briarty 1991); and the presence of vacuoles with flocculent materials, membrane fragments, and weak electron-scattering bodies in the cells of the globular embryo of Viciafaba (Johansson and Walles 1993b). Cells of the heart-shaped and older embryos have the normal complement of organelles typical of meristematic cells, and no generally characteristic ultrastructural changes augur the onset of tissue and organ differentiation in them. Clearly, in a situation in which newly formed cells are constituted into tissues in a developing organ, it is unrealistic to expect major upheavals in cell structure. Some exceptions are noteworthy, such as a 4- to 16-fold proliferation of the rough ER found in the cotyledons of Phaseolus vulgaris during the cell elongation phase, coincident with storage protein accumulation (Briarty 1980). At the histochemical level, tests on embryos of various plants have encountered higher concentrations of RNA in the cells of the presumptive cotyledons than in the rest of the embryo, though it is not surprising that transcriptional activity of the cells increases preparatory to differentiation (Rondet 1962; Pritchard 1964b; Schulz and Jensen 1968c; Vallade 1970; Norreel 1972; Syamasundar and Panchaksharappa 1976; Shah and Pandey 1978). Another measure of activity of developing embryos is the distribution of enzymes of respiratory metabolism. Determinations made in the embryos of Gossypium hirsutum show a striking correlation between tissue and organ differentiation and the activity of succinic dehydrogenase, a key enzyme in aerobic respiration. A high enzyme titer is found in all recognizable organs of the embryo, such as the suspensor, cotyledons, hypocotyl, shoot apex, subapical region, and the elongating radicle, during the period of their active growth (Forman and Jensen 1965). Alcohol dehydrogenase is involved in anaerobic metabolism, and multiple forms of the enzyme are generally present in cells. Differentiation of the axis and cotyledons of embryos of Phaseolus vulgaris is associated with the appearance of a fastmoving isozyme of alcohol dehydrogenase in these organs (Boyle and Yeung 1983). Since the embryo grows in an environment insulated from atmospheric oxygen, perhaps a case can be made for a functional role in normal cellular activities for organ-specific isozymes of the relevant enzymes of anaerobic metabolism. These observations not only give meaning to the process of embryogenesis, but also suggest that a cytochemical examination of embryogenesis with more refined techniques should prove instructive.
Zygotic Embryogenesis
(ii) Cell Division and Cell Elongation A generally accepted notion is that cell divisions predominate during the early period of embryogenesis and that cell enlargement prevails during the later period. A variation of this theme is observed in the embryos of Capsella bursa-pastoris and Gossypium hirsutum, in which the increase in cell number is interrupted by a lag period during cotyledon initiation; this results in a conventional S-shaped curve for cell number from zygote to the globular embryo, and another S-shaped curve from the heart-shaped to the mature embryo (Pollock and Jensen 1964). In the globular embryos of Hordeum vulgare, Oryza saliva, and Triticum aes-
tivum, an exponential increase in cell number is maintained only up to the 20-40-celled stage, after which the rate of increase declines up to the 100-200-celled stage and again up to the 1,000celled stage. This has been attributed to subtle changes in cell metabolism preparatory to changes in form in the embryos (Nagato 1978). Changes in cell size are tightly linked to embryogenic divisions and, notwithstanding the lack of extensive quantitative studies, it is clear that a reduction in cell size accompanies the first several rounds of division of the zygote. In G. hirsutum and C. bursa-pastoris, the zygote remains the largest cell in the life of the embryo, and the average cell size is reduced to about one-fifth that of the zygote even as early as the first division. Cell size drops even further up to the 100-celled embryo stage in G. hirsutum and up to the young globular embryo stage in C. bursapastoris (Pollock and Jensen 1964). A distinctive reduction in cell size up to the midglobular stage of the embryo also occurs in Arabidopsis. The reduction in cell size is accompanied by progressive decreases in cell-to-nuclear volume ranging from 365 in the zygote to about 13 in the midglobular embryo (Mansfield and Briarty 1991). These changes probably help to establish a favorable balance between the nuclear size and cytoplasmic volume for rapid divisions. The distribution of mitotic figures is uniform in very young embryos of G. hirsutum and C. bursapastoris and continually changes to reflect the onset of tissue and organ differentiation, assuming a Ushaped pattern in heart-shaped embryos and ending up in a Y-shaped pattern in torpedo-shaped embryos. Apparently, the onset of specific histogenic patterns during embryo development depends upon the attainment of a critical cell number before the shape change occurs, and is followed by asymmetric distribution of cell divisions. Thus,
Embryogenesis and Physiology of Growth of Embryos
the transition of the globular embryo to the heartshaped form seems to result from an increased mitotic activity in the distal part of the embryo (Ushaped pattern) that gives rise to the cotyledons. A different pattern of distribution of cell divisions (Yshaped pattern), accompanied by cell elongation, might account for the transition of the heartshaped embryo to the torpedo-shaped stage (Pollock and Jensen 1964). The key facts that may provide clues to a more explicit understanding of the factors that control differential cell divisions in a homogeneous tissue are, however, not available. One might suspect that some cytoplasmic gradient is being set up by cell division factors diffusing in the direction of the embryo where potential divisions occur, but as yet we have no evidence for this.
373
1967), Vicia faba (Brunori 1967), and Brassica napus (Silcock et al 1990) in the middle or later periods of embryogenesis indicate that the DNA content of the cells does not deviate from the 2C level even as dehydration of embryos preparatory to quiescence of seeds sets in. It is possible that with progressive desiccation of the embryo, other factors, such as available water supply, might limit DNA synthesis. A great deal of our present knowledge of the dynamics of DNA synthesis and accumulation in developing embryos has been acquired from analyses of cotyledons of legumes and other plants that accumulate storage proteins (Figure 13.7). After an initial period of growth by cell divisions, the subsequent growth of cotyledons is marked by cell expansion accompanied by the accumulation of enormous quantities of storage products, mostly proteins. Scharpe and van Parijs (1973) showed that (iii) DNA Synthesis cells of developing cotyledons of Pisum sativum The concept of constancy of DNA in a cell pre- continue to synthesize DNA after cell divisions dicts that DNA synthesis must keep pace with the have ceased; this leads to a progressive increase in production of cells so that each new cell will have a the DNA amount in the cells of full-grown cotylefixed amount of DNA. One might hope that the dons to a level as high as 64C. The general pattern precise division of the zygote to produce a multi- of changes in DNA levels of cotyledonary cells of P. cellular embryo will depend upon an efficient arvense closely resembles that described in P. DNA duplication mechanism, followed by cytoki- sativum (D. L. Smith 1973). These two studies have nesis. This picture is obscured in some studies been augmented by reports of a continued increase because of complications introduced by endoredu- in the nuclear DNA content of the parenchymatous plication. The DNA content of the zygote has been cells of cotyledons of Vicia faba (Millerd and shown to range from the expected 2C-4C amounts Whitfeld 1973), Gossypium hirsutum (Walbot and in Tradescantia paludosa (Woodard 1956), Hordeum Dure 1976), Glycine max (Dhillon and Miksche vulgare, and H. vulgare X H. bulbosum (Bennett and 1983), Phaseolus vulgaris (Briarty 1980; Johnson and Smith 1976; Mogensen and Holm 1995), to attain a Sussex 1990), and Sinapis alba (Fischer et al 1988). In 3C-6C level in Petunia hybrida (Vallade et al 1978), the embryos of V. faba (Millerd and Whitfeld 1973; or to become as high as 16C in H. distichum (Mericle Borisjuk et al 1995) and G. hirsutum (Walbot and and Mericle 1970). DNA amounts in the nuclei of Dure 1976), it has been established that the increase two- to four-celled embryos of these species show a in DNA content of cells following completion of similar range, from a 2C amount in T. paludosa to 8C division is by endoreduplication; that endoredupliin H. distichum. In the globular stage embryo of cation occurs at the same time that cell elongation Vanda sanderiana, there is a gradient of increasing and storage protein accumulation proceed adds DNA values from 2C to above 8C in cells from the interest to the possible relationship among the distal meristematic tip to cells in the proximal three events. parenchymatous region (Alvarez 1968). With furThe question as to whether the increase in DNA ther development, as in heart-shaped and torpedo- content of cotyledonary cells represents endoredushaped embryos of Glyine max and Vitis vinifera, plication of the entire genome or selective amplifithere is a stabilization of DNA content in the cells at cation of certain sequences has engendered some the 2C-4C levels (Chamberlin, Homer and Palmer discussion. According to Walbot and Dure (1976), 1993b; Faure and Nougarede 1993). In Gossypium the presence of identical reassociation kinetics of hirsutum, the amount of DNA per cell remains nonrepetitive DNA from cotton cotyledons before fairly constant from the early heart-shaped to the and after endoreduplication, the absence of satelcotyledonary stage embryos, although there is lites after analytical ultracentrifugation of the two nearly a 30-fold increase in cell number during this DNA samples, and the constancy in the number of transition (Fisher and Jensen 1972). Analyses of rRNA cistrons per unit of the two DNA samples embryos of Lactuca sativa (Brunori and D'Amato argue against selective amplification. Results in
374
Zygotic Embryogenesis
Cell Num ber
Root Tip
•1
2C
4C
8C
16C
32C
64C
128C 256C
1 2C
f 4C
1 8C
1 16C
32C
64C
128C
256C
32C
64C
128C
2S6C
Stage V
,J 2C
•
| ^ _ 4C
8C
tec
Stage VI
4-
r- • - 1 4C 2C 1
•
- • 1 • i • F • 32C 8C
16C
i_ ^—
- ^ 128C
—I 2S6C
64C
Stage VII -
2C
sc
4C
tec
MI\i u I
•_
I 64C
1 128C
1 256C
1 32C
1 64C
1 128C
1 256C
32C
64C
128C 2S6C
32C
12"
••
8'
Stage VIII
4"
Ju 2C
6 4 2 "
4C
.---F-J
8C
16C
Stage VIII Axis
2C
4C
• 8C
'6C
Figure 13.7. Distribution of nuclear DNA in the root tip cells, cells of developing cotyledons (stages IV to VIII), and stage VIII embryo axis of Phaseolus vulgaris. Stage IV is the youngest and stage VIM is the oldest embryo analyzed. Nuclear DNA contents are given in C values. (From Johnson and Sussex 1990. Chromosoma 99:223-230. © Springer-Verlag, Berlin.)
accord with this finding have come from a comparison of the hybridization kinetics of cloned soybean embryo storage protein gene with an excess of DNA isolated from embryo and leaf. The presence of identical smooth curves resulting from the annealing of the two DNA samples attests to the absence of selective gene amplification during continued DNA synthesis in the embryo (Goldberg et al 1981a). Additionally, Southern blot hybridization of DNA from the embryos and leaves of soybean with probes for glycinin, the major soybean storage protein, and for a 15-kDa protein showed similar patterns, confirming that selective gene amplification is not a factor to be reckoned with during
endoreduplication (Fischer and Goldberg 1982). Results from the use of a wide selection of molecular probes in hybridization experiments have also discounted gene amplification and random synthesis of DNA as factors in DNA accumulation in the cells of cotyledons of Phaseolus vulgaris Qohnson and Sussex 1990). If storage protein genes or genes for other essential proteins are not selectively amplified in the cotyledons, it is a mystery why the cells contain a lot more genome than all the proteins it can code for. Conceivably, the extra DNA copies may be of restricted but strategic significance as a storage form of nucleotides during germination. 4. TISSUE AND MERISTEM DIFFERENTIATION DURING EMBRYOGENESIS
To give legitimacy to the embryo as the future sporophyte, the three primary meristems - namely, the protoderm, ground meristem, and procambium and the shoot and root apical meristems are blocked out sequentially at different stages of embryogenesis. In the embryo of Capsella bursa-pastoris, the first meristem to differentiate is the protoderm, which is formed by a periclinal division of cells of the octant embryo. Because of the nature of this division, the packet of cells of the globular embryo can be neatly separated into eight external cells of the protoderm, sheltering an internal group of eight cells. After they are cut off, cells of the protoderm continue to divide anticlinally to keep pace with the increasing volume of the embryo and they gradually begin to acquire characters of the epidermis (Schulz and Jensen 1968a). That early during embryogenesis the peripheral layer Of cells becomes superficially similar to epidermal cells is attested by the presence of cutin on globular and heart-shaped embryos (Rodkiewicz, Fyk and Szczuka 1994). The expression of a gene that encodes a candidate protein involved in cutin biosynthesis on the protoderm of early stage carrot embryos is consistent with this observation (Sterk et al 1991). At the heart of the timing of initiation of the protoderm lies the question whether all cells in the embryo are undetermined until the protoderm is carved out or whether genes for protoderm formation are turned on at an earlier stage. In Citrus jambhiri, the commitment to differentiate an epidermis is made by the zygote itself through the development of a cuticle around its outer surface; this trait is faithfully transmitted to the external layer of cells of the embryo. Failure of epidermal re-forma-
375
Embryogenesis and Physiology of Growth of Embryos
tion in embryos whose epidermis had been surgically removed has shown that both epidermal and subepidermal cell layers are already locked into their developmental fate along divergent pathways early during embryogenesis (Bruck and Walker 1985a, b). The ground meristem and the procambium are derived from the inner core of cells of the globular embryo. The cells of the ground meristem may differentiate into a cortex or into both cortex and pith. In either case, the peripheral layer of cells of the central core of the proembryo is the progenitor of the cortex of the mature embryo. During the elongation of the embryo that begins around the torpedo-shaped stage, the cells of the peripheral layer divide periclinally to form three or four concentric layers of cells of the embryonic cortex. Formation of the pith from the ground meristem is foreshadowed by repeated divisions of the cells of the central region of the globular embryo, which continue as the pith of the hypocotyl part of the mature embryo. It was mentioned earlier that the presence of transcripts of the KTI-3 gene is a molecular marker of the micropylar end of the soybean globular embryo. Because the transcripts persist in the cells of the ground meristem of heart-shaped and cotyledon stage embryos, it is conceivable that these cells have their origin in cells of the globular embryo harboring KTI-3 transcripts (Perez-Grau and Goldberg 1989).
the analysis of tissue differentiation in the embryos of several other plants, but limitations of space do not permit us to take up these examples. On the whole, the pattern of tissue formation is a crucial process of embryogenesis and one of enduring histogenetic interest, but one about whose mechanisms we know very little. (i) Shoot and Root Apical Meristems
Differentiation of the shoot and root apical meristems in the embryo makes the preexisting, but cryptic, bipolarity of the embryo more obvious. Investigations on the histogenesis of embryos of Nerium oleander (Mahlberg 1960), Downingia baci-
galupi, and D. pulchella (Campanulaceae) (Kaplan 1969) have shown that a precise pattern of cell divisions and differentiation distinguishes the formation of the shoot and root apical meristems. In N. oleander, the earliest indication of the shoot and root apices is the appearance of pockets of partially differentiated meristematic cells at both poles of the globular embryo. These cells are interpreted as direct descendants of the octant stage embryo rather than as newly generated cells with meristematic potential. In all three species, differentiation of the shoot apex is also augured by a change from the random distribution of cell divisions to their concentration in the apical half and by cytohistological differentiation into a peripheral and a Procambium initiation is a critical feature of his- central zone. It is to be noted that these changes togenesis of the embryo and occurs in the cells of occur before the appearance of the cotyledons, sugthe cortex or pith of globular embryos before the gesting that the cotyledons are the first formative primordia of the cotyledons are formed. As shown organs produced by the shoot apex. The visible difin Nerium oleander (Apocynaceae), elements of the ferentiation of the shoot apex generally coincides procambium are first seen in the innermost part of with the change in the shape of the embryo from the cortex of the early heart-shaped embryo at the globular to heart shaped and is attributed to the level of the presumptive cotyledons. More initials continued meristematic activity of the protoderm are laid down on either side of the original file and and the underlying layer of cells in the apical part this leads to the formation of a complete ring of pro- of the embryo. The formation of the central zone of cambial cells. After the procambium is established the shoot apex is prefigured by the cytological difin the hypocotyl of the developing embryo, it ferentiation of three or four layers of cells below the extends acropetally into the cotyledons and shoot protoderm. Differentiation of the shoot apex is apex, and basipetally into the root apex. Thus, by complete when it becomes dome shaped and the time the embryo becomes torpedo shaped, it has lapses into a state of quiescence or dormancy in the a solid cylinder of procambium extending through mature embryo. its entire length and into the cotyledons (Mahlberg Depending upon the types of embryo develop1960). In Rhizophora mangle (Rhizophoraceae), files ment (described in a previous section), the root of procambium linking the shoot and root apices are meristem and the root cap have their origin in the apparently initiated independently, because they terminal cell or basal cell of the two-celled proemappear discontinuous for a brief period until they bryo. For illustrative purposes, an example of the are connected to the hypocotyl procambium by embryo in which the root apex is derived wholly cotyledonary strands (Juncosa 1982). Besides these from the terminal cell of a two-celled proembryo is selected examples, contributions have been made to that of D. pulchella. Limited transverse and longitu-
Zygotic Embryogenesis
376
dinal divisions of the terminal cell lead to the formation of a large proembryo consisting of four superimposed tiers of one to four cells in each tier. During the organization of the root apex, cells of the most basal tier divide periclinally to give rise to the root cap. Protodermal cells of the next tier, toward the apical part of the proembryo, also divide periclinally to contribute to the root cap. In addition, cells of this second tier generate derivatives that form the root cortex and meristem (Kaplan 1969). In C. bursa-pastoris, in which the embryonic root apex is derived from the hypophysis, the initial division of this cell is in the transverse plane. By further transverse and longitudinal divisions, the daughter cell close to the suspensor gives rise to the root cap and root epidermis, and the one close to the organogenetic part of the embryo generates the rest of the root apex, including the root meristem (Schaffner 1906; Soueges 1919). Whatever factors determine the extent to which the terminal and basal cells of the two-celled proembryo are involved in the formation of the root apex, the embryonic root apex - compared to the shoot apex - displays a remarkable degree of architectural plasticity that allows it to generate closely integrated final products. (ii) Origin of the Quiescent Center A feature of the angiosperm root meristem is the presence of a group of cells that divide rather infrequently or not at all. This packet of cells, known as the quiescent center, is characterized by low DNA, RNA, and protein synthetic activities; the combination of cellular properties has led to the suggestion that the quiescent center originates as a result of the decreasing metabolic activity of certain cells in the embryonic root meristem. Despite the merit of this suggestion, and the use of a variety of methods to monitor the formation of the quiescent center in developing embryos, the evidence in favor of it is at best ambiguous. Relevant data typically invoked to characterize the appearance and disappearance of the quiescent center in developing embryos include the tempo in mitotic activity, staining reactions, and the pattern of 3H-thymidine incorporation (Sterling 1955; Rondet 1962; Vallade 1972; Jones 1977; Clowes 1978a, b). Localization of mRNA in developing embryos of C. bursa-pastoris using 3H-poly(U) as a probe has shown that autoradiographic silver grains caused by annealing of the label are plentiful in the cells of the developing embryo, except in the hypophysis and its derivatives, which constitute the quiescent center in the root apex of the mature
embryo. In seedling roots, the same cells that fail to anneal 3H-poly(U) also fail to incorporate 3H-thymidine, indicating their identity as the quiescent center. These correlations support the conclusion that the quiescent center in the root of C. bursa-pastoris has its origin in the hypophysis of the embryonic suspensor (Raghavan 1990a). 5. THE SUSPENSOR
In the study of angiosperm embryogenesis, interest at the morphological to molecular levels has been bestowed largely on the organogenetic part of the embryo, with little or no attention beyond anatomical descriptions given to the suspensor. This is mainly due to the well-grounded notion that the role of the suspensor is purely of a mechanical nature, namely, to push the growing embryo deep into the chalazal end of the embryo sac, close to the source of nutrients. During the past three decades, a number of ultrastructural and cytological studies, along with a few physiological and biochemical investigations, have led to a revival of interest in the structure and physiology of the suspensor and to a refreshing departure from the classical views about its function. The current view of the structure-function relations of the suspensor is that the suspensor is a dynamic part of the embryo complex, functioning in the absorption and short-distance translocation and exchange of metabolites necessary for the growth of the embryo. Although a filament of 8-10 cells attached to the organogenetic part of the embryo, as seen in Capsella bursa-pastoris or Arabidopsis, is one of the
most familiar images of the suspensor, comparative studies of embryogenesis in angiosperms have shown that suspensors come in various sizes and shapes. A few of these interesting variations in suspensor morphology have been collated, illustrated, and described (Natesh and Rau 1984; Raghavan 1986c). In many plants, the morphological complexity of the suspensor is attributed to the extension growth of its cells, which reach the extraembryonal tissues of the ovule and even the ovary. In embryology literature, these outgrowths are frequently referred to as haustoria, implying that they are specialized structural adaptations for the absorption of nutrient substances. In some families, such as Orchidaceae, Podostemaceae, and Trapaceae, in which complex suspensors with haustoria have been described, the ovules are characterized by the complete absence of endosperm or by the presence of a reduced amount of endosperm. This indicates that nutrients absorbed
Embryogenesis and Physiology of Growth of Embryos
by the suspensor haustoria may provide an important source of energy for the growth of the embryo. However, direct evidence has been difficult to obtain, as most observations are based on the intimate association of the haustoria with the extraembryonal tissues and do not bear directly on the actual transfer of nutrients or on the ultrastructural modifications of the haustoria for nutrient uptake and transfer. As will be described, ultrastructural modifications suggest such a function for the regular, nonhaustorial suspensor cells of several plants.
377
proteins in the heart-shaped embryo. Gradually, the middle lamella of cells begins to loosen, and eventually the suspensor is obliterated by the burgeoning embryo. The basal cell generally remains active for some time after the collapse of the other cells (Schulz and Jensen 1969). The presence of a series of invaginations on the walls of cells of the suspensor of C. bursa-pastoris has been a focal point of ultrastructural study. The invaginations arise from the outer lateral walls of the mid-suspensor cells of the globular embryo and increase in number and complexity at the heartshaped stage. The wall projections are a feature mat (i) Ultrastructure of the Suspensor Cells the cells of the suspensor share with the basal cell. Despite the fact that the suspensor is an The inner wall of the micropylar and lateral parts ephemeral organ that does not become a part of the of the basal cell are also thrown up into an elabomature embryo, structural adaptations for specific rate network of invaginations projecting into the functions find their highest expression in the cells cytoplasm. In fact, the projections first appear in of the suspensor. With regard to the alignment of the wall at the micropylar end of the zygote and the organelles and the density of their distribution, increase in number, peaking in the basal suspensor suspensors of different species thus far investi- cell of the heart-shaped embryo (Schulz and Jensen gated, as well as the different cells of a suspensor, 1968c, 1969). are astonishingly diverse. Other aspects of the susAs will be seen later, Phaseolus is a paradigm for pensor, such as the number of cells present and defining and interpreting the nuclear cytology of their life span, also vary in the different species. suspensor cells in an angiosperm. In P. coccineus, The first detailed account of the ultrastructural the suspensor is polarized from the proembryo cytology of the suspensor was provided in Capsella stage onwards with respect to mitotic activity. bursa-pastoris. For purposes of discussion, three Although divisions mostly cease in the more basal stages have been identified in the life of the suspen- cells of the suspensor by the proembryo stage, they sor: (a) the octant embryo, whose suspensor con- continue in the cells toward the chalazal region. sists of 6 cells; (b) the globular embryo, whose The suspensor attains its maximum number of suspensor attains the maximum cell number and about 200 cells in the heart-shaped embryo; at the appears as a filament of 10 cells interposed between cotyledon stage, it reaches its maximum size. The a large, vacuolate basal cell and the embryo; and (c) presence of a large number of cells in the suspenthe heart-shaped embryo, whose suspensor attains sor, however, does not enhance its longevity the maximum length and lives out its genetically (Yeung and Clutter 1978). permissible life span. Throughout its development, Cell specialization in the suspensor cells of P. the suspensor is functionally connected with both coccineus begins with the appearance of wall infoldthe embryo and the basal cell by numerous ER-con- ings as early as the proembryo stage. Soon after, the taining plasmodesmata. Although cells of the sus- cells become committed to a developmental pathpensor harbor the usual array of organelles, way characterized by a disproportionate increase ribosomes appear to be specially concerned with in the number of ribosomes, plastids, mitochoncellular interactions of this organ. In line with this dria, dictyosomes, and smooth ER, which continview of the seemingly special role of ribosomes, it ues until the cotyledon stage. In considering the has been observed that at the octant stage of the ultrastructure of the wall of developing suspensor embryo, the suspensor cells have a respectable cells, it is significant that an increase in the number number of ribosomes, which show an increasing of invaginations parallels the increase in organelle gradient toward the chalazal end. In the globular density and that a raft of organelles, particularly embryo, ribosomes gradually dissipate in the cyto- ER, dictyosomes, and mitochondria, lie in close plasm; they completely disappear from the suspen- proximity to the wall ingrowths. These observasor cells of the heart-shaped embryo. Symptomatic tions further strengthen a role for the wall projecof the inevitable failure of the cellular machinery of tions in nutrient exchange (Yeung and Clutter the suspensor is the onset of cytoplasmic degener- 1978). ation accompanied by the loss of nucleic acids and Although the suspensor in Stellaria media is
378
Figure 13.8. Morphology and infrastructure of the suspensor of Medicago scutellata. (a) A globular embryo showing vacuolation in the chalazal and basal suspensor cells, (b) Electron micrograph of a portion of the basal suspensor cell of (a), showing wall ingrowths, (c) A heart-shaped embryo with suspensor. (d) Electron micrograph of a portion of the basal suspensor cell of (c), showing mitochondria associ-
Zygotic Embryogenesis
ated with wall ingrowths. BS, basal cell; BSC, basal suspensor cell; END, endosperm; M, mitochondria; S, starchcontaining integumentary cells; Wl, wall ingrowth. Scale bar = 10 urn (a), 1um (b, d), 100 um (c). (From Sangduen, Kreitner and Sorensen 1983; photographs supplied by Dr. E. L. Sorensen.)
formed from the terminal cell of the two-celled Kiihner 1976), Medicago sativa, M. scutellata proembryo (Caryophyllad type), the undivided (Sangduen, Kreitner and Sorensen 1983a), Alisma basal cell presents a profile similar to, and no less lanceolatum, and A. plantago-aquatica (Alismaceae) complex than, that of the cells of the suspensor. (Bohdanowicz 1987) give every appearance of havOne manifestation of the complexity portrayed in ing undergone some degree of specialization the basal cell is the appearance of an array of mas- unique to each species. sive wall projections on the inner surface, along Wall labyrinths of the type just described appear with a preponderance of mitochondria near them. to be a widespread phenomenon (Figure 13.8) and Microbodies, plastids, and extensive whorls of have been documented in suspensor cells of P. vultubular ER, which are generally absent from the garis (Schnepf and Nagl 1970), Diplotaxis erucoides cells of the embryo, become a conspicuous part of (Simoncioli 1974), T. majus (Nagl 1976b), Brassica the organelle complement of the basal cell and the napus (Tykarska 1979), Vigna sinensis (Hu, Zhu and cells of the suspensor. The plastids of the suspensor Zee 1983), M. sativa, M. scutellata (Sangduen, cells have an unusual morphology with numerous Kreitner and Sorensen 1983a), Alyssum maritimum tubules and electron-translucent inclusions. There (Brassicaceae) (Prabhakar and Vijayaraghavan is also a decreasing gradation in the complexity of 1983), Alisma lanceolatum, A. plantago-aquatica plastids and in the number of microbodies from the (Bohdanowicz 1987), Glycine max (Dute, Peterson basal cell to the chalazal suspensor cells (Newcomb and Rushing 1989; Chamberlin, Homer and Palmer and Fowke 1974). Plastids found in the suspensor 1993a), and Viciafaba (Johansson and Walles 1993b). cells of Pisum sativum (Marinos 1970b), Phaseolus There is now considerable and varied evidence vulgaris, P. coccineus (Schnepf and Nagl 1970; Yeungbased on the transfer cell morphology of the susand Clutter 1979), Ipomoea purpurea (Ponzi and pensor cells to support the solute absorption theory Pizzolongo 1973), Tropaeolum majus (Nagl and of suspensor function first articulated by Schulz
Embryogenesis and Physiology of Growth of Embryos
and Jensen (1969). According to this view, metabolites from the endosperm or from the surrounding diploid cells of the ovule are delivered to the embryo through the suspensor cells, with the wall imaginations facilitating the transfer by increasing the surface area of absorption. The plasmodesmata perforating the embryo-suspensor boundary may serve to maintain an open channel of communication for the flow of solutes between the suspensor and the embryo. That the path of nutrients to the developing heart-shaped embryos of P. vulgaris and P. coccineus is through the suspensor has become evident from the work of Yeung (1980). The results showed that if 14C-sucrose is administered through pods or to isolated embryos, much of the radioactivity appears in the suspensor and in the suspensor pole of the embryo. Somewhat similar results were obtained in the transport of a labeled polyamine, putrescine, in the ovules of P. coccineus (Nagl 1990). Although the new knowledge of the wall architecture of the cells of the suspensor of angiosperm embryos has pushed the frontiers of our understanding of the function of this organ to a higher level, until more is known about the gradient properties of the suspensor and its surrounding cells, the mechanism of nutrient incorporation into the embryo through the suspensor will remain unclear.
(ii) Nuclear Cytology of the Suspensor Cells
379
gradient of DNA amounts, with low values in the cells close to the embryo, intermediate values in the neck region of the suspensor, and very high values in the cells of the basal region. Endoreduplication in the suspensors of Alisma lanceolatum and A. plantago-aquatica is nonconventional since it affects only the large basal cell. The nucleus of the cell rhythmically bloats in size and, based on the increase in nuclear volume, it has been estimated that in the latter species the nucleus attains a ploidy level as high as l,024N (Bohdanowicz 1973, 1987). Although the chromosomal constitution of endoreduplicated suspensor cells of several species has revealed polyteny, the phenomenon has been most extensively studied in P. coccineus (Nagl 1967) and P. vulgaris (Nagl 1969a, b). Like polytene chromosomes in the salivary glands of dipteran larvae, those of the suspensor cells of the two species of Phaseolus also display puffs, which represent sites of DNA synthesis and transcription. However, some DNA-synthesizing puffs are found on chromosomes only at specific stages of embryogenesis, indicating the possibility of a stage-specific variation in DNA replication in the same polytene chromosome (Tagliasacchi et al 1984). A gene identified in the polytene chromosome of P. coccineus by in situ hybridization using a cloned probe encodes a protein that inhibits fungal endopolygalacturonase (Frediani et al 1993). Since this is a stress-related protein, it is too early to assess the significance of this observation.
As shown by the molecular hybridization of Investigations on the nuclear cytology of suspensor cells of various angiosperms have shown 25S, 18S, and 5S 3H-labeled rRNA fractions to DNA that endoreduplication coupled with the presence of suspensor cells of P. coccineus, genes for rRNA of polytene chromosomes distinguishes their are clustered at the nucleolus-organizing regions of development. A recurrent theme that has since polytene chromosomes (Durante et al 1977). emerged from these studies is that despite its short However, only the ribosomal cistrons embedded in life span, the suspensor is made up of transcrip- random nucleolus-organizing centers are functiontionally active cells that may play a regulatory role ally active and synthesize RNA in a given chromoin the development of the embryo. Research on the some. In general, compared to the heterochromatic cytology of suspensor cells has been reviewed regions of the chromosomes, the bands engaged in before and will not be considered here again RNA puffing are undermethylated; consequently, (Raghavan 1986c). transcriptional activity of polytene chromosomes On the basis of nuclear volume or quantitative may rely on specific molecular components of Feulgen microspectrophotometry, values for the DNA (Frediani et al 1986a, b). When the activities DNA content of suspensor cells so far reported of the same chromosome puffs at different stages of are analyzed by autoradiography range from 16C in Sophora flavescens (Fabaceae) embryogenesis 3 (Nagl 1978) to 8,192C in Phaseolus coccineus (Brady of H-uridine incorporation, they are found to 1973). In P. coccineus (Brady 1973; Nagl 1974) and engage in RNA synthesis with different amplitudes Eruca sativa (Corsi, Renzoni and Viegi 1973), the (Forino et al 1992). Another intriguing observation outcome of endoreduplication in cells in the differ- is that differing labeling patterns appear in the ent regions of the suspensor is somewhat different, polycistronic transcription units for 25S, 18S, and but its consequences in terms of gene regulation are 5S rRNA genes when polytene chromosomes are probably the same. In both these species, there is a annealed with fractions of DNA banded in the ana-
380
lytical ultracentrifuge, showing that rRNA cistrons of the same class are interspersed with different DNA sequences (Durante et al 1987). Given the nottoo-low number of the chromosomes in the suspensor cells and given the complex, convoluted nature of the chromosomes, these cytological observations are illuminating and provide a good beginning toward the goal of achieving a complete understanding of the transcriptional apparatus of the polytene chromosomes. The question as to whether polytene chromosomes undergo selective DNA amplification has been subject to intensive study. Autoradiography of 3H-thymidine incorporation into suspensor cells of P. coccineus showed that in most cells the label is found exclusively in the heterochromatic region of the polytene chromosomes although in a small number of cells the label decorates the entire nucleus. These results have been interpreted to indicate that certain genes of the heterochromatic regions undergo DNA amplification, in addition to the scheduled DNA synthesis, due to endoreduplication (Avanzi, Cionini and D'Amato 1970). Consistent with the observation that cells in the different regions of the suspensor display varying degrees of endoreduplication, selective DNA amplification is found to prevail in the puffs of the highly polytene chromosomes of cells in the micropylar region of the suspensor, whereas cells produced during the early stages of suspensor development show uniform labeling over the nuclei due to endoreduplication (Cremonini and Cionini 1977). The occurrence of selective DNA amplification has been confirmed in the analytical ultracentrifuge from the density shifts in DNA isolated from suspensor cells of P. coccineus (Lima-deFaria et al 1975). This work showed the presence in the suspensor cells of a class of DNA with buoyant density that has no parallel with DNA isolated from other parts of the plant. As shown in the same work by a lower saturation value in hybridization experiments with total suspensor DNA and rRNA, certain DNA sequences, particularly the genes for rRNA, are underreplicated in the genome of the suspensor cells. Thus, a seemingly fragile balance between amplification of certain genes and underreplication of other genes governs the nuclear activities of cells of the suspensor during a major part of its life. Although these observations stress a regulatory role for the polytene chromosomes of the suspensor cells, precise determination of the function of their gene products, given the small size and transient nature of the suspensor, requires a protracted study.
Zygotic Embryogenesis (iii) Physiology and Biochemistry of the Suspensor
Based on cytochemical, physiological, and biochemical information, the suspensor cells are considered to be metabolically active throughout their life span; the activities may be geared to the absorption of nutrients or for self-autolysis. In Brassica campestris, several enzymes of oxidative metabolism are found to be maximally active in the suspensor cells between the globular and torpedo-shaped embryos and to diminish thereafter (Malik, Singh and Thapar 1976; Malik, Vermani and Bhatia 1976; Bhalla, Singh and Malik 1980a, 1981b). The preponderance of enzymes in the actively growing suspensors presumably indicates the need for a high metabolism connected with nutrient uptake. However, the cessation of growth of the suspensor promptly initiates hydrolytic enzyme activities beginning in the cells at the micropylar end. The presence of specialized plastids in the suspensor cells of Phaseolus coccineus and P. vulgaris (Nagl 1977;
Gartner and Nagl 1980) has led to the view that the acid phosphatase of these organelles may possibly play a role in the autolysis of the suspensor. In Tropaeolum majus, whose suspensor produces haustoria, self-digestion of the suspensor is initiated by degenerative changes in the mitochondria coupled with the presence of high acid phosphatase activity in the haustorium (Nagl 1976c; Gartner and Nagl 1980). Other enzymes, such as alkaline phosphatase and acetylesterase, are also active in the cells of the suspensor and in the haustorium, becoming maximally effective at the early cotyledon stage of the embryo when suspensor autolysis is well under way (Singh, Bhalla and Malik 1980). Comparative analyses of RNA and protein metabolism of the suspensor and embryo of P. coccineus have provided evidence of higher transcriptional and translational activities in the suspensor cells than in the cells of the embryo. Expressed on a per cell basis, RNA content and the rate of RNA synthesis are low during the early development of the suspensor, increasing to a maximum at the late heart-shaped or early cotyledon embryo stage, and then declining. Protein content and the rate of protein synthesis also register low values in the suspensor cells of early stage embryos but increase substantially in the suspensors of late stage embryos. Characteristically, at most stages analyzed, cells of the suspensor contain more RNA and proteins and synthesize them more efficiently than do the cells of embryos of comparable age (Walbot et al 1972; Sussex et al 1973). Somewhat similar
Embryogenesis and Physiology of Growth of Embryos
381
One prevalent notion about the role of hormones results have been obtained for the comparative biosynthetic activities of the embryo and suspensor in the suspensor is that in the early stages of develcells of T. majus (Bhalla, Singh and Malik 1981a). opment, the embryo does not bear the burden of Calculations based on the DNA content and rates of synthesizing the hormones necessary for its growth, RNA synthesis in suspensor cells of P. coccineus at but that this function is taken up by the suspensor different stages of embryogeny have shown that the (Cionini et al 1976; Bennici and Cionini 1979; Yeung template activity of these cells in the late heart- and Sussex 1979). A role attributed to gibberellins is shaped or early cotyledon stage embryos is higher that, in the suspensor of P. vulgaris, the hormone than is expected of the diploid gene copies present increases the translational activity and maintains in them (Clutter et al 1974). A clear expression of the the protein level in the embryo. The role of the susprotein synthetic activity of the suspensor geared to pensor in enhancing translational activity was embryo nutrition is seen in the transient accumula- shown by a decrease in protein synthesis in tion of the storage proteins, legumin and vicilin, in embryos detached from the suspensor and cultured the suspensor cells of globular embryos of Viciafaba (Brady and Walthall 1985; Walthall and Brady 1986). even before these proteins begin their massive, A convincing demonstration of the effect of gibberellins in enhancing transcriptional activity of embryo-specific accumulation (Panitz et al 1995). The focus on the synthesis of growth hormones suspensor cells is the hormone-induced formation by the suspensor and their relevance for the func- of RNA puffs and the increased RNA synthesis in tion of this organ represents a major and essentially certain regions of the polytene chromosomes of P. novel addition to our knowledge of its physiology coccineus (Nagl 1974; Forino et al 1992). and biochemistry during the last 20 years. Auxins The significance of the presence of other growth (Alpi, Tognoni and D'Amato 1975; Przybyllok and hormones in the suspensor remains largely Nagl 1977), gibberellins (Alpi, Tognoni and unknown. In Tropaeolum majus, changes in the conD'Amato 1975; Alpi et al 1979; Picciarelli et al 1984; centrations of peroxidase in the suspensor and the Picciarelli and Alpi 1987; Piaggesi et al 1989; embryo are believed to ensure the regulation of Picciarelli, Piaggesi and Alpi 1991), cytokinins endogenous auxin diffusing from the suspensor (Lorenzi et al 1978), and ABA (Perata, Picciarelli into the embryo. This correlation is somewhat typand Alpi 1990) have been identified in the suspen- ical of organs that synthesize or utilize IAA for sors of P. coccineus and other plants. The suspensor growth (Singh, Bhalla and Malik 1979). of P. coccineus has 10 times more gibberellin activity In conclusion, it is clear that the suspensor is an than the heart-shaped embryo and contains as early embryonic organ that is programmed for termany as six different gibberellins (Picciarelli and minal differentiation and, except in a few cases, it Alpi 1986; Piaggesi et al 1989). The high concentra- does not contribute any cells to the sporophytic tion of gibberellins is reflected in the ability of a generation of the plant. Yet, it cannot be naively cell-free extract prepared from homogenized sus- assumed that the primary function of the suspenpensors to synthesize precursors of GA, as well as sor is to orient the embryo in close proximity to the GAj, GAy and GA8, from radioactively labeled sub- source of nutrition in the embryo sac or in the strates (Ceccarelli, Lorenzi and Alpi 1979,1981a, b). ovule. Important findings about the structure of In the heart-shaped embryo of P. coccineus, the walls of suspensor cells implicate these cells in cytokinins are also present in higher concentrations the absorption of metabolites from the surroundin the suspensor cells than in the cells of the ing tissues and in the transport of metabolites to embryo proper; a complete reversal of the status of the embryo. The cells of the suspensor, and espeboth gibberellins and cytokinins occurs in the sus- cially their polytene chromosomes, are clearly pensor and embryo at the cotyledon stage (Alpi, involved in transcriptional activity, although the Tognoni and D'Amato 1975; Lorenzi et al 1978). physiological role of the gene products syntheChanges in the auxin content of the suspensor have sized is not understood. Finally, the accumulation drawn little attention, although it appears that the of growth hormones by the suspensor has genersuspensor contains higher amounts of auxin than ated great interest in elucidating the possible funcdoes the embryo (Alpi, Tognoni and D'Amato 1975; tion of hormones in promoting embryo growth. Przybyllok and Nagl 1977). Very little ABA is pre- Despite these advances in our understanding of sent in the suspensor of P. coccineus during the ini- the suspensor, further studies are needed to gain tial stages of its development, but some increase, new insights into the physiological and molecular with two peaks, is noted subsequently (Perata, bases for its terminally differentiated state and for Picciarelli and Alpi 1990). its interaction with the embryo. The future direc-
382
tions in the study of the suspensor are briefly considered by Yeung and Meinke (1993) in a thoughtful essay. 6. NUTRITION OF THE EMBRYO To a large extent, continued development of the embryo depends upon its interaction with the milieu of the embryo sac, the endosperm, and the cells of the ovule surrounding the embryo sac. Thus, in contrast to the static image conjured by pictures of embryos in isolation, it is reasonable to assume that developing embryos maintain a channel of communication with the external cellular environment. This traffic has in turn generated the reprise of a cherished objective of experimental plant embryologists, namely, to identify the substances that provide the nurture and nutriments to developing embryos, but this goal remains largely unattained. (i) Nutrient Supply to the Zygote It is noted that the primary endosperm nucleus generally begins to divide ahead of the zygote, thus ensuring that a mass of cells charged with storage products is in place when the zygote begins to divide. However, few would agree that the nutrition of a specialized cell like the zygote is, in the final analysis, grounded on the advance performance of another cell of a different genotype. This has led to a renewed interest in the ultrastructural analysis of the embryo sac, because an elucidation of the mechanism of zygote nutrition logically requires a consideration of the structure of the embryo sac before fertilization. It now appears that in a broad sense, the structural modifications of the embryo sac devoted to the nutrition of the egg persist after fertilization or find their counterpart in the embryo sac of the fertilized egg. It is likely that several other modifications of the embryo sac and its surrounding tissues for the nutrition of the female gametophyte described in Chapter 6 are carried forward unchanged after fertilization and have the same primary consequences for the zygote.
Zygotic Embryogenesis
and Palmer 1984; Chamberlin, Horner and Palmer 1993a), Arabidopsis (Mansfield and Briarty 1991), and Butomus umbellatus (Fernando and Cass 1996), or adjacent to the degenerating nucellus in wheat (Morrison, O'Brien and Kuo 1978). In Vicia faba, invaginations from the embryo sac wall are formed close to the zygote, as well as at the chalazal end opposite the surviving cells of the nucellus, referred to as the nucellar cap (Johansson and Walles 1993b). In an investigation to trace autoradiographically the fate of 14C-labeled photosynthates in the ovules of soybean, it was found that at the zygotic stage the label was concentrated over the wall projections of the embryo sac and in the hypostase at the chalazal end. The interesting conclusion drawn from this study - that two pathways of solute transport, a primary one from the outer integument to the base of the zygote and to the central cell, and a secondary one through the hypostase, operate to supply nutrients to the zygote - obviously bolsters a role for the wall invaginations in the process of nutrient supply (Chamberlin, Horner and Palmer 1993a). Zygote nutrition is a problem with wide ramifications and of considerable interest; the stimuli and nutrients that this cell normally receives from the rest of the plant body deserve to be characterized.
(ii) Embryo-Endosperm Relations From the time embryologists first began to understand the relevance of the endosperm in seed formation, the function of nurturing the embryo has traditionally been assigned to this tissue. Although evidence in favor of the role of the endosperm as a nurse tissue for the developing embryo is more circumstantial than experimental, the idea has gained great credibility over the years. The absence of the endosperm in mature seeds of some plants known to produce this tissue at early stages of ovule development seems to indicate that the endosperm is not a permanent source of nutrition but is apparently consumed by the growing embryo. In seeking structural adaptations in the proembryos that mediate the transfer of nutrients from the endosperm, as discussed in the previous section, much attention has centered on wall proliferations from the suspensor cells. The invaginations from the outer wall of the suspensor cells in contact with the endosperm, as seen in Capsella bursa-pastoris (Schulz and Jensen 1969, 1974),
The number of investigations that encompass the structure of the embryo sac wall after fertilization is comparatively small, and invariably all refer to the presence of wall projections akin to transfer cells. The projections originate from the inner wall of the embryo sac at the micropylar end close to the Diplotaxis erucoides (Simoncioli 1974), Phaseolus coczygote in Euphorbia helioscopia (Gori 1977), cottoncineus (Yeung and Clutter 1979), Vigna sinensis (Hu, (Schulz and Jensen 1977), soybean (Tilton, Wilcox Zhu and Zee 1983), and Glycine max (Chamberlin,
Embryogenesis and Physiology of Growth of Embryos
Homer and Palmer 1993a), and from the inner wall of the embryo sac at the micropylar or chalazal pole, as seen in pea (Marinos 1970a; Harris and Chaffey 1986), Haemanthus katherinae (Newcomb 1978), Medicago sativa (Sangduen, Kreitner and Sorensen 1983a), and Vigna sinensis (Hu, Zhu and Zee 1983), seem in particular to invite considerable speculation about the possible functions of the imaginations in nutrient transport into or from the endosperm. One investigation has shown that as the zygote of G. max is phased into a multicellular embryo, there is a sequential development of invaginations in at least four different sites in the embryo sac wall in addition to some wall ingrowths that appear in the cells of the inner integument and nucellus. This observation has drawn attention to the possibility that the wall of the embryo sac functions both as a sink and a source of apoplastic transport to nourish the growing embryo (Folsom and Cass 1986). In view of the postulated linkage between wall ingrowths and short-distance transport of solutes, it does not seem unreasonable to find evidence for the presence of ingrowths on the walls of proembryos (Dute, Peterson and Rushing 1989) or the walls of the embryonic cotyledons (Johansson and Walles 1993a). It is relevant to point out that a plasma membrane almost invariably follows the contour of the wall projections; this may not be coincidental, since the membrane greatly increases the surface area of the invaginations. Occasionally, concentrations of various organelles in the vicinity of the wall proliferations have been described with the implication that they serve to increase the active transport of solutes across the plasma membrane. However, the possible direction of the flow of solutes remains unclear because there is nothing to preclude the absorption of nutrients from the endosperm into the suspensor or the transport of metabolites from the degenerating suspensor into the endosperm. A sense of uncertainty prevails also about the direction of solute flow through the invaginations that arise from the inner wall of the suspensor cells. The timing of the direct absorption of endosperm nutrients by the embryo may also be worth emphasizing at this point. Direct absorption generally happens only after the suspensor has degenerated to the point that individual cells are separate from one another. The direct utilization of the endosperm by the growing embryo is indicated by the appearance of a conspicuous zone of destroyed endosperm cells around the cotyledons, as in dicots (Erdelska 1980,1985,1986), or ahead of
383
the tip of the scutellum and coleoptile, as in cereals (Smart and O'Brien 1983; Jones and Rost 1989b). In a similar way, the accumulation of legumin and vicilin in the endosperm cells surrounding the embryo of Vicia faba and their subsequent decay may be looked upon as processes involved in the supply of nutrients to the growing embryo (Panitz et al 1995). Significant increases in the cationdependent phosphatase activity during tobacco seed development as detected biochemically and by ultracytochemical localization methods have been attributed to the active transport of assimilates to the endosperm and embryo (Mogensen and Pollak 1983; Mogensen 1985). To the extent that the cytoplasm of lysed cells can serve as a source of metabolites, the expanding growth of the embryo sac after fertilization has profound effects on the nutrition of the embryo. Expansion of the embryo sac is evident mostly at its micropylar or chalazal end when cells of the nucellus or integument begin to disintegrate and release free cytoplasm and nuclei; some of the newly liberated organelles are even incorporated into the endosperm (Pacini, Cresti and Sarfatti 1972; Norstog 1974; Schulz and Jensen 1974; Pacini, Simoncioli and Cresti 1975; Hardham 1976). In Capsella bursa-pastoris, cells of the residual intact chalazal nucellus enlarge to form a tissue (chalazal proliferating tissue) that protrudes into the embryo sac. The cells of this tissue actively incorporate precursors of both DNA and RNA and accumulate poly(A)-RNA during the period of their active growth. However, by the time the embryo attains the globular stage, degeneration in the chalazal proliferating tissue is initiated in the cells close to the embryo sac and continues until all the cells have broken down (Pollock and Jensen 1967; Schulz and Jensen 1971; Raghavan 1990b). The general interpretation of these various observations is that proliferations on the embryo sac wall aid in the passage of cytoplasmic nutrients from the lysed ovular cells into the central cell. (iii) Nutrient Supply from the Mother Plant
It is difficult to find structural evidence for the supply of nutrients from the vegetative parts of the plant to the growing embryo. From physiological experiments it is known that seeds that accumulate large quantities of storage reserves in the embryo or in the endosperm act as powerful sinks for metabolites from other parts of the plant. With legume seeds, the use of 14CO2, which is incorporated into photosynthates, has given the most con-
384
vincing demonstration of the accumulation in the embryo of sugars manufactured in the leaf. Considering the amount of storage reserves synthesized by developing seeds, the vascular system of the ovule might be expected to play a role in translocating precursors to the embryo or to the endosperm. In developing pea seeds, it has been found that by the time storage protein synthesis is initiated in the cotyledons, most of the endosperm has been consumed. Storage protein synthesis coincides with a marked increase in the amount of vascular tissues in the ovule and funiculus, as well as a marked increase in the number of phloem transfer cells in the latter (Hardham 1976). However, it is uncertain whether nutrients transported through the vascular tissues from the vegetative parts of the mother plant are involved in the nutrition of the embryo or are transformed into storage products. An investigation that traced the fate of 14C-photosynthate into the embryo sac of soybean at the zygote stage was referred to earlier. Extension of this study has provided evidence for the flow of labeled photosynthate through the vascular system into the micropylar, chalazal, and lateral poles of the embryo sac for the nurture of developing globular to heart-shaped embryos (Chamberlin, Horner and Palmer 1993a). Yaklich et al (1992) have shown that the outermost layer of cells of the endosperm of soybean (considered analogous to aleurone cells) is reinforced by the generation of a significant number of Golgi vesicles and an abundance of ER coincident with the period of storage protein accumulation in the embryo. This observation has suggested the possibility of the translocation of photosynthates from the seed coat to the aleurone cells and their secretion into the milieu of the embryo for purposes of nutrition and storage protein synthesis. The main ideas and results that have flowed from several laboratories have been presented here to show that embryos developing in vivo employ different strategies for their nurture. These strategies can be related to specific stages of embryogenesis, with the wall ingrowths' having a prominent role at the zygote or proembryo stage. An immunochemical localization of high levels of a GTP-binding protein (G-protein) in developing embryos of Arabidopsis has fostered the argument that nutritional relations of embryos involve a source-sink interaction (Weiss, Huang and Ma 1993). Because G-proteins are involved in a variety of signal transduction pathways, growth hormones that pass into embryos from various sources can be considered as good candidates for signals transduced by these proteins. The nutrition of embryos is a subject that
Zygotic Embryogenesis
inevitably involves whole-plant physiology, ultrastructure, and molecular biology, with an array of interacting variables impinging on the problem, so only general conclusions are possible at this stage. As described in the following section, it seems easier to obtain information about the chemical nature of the regulatory substances necessary for embryo growth from in vitro culture than from an assault on embryos growing in vivo. 7. EMBRYO CULTURE
The cultivation of embryos excised from ovules and seeds under aseptic conditions in a medium of known chemical composition constitutes the basic methodology of embryo culture. By growing embryos outside the environment of the ovule it is possible to identify the special nutrient or hormonal requirements essential for their continued growth, differentiation, and morphogenesis. Ideally, one would hope to follow progressive embryogenesis in vitro in a defined medium, starting with a zygote, but in a true sense this aim is still a distant goal. Most of the work has been done on the culture of relatively mature and differentiated embryos excised from quiescent seeds. The logic behind this was the hope that a knowledge of the nutrient requirements for the growth of excised seed embryos under controlled conditions would provide clues about the nature of the storage materials utilized during seed germination. Another trend seen in this area was toward the culture of progressively younger embryos with a view to determining their changing nutritional requirements. From a theoretical angle, embryo culture draws attention to the degree to which growth and behavior of embryos of different ages may be regulated exogenously by nutritional or hormonal substances or by physical modifications of the culture environment; from an applied perspective, the technique offers a means of obtaining seedlings from inviable hybrids and overcoming dormancy of seeds.
(i) Seed Embryo Culture
Embryo culture has a long history dating back to the work of Hannig (1904), who obtained transplantable seedlings from embryos excised from the seeds of Raphanus sativus, R. landra, R. caudatus, and Cochlearia danica (Brassicaceae) grown in a medium containing macronutrient salts and cane sugar. Since the publication of this work, the subject has developed rapidly, generating a long list of plants whose embryos have been successfully cultured
Embryogenesis and Physiology of Growth of Embryos
under aseptic conditions in defined media as a means of providing useful information on the physiology of embryo growth. Rather than deal encyclopedically with these studies, the account given here is in large part based on the work done on a few selected species to elucidate the requirements for growth and differentiation of embryos in culture. Carbon and Nitrogen Nutrition. Although supplementation of the culture medium with a carbon energy source is not essential for inducing the growth of isolated seed embryos, it is the general experience that the growth and survival of embryos are markedly enhanced by its presence. A caveat should be added here: Even with the ideal sugar in the medium, the growth of embryos may be influenced by other factors, such as the genotype of the donor plant (Dunwell 1981). At the morphogenetic level, addition of sucrose to the medium has been shown to promote root growth or the growth of both shoot and root primordia of cultured embryos (Rietsema, Satina and Blakeslee 1953; Buffard-Morel 1968). Embryos of Datura stramonium (Rietsema, Satina and Blakeslee 1953) and pea (Davis 1983) require a higher sucrose concentration for optimum root growth than that required for optimum shoot growth, and that may partly explain the reason why a concentration of sucrose effective in enhancing root growth inhibits shoot growth. The favorable effect of dextrose on the root growth of embryos of Cocos nucifera is dependent upon the composition of the mineral-salt medium, because a high nitrogen-containing medium like Murashige-Skoog medium is particularly effective in this respect (Del Rosario and de Guzman 1976). When one realizes that shoot and root primordia are already laid down in the embryo at the time of culture, it becomes clear that the role of sugar in the morphogenesis of embryonic organs has to be viewed in a limited context. To the extent that supplementation of the medium with various nitrogen compounds tells us something about the amino acid metabolism of cultured embryos, a few generalizations have emerged. One is that embryos are able to grow moderately well in a medium containing nitrates or ammonium salts as the sole source of nitrogen. Secondly, the amide glutamine has proved to be an efficient source of nitrogen for the growth of embryos of several plants (Paris et al 1953; Rijven 1956; Matsubara 1964). Thirdly, mutual antagonism and synergism exist between different amino acids in the growth of cultured embryos. For example, a mixture of 20 amino acids put together in the proportions present in casein hydrolyzate is as effective
385
as casein hydrolyzate in promoting growth of preheart-shaped and early heart-shaped embryos of D. innoxia and D. stramonium. This occurs despite the fact that the mixture contains readily utilized, partially utilized, and even relatively toxic compounds. Moreover, favorable effects of the complete amalgam are not reproduced by incomplete mixtures of amino acids or by single amino acids (Sanders and Burkholder 1948). The growth of embryos of cereals such as wheat, barley, and maize is inhibited by the simultaneous addition of lysine and threonine to the medium; the inhibition is alleviated by methionine, homoserine, or homocysteine (Bright, Wood and Miflin 1978; Green and Donovan 1980). Since these observations indicate that embryos are starved for methionine and its metabolic intermediates by feedback inhibition of enzymes of the aspartate pathway, they are of interest for possible selection of feedback-sensitive mutants. Hormonal Effects. Following the discovery of plant hormones, especially auxins, gibberellins, and cytokinins, concerted efforts have been made to study their effects on the growth and morphogenesis of cultured embryos. The general conclusion is that although the addition of hormones modifies the growth of cultured embryos, there is no evidence to indicate the need for a specific hormone for the growth of embryos of a wide range of plants. As far as the effects of hormones on the growth of embryonic organs are concerned, they are quite similar to those displayed by the corresponding organs of a seedling plant. As seen in cultured embryos of Capsella bursa-pastoris, the effect of IAA is manifest in the promotion of growth of the radicle at low concentrations; the inhibition of growth of the radicle, hypocotyl, and shoot at intermediate concentrations; and callus initiation at high concentrations (Raghavan and Torrey 1964). However, it appears that globally the most effective auxin for inducing callus growth in cultured embryos is 2,4-D and, as will be seen in Chapter 16, in many instances calluses thus produced provide the starting material for the induction of somatic embryos. Among the various parameters of auxininduced callus growth on cultured embryos, the genotypic effect of the donor plant has come under intense scrutiny. In oat (Cummings, Green and Stuthman 1976), wheat (Maddock et al 1983), and Poa pratensis (McDonnell and Conger 1984), generic differences account for the ability of cultured embryos of more or less the same age to produce differentiating cultures; in maize, the response of embryos of the same and different ages to 2,4-D is
386 dependent on the particular genotype (Green and Phillips 1975). Varietal differences, as well as the growing conditions of the plant from which embryos are excised, seem to determine the course of auxin-induced regeneration of callus from cultured barley embryos (Bayliss and Dunn 1979). The organ that differentiates the callus in response to auxin application appears to vary in embryos of different species, as well as in embryos of different ages of the same species. In mature oat embryos, the radicle seems to be the most responsive organ and the scutellum undergoes necrosis, whereas whole young embryos are transformed into a callus without any necrosis of tissues (Cummings, Green and Stuthman 1976). In maize, irrespective of the age of embryos at excision, calluses are most successfully induced from the scutellum (Green and Phillips 1975). This organ is also the most proliferative in immature embryos of barley, whereas growth in the more mature embryos occurs from the mesocotyl-radicle region (Bayliss and Dunn 1979; Dale and Deambrogio 1979). The ease of callus regeneration from specific organs of cultured embryos does not mean that other organs of the embryo are incapable of regeneration; apparently, in embryos of different plants, cells of certain organs with an increased potential for uncontrolled growth possess a selective advantage and tend to overgrow cells of other organs. Contrary to the general experience that roots are relatively insensitive to GA, the hormone promotes root elongation and the production of lateral roots in cultured embryos of C. bursa-pastoris. If one considers the effects of GA on embryos of different ages, older embryos are seen to produce roots with abundant laterals whereas younger embryos form roots with fewer laterals (Veen 1963; Raghavan and Torrey 1964). Cultured embryos of two different ages of cotton also display differences in GA sensitivity, such as the enhancement of cotyledon maturation and cell elongation in immature embryos, and the acceleration of cell elongation, cell division, and axis growth in older embryos (Dure and Jensen 1957). These observations are in accordance with a speculative view that GA effectively induces growth in embryos in which the meristems are sufficiently well developed. The limited information available on the morphogenetic effects of cytokinins concerns kinetininduced inhibition of root growth in cultured embryos of C. bursa-pastoris. The degree of inhibition depends on the age of embryos and the conditions of culture. In cultures grown in complete darkness, relatively low concentrations of kinetin,
Zygotic Embryogenesis
which inhibit root elongation in heart-shaped embryos, are marginally inhibitory in slightly older embryos and are without effect in torpedo-shaped and mature embryos. In light-grown cultures, kinetin inhibits root elongation in embryos of all ages but promotes precocious leaf expansion and callus growth (Veen 1963; Raghavan and Torrey 1964). An indirect role for an auxin-cytokinin mixture in inducing the growth of embryos is seen in the observation that seed embryos of Colocasia esculenta (taro; Araceae) develop into plantlets only when they are cultured in the presence of the endosperm. Indicative of a role for hormonal substances of endospermic origin in the growth of the embryo, it was found that NAA and 6-methylaminopurine can replace the endosperm effect on cultured embryos (Nyman et al 1986,1987). From this brief survey, it appears that despite the different growth modifications induced by hormones in cultured seed embryos, there is no evidence to demonstrate or exclude the involvement of a specific hormone in the morphogenesis of embryos. This indicates that mature embryos excised from quiescent seeds have a sufficient supply of endogenous hormones for the initiation of growth when they encounter a relatively simple nutrient medium, a carbon energy source, the normal composition of the atmosphere, and a favorable temperature. As will be seen, a role for hormones in the growth of immature embryos and proembryos is more explicit than their role for mature embryos. (ii) Culture of Immature Embryos and Precocious Germination
During the early period of embryo culture investigations, Dieterich (1924) discovered that whereas mature embryos grown in culture generally produce seedlings that look normal in most respects, cultured immature embryos inexplicably skip the later part of the embryogenesis program and grow into rudimentary, weak seedlings. These malformed seedlings were referred to as "kiinstliche Fruhgeburt" and the process whereby the embryo becomes competent to respond to germination cues before it has fully played out its maturation program is now well known as precocious germination. This apparent dichotomy in the growth patterns of mature and immature embryos has repeatedly been noted in subsequent years by various investigators and has invited frequent comment. In developmental terms, during precocious germination, the normal process of embryogenesis that would lead to a period of arrested
Embryogenesis and Physiology of Growth of Embryos
growth of the embryo in the seed resulting in quiescence or dormancy, is replaced by an alternate program that leads to germinative growth. In most plants there is a vulnerable period during embryogenesis when embryos germinate precociously upon transfer to a culture medium; those cultured earlier than this period fail to grow, whereas older embryos germinate normally. One approach employed to investigate the causes of precocious germination and its control is by manipulation of the medium composition. This is based on the view that immature embryos are already predisposed to evolve into full-term embryos when they complete their development within the ovule, but upon excision and culture, they are prevented from doing so by the limitations imposed by the nutrient medium. The first noteworthy attempt at manipulation of the culture medium to control precocious germination utilized barley embryos as the test material. Normal growth of mature barley embryos cultured in an agar medium containing mineral salts and sucrose is characterized by the enlargement of the scutellum and the development of the first leaf and root within their respective sheathing structures, whereas precocious germination of immature embryos involves the formation of short, spindly seedlings. If the medium is additionally fortified with casein hydrolyzate, tomato juice, or other natural plant extracts, precocious germination is forestalled in the immature embryos and an embryogenic program is reinstated (Kent and Brink 1947). The effectiveness of casein hydrolyzate in preventing precocious germination of barley embryos suggested a role in the process for one or more components of this nutrient mixture, such as the inorganic phosphate, sodium chloride, or amino acids. When the growth of immature embryos of barley in media to which the three groups of components of casein hydrolyzate were added singly or in combination was monitored, it was found that amino acids and sodium chloride, mixed in the same proportion in which they are present in casein hydrolyzate, effectively duplicate the activity of casein hydrolyzate in suppressing germination. Surprisingly, immature embryos completed the embryogenic program in a medium containing sucrose or mannitol at concentrations providing an osmotic pressure equal to that of casein hydrolyzate; this supports the view that inhibition of precocious germination results from the high osmotic pressure of the medium and is not due to a specific nutrient requirement satisfied by casein hydrolyzate (Ziebur et al 1950). The tendency of immature barley embryos to germinate
387
precociously is diminished to a considerable degree by high intensities of light, moderately high temperatures, and reduced oxygen tension in the culture milieu (Norstog and Klein 1972). An arresting observation by Norstog (1972b) that precocious germination of cultured immature barley embryos is inhibited by ABA is relevant to this discussion. This result, which pointed to a role for endogenous ABA as an inhibitor of precocious germination during normal seed development, has been confirmed in experiments with embryos of Brassica napus (Crouch and Sussex 1981; Wilen et al 1991), Gossypium hirsutum (Choinski, Trelease and Doman 1981), Triticum aestivum (Triplett and Quatrano 1982; Morris et al 1985; Hess and Carman 1993), Glycine max (Ackerson 1984a; Eisenberg and Mascarenhas 1985), Oryza sativa (Stinissen, Peumans and de Langhe 1984), Hordeum vulgare (Morris et al 1985), Medicago sativa (Xu, Coulter and Bewley 1990), Zea mays (Rivin and Grudt 1991), and Linum usitatissimum (Wilen et al 1991). Results of a typical experiment are shown in Figure 13.9. More definitive evidence favoring a role for ABA in the control of precocious germination of embryos has come from other investigations, lhle and Dure (1970,1972) examined the activity of certain proteases involved in the mobilization of stored reserves of cotyledons of cotton embryos as an event connected with germination and showed that culture of immature embryos triggers precocious germination as well the development of enzyme activity by premature transcription of mRNAs. Because removal of the embryo from the ovule promotes precocious germination, it appeared that the ovule tissues might supply to the embryo some regulatory molecules that prevent germination. This was found to be the case with the discovery that addition of an aqueous extract of the ovule or ABA to the medium halts precocious germination of immature embryos and the development of enzyme activity in them. In attempts to correlate the levels of ABA in the reproductive tissues to the patterns of growth of cultured embryos, it was found that embryo axes excised from mature soybean seeds containing low levels of ABA germinate soon after culture, whereas those from immature seeds with high levels of endogenous hormone do not germinate or require a long lag period before germination. When the endogenous ABA is removed from immature embryos by washing them in water or by drying them in the pod, embryos germinate precociously (Ackerson 1984a, b). According to King (1976), embryos excised from immature wheat grains containing high concentra-
388
Zygotic Embryogenesis
100-
10' 3
ABA concentration (mol-m ) in culture medium Figure 13.9. Germination index of developing embryos of Medicago sativa (stages III, youngest, to IX, oldest) on a medium containing 1% sucrose and different concentrations of ABA. Germination of embryos of stages Ill-V was inhibited by low concentrations of the hormone. (From Xu and Bewley 1991. J. Expt. Bot. 42:821-826, by permission of Oxford University Press.)
tions of ABA germinate infrequently as compared to the high germination percentages of mature embryos excised from dry grains containing low levels of the hormone. High concentrations of ABA are found in the embryos of Phaseolus vulgaris when they are just phased out of the stage vulnerable to precocious germination, and again at seed maturity as they begin to desiccate (Morris 1978; Hsu 1979). We see here a curious situation in which a sustained supply of ABA from the ovule tissue checks the germination instincts of the developing embryo and commits it to a pathway of maturation and dehydration inside the seed. At this time, the embryo is liberated from the chemical influence of the ovule tissue and germinates in a normal fashion when confronted with favorable conditions. Investigations on the protein metabolism of germinating embryos also cast some light on the role of ABA in the regulation of precocious germination. As reviewed in Chapters 14 and 15, synthesis of a variety of proteins, including LEA proteins, storage proteins, and defense-related proteins, is slowed down or inhibited in precociously germinating embryos, whereas in the presence of a concentration of ABA in the medium appropriate to suppress precocious germination, cultured embryos accumulate these pro-
teins to the same extent as those growing in situ. Other investigations have, however, questioned a role for ABA in preventing precocious germination. In addition to ABA, osmotically active agents such as mannitol and sucrose can also inhibit precocious germination and, in some cases, a high osmotic pressure is more effective in this role than exogenous ABA. Moreover, the results of experiments to determine the levels of endogenous ABA in embryos cultured in high osmoticum are consistent in showing that inhibition of precocious germination is not necessarily accompanied by a rise in ABA levels (Finkelstein and Crouch 1986; Morris et al 1988, 1990; Barratt et al 1989). Whereas both ABA and osmoticum effectively prevent precocious germination of immature embryos of M. sativa, proteins that are the hallmarks of embryos developing in situ in the pod are synthesized only in the presence of the osmoticum (Xu, Coulter and Bewley 1990). Another work showed that ABA and osmoticum are differently effective at different stages of embryogenesis and that when both agents are present, the combined inhibitory effect is greater than the sum of the individual effects (Xu and Bewley 1991). These various observations are in line with the view that ABA is not the universal regulator of precocious germination of embryos; even when a distinct effect of the hormone is documented, rather than being the primary effector of the process, ABA might inhibit precocious germination secondarily by blocking water uptake by the embryos. One of the points to be considered in assessing the control of precocious germination is the importance of the constraint imposed by the seed environment and the associated seminal tissues. Because two potential regulatory factors, namely, ABA and osmoticum, have been studied with inconclusive results, it seems likely that other specific signals supplied by the ovular tissues may predispose the embryo to complete its maturation program rather than opt for the germinative mode of development. Release of the embryo from the influence of the maternal environment has been achieved by premature desiccation of the seed; subsequent hydration of the seed switches the embryo from a developmental to a germination mode. As in precocious germination, embryos of prematurely desiccated seeds are most prone to germinate during a narrow developmental window restricted to the period before they naturally desiccate and pass into the quiescent or dormant state (Kermode and Bewley 1985,1988). At the biochemical level, premature drying suppresses developmental protein synthesis in the embryo and
Embryogenesis and Physiology of Growth of Embryos
induces germination protein synthesis; a switch in the synthesis of mRNAs from those coding for developmental proteins to those encoding germination proteins also occurs during premature drying and germination (Dasgupta and Bewley 1982; Misra and Bewley 1985a; Oishi and Bewley 1992; Xu and Bewley 1995a). Precocious germination and germination after premature drying are two faces of the same coin, reflecting on the need for completion of the embryo maturation program for successful seedling development. Furthermore, one may assume that cultured immature embryos germinate precociously because they have lost access to the pool of endogenous signals from the maternal tissues and premature drying depletes the maternal tissues of the signals necessary for inducing embryo maturation. (iii) Culture of Proembryos
During their growth within the ovule, proembryos, including the zygote, are heterotrophic in nature and are dependent not only upon the metabolites present in their cells, but also upon those diffusing from the endosperm and the surrounding maternal tissues. From the embryo culture perspective, this suggests that the nutritional requirements for the culture of proembryos are bound to be more exacting than those for immature embryos, and mature, differentiated embryos. In addition, the technical difficulties involved in isolating uninjured zygotes and proembryos have made it impossible to study the initiation and progress of cell divisions in culture in all but a few species. To follow the course of development in this field, the topic will be reviewed here under the following headings: (a) effect of nutrients of endospermic origin; (b) effect of the physical parameters of culture; (c) role of hormones; (d) role of the suspensor; (e) zygote culture; and (f) culture of fertilized ovules and embryo sacs.
389
pantothenic acid) and an assortment of organic substances (adenine, glycine, and succinic acid). Although still smaller embryos (200-500 [im long) failed to grow, or grew feebly in this medium before turning into undifferentiated callus, a dramatic recrudescence of growth occurred in these embryos when the medium containing vitamins and organic addenda was further supplemented with nonautoclaved coconut milk. This was attributed to the presence of a hormonal factor, designated as the "embryo factor," which was later obtained in a form sufficiently pure enough to promote growth of embryos in a dilution as low as 1:19,000 on a dry weight basis as compared to 1:110 for the crude coconut milk (van Overbeek, Siu and Haagen-Smit 1944). In later years, these studies spawned several successful attempts to culture proembryos of other plants by the use of plant extracts of endospermic or nonendospermic origin, most notably of maize by the use of water extracts of date, banana, wheat gluten hydrolyzate, and tomato juice (Kent and Brink 1947); D. tatula by diffusates from young seeds of Lupinus luteus and old seeds of Sechium edule (Cucurbitaceae) (Matsubara 1962); Cucumis sativus, Cucurbita maxima, and C. moschata by extracts of their respective endosperms (Nakajima 1962); and Secale cereale by use of its own endosperm as a nurse tissue (Stefaniak 1987).
Nutrient media containing undefined natural plant extracts cannot be regarded as chemically defined until the essential functions of the latter can be duplicated by, and restricted to, known compounds. An important step toward the formulation of a completely synthetic medium for the culture of proembryos of barley is noteworthy in this context. Norstog (1961) showed that undifferentiated barley embryos, which generally fail to survive in White's mineral-salt medium, can be successfully cultured by supplementing the medium with coconut milk. The smallest embryo thus cultured had probably no more than 100 cells (about 60 fim long), and its Effect of Nutrients of Endospermic Origin. An survival and subsequent growth in culture were early lead in the successful culture of proembryos enhanced by the addition to the medium of glutacame from the work of van Overbeek, Conklin and mine or a mixture of amino acids besides coconut Blakeslee (1942) on the physiology of growth of milk. In a further modification of this medium, a embryos of Datura stramonium. From previous rigorous definition of growth requirements for the studies of these investigators, there was some gen- culture of barley embryos as small as 90 ^m was eral agreement that it was possible to grow heart- achieved with the demonstration that a phosphateshaped and torpedo-shaped embryos of D. enriched White's medium at an optimum pH of 4.9 stramonium to the plantlet stage in a mineral-salt fortified with glutamine and alanine (400 mg/1 medium enriched with a mixture of vitamins (nico- each) as major nitrogen sources and lesser amounts tinic acid, thiamine, pyridoxine, ascorbic acid, and of leucine (20.0 mg/1), tyrosine, phenylalanine, cys-
390 teine, and tryptophan (all at 10.0 mg/1) is an effective substitute for coconut milk (Norstog and Smith 1963). The effectiveness of endosperm extracts of Cucumis sativus, Cucurbita maxima, and C. moschata
in promoting the growth of their respective embryos, mentioned earlier, is similarly matched by manipulation of the hormonal and organic additives to the medium, because a mixture of IAA, 1,3diphenylurea, and casein hydrolyzate supports growth of embryos to the same extent as does a medium containing the embryo factor (Nakajima 1962). The conclusion from these results is that growth induction in cultured proembryos by endosperm extracts is a measure of the effect of specific chemical components of the endosperm. Effect of the Physical Parameters of Culture.
Interest in the effect of the physical conditions of culture on the continued growth of proembryos was stimulated by the common observation that the amorphous liquid endosperm in which the proembryos are constantly bathed has a high osmolarity. Measurements have shown that the ovular sap surrounding proembryos has a very low osmotic potential value (negative), which substantially decreases (becomes more negative) with embryo maturity (Ryczkowski 1960; J. G. Smith 1973; Yeung and Brown 1982). This has led to the view that the osmotic pressure of the medium might play a role in promoting the growth of proembryos in vitro. The importance of this parameter of the culture medium in the control of the growth of proembryos has not been critically tested, although the balance of evidence is in favor of adjustment of the medium to be isotonic with the ovular sap for this purpose. As indicated earlier, the embryo factor from coconut milk fully met the requirements for continued growth of pre-heartshaped embryos of D. stramonium. However, another work showed that whereas mature embryos of this species grow in the complete absence of sucrose in the medium, earlier-stage embryos require progressively higher concentrations of sucrose. For example, a medium containing 8%-12% sucrose is found optimal for continued growth and differentiation of globular and pre-heart-shaped embryos. Interestingly enough, these embryos grow to the same extent in a medium containing 2% sucrose plus enough mannitol to be isotonic with 8%-12% sucrose (Rietsema, Satina and Blakeslee 1953). From this it appears that the essentiality of sucrose in embryo culture medium is to be accounted for wholly or in part as an osmotic stabilizer rather than as a carbon energy
Zygotic Embryogenesis
source. The successful culture of proembryos of Capsella bursa-pastoris (Rijven 1952; Veen 1963), D. tatula (Matsubara 1964), Linum usitatissimum (Pretova 1974), and Triticum aestivum (Fischer and Neuhaus 1995) has also been linked to the provision of a high osmolarity in the medium. Although success has been sporadic, a new small-size-limit in proembryo culture in a medium of high osmolarity has been attained with 25^45-^m-diameter proembryos of Arabidopsis (Wu et al 1992). The need to gradually reduce the osmolarity of the medium during the growth of embryos without sequential transfer from one medium to another has resulted in two technical modifications of the culture system. For the culture of embryos of C. bursa-pastoris as small as 50 ium in diameter, a continual change in the osmolarity of the medium was obtained by using two media of different compositions, solidified in juxtaposition in a Petri dish. During growth in the medium of high osmolarity, the embryo was subjected to continually changing osmotic conditions due to diffusion from the medium of low osmolarity (Monnier 1976a). A culture system comprised of two agar layers, with the top layer having a higher osmolarity than the bottom layer, was used to obtain continued growth and differentiation of 8-36-celled proembryos of Brassica juncea. Proembryos were embedded in the top layer, whose osmolarity decreased during culture (Liu, Xu and Chua 1993a). Although no illuminating ideas as to how osmolarity of the medium regulates the cells of the tiny embryos to engage their intrinsic ability for division and growth have been generated by these investigations, an effective control of the flow of metabolites and inorganic ions into the cells of the embryo by the osmoticum appears a distinct possibility. Role of Hormones. Because of the widespread use of hormones as adjuvants in tissue culture media for growth induction in organ, tissue, and cell cultures of plants, it is not unexpected that proembryo culture has come under the spell of hormones. Embryos of C. bursa-pastoris have figured prominently in the investigations using hormones. Although some workers (Rijven 1952; Veen 1963) had successfully cultured heart-shaped embryos in an inorganic liquid medium of high osmolarity secured by the addition of 12%-18% sucrose, a later work showed that it is possible to culture these same embryos in an agar-solidified mineral-salt medium supplemented with 2% sucrose. Growth of still smaller embryos (up to about 55 \xm long) was secured by fortifying this medium with IAA,
Embryogenesis and Physiology of Growth of Embryos
kinetin, and adenine sulfate (Raghavan and Torrey 1963). The addition of the hormone mixture eliminated the need for a medium of high osmolarity, suggesting that the picture of osmotic regulation of the growth of proembryos is somewhat misleading. Nonetheless, the need for a properly balanced mixture of hormones for the growth of these embryos is in a large part met by an osmoticum or by an increase in the concentration of major salts by a factor of 10. The more the mineral salts of the medium became fully competent to support the growth of proembryos, the more they appeared likely to be the governing factor in regulating proembryo growth. From a comparative analysis of different inorganic salt media, Monnier (1976a) found that the high-salt medium of MurashigeSkoog is superior for the culture of proembryos of C. bursa-pastoris. A concurrent experiment also revealed the interesting point that excellent growth of proembryos ensued when they were cultured in a modification of this medium with increased levels of Ca+ and K+ and a decreased concentration of NH4NO3. In other approaches, enhanced growth of proembryos of C. bursa-pastoris was obtained by supporting them on a bed of polyacrylamide instead of agar (Monnier 1975) or by increasing the partial pressure of O2 in the medium (Monnier 1976b). Although C. bursa-pastoris has dominated the field of proembryo culture involving manipulation of hormonal and other additives to the medium, a successful report of culture of proembryos of Linum usitatissimum by kinetin (Pretova', 1986) is noteworthy. As experiments in the culture of early division phase proembryos and of the single-celled zygotes of these plants are contemplated, simultaneous changes in the chemical and physical conditions of culture offer a means of channeling the work in new directions. Role of the Suspensor. As discussed earlier, in the division of labor of the whole embryo, the suspensor with its wall projections for solute absorption and transport is presumed to channel exogenous stimuli for the growth of the organogenetic part of the embryo. This is especially true at the proembryo stage, when the suspensor is at the prime of its life before degradative changes set in. This has raised the question as to whether it is possible to demonstrate a causal relationship between the presence of the suspensor and the growth of proembryos in culture. The simple approach to interfering with the influence of the suspensor is to sever its connection with the embryo and follow the growth of the embryo in culture. This is no
391
doubt an exacting task; it has been performed with only a few species, perhaps involving a limited number of embryos. From experiments using this approach, it was found that continued growth of proembryos of Eruca sativa (Corsi 1972), Phaseolus coccineus (Nagl 1974; Yeung and Sussex 1979), and C. bursa-pastoris (Monnier 1984) is more enhanced in the presence of an attached suspensor than in its absence. Growth of embryos of P. coccineus is promoted even by the presence of a detached suspensor kept in close proximity in the medium (Yeung and Sussex 1979). A string of experiments with embryos of P. coccineus involving supplementation of the medium with growth hormones and determination of the growth hormone levels separately in the embiyo and suspensor has provided indirect evidence showing that the alleged suspensor effect is due to the production of GA and cytokinins. Whereas intact pre-cotyledonary stage embryos grow in a mineral-salt medium, the chance of survival of embryos of the same age deprived of the suspensor is very remote unless GA is present in the medium. GA is, however, inimical to the growth of suspensorless post-cotyledonary stage embryos as compared to intact embryos of the same age grown in a hormone-free medium (Cionini et al 1976; Yeung and Sussex 1979). These responses of embryos to GA are correlated with a falling level of endogenous GA in the organogenetic part and a rising level in the suspensor of the pre-cotyledonary stage embryos. As the suspensor begins to undergo senescence in the post-cotyledonary stage embryos, there is a sharp decline in the concentration of GA in the suspensor and a corresponding increase in its titer in the embryo proper (Alpi, Tognoni and D'Amato 1975). Likewise, addition of zeatin or zeatin riboside fosters the growth of suspensorless pre-cotyledonary stage embryos of P. coccineus, but post-cotyledonary stage embryos scarcely respond to the cytokinins (Bennici and Cionini 1979). As prototypes of hormones that promote cell divisions, cytokinins like zeatin and its riboside might be expected to be present in the organogenetic part of the post-cotyledonary stage embryos in a quantity sufficient to promote embryo growth. Indeed, this has been found to be the case, in a manner that is comparable to the presence of the same cytokinins in the suspensor part of the pre-cotyledonary stage embryos (Lorenzi et al 1978). These results provide a strong argument for a role for hormonal gradients from the suspensor in promoting growth of proembryos. This concept might serve as baseline infor-
392
Zygotic Embryogenesis
Figure 13.10. Culture of isolated zygotes of Zea mays, (a) Section of a polarized zygote 3 hours after culture. (b)Zygote showing first division 19 hours after culture; arrowheads point to the newly formed cell wall, (c) Section through an
embryo formed from a zygote 11 days after culture, showing the embryo proper and suspensorlike part. N, nucleus. Scale bars = 10 urn (a, b) and 50 urn (c). (From Leduc et al 1996; photographs supplied by Dr. C. Dumas.)
mation for formulating a medium for growing in vitro still smaller embryos, and even the zygote,
favorable effect of the nurse tissue may conceivably depend upon the release of undefined nutrients into the medium, these results indicate that, analogous to the culture of isolated proembryos, requirements for regenerating plants from zygotes are likely to be exacting and varied. The potential of the maize system for the transfer of foreign genes has been underscored by preliminary experiments showing transient expression of microinjected reporter genes.
of P. coccineus.
Zygote Culture. Up to the present time, only two successful attempts at zygote culture have been reported. Zygotes extruded from embryo sacs of barley developed into embryolike structures that yielded fully fertile plants when cocultured with embryogenic microspores of barley (Holm et al 1994). Through the use enzyme digestion and microdissection, zygotes of maize excised from embryo sacs were similarly induced to form fertile plants by simulating typical in vivo embryonic development (Leduc et al 1996). The unexpected finding was that the same nurse tissue of embryogenic microspores of barley was very effective in inducing continued growth and divisions of explanted maize zygotes (Figure 13.10). As the
Culture of Fertilized Ovules and Embryo Sacs. Since embryos must be isolated in order to be cultured, the difficulty of extracting them from the inner sanctum of the ovule could well be one of the reasons why there are so few reports of successful culture of proembryos. From this perspective, the culture of ovary, ovule, and embryo sac has proved to be particularly rewarding for the analysis of the
Embryogenesis and Physiology of Growth of Embryos
393
growth requirements of proembryos and zygotes of transfer of embryos to a medium of a different certain plants that have hitherto defied attempts at composition is a relatively simple protocol for excision and culture. As in embryo culture studies, regenerating plants in vitro from the zygote (M61, much effort in ovule culture has gone to devising an Matthys-Rochon and Dumas 1993,1995). appropriate medium. Some early investigations reported only limited success in the culture of isolated ovules of Papaver somniferum containing the 8. GENERAL COMMENTS zygote or two-celled proembryo in a mineral-salt Embryogenesis in angiosperms, in whatever conmedium containing 5% sucrose; rapid growth of the text of study, leads ultimately to the analysis of celluenclosed nascent sporophyte was obtained in a lar mechanisms using histological, cytological, medium supplemented with casein hydrolyzate, genetic, biochemical, molecular, and tissue culture yeast extract, or kinetin (Maheshwari 1958; techniques. The distinction between dicots and Maheshwari and Lai 1961). Other examples, such as monocots offers little difficulty because differences in Zephyranthes sp. (Kapoor 1959), Capsella bursa-pas-embryogenic development between the two classes toris (Lagriffol and Monnier 1985), Hordeum vulgare relate mainly to the division patterns that herald his(Topfer and Steinbiss 1985; Holm et al 1995), togenesis and organogenesis. Although different patTriticum aestivum (Zenkteler and Nitzsche 1985; terns of early embryogenesis have been described, Comeau et al 1992), and Zea mays (Schel and Kleft there is little or no indication of the functional mean1986), can also be cited to illustrate the successful ing of these division pathways. Admittedly, many application of ovary and ovule culture to promoting materials that are needed by the zygote to commence the growth of the enclosed zygote or proembryo. divisions are stockpiled or synthesized by this cell Accelerated growth of the ovular tissues in rela- during a period of growth preceding division; howtion to the growth of the enclosed embryo or vice ever, there is no consistent pattern of distribution of versa may also determine the success or failure of these materials in the zygote that foreshadows the ovule culture. This problem has been analyzed in a major parts of the embryo. The zygote shows maxigeneral consideration of the need to formulate a mum dependence for its growth on the endosperm, medium for the culture of cotton ovules at the both for organic nutrients and for cell division and zygote stage. Working from the experience of pre- morphogenetic stimuli. vious investigators, who obtained perfectly During embryogenesis, a variety of important respectable seeds with underdeveloped embryos changes occur, commonly including activation of from cultured proembryo stage ovules of cotton, many metabolic pathways associated with the proStewart and Hsu (1977) found that postfertilization duction of new tissues and organs. This is reflected ovules of cotton can be cultured with a reasonable in the progressive transition from heterotrophy of degree of success in a medium containing low con- the proembryos to autotrophy of mature and difcentrations of IAA, kinetin, and GA along with 15 ferentiated embryos. Interaction between the mM NH4+. The key to the combined use of hor- embryo and surrounding endosperm has a role in mones and ammonium ions was the observation determining the acquisition of metabolic compethat the addition of hormones alone enables the tence. Taken as a whole, the physiological and cytoovules to attain a normal size, whereas NH4+ pro- logical properties of the suspensor serve to motes growth and division of the zygote. illustrate one of the few possible ways hetIntermediate systems between embryo culture erotrophic nutrition of the embryo is sustained. and ovule culture have been developed for some Several technical innovations suggest that methplants to obtain growth in vitro of zygotes and ods are at hand to monitor the requirements for proembryos. Such systems include the culture of growth of the fertilized egg and proembryos outzygote-containing embryo sacs of Z. mays sur- side the confines of the embryo sac. In successful rounded by the nucellus with or without a block of experiments designed to nurture the zygote and the endosperm (van Lammeren 1988b; Campenot proembryo in defined media, it might even be poset al 1992; M61, Matthys-Rochon and Dumas 1993, sible to identify the nutrient substances provided 1995; Leduc, Matthys-Rochon and Dumas 1995) by the endosperm and those present in the zygote and the culture of embryo sacs of Cucumis sp. con- and proembryo. Availability of this information taining proembryos and free nuclear endosperm will permit more specific conclusions to be drawn (Custers and Bergervoet 1990). Initial culture of about the biosynthetic pathways activated during embryo sacs of Z. mays enclosing zygotes or two- progressive embryogenesis in angiosperms. celled proembryos combined with a subsequent
Chapter 14
Genetic and Molecular Analysis of Embryogenesis 1. Genetic Control of Embryogenesis (i) Embryo Abortion in Wide Hybrids (ii) Radiation-induced Mutations (iii) Spontaneous Embryo Mutants 2. Gene Expression Patterns during Embryogenesis (i) Diversity and Changes of mRNA Population during Embryogenesis (ii) Gene Activity during Fertilization and Early Zygote Divisions (iii) RNA Metabolism in Developing Embryos (iv) Protein Metabolism in Developing Embryos 3. Characterization of Genes Essential for Embryogenesis (i) Embryo-lethal Mutants (ii) Genetic Control of Embryonic Pattern Formation 4. Expression of Genes for Housekeeping and Structural Proteins (i) Genes for Glyoxylate Cycle and Related Enzymes (ii) Heat-Shock Protein Genes (iii) Expression of Other Genes 5. Expression of Genes Related to Desiccation Tolerance and Dormancy (i) LEA Genes from Cotton and Other Dicots (ii) LEA Genes from Wheat and Other Monocots (iii) Is Embryo Desiccation Developmentally Regulated by ABA? (iv) Genes Involved in Dormancy Regulation and Seedling Growth 6. Regulation of Expression of Defense-related Genes (i) Lectin Gene Expression in Embryos of Cereals and Grasses (ii) Expression of Defense-related Genes in Embryos of Other Plants 7. General Comments
394
Completion of embryogenesis in angiosperms results in the formation of a full-term embryo, which is an assemblage of organs whose cells and tissues are specialized for functions essential for the initiation and continuation of the sporophytic life of the plant. As seen in Chapter 13, our current understanding of embryogenesis owes much to extensive descriptive accounts gained from light and electron microscopic studies and experimental work relating to the culture of embryos. Based on these accounts, it is convenient to consider embryogenesis in terms of a series of developmental processes aimed at creating a recognizable morphological structure. Most important among these processes are the establishment of the precise spatial organization of cells derived from the first few rounds of division of the zygote, the differentiation of cells of common origin to create diversity in the resulting embryo, and biochemical preparations for embryo maturation, desiccation, and dormancy. As part of the mature seed, the embryo unleashes another developmental program to initiate germination; however, since germination-related events are not considered in this book, the topic will not be a part of the discussion here. The developmental episodes during embryogenesis have little in common, yet there is a common background because they are at the root of many multidimensional processes involving different levels of interactions. In arbitrary terms we can conceptualize embryo development by stating that the critical aspects of morphogenesis and differentiation are completed relatively early in the time course of embryogenesis, whereas changes connected with maturation are relegated to later periods. The primary objective of this chapter is to cast the major events of embryogenesis within a genetic and molecular perspective and to explain them in
Genetic and Molecular Analysis of Embryogenesis
terms of changes in gene expression; an overview of the research efforts that reflect on the role of genomic information during maturation, desiccation, and dormancy of the embryo is also the subject matter of this chapter. Some of the questions that will be considered here are, How is the basic body plan of the embryo generated and what are the mechanisms of cell specification that occur during the initial stages of embryo development? How do the different parts of a developing seed interact in the formation of a mature embryo? What are the genes and processes that regulate embryogenesis? What are the molecular changes that direct polarity and organ differentiation in the embryo? How does the embryo acquire desiccation tolerance? It must be admitted at the outset that complete answers to these questions and to others of equal significance are not presently at hand, but investigators have begun to tackle such questions. Moreover, considering the fact that on a conservative estimate at least 20,000 diverse genes are believed to be active during embryogenesis, the problem of sorting through them to identify genes active at landmark stages should prove to be a formidable task. Although plant embryo development differs in many striking and fundamental respects from embryogenesis in animal systems, some comments about the approaches used to study the controlling mechanisms in the construction of the animal from the egg are appropriate here. The backbone of research in the early part of this century was provided by classical genetic studies using mutant genes as markers to study pattern formation, cellular determination, and the origin of diversely committed cell lineages. An alternative strategy was to isolate mutants that lead to developmental defects often reflected in anomalous patterns of morphogenesis and differentiation. The abnormalities were attributed to lesions in information flow, namely, the flow from genes to proteins. Blossoming of the field of developmental genetics led to the discovery of lethal mutants, which die after completing parts of the embryogenic program. Lethal mutants were of particular interest to embryologists because the availability of organisms whose development was arrested at relatively early stages of embryogenesis made it possible to identify genes that perform essential functions during this critical period. The advent of molecular biology has led to investigations on the expression of specific genes during embryogenesis by a multifaceted approach involving nucleic acid hybridization, construction of stagespecific cDNA libraries, analysis of the nucleotide sequence of developmentally regulated genes, identification of their protein products, and transforma-
395
tion of eggs and embryos with cloned genes and antisense RNAs. Following the lead of animal developmental biology, genetic analyses, aided mostly by mutational dissection and combined with molecular characterization at different levels of complexity, have been increasingly applied to study the fundamental genomic strategies utilized by developing angiosperm embryos. Indeed, few organs of the angiosperm plant body have recently benefited as much as embryos from the integrated genetic and molecular approach of investigators. However, identification and characterization of genes or gene products associated with embryogenesis have barely begun. Useful complements to the contents of this chapter are the reviews by Dure (1985), Meinke (1986, 1991a, b, c), Sheridan and Clark (1987), Goldberg, Barker and Perez-Grau (1989), Reinbothe, Tewes and Reinbothe (1992a), Lindsey and Topping (1993), de Jong, Schmidt and de Vries (1993), Goldberg, de Paiva and Yadegari (1994), Jurgens, Torres Ruiz and Berleth (1994), and Jurgens (1995), from which this chapter has drawn heavily to achieve a broad coverage and to present the newer developments in their proper historical perspective. 1 . GENETIC CONTROL OF EMBRYOGENESIS
A pervading role for genes in embryo development in angiosperms has been inferred in the past from analyses of embryos following wide hybridizations between plants, from analyses of the administration of ionizing radiations to developing ovules, and from investigations of spontaneous mutants with altered patterns of embryo development. Although studies described in minute detail the morphological and cytological abnormalities in developing embryos, they did not attempt to analyze gene action at the molecular level in arresting embryo growth. Moreover, many cases of cytological damage were not supported by breeding experiments designed to help understand the control of the aberrant morphology at the genetic level. Despite these limitations, results have provided important clues as to how the genotype of participating gametes determines the growth of embryos and how perturbation of the genome of certain cells affects the replicative activity and the final form of embryos. (i) Embryo Abortion in Wide Hybrids
One of the key problems in present-day horticultural and breeding practices is that crosses between distantly related species and genera of
396
plants (wide hybridizations) have not yielded many agriculturally beneficial hybrids. From a genetic point of view, this is primarily due to the fact that in attempts to transfer beneficial alien genes across interspecific and intergeneric barriers, many deleterious genes that are difficult to eliminate also find their way into the hybrid genome. The resulting disturbance in the equilibrium between growth of the maternal tissues, embryo, and endosperm leads to embryo lethality and seed collapse. In an extensive review of embryo-endosperm relations in plants, Brink and Cooper (1947b) listed six types of genetic disturbances that cause embryo abortion in hybrids. These are (a) enforced self-fertilization in a normally cross-fertilizing species; (b) crosses involving parents of the same ploidy level but belonging to different species and genera; (c) crosses between parents of the same species but of different ploidy levels; (d) crosses between parents of different species and ploidy levels; (e) unbalanced chromosome conditions, especially aneuploidy in the endosperm; and (f) maternal genotypes causing embryo arrest irrespective of the source of the male gamete. For effective hybridization, union of the egg and sperm from members of the same species with identical chromosome numbers appears a basic necessity, and even a slight asymmetry at the species level or chromosomal level can cause loss of embryos. Numerous examples of ovule development in wide hybrids have been described in previous publications (Raghavan 1976b, 1977b) and so only representative cases will be considered here. What are the kinds of disturbances that occur during wide hybridization and how do they modify the normal development of the ovule and result in embryo abortion? Of interest is the fact that the deleterious genes affect not only the growth of the embryo but also that of the endosperm. Although accounts of the ontogeny of hybrid ovules differ in details, there are some close parallels in the patterns of embryogenesis, and virtually all investigators seem to agree in identifying an early stage when developmental anomalies in the embryo begin to surface. This was first shown in reciprocal crosses between Oenothera biennis X O. muricata and O. biennis X O. lamarckiana, in which the initial
growth of embryos was not appreciably affected; however, the growth slowed down later, and eventually the embryo did not advance beyond a few cells. Development progressed further in some afflicted embryos, but survival was generally erratic and ovules enclosing underdeveloped embryos matured into the familiar, shrunken, aborted seeds (Renner 1914). Quantitative, time-
Zygotic Embryogenesis
course analysis of the growth of embryos and descriptive accounts of cellular changes associated with embryo abortion have further characterized the effects of genomic instability in hybrids. Seed development in a cross between Trifolium ambiguum and T. repens exemplifies the time-course changes associated with embryo growth in normal and hybrid plants. In selfed T. ambiguum, the zygote embarks on the embryogenic pathway soon after fertilization and proceeds through the globular, heart-shaped, torpedo-shaped, and mature stages of embryogenesis with precise regularity. Early growth of the hybrid embryo almost betrays its developmental fate, since it outgrows the selfed embryo during the first two days after pollination in the number of cells formed. Growth of the embryo subsequently slows down and, in the hybrid seeds, the embryo produces at best a tightly knit mass of about 1,000 cells and does not differentiate beyond the heart-shaped stage (Williams and White 1976). Hybrid embryos generated from T. semipilosum X T. repens cross maintain the same rate of growth as selfed T. semipilosum embryos up to four days after pollination, but their further growth bogs down by the globular stage (White and Williams 1976). Although embryos from T. repens X T. medium hybrid exhibit a regular pattern of growth up to four or five days after pollination, mitotic activity is slower in the hybrid embryos than in embryos of both parental species (Kazimierska 1978). The view that the action of the deleterious genes is initiated in the hybrid at the time of fertilization or shortly thereafter seems to be supported by this observation. In crosses involving different species of Phaseolus, a gradation is found in the stages at which deleterious genes become active and block embryo growth. In P. vulgaris X P. lunatus and the reciprocal cross, embryos cease to grow at the pre-heart-shaped and four-celled stages, respectively (Mok, Mok and Rabakoarihanta 1978; Rabakoarihanta, Mok and Mok 1979). Embryos of hybrids from P. vulgaris X P. acutifolius and P. coccineus X P. vulgaris crosses begin development in the normal way and reach the adult form, but gene action prevents embryo maturation (Mok, Mok and Rabakoarihanta 1978; Shii et al 1982). A further extreme stage of hybrid embryo development is seen in P. vulgaris X P. coccineus cross, in which embryos escape the action of deleterious genes and attain maturity (Shii et al 1982). In wide hybridization studies using various lines of Hordeum vulgare, Secale cereale, and Triticum aestivum as female par-
ents and different or closely related genera of Poaceae as pollen donors, embryos succumb to the
Genetic and Molecular Analysis of Embryogenesis
action of alien genes mostly at the globular stage (Zenkteler and Nitzsche 1984). These observations suggest that embryo abortion in wide hybrids is clearly linked in some way to postfertilization events in the embryo sac. Among the questions of general importance concerning gene action is, what are the cellular changes in the hybrid embryos that lead to their demise? In a comparative ultrastructural study of embryo development in selfed Hibiscus costatus and H. aculeatus and H. costatus X H.furcellatus hybrids,
Ashley (1972) found that the embryogenic development of the selfed parent begins with a pronounced shrinkage of the zygote and a marked segregation of organelles toward the chalazal end. In contrast, the zygote of the hybrid hardly changes physically or cytologically. However, it is with regard to the subsequent fate of the organelles that the hybrid embryo presents in a microcosm a good example of the damage inflicted by deleterious genes. Whereas the cells of the selfed embryo are endowed with a generous supply of organelles, cells of hybrid embryos are characterized by a paucity of organelles and extensive vacuolation, so that cells of the two types of embryos do not bear even the most tenuous resemblance to each other. Early degenerative changes observed in embryos of Medicago sativa X M. scutellata hybrid are con-
fined to the suspensor, which is much more reduced than suspensors of intraspecific hybrids in the number of cells formed and in the distribution of organelles in each cell. Gene action also affects the functional capacity of the suspensor, since the characteristic wall invaginations mediating in metabolic interchange at the cell surface are absent (Sangduen, Kreitner and Sorensen 1983b). These results provide a different perspective of the development of the hybrid embryo due to cellular and physiological malfunctions. As described in Chapter 12, by and large, the most general problems relating to the genetics of wide hybrids are posed by the deterioration of the endosperm and by successful embryo rescue operations. In this sense, the alien genes do not cause lethality of hybrid embryos and their effects can be considered as transient. Unfortunately, genetic analysis of hybrids to identify the individual genes that account for embryo or endosperm abnormalities has not been possible because of the many genetic changes found in wide crosses. (ii) Radiation-induced Mutations Because ionizing radiations such as X-rays and y-rays are powerful mutagenic agents, they have
397
been used in attempts to induce chromosomal and gene mutations in embryos. This approach is based on the premise that a dose of radiation that destroys the integrity of cellular DNA will cause abnormalities in the orderly development of the embryo and that these can be related to gene action. However, this objective has hardly been realized, as investigations on the effects of ionizing radiations on embryos have yielded a bag of mixed results. A large body of research on the irradiation of seeds resulted in many chromosomal mutations and in the formation of seedlings with altered morphology of the vegetative parts. Irradiation of ovules with immature embryos led to the formation of full-term embryos with abnormal organs, whereas irradiation of ovules enclosing very young embryos resulted in embryo lethality associated with abnormal cellular changes. What was lacking in these studies was a focus on specific genetic changes in the embryos so that their relationship to gene function could be understood. Impressive radiobiological contributions to embryogenesis have come from work with Hordeum vulgare. In these studies, spikes of an inbred line were exposed to acute X- or y-irradiation and subsequent growth of embryos under normal conditions was monitored. Although the level of radiation used did not lead to lethality, the histological heterogeneity of embryos made them prime targets for radiation; indeed, the relatively homogeneous early stage proembryos were less radiosensitive than the midstage proembryos and differentiating embryos. Depending upon the developmental stage of embryos at the time of irradiation, X-rays induced profound cell and tissue abnormalities that led to aberrant growth of the scutellum, root, and shoot. These disturbances can be traced to differential cell behavior in specific organs, where certain cells are killed, others begin to divide at a low amplitude, and still others continue to divide at the same rate as before. Neither the radiosensitivity of specific embryonic organs nor the drastic nature of the cellular damage seems to be anticipated during the early period of recovery of embryos because many cell generations intervene between the actual exposure and the occurrence of microscopically detectable symptoms. A particular abnormality resulting in anomalous coleoptile formation, designated as "cleft coleoptile," was extremely stage specific and appeared only following irradiation of the spike at the midproembryo stage; irradiation before or after the critical period failed to produce the cleft. In the midproembryo stage, the scutellum and shoot are most radiosensitive, whereas in the differentiating
398
embryos, these organs become insensitive and the ontogenetically younger root moves up on the list (Eunus 1955; Mericle and Mericle 1957, 1961). These observations are in line with the view that genes for the initiation of specific organs are activated at different times during embryo ontogeny, with those for coleoptile growth being very stage specific. Results similar to those obtained with H. vulgare have been reported for embryo abortion and developmental abnormalities in other plants following irradiation. In Arabidopsis, by timing the stage at which flowers or ovules are X-irradiated, it is possible to control organ formation in embryos at any developmental period between fertilization of the egg and embryonic tissue differentiation. Embryo abortion occurs when flowers harboring ovules with egg, zygote, or two-celled proembryo are irradiated. The transition of the young embryo from radial to bilateral symmetry marks a sharp turning point in embryogenesis as this is the most radiosensitive phase. Among the abnormalities resulting from irradiation at this period are changes in the number and orientation of cotyledons, which lead to syncotyly, tricotyly, anisocotyly, tetracotyly, and the formation of multiple embryos. Surprisingly, full-grown embryos escape the debilitating effects of a dose of radiation that is effective in embryos at earlier stages (Reinholz 1959). Genetic alterations thus cause the loss of growth control by cells only at a vulnerable period, and the result is abnormal morphogenesis in the developing embryos.
Zygotic Embryogenesis
establish new centers of mitosis as a prelude to forming adventive embryos. The requirement for a gene product for continued growth of embryos that was suggested by the irradiation experiments has been confirmed in recessive embryo-lethal mutants of Arabidopsis obtained by EMS mutagenesis and TDNA insertional mutagenesis following transformation with Agrobacterium tumefaciens. In a few
mutants, arrest of embryo growth is followed by abnormal divisions of the suspensor, resulting in a multi-tiered structure with as many as 150 cells in place of a normal suspensor of 6-8 cells (Marsden and Meinke 1985; Yadegari et al 1994). When the development of a mutant embryo is followed in concert with that of the suspensor, morphogenetic defects in the embryo appear to precede visible defects in the suspensor. This observation is consistent with the view that disruption of the growth of the embryo can cause anarchy in the suspensor, leading to uncontrolled growth. In these mutants, designated as suspensor (sus), defects are noted in the development of the embryo protoderm (SMS-2 and s«s-3) and in the division of cells of the inner tier of the globular embryo (sus-1, sus-2, and sws-3) (Schwartz, Yeung and Meinke 1994). Another embryo-defective mutant is twin (twn), which produces a high percentage of twin seedlings (Figure 14.1); here, secondary embryos capable of surviving seed desiccation and germinating into seedlings originate by continued division of cells of the suspensor (Vernon and Meinke 1994). Two alternative predictions made by genetic models are that SUS genes are required for normal embryo development Indirect evidence in support of the view that a and suspensor identity and that these genes are gene product is required for continued growth of involved in the production of an undefined signal the organogenetic part of the embryo but not of the for inducing the same morphogenetic changes. In suspensor has been obtained from experiments any case, it is clear that continued growth of the susinvolving exposure of immature ovules to ionizing pensor is inhibited by the organogenetic part of the radiations. In general, the suspensor cells of irradi- embryo and that when the inhibitory effect is ated embryos show a tendency to divide under con- removed by mutation or by treatments that arrest ditions that arrest the growth of the embryonal part embryo growth, the suspensor is able to express its itself. Induction of additional divisions in the sus- full developmental potential. In the case of the twn pensor cells is especially rapid when prefertilization mutation, such a simplistic explanation is probably ovules or ovules harboring proembryos are irradi- untenable because the mutation, besides promoting ated. Classic responses of the suspensor to irradia- embryogenic potential of the suspensor, disrupts the tion are seen in Nicotiana rustica (Devreux and normal development of the embryo in various ways. Scarascia Mugnozza 1962) and Capsella bursa-pas- In order to appreciate the mutagenic effects of toris (Devreux 1963), in which the abnormal sus- ionizing radiations on embryos, it is worthwhile to pensors contain a file of 8-17 cells rather than the remember that, in the experiments described, usual 4-7 cells; in Arabidopsis, with unusually long embryos were shielded during irradiation by the suspensors composed of several files of cells ovular tissues and endosperm. To what extent do (Gerlach-Cruse 1969; Akhundova, Grinikh and the abnormalities result from the direct action of Shevchenko 1978); and in Eranthis hiemalis radiation on the embryo itself and how much is (Ranunculaceae) (Haccius and Reichert 1964), in indirect? In experiments in which embryos excised which cells of the suspensor burst into growth and
399
Genetic and Molecular Analysis of Embryogenesis
Figure 14.1. Suspensor development in mutants of Arabidopsis. (a) Initiation of a supernumerary embryo from the suspensor of a mutant seed. The cell adjacent to the basal cell of the suspensor has divided to form an octant stage embryo (e2). (b) Twin embryos in a mutant seed. The embryos are not linked with the suspensor. Scale bars = 20 urn. (From Vernon and Meinke 1994; photographs supplied by Dr. D. W. Meinke.)
from unirradiated cereal grains were transplanted into irradiated endosperm, one finds that the irradiated tissue is endowed with the same mutagenic power as direct irradiation to reduce the survival period of embryos and to cause chromosomal aberrations in their cells (Avanzi et al 1967; Floris, Meletti and D'Amato 1970). Along the same lines, other experiments have shown that the use of irradiated nutrient medium or the addition of irradiated sucrose to the nutrient medium causes cytological and morphological changes reminiscent of mutagenic effects in cultured seed embryos and somatic embryos (Swaminathan, Chopra and Bhaskaran 1962; Ammirato and Steward 1969). Another approach in this research area has taken advantage of the finding that embryo growth is impaired following fertilization of a normal egg with sperm originating from irradiated pollen grains. Despite the fact that mitosis in the nucleus of the generative cell of the irradiated pollen grain is often abnormal, pollen grains germinate and produce pollen tubes, and sperm cells participate in double fertilization in the embryo sac. Cytological abnormalities are common in the cells of embryos developing from the union of one set of normal chromosomes of the egg and one set of irradiated chromosomes of the sperm (Brewbaker and Emery 1962). Irradiation of pollen grains of Lilium formosanum with a high dose of X-rays did not impair pollen germination, pollen tube growth, or production of functional sperm cells, but the treatment induced dominant lethality that results in the failure of normal zygote formation upon fertilization of an unirradiated egg. Although the endosperm
nucleus undergoes a few rounds of division and thus succumbs later than the incipient zygote, the cells generated show more cytological abnormalities than the zygote (Brown and Cave 1953, 1954; Cave and Brown 1954). Comparative studies of the effects of irradiation of the male and female gametophytes on the development of the embryo in Triticum durum showed that irradiated sperm affected embryo development more than an irradiated egg did (Donini and Hussain 1968). Mutations did not result in the immediate death of the zygote, but degeneration of the proembryo due to necrosis of cells was frequently observed after fertilization of T. durum egg with irradiated sperm. The usefulness of these lethals to the study of the developmental effects of mutations is limited because they cannot be used for progeny tests. (iii)
Spontaneous Embryo Mutants
Spontaneous embryo mutants have been identified with precise regularity in some crop plants and are thus highly suitable for studying developmental problems of embryogenesis. Moreover, compared to the lethal mutants generated by radiation treatments, most defective embryo mutants are viable when maintained as heterozygotes. Many spontaneous embryo mutations have been described, ranging from those that impair basic physiological processes of embryos, such as their lapse into quiescence or dormancy, to those that affect biochemical characteristics such as pigment or storage protein accumulation. Mutations that interfere with the quiescence of embryos induce premature embryo germination. This happens during the maturation of the seed when the embryo begins to germinate and form a seedling although the seed is still attached to the mother plant. Such mutants are known as viviparous (vp), and the phenomenon is known as vivipary. Vivipary. Due to its widespread occurrence, vivipary in maize is better understood than in other plants. Not only has the genetic control of vivipary been clearly demonstrated in maize, the system has also proved to be an attractive one for analyzing the various ramifications of gene action in regulating the maturation pathway of the embryo and endosperm in the grain. In maize, at least nine recessive genes (vp-1, vp-2, vp-5, vp-7, vp8, vp-2, w-3, y-9, and al) that can induce vivipary are known and their respective genotypes have been described in detail. Genetic and physiological studies have shown that besides causing premature ger-
400
mination, most viviparous genes interfere with other facets of grain maturation, such as ABAmediated responses, pigment synthesis, and storage protein accumulation. All except one (vp-1) of the nine viviparous genotypes are deficient in ABA; all ABA-deficient mutants except one (vp-8) share an inability to form carotene, with the result that chlorophyll synthesis is also suppressed in the seedlings. Mutation in the vp-1 locus also suppresses the accumulation of anthocyanin in the embryo and in the aleurone cells of the endosperm (Robertson 1955; Neill, Horgan and Parry 1986; McCarty and Carson 1991). Another effect of vp mutation is the failure of embryos to accumulate storage protein globulins encoded by GLB-1 and GLB-2 genes; however, globulin synthesis can be reinstated in vp-5 mutant embryos that are deficient in ABA by exogenous application of the hormone, but this treatment is ineffective in embryos of the vp-1 mutant (Rivin and Grudt 1991). In addition, use of a transposon tagging approach has led to the characterization of mutant alleles in which vp-1 function is altered in dramatic ways, such that the embryo attains a near normal state of dormancy but neither the embryo nor the endosperm develops anthocyanin (McCarty et al 1989a). It is possible that what we see here is the pleiotropic action of a single gene on germination, pigment synthesis, and the hormone level of embryos, but the nature of the metabolic blocks involved, or how blocks in the synthesis of pigments or hormones might induce vivipary, is not known.
Zygotic Embryogenesis
phenotype. Of the several physiological differences between the patterns of development of wild-type and vp-1 mutant embryos, one that may prove useful in the elucidation of the possible role of ABA in vivipary is that viviparous embryos do not desiccate at the time when wild-type embryos begin to undergo dehydration, but continue to increase in fresh weight (Wilson, Rhodes and Dickinson 1973; Neill, Horgan and Parry 1986; Neill, Horgan and Rees 1987). When the different mutants are screened for their sensitivity to ABA inhibition of germination and in vitro embryo growth, embryos of the vp-1 mutant are easily coaxed to grow in culture in the presence of ABA in the medium, indicating that the mutation in some way confers resistance to the inhibitory action of the hormone (Robichaud, Wong and Sussex 1980; Robichaud and Sussex 1986; Sturaro et al 1996). However, this does not appear to be caused by a defect in the actual mechanism of action of the hormone, as the patterns of uptake and metabolism of ABA are remarkably similar in mutant and wild-type embryos (Robichaud and Sussex 1987). As a tentative conclusion from these results, the reduced sensitivity of mutant embryos to growth inhibition by ABA may be considered to involve a stable genetic change in their signal transduction pathway concerned with ABA perception.
A different type of experiment has revealed that interfering with the ABA metabolism of wild-type maize caryopses can induce vivipary. In this experiment, developing ears are treated with fluridone, Are the effects of the VP gene mediated by an inhibitor of carotenoid biosynthesis, and the fate changes in the embryo, in the endosperm, or in of the caryopses is followed. When the treatment is both? Although most routine crosses have yielded applied between 9 and 11 days after pollination, results too ambiguous to permit the assignment of premature plumule elongation characteristic of genetic effects to either the embryo or endosperm, vivipary is found in a high percentage of the carythe use of pollen grains carrying an A-B chromo- opses. Since carotenoids are probable precursors of some interchange in crossing plants heterozygous ABA, an obvious conclusion from this work is that for a VP gene has shown that vivipary is deter- vivipary is due to a lesion in ABA synthesis (Fong, mined solely by the genotype of the embryo. This Smith and Koehler 1983). conclusion has relied on the observation that in all Some interesting insight into the processes that cases, genetically viviparous embryos germinate control premature germination and block in the prematurely whether they are surrounded by a anthocyanin pathway in the embryo has been wild-type or a mutant endosperm in the caryopsis derived from the isolation and characterization of (Robertson 1952). Anthocyanin expression in the the VP-1 gene. McCarty et al (1989b) cloned the VPaleurone, however, is independent of the embryo 1 gene by transposon tagging and showed that it is genotype and results from the coordinate control of a single-copy gene with a 2,500-nucleotide tranthe synthesis of critical enzymes in the endosperm script. RNA blot hybridizations revealed that tranby many genes, including the one that determines scripts of the gene are present in the embryo and vivipary (Dooner and Nelson 1979; Dooner 1985). endosperm tissues of the wild-type caryopses but Since vp-1 is an exceptional mutation that causes not in the caryopses of the mutant stock; consistent vivipary without altering the ABA level of with the effects of the mutation on the embryo and embryos, considerable effort has been made to endosperm, seedling tissues such as the shoot and study the biochemical and molecular basis for its root also do not express the gene. Finally, a surpris-
Genetic and Molecular Analysis of Embryogenesis
ing observation that emerged from RNA blot hybridizations using probes made to genes implicated in VT-2-regulated processes is that expression of C-l, a regulatory gene in the anthocyanin pathway, is selectively blocked in both the embryo and the endosperm of the mutant vp-1 caryopses. Reporter Plasmids:
GUS tni-oUo
i—— Em promoter
35S-Sh-GUS
GUS
jll UMVWSpremour
Effector plasmid: 35S-Sh-Vpi
.SM^I
Vpi cONA
NOS 3
CaHVKSprantour
Reporter Plasmid
Effector Plasmid
GUS Activity pmoth'ujVo'
none
4.27+0.51
35S-Sh-Vp1
587+68
Em-GUS
(137X)
none
2750+320
35S-Sh-Vp1
4170+170
35S-Sh1-GUS (1.5X)
-6 -5 log[ABA]
-4
Figure 14.2. Transactivation of EM gene by VP-1 gene in a maize protoplast assay, (a) Schematic representations of reporter and effector plasmids. Em-CUS plasmid contains 554 bp of the 5'-flanking sequence of the f/Vf gene and the nopaline synthase (NOS) gene fused to GUS gene. The 35S-Sh-GUS plasmid includes the CaMV 35S promoter, the first intron of the maize SH-1 gene, the C4Igene, and the 3'-sequence of the NOS gene fused to GUS gene. The 35S-Sh-Vp1 plasmid has the cDNA clone containing the complete protein coding sequence of VP-1 gene. Maize protoplasts were cotransformed by electroporation with
401
In an attempt to define the interaction between VP1 and C-l genes in controlling the anthocyanin pathway, Hattori et al (1992) found that overexpression of VP-1 protein in maize protoplasts of embryogenic cell origin activates a reporter gene driven by the C-l promoter. At low concentrations of ABA, VP-1 protein also strongly augments the activation of C-l; however, at high concentrations of the hormone, the protein does not stimulate reporter gene expression. Whatever may be the significance of this result in the broader role of the VP-1 gene in caryopsis development, a conclusion that can be drawn is that this gene may control the anthocyanin pathway by translational or posttranslational regulation of C-l; it is, however, a moot question whether VP-1 protein plays an essential role in the ABA-signal transduction pathway. The amino acid sequence predicted from a cDNA clone indicates that the VP-1 gene encodes a 73-kDa protein that has no detectable similarity to known proteins (McCarty et al 1991). Although this protein appears otherwise undistinguished, it is significant that its acidic domain apparently functions as a transcription factor in regulating the expression of other genes and induces strong activation in maize protoplasts of a cotransformed GUS reporter gene fused to the promoter of a presumptive target gene such as the ABA-regulated EM (for early methionine-labeled) gene from wheat embryos. Another observation pertinent to the role of ABA in the maturation program of the embryo is that addition of the hormone causes a synergistic effect on the activation of the target gene (Figure 14.2). The key implication of these results is that proteins encoded by the VP-1 gene are necessary for dormancy induction in the embryos as well as for ABA response, perhaps as transcription factors. Inhibition of precocious induction of germination-specific oc-amylase genes in the aleurone cells of the endosperm provides another setting in which the VP-1 gene might act during caryopsis development (Hoecker, Vasil and McCarty 1995). By combining both activator and repressor functions in the same protein and in cells each of the effector and reporter plasmid DNAs as shown. GUS gene activity is expressed as the mean of three readings ± standard error. Values in parentheses show times activation over basal expression, (b) Synergistic interaction of VP-1 gene with ABA. Protoplasts from the electroporation experiments were cultured in media containing different concentrations of ABA. Squares represent protoplasts electroporated with Em-CUS construct alone and triangles represent protoplasts electroporated with Em-CUS construct together with 35S-Sh-Vp1 construct. (From McCarty et al 1991; ©Cell Press.)
402 of two different genotypes, the VP-1 gene seems to integrate the control of embryogenesis and germination in an unusual way. It was mentioned earlier that the synthesis of globulins in viviparous embryos is related to the actions of the mutant gene and ABA. Northern blot analysis showed that whereas GLB-1 and GLB-2 transcripts are not detected in vp-1 mutant embryos, they are present at a slightly reduced level in embryos of certain viviparous lines deficient in ABA (for example, vp-5). Since the expression of GLB-1 and GLB-2 transcripts is enhanced in the wild-type embryos by ABA, the function of the VP1 gene product could simply be to ensure the progress and completion of an ABA-regulated process in the maturation of the embryo (Kriz, Wallace and Paiva 1989; Paiva and Kriz 1994). Synthesis and accumulation of several LEA proteins in embryos of vp mutants are modulated differently than are those of globulins. The list includes a group of 23-25-kDa LEA proteins that appear in embryos of wild-type and certain ABA-deficient mutants (vp2, vp-5, and vp-7) of maize. The proteins and their mRNAs are also precociously induced by ABA in young wild-type and mutant embryos (Pla et al 1989; Mao et al 1995). In the vp-2 mutant, the pattern of transcript expression of the gene RAB-28 (for responsive to ABA), which encodes a 27-kDa protein, is strikingly similar to that found in wild-type embryos. Although RAB-28 mRNA is not detected in the embryos of vp-1 mutant, it is inducible by low concentrations of ABA (Pla et al 1991). Another LEA protein, MLG-3, is detected in negligible amounts in vp-5 mutant embryos, but the latter accumulate the protein when they are cultured in the presence of ABA. However, culturing vp-1 mutant embryos in ABA does not result in the accumulation of MLG-3, whereas nongerminating vp-1 embryos retained on the ear or cultured in a high osmoticum can accumulate this protein (Thomann et al 1992). The most direct interpretation of these results is that MLG-3 proteins, unlike globulins, have no specific requirement for the VP-1 gene product, but other signals, such as ABA or osmotic changes, contribute to their expression in the mature embryo. On the other hand, the regulation of expression of the EM gene in vp-1 mutant embryos is rigorously dependent upon the presence of ABA in the medium and upon its perception through a mechanism mediated by the VP-1 gene, because no EM mRNA accumulates in mutant embryos of any age growing on the ear, or cultured in ABA, or subjected to an acute osmotic stress (Butler and Cuming 1993; Paiva and Kriz 1994).
Zygotic Embryogenesis
Studies on vivipary in other plants have added few new dimensions to our present understanding of this phenomenon. In mangroves such as Rhizophora mangle, in which vivipary is a natural phenomenon, embryos have a high water content and are generally insensitive to inhibition by ABA in culture (Sussex 1975). This probably facilitates transition of the embryo from a developmental mode to a germination mode while still contained within the seed. The relationship between a low ABA level in the seed and vivipary was confirmed in an ABA-deficient mutant of Arabidopsis in which the seeds were induced to germinate in the fruit in an atmosphere of high humidity (Karssen et al 1983). In other mutants in which embryos maintain a high water content, seeds often showed vivipary without any further treatment (Koornneef et al 1989). This is what one would expect of embryos growing in ovules deficient in ABA. To the extent that water stress in plants is known to cause a rise in the ABA level of cells, it is of interest to note that vivipary is associated with a high water content in the embryos. A general conclusion is that ABA might regulate the metabolism of the embryo during the transitional period between the end of embryogenic development and onset of germination by its effects on the water potential of the cells. Screening a rice genomic library with maize VP-1 gene as a probe has led to the isolation of a gene functionally homologous to the maize gene in transient expression experiments using rice protoplasts; transcripts of the gene are detected in rice embryos as early as 10 days after flowering and decrease toward maturity (Hattori, Terada and Hamasuna 1994). Isolation of genes for vivipary in rice has not hitherto been accomplished and offers great promise in broader studies. Seed Mutants Deficient in Storage Products. Growers have known for a long time that inbred lines of some cultivated crops produce seeds that are genetically defective in the expression of specific storage products in the embryo. Indeed, it was the character of the pea seeds that affects the accumulation of storage products in the cotyledons (wrinkled and green seeds) that Mendel used in the famous experiments on the inheritance pattern involving the independent assortment of more than one gene. It is appropriate that this natural collection of mutant seeds has been studied to clarify the molecular basis for the derangement in storage product accumulation. In pea, the production of wrinkled seeds is governed by two genetic loci, r (for rugosus) and rb,
Genetic and Molecular Analysis of Embryogenesis
403
and it has often been argued that these genes sepa- the embryos of rr seeds is the cloning of the r locus. rately affect starch metabolism of the embryo in a Bhattacharyya et al (1990) used an antibody to manner that results in a reduced starch content the isoform of the starch-branching enzyme to with a concomitant reduction in starch granule size clone a cDNA made to poly(A)-RNA of embryos of and an increase in the proportion of the amylose RR seeds and showed that transcripts encoding content of the starch; alleles at the rb locus also pro- the enzyme are about 10-fold less abundant in foundly affect the growth of the testa. Comparative the mutant embryos than in the normal ones. analyses of the effects of the r locus on the growth Furthermore, the absence of starch-branching and development of embryos of near-isogenic enzyme in the mutant embryos has been attributed round (RR) and wrinkled (rr) lines of peas have to an insertion in the gene encoding the enzyme revealed several physiological changes leading to a and, because of the presence of inverted repeats at derangement in starch metabolism. One invariable its termini, the inserted sequence is considered to change is the overall high fresh weight of embryos be transposonlike. of the mutant line compared to embryos homozyLectins are sugar-binding proteins and glycogous for the R-allele (Hedley et al 1986, 1994; proteins of wide occurrence in plants. Genetic Lloyd, Wang and Hedley 1996). Throughout most defects that interfere with embryo lectin contents of the developmental period studied, cells of are common among cultivated varieties of soymutant embryos also showed a notable effect of the beans. Molecular studies on a lectinless line of soylesion in possessing a higher water content and in bean have attributed the phenotypic condition to a having a higher osmotic pressure than normal virtual absence of lectin transcripts in the genome. embryos (Wang et al 1987b). A genetic effect on cell The transcription lesion in the lectinless phenotype size appeared somewhat late in embryogenesis and is apparently due to a silencing of the lectin gene was reflected in the embryos of the wrinkled line caused by the transposonlike insertion of a 3.4-kb having larger cells than those of the normal round DNA segment (Vodkin 1981; Goldberg, Hoschek line (Ambrose et al 1987). Other side effects of the and Vodkin 1983; Vodkin, Rhodes and Goldberg mutant gene are seen in the lipid content and in the 1983). It is of particular interest that seeds of transcomposition of the storage proteins of the embryos; genic tobacco plants harboring a chimeric lectin generally, mature rr embryos contain more lipid gene containing the 5'-half of the mutant gene and and less of the storage protein legumin than RR its promoter region and the 3'-half of the wild-type embryos (Davies 1980; Coxon and Davies 1982). gene synthesize both lectin mRNA and protein. Some of the parameters of wrinkled seeds, such as This shows that the control region of the mutant starch content and composition and lipid content, gene is transcriptionally competent but that insercould be reproduced in chemically induced tion of the transposonlike element blocks gene mutants of peas (T. L. Wang et al 1990). transcription by silencing the promoter (Okamuro An important insight into the basis for the dif- and Goldberg 1992). How the promoter functions ference in the osmotic potential between RR and rr are impaired by the transposon element remains embryos was gained when it was found that wrin- unknown and should prove to be an exciting area kled seeds have higher levels of free sucrose than of future research. normal seeds (Stickland and Wilson 1983). Since Mutation in a naturally occurring variety of soythe increased sucrose content will lead to a high bean that lacks the a'-subunit of the 7S seed storage osmotic pressure in the cells, questions concerning protein, p-conglycinin, has been found to be caused this phenomenon center on the defect in the con- by a deletion that extends through most of the codversion of sucrose to starch. At the biochemical ing sequences of the gene (Ladin, Doyle and level, the r mutation is believed to lead to a reduc- Beachy 1984). Another soybean seed protein mutation in the activity of the starch-branching enzyme tion includes an inversion that separates the 5'- and (1,4-oc-D-glucan, l,4-a-D-glucan-6-glycosyltrans- 3'-regions of an aberrant glycinin gene, GY-3 (Cho ferase), which results in the complete absence of et al 1989). Because the synthesis of the respective one isoform of this enzyme (Matters and Boyer transcripts predicted in the wild-type embryos is 1982; Smith 1988). This may cause a block in the inhibited in the mutants by lesions in the correpathway of starch synthesis in embryos of wrin- sponding genes, it follows that the mutations affect kled peas and may lead to a consequent accumula- gene action at the transcriptional level. tion of sucrose. A different picture emerges in considering the Among other attempts made to relate metabolic molecular basis of a null allele of a glycinin gene consequences of gene action to starch synthesis in (GY-4) and a mutation in a KTI gene in soybean
Zygotic Embryogenesis
404
KTi +
KIT
A GCTA
AG C T A
-+510
-+480
KTi3
+
KTi3"
+ 478 j | J +498 CTT GAG AGA GTT TCT GAT GAT Leu Glu Arg Val Ser Asp Asp aa!19 aal25
CTT TAG AG—T TTC TGA TGA TGA • • Leu Phe • Ser
b Figure 14.3. Nucleotide and amino acid sequence of Kunitz trypsin inhibitor genes in wild-type (KTi+) and mutant (KTi~) embryos of soybean, (a) DNA sequence ladders represent nucleotides +460 to +520 of the two genes. Nucleotides +481 to +490 of the wild type and the corresponding mutant sequences are shown to the left. Squares show the nucleotides that are conserved between the two genes; asterisks indicate nucleotides that are mutated in the KTi" gene, (b) Nucleotide and amino acid sequence comparisons of KTI genes of wild-type and mutant embryos. Only nucleotides +478 to +498 and translated amino acids 119 to 125 are shown. Arrows refer to the nucleotides mutated in the KTi" gene (shown as asterisks in [a]). The asterisks designate translational stop codons that result from the mutations. (From Jofuku, Schipper and Goldberg 1989. Plant Cell 1:427-435. © American Society of Plant Physiologists. Print supplied by Dr. R. B. Goldberg.)
embryos. The null phenotype is attributed to a point mutation that leads to the elimination of the translation start codon due to a single base change. Although this does not interfere with transcription, mRNA is not effectively retained on the polysomes like the functional mRNA (Scallon, Dickinson and Nielsen 1987). In the mutant defective in the gene encoding KTI, a frame-shift mutation due to alter-
ation of three nucleotides causes drastically reduced accumulation of KTI mRNA and KTI in the embryos (Figure 14.3); it has also been found that the KTI gene is transcribed at the same relative rate in wild-type and mutant embryos (Jofuku, Schipper and Goldberg 1989). Evidently, some posttranscriptional modifications impose restraints on the stability of mRNAs and reduce the prevalence of mRNAs or of the protein products in the mutant embryos. The lectin-nonproducing mutant seeds of soybean have a counterpart in certain genetic mutants of Phaseolus vulgaris. In cultivars like Pinto, mutations of the genes encoding the lectin, PHA, have the effect of producing embryos that are deficient in lectins. At the biochemical level, the effects of the mutation have been interpreted as the consequence of a reduced synthesis of PHA and an equally low level of PHA gene transcription (Staswick and Chrispeels 1984; Horowitz 1985; Chappel and Chrispeels 1986). A partial explanation for the molecular basis of the mutation was provided by the cloning and sequencing of the two PHA genes of Pinto called PDLEC-1 and PDLEC-2. The latter follows very closely the sequence of the genes of the wild type and of PDLEC-1 up to -240 bp relative to the transcription site. Compared to other genes, farther upstream, PDLEC-2 suffers a 100-bp deletion that contains a 60-bp tandem repeat. Since the deleted segment might function as an enhancerlike element, this deletion could account for the low level of expression of PDLEC-2 (Voelker, Staswick and Chrispeels 1986). In support of this hypothesis, it has been found that embryos of transgenic tobacco seeds harboring a PDLEC-2 allele accumulate about 50 times less PHA than embryos of plants transformed with a gene from the normal cultivar (Voelker, Strum and Chrispeels 1987). But this clearly is not the whole explanation for the effect of the mutation, given the lack of knowledge about the causes for the low expression of PDLEC-1.
2. GENE EXPRESSION PATTERNS DURING EMBRYOGENESIS
Included among the genes expressed during embryogenesis are those that encode enzymes for housekeeping activities and mobilization of food reserves, structural proteins of the cytoskeletons and membranes, and storage proteins and embryospecific genes that regulate specific aspects of embryo development. Although we are a long way from characterizing these genes and their products and analyzing the regulatory pathways of gene
Genetic and Molecular Analysis of Embryogenesis
action, current experimental approaches have led to new insights into the molecular processes controlling gene expression pattern during embryogenesis. (i) Diversity and Changes of mRNA Population during Embryogenesis
The modern global view of quantitative changes in the gene expression program during embryogenesis was developed by Goldberg et al (1981b), Galau and Dure (1981), and Morton et al (1983) from reassociation kinetics of nucleic acids of embryos of soybean, cotton, and pea, respectively. Because of the small size of the early stage embryos, which precludes their isolation in ample quantities for preparation of nucleic acids, these studies were undertaken with embryos in advanced stages of development; yet the power of this approach to the study of gene regulation is unmistakably demonstrated in these investigations. As an intermediate between the gene and the protein it encodes, mRNA populations in embryos at different stages of development should provide an overview, down to the details of changing complexity, abundance distribution, and diversity of the genes active during embryogenesis. Hybridization experiments using an excess of mRNA from cotyledon stage (30 days after flowering) and midmaturation stage (75 days after flowering) embryos of soybean with cDNA prepared from corresponding mRNAs have shown stagespecific changes in mRNA complexity, ranging from 15,000 average-sized mRNAs in the cotyledon stage embryos to more than 30,000 in midmaturation stage embryos. Considering the simple morphology of the cotyledon stage embryos, the degree of specification at the level of mRNAs is apparently enormous. In terms of the number of molecules of mRNA present per cell per sequence, it is possible to delineate superabundant (>19,000 molecules), moderately prevalent (550-800 molecules), and rare (3-17 molecules) classes of mRNAs in the soybean embryo cells. Attesting to the diversity in the distribution of the different classes of mRNA is the fact that in the cotyledon stage embryos, mRNAs belong to the moderately prevalent and rare classes, whereas in the later-stage embryos they are predominantly of the superabundant class, with small amounts of the moderately prevalent and rare classes. Interestingly, the superabundant class, which accounts for nearly 50% of the mRNA mass of midmaturation stage embryos, contains only 6-10 diverse mRNAs. Despite the fact that morphological complexity of the embryo increases with
405 development, a surprising finding is that cotyledon stage embryos and midmaturation stage embryo axes and cotyledons contain approximately the same total number (14,000-18,000) of diverse mRNAs. This signifies that in terms of the absolute number of structural genes expressed, there are no detectable changes in gene expression during morphological differentiation of the embryo and within morphologically distinct parts of the embryo (Goldberg et al 1981b). Another approach to the analysis of mRNA changes during soybean embryo development was by the hybridization of kinetic fractions of midmaturation stage embryo-abundant and rare cDNA (selected by repeated cycles of hybridization with homologous mRNA, followed by hydroxyapatite fractionation) to an excess of cotyledon stage and midmaturation stage embryo mRNA. Comparison of the mRNA sequences present in embryos of the two developmental stages showed that almost the entire sequence comprising the mass and diversity of mRNAs of the older embryos is contained in the young embryos. Although this comparison is sensitive only to changes in at least hundreds of rare transcripts, the logical conclusion is that mRNAs of embryos of different ages tend to be qualitatively similar. The interesting possibility from these results is that in soybean the entire set of structural genes that encode proteins of embryogenesis is transcribed at a very early stage, regardless of whether the genes are translated or not. Finally, an intriguing feature of the fate of embryogenic mRNA sequences is that most of them persist throughout development and are stored in the cells of the mature seed embryo (Goldberg et al 1981b). Quantitative changes in the gene expression program during embryogenesis in cotton have been followed by RNA-excess DNA-RNA hybridization studies that have identified mRNA sets from twodimensional SDS-PAGE analyses of extant proteins, proteins synthesized in vivo and in vitro, and the changing mRNA populations in embryos of different ages. As in soybean, DNA-RNA hybridization with mRNA populations from young (50 mg fresh weight), maturation phase (110 mg fresh weight), mature (from seeds), and germinated embryos implies that 15,000-20,000 genes are expressed at the mRNA level during embryogenesis. However, only a few hundred genes encode mRNAs that are abundant enough to be detected in profiles of twodimensional gels of in vitro translation products. The analysis of extant, in vivo and in vitro synthesized proteins allowed the changes in the abundant mRNAs and abundant proteins to be followed with
406
Zygotic Embryogenesis
EMBRYOGENESIS EARLY
LATE
MID
matuf,
EARLY GERMINATION
seed
SB
Figure 14.4. A diagrammatic representation of the groups of abundant mRNAs demonstrated in developing and germinating cotton embryos. (From Dure 1985. Oxford
Surv. Plant Mol. Cell Biol. 2:179-197, by permission of Oxford University Press.)
a focus on two major questions. Does the population of extant proteins in the young embryo include some proteins whose synthesis is not required for continued differentiation of the embryo? Does a mature embryo synthesize proteins that are not needed for the survival of the young seedling? When the soluble protein fractions from cotyledons of embryos of different ages are analyzed, substantial modulations are found to occur in several sets of proteins whose synthesis is developmentally regulated; gene sets expressed in embryos of the same ages are deduced from in vitro translation analysis of their mRNAs. These data are diagrammatically presented in Figure 14.4 as groups of abundant mRNAs expressed during embryogenesis. mRNAs of groups 1 and 2 are present throughout embryogenesis; whereas group 1 mRNAs are expressed during germination as well, those of group 2 are turned off after embryogenesis is completed. Compared to group 2 mRNAs, which encode for only six proteins, group 1 mRNAs encode an impressively large set of proteins, including actin and tubulin. Another point of comparison is between the products of cell-free translation of mRNAs isolated from embryos of different ages and proteins synthesized in vivo by embryos of the same age; this comparison has shown that many translatable mRNAs of group 1 are detected only when their corresponding proteins are synthesized in vivo, indicating that gene activity with regard to the synthesis of these proteins is regulated at the transcriptional level. However, proteins
encoded by group 2 mRNAs are synthesized in vitro by embryos of all ages. Expression of group 3 mRNAs occurs coincident with the burst of mitotic activity during early embryogenesis but, as cell divisions cease, their genes are turned off in the maturation phase and in seed embryos. Proteins encoded by these transcripts accumulate extensively in embryos of all developmental stages, although they are synthesized only in young and maturation stage embryos. The most extensively changing mRNAs are those that code for storage proteins (group 4). These mRNAs are found in high levels in maturation stage embryos but vanish completely in mature embryos. Two other groups of mRNAs (groups 5 and 6), which become abundant during late embryogenesis and accumulate in seed embryos, differ in one important respect. Group 5 mRNAs, which consist of two subsets and their corresponding proteins, disappear from the abundant class during the early hours of germination but can be reinstated by incubating excised immature embryos in ABA. This is just what one would expect of a group of transcripts and their protein products whose synthesis in the embryos is regulated by endogenous ABA. The other group of mRNAs (group 6) and their proteins persist during normal germination of seed embryos, as well as during precocious germination of immature embryos. There is also one final group of mRNAs (group 7), which is not expressed during embryogenesis but appears during germination (Dure,
Genetic and Molecular Analysis of Embryogenesis
Greenway and Galau 1981; Galau and Dure 1981; Chlan and Dure 1983). Comparisons of reassociation kinetics between cDNAs from cotton embryos of all three developmental stages and an excess of homologous mRNAs have identified the same groups of mRNAs whose abundant proteins are synthesized in vivo and in vitro, as well as several less abundant mRNAs. These mRNAs belong to abundance classes that show more than a twofold increase in their sequences in the seed embryos. Another type of analysis has involved hybridization of each cDNA clone to mRNAs isolated from other embryo stages. Such heterologous reassociations have shown that embryos of all ages contain most of the mRNA sequences, although sequences prevalent at one stage may be present at a reduced level in other stages. Studies of this type have also focused on some major differences in the concentrations of the more abundant mRNAs of embryos of different stages of development, the most dramatic being a stark reduction in the presumptive storage protein mRNAs in the mature embryo (Galau and Dure 1981). The hybridization kinetics of mRNAs from cotyledons of developing pea embryos to cDNAs show much that is common with, and much that is different from, those of soybean and cotton. The total complexity of mRNA sequences of different abundance classes at the early stage of cotyledon development in pea (ca. 20,000) compares favorably with that of cotton (ca. 25,000) and soybean (ca. 32,000) embryos of slightly different ages. As noted in soybean, there is an increase in the very abundant class of mRNAs during embryo development when storage protein synthesis is initiated and sustained. However, compared to soybean and cotton, in pea the rare class of mRNAs completely disappears as embryogenesis proceeds, resulting in a 10-fold decrease in the number of mRNA sequences in embryos at the late stage of cotyledon development (Morton et al 1983). The obvious adaptive value of this strategy is that genes for the production of enzymes and structural proteins are no longer required and are switched off as the embryo begins to accumulate storage proteins. Since this reduces the number of active genes in the cell to about 200, some questions have been raised about the survival capacity of such cells (Dure 1985). (ii) Gene Activity during Fertilization and Early Zygote Divisions
There have been very few investigations on the molecular biology of fertilization in gymnosperms
407
and angiosperms, and the mechanism of gene activation during the early division phase of the zygote has completely eluded us. A major impediment to these studies is that events of fertilization and partitioning of the zygote take place within the confines of the archegonium in gymnosperms and the embryo sac in angiosperms. Reference was made in Chapter 13 to studies that seemed to indicate that fertilization of the egg and subsequent divisions of the zygote are accompanied by changes in the RNA content of the cells. Unfortunately, these investigations did not address issues such as the role of the maternal genome in fertilization and early embryo development. The concept of maternal message, which has its roots in animal embryology, assumes that the mature egg contains large stores of mRNAs that become available after fertilization to code for the first proteins of the zygote. There is no hard evidence for the existence of a legacy of template information in the gymnosperm or angiosperm egg, and the source of the genetic information parceled out for the division of the zygote is unknown. One study has shown that when rice and barley florets are cultured in a medium containing 3H-uridine, no autoradiographically detectable incorporation of the label occurs in developing embryos of less than 100 cells. At the same time, 3H-leucine is readily incorporated into embryos of all ages. Because the eggs are fertilized after the florets are placed in culture, assuming that the cells of the globular embryo do not possess a large uridine pool, it has been reasoned that protein synthesis in the absence of concurrent RNA synthesis probably involves the use of stored maternal mRNA (Nagato 1979). Based on in vitro translation of poly(A)-RNA isolated from ovules of Pinus strobus before and after fertilization, evidence was obtained for the transcription of a prominent mRNA at the time of fertilization and at the beginning of embryogenesis. A cDNA clone isolated from ovules containing globular embryos was found to encode a protein homologous to the US storage protein globulin. Since transcripts of this cDNA clone are found to accumulate in the megagametophytes of fertilized, but not of unfertilized, ovules, the trigger of the fertilization event in causing transcript abundance is obvious. Although transcripts are not detected in the developing embryos, this does not preclude a possible role for fertilization-induced signals from the embryo in transcript accumulation in the megagametophyte (Whitmore and Kriebel 1987; Baker et al 1996). Analysis of genes expressed in the embryo, with the exclusion of the female gametophyte and the sporophytic tissues that constitute the bulk of the ovule, may perhaps provide a clear picture of the embryo gene activation program during fertilization.
408
Based largely on the end result, namely, activation of the egg in the absence of fertilization (parthenogenesis), many claims have been made for an independent role for the maternal genome in directing early embryo development. These claims include the spontaneous division of the egg in the absence of a stimulus of fusion with the sperm, as in apomicts (Nogler 1984; Ramachandran and Raghavan 1992a); use of physical and chemical agents to inactivate the sperm without hindering the growth and penetration of the pollen tube and thus to stimulate the division of the egg (Lacadena 1974); and the purported origin of embryos from the eggs of cultured unfertilized ovules or ovaries (Yang and Zhou 1982; Yan, Yang and Jensen 1989; Ferrant and Bouharmont 1994). Haploid plants that arise by chromosome elimination in interspecific crosses between Hordeum vulgare and H. bulbosum
also offer a basis for the assumption that maternal templates alone can support embryo development, albeit under somewhat abnormal conditions (Kasha and Kao 1970). Since the male chromosomes disappear even before they are integrated into the genome of the egg, the haploid embryo essentially develops by transcribing the stored maternal information. Although the examples considered here do not constitute a rigorous demonstration of the presence of stored mRNAs, there is probably some truth in the assertion that the angiosperm egg cell, whether it is fertilized or not, has the potential to initiate the developmental program of the embryo using maternal transcripts.
(iii) RNA Metabolism in Developing Embryos
Zygotic Embryogenesis
Rosie 1975), Oryza sativa (Zhu, Shen and Tang 1980a), Triticum vulgaris (Zhu, Shen and Tang 1980b), and Abelmoschus esculentus (Malvaceae) (Bhalla, Singh and Malik 1980b), among others, are examples that illustrate a rising level of RNA through most periods of growth of the embryos. Investigations on P. vulgaris implicate the high water content of the embryo as the stimulus for increased RNA synthesis; as the water content of the embryo begins to decrease, the rate of RNA synthesis falls to a low value and remains at this level through the period of desiccation and dormancy (Walbot 1971, 1972). Other studies have shown that the RNA metabolism of embryos of different ages and that of specific parts of the same embryo are profoundly interrelated. When just the organogenetic part of embryos of different ages of P. coccineus, consisting of rapidly dividing cells, is analyzed, RNA content per cell and the rate of RNA synthesis per cell are highest in the early heart-shaped embryo and decline thereafter. In the numerically stabilized cell population of the suspensor, these parameters are low during early embryogenesis. The decrease in the RNA synthetic activity of the embryo cells probably reflects the inability of the nucleolus to synthesize new RNA to keep up with the rapid tempo of mitosis (Walbot et al 1972; Sussex et al 1973). Although the rate of RNA synthesis declines to near zero during desiccation of the embryo, protein accumulation is unabated; the possibility exists that the proteins are encoded by mRNAs synthesized at an earlier period in embryogenesis and stored for later use (Walbot 1971).
Investigations on the RNA metabolism of cotyledons of developing embryos have generally Several early investigations have shown that supported the results obtained with whole progressive embryo development signals consider- embryos. In the cotyledons of different species able changes in the temporal and spatial patterns of there is close agreement in the time of initiation of accumulation and synthesis of bulk RNA. In the RNA synthesis, which generally follows DNA synwake of the isolation and characterization of thesis. RNA accumulates in cotyledons of near-norembryo-specific genes, these studies can be consid- mal water content in parallel with the period of ered to represent at best a gross approximation of rapid cell division activity; this is followed by a cesgene activity in developing embryos. For this rea- sation or decrease of accumulation during desiccason, only a small portion of this work will be dis- tion. The major portion, if not all, of the RNA synthesized appears to be rRNA (Galitz and cussed here. One manifestation of RNA metabolism during Howell 1965; Millerd and Whitfeld 1973; Poulson embryogenesis is the continued accumulation of and Beevers 1973; Scharpe and van Parijs 1973; D. RNA, so that by the completion of embryogenesis, L. Smith 1973; Walbot 1973; Millerd and Spencer the amount of RNA in the mature embryo will have 1974). Considered together, the RNA metabolism of increased to many times the initial value of the developing whole embryos and isolated cotyleearly stage embryo. Viciafaba (Wheeler and Boulter dons invariably indicates a changing biochemical 1967; Manteuffel et al 1976), Phaseolus vulgaris pattern with time. (Walbot 1971), Hordeum distichum (Duffus and The rise and decline in RNA accumulation and
Genetic and Molecular Analysis of Embryogenesis
synthesis during embryogenesis reflect, to some extent, the availability of sufficient mRNA. Poulson and Beevers (1973) found that although the RNA content of pea cotyledons remains fairly constant during the seed maturation period, the capacity of the cotyledons for RNA synthesis declines considerably during this period. The abruptness of this change correlates well with the observation that ribosomal preparations made from cotyledons of mature embryos show a decreasing abundance of polysomes and are practically inefficient in supporting cell-free translation. Since the full cell-free translation potential of ribosomes is realized by the addition of an artificial messenger, experimental results are consistent with a declining mRNA component in the cotyledons with age (Beevers and Poulson 1972). A genetic approach was used by Davies and Brewster (1975) to analyze RNA synthetic activity in developing pea cotyledons. The basis for this work was the observation that cotyledons of certain genotypes of pea show appreciable differences in their RNA contents. However, when plants with widely different rRNA contents are crossed, cotyledons of the Fl hybrid embryos are found to have an rRNA content identical to that of the maternal parent. Of the two models proposed to account for the essential silencing of a set of rRNA cistrons, one supposes that some aspect of gene activity, such as transcription or processing of RNA, is primarily under the control of the maternal genome. Alternatively, a selective activation of the maternal alleles might occur and result in an unusual situation in which one set of rRNA cistrons produces the same amount of RNA in the hybrid as two sets do in the parent. Clearly, both models are valid, but further work is necessary to choose between them. Chapter 13 discussed the fact that in the cotyledons of legumes, the DNA content of cells increases during the first few days of cell expansion growth. Since the large increase in RNA content occurs almost in parallel with the increase in nuclear DNA amounts, a logical question is whether there is an apparent gene dosage effect on RNA synthesis in the cotyledons. One measure of the transcriptional activity of DNA is a change in the level of RNA polymerases. Experiments by Millerd and Spencer (1974) have proved illuminating by showing that cotyledons of a variety of pea do not synthesize additional RNA polymerase in proportion to the increasing amount of nuclear DNA template available. Since this occurs even when the template activity of chromatin in the cotyledons remains constant, a role for gene dosage, in the sense of
409
loading the cytoplasm with transcription products in a quantitative way, does not appear to be valid. Investigations of Cullis (1976) with two other varieties of pea showed that an increase in the endogenous RNA polymerase activity of chromatin occurs even before the average DNA content increases in the majority of the cells and continues well into the period of DNA endoreduplication in most cells. This might suggest that some of the endoreduplicated DNA is used as a template for RNA synthesis. A new element of uncertainty was introduced into this issue when it was found in a later work that the extent of utilization of extra copies of DNA is associated with experimental conditions that affect the rate of growth of cotyledons. For example, in plants raised at 15 °C, which permits a tardy growth of cotyledons, RNA polymerase activity increases in proportion to the DNA content of cells over a wide developmental window. When cotyledons develop rapidly, as in plants grown at 30 °C, there is no increase in the level of chromatin-bound RNA polymerase activity in proportion to the DNA (Cullis 1978). The differences in the template activity of chromatin from two sets of cotyledons are interpretable in terms of differences in the number of transcriptionally active cistrons available or in the specific activity of the bound polymerase. These views are not inconsistent with the currently known mechanisms of regulation of DNA template activity. In summary, it appears that there is no firm ground for the idea that, except under special conditions, transcriptional activity in the embryonic cotyledons is related to the availability of extra copies of DNA. Endoreduplication results in a multifold increase in the informational pool of the cells of the cotyledons, but the ultimate function and disposition of this information remain to be fully understood. Gene expression pattern during embryogenesis has been deduced from changes in the most prevalent mRNA populations seen in cell-free translation profiles of embryos of different ages (Figure 14.5). The changes in the proteins synthesized and in the expression of their mRNAs during embryogenesis in cotton were described earlier. The existence of in vivo and in vitro synthesized polypeptide subsets that exhibit stage-specific expression appears to be a consistent feature of embryogenesis in maize (Sanchez-Martinez, Puigdomenech and Pages 1986; Boothe and Walden 1990; Oishi and Bewley 1992), radish (Aspart et al 1984; Laroche-Raynal et al 1984), Phaseolus vulgaris (Misra and Bewley 1985b), rice (Mundy and Chua, 1988), and soybean
4.3
•8.5
a
-82
8-B-43
4-a 5-B '
'B-1 B-2
-25
•
-18
7-B
12
*IP
-v-
34
50
•-. 36
37 Jt 17
-«
4.3 -82
*' t
-43 •33
34 "
.
.' • * . J
-18
ra
30=> 29o
• 4 a is sB 30> 29.
«39
Genetic and Molecular Analysis of Embryogenesis
(Rosenberg and Rinne 1988). In both maize and radish, although the prevalent mRNAs show numerous changes during embryogenesis, two main subsets, one synthesized in young and immature embryos and the other accumulating in mature embryos, have been identified. Maturation of rice embryo is associated with the accumulation of mRNAs encoding about seven polypeptides ranging in molecular mass from 20 to 45 kDa. Most of the mRNA sequences of the mature embryos of these plants persist as stored mRNAs and, consequently, they are considered to be products of genes transcribed during embryogenesis. The concept of stored mRNAs in embryos of quiescent and dormant seeds of other angiosperms has been supported by a variety of experimental approaches. One important consideration about stored mRNAs is that they serve as templates for the synthesis of proteins of early germination, until the growing embryo begins to synthesize new mRNAs. In vitro translation products of mRNAs isolated from quiescent embryos of wheat (Helm and Abemethy 1990) and sorghum (Howarth 1990) have been shown to include transcripts of HSP genes. Additionally, cDNA libraries made to poly(A)RNA of embryos excised from dry seeds of radish (Raynal et al 1989), Vigna unguiculata (Ishibashi and Minamikawa 1990; Ishibashi, Yamauchi and Minamikawa 1990), V. radiata (Manickam et al 1996), Pinus thunbergii (Masumori, Yamamoto and Sasaki 1992), and sunflower (Almoguera and Jordano 1992), and embryos excised from caryopses of rice (Ramachandran and Raghavan 1992b) and wheat (Kawashima et a 1992) have been differentially screened, and developmentally significant cDNA clones have been described. Of special interest is a gene that encodes a Zn-containing metallothionein-like protein isolated from mature wheat embryos. The expression of this protein is gene modulated in developing embryos by ABA and not by Zn2+. In spite of much information about the structure and ubiquity of metallothioneins in animal cells, little is known about their function during plant embryogenesis.
Figure 14.5. Fluorographs showing the in vitro translation profiles of proteins synthesized by RNA from embryo axes of Phaseolus vulgaris of different stages of development. Schematic representations of the fluorographs are given on the right. E indicates the major translation products from the endogenous mRNA. Spots enclosed in boxes correspond to permanent polypeptides. (a) Embryo axes
••
3L 2L 31
•«« Awr
IM9K* 4.
4L
.«
^
IS 4L
».
9fiB* iMMN 3)-|3-glucan-specific monoclonal antibody. Planta Mericle, L.W., and Mericle, R.P. 1970. Nuclear DNA com185:1-8. plement in young proembryos of barley. Mutation Meinke, D.W. 1982. Embryo-lethal mutants of Arabidopsis Res. 10:515-518. thaliana: evidence for gametophytic expression of the Merkle, S.A., Parrott, W.A., and Flinn, B.S. 1995. mutant genes. Theor. Appl. Genet. 63:381-386. Morphogenetic aspects of somatic embryogenesis. Meinke, D.W. 1985. Embryo-lethal mutants of Arabidopsis In In Vitro Embryogenesis in Plants, ed. T.A. Thorpe, thaliana: analysis of mutants with a wide range of pp. 155-203. Dordrecht: Kluwer Academic lethal phases. Theor. Appl. Genet. 69:543-552. Publishers. Meinke, D.W. 1986. Embryo-lethal mutants and the study of plant embryo development. Oxford Surv. Plant Mertz, E.T., Bates, L.S., and Nelson, O.E. 1964. Mutant gene that changes protein composition and increases Mol. Cell Biol. 3:122-165. lysine content of maize endosperm. Science Meinke, D.W. 1991a. Embryonic mutants of Arabidopsis thaliana. Devel Genet. 12:382-392. 145:279-280. Meinke, D.W. 1991b. Genetic analysis of plant develop- Meurs, C, Basra, A.S., Karssen, CM., and van Loon, L.C. ment. In Plant Physiology. A Treatise, Vol. X, ed. F.C. 1992. Role of abscisic acid in the induction of desicSteward and R.G.S. Bidwell, pp. 437-490. San Diego: cation tolerance in developing seeds of Arabidopsis thaliana. Plant Physiol. 98:1484-1493. Academic Press. Meinke, D.W. 1991c. Perspectives on genetic analysis of Meyer, K., Leube, M.P., and Grill, E. 1994. A protein phosphatase 2C involved in ABA signal transduction in plant embryogenesis. Plant Cell 3:857-866. Arabidopsis thaliana. Science 264:1452-1455. Meinke, D.W. 1992. A homeotic mutant of Arabidopsis thaliana with leafy cotyledons. Science 258:1647-1650. Meyer, V.G. 1969. Some effects of genes, cytoplasm, and environment on male sterility of cotton (Gossypium). Meinke, D.W., Chen, J., and Beachy, R.N. 1981. Expression Crop Sci. 9:237-242. of storage-protein genes during soybean seed develMeyer, V.G., and Buffet, M. 1962. Cytoplasmic effects on opment. Planta 153:130-139. external-ovule production in cotton. /. Hered. Meinke, D.W., and Sussex, I.M. 1979a. Embryo-lethal 53:251-253. mutants of Arabidopsis thaliana. A model system for genetic analysis of plant embryo development. Devel Meyerowitz, E.M. 1994. The genetics of flower development. Sci. Am. 271(5):40-47. Biol. 72:50-61. Meinke, D.W., and Sussex, I.M. 1979b. Isolation and char- Meyerowitz, E.M., Bowman, J.L., Brockman, L.L., Drews, G.N., Jack, T., Sieburth, L., and Weigel, D. 1991. A acterization of six embryo-lethal mutants of genetic and molecular model for flower developArabidopsis thaliana. Devel Biol. 72:62-72. ment in Arabidopsis thaliana. Development Suppl. Meinke, D.W., Franzmann, L.H., Nickle, T.C., and Yeung, 1:157-167. E.C. 1994. Leafy cotyledon mutants of Arabidopsis. Mian, H.R., Kuspira, J., Walker, G.W.R., and Muntjewerff, Plant Cell 6:1049-1064. N. 1974. Histological and cytochemical studies on Melcher, U. 1979. In vitro synthesis of a precursor to the five genetic male-sterile lines of barley (Hordeum vitlmethionine-rich polypeptide of the zein fraction of gare). Can. ]. Genet. Cytol. 16:355-379. corn. Plant Physiol. 63:354-358. Melchers, G., Mohri, Y, Watanabe, K., Wakabayashi, S., Miao, S.-H., Kuo, C.-S., Kwei, Y.-L., Sun, A.-T., Ku, S.-Y, Lu, W.-L., Wang, Y.-Y, Chen, M.-L., Wu, M.-K., and and Harada, K. 1992. One-step generation of cytoplasmic male sterility by fusion of mitochondrHang, L. 1978. Induction of pollen plants of maize ial-inactivated tomato protoplasts with nuclear-inacand observations on their progeny. In Proceedings of tivated Solanum protoplasts. Proc. Natl Acad. Sci. Symposium on Plant Tissue Culture, pp. 23-33. Peking: USA 89:6832-6836. Science Press. Mena, M., Mandel, M.A., Lerner, D.R., Yanofsky, M.F., Michalczuk, L., Cooke, T.J., and Cohen, J.D. 1992. Auxin and Schmidt, R.J. 1995. A characterization of the levels at different stages of carrot somatic embryogeMADS-box gene family in maize. Plant ]. 8:845-854. nesis. Phytochemistry 31:1097-1103.
608
References
Michalczuk, L., Ribnicky, D.M., Cooke, T.J., and Cohen, deficient miniature-1 seed mutation is associated J.D. 1992. Regulation of indole-3-acetic acid biosynwith aberrant pedicel and endosperm development. thetic pathways in carrot cell cultures. Plant Physiol. Plant Cell 4:297-305. 100:1346-1353. Miller, R.J., and Koeppe, D.E. 1971. Southern corn leaf Michel, D., Salamini, R, Bartels, D., Dale, P., Baga, M., and blight: susceptible and resistant mitochondria. Szalay, A. 1993. Analysis of a desiccation and ABAScience 173:67-69. responsive promoter isolated from the resurrection Millerd, A. 1975. Biochemistry of legume seed proteins. plant Craterostigma planlagineum. Plant J. 4:29-40. Annu. Rev. Plant Physiol. 26:53-72. Miernyk, J.A., and Trelease, R.N. 1981. Control of enzyme Millerd, A., Simon, M., and Stern, H. 1971. Legumin synactivities in cotton cotyledons during maturation thesis in developing cotyledons of Viciafaba L. Plant and germination. IV. fi-Oxidation. Plant Physiol. Physiol. 48:419-425. 67:341-346. Millerd, A., and Spencer, D. 1974. Changes in RNA-synMiflin, B.J., Burgess, S.R., and Shewry, P.R. 1981. The thesizing activity and template activity in nuclei development of protein bodies in the storage tissues from cotyledons of developing pea seeds. Aust. ]. of seeds: subcellular separations of homogenates of Plant Physiol. 1:331-341. barley, maize, and wheat endosperms and of pea Millerd, A., Thomson, J.A., and Schroeder, H.E. 1978. cotyledons. /. Expt. hot. 32:199-219. Cotyledonary storage proteins in Pisum sativum. III. Mignouna, H., Virmani, S.S., and Briquet, M. 1987. Patterns of accumulation during development. Aust. J. Plant Physiol. 5:519-534. Mitochondrial DNA modifications associated with cytoplasmic male sterility in rice. Theor. Appl. Genet. Millerd, A., and Whitfeld, P.R. 1973. Deoxyribonucleic 74:666-669. acid and ribonucleic acid synthesis during the cell Mii, M. 1973. Studies on the haploid formation from expansion phase of cotyledon development in Vicia immature pollen grain. I. Relation between pollen faba L. Plant Physiol. 51:1005-1010. stage and the plantlet formation and the process of Millerd, A., Spencer, D., Dudman, W.F., and Stiller, M. embryoid formation in Nicotiana tabacum. ]ap. ]. 1975. Growth of immature pea cotyledons in culture. Breed. 23:27-34. Aust. J. Plant Physiol. 2:51-59. Mikami, T., Kishima, Y, Sugiura, M., and Kinoshita, T. Milligan, S.B., and Gasser, C.S. 1995. Nature and regula1985. Organelle genome diversity in sugar beet with tion of pistil-expressed genes in tomato. Plant Mol. normal and different sources of male sterile cytoBiol. 28:691-711. plasms. Theor. Appl. Genet. 71:166-171. Minocha, S.C, Papa, N.S., Khan, A.J., and Samuelsen, A.I. Mikami, T., Sugiura, M., and Kinoshita, T. 1984. Molecular 1991. Polyamines and somatic embryogenesis in carheterogeneity in mitochondrial and chloroplast rot. III. Effects of methylglyoxal bis(guanylhydraDNAs from normal and male sterile cytoplasms in zone). Plant Cell Physiol. 32:395-402. sugar beets. Curr. Genet. 8:319-322. Mis£ra, S., Miiller, A.J., Weiland-Heidecker, U., and Miki, H. 1964. A study of the tropism of pollen tubes to the Jurgens, G. 1994. The FUSCA genes of Arabidopsis: pistil. I. Tropism in Lilium. Bot. Mag. Tokyo 67:143-147. negative regulators of light responses. Mol. Gen. Miki-Hirosige, H., Hoek, I.H.S., and Nakamura, S. 1987. Genet. 244:242-252. Secretions from the pistil of Lilium longiflorum. Am. J. Mishkind, M., Raikhel, N.V., Palevitz, B.A., and Keegstra, Bot. 74:1709-1715. K. 1982. Immunocytochemical localization of wheat germ agglutinin in wheat. /. Cell Biol. 92:753-764. Miki-Hirosige, H., and Nakamura, S. 1982. Process of metabolism during pollen tube wall formation. /. Mishkind, M.L., Palevitz, B.A., Raikhel, N.V., and Electron Micros. 31:51-62. Keegstra, K. 1983. Localization of wheat germ aggluMike-Hirosige, H., Nakamura, S., and Tanaka, I. 1988. tinin-Iike lectins in various species of the Gramineae. Ultrastructural research on cell wall regeneration by Science 220:1290-1292. cultured pollen protoplasts of Lilium longiflorum. Sex. Misoo, S., Yoshida, K., and Mastubayashi, M. 1979. Plant Reprod. 1:36-45. Studies on the mechanisms of pollen embryogenesis. III. Mitotic responses of the pollen to varied sucrose Miki-Hirosige, H., Nakamura, S., Yasueda, H., Shida, T., concentrations and the process of embryoid formaand Takahashi, Y. 1994. Immunocytochemical localtion in tobacco anther culture. Sci. Rep. Fac. Agr. Kobe ization of the allergenic proteins in the pollen of Univ. 13:193-202. Cryptomeria japonica. Sex. Plant Reprod. 7:95-100. Mikulska, E., and Rodkiewicz, B. 1967a. infrastructure of Misra, S., and Bewley, J.D. 1985a. Reprogramming of prothe maturing embryo sac of Lilium regale. Ada Soc. tein synthesis from a developmental to a germinaBot Polon. 36:555-566. tive mode induced by desiccation of the axes of Phaseolus vulgaris. Plant Physiol. 78:876-882. Mikulska, E., and Rodkiewicz, B. 1967b. Fine structure of four-nucleate stages and the central cell of Lilium Misra, S., and Bewley, J.D. 1985b. The messenger RNA regale embryo sac. Flora 157:365-372. population in the embryonic axes of Phaseolus vidgaris during development and following germinaMiller, D.D., Scordilis, S.P., and Hepler, P.K. 1995. tion. /. Expt. Bot. 36:1644-1652. Identification and localization of three classes of myosins in pollen tubes of Lilium longiflorum and Misra, S., and Green, M.J. 1991. Developmental gene Nicotiana alata. J. Cell Sci. 108:2549-2563. expression in conifer embryogenesis and germination. II. Crystalloid protein synthesis in the developMiller, D.D., Callaham, D.A., Gross, D.J., and Hepler, P.K. 2+ ing embryo and megagametophyte of white spruce 1992. Free Ca gradient in growing pollen tubes of (Picea glauca [Moench] Voss.). Plant Sci. 78:61-71. Lilium. J. Cell Sci. 101:7-12. Miller, E.C. 1919. Development of the pistillate spikelet Misra, S., Attree, S.M., Leal, I., and Fowke, L.C. 1993. Effect of abscisic acid, osmoticum, and desiccation on synand fertilization in Zea mays L. /. Agr. Res. thesis of storage proteins during the development of 18:255-265. white spruce somatic embryos. Ann. Bot. 71:11-22. Miller, M.E., and Chourey, P.S. 1992. The maize invertase-
References
609
Misset, M.T., and Gourret, J.P. 1984. Accumulation of Moens, P.B. 1968. The structure and function of the synapsmooth cisternae in the tapetal cells of Ulex europaeits tonemal complex in Lilium longiflorum sporocytes. L. (Papilionoideae). /. Cell Sci. 72:65-74. Chromosoma 23:418-451. Mitchell, J.C., and Petolino, J.F. 1988. Heat stress effects on Moewus, F. 1950. Zur Physiologie und Biochemie der Selbststerilitat bei Forsythia. Biol. Zentral. 69:181-197. isolated reproductive organs of maize. /. Plant Moffatt, B., Pethe, C, and Laloue, M. 1991. Metabolism of Physiol. 133:625-628. benzyladenine is impaired in a mutant of Arabidopsis Mitten, D.H., Sato, S.J., and Skokut, T.A. 1984. In vitro thaliana lacking adenine phosphoribosyltransferase regenerative potential of alfalfa germplasm sources. activity. Plant Physiol. 95:900-908. Crop Sci. 24:943-945. Mittermann, I., Swoboda, I., Pierson, E., Eller, N., Kraft, Moffatt, B., and Somerville, C. 1988. Positive selection for D., Valenta, R., and Heberle-Bors, E. 1995. Molecular male-sterile mutants of Arabidopsis lacking adenine cloning and characterization of profilin from tobacco phosphoribosyl transferase activity. Plant Physiol. (Nicotiana tabacum): increased profilin expression 86:1150-1154. during pollen maturation. Plant Mol. Biol. Mogensen, H.L. 1972. Fine structure and composition of 27:137-146. the egg apparatus before and after fertilization in Mix, G., Wilson, H.M., and Foroughi-Wehr, B. 1978. The Quercus gambelii: the functional ovule. Am. ]. Bot. cytological status of plants of Hordeum vulgare L. 59:931-941. regenerated from microspore callus. Z. Pflanzen- Mogensen, H.L. 1973. Some histochemical, ultrastruczilchtg 80:89-99. tural, and nutritional aspects of the ovule of Quercus gambelii. Am. J. Bot. 60:48-54. Miyake, T., Kuroiwa, H., and Kuroiwa, T. 1995. Differential mechanisms of movement between a generative cell Mogensen, H.L. 1977. Ultrastructural analysis of female and a vegetative nucleus in pollen tubes of Nicotiana pachynema and the relationship between synaptonemal complex length and crossing-over in Zea tabacum as revealed by additions of colchicine and mays. Carlsberg Res. Commun. 42:475-497. nonanoic acid. Sex. Plant Reprod. 8:228-230. Miyamura, S., Kuroiwa, T., and Nagata, T. 1987. Mogensen, H.L. 1978a. Synergids of Proboscidea louisianica before fertilization. Phytomorphology 28:114-122. Disappearance of plastid and mitochondrial nucleoids during the formation of generative cells of Mogensen, H.L. 1978b. Pollen tube-synergid interactions in Proboscidea louisianica (Martineaceae). Am. ]. Bot. higher plants revealed by fluorescence microscopy. Protoplasma 141:149-159. 65:953-964. Miyoshi, H., Usami, T., and Tanaka, I. 1995. High level of Mogensen, H.L. 1981a. Ultrastructural localization of GUS gene expression driven by pollen-specific proadenosine triphosphatase in the ovules of Saintpaulia moters in electroporated lily pollen protoplasts. Sex. ionantha (Gesneriaceae) and its relation to synergid Plant Reprod. 8:205-209. function and embryo sac nutrition. Am. ]. Bot. 68:183-194. Mizelle, M.B., Sethi, R., Ashton, M.E., and Jensen, W.A. 1989. Development of the pollen grain and tapetum Mogensen, H.L. 1981b. Translocation of uranin within the of wheat {Triticwn aestivum) in untreated plants and living ovules of selected species. Am. ]. Bot. plants treated with chemical hybridizing agent 68:195-199. RH0007. Sex. Plant Reprod. 2:231-253. Mogensen, H.L. 1982. Double fertilization in barley and the cytological explanation for haploid embryo forMizukami, Y., and Ma, H. 1992. Ectopic expression of the mation, embryoless caryopses, and ovule abortion. floral homeotic gene AGAMOUS in transgenic Arabidopsis plants alters floral organ identity. Cell Carlsberg Res. Commun. 47:313-354. 71:119-131. Mogensen, H.L. 1984. Quantitative observations on the Mizukami, Y., and Ma, H. 1995. Separation of AG function pattern of synergid degeneration in barley. Am. ]. in floral meristem determinacy from that in reproBot. 71:1448-1451. ductive organ identity by expressing antisense AG Mogensen, H.L. 1985. Ultracytochemical localization of RNA. Plant Mol. Biol. 28:767-784. plasma membrane-associated phosphatase activity Mo, L.H., and von Arnold, S. 1991. Origin and development in developing tobacco seeds. Am. ]. Bot. 72:741-754. of embryogenic cultures from seedlings of Norway Mogensen, H.L. 1986. On the male germ unit in an spruce (Picea abies). ]. Plant Physiol. 138:223-230. angiosperm with bicellular pollen, Hippeastrum vitaMo, L.H., Egertsdotter, U., and von Arnold, S. 1996. turn. In Biotechnology and Ecology of Pollen, ed. D.L. Secretion of specific extracellular proteins by Mulcahy, G.B. Mulcahy and E. Ottaviano, pp. somatic embryos of Picea abies is dependent on 297-305. New York: Springer-Verlag. embryo morphology. Ann. Bot. 77:143-152. Mogensen, H.L. 1988. Exclusion of male mitochondria Mo, Y, Nagel, C, and Taylor, L.P. 1992. Biochemical comand plastids during syngamy in barley as a basis for plementation of chalcone synthase mutants defines a maternal inheritance. Proc. Natl Acad. Sci. USA role for flavonols in functional pollen. Proc. Natl 85:2594-2597. Acad. Sci. USA 89:7213-7217. Mogensen, H.L. 1992. The male germ unit: concept, comMo, Y.-S., and Yang, H.-Y. 1991. Mass isolation and preserposition, and significance. Int. Rev. Cytol. vation of viable sperm cells in Brassica campestris var. 140:129-147. purpurea. Ada Bot. Sinica 33:649-657. Mogensen, H.L. 1996. The hows and whys of cytoplasmic inheritance in seed plants. Am. ]. Bot. 83:383-404. Mo, Y.-S., and Yang, H.-Y. 1992. Isolation and fusion of sperm cells in several bicellular pollen species. Ada Mogensen, H.L., and Holm, P.B. 1995. Dynamics of Bot. Sinica 34:688-697. nuclear DNA quantities during zygote development Modrusan, Z., Reiser, L., Feldmann, K.A., Fischer, R.L., in barley. Plant Cell 7:487-494. and Haughn, G.W. 1994. Homeotic transformation Mogensen, H.L., and Pollak, P.E. 1983. Mg2+, ^-depenof ovules into carpel-like structures in Arabidopsis. dent phosphatase activity during seed development in Nicotiana tabacum. Am. J. Bot. 70:1105-1108. Plant Cell 6:333-349.
670
References
Mogensen, H.L., and Rusche, M.L. 1985. Quantitative ultrastructural analysis of barley sperm. I. Occurrence and mechanism of cytoplasm and organelle reduction and the question of sperm dimorphism. Protoplasma 128:1-13. Mogensen, H.L., and Suthar, H.K. 1979. Ultrastructure of the egg apparatus of Nicotiana tabacum (Solanaceae) before and after fertilization. Bot. Gaz. 140:168-179. Mogensen, H.L., and Wagner, V.T. 1987. Associations among components of the male germ unit following in vivo pollination in barley. Protoplasma 138:161-172. Mogensen, H.L., Wagner, V.T., and Dumas, C. 1990. Quantitative, three-dimensional ultrastructure of isolated corn (Zea mays) sperm cells. Protoplasma 153:136-140. Mogensen, H.L., Leduc, N., Matthys-Rochon, E., and Dumas, C. 1995. Nuclear DNA amounts in the egg and zygote of maize (Zea mays L.). Planta 197:641-645. Mohan Ram, H.Y., and Jaiswal, V.S. 1971. Feminization of male flowers of Cannabis sativa L. by a morphactin.
Monfort, S. 1985. Androgenesis of coconut: embryos from anther culture. Z. Pflanzenzilchtg 94:251-254. Monnier, M. 1975. Action d'un gel de polyacrylamide employe' comme support pour la culture de l'embryon immature de Capsella bursa-pastoris. Compt. Rend. Acad. Sci. Paris 280D:705-708.
Monnier, M. 1976a. Culture in vitro de l'embryon immature de Capsella bursa-pastoris Moench. Rev. Cytol. Biol. Veget. 39:1-120.
Monnier, M. 1976b. Action de la pression partielle d'oxygene sur le deVeloppement de l'embryon de Capsella bursa-pastoris cultiv£ in vitro. Compt. Rend. Acad. Sci. Paris 282D:1009-1012. Monnier, M. 1984. Survival of young immature Capsella
embryos cultured in vitro. ]. Plant Physiol. 115:105-113. Montague, M.J., Armstrong, T.A., and Jaworski, E.G. 1979. Polyamine metabolism in embryogenic cells of Daucus carota. II. Changes in arginine decarboxylase activity. Plant Physiol. 63:341-345. Montague, M.J., Koppenbrink, J.W., and Jaworski, E.G. Naturwissenschaften 58:149-150. 1978. Polyamine metabolism in embryogenic cells of Daucus carota. I. Changes in intracellular content and Mohapatra, S.S., and Knox, R.B., ed. 1996. Pollen rates of synthesis. Plant Physiol. 62:430-433. Biotechnology. Gene Expression and Allergen Characterization. New York: Chapman & Hall. Moore, C.W., and Creech, R.G. 1972. Genetic fine structure analysis of the amylose-extender locus in Zea mays Mohapatra, S.S., Hill, R., Astwood, }., Ekramoddoullah, L. Genetics 70:611-619. A.K.M., Olsen, E., Silvanovitch, A., Hatton, T., Kisil, F., and Sehon, A.H. 1990. Isolation and characteriza- Moore, H.M., and Nasrallah, J.B. 1990. A Brassica selftion of a cDNA clone encoding an IgE-binding proincompatibility gene is expressed in the stylar transtein from Kentucky bluegrass (Poa pratensis) pollen. mitting tissue of transgenic tobacco. Plant Cell Int. Arch. Allergy Appl. Immunol. 91:362-368. 2:29-38. Mohr, S., Schulte-Kappert, E., Oldenbach, W., Oettler, G., Moore, P.J., Moore, A.J., and Collins, G.B. 1994. Genotypic and Kiick, U. 1993. Mitochondrial DNA of cytoplasand developmental regulation of transient expresmic male-sterile Tritkum timopheevi: rearrangement sion of a reporter gene in soybean zygotic cotyleof upstream sequences of the atp6 and or/25 genes. dons. Plant Cell Rep. 13:556-560. Theor. Appl. Genet. 86:259-268. Moreira, M.A., Hermodson, M.A., Larkins, B.A., and Mok, D.W.S., Mok, M.C., and Rabakoarihanta, A. 1978. Nielsen, N.C. 1979. Partial characterization of the Interspecific hybridization of Phaseolus vulgaris with acidic and basic polypeptides of glycinin. /. Biol. P. lunatus and P. acutifolius. Theor. Appl. Genet. Chem. 254:9921-9926. 52:209-215. Moreira, M.A., Hermodson, M.A., Larkins, B.A., and M61, R. 1986. Isolation of protoplasts from female gameNielsen, N.C. 1981. Comparison of the primary tophytes of Torenia fournieri. Plant Cell Rep. structure of the acidic polypeptides of glycinin. Arch. Biochem. Biophys 210:633-642. 5:202-206. M61, R., Betka, A., and Wojciechowicz, M. 1995. Induction Moreno, J., and Chrispeels, M.J. 1989. A lectin gene of autonomous endosperm in Lupinus luteus, encodes the a-amylase inhibitor of the common Helleborus niger and Melandrium album by in vitro cul-
ture of unpollinated ovaries. Sex. Plant Reprod. 8:273-277. M61, R., Matthys-Rochon, E., and Dumas, C. 1993. In vitro culture of fertilized embryo sacs of maize: zygotes and two-celled proembryos can develop into plants. Planta 189:213-217. M61, R., Matthys-Rochon, E., and Dumas, C. 1994. The kinetics of cytological events during double fertilization in Zea mays L. Plant ]. 5:197-206. M61, R., Matthys-Rochon, E., and Dumas, C. 1995. Embryogenesis and plant regeneration from maize zygotes by in vitro culture of fertilized embryo sacs. Plant Cell Rep. 14:743-747.
Moneger, F., Smart, C.J., and Leaver, C.J. 1994. Nuclear restoration of cytoplasmic male sterility in sunflower is associated with the tissue-specific regulation of a novel mitochondrial gene. EMBO ]. 13:8-17. Moneger, E, Mandaron, P., Niogret, M.-F., Freyssinet, G., and Mache, R. 1992. Expression of chloroplast and mitochondrial genes during microsporogenesis in maize. Plant Physiol. 99:396-100.
bean. Proc. Natl Acad. Sci. USA 86:7885-7889.
Morgan, J.M. 1980. Possible role of abscisic acid in reducing seed set in water-stressed wheat plants. Nature 285:655-657. Morris, D.A. 1978. Germination inhibitors in developing seeds of Phaseolus vulgaris L. Z. Pflanzenphysiol. 86:433-441. Morris, PC, Kumar, A., Bowles, D.J., and Cuming, A.C. 1990. Osmotic stress and abscisic acid induce expression of the wheat Em genes. Eur. J. Biochem. 190:625-630. Morris, PC, Maddock, S.E., Jones, M.G.K., and Bowles, D.J. 1985. Changes in the levels of wheat- and barleygerm agglutinin during embryogenesis in vivo, in vitro and during germination. Planta 166:407-413. Morris, PC, Weiler, E.W., Maddock, S.E., Jones, M.G.K., Lenton, J.R., and Bowles, D.J. 1988. Determination of endogenous abscisic acid levels in immature cereal embryos during in vitro culture. Planta 173:110-116. Morrison, I.N., Kuo, J., and O'Brien, T.P. 1975. Histochemistry and fine structure of developing wheat aleurone cells. Planta 123:105-116.
References
Morrison, I.N., and O'Brien, T.P. 1976. Cytokinesis in the developing wheat grain; division with and without a phragmoplast. Planta 130:57-67. Morrison, I.N., O'Brien, T.P., and Kuo, J. 1978. Initial cellularization and differentiation of the aleurone cells in the ventral region of the developing wheat grain. Planta 140:19-30. Morrison, R.A., and Evans, D.A. 1987. Gametoclonal variation. Plant Breed. Rev. 5:359-391. Morrison, R.A., and Evans, D.A. 1988. Haploid plants from tissue culture: new plant varieties in a shortened time frame. Biotechnology 6:684-690. Morton, CM., Lawson, D.L., and Bedinger, P. 1989. Morphological study of the maize male sterile mutant ms7. Maydica 34:239-245. Morton, H., Evans, I.M., Gatehouse, J.A., and Boulter, D. 1983. Sequence complexity of messenger RNA in cotyledons of developing pea (Pisum sativum) seed. Phytochemistry 22:807-812. Morton, R.K., Palk, B.A., and Raison, J.K. 1964. Intracellular components associated with protein synthesis in developing wheat endosperm. Biochem. J. 91:522-528. Moscatelli, A., Del Casino, C, Lozzi, L., Cai, G., Scali, M., Tiezzi, A., and Cresti, M. 1995. High molecular weight polypeptides related to dynein heavy chains in Nicotiana tabacum pollen tubes. /. Cell Sci. 108:1117-1125. Moses, M.J., and Taylor, J.H. 1955. Desoxypentose nucleic acid synthesis during microsporogenesis in Tradescantia. Expt. Cell Res. 9:474-488. Moss, G.I., and Heslop-Harrison, J. 1967. A cytochemical study of DNA, RNA, and protein in the developing maize anther. II. Observations. Ann. Bot. 31:555-572. Moss, G.I., and Heslop-Harrison, J. 1968. Photoperiod and pollen sterility in maize. Ann. Bot. 32:833-846. Motto, M., Maddaloni, M., Ponziani, G., Brembilla, M., Marotta, R., di Fonzo, N., Soave, C, Thompson, R., and Salamini, F. 1988. Molecular cloning of the o2-m5 allele of Zea mays using transposon marking. Mol. Gen. Genet. 212:488^94. Mouritzen, P., and Holm, P.B. 1995. Isolation and culture of barley megasporocyte protoplasts. Sex. Plant Reprod. 8:321-325. Mu, J.-R, Lee, H.-S., and Kao, T.-H. 1994. Characterization of a pollen-expressed receptor-like kinase gene of Petunia inflata and the activity of its encoded kinase. Plant Cell 6:709-721. Mu, J.-H., Stains, J.P., and Kao, T.-H. 1994. Characterization of a pollen-expressed gene encoding a putative pectin esterase of Petunia inflata. Plant Mol. Biol. 25:539-544. Mulcahy, D.L., and Mulcahy, G.B. 1983. Gametophytic selfincompatibility reexamined. Science 220:1247-1251. Mulcahy, D.L., Mulcahy, G.B., and Robinson, R.W. 1979. Evidence for postmeiotic genetic activity in pollen of Cucurbita species. /. Hered. 70:365-368. Mulcahy, D.L., Robinson, R.W., Ihara, M., and Kesseli, R. 1981. Gametophytic transcription for acid phosphatases in pollen of Cucurbita species hybrids. /. Hered. 72:353-354. Mulcahy, G.B., and Mulcahy, D.L. 1985. Ovarian influence on pollen tube growth, as indicated by the semivivo technique. Am. ]. Bot. 72:1078-1080. Mulcahy, G.B., and Mulcahy, D.L. 1987. Induced pollen tube directionality. Am. J. Bot. 74:1458-1459. Miiller, A.J. 1963. Embryonentest zum Nachweis rezes-
611 siver Letalfaktoren bei Arabidopsis thaliana. Biol. Zentral. 82:133-163. Miiller-Rober, B., la Cognata, U., Sonnewald, U., and Willmitzer, L. 1994. A truncated version of an ADPglucose pyrophosphorylase promoter from potato specifies guard cell-selective expression in transgenic plants. Plant Cell 6:601-612. Mullins, M.G., Tang, F.C.A., and Facciotti, D. 1990. Agrobacterium-mediated genetic transformation of grapevines: transgenic plants of Vitis rupestris Scheele and buds of Vitis vinifera L. Biotechnology 8:1041-1045. Mundy, J., and Chua, N.-H. 1988. Abscisic acid and waterstress induce the expression of a novel rice gene. EMBO ]. 7:2279-2286. Munjal, S.V., and Narayan, R.K.J. 1995. Restriction analysis of the mitochondrial DNA of male-sterile and maintainer lines of pearl millet, Pennisetum americanum L. Plant Breed. 114:256-258. Munksgaard, D., Mattsson, O., and Okkels, F.T. 1995. Somatic embryo development in carrot is associated with an increase in levels of S-adenosylmethionine, S-adenosylhomocysteine and DNA methylation. Physiol. Plant. 93:5-10. Mufioz, C.A.P., Webster, B.D., and Jernstedt, J.A. 1995. Spatial congruence between exine pattern, microtubules and endomembranes in Vigna pollen. Sex. Plant Reprod. 8:147-151. Muntz, K. 1987. Developmental control of storage protein formation and its modulation by some internal and external factors during embryogenesis in plant seeds. Biochem. Physiol. Pflanz. 182:93-116. Murai, N., Sutton, D.W., Murray, M.G., Slightom, J.L., Merlo, D.J., Reichert, N.A., Sengupta-Gopalan, C, Stock, C.A., Barker, R.F., Kemp, J.D., and Hall, T.C. 1983. Phaseolin gene from bean is expressed after transfer to sunflower via tumor-inducing plasmid vectors. Science 222:476-482. Murfett, J., Bourque, J.E., and McClure, B.A. 1995. Antisense suppression of S-RNase expression in Nicotiana using RNA polymerase II- and Ill-transcribed gene constructs. Plant Mol. Biol. 29:201-212. Murfett, J., Atherton, T.L., Mou, B., Gasser, C.S., and McClure, B.A. 1994. S-RNase expressed in transgenic Nicotiana causes S-allele-specific pollen rejection. Nature 367:563-566. Murfett, J., Cornish, E.C., Ebert, P.R., Bonig, I., McClure, B.A., and Clarke, A.E. 1992. Expression of a selfincompatibility glycoprotein (S2-ribonuclease) from Nicotiana alata in transgenic Nicotiana tabncum. Plant Cell 4:1063-1074. Murfett, J., Ebert, PR., Haring, V, and Clarke, A.E. 1995. An S-RNase promoter from Nicotiana alata functions in transgenic N. alata plants but not in Nicotiana tabacum. Plant Mol. Biol. 28:957-963. Murfett, J., Strabala, T.J., Zurek, D.M., Mou, B., Beecher, B., and McClure, B.A. 1996. S RNase and interspecific pollen rejection in the genus Nicotiana: multiple pollen-rejection pathways contribute to unilateral incompatibility between self-incompatible and selfcompatible species. Plant Cell 8:943-958. Murgia, M., and Wilms, H.J. 1988. Three-dimensional image and mitochondrial distribution in sperm cells of Euphorbia dulcis. In Plant Sperms Cells as Tools for Biotechnology, ed. H.J. Wilms and C.J. Keijzer, pp. 75-79. Wageningen: Pudoc. Murgia, M., Charzyrlska, M., Rougier, M., and Cresti, M.
672
References
1991a. Secretory tapetum of Brassica oleracea L.: agriculturally important lupin species (Lupinus polarity and ultrastructural features. Sex. Plant angustifolius, L. albus, L. mutabilis). Plant Cell Tissue Reprod. 4:28-35. Organ Cult. 28:19-25. Murgia, M., Detchepare, S., van Went, J.L., and Cresti, M. Nagahashi, J., and Beevers, L. 1978. Subcellular localization of glycosyl transferases involved in glycopro1991b. Brassica napus pollen development during generative cell and sperm cell formation. Sex. Plant tein biosynthesis in the cotyledons of Pisum sativum L. Plant Physiol. 61:451-i59. Reprod. 4:176-181. Murgia, M., Huang, B.-Q., Tucker, S.C., and Musgrave, Nagata, R., Wada, M., Satoh, S., Kamada, H., and Harada, M.E. 1993. Embryo sac lacking antipodal cells in H. 1993. A cDNA differentially expressed between Arabidopsis thaliana (Brassicaceae). Am. J. Bot. embryogenic and nonembryogenic carrot (Daucus 80:824-838. carota L.) cells. Plant Physiol. 103:1021. Murgia, M., Wilms, H.J., Cresti, M., and Cesca, G. 1986. Nagato, Y. 1978. Analyses on the growth of embryo in Infrastructure of pollen development in Euphorbia some Gramineae. Jap. }. Breed. 28:97-105. dulcis L. 1. Diploid plants. Ada Bot. Need. 35:405-424. Nagato, Y, 1979. Incorporation of 3H-uridine and 3HMurphy, D.J., and Cummins, I. 1989a. Purification and leucine during early embryogenesis of rice and barimmunogold localisation of the major oil-body ley in caryopsis culture. Plant Cell Physiol. membrane protein of oilseed rape. Plant Sci. 20:765-773. 60:47-54. Nagato, Y., Kitano, Y, Kamijima, O., Kikuchi, S., and Murphy, D.J., and Cummins, I. 1989b. Biosynthesis of Satoh, H. 1989. Developmental mutants showing seed storage products during embryogenesis in abnormal organ differentiation in rice embryos. rapeseed, Brassica napus. J. Plant Physiol. 135:63-69. Theor. Appl. Genet. 78:11-15. Murphy, D.J., Cummins, I., and Kang, A.S. 1989. Synthesis Nagl, W. 1967. Die Riesenchromosomen von Phaseolus coccineus L.: Baueigentumlichkeiten, Strukturmodifikationen, of the major oil-body membrane protein in developzusatzliche Nukleolen und Vergleich mit den mitoing rapeseed (Brassica napus) embryos. Biochem. J. tischen Chromosomen. Osterr. Bot. Z. 114:171-182. 258:285-293. Murphy, D.J., Cummins, I., and Ryan, AJ. 1989. Nagl, W. 1969a. Banded polytene chromosomes in the Immunocytochemical and biochemical study of the legume Phaseolus vulgaris. Nature 221:70-71. biosynthesis and mobilisation of the major seed stor- Nagl, W. 1969b. Puffing of polytene chromosomes in a plant age proteins of Brassica napus. Plant Physiol. Biochem. (Phaseolus vulgaris).Naturwissenscliaften 56:221-222. 27:647-657. Nagl, W. 1974. The Phaseolus suspensor and its polytene chromosomes. Z. Pflanzenphysiol. 73:1-44. Murphy, J.J., and Dalby, A. 1971. Changes in the protein fractions of developing normal and opaque-2 maize Nagl, W. 1976a. The polytenic antipodal cells in Scilla bifoendosperm. Cereal Chem. 48:336-349. lia: DNA replication pattern and possibility of nucleMurray, B.G. 1990. Heterostyly and pollen-tube interacolar DNA amplification. Cytobiologie 14:165-170. tions in Luculia gratissima (Rubiaceae). Ann. Bot. Nagl, W. 1976b. Early embryogenesis in Tropaeolum majus 65:691-698. L.: ultrastructure of the embryo-suspensor. Biochem. Murray, M.G., and Kennard, W.C. 1984. Altered chroPhysiol. Pflanz. 170:253-260. matin conformation of the higher plant gene phase- Nagl, W. 1976c. Ultrastructural and developmental olin. Biochemistry 23:4225-4232. aspects of autolysis in embryo-suspensors. Ber. Deut. Bot. Ges. 89:301-311. Murty, N.S.R., and Lakshmi, N. 1979. Male sterile mutant Nagl, W. 1977. "Plastolysomes" - plastids involved in the in Capsicum annuum L. Curr. Sci. 48:312. Muschietti, J., Dircks, L., Vancanneyt, G., and McCormick, autolysis of the embryo-suspensor in Phaseolus. Z. Pflanzenphysiol. 85:45-51. S. 1994. LAT52 protein is essential for tomato pollen development: pollen expressing antisense LAT52 Nagl, W. 1978. Endopolyploidy and Polyteny in RNA hydrates and germinates abnormally and canDifferentiation and Evolution. Amsterdam: Northnot achieve fertilization. Plant J. 6:321-338. Holland Publishing Co. Musgrave, M.E., Antonovics, J., and Siedow, J.N. 1986. Is Nagl, W. 1990. Translocation of putrescine in the ovule, male-sterility in plants related to lack of cyanidesuspensor and embryo of Phaseolus coccineus. J. Plant resistant respiration in tissues? Plant Sci. 44:7-11. Physiol. 136:587-591. Mutschler, M.A., Bliss, F.A., and Hall, T.C. 1980. Variation Nagl, W. 1992. The polytenic endosperm haustorium of in the accumulation of seed storage protein among Rhinanthus minor (Scrophulariaceae): functional genotypes of Phaseolus vulgaris (L.). Plant Physiol. ultrastructure. Can. J. Bot. 70:1997-2004. 65:627-630. Nagl, W., and Kiihner, S. 1976. Early embryogenesis in Tropaeolum majus L.: diversification of plastids. Nacken, W.K.F., Huijser, P., Beltran, J.-P, Saedler, H., and Planta 133:15-19. Sommer, H. 1991a. Molecular characterization of Naito, S., Dube, PH., and Beachy, R.N. 1988. Differential two stamen-specific genes, tapl and fill, that are expression of conglycinin a' and |3 subunit genes in expressed in the wild type, but not in the deficiens transgenic plants. Plant Mol. Biol. 11:109-123. mutant of Antirrhinum majus. Mol. Gen. Genet. Naito, S., Hirai, M.Y., Chino, M., and Komeda, Y. 1994a. 229:129-136. Expression of a soybean (Glycine max [L.] Merr.) seed storage protein gene in transgenic Arabidopsis Nacken, W.K.F., Huijser, P., Saedler, H., and Sommer, H. 1991b. Molecular analysis of tap2, an anther-specific thaliana and its response to nutritional stress and to gene from Antirrhinum majus. FEBS Lett. 280:155-158. abscisic acid mutations. Plant Physiol. 104:497-503. Nadeau, J.A., Zhang, X.S., Li, J., and O'Neill. S.D. 1996. Naito, S., Inaba-Higano, K., Kumagai, T., Kanno, T., Ovule development: identification of stage-specific Nambara, E., Fujiwara, T., Chino, M., and Komeda, and tissue-specific cDNAs. Plant Cell 8:213-239. Y. 1994b. Maternal effects of mtol mutation, that causes overaccumulation of soluble methionine, on Nadolska-Orczyk, A. 1992. Somatic embryogenesis of
References
613
the expression of a soybean (3-conglycinin gene pro- Nasrallah, J.B., and Nasrallah, M.E. 1984. Electrophoretic heterogeneity exhibited by the S-allele specific glymoter-GUS fusion in transgenic Arabidopsis thaliana. coproteins of Brassica. Experientia 40:279-281. Plant Cell Physiol. 35:1057-1063. Nakagawa, H., Ohmiya, Kv and Hattori, T. 1996. A rice Nasrallah, J.B., and Nasrallah, M.E. 1989. The molecular genetics of self-incompatibility in Brassica. Annu. bZIP protein, designated OSBZ8, is rapidly induced Rev. Genet. 23:123-139. by abscisic acid. Plant J. 9:217-227. Nakajima, T. 1962. Physiological studies of seed develop- Nasrallah, J.B., and Nasrallah, M.E. 1993. Pollen-stigma signaling in the sporophytic self-incompatibility ment, especially embryonic growth and endosperm response. Plant Cell 5:1325-1335. development. Bull. Univ. Osaka Pref. Ser. B 13:13-48. Nakamura, N., Sado, M., and Arai, Y. 1980. Sucrose Nasrallah, J.B., Nishio, T., and Nasrallah, M.E. 1991. The self-incompatibility genes of Brassica: expression and metabolism during the growth of Camellia japonica use in genetic ablation of floral tissues. Annu. Rev. pollen. Phytochemistry 19:205-209. Plant Physiol. Plant Mol. Biol. 42:393^122. Nakamura, N., and Suzuki, H. 1981. Sugar composition of pollen grain and pollen tube cell walls. Nasrallah, J.B., Rundle, S.J., and Nasrallah, M.E. 1994. Genetic evidence for the requirement of the Brassica Phytochemistry 20:981-984. S-locus receptor kinase gene in the self-incompatiNakamura, N., Fukushima, A., Iwayama, H., and Suzuki, bility response. Plant ]. 5:373-384. H. 1991. Electrotropism of pollen tubes of camellia Nasrallah, J.B., Yu, S.-M., and Nasrallah, M.E. 1988. Selfand other plants. Sex. Plant Repwd. 4:138-143. incompatibility genes of Brassica oleracea: expression, Nakamura, S. 1979. Development of the pollen grain wall in Lilium longiflorum. ]. Electron Micros. 28:275-284. isolation and structure. Proc. Natl Acad. Sci. USA Nakamura, S., and Miki-Hirosige, H. 1982. Coated vesi85:5551-5555. cles and cell plate formation in the microspore Nasrallah, J.B., Kao, T.-H., Chen, C.-H., Goldberg, M.L., mother cell. /. Ultrastr. Res. 80:302-311. and Nasrallah, M.E. 1987. Amino-acid sequence of glycoproteins encoded by three alleles of the S locus Nakamura, S., and Miki-Hirosige, H. 1985. Fine-structural of Brassica oleracea. Nature 326:617-619. study on the formation of the generative cell wall and intine-3 layer in a growing pollen grain of Lilium Nasrallah, J.B., Kao, T.-H., Goldberg, M.L., and Nasrallah, longiflorum. Am. J. Bot. 72:365-375. M.E. 1985. A cDNA clone encoding an S-locus-specific glycoprotein from Brassica oleracea. Nature Nakamura, S., Ikehara, T., Uchida, H., Suzuki, T., and 318:263-267. Sodmergen. 1992. Fluorescence microscopy of plastid nucleoids and a survey of nuclease C in higher Nasrallah, J.B., Stein, J.C., Kandasamy, M.K., and plants with respect to mode of plastid inheritance. Nasrallah, M.E. 1994. Signaling the arrest of pollen Protoplasma 169:68-74. tube development in self-incompatible plants. Science 266:1505-1508. Nakanishi, T., Esashi, Y, and Hinata, K. 1969. Control of self-incompatibility by CO2 gas in Brassica. Plant Cell Nasrallah, M.E. 1974. Generic control of quantitative variation in self-incompatibility proteins detected by Physiol. 10:925-927. immunodiffusion. Genetics 76:45-50. Nakanishi, T., and Hinata, K. 1973. An effective time for CO2 gas treatment in overcoming self-incompatibil- Nasrallah, M.E. 1979. Self-incompatibility antigens and S ity in Brassica. Plant Cell Physiol. 14:873-879. gene expression in Brassica. Heredity 43:259-263. Nakanishi, T.M., Saeki, N., Maeno, M., Ozaki, T., Kawai, Nasrallah, M.E., Barber, J.T., and Wallace, D.H. 1970. Selfincompatibility proteins in plants: detection, genetY., and Ichii, T. 1991. Ultrastructural study on the ics, and possible mode of action. Heredity 25:23-27. stylar transmitting tissue in Japanese pear. Sex. Plant Reprod. 4:95-103. Nasrallah, M.E., Kandasamy, M.K., and Nasrallah, J.B. 1992. A genetically defined trans-acting locus reguNakano, M., Hoshino, Y, and Mii, M. 1994. Regeneration lates S-locus function in Brassica. Plant /. 2:497-506. of transgenic plants of grapevine (Vitis vinifera L.) via Agrobacterium r/iizogenes-mediated transformation of Nasrallah, M.E., and Nasrallah, J.B. 1986. Molecular biolembryogenic calli. /. Expt. Bot. 45:649-656. ogy of self-incompatibility in plants. Trends Genet. 2:239-244. Nambara, E., Naito, S., and McCourt, P. 1992. A mutant of Arabidopsis which is defective in seed development Nasrallah, M.E., and Wallace, D.H. 1967a. Immunoand storage protein accumulation is a new abi3 genetics of self-incompatibility in Brassica oleracea L. allele. Plant J. 2:435-441. Heredity 22:519-527. Nambara, E., Keith, K., McCourt, P., and Naito, S. 1995. A Nasrallah, M.E., and Wallace, D.H. 1967b. Immunoregulatory role for the AB13 gene in the establishchemical detection of antigens in self-incompatibilment of embryo maturation in Arabidopsis thaliana. ity genotypes of cabbage. Nature 213:700-701. Nasrallah, M.E., Wallace, D.H., and Savo, R.M. 1972. Development 121:629-636. Genotype, protein, phenotype relationships in selfNarayanaswamy, S., and George, L. 1972. Morphogenesis incompatibility of Brassica. Genet. Res. 20:151-160. of belladonna (Atropa belladonna L.) plantlets from pollen in culture. Indian J. Expt. Bid. 10:382-384. Natesh, S., and Rau, M.A. 1984. The embryo. In Nasatir, M., and Stern, H. 1959. Changes in the activities Embryology of Angiosperms, ed. B.M. Johri, pp. of aldolase and D-glyceraldehyde-3-phosphate 377-443. Berlin: Springer-Verlag. dehydrogenase during the mitotic cycle in Nath, J., and Watson, C.V. 1980. Acid phosphatase microspores of Lilium longiflorum.}. Biophys. Biochem. changes associated with development of male sterile Cytol. 6:189-192. and fertile maize (Zea mays L.). Biochem. Genet. Nasrallah, J.B., Doney, R.C., and Nasrallah, M.E. 1985. 18:377-387. Biosynthesis of glycoproteins involved in the Naumova, T.N. 1993. Apomixis in Angiosperms. Nucellar pollen-stigma interaction of incompatibility in and Integumentary Embryony. Boca Raton: CRC Press. developing flowers of Brassica oleracea L. Planta Naumova, T.N., and Willemse, M.T.M. 1995. Ultrastructural characterization of apospory in 165:100-107.
614
References
Panicum maximum. Sex. Plant Reprod. 8:107-204. A., Blagrove, R.J., March, J.F., Kortt, A.A., and Nave, E.B., and Sawhney, V.K. 1986. Enzymatic changes Higgins, T.J.V. 1990. Pea convicilin: structure and in post-meiotic anther development in Petunia primary sequence of the protein and expression of a hybrida. I. Anther ontogeny and isozyme analyses. /. gene in the seeds of transgenic tobacco. Planta Plant Physiol. 125:451^65. 180:461-170. Nawaschin, S.G. 1898. Resultate einer Revision der Newcomb, W. 1973a. The development of the embryo sac Befruchtungsvorgange bei Lilium martagon und of sunflower Helianthus annuus before fertilization. Fritillaria tenella. Bull. Acad. Imp. Sci. St. Petersburg Can. J. Bot. 51:863-878. 9:377-382. Newcomb, W. 1973b. The development of the embryo sac of sunflower Helianthus annuus after fertilization. Negruk, V.I., Cherny, D.I., Nififorova, I.D., Aleksandrov, A.A., and Butenko, R.G. 1982. Isolation and characCan. J. Bot. 51:879-890. terization of minicircular DNAs found in mitochon- Newcomb, W. 1978. The development of cells in the drial fraction of Viciafaba. FEBS Lett. 142:115-117. coenocytic endosperm of the African blood lily Haemanthus katherinae. Can. J. Bot. 56:483-501. Nehra, N.S., Chibbar, R.N., Leung, N., Caswell, K., Mallard, C, Steinhauer, L., Baga, M., and Kartha, Newcomb, W., and Fowke, L.C. 1973. The fine structure of K.K. 1994. Self-fertile transgenic wheat plants regenthe change from the free-nuclear to cellular condierated from isolated scutellar tissues following tion in the endosperm of chickweed Stellaria media. microprojectile bombardment with two distinct gene Bot. Gaz. 134:236-241. constructs. Plant ]. 5:285-297. Newcomb, W., and Fowke, L.C. 1974. Stellaria media Neill, J.D., Litts, J.C., Anderson, O.D., Greene, F.C., and embryogenesis: the development and ultrastructure Stiles, J.I. 1987. Expression of a wheat a-gliadin gene of the suspensor. Can. ]. Bot. 52:607-614. in Saccharomyces cerevisiae. Gene 55:303-317. Newcomb, W., and Steeves, T.A. 1971. Helianthus annuus embryogenesis: embryo sac wall projections before Neill, S.J., Horgan, R., and Parry, A.D. 1986. The and after fertilization. Bot. Gaz. 132:367-371. carotenoid and abscisic acid content of viviparous kernels and seedlings of Zea mays L. Planta 169:87-96. Newton, C.H., Flinn, B.S., and Sutton, B.C.S. 1992. VicilinNeill, S.J., Horgan, R., and Rees, A.F. 1987. Seed developlike seed storage proteins in the gymnosperm intement and vivipary in Zea mays L. Planta 171:358-364. rior spruce (Picea glauca/engelmanii). Plant Mol. Biol. Nelson, O.E., Chourey, P.S., and Chang, M.T. 1978. 20:315-322. Nucleoside diphosphate sugar-starch glucosyl trans- Ng, J.D., Ko, T.-Z., and McPherson, A. 1993. Cloning, ferase activity of wx starch granules. Plant Physiol. expression, and crystallization of jack bean 62:383-386. (Canavalia ensiformis) canavalin. Plant Physiol. 101:713-728. Nelson, O.E., Mertz, E.T., and Bates, L.S. 1965. Second mutant gene affecting the amino acid pattern of Nielsen, K.A., and Hansen, I.B. 1992. Appearance of extramaize endosperm proteins. Science 150:1469-1470. cellular proteins associated with somatic embryogeNelson, O.E., and Rines, H.W. 1962. The enzymatic definesis in suspension cultures of barley (Hordeum ciency in the waxy mutant of maize. Biochem. vulgare L.). /. Plant Physiol. 139:489-497. Biophys. Res. Commun. 9:297-300. Nielsen, N.C., Dickinson, CD., Cho, T.-J., Thanh, V.H., Nelson, O.E., and Tsai, C.Y. 1964. Glucose transfer from Scallon, B.J., Fischer, R.L., Sims, T.L., Drews, G.N., adenosine diphosphate-glucose to starch in preparaand Goldberg, R.B. 1989. Characterization of the tions of waxy seeds. Science 145:1194-1195. glycinin gene family in soybean. Plant Cell 1:313-328. Nelson, P.M., and Rossman, E.C. 1958. Chemical induc- Niimi, Y. 1970. In vitro fertilization in the self-incompatition of male sterility in inbred maize by use of gibble plant, Petunia hybrida. J. Jap. Soc. Hort. Sci. berellins. Science 127:1500-1501. 39:346-352. Nepi, M., Ciampolini, F., and Pacini, E. 1995. Niitsu, T., Hanaoka, A., and Uchiyama, K. 1972. Development of Cucurbita pepo pollen: ultrastructure Reinvestigations of the spindle body and related and histochemistry of the sporoderm. Can. J. Bot. problems. II. Effects of cycloheximide upon the 73:1046-1057. development of kinetochore fibers studied in the pollen mitosis of Ornithogalum virens in vivo. Neto, G.C., Yunes, J.A., da Silva, M.J., Vettore, A.L., Cytologia 37:143-154. Arruda, P., and Leite, A. 1995. The involvement of opaque 2 on pS-prolamin gene regulation in maize Niizeki, H., and Oono, K. 1968. Induction of haploid rice and Coix suggests a more general role for this tranplant from anther culture. Proc. ]ap. Acad. 44:554-557. scriptional activator. Plant Mol. Biol. 27:1015-1029. Ninnemann, H., and Epel, B. 1973. Inhibition of cell diviNeuffer, M.G., and Sheridan, W.F. 1980. Defective kernel sion by blue light. Expt. Cell Res. 79:318-326. mutants of maize. I. Genetic and lethality studies. Niogret, M.F., Culianez-Macia, Goday, A., Alba, M.M., Genetics 95:929-944. and Pages, M. 1996. Expression and cellular localizaNeuhaus, G., Spangenberg, G., Scheid, O.M., and tion of rab28 mRNA and RAB28 protein during Schweiger, H.-G. 1987. Transgenic rapeseed plants maize embryogenesis. Plant ]. 9:549-557. obtained by the microinjection of DNA into Niogret, M.-F., Dubald, M., Mandaron, P., and Mache, R. microspore-derived embryoids. Theor. Appl. Genet. 1991. Characterization of pollen polygalacturonase 75:30-36. encoded by several cDNA clones in maize. Plant Neumann, D., and Weber, E. 1978. Formation of protein Mol. Biol. 17:1155-1164. bodies in ripening seeds of Vicia faba L. Biochem. Nirmala, C, and Kaul, M.L.H. 1994. Male sterility in pea. Physiol. Pflanz. 173:167-180. V. Gene action during heterotypic and homotypic Newbigin, E., Anderson, M.A., and Clarke, A.E. 1993. divisions. Cytologia 59:43-50. Gametophytic self-incompatibility systems. Plant Nishihara, M., Ito, M., Tanaka, I., Kyo, M., Ono, K., Cell 5:1315-1324. Irifune, K., and Morikawa, H. 1993. Expression of the p-glucuronidase gene in pollen of lily (Lilium Newbigin, E.J., deLumen, B.O., Chandler, P.M., Gould,
References
615
longiflorum), tobacco (Nicotiana tabacum), Nicotiana megagametogenesis in two species of Oenothera. rustica, and peony (Paeonia lactiflora) by particle Ann. Bot. 46:473-477. bombardment. Plant Physiol. 102:357-361. Noher de Halac, I., Cismondi, I.A., and Harte, C. 1990. Pollen ontogenesis in Oenothera: a comparison of Nishio, T., and Hinata, K. 1977. Analysis of S-specific progenotypically normal anthers with the male-sterile teins in stigma of Brassica olemcea L. by isoelectric mutant sterilis. Sex. Plant Reprod. 3:41-53. focusing. Heredity 38:391-396. Nishio, T., and Hinata, K. 1982. Comparative studies on S- Noher de Halac, I., and Harte, C. 1977. Different patterns glycoproteins purified from different S-genotypes in of callose wall formation during megasporogenesis self-incompatible Brassica species. I. Purification and in two species of Oenothera (Onagraceae). Plant Syst. Evol. 127:23-38. chemical properties. Genetics 100:641-647. Nishio, T., Toriyama, K., Sato, T., Kandasamy, M.K., Noma, M., Huber, J., Ernst, D., and Pharis, R.P. 1982. Paolillo, D.J., Nasrallah, J.B., and Nasrallah, M.K. Quantitation of gibberellins and the metabolism of [3H]gibberellin A, during somatic embryogenesis 1992. Expression of S-locus glycoprotein genes from Brassica oleracea and B. campestris in transgenic plants in carrot and anise cell cultures. Planta of self-compatible B. napus cv Westar. Sex. Plant 155:369-376. Reprod. 5:101-109. Nomura, K., and Komamine, A. 1985. Identification and isolation of single cells that produce somatic Nishizawa, N.K., Mori, S., Watanabe, Y, and Hirano, H. embryos at a high frequency in a carrot suspension 1994. Ultrastructural localization of the basic 7S culture. Plant Physiol. 79:988-991. globulin in soybean (Glycine max) cotyledons. Plant Cell Physiol. 35:1079-1085. Nomura, K., and Komamine, A. 1986a. Polarized DNA Nissen, P. 1994. Stimulation of somatic embryogenesis in synthesis and cell division in cell clusters during carrot by ethylene: effects of modulators of ethylsomatic embryogenesis from single carrot cells. Neiu Phytol. 104:25-32. ene biosynthesis and action. Physiol. Plant. 92:397-403. Nomura, K., and Komamine, A. 1986b. In situ hybridization on tissue sections. A method for detection of Nitsch, C. 1974a. Pollen culture - a new technique for poly(A)+RNA. Plant Tissue Cult. Lett. Japan 3:92-94. mass production of haploid and homozygous plants. In Haploids in Higher Plants. Advances and Nomura, K., and Komamine, A. 1995. Physiological and biochemical aspects of somatic embryogenesis. In In Potential, ed. KJ. Kasha, pp. 123-135. Guelph: The Vitro Embryogenesis in Plants, ed. T.A. Thorpe, pp. University of Guelph. 249-265. Dordrecht: Kluwer Academic Publishers. Nitsch, C. 1974b. La culture de pollen isol£ sur milieu synth&ique. Compt. Rend. Acad. Set Paris 278D:1031-1034. Nong, V.H., Schlesier, B., Bassiiner, R., Repik, A., Horstmann, C, and Miintz, K. 1995. Narbonin, a Nitsch, C, and Norreel, B. 1973. Effet d'un choc thernovel 2S protein from Vicia narbonensis L. seeds: mique sur le pouvoir embryogene du pollen de Datura innoxia cultiv6 dans l'anthere ou isole de l'ancDNA, gene structure and developmentally reguthere. Compt. Rend. Acad. Sci. Paris 276D:303-306. lated formation. Plant Mol. Biol. 28:61-72. Nitsch, J.P. 1972. Haploid plants from pollen. Z. Norgaard, J.V., and Krogstrup, P. 1991. Cytokinin induced somatic embryogenesis from immature embryos of Pflanzenzuchtg 67:3-18. Abies nordmanniana Lk. Plant Cell Rep. 9:509-513. Nitsch, J.P., and Nitsch, C. 1969. Haploid plants from pollen grains. Science 163:85-87. Norreel, B. 1972. Etude comparative de la repartition des Nivison, H.T., and Hanson, M.R. 1989. Identification of a acides ribonucl&ques au cours de 1'embryogenese mitochondrial protein associated with cytoplasmic zygotique et de 1'embryogenese androge'ne'tique male sterility in Petunia. Plant Cell 1:1121-1130. chez le Nicotiana tabacum L. Compt. Rend. Acad. Sci. Nivison, H.T., Sutton, C.A., Wilson, R.K., and Hanson, Paris 275D:1219-1222. M.R. 1994. Sequencing, processing, and localization Norstog, K. 1961. The growth and differentiation of culof the petunia cms-associated mitochondrial protein. tured barley embryos. Am. }. Bot. 48:876-884. Plant ]. 5:613-623. Norstog, K. 1972a. Early development of the barley Nobiling, R., and Reiss, H.-D. 1987. Quantitative analysis embryo: fine structure. Am. J. Bot. 59:123-132. of calcium gradients and activity in growing pollen Norstog, K. 1972b. Factors relating to precocious germitubes of Lilium longiflorum. Protoplasma 139:20-24. nation in cultured barley embryos. Phytomorphology Nogler, G.A. 1984. Gametophytic apomixis. In Embryology 22:134-139. of Angiosperms, ed. B.M. Johri, pp. 475-518. Berlin: Norstog, K. 1974. Nucellus during early embryogeny in Springer-Verlag. barley: fine structure. Bot. Gaz. 135:97-103. Noguchi, T. 1990. Consumption of lipid granules and for- Norstog, K., and Klein, R.M. 1972. Development of culmation of vacuoles in the pollen tube of Tradescantia tured barley embryos. II. Precocious germination reflexa. Protoplasma 156:19-28. and dormancy. Can. }. Bot. 50:1887-1894. Noguchi, X, and Morr£, D.J. 1991. Membrane flow in Norstog, K., and Smith, J.E. 1963. Culture of small barley plants: preparation and kinetics of labelling of embryos on defined media. Science 142:1655-1656. plasma membranes from growing pollen tubes of Nozue, K., and Wada, M. 1993. Electrotropism of tobacco. Protoplasma 163:34-42. Nicotiana pollen tubes. Plant Cell Physiol. Noguchi, T., and Ueda, K. 1990. Structure of pollen grains 34:1291-1296. of Tradescantia reflexa with special reference to theNunberg, A.N., Li, Z., Bogue, M.A., Vivekananda, J., generative cell and the ER around it. Cell Struct. Reddy, A.S., and Thomas, T.L. 1994. Developmental Fund. 15:379-384. and hormonal regulation of sunflower helianthinin Noher de Halac, I. 1980a. Fine structure of nucellar cells genes: proximal promoter sequences confer regionduring development of the embryo sac in Oenothera alized seed expression. Plant Cell 6:473-486. biennis L. Ann. Bot. 45:515-521. Nuti Ronchi, V., Caligo, M.A., Nozzolini, M., and Noher de Halac, I. 1980b. Callose deposition during Luccarini, G. 1984. Stimulation of carrot somatic
676
References
embryogenesis by proline and serine. Plant Cell Rep. allows endosperm development without fertiliza3:210-214. tion. Proc. Natl Acad. Sci. USA 93:5319-5324. Nuti Ronchi, V., Giorgetti, L., Tonelli, M., and Martini, G. Ohl, S., Hedrick, S.A., Chory, J., and Lamb, C.J. 1990. 1992a. Ploidy reduction and genome segregation in Functional properties of a phenylalanine ammoniacultured carrot cell lines. I. Prophase chromosome lyase promoter from Arabidopsis. Plant Cell reduction. Plant Cell Tissue Organ Cult. 30:107-114. 2:837-848. Nuti Ronchi, V., Giorgetti, L., Tonelli, M. and Martini, G. Ohshika, K., and Ikeda, H. 1994. Isolation and preserva1992b. Ploidy reduction and genome segregation in tion of the living embryo sac of Crinum asiaticum L. cultured carrot cell lines. II. Somatic meiosis. Plant var. japonicum Baker. /. Plant Res. 107:17-21. Cell Tissue Organ Cult. 30:115-119. Ohta, Y. 1986. High-efficiency genetic transformation of maize by a mixture of pollen and exogenous DNA. Nygaard, P. 1977. Utilization of exogenous carbohydrates Proc. Natl Acad. Sci. USA 83:715-719. for tube growth and starch synthesis in pine pollen suspension cultures. Physiol. Plant. 39:206-210. Ohyama, T., Iwaikawa, Y, Kobayashi, T., Hotta, Y., and Tabata, S. 1992. Isolation of synaptonemal complexes Nyman, L.P., Webb, E.L., Gu, Z., and Arditti, J. 1986. from lily microsporocytes. Plant Sci. 86:115-124. Structure and in vitro growth of zygotic embryos of taro (Colocasia esculenta var. antiquorum). Ann. Bot.Oishi, M.Y., and Bewley, J.D. 1992. Premature drying, fluri57:623-630. done-treatment, and embryo isolation during development of maize kernels (Zea mays L.) induce Nyman, L.P., Webb, E.L., Gu, Z., and Arditti, J. 1987. germination, but the protein synthetic responses are Effects of growth regulators and glutamine on in different. Potential regulation of germination and provitro development of zygotic embryos of taro tein synthesis by abscisic acid. /. Expt. Bot. 43:759-767. (Colocasia esculenta var. antiquorum). Ann. Bot. 59:517-523. Okada, K., Ueda, J., Komaki, M.K., Bell, C.J., and Shimura, Y. 1991. Requirement of the auxin polar transport system in early stages of Arabidopsis floral O'Brien, S.P. 1994. Pistil structure and pollen tube pathways in Leptospermum myrsinoides and L. continentale bud formation. Plant Cell 3:677-684. (Myrtaceae). Ann. Bot. 73:225-230. Okamoto, H., Yano, A., Shiraishi, H., Okada, K., and O'Driscoll, D., Hann, C., Read, S.M., and Steer, M.W. 1993. Shimura, Y. 1994. Genetic complementation of a floEndocytotic uptake of fluorescent dextrans by pollen ral homeotic mutation, apelalai, with an Arabidopsis tubes grown in vitro. Protoplasma 175:126-130. thaliana gene homologous to DEFICIENS of Antirrhinum majus. Plant Mol. Biol. 26:465-472. O'Keefe, D.P., Tepperman, J.M., Dean, C., Leto, K.J., Erbes, D.L., and Odell, J.T. 1994. Plant expression of a bacter- Okamuro, J.K., den Boer, B.G.W., and Jofuku, K.D. 1993. Regulation of Arabidopsis flower development. Plant ial cytochrome P450 that catalyzes activation of a sulCell 5:1183-1193. fonylurea pro-herbicide. Plant Physiol. 105:473-482. O'Neill, P., Singh, M.B., and Knox, R.B. 1988. Cell biology Okamuro, J.K., and Goldberg, R.B. 1992. A mutant lectin gene is rescued from an insertion element that blocks of the stigma of Brassica campestris in relation to CO2 its expression. Plant Cell 4:1141-1146. effects on self-pollination. /. Cell Sci. 89:541-549. O'Neill, P.M., Singh, M.B., and Knox, R.B. 1989. Okamuro, J.K., Jofuku, K.D., and Goldberg, R.B. 1986. Biosynthesis of S-associated proteins following selfSoybean seed lectin gene and flanking nonseed proand cross-pollinations in Brassica campestris L. var. tein genes are developmentally regulated in transT.15'. Sex. Plant Reprod. 2:103-108. formed tobacco plants. Proc. Natl Acad. Sci. USA 83:8240-8244. O'Neill, S.D., Nadeau, J.A., Zhang, X.S., Bui, A.Q., and Halevy, A.H. 1993. Interorgan regulation of ethylene Okita, T.W. 1984. Identification and DNA sequence analybiosynthetic genes by pollination. Plant Cell sis of a y-type gliadin cDNA plasmid from winter 5:419-432. wheat. Plant Mol. Biol. 3:325-332. Obermeyer, G., and Blatt, M.R. 1995. Electrical properties Okita, T.W., Cheesbrough, V., and Reeves, CD. 1985. Evolution and heterogeneity of the a-/P-type and yof intact pollen grains of Lilium longiflorum: charactype gliadin DNA sequences./. Biol. Chem. 260:8203-8213. teristics of the non-germinating pollen grain. /. Expt. Bot. 46:803-813. Okita, T.W., and Greene, F.C. 1982. Wheat storage proteins. Isolation and characterization of the gliadin Obermeyer, G., and Weisenseel, M.H. 1991. Calcium messenger RNAs. Plant Physiol. 69:834-839. channel blocker and calmodulin antagonists affect the gradient of free calcium ions in lily pollen tubes. Okita, T.W., Hwang, Y.S., Hnilo, J., Kim, W.T., Aryan, A.P., Larson, R., and Krishnan, H.B. 1989. Structure and Eur. J. Cell Biol. 56:319-327. expression of the rice glutelin multigene family. /. Obermeyer, G., Liitzelschwab, M., Heumann, H.-G., and + Biol. Chem. 264:12573-12581. Weisenseel, M.H. 1992. Immunolocalization of H ATPases in the plasma membrane of pollen grains Okonkwo, S.N.C., and Raghavan, V. 1982. Studies on the and pollen tubes of Lilium longiflorum. Protoplasma germination of seeds of the root parasites, Alectra 171:55-63. vogelii and Striga gesnerioides. I. Anatomical changes in the embryos. Am. }. Bot. 69:1636-1645. Ockendon, D.J. 1968. Biosystematic studies in the Linum Oldknow, J., and Trick, M. 1995. Genomic sequence of an perenne group. New Phytol. 67:787-813. SRK-Iike gene linked to the S-locus of a self-incomOckendon, D.J., and Gates, P.J. 1975. Growth of cross- and patible Brassica oleracea line. Sex. Plant Reprod. self-pollen tubes in the styles of Brassica oleracea. Neiu 8:247-253. Phytol. 75:155-160. Ogur, M., Erickson, R.O., Rosen, G.U., Sax, K.B., and Oldknow, J., Franklin, T.M., Allard, S., Robert, L.S., and Trick, M. 1995. DNA sequences of the two homoeolHolden, C. 1951. Nucleic acids in relation to cell diviogous SLR1 genes of Brassica napus cv Westar. Sex. sion in Lilium longiflorum. Expt. Cell Res. 2:73-89. Plant Reprod. 8:254-255. Ohad, N., Margossian, L., Hsu, Y.-C, Williams, C, Repetti, P., and Fischer, R.L. 1996. A mutation that Olesen, P., and Bruun, L. 1990. A structural investigation
References
617
of the ovule in sugar beet, Beta vulgaris: integuments Ottaviano, E., and Mulcahy, D.L. 1989. Genetics of angiosperm pollen. Adv. Genet. 26:1-64. and micropyle. Nord. J. Bot. 9:499-506. Olff, H v Kuiper, D., van Damme, J.M.M., and Kuiper, Ottoboni, L.M.M., Leite, A., Yunes, J.A., Targon, M.L.P.N., P.J.C. 1989. Gynodioecy in Plantago lanceolata. VI. de Souza Filho, G.A., and Arruda, P. 1993. Sequence Functions of cytokinins in growth, development, analysis of 22 kDa-like a-coixin genes and their comand reproduction of two sex types. Can. ]. Bot. parison with homologous zein and kafirin genes 67:2765-2769. reveals highly conserved protein structure and regulatory elements. Plant Mol. Biol. 21:765-778. Olmedilla, A., Schrauwen, J.A.M., and Wullems, GJ. 1991. Visualization of starch-synthase expression by in situ Ouellet, T., Rutledge, R.G., and Miki, B.L. 1992. Members hybridization during pollen development. Planta of the acetohydroxyacid synthase multigene family 184:182-186. of Brassica napus have divergent patterns of expression. Plant J. 2:321-330. Olsen, O.-A., Jakobsen, K.S., and Schmelzer, E. 1990. Development of barley aleurone cells: temporal and Ouyang, T.-W., Hu, H., Chuang, C.-C., and Tseng, C.-C. spatial patterns of accumulation of cell-specific 1973. Induction of pollen plants from anthers of mRNAs. Planta 181:462-466. Triticum aestivum L. cultured in vitro. Sci. Sinica 16:79-95. Olsen, O.-A., Potter, R.H., and Kalla, R. 1992. Histo-differentiation and molecular biology of developing cereal Overman, M.A., and Warmke, H.E. 1972. Cytoplasmic male sterility in Sorghum. II. Tapetal behavior in ferendosperm. Seed Sci. Res. 2:117-131. Olson, A.R. 1980. Seed morphology of Monotropa uniflora tile and sterile anthers. /. Hered. 63:226-234. L. (Ericaceae). Am. J. Bot. 67:968-974. Owen, H.A., and Makaroff, C.A. 1995. Ultrastructure Olson, A.R. 1991. Postfertilization changes in ovules of of microsporogenesis and microgametogenesis Monotropa uniflora L. (Monotropaceae). Am. ]. Bot. in Arabidopsis thaliana (L.) Heynh. ecotype 78:99-107. Wassilewskija (Brassicaceae). Protoplasma 185:7-21. Olson, A.R. 1993. Patterns of embryo and endosperm for- Owens, SJ. 1981. Self-incompatibility in the Commelmation in Monotropa hypopitys (Monotropaceae) inaceae. Ann. Bot. 47:567-581. from North America and western Sweden. Am. J. Owens, S.J., and Dickinson, H.G. 1983. Pollen wall develBot. 80:839-846. opment in Gibasis (Commelinaceae). Ann. Bot. Olson, A.R. 1994. Pollen tube pathway through the 51:1-15. gynoecium of Monotropsis odorata (Monotropaceae). Owens, S.J., and Horsfield, NJ. 1982. Light and electron Am. J. Bot. 81:718-725. microscopic study of stigmas in Aneilema and Olson, A.R., and Cass, D.D. 1981. Changes in megagameCommelina species (Commelinaceae). Protoplasma tophyte structure in Papaver nudicaule L. 112:26-36. (Papaveraceae) following in vitro placental pollina- Owens, S.J., and Kimmins, F.M. 1981. Stigma morphology tion. Am. J. Bot. 68:1333-1341. in Commelinaceae. Ann. Bot. 47:771-783. Oommen, A., Dixon, R.A., and Paiva, N.L. 1994. The elic- Ozias-Akins, P., and Vasil, I.K. 1982. Plant regeneration itor-inducible alfalfa isoflavone reductase promoter from cultured immature embryos and inflorescences confers different patterns of developmental expresof Triticum aestivum L. (wheat): evidence for somatic sion in homologous and heterologous transgenic embryogenesis. Protoplasma 110:95-105. plants. Plant Cell 6:1789-1803. Ozias-Akins, P., and Vasil, I.K. 1983. Improved efficiency and normalization of somatic embryogenesis in Ooms, J.J.J., Leon-Kloosterziel, K.M., Bartels, D., Koornneef, M., and Karssen, CM. 1993. Acquisition Triticum aestivum (wheat). Protoplasma 117:40-44. of desiccation tolerance and longevity in seeds of Arabidopsis thaliana. A comparative study using Pacini, E., Bellani, L.M., and Lozzi, R. 1986. Pollen, tapeabscisic acid-insensitive abi3 mutants. Plant Physiol. turn and anther development in two cultivars of 102:1185-1191. sweet cherry (Prunus avium). Phytomorphology Oparka, K.J., and Harris, N. 1982. Rice protein-body for36:197-210. mation: all types are initiated by dilation of the Pacini, E., and Casadoro, G. 1981. Tapetum plastids of endoplasmic reticulum. Planta 154:184-188. Olea europaea L. Protoplasma 106:289-296. Or, E., Boyer, S.K., and Larkins, B.A. 1993. Opaque! modi- Pacini, E., and Cresti, M. 1976. Close association between fiers act post-transcriptionally and in a polar manner plastids and endoplasmic reticulum cisterns during on y-zein gene expression in maize endosperm. Plant pollen grain development in Lycopersicon peruCell 5:1599-1609. vianum. ]. Ultrastr. Res. 57:260-265. Ori, N., Sessa, G., Lotan, T., Himmelhoch, S., and Fluhr, R. Pacini, E., and Cresti, M. 1978. Ultrastructural characteris1990. A major stylar matrix polypeptide (sp41) is a tics of tapetum and microspore mother cells in member of the pathogenesis-related proteins superLycopersicon peruvianum during meiotic prophase. class. EMBO }. 9:3429-3436. Soc. Bot. Fr. Actual. Bot. 125:121-128. Ornduff, R. 1964. The breeding system of Oxalis suksdorfii. Pacini, E., Cresti, M., and Sarfatti, G. 1972. Incorporation Am. J. Bot. 51:307-314. of integumentary nuclei in Eranthis hiemalis Ornduff, R. 1966. The breeding system of Pontederia corendosperm and their disaggregation by the endoplasmic reticulum. /. Submicros. Cytol. 4:19-31. data L. Bull. Torrey Bot. Cl. 93:407-416. Osborn, T.C., Alexander, D.C., Sun, S.S.M., Cardona, C, Pacini, E., Franchi, G.G., and Hesse, M. 1985. The tapeand Bliss, F.A. 1988. Insecticidal activity and lectin tum: its form, function, and possible phylogeny in homology of arcelin seed protein. Science Embryophyta. Plant Syst. Evol. 149:155-185. 240:207-210. Pacini, E., and Juniper, B.E. 1979a. The ultrastructure of 0sterbye, U. 1975. Self-incompatibility in Ranunculus acris pollen-grain development in the olive (Olea L. I. Genetic interpretation and evolutionary aspects. europaea). 1. Proteins in the pore. Neiu Phytol. Hereditas 80:91-112. 83:157-163.
678
References
Pacini, Ev and Juniper, B.E. 1979b. The ultrastructure of Pandey, K.K. 1958. Time of the S allele action. Nature pollen grain development in the olive (Oka 181:1220-1221. europaea). 2. Secretion by the tapetal cells. New Phytol. Pandey, K.K. 1960. Incompatibility in Abutilon 'hybridum.' 83:165-174. Am. ]. Bot. 47:877-883. Pacini, E., and Juniper, B.E. 1983. The ultrastructure of Pandey, K.K. 1964. Elements of the S-gene complex. Genet. the formation and development of the amoeboid Res. 5:397-409. tapetum in Arum italicum Miller. Protoplasma Pandey, K.K. 1967. Origin of genetic variability: combina117:116-129. tions of peroxidase isozymes determine multiple allelism of the S gene. Nature 213:669-672. Pacini, E., and Keijzer, C.J. 1989. Ontogeny of intruding nonperiplasmodial tapetum in the wild chicory, Chichorium Pandey, K.K. 1970. Time and site of the S-gene action, intybus (Compositae). Plant Syst. Evol. 167:149-164. breeding systems and relationships in incompatibility. Euphytica 19:364-372. Pacini, E., Simoncioli, C, and Cresti, M. 1975. Ultrastructure of nucellus and endosperm of Pandey, K.K. 1973. Heat sensitivity of esterase isozymes Diplotaxis erucoides during embryogenesis. Caryologia in the styles of Lilium and Nicotiana. New Phytol. 28:525-538. 72:839-850. Pacini, E., Taylor, P.E., Singh, M.B., and Knox, R.B. 1992. Pandey, K.K. 1977. Mentor pollen: possible role of wallheld pollen growth promoting substances in overDevelopment of plastids in pollen and tapetum of coming intra- and interspecific incompatibility. rye-grass, Lolium perenne L. Ann. Bot. 70:179-188. Genetica 47:219-229. Pais, M.S., and Feijo, J.A. 1987. Microbody proliferation during the microsporogenesis of Ophrys lutea Cav. Pang, P.P., Pruitt, R.E., and Meyerowitz, E.M. 1988. (Orchidaceae). Protoplasma 138:149-155. Molecular cloning, genomic organization, expression and evolution of 12S seed storage protein genes Paiva, R., and Kriz, A.L. 1994. Effect of abscisic acid on of Arabidopsis thaliana. Plant Mol. Biol. 11:805-820. embryo-specific gene expression during normal and precocious germination in normal and viviparous Panitz, R., Borisjuk, L., Manteuffel, R., and Wobus, U. 1995. maize (Zea mays) embryos. Planta 192:332-339. Transient expression of storage-protein genes during early embryogenesis of Vicia faba: synthesis and Pakendorf, K.W. 1970. Male sterility in Lupinus mutabilis metabolization of vicilin and legumin in the embryo, Sweet. Z. Pflanzenzilchtg 63:227-236. suspensor and endosperm. Planta 196:765-774. Palevitz, B.A. 1990. Kinetochore behavior during generative cell division in Tradescantia virginiana. Protoplasma Pankow, H. 1957. Uber den Pollenkitt bei Galanthus nivalis 157:120-127. L. Flora 146:240-253. Palevitz, B.A. 1993. Organization of the mitotic apparatus Parchman, L.G., and Lin, K.-C. 1972. Nucleolar RNA synduring generative cell division in Nicotiana tabacum. thesis during meiosis of lily microsporocytes. Nature Protoplasma 174:25-35. New Biol. 239:235-237. Palevitz, B.A., and Cresti, M. 1988. Microtubule organiza- Parchman, L.G., and Roth, T.F. 1971. Pachytene synaption in the sperm of Tradescantia virginiana. tonemal complexes and meiotic achiasmatic chroProtoplasma 146:28-34. mosomes. Chromosoma 33:129-145. Palevitz, B.A., and Cresti, M. 1989. Cytoskeletal changes Parchman, L.G., and Stern, H. 1969. The inhibition of protein synthesis in meiotic cells and its effect on chroduring generative cell division and sperm formation mosome behavior. Chromosoma 26:298-311. in Tradescantia virginiana. Protoplasma 150:54—71. Palevitz, B.A., and Liu, B. 1992. Microfilaments (F-actin) Parcy, F., Valon, C, Raynal, M., Gaubier-Comella, P., in generative cells and sperm: an evaluation. Sex. Delseny, M., and Giraudat, J. 1994. Regulation of gene expression programs during Arabidopsis seed Plant Reprod 5:89-100. development: roles of the ABI3 locus and of endogePalevitz, B.A., and Tiezzi, A. 1992. Organization, componous abscisic acid. Plant Cell 6:1567-1582. sition, and function of the generative cell and sperm Pareddy, D.R., and Greyson, R.I. 1985. In vitro culture of cytoskeleton. Int. Rev. Cytol. 140:149-185. immature tassels of an inbred field variety of Zea Palmer, J.D., Shields, C.R., Cohen, D.B., and Orton, T.J. mays, cv. Oh43. Plant Cell Tissue Organ Cult. 1983. An unusual mitochondrial DNA plasmid in 5:119-128. the genus Brassica. Nature 301:725-728. Palmer, R.G. 1971. Cytological studies of ameiotic and Pareddy, D.R., and Petolino, J.F. 1992. Maturation of maize pollen in vitro. Plant Cell Rep. 11:535-539. normal maize with reference to premeiotic pairing. Paris, D., Rietsema, J., Satina, S., and Blakeslee, A.F. 1953. Chromosoma 35:233-246. Effect of amino acids, especially aspartic and gluPalmer, R.G., Johns, C.W., and Muir, PS. 1980. Genetics tamic acid and their amides, on the growth of Datura and cytology of the ms3 male-sterile soybean. /. stramonium embryos in vitro. Proc. Natl Acad. Sci. Hered. 71:343-348. USA 39:1205-1212. Palser, B.F., Rouse, J.L., and Williams, E.G. 1992. A scanning electron microscope study of the pollen tube Parker, M.L. 1982. Protein accumulation in developing endosperm of a high-protein line of Triticum dicocpathway in pistils of Rhododendron. Can. ]. Bot. coides. Plant Cell Environ. 5:37-43. 70:1039-1060. Pan, D., and Nelson, O.E. 1984. A debranching enzyme Parker, M.L., and Hawes, C.R. 1982. The Golgi apparatus in developing endosperm of wheat (Triticum aesdeficiency in endosperms of the sugary-1 mutants of tivum L.). Planta 154:277-283. maize. Plant Physiol. 74:324-238. Pan, J.-L., Gao, G.-H., and Ban, H. 1983. Initial patterns of Patton, D.A., Franzmann, L.A., and Meinke, D.W. 1991. Mapping genes essential for embryo development in androgenesis in wheat anther culture. Ada Bot. Arabidopsis thaliana. Mol. Gen. Genet. 227:337-347. Sinica 25:34-39. Panchaksharappa, M.G., and Syamasundar, J. 1975. A Patton, D.A., and Meinke, D.W. 1990. Ultrastructure of arrested embryos from lethal mutants of Arabidopsis cytochemical study of ovule development in Divcadi thaliana. Am. J. Bot. 77:653-661. montanum Dalz. Cytologia 40:141-149.
References
619
Patton, D.A., Volrath, S., and Ward, E.R. 1996. Pennell, R.I., Geltz, N.R., Koren, E., and Russell, S.D. 1987. Production and partial characterization of Complementation of an Arabidopsis thaliana biotin hybridoma antibodies elicited to the sperm of auxotroph with an Escherichia coli biotin biosynthetic Plumbago zeylanica. Bot. Gaz. 148:401-406. gene. Mol. Gen. Genet. 251:261-266. Paul, A.-L., and Ferl, RJ. 1994. In vivo footprinting identifies Pennell, R.I., Janniche, L., Kjellbom, P., Scofield, G.N., Peart, J.M., and Roberts, K. 1991. Developmental an activating element of the maize Adh2 promoter speregulation of a plasma membrane arabinogalactan cific for root and vascular tissues. Plant ]. 5:523-533. protein epitope in oilseed rape flowers. Plant Cell Paul, W., Hodge, R., Smartt, S., Draper, J., and Scott, R. 3:1317-1326. 1992. The isolation and characterisation of the tapetum-specific Arabidopsis thaliana A9 gene. Plant Mol.Pennell, R.I., Janniche, L., Scofield, G.N., Booij, H., de Biol. 19:611-622. Vries, S.C., and Roberts, K. 1992. Identification of a transitional cell state in the developmental pathway PavingerovS, D., Dostal, J., Biskov<S, R., and Benetka, V. to carrot somatic embryogenesis. /. Cell Biol. 1994. Somatic embryogenesis and Agrobacterium119:1371-1380. mediated transformation of chrysanthemum. Plant Pennell, R.I., Knox, J.P., Scofield, G.N., Selvendran, R.R., Sci. 97:95-101. and Roberts, K. 1989. A family of abundant plasma Pavlov, A.N., Konarev, V.G., Kolesnik, T.I., and membrane-associated glycoproteins related to the Shayakhmetov, I.F. 1975. Gliadins of the caryopsis of arabinogalactan proteins is unique to flowering wheat during the process of its development. Soviet plants. /. Cell Biol. 108:1967-1977. Plant Physiol. 22:63-67. Payne, G., Kono, Y, and Daly, J.M. 1980. A comparison of Pental, D., Mukhopadhyay, A., Grover, A., and Pradhan, A.K. 1988. A selection method for the synthesis of purified host specific toxin from Helminthosporium triploid hybrids by fusion of microspore protoplasts maydis, race T, and its acetate derivative on oxidation (n) with somatic cell protoplasts (2n). Theor. Appl. by mitochondria from susceptible and resistant Genet. 76:237-243. plants. Plant Physiol. 65:785-791. Pearson, O.H. 1981. Nature and mechanisms of cytoplas- Penza, R., Lurquin, P.F., and Filippone, E. 1991. Gene mic male sterility in plants. HortScience 16:482-487. transfer by cocultivation of mature embryos with Agrobacterium tumefaciens: application to cowpea Pechan, P.M. 1989. Successful cocultivation of Brassica napus microspores and proembryos with Agro(Vigna unguiculata Walp). /. Plant Physiol. 138:39-13. bacterium. Plant Cell Rep. 8:387-390. Perata, P., Picciarelli, P., and Alpi, A. 1990. Pattern of variPechan, P.M., and Keller, W.A. 1988. Identification of ations in abscisic acid content in suspensors, potentially embryogenic microspores in Brassica embryos, and integuments of developing Phaseolus coccineus seeds. Plant Physiol. 94:1776-1780. napus. Physiol. Plant. 74:377-384. Pechan, P.M., Bartels, D., Brown, D.C.W., and Schell, J. Perdue, T.D., Loukides, C.A., and Bedinger, PA. 1992. The 1991. Messenger-RNA and protein changes associformation of cytoplasmic channels between tapetal ated with induction of Brassica microspore embryocells in Zea mays. Protoplasma 171:75-79. genesis. Planta 184:161-165. Perdue, T.D., and Parthasarathy, M.V. 1985. In situ localization of F-actin in pollen tubes. Enr. J. Cell Biol. Peddada, L.B., and Mascarenhas, J.P. 1975. The synthesis 39:13-20. of 5S ribosomal RNA during pollen development. Devel. Growth Diffn 17:1-8. Pereira, A.S.R., and Linskens, H.F. 1963. The influence of Pedersen, K., Argos, P., Naravana, S.V.L., and Larkins, glutathione and glutathione antagonists on meiosis B.A. 1986. Sequence analysis and characterization of in excised anthers of Lilium henryi. Ada Bot. Neerl. a maize gene encoding a high-sulfur zein protein of 12:302-314. Mr 15,000. /. Biol. Chem. 261:6279-6284. Perez, M., Ishioka, G.Y., Walker, L.E., and Chesnut, R.W. 1990. cDNA cloning and immunological characteriPedersen, K., Bloom, K.S., Anderson, J.N., Glover, D.V., zation of the rye grass allergen Lol p I. /. Biol. Client. and Larkins, B.A. 1980. Analysis of the complexity 265:16210-16215. and frequency of zein genes in the maize genome. Biochemistry 19:1644-1650. Perez-Grau, L., and Goldberg, R.B. 1989. Soybean seed Pedersen, K., Devereux, J., Wilson, D.R., Sheldon, E., and protein genes are regulated spatially during Larkins, B.A. 1982. Cloning and sequence analysis embryogenesis. Plant Cell 1:1095-1109. reveal structural variation among related zein genes P6rez-Grau, L., Cortadas, J., Puigdomenech, P., and Palau, in maize. Cell 29:1015-1026. J. 1986. Accumulation and subcellular localization of Pedersen, S., and Andersen, S.B. 1993. Developmental glutelin-2 transcripts during maturation of maize expression of isozymes during embryogenesis in endosperm. FEBS Lett. 202:145-148. Perez-Munoz, C.A., Jernstedt, J.A., and Webster, B.D. barley anther culture. Plant Sci. 91:75-86. Pedersen, S., Simonsen, V., and Loeschcke, V. 1987. 1993a. Pollen wall development in Vigna vexillata. Overlap of gametophytic and sporophytic gene I. Characterization of wall layers. Am. ]. Bot. expression in barley. Theor. Appl. Genet. 75:200-206. 80:1183-1192. Peirson, B.N., Owen, H.A., Feldmann, K.A., and Perez-Munoz, C.A., Jernstedt, J.A., and Webster, B.D. Makaroff, C.A. 1996. Characterization of three male1993b. Pollen wall development in Vigna vexillata. II. sterile mutants of Arabidopsis thaliana exhibiting Ultrastructural studies. Am. j. Bot. 80:1193-1202. alterations in meiosis. Sex. Plant Reprod. 9:1-10. Periasamy, K., and Amalathas, J. 1991. Absence of callose Pelletier, G., and Ilami, M. 1972. Les facteurs de l'anand tetrad in the microsporogenesis of Pandanus drog^nese in vitro chez Nicotiana tabacum. Z. odoratissimus with well-formed pollen exine. Ann. Pflanzenphysiol. 68:97-114. Bot. 67:29-33. Pennell, R.I., and Roberts, K. 1990. Sexual development Pernollet, J.-C, and Vaillant, V. 1984. Characterization and complexity of wheat developing endosperm in the pea is presaged by altered expression of aramRNAs. Plant Physiol. 76:187-190. binogalactan protein. Nature 344:547-549.
620
References
fusion and tip extension in pollen tubes. Planta Pescitelli, S.M., and Sukhapinda, K. 1995. Stable transfor163:20-26. mation via electroporation into maize type II callus and regeneration of fertile transgenic plants. Plant Pierson, E.S. 1988. Rhodamine-phalloidin staining of FCell Rep. 14:712-716. actin in pollen after dimethylsulphoxide permeabilization. Sex. Plant Reprod. 1:83-87. Peterson, D.M., Saigo, R.H., and Holy, J. 1985. Development of oat aleurone cells and their protein Pierson, E.S., Derksen, J., and Traas, J.A. 1986. bodies. Cereal Cliem. 62:366-371. Organization of microfilaments and microtubules in pollen tubes grown in vitro or in vivo in various Peterson, P.A., Flavell, R.B., and Barratt, D.H.P. 1975. angiosperms. Eur. J. Cell Biol. 41:14-18. Altered mitochondrial membrane activities associated with cytoplasmically-inherited disease sensitiv- Pierson, E.S., Kengen, H.M.P., and Derksen, J. 1989. ity in maize. Theor. Appl. Genet. 45:309-314. Microtubules and actin filaments co-localize in pollen tubes of Nicotiana tabacum L. and Lilium longiPetolino, J.F., and Jones, A.M. 1986. Anther culture of elite florum Thunb. Protoplasma 150:75-77. genotypes of maize. Crop Sci. 26:1072-1074. Petolino, J.F., and Thompson, S.A. 1987. Genetic analysis Pierson, E.S., Lichtscheidl, I.K., and Derksen, J. 1990. Structure and behaviour of organelles in living of anther culture response in maize. Theor. Appl. pollen tubes of Lilium longiflorum. f. Expt. Bot. Genet. 74:284-286. 41:1461-1468. Pettitt, J.M. 1976. Pollen wall and stigma surface in the marine angiosperms Thalassia and Thalassodendron.Pierson, E.S., Li, Y.Q., Zhang, H.Q., Willemse, M.T.M., Linskens, H.F., and Cresti, M. 1995. Pulsatory Micron 7:21-32. growth of pollen tubes: investigation of a possible Pettitt, J.M. 1980. Reproduction in seagrasses: nature of relationship with the periodic distribution of cell the pollen and receptive surface of the stigma in the wall components. Ada Bot. Need. 44:121-128. Hydrocharitaceae. Ann. Bot. 45:257-271. Pettitt, J.M. 1981. Reproduction in seagrasses: pollen Pierson, E.S., Miller, D.D., Callaham, D.A., Shipley, A.M., development in Thalassia hemprichii, Halophila stipu- Rivers, B.A., Cresti, M., and Hepler, P.K. 1994. Pollen tube growth is coupled to the extracellular calcium lacea and Thalassodendron ciliatum. Ann. Bot. ion flux and the intracellular calcium gradient: effect 48:609-622. of BAPTA-type buffers and hypertonic media. Plant Phatak, S.C., Wittwer, S.H., Honma, S., and Bukovac, M.J. Cell 6:1815-1828. 1966. Gibberellin-induced anther and pollen development in a stamen-less tomato mutant. Nature Pierson, E.S., Willekens, P.G.M., Maessen, M., and 209:635-636. Helsper, J.P.F.G. 1986. The effect of lectins on germinating pollen of Lilium longiflorum. I. Effect on Piaggesi, A., Picciarelli, P., Lorenzi, R., and Alpi, A. 1989. pollen germination, pollen tube growth and organiGibberellins in embryo-suspensor of Phaseolus coczation of microfilaments. Acta Bot. Neerl. 35:249-256. cineus seeds at the heart stage of embryo developPipkin, J.L., Jr., and Larson, D.A. 1972. Characterization of ment. Plant Physiol. 91:362-366. the Very' lysine-rich histones of active and quiesPicciarelli, P., and Alpi, A. 1986. Gibberellins in suspencent anther tissues of Hippeastrum belladonna. Expt. sors of Phaseolus coccinens L. seeds. Plant Physiol. Cell Res. 71:249-260. 82:298-300. Picciarelli, P., and Alpi, A. 1987. Embryo-suspensor of Pipkin, J.L., Jr., and Larson, D.A. 1973. Changing patterns of nucleic acids, basic and acidic proteins in generaTropaeolum majus: identification of gibberellin A63. tive and vegetative nuclei during pollen germinaPhytochemistry 26:329-330. tion and pollen tube growth in Hippeastrum Picciarelli, P., Piaggesi, A., and Alpi, A. 1991. Gibberellins belladonna. Expt. Cell Res. 79:28-42. in suspensor, embryo and endosperm of developing seeds of Cytisus laburnum. Phytochemistry Pirrie, A., and Power, J.B. 1986. The production of fertile, triploid somatic hybrid plants (Nicotiana glutinosa (n) 30:1789-1792. + N. tabacum (2n)) via gameticsomatic protoplast Picciarelli, P., Alpi, A., Pistelli, L., and Scalet, M. 1984. fusion. Theor. Appl. Genet. 72:48-52. Gibberellin-like activity in suspensors of Tropaeolum majus L. and Cytisus laburnum L. Planta 162:566-568. Pitto, L., Lo Schiavo, F., and Terzi, M. 1985. a-amanitin resistance is developmentally regulated in carrot. Picton, J.M., and Steer, M.W. 1981. Determination of secreProc. Natl Acad. Sci. USA 82:2799-2803. tory vesicle production rates by dictyosomes in pollen tubes of Tradescantia using cytochalasin D. /. Pitto, L., Lo Schiavo, F., Giuliano, G., and Terzi, M. 1983. Analysis of the heat-shock protein pattern during Cell Sci. 49:261-272. somatic embryogenesis of carrot. Plant Mol. Biol. Picton, J.M., and Steer, M.W. 1982. A model for the mech2:231-237. anism of tip extension in pollen tubes. /. Theor. Biol. Pla, M., Goday, A., Vilardell, J., G6mez, J., and Pages, M. 98:15-20. 1989. Differential regulation of ABA-induced 23-25 Picton, J.M., and Steer, M.W. 1983a. Evidence for the role kDa proteins in embryo and vegetative tissues of the of Ca2+ ions in tip extension in pollen tubes. viviparous mutants of maize. Plant Mol. Biol. Protoplasma 115:11-17. 13:385-394. Picton, J.M., and Steer, M.W. 1983b. The effect of cycloheximide on dictyosome activity in Tradescantia Pla, M., Gomez, J., Goday, A., and Pages, M. 1991. Regulation of the abscisic acid-responsive gene rablS pollen tubes determined using cytochalasin D. Eur. J. in maize viviparous mutants. Mol. Gen. Genet. Cell Biol. 29:133-138. 230:394-400. Picton, J.M., and Steer, M.W. 1983c. Membrane recycling and the control of secretory activity in pollen tubes. Pla, M., Vilardell, J., Guiltinan, M.J., Marcotte, W.R., Niogret, M.F., Quatrano, R.S., and Pages, M. 1993. /. Cell Sci. 63:303-310. The Cfs-regulatory element CCACGTGG is involved Picton, J.M., and Steer, M.W. 1985. The effects of ruthein ABA and water-stress responses of the maize gene nium red, lanthanum, fluorescein isothiocyanate rab28. Plant Mol. Biol. 21:259-266. and trifluoperazine on vesicle transport, vesicle
References
621
induced male and female sterility in canola (Brassica Plant, A.L., van Rooijen, G.J.H., Anderson, C.P., and napus L.). Ann. Bot. 62:83-86. Moloney, M.M. 1994. Regulation of an Arabidopsis oleosin gene promoter in transgenic Brassica napus. Polowick, PL., and Sawhney, V.K. 1990. Microsporogenesis Plant Mol. Biol. 25:193-205. in a normal line and in the ogu cytoplasmic male-sterile line of Brassica napus. I. The influence of high temPlatt-Aloia, K.A., Lord, E.M., DeMason, D.A., and peratures. Sex. Plant Reprod. 3:263-276. Thomson, W.W. 1986. Freeze-fracture observations on membranes of dry and hydrated pollen from Polowick, P.L., and Sawhney, V.K. 1991. Microsporogenesis in a normal line and in the ogu cytoplasmic Collomia, Phoenix and Zea. Planta 168:291-298. male-sterile line of Brassica napus. II. The influence of Plaut, W.S. 1953. DNA synthesis in the microsporocytes of intermediate and low temperatures. Sex. Plant Lilium henryi. Hereditas 39:438-444. Reprod. 4:22-27. Pnueli, L., Abu-Abeid, M., Zamir, D., Nacken, W., SchwarzSommer, Z., and Lifschitz, E. 1991. The MADS box Polowick, P.L., and Sawhney, V.K. 1992. Ultrastructural changes in the cell wall, nucleus and cytoplasm of gene family in tomato: temporal expression during pollen mother cells during meiotic prophase I in floral development, conserved secondary structures Lycopersicon esculentum (Mill.). Protoplasma and homology with homeotic genes from Antirrhinum 169:139-147. and Arabidopsis. Plant J. 1:255-266. Pnueli, L., Hareven, D., Broday, L., Hurwitz, C, and Polowick, PL., and Sawhney, V.K. 1993a. Differentiation of the tapetum during microsporogenesis in tomato Lifschitz, E. 1994a. The TM5 MADS box gene medi(Lycopersicon esculentum Mill.), with special reference ates organ differentiation in the three inner whorls of to the tapetal cell wall. Ann. Bot. 72:595-605. tomato flowers. Plant Cell 6:175-186. Pnueli, L., Hareven, D., Rounsley, S.D., Yanofsky, M.F., Polowick, PL., and Sawhney, V.K. 1993b. An ultrastrucand Lifschitz, E. 1994b. Isolation of the tomato AGAtural study of pollen development in tomato MOUS gene TAG1 and analysis of its homeotic role (Lycopersicon esculentum). I. Tetrad to early binucleate in transgenic plants. Plant Cell 6:163-173. microspore stage. Can. J. Bot. 71:1039-1047. Poddubnaya-Arnoldi, V.A. 1967. Comparative embryol- Polowick, PL., and Sawhney, V.K. 1993c. An ultrastrucogy of the Orchidaceae. Phytomorphology 17:312-320. tural study of pollen development in tomato (Lycopersicon esculentum). II. Pollen maturation. Can. Poethig, R.S., Coe, E.H., Jr., and Johri, M.M. 1986. Cell linJ. Bot. 71:1048-1055. eage patterns in maize embryogenesis: a clonal analysis. Devel Biol. 117:392-404. Polowick, P.L., and Sawhney, V.K. 1995. Ultrastructure of Polito, V.S. 1983. Membrane-associated calcium during the tapetal cell wall in the stamenless-2 mutant of pollen grain germination: a microfluorometric tomato (Lycopersicon esculentum): correlation analysis. Protoplasma 117:226-232. between structure and male sterility. Protoplasma 189:249-255. Pollak, P.E. 1992. Cytological differences between a cytoplasmic male sterile tobacco cybrid and its fertile Polsoni, L., Kott, L.S., and Beversdorf, W.D. 1988. Largecounterpart during early anther development. Am. ]. scale microspore culture technique for mutationBot. 79:937-945. selection studies in Brassica napus. Can. f. Bot. 66:1681-1685. Pollak, P.E., Hansen, K., Astwood, J.D., and Taylor, L.P. 1995. Conditional male fertility in maize. Sex. Plant Polya, G.M., Micucci, V, Rae, A.L., Harris, P.J., and Reprod. 8:231-241. Clarke, A.E. 1986. Ca2+-dependent protein phosphoPollak, P.E., Vogt, T, Mo, Y., and Taylor, L.P. 1993. rylation in germinated pollen of Nicotiana alata, an Chalcone synthase and flavonol accumulation in ornamental tobacco. Physiol. Plant. 67:151-157. stigmas and anthers of Petunia hybrida. Plant Physiol.Pomeroy, M.K., Kramer, J.K.G., Hunt, D.J., and Keller, 102:925-932. W.A. 1991. Fatty acid changes during development of zygotic and microspore-derived embryos of Pollock, E.G., and Jensen, W.A. 1964. Cell development Brassica napus. Physiol. Plant. 81:447-454. during early embryogenesis in Capsella and Gossypium. Am. /. Bot. 51:915-921. Ponte, I., Guillen, P., Debon, R.M., Reina, M., Aragay, A., Pollock, E.G., and Jensen, W.A. 1967. Ontogeny and cytoEspel, E., di Fonzo, N., and Palau, J. 1994. Narrow chemistry of the chalazal proliferating cells of A/T-rich zones present at the distal 5-flanking Capsella bursa-pastoris (L.) Medic. New Phytol. sequences of the zein genes Zcl and Zc2 bind a 66:413-417. unique 30 kDa HMG-like protein. Plant Mol. Biol. 26:1893-1906. Polowick, PL., and Greyson, R.I. 1982. Anther development, meiosis and pollen formation in Zea tassels Ponz, F., Paz-Ares, J., Hernandez-Lucas, C, Carbonero, P., cultured in defined liquid medium. Plant Sci. Lett. and Garci'a-Olmedo, F. 1983. Synthesis and process26:139-145. ing of thionin precursors in developing endosperm Polowick, PL., and Greyson, R.I. 1984. The relative effifrom barley (Hordeum vulgare L.). EMBO /. ciency of cytokinins in the development of normal 2:1035-1040. spikelets on cultured tassels of Zea mays. Can. J. Bot.Ponzi, R., and Pizzolongo, P. 1973. Ultrastructure of plas62:830-834. tids in the suspensor cells of Ipomoea purpurea Roth. Polowick, PL., and Sawhney, V.K. 1985. Temperature /. Submicros. Cytol. 5:257-263. effects on male fertility and flower and fruit devel- Ponzi, R., and Pizzolongo, P. 1976. Cytinus hypocistis L. opment in Capsicum annuum L. Sci. Hort. embryogenesis: ultrastructural aspects of megas25:117-127. porogenesis and megagametogenesis. /. Submicros. Polowick, PL., and Sawhney, V.K. 1987. A scanning elecCytol. 8:327-336. tron microscopic study on the influence of tempera- Ponzi, R., and Pizzolongo, P. 1984. Ultrastructural study ture on the expression of cytoplasmic male sterility of change from free-nuclear to cellular endosperm in in Brassica napus. Can. ]. Bot. 65:807-814. Ipomoea purpurea Roth, and Cytinus hypocistis L. Polowick, P.L., and Sawhney, V.K. 1988. High temperature Giorn. Bot. Hal. 118:147-154.
622 Porath, D., and Galun, E. 1967. In vitro culture of hermaphrodite floral buds of Cucumis melo L.: Microsporogenesis and ovary formation. Ann. Bot. 31:283-290. Porter, E.K. 1981. Origins and genetic nonvariability of the proteins which diffuse from maize pollen. Environ. Health Pers. 37:53-59. Porter, E.K., Bird, J.M., and Dickinson, H.G. 1982. Nucleic acid synthesis in microsporocytes of Lilium cv. Cinnabar: events in the nucleus. /. Cell Sci. 57:229-246. Porter, E.K., Parry, D., and Dickinson, H.G. 1983. Changes in poly(A)+RNA during male meiosis in Lilium. J. Cell Sci. 62:177-186. Porter, E.K., Parry, D., Bird, J., and Dickinson, H.G. 1984. Nucleic acid metabolism in the nucleus and cytoplasm of angiosperm meiocytes. Symp. Soc. Expt. Biol. 38:363-379. Portnoi, L., and Horovitz, A. 1977. Sugars in natural and artificial pollen germination substrates. Ann. Bot. 41:21-27. Potrykus, 1.1990. Gene transfer to plants: assessment and perspectives. Physiol. Plant. 79:125-134. Potz, H., and Tatlioglu, T. 1993. Molecular analysis of cytoplasmic male sterility in chives (Allium schoenoprasum L.). Theor. Appl. Genet. 87:439-445. Poulson, R., and Beevers, L. 1973. RNA metabolism during the development of cotyledons of Pisum sativum L. Biochim. Biophys. Ada 308:381-389. Povvling, A. 1981. Species of small DNA molecules found in mitochondria from sugarbeet with normal and male sterile cytoplasms. Mol. Gen. Genet. 183:82-84. Powling, A. 1982. Restriction endonuclease analysis of mitochondrial DNA from sugarbeet with normal and male-sterile cytoplasms. Heredity 49:117-120. Powling, A., and Ellis, T.H.N. 1983. Studies on the organelle genomes of sugarbeet with male-fertile and male-sterile cytoplasms. Theor. Appl. Genet. 65:323-328. Prabhakar, K., and Vijayaraghavan, M.R. 1983. Histochemistry and ulrrastructure of suspensor cells in Alyssum maritimum. Cytologia 48:389-402. Prahl, A.-K., Springstubbe, H., Grumbach, K., and Wiermann, R. 1985. Studies on sporopollenin biosynthesis: the effect of inhibitors of carotenoid biosynthesis on sporopollenin accumulation. Z. Natnrfars. 40c:621-626. Pramanik, S.K., and Bewley, J.D. 1993. A ubiquitin carrier protein cDNA from developing alfalfa embryos. Plant Physiol. 102:1049-1050. Pramanik, S.K., Krochko, J.E., and Bewley, J.D. 1992. Distribution of cytosolic mRNAs between polysomal and ribonucleoprotein complex fractions in alfalfa embryos. Plant Physiol. 99:1590-1596. Prandl, R., Kloske, E., and Schoffl, F. 1995. Developmental regulation and tissue-specific differences of heat shock gene expression in transgenic tobacco and Arabidopsis plants. Plant Mol. Biol. 28:73-82. Prasad, K. 1977. Histochemistry of anther and ovule in Farsetia hamiltonii (Royle) and Ernca sativa Mill (Cruciferae). /. Indian Bot. Soc. 56:90-99. Prat, S., Perez-Grau, L., and Puigdomenech, P. 1987. Multiple variability in the sequence of a family of maize endosperm proteins. Gene 52:41-49. Prat, S., Cortadas, J., Puigdomenech, P., and Palau, J. 1985. Nucleic acid (cDNA) and amino acid sequences of the maize endosperm protein glutelin-2. Nucl. Acids Res. 13:1493-1504.
References
Preiss, J., Lammel, C, and Sabraw, A. 1971. A unique adenosine diphosphoglucose pyrophosphorylase associated with maize embryo tissue. Plant Physiol. 47:104-108. Pressey, R. 1991. Polygalacturonase in tree pollens. Phytochemistry 30:1753-1755. Pressey, R., and Reger, B.J. 1989. Polygalacturonase in pollen from corn and other grasses. Plant Sci. v 59:57-62. Pretov3, A. 1974. The influence of the osmotic potential of the cultivation medium on the development of excised flax embryos. Biol. Plant. 16:14-20. Pretov3, A. 1986. Growth of zygotic flax embryos in vitro and influence of kinetin. Plant Cell Rep. 3:210-211. Preuss, D., Rhee, S.Y., and Davis, R.W. 1994. Tetrad analysis possible in Arabidopsis with mutation of the QUARTET (QRT) genes. Science 264:1458-1460. Preuss, D., Lemieux, B., Yen, G., and Davis, R.W. 1993. A conditional sterile mutation eliminates surface components from Arabidopsis pollen and disrupts cell signaling during fertilization. Genes Devel. 7:974-985. Prevost, I., and Le Page-Degivry, M.T. 1985. Changes in abscisic acid content in axis and cotyledons of developing Phaseolus vulgaris embryos and their physiological consequences. /. Expt. Bot. 36:1900-1905. Price, S.D., and Barrett, S.C.H. 1982. Tristyly in Pontederia cordata (Pontederiaceae). Can. J. Bot. 60:897-905. Priestley, D.A., and de Kruijff, B. 1982. Phospholipid motional characteristics in a dry biological system. A 31 P-nuclear magnetic resonance study of hydrating Typha latifolia pollen. Plant Physiol. 70:1075-1078. Pring, D.R., Conde, M.R, and Levings, C.S., III. 1980. DNA heterogeneity within the C group of maize male-sterile cytoplasms. Crop Sci. 20:159-162. Pring, D.R., Conde, M.R, and Schertz, K.R 1982. Organelle genome diversity in Sorghum: male-sterile cytoplasms. Crop Sci. 22:414-421. Pring, D.R., and Levings, C.S., III. 1978. Heterogeneity of maize cytoplasmic genomes among male-sterile cytoplasms. Genetics 89:121-136. Pring, D.R., Conde, M.R, Schertz, K.R, and Levings, C.S., III. 1982. Plasmid-like DNAs associated with mitochondria of cytoplasmic male-sterile Sorghum. Mol. Gen. Genet. 186:180-184. Pring, D.R., Levings, C.S., III., Hu, W.W.L., and Timothy, D.H. 1977. Unique DNA associated with mitochondria in the "S"-type cytoplasm of male-sterile maize. Proc. Natl Acad. Sci. USA 74:2904-2908. Pritchard, A.J., and Hutton, E.M. 1972. Anther and pollen development in male-sterile Phaseolus atropurpureus. J. Hered. 63:280-282. Pritchard, H.N. 1964a. A cytochemical study of embryo sac development in Stellaria media. Am. J. Bot. 51:371-378. Pritchard, H.N. 1964b. A cytochemical study of embryo development in Stellaria media. Am. ]. Bot. 51:472-479. Pruitt, K.D., and Hanson, M.R. 1989. Cytochrome oxidase subunit II sequences in Petunia mitochondria: two intron-containing genes and an intron-less pseudogene associated with cytoplasmic male sterility. Curr. Genet. 16:281-291. Pruitt, K.D., and Hanson, M.R. 1991. Transcription of the Petunia mitochondrial CMS-associated Pcf locus in male sterile and fertility-restored lines. Mol. Gen. Genet. 227:348-355. Przybyllok, T., and Nagl, W. 1977. Auxin concentration in the embryo and suspensors of Tropaeolum majus, as
References
623
determined by mass fragmentation (single ion deteclated expression of a reintroduced napin gene. Theor. tion). Z. Pflanzenphysiol. 84:463-465. Appl. Genet. 75:685-694. Punnett, H.H. 1953. Cytological evidence of hexaploid Radlowski, M., Kalinowski, A., Krblikowski, Z., and Bartkowiak, S. 1994. Protease activity from maize cells in maize endosperm. /. Hered. 44:257-259. pollen. Phytochemistry 35:853-856. Puupponen-Pimia, R., Saloheimo, M., Vasara, T., Ra, R., Gaugecz,J., KurteYi, U., Knowles, J.K.C., Keranen, S., Radojevic, L., Zylberberg, L., and Kovoor, J. 1980. Etude ultrastructurale des embryons androgenetiques and Kauppinen, V. 1993. Characterization of a birch d'Aesculus hippocastannm L. Z. Pflanzenphysiol. (Betula pendula Roth.) embryogenic gene, BP8. Plant Mol. Biol. 23:423-428. 98:255-261. Pwee, K.-Hv and Gray, J.G. 1993. The pea plastocyanin Rae, A.L., Harris, P.J., Bade, A., and Clarke, A.E. 1985. promoter directs cell-specific but not full light-reguComposition of the cell walls of Nicotiana alata Link et Otto pollen tubes. Planta 166:128-133. lated expression in transgenic tobacco plants. Plant ]. 3:437-449. Raeber, J.G., and Bolton, A. 1955. A new form of male Pysh, L.D., Aukerman, M.J., and Schmidt, R.J. 1993. sterility in Nicotiana tabacum L. Nature 176:314-315. OHP1: a maize basic domain/ leucine zipper protein Rafalski, J.A. 1986. Structure of wheat gamma-gliadin genes. Gene 43:221-229. that interacts with opaque2. Plant Cell 5:227-236. Rafalski, J.A., Scheets, K., Metzler, M., Peterson, D.M., Hedgcoth, C, and Soil, D.G. 1984. Developmentally Qiu, X., and Erickson, L. 1996. A pollen-specific polyregulated plant genes: the nucleotide sequence of a galacturonase-like cDNA from alfalfa. Sex. Plant wheat gliadin genomic clone. EMBO ]. 3:1409-1415. Reprod. 9:123-124. Qu, R., and Huang, A.H.C. 1990. Oleosin KD 18 on the Raff, J.W., and Clarke, A.E. 1981. Tissue-specific antigens secreted by suspension-cultured callus cells of surface of oil bodies in maize. Genomic and cDNA Primus avium L. Planta 153:115-124. sequences and the deduced protein structure. /. Biol. Chem. 265:2238-2243. Raff, J.W., Knox, R.B., and Clarke, A.E. 1981. Style antiQu, R., Vance, V.B., and Huang, A.H.C. 1990. Expression gens of Primus avium L. Planta 153:125-129. of genes encoding oleosin isoforms in the embryos Raff, J.W., Pettitt, J.M., and Knox, R.B. 1981. of maturing maize kernels. Plant Sci. 72:223-232. Cytochemistry of pollen tube growth in stigma and style of Primus avium. Phytomorphology 31:214-231. Qu, R., Wang, S.-M., Lin, Y.-H., Vance, V.B., and Huang, A.H.C. 1986. Characteristics and biosynthesis of Rafnar, T., Griffith, I.J., Kuo, M.-C, Bond, J.F., Rogers, B.L., membrane proteins of lipid bodies in the scutella of and Klapper, D.G. 1991. Cloning of Amb a I (antigen maize (Zea mays L.). Biochem. ]. 235:57-65. E), the major allergen family of short ragweed Qu, R.-D., and Chen, Y. 1984. Pathways of androgenesis pollen. /. Biol. Chem. 266:1229-1236. and observations on cultured pollen grains in rice Raghavan, P., and Philip, V.]. 1982. Morphological and (Oryza sativa subsp. Keng). Ada Bot. Sinica histochemical changes in the egg and zygote of 26:580-587. Lagerstroemia speciosa. I. Cell size, vacuole and insoluble polysaccharides. Proc. Indian Acad. Sci. (Plant Quattrocchio, R, Tolk, M.A., Coraggio, I., Mol. J.N.M., Sci.) 91:465^172. Viotti, A., and Koes, R.E. 1990. The maize zein gene zE19 contains two distinct promoters which are Raghavan, V. 1976a. Role of the generative cell in androindependently activated in endosperm and anthers genesis in henbane. Science 191:388-389. of transgenic Petunia plants. Plant Mol. Biol. Raghavan, V. 1976b. Experimental Embryogenesis in 15:81-93. Vascular Plants. London: Academic Press. Quayle, T.J.A., Hetz, W., and Feix, G. 1991. Raghavan, V. 1977a. Patterns of DNA synthesis during Characterization of a maize endosperm culture pollen embryogenesis in henbane. /. Cell Biol. expressing zein genes and its use in transient trans73:521-526. formation assays. Plant Cell Rep. 9:544-548. Raghavan, V. 1977b. Applied aspects of embryo culture. Que, Q.-D., and Tang, X.-H. 1988. The changes of protein In Applied and Fundamental Aspects of Plant Cell, content and synthetic activity in squash pollen durTissue, and Organ Culture, ed. J. Reinert and Y.P.S. ing germination. Ada Bot. Sinica 30:501-507. Bajaj, pp. 375-397. Berlin: Springer-Verlag. Raghavan, V. 1978. Origin and development of pollen Rabakoarihanta, A., Mok, D.W.S., and Mok, M.C. 1979. embryoids and pollen calluses in cultured anther Fertilization and early embryo development in recisegments of Hyoscyamus niger (henbane). Am. ]. Bot. procal interspecific crosses of Phaseolus. Theor. Appl. 65:984-1002. Genet. 54:55-59. Raghavan, V. 1979a. Embryogenic determination and Racchi, M.L., Gavazzi, G., Monti, D., and Manitto, P. 1978. ribonucleic acid synthesis in pollen grains of An analysis of the nutritional requirements of the pro Hyoscyamus niger (henbane). Am. J. Bot. 66:36-39. mutant in Zea mays. Plant Sci. Lett. 13:357-364. Raghavan, V. 1979b. An autoradiographic study of RNA Racusen, R.H., and Schiavone, F.M. 1988. Detection of synthesis during pollen embryogenesis in spatially- and stage-specific proteins in extracts from Hyoscyamus niger (henbane). Am. ]. Bot. 66:784-795. single embryos of the domesticated carrot. Raghavan, V. 1981a. A transient accumulation of poly(A)Development 103:665-674. containing RNA in the tapetum of Hyoscyamus niger Radding, CM. 1982. Homologous pairing and strand during microsporogenesis. Devel Biol. 81:342-348. exchange in genetic recombination. Annu. Rev. Genet. Raghavan, V. 1981b. Distribution of poly(A)-containing 16:405-437. RNA during normal pollen development and during Radke, S.E., Andrews, B.M., Moloney, M.M., Crouch, induced pollen embryogenesis in Hyoscyamus niger. M.L., Kridl, J.C, and Knauf, V.C. 1988. }. Cell Biol. 89:593-606. Transformation of Brassica napus L. using Raghavan, V. 1984. Protein synthetic activity during norAgrobaderium tumefaciens: developmentally regumal pollen development and during induced pollen
624
References
embryogenesis in Hyoscyamus niger. Can. J. Bot. Raghavan, V., and Sharma, K.K. 1995. Zygotic embryoge62:2493-2513. nesis in gymnosperms and angiosperms. In In Vitro Raghavan, V. 1985. The applications of embryo rescue in Embryogenesis in Plants, ed. T.A. Thorpe, pp. 73-115. agriculture. In Biotechnology in International Dordrecht: Kluwer Academic Press. Agricultural Research, pp. 189-197. Manila: Raghavan, V., and Torrey, J.G. 1963. Growth and morphoInternational Rice Research Institute. genesis of globular and older embryos of Capsella in culture. Am. }. Bot. 50:540-551. Raghavan, V. 1986a. Pollen developmental biology in cultured anthers. In Cell Culture and Somatic Cell GeneticsRaghavan, V, and Torrey, J.G. 1964. Effects of certain of Plants, Vol. 3, ed. I.K. Vasil, pp. 275-304. Orlando: growth substances on the growth and morphogeneAcademic Press. sis of immature embryos of Capsella in culture. Plant Raghavan, V. 1986b. Variability through wide crosses and Physiol. 39:691-699. embryo rescue. In Cell Culture and Somatic Cell Rahman, S., Shewry, P.R., and Miflin, B.J. 1982. Genetics of Plants, Vol. 3, ed. I.K. Vasil, pp. 612-633. Differential protein accumulation during barley Orlando: Academic Press. grain development. /. Expt. Bot. 33:717-728. Raghavan, V. 1986c. Embryogenesis in Angiosperms. A Rahman, S., Kreis, M., Forde, B.G., Shewry, PR., and Developmental and Experimental Study. Cambridge: Miflin, B.J. 1984. Hordein-gene expression during Cambridge University Press. development of the barley (Hordeum vulgare) endosperm. Biochem. J. 223:315-322. Raghavan, V. 1988. Anther and pollen development in rice (Oryza sativa). Am. J. Bot. 75:183-196. Raikhel, N.V., Bednarek, S.Y., and Wilkins, T.A. 1988. Celltype-specific expression of a wheat-germ agglutinin Raghavan, V. 1989. mRNAs and a cloned histone gene are gene in embryos and young seedlings of Triticum aesdifferentially expressed during anther and pollen tivum. Planta 176:406-414. development in rice (Oryza sativa L.). /. Cell Sci. 92:217-229. Raikhel, N.V., and Lerner, D.R. 1991. Expression and regulation of lectin genes in cereals and rice. Devel Raghavan, V. 1990a. Origin of the quiescent center in the Genet. 12:255-260. root of Capsella bursa-pastoris (L.) Medik. Planta 181:62-70. Raikhel, N.V., and Quatrano, R.S. 1986. Localization of Raghavan, V. 1990b. Spatial distribution of mRNA during wheat-germ agglutinin in developing wheat embryos and those cultured in abscisic acid. Planta pre-fertilization ovule development in Capsella bursa-pastoris. Sex. Plant Reprod. 3:170-178. 168:433-440. Raghavan, V. 1990c. Hyoscyamus spp.: anther culture stud- Raikhel, N.V., and Wilkins, T.A. 1987. Isolation and ies. In Biotechnology in Agriculture and Forestry 12. characterization of a cDNA clone encoding wheat Haploids in Crop Improvement I, ed. Y.P.S. Bajaj, pp. germ agglutinin. Proc. Natl Acad. Sci. USA 290-305. Berlin: Springer-Verlag. 84:6745-6749. Raghavan, V. 1991. Some perspectives on the develop- Raina, A., and Datta, A. 1992. Molecular cloning of a gene ment of embryos in angiosperms. A synthesis of old encoding a seed-specific protein with nutritionally ideas and new concepts. Rev. Cytol. Biol. Veget.-Bot. balanced amino acid composition from Amaranthus. 14:163-174. Proc. Natl Acad. Sci. USA 89:11774-11778. Raghavan, V. 1993. Embryo culture: methods and appli- Raineri, D.M., Bottino, P., Gordon, M.P., and Nester, E.W. cations. In Plant Biotechnology. Commercial Prospects 1990. Agrobacterium-mediated transformation of rice and Problems, ed. J. Prakash and R.L.M. Pierik, pp. (Oryza sativa L.). Biotechnology 8:33-38. 143-168. New Delhi: Oxford & IBH Publishing Co. Rajam, M.V. 1989. Restriction of pollen germination and Raghavan, V. 1994. In vitro methods for the control of fertube growth in lily pollen by inhibitors of polyamine tilization and embryo development. In Plant Cell and metabolism. Plant Sci. 59:53-56. Tissue Culture, ed. I.K. Vasil and T.A. Thorpe, pp. Rajasekaran, K., Vine, J., and Mullins, M.G. 1982. 173-194. Dordrecht: Kluwer Academic Publishers. Dormancy in somatic embryos and seeds of Vitis: Raghavan, V. 1995. Manipulation of pollen grains for changes in endogenous abscisic acid during embryogametophytic and sporophytic types of growth. geny and germination. Planta 154:139-144. Meth. Cell Biol. 49:367-375. Rajasekaran, K., Hein, M.B., Davis, G.C., Carnes, M.G., Raghavan, V, and Goh, C.J. 1994. DNA synthesis and and Vasil, I.K. 1987. Endogenous growth regulators mRNA accumulation during germination of in leaves and tissue cultures of Pennisetum purpureum Schum. /. Plant Physiol. 130:13-25. embryos of the orchid Spathoglottis plicata. Protoplasma 183:137-147. Rajasekhar, E.W. 1973. Nuclear divisions in protoplasts isolated from pollen tetrads of Datura metel. Nature Raghavan, V., Jiang, C, and Bimal, R. 1992. Cell- and tis246:223-224. sue-specific expression of rice histone gene transcripts during anther and pollen development in Rajora, O.P., and Zsuffa, L. 1986. Sporophytic and gamehenbane (Hyoscyamus niger). Am. J. Bot. 77:778-783. tophytic gene expression in Populus deltoides Marsh., P. nigra L., and P. maximowiczii Henry. Can. ]. Genet. Raghavan, V., and Kamalay, J.C. 1993. Expression of two Cytol. 28:476^182. cloned mRNA sequences during development and germination of spores of the sensitive fern, Onoclea Ramachandran, C, and Raghavan, V. 1989. Changes in sensibilis L. Planta 189:1-9. nuclear DNA content of endosperm cells during grain development in rice (Oryza sativa). Ann. Bot. Raghavan, V, and Nagmani, R. 1989. Cytokinin effects on 64:459-468. pollen embryogenesis in cultured anthers of Ramachandran, C, and Raghavan, V. 1990. Intracellular Hyoscyamus niger. Can. J. Bot. 67:247-257. localization of glutelin mRNA during grain developRaghavan, V, and Olmedilla, A. 1989. Spatial patterns of ment in rice. /. Expt. Bot. 41:393-399. histone mRNA expression during grain development and germination in rice. Cell Diffn Devel. Ramachandran, C, and Raghavan, V. 1992a. Apomixis in 27:183-196. distant hybridization. In Distant Hybridization in
References
625
par culture d'antheres. Analyse cytophotomiode favorable a Blakeslee: The Genus Datura, ed. A.G. Avery, S. Satina l'androgenese. Can. J. Bot. 57:450-457. and J. Rietsema, pp. 181-195. New York: Ronald Sangwan-Norreel, B.S. 1981. Evolution in vitro du contenu Press Co. en ADN nucl£aire et de la ploi'die des embryons Sato, S., Willson, C, and Dickinson, H.G. 1989. The RNA polliniques du Datura innoxia. Can. ]. Bot. 59:508-517. content of the nucleolus and nucleolus-like inclusions Sangwan-Norreel, B.S. 1983. Male gametophyte nuclear in the anther of Lilium estimated by an improved DNA content evolution during androgenic inducRNase-gold labelling method. /. Cell Sci. 94:675-683. tion in Datura innoxia Mill. Z. Pflanzenphysiol. Sato, S., Jones, K., de Los Dios Alche, J., and Dickinson, 111:47-54. H.G. 1991. Cytoplasmic nucleoloids of Lilium male Sano, Y. 1984. Differential regulation of waxy gene expresreproductive cells contain rDNA transcripts and sion in rice endosperm. Theor. Appl. Genet. share features of development with nucleoli. /. Cell 68:467-473. Sci. 100:109-118. Santanen, A., and Simola, L.K. 1992. Changes in Sato, S., Newell, C, Kolacz, K., Tredo, L., Finer, J., and polyamine metabolism during somatic embryogeneHinchee, M. 1993. Stable transformation via particle sis in Picea abies. J. Plant Physiol. 140:475-480. bombardment in two different soybean regeneration systems. Plant Cell Rep. 12:408-413. Sari Gorla, M., Frova, C, Binelli, G., and Ottaviano, E. 1986. The extent of gametophytic-sporophytic gene Sato, S., Toya, T., Kawahara, R., Whittier, R.F., Fukuda, H., expression in maize. Theor. Appl. Genet. 72:42-47. and Komamine, A. 1995. Isolation of a carrot gene Sarker, R.H., Elleman, C.J., and Dickinson, H.G. 1988. expressed specifically during early-stage somatic Control of pollen hydration in Brassica requires conembryogenesis. Plant Mol. Biol. 28:39-46. tinued protein synthesis, and glycosylation is neces- Sato, X, Thorsness, M.K., Kandasamy, M.K., Nishio, T., sary for intraspecific incompatibility. Proc. Natl Acad. Hirai, M., Nasrallah, J.B., and Nasrallah, M.E. 1991. Sci. USA 85:4340-4344. Activity of an S-locus gene promoter in pistils and Sasaki, K., Shimomura, K., Kamada, H., and Harada, H. anthers of transgenic Brassica. Plant Cell 3:867-876. 1994. IAA metabolism in embryogenic and non- Sato, Y, Hong, S.-K., Tagiri, A., Kitano, H., Yamamoto, N., embryogenic carrot cells. Plant Cell Physiol. Nagato, Y, and Matsuoka, M. 1996. A rice homeobox 35:1159-1164. gene, OSH1, is expressed before organ differentiaSasaki, Y., and Harada, H. 1991. Binding form of pollen tion in a specific region during early embryogenesis. mother cell protein in the nucleosomes of lily. Plant Proc. Natl Acad. Sci. USA 93:8117-8122. Physiol. 96:1161-1166. Satoh, H., and Omura, T. 1981. New endosperm mutaSasaki, Y, Yasuda, H., Ohba, Y, and Harada, H. 1990. tions induced by chemical mutagens in rice, Oryza Isolation and characterization of a novel nuclear prosativa L. Jap. J. Breed. 31:316-326. tein from pollen mother cells of lily. Plant Physiol. Satoh, S., Kamada, H., Harada, H., and Fujii, T. 1986. 94:1467-1471. Auxin-controlled glycoprotein release into the Sass, J.E., and Sprague, G.F. 1950. The embryology of medium of embryogenic carrot cells. Plant Physiol. "germless" maize. Iowa State Coll. ]. Sci. 24:209-218. 81:931-933. Sassa, H., Hirano, H., and Ikehashi, H. 1992. Self-incom- Saunders, J.W., and Bingham, E.T. 1972. Production of
634
References
alfalfa plants from callus tissue. Crop Sci. 12:804-808. Scalla, R., Due, G., Rigaud, J., Lefebvre, A., and Meignoz, Sauter, J.J. 1968. Histoautoradiographische Untersuchungen R. 1981. RNA containing intracellular particles in cytoplasmic male sterile faba bean (Vicia faba L.). zur Ribonudeinsaure-Synthese wahrend der Meiosis bei Paeonia teniiifolia L. Naturwissenschaften 55:236.
Sauter, J.J. 1969a. Autoradiographische Untersuchungen zur RNS- und Protein-synthese in Pollenmutterzellen, jungen Pollen und Tapetumzellen wahrend der
Plant Sci. Lett. 22:269-277.
Scallon, B., Thanh, V.H., Floener, L.A., and Nielsen, N.C. 1985. Identification and characterization of DNA clones encoding group-II glycinin subunits. Theor. Mikrosporogenese von Paeonia teniiifolia. Z. Appl. Genet. 70: 510-519. Pflatizenphysiol. 61:1-19. Scallon, B.J., Dickinson, CD., and Nielsen, N.C. 1987. Sauter, J.J. 1969b. Cytochemische Untersuchung der Characterization of a null-allele for the Gyt glycinin Histone in Zellen mit unterschiedlicher RNS- und gene from soybean. Mol. Gen. Genet. 208:107-113. Protein-Synthese. Z. Pflanzenphysiol. 60:434-449. Scanlon, M.J., Stinard, P.S., James, M.G., Myers, A.M., and Robertson, D.S. 1994. Genetic analysis of 63 mutaSauter, J.J., and Marquardt, H. 1967a. Nucleohistone und tions affecting maize kernel development isolated Ribonukleinsaure-Synthese wahrend der Pollenfrom mutator stocks. Genetics 136:281-294. entwicklung. Naturwissenschaften 54:546. Sauter, J.J., and Marquardt, H. 1967b. Die Rolle des Schaffner, M. 1906. The embryology of the shepherd's Nukleohistons bei der RNS- und Proteinsynthese purse. Ohio Naturl. 7:1-8. wahrend der Mikrosporogenese von Paeonia teniiifo- Schardl, C.L., Pring, D.R., and Lonsdale, D.M. 1985. lia L. Z. Pflanzenphysiol. 58:126-137. Mitochondrial DNA rearrangements associated with fertile revertants of S-type male-sterile maize. Cell Sautter, C, Leduc, N., Bilang, R., Iglesias, V.A., Gisel, A., 43:361-368. Wen, X., and Potrykus, I. 1995. Shoot apical meristems as a target for gene transfer by microballistics. Schardl, C.L., Lonsdale, D.M., Pring, D.R., and Rose, K.R. Euphytica 85:45-51. 1984. Linearization of maize mitochondrial chromosomes by recombination with linear episomes. Savidge, B., Rounsley, S.D., and Yanofsky, M.F. 1995. Nature 310:292-296. Temporal relationship between the transcription of two Arabidopsis MADS box genes and the floral Scharpe, A., and van Parijs, R. 1973. The formation of organ identity genes. Plant Cell 7:721-733. polyploid cells in ripening cotyledons of Pisum sativum L. in relation to ribosome and protein synSawhney, V.K. 1974. Morphogenesis of the stamenless-2 thesis. /. Expt. Bot. 24:216-222. mutant in tomato. III. Relative levels of gibberellins in the normal and mutant plants. /. Expt. Bot. Scheer, U., and Franke, W.W. 1972. Annulate lamellae in 25:1004-1009. plant cells: formation during microsporogenesis and pollen development in Canna generalis Bailey. Planta Sawhney, V.K. 1981. Abnormalities in pepper {Capsicum 107:145-159. annuum) flowers induced by gibberellic acid. Can. ]. Scheets, K., and Hedgcoth, C. 1988. Nucleotide sequence Bot. 59:8-16. of a ygliadin gene: comparisons with other y gliadin Sawhney, V.K. 1983. Temperature control of male sterility sequences show the structure of y gliadin genes and in a tomato mutant. /. Hered. 74:51-54. the general primary structure of y gliadins. Plant Sci. Sawhney, V.K., and Bhadula, S.K. 1987. Characterization 57:141-150. and temperature regulation of soluble proteins of a male sterile tomato mutant. Biochem. Genet. Scheike, R., Gerold, E., Brennicke, A., Mehring-Lemper, M., and Wricke, G. 1992. Unique patterns of mito25:717-728. chondrial genes, transcripts and proteins in different Sawhney, V.K., and Bhadula, S.K. 1988. Micromale-sterile cytoplasms of Daucus carota. Theor. Appl. sporogenesis in the normal and male-sterile staGenet. 83:419-427. menless-2 mutant of tomato (Lycopersicon esculenSchel, J.H.N., and Kieft, H. 1986. An ultrastructural study tum). Can. J. Bot. 66:2013-2021. of embryo and endosperm development during in Sawhney, V.K., Chen, K., and Sussex, I.M. 1985. Soluble vitro culture of maize ovaries (Zea mays). Can. ]. Bot. proteins of the mature floral organs of tomato 64:2227-2238. (Lycopersicon esculentum Mill.). /. Plant Physiol. 121:265-271. Schel, J.H.N., Kieft, H., and van Lammeren, A.A.M. 1984. Interactions between embryo and endosperm durSawhney, V.K., and Greyson, R.I. 1973a. Morphogenesis ing early developmental stages of maize caryopses of the stamenless-2 mutant in tomato. I. (Zea mays). Can. J. Bot. 62:2842-2853. Comparative description of the flowers and ontogeny of stamens in the normal and mutant Scherer, D., Sato, A., McCarter, D.W., Radke, S.E., Kridl, J.C., and Knauf, V.C. 1992. Non-essential repeats in plants. Am. ]. Bot. 60:514-523. the promoter region of a Brassica rapa acyl carrier Sawhney, V.K., and Greyson, R.I. 1973b. Morphogenesis of protein gene expressed in developing embryos. the stamenless-2 mutant in tomato. II. Modifications Plant Mol. Biol. 18:591-594. of sex organs in the mutant and normal flowers by Scheres, B., Wolkenfelt, H., Willemsen, V., Terlouw, M., plant hormones. Can. ]. Bot. 51:2473-2479. Lawson, E., Dean, C, and Weisbeek, P. 1994. Sawhney, V.K., and Greyson, R.I. 1979. Interpretations of Embryonic origin of the Arabidopsis primary root determination and canalisation of stamen developand root meristem initials. Development ment in a tomato mutant. Can. J. Bot. 57:2471-2477. 120:2475-2487. Sawhney, V.K., and Polowick, PL. 1986. Temperatureinduced modifications in the surface features of sta- Schernthaner, J.P., Matzke, M.A., and Matzke, A.J.M. 1988. Endosperm-specific activity of a zein gene promens of a tomato mutant: an SEM study. Protoplasma moter in transgenic tobacco plants. EMBO ]. 131:75-81. 7:1249-1255. Sawhney, V.K., and Shukla, A. 1994. Male sterility in flowering plants: are plant growth substances involved? Schiavone, F.M. 1988. Microamputation of somatic embryos of the domestic carrot reveals apical control Am. }. Bot. 81:1640-1647.
References
635
male sterility in interspecific hybrids of Epilobium. 2. of axis elongation and root regeneration. Development 103:657-664. Expression of mitochondrial genes and structure of the mitochondrial DNA. Theor. Appl. Genet. Schiavone, F.M., and Cooke, T.J. 1987. Unusual patterns of 76:565-569. somatic embryogenesis in the domesticated carrot: developmental effects of exogenous auxins and Schmulling, T., Schell, J., and Spena, A. 1988. Single genes auxin transport inhibitors. Cell Diffn 21:53-62. from Agrobacterium rhizogenes influence plant develSchiavone, F.M., and Racusen, R.H. 1990. Microsurgery opment. EMBO /. 7:2621-2629. reveals regional capabilities for pattern reestablish- Schmulling, T., Rohrig, H., Pilz, S., Walden, R., and Schell, ment in somatic carrot embryos. Devel Bio/. J. 1993. Restoration of fertility by antisense RNA in 141:211-219. genetically engineered male sterile tobacco plants. Mol. Gen. Genet. 237:385-394. Schiavone, F.M., and Racusen, R.H. 1991. Regeneration of the root pole in surgically transected carrot embryos Schnall, J.A., Cooke, T.J., and Cress, D.E. 1988. Genetic analysis of somatic embryogenesis in carrot cell culoccurs by position-dependent, proximodistal ture: initial characterization of six classes of temperreplacement of missing tissues. Development 113:1305-1313. ature-sensitive variants. Devel Genet. 9:49-67. Schick, R., and Stubbe, H. 1932. Die Gene von Antirrhinum Schnall, J.A., Hwang, C.H., Cooke, T.J., and Zimmerman, J.L. 1991. An evaluation of gene expression during majus. II. Z. Indukt. Abstamm. Vererbungs. 62:249-290. somatic embryogenesis of two temperature-sensiSchickler, H., Benner, M.S., and Messing, J. 1993. tive carrot variants unable to complete embryo Repression of the high-methionine zein gene in the development. Physiol. Plant. 82:498-504. maize inbred line Mol7. Plant J. 3:221-229. Schiefelbein, J., Galway, M., Masucci, J., and Ford, S. 1993. Schneider, T., Dinkins, R., Robinson, K., Shellhammer, J., Pollen tube and root-hair tip growth is disrupted in and Meinke, D.W. 1989. An embryo-lethal mutant of Arabidopsis thaliana is a biotin auxotroph. Devel Biol. a mutant of Arabidopsis thaliana. Plant Physiol. 131:161-167. 103:979-985. Schlag, M., and Hesse, M. 1992. The formation of the gen- Schneitz, K., Hiilskamp, M., and Pruitt, R.E. 1995. Wildtype ovule development in Arabidopsis thaliana: a erative cell in Polystachia pubescens (Orchidaceae). Sex. Plant Reprod. 5:131-137. light microscope study of cleared whole-mount tissue. Plant J. 7:731-749. Schlappi, M., and Hohn, B. 1992. Competence of immature maize embryos for Agrobacterium-mediated Schnepf, E., and Nagl, W. 1970. Uber einige gene transfer. Plant Cell 4:7-16. Strukturbesonderheiten der Suspensorzellen von Schlosser, K. 1961. Cytologische und cytochemische Phaseolus vulgaris. Protoplasma 69:133-143. Untersuchungen uber das Pollenschlauch wachstum Schopke, C, Taylor, N., Cdrcamo, R., Konan, N.K., selbststeriler Petunien. Z. Bot. 49:266-288. Marmey, P., Henshaw, G.G., Beachy, R.N., and Fauquet, C. 1996. Regeneration of transgenic cassava Schmidt, H., and Schmidt, V. 1981. Untersuchungen an plants (Manihot esculenta Crantz) from microbompollensterilen, stamenless-ahnlichen Mutanten von barded embryogenic suspension cultures. Nature Lycopersicon esculentum Mill. II. Normalisierung von Biotech. 14:731-735. ms-15 und ms-33 mit Gibberellinsaure (GA3). Biol. Zentral. 100:691-696. Schou, 0.1984. The dry and wet stigmas of Primula obconica: ultrastructural and cytochemical dimorphisms. Schmidt, R.J., Burr, F.A., and Burr, B. 1987. Transposon Protoplasma 121:99-113. tagging and molecular analysis of the maize regulaSchou, O., and Mattsson, 0.1985. Differential localization tory locus opaque-2. Science 238:960-963. of enzymes in the stigmatic exudates of Primula Schmidt, R.J., Burr, F.A., Aukerman, M.J., and Burr, B. 1990. Maize regulatory gene opaque-2 encodes a obconica. Protoplasma 125:65-74. protein with a "leucine-zipper" motif that binds to Schrauwen, J., and Linskens, H.F. 1972. Ribonuclease in styles. Planta 102:277-285. zein DNA. Proc. Natl Acad. Sci. USA 87:46-50. Schmidt, R.J., Kerudat, M., Aukerman, M.J., and Hoschek, Schrauwen, J.A.M., de Groot, P.F.M., van Herpen, M.M.A., van der Lee, T., Reynen, W.H., Weterings, G. 1992. Opaque-2 is a transcriptional activator that K.A.P., and Wullems, G.J. 1990. Stage-related expresrecognizes a specific target site in 22-kD zein genes. Plant Cell 4:689-700. sion of mRNAs during pollen development in lily and tobacco. Planta 182:298-304. Schmidt, R.J., Veit, B., Mandel, M.A., Mena, M., Hake, S., and Yanofsky, M.F. 1993. Identification and molecu- Schrauwen, J.A.M., Mettenmeyer, T., Croes, A.F., and lar characterization of ZAG1, the maize homolog of Wullems, G.J. 1996. Tapetum-specific genes: what the Arabidopsis floral homeotic gene AGAMOUS. role do they play in male gametophyte developPlant Cell 5:729-737. ment? Acta Bot. Neerl. 45:1-15. Schmit, A.-C, and Lambert, A.-M. 1987. Characterization Schrauwen, J.A.M., Reijnen, W.H., de Leeuw, H.C.G.M., and van Herpen, M.M.A. 1986. Response of pollen to and dynamics of cytoplasmic F-actin in higher plant heat stress. Acta Bot. Neerl. 35:321-327. endosperm cells during interphase, mitosis, and cytokinesis. /. Cell Biol. 105:2157-2166. Schroder, M.-B. 1984. Ultrastructural studies on plastids Schmit, A.-C, and Lambert, A.-M. 1990. Microinjected of generative and vegetative cells in the family fluorescent phalloidin in vivo reveals the F-actin Liliaceae. 1. Lilium martagon L. Biol. Zentral. dynamics and assembly in higher plant mitotic cells. 103:547-555. Schroder, M.-B. 1985a. Ultrastructural studies on plastids Plant Cell 2:129-138. of generative and vegetative cells in the family Schmitz, U.K., and Kowallik, K.V. 1987. Why are plastids Liliaceae. 2. Pritillaria imperialis and F. meleagris. Biol. maternally inherited in Epilobium? Ultrastructural observations during microgametogenesis. Plant Sci. Zentral. 104:21-27. Schroder, M.-B. 1985b. Ultrastructural studies on plastids 53:139-145. of generative and vegetative cells in Liliaceae. 3. Schmitz, U.K., and Michaelis, G. 1988. Dwarfism and
636
References
Plastid distribution during the pollen development the central cell. /. Cell Sci. 12:741-763. in Gasteria verrucosa (Mill.) Duval. Protoplasma Schulz, P., and Jensen, W.A. 1974. Capsella embryogenesis: 124:123-129. the development of the free nuclear endosperm. Schroder, M.-B. 1986a. Ultrastructural studies on plastids Protoplasma 80:183-205. of generative and vegetative cells in Liliaceae. 4. Schulz, P., and Jensen, W.A. 1977. Cotton embryogenesis: Plastid degeneration during generative cell maturathe early development of the free nuclear tion in Convallaria majalis L. Biol. Zentral. endosperm. Am. }. Bot. 64:384-394. 105:427^33. Schulz, P., and Jensen, W.A. 1981. Pre-fertilization ovule development in Capsella: ultrastructure and ultracySchroder, M.-B. 1986b. Ultrastructural studies on plastids of generative and vegetative cells in Liliaceae. 5. The tochemical localization of acid phosphatase in the behaviour of plastids during pollen development in meiocyte. Protoplasma 107:27-45. Chlorophytum comosum (Thunb.) Jacques. Theor. Appl.Schulz, P., and Jensen, W.A. 1986. Prefertilization ovule Genet. 72:840-844. development in Capsella: the dyad, tetrad, developSchroder, M.-B., and Hagemann, R. 1986. Ultrastructural ing megaspore, and two-nucleate gametophyte. Can. studies on plastids of generative and vegetative cells J. Bot. 64:875-884. in Liliaceae. 6. Patterns of plastid distribution during Schulz, P.J., Cross, J.W., and Almeida, E. 1993. Chemical generative cell formation in Aloe secundiflora and A. agents that inhibit pollen development: effects of the jucunda. Ada Bot. Neerl. 35:243-248. phenylcinnoline carboxylates SC-1058 and SC-1271 on the ultrastructure of developing wheat anthers Schroder, M.-B., and Oldenburg, H. 1990. Ultrastructural (Triticum aestivum L. var. Yecora rojb). Sex. Plant studies on plastids of generative and vegetative cells Reprod. 6:108-121. in Liliaceae. 7. Plastid distribution during generative cell development in Tulbaghia violacea Harv. FloraSchulz, R., and Jensen, W.A. 1968a. Capsella embryogene184:131-136. sis: the early embryo. /. Ultrastr. Res. 22:376-392. Schroeder, H.E. 1982. Quantitative studies on the cotyle- Schulz, R., and Jensen, W.A. 1968b. Capsella embryogenesis: the synergids before and after fertilization. Am. J. donary proteins in the genus Pisum. ]. Sci. Food Agr. 33:623-633. Bot. 55:541-552. Schroeder, H.E. 1984. Effects of applied growth regulators Schulz, R., and Jensen, W.A. 1968c. Capsella embryogenesis: the egg, zygote, and young embryo. Am. ]. Bot. on pod growth and seed protein composition in Pisum sativum L. /. Expt. Bot. 35:813-821. 55:807-819. Schroeder, H.E., Schotz, A.H., Wardley-Richardson, X, Schumann, CM., and Hancock, J.F. 1989. Paternal inheriSpencer, D., and Higgins, T.J.V. 1993. Transformation tance of plastids in Medicago sativa. Theor. Appl. Genet. 78:863-866. and regeneration of two cultivars of pea (Pisum sativum L.). Plant Physiol. 101:751-757. Schuster, W. 1961. Untersuchungen iiber kiinstlich induzierte Pollensterilitat bei Sonnenblumen Schubert, R., Baumlein, H., Czihal, A., and Wobus, U. (Helianthus annuus L.). Z. Pflanzenzuchtg 46:389-404. 1990. Genomic sequence of a 12S seed storage protein gene from oat (Avena sativa L. cv. ~Solidor'). Schwab, C.A. 1971. Callose in megasporogenesis of Nucl. Acids Res. 18:377. Diarrhena (Gramineae). Can. }. Bot. 49:1523-1524. Schubert, R., Panitz, R., Manteuffel, R., Nagy, I., Wobus, Schwall, M., and Feix, G. 1988. Zein promoter activity in transiently transformed protoplasts from maize. U., and Baumlein, H. 1994. Tissue-specific expression of an oat 12S seed globulin gene in developing Plant Sci. 56:161-166. tobacco seeds.: differential mRNA and protein accu- Schwartz, B.W., Yeung, E.C., and Meinke, D.W. 1994. mulation. Plant Mol. Biol. 26:203-210. Disruption of morphogenesis and transformation of Schuler, M.A., Schmitt, E.S., and Beachy, R.N. 1982. the suspensor in abnormal suspensor mutants of Closely related families of genes code for the a and Arabidopsis. Development 120:3235-3245. a' subunits of the soybean 7S storage protein com- Schwarz-Sommer, Z., Hue, I., Huijser, P., Flor, P.J., plex. Nucl. Acids Res. 10:8225-8244. Hansen, R., Tetens, F., Lonnig, W.-E., Saedler, F., and Sommer, H. 1992. Characterization of the Schuler, M.A., Ladin, B.R, Pollacco, J.C., Freyer, G., and Antirrhinum floral homeotic MADS-box gene defiBeachy, R.N. 1982. Structural sequences are conciens: evidence for DNA binding and autoregulation served in the genes coding for the a, a' and (5-subof its persistent expression throughout flower develunits of the soybean 7S seed storage protein. Nucl. opment. EMBOJ. 11:251-263. Acids Res. 10:8245-8261. Schultz, E.A., and Haughn, G.W. 1991. LEAFY, a homeotic Schwarz-Sommer, Z., Huijser, P., Nacken, W., Saedler, H., and Sommer, H. 1990. Genetic control of flower gene that regulates inflorescence development in development by homeotic genes in Antirrhinum Arabidopsis. Plant Cell 3:771-781. majus. Science 250:931-936. Schultz, E.A., and Haughn, G.W. 1993. Genetic analysis of the floral initiation process (FLIP) in Arabidopsis. Scofield, S.R., and Crouch, M.L. 1987. Nucleotide Development 119:745-765. sequence of a member of the napin storage proSchultz, E.A., Pickett, F.B., and Haughn, G.W. 1991. The tein family from Brassies napus. ]. Biol. Chem. FLOW gene product regulates the expression 262:12202-12208. domain of homeotic genes AP3 and PI in Arabidopsis Scoles, G.J., and Evans, L.E. 1979. Pollen development in male-fertile and cytoplasmic male-sterile rye. Can. J. flowers. Plant Cell 3:1221-1237. Bot. 57:2782-2790. Schulz, P., and Jensen, W.A. 1969. Capsella embryogenesis: the suspensor and the basal cell. Protoplasma Scorza, R., Cordts, J.M., Ramming, D.W., and Emershad, R.L. 1995. Transformation of grape (Vitis vinifera L.) 67:139-163. zygotic-derived somatic embryos and regeneration Schulz, P., and Jensen, W.A. 1971. Capsella embryogenesis: of transgenic plants. Plant Cell Rep. 14:589-592. the chalazal proliferating tissue. /. Cell Sci. 8:201-227. Schulz, P., and Jensen, W.A. 1973. Capsella embryogenesis: Scott, P., Lyne, R.L., and ap Rees, T. 1995. Metabolism of
References
637
maltose and sucrose by microspores isolated from Sedgley, M., Blesing, M.A., Bonig, I., Anderson, M.A., and Clarke, A.E. 1985. Arabinogalactan-proteins are barley (Hordeum vulgare L.). Planta 197:435-441. localized extracellularly in the transmitting tissue of Scott, R., Dagless, E., Hodge, R., Paul, W., Soufleri, I., and Nicotiana alata Link and Otto, an ornamental Draper, J. 1991. Patterns of gene expression in develtobacco. Micron Micros. Ada 16:247-254. oping anthers of Brassica napits. Plant Mol. Biol. 17:195-207. Seffens, W.S., Almoguera, C, Wilde, H.D., Vonder Haar, R.A., and Thomas, T.L. 1990. Molecular analysis of a Scribailo, R.W., and Barrett, S.C.H. 1991a. Pollen-pistil phylogenetically conserved carrot gene: developinteractions in tristylous Pontederia sagittata mental and environmental regulation. Devel Genet. (Pontederiaceae). I. Floral heteromorphism and 11:65-76. structural features of the pollen tube pathway. Am. ]. Bot. 78:1643-1661. Sehgal, C.B., and Gifford, E.M., Jr. 1979. Developmental and histochemical studies of the ovules of Nicotiana Scribailo, R.W., and Barrett, S.C.H. 1991b. Pollen-pistil mstica L. Bot. Gaz. 140:180-188. interactions in tristylous Pontederia sagittata (Pontederiaceae). II. Patterns of pollen tube growth. Sen, S.K. 1969. Chromatin-organisation during and after synapsis in cultured microsporocytes of Liliutn in Am. ]. Bot. 78:1662-1682. presence of mitomycin C and cycloheximide. Expt. Scutt, C.P., and Croy, R.R.D. 1992. An S5 self-incompatiCell Res. 55:123-127. bility allele-specific cDNA sequence from Brassica oleracea shows high homology to the SLR2 gene. Mol. Sen, S.K., Kundu, S.C., and Gaddipati, J.P. 1977. High Gen. Genet. 232:240-246. speed scintillation autoradiography of DNA fibres undergoing DNA synthesis at zygotene and Scutt, C.P., Fordham-Skelton, A.P., and Croy, R.R.D. 1993. pachytene in the lily. Expt. Cell Res. 108:471-i73. Okadaic acid causes breakdown of self-incompatibility in Brassica oleracea: evidence for the involve-Senda, M., Mikami, T., and Kinoshita, T. 1993. The sugar ment of protein phosphatases in the incompatible beet mitochondrial gene for the ATPase alpha-subunit: response. Sex. Plant Reprod. 6:282-285. sequence, transcription and rearrangements in cytoplasmic male-sterile plants. Curr. Genet. 24:164-170. Scutt, C.P., Gates, P.J., Gatehouse, J.A., Boulter, D., and Croy, R.R.D. 1990. A cDNA encoding an S-locus spe- Senda, M., Harada, T., Mikami, T., Sugiura, M., and cific glycoprotein from Brassica oleracea plants conKinoshita, T. 1991. Genomic organization and taining the S5 self-incompatibility allele. Mol. Gen. sequence analysis of the cytochrome oxidase subunit II gene from normal and male-sterile mitochondria Genet. 220:409-113. in sugar beet. Curr. Genet. 19:175-181. Seavey, S.R., and Bawa, K.S. 1986. Late-acting self-incomSengupta, C, and Raghavan, V. 1980a. Somatic embryopatibility in angiosperms. Bot. Rev. 52:195-219. Secor, D.L., and Russell, S.D. 1988. Megagametophyte genesis in carrot cell suspension. I. Pattern of protein organization in a polyembryonic line of Linum usiand nucleic acid synthesis. /. Expt. Bot. 31:247-258. tatissimum. Am. J. Bot. 75:114-122. Sengupta, C, and Raghavan, V. 1980b. Somatic embryogenesis in carrot cell suspension. II. Synthesis of Sedgley, M. 1974. The concentration of S-protein in stigribosomal RNA and poly (A)+ RNA. /. Expt. Bot. mas of Brassica oleracea plants homozygous and heterozygous for a given S-allele. Heredity 33:412-416. 31:259-268. Sedgley, M. 1975. Flavanoids in pollen and stigma of Sengupta, G, Deluca, V., Bailey, D.S., and Verma, D.P.S. Brassica oleracea and their effects on pollen germina1981. Post-translational processing of 7S and US tion in vitro. Ann. Bot. 39:1091-1095. components of soybean storage proteins. Plant Mol. Biol. 1:19-33. Sedgley, M. 1979. Structural changes in the pollinated and unpollinated avocado stigma and style. /. Cell Sci. Sengupta-Gopalan, C, Reichert, N.A., Barker, R.F., Hall, 38:49-60. T.C., and Kemp, J.D. 1985. Developmentally regulated expression of the bean fl-phaseolin gene in Sedgley, M. 1981a. Ultrastructure and histochemistry of tobacco seed. Proc. Nail Acad. Sci. USA 82:3320-3324. the watermelon stigma. /. Cell Sci. 48:137-146. Sedgley, M. 1981b. Anatomical aspects of compatible Serrano, L., Rochange, F., Semblat, J.P., Marque, C, pollen-stigma interaction. Phytomorphology Teulieres, C, and Boudet, A.-M. 1996. Genetic transformation of Eucalyptus globulus through biolistics: 31:158-165. Sedgley, M. 1981c. Early development of the Macadamia complementary development of procedures for ovary. Aust. ]. Bot. 29:185-193. organogenesis from zygotic embryos and stable Sedgley, M. 1982. Anatomy of the unpollinated and pollitransformation of corresponding proliferating tisnated watermelon stigma. /. Cell Sci. 54:341-355. sue. /. Expt. Bot. 47:285-290. Sedgley, M., and Blesing, M.A. 1982. Foreign pollination Sessa, G., and Fluhr, R. 1995. The expression of an abunof the stigma of watermelon (Citrullus lanatus dant transmitting tract-specific endoglucanase [Thunb.] Matsum and Nakai). Bot. Gaz. 143:210-215. (Sp41) is promoter-dependent and not essential for Sedgley, M., and Blesing, M.A. 1983. Developmental the reproductive physiology of tobacco. Plant Mol. anatomy of the avocado stigma papilla cells and Biol. 29:969-982. their secretion. Bot. Gaz. 144:185-190. Sessions, R.A., and Zambryski, P.C. 1995. Arabidopsis gynoecium structure in the wild type and in ettin Sedgley, M., and Buttrose, M.S. 1978. Structure of the mutants. Development 121:1519-1532. stigma and style of the avocado. Aust. J. Bot. 26:663-682. Sethi, R.S., and Jensen, W.A. 1981. Ultrastructural eviSedgley, M., and Clarke, A.E. 1986. Immuno-gold localidence for secretory phases in the life history of stigsation of arabinogalactan protein in the developing matic hairs of cotton: a dry type stigma. Am. }. Bot. style of Nicotiana alata. Nord. J. Bot. 6:591-598. 68:666-674. Sedgley, M., and Scholefield, P.B. 1980. Stigma secretion in Seurinck, J., Truettner, J., and Goldberg, R.B. 1990. The the watermelon before and after pollination. Bot. nucleotide sequence of an anther-specific gene. Nucl. Gaz. 141:428-434. Acids Res. 18:3403.
638
References
Shah, C.K., and Pandey, S.N. 1978. Histochemical studies mutants of maize. II. Morphological and embryo during embryogenesis in Limnophyton obtusifolium. culture studies. Genetics 95:945-960. Phytomorphology 28:31-42. Sheridan, W.F., and Neuffer, M.G. 1982. Maize developShannon, L.M., and Chrispeels, M.J. 1986. Molecular mental mutants. Embryos unable to form leaf priBiology of Seed Storage Proteins and Lectins. Rockville: mordia. /. Hered. 73:318-329. American Society of Plant Physiologists. Sheridan, W.F., and Stern, H. 1967. Histones of meiosis. Shannon, T.M., Picton, J.M., and Steer, M.W. 1984. The Expt. Cell Res. 45:323-335. inhibition of dictyosome vesicle formation in higher Sheridan, W.F., and Thorstenson, Y.R. 1986. plant cells by cytochalasin D. Eur. ]. Cell Bid. Developmental profiles of three embryo-lethal 33:144-147. maize mutants lacking leaf primordia: ptd*-1130, cp*Sharma, N., Bajaj, M., and Shivanna, K.R. 1985. 1418, and bno*-747B. Devel Genet. 7:35-19. Overcoming self-incompatibility through the use of Sheridan, W.F., Avalkina, N.A., Shamrov, I.I., Batygina, lectins and sugars in Petunia and Eruca. Ann. Bot. T.B., and Golubovskaya, I.N. 1996. The macl gene: 55:139-141. Controlling the commitment to the meiotic pathway Sharma, N., and Shivanna, K.R. 1982. Effects of pistil in maize. Genetics 142:1009-1020. extracts on in vitro responses of compatible and Sherwood, R.T. 1995. Nuclear DNA amount during incompatible pollen in Petunia hybrida Vilm. Indian }. sporogenesis and gametogenesis in sexual and Expt. Biol. 20:255-256. aposporous buffelgrass. Sex. Plant Reprod. 8:85-90. Sharma, N., and Shivanna, K.R. 1983. Lectin-like compo- Shevell, D.E., Leu, W.-M., Gillmor, C.S., Xia, G., nents of pollen and complementary saccharide moiFeldmann, K.A., and Chua, N.-H. 1994. EMB30 is ety of the pistil are involved in self-incompatibility essential for normal cell division, cell expansion, and recognition. Curr. Sci. 52:913-916. cell adhesion in Arabidopsis and encodes a protein that has similarity to Sec7. Cell 77:1051-1062. Sharma, N., and Shivanna, K.R. 1986. Treatment of the stigma with an extract of a compatible pistil over- Shewry, PR. 1995. Plant storage proteins. Biol. Rev. comes self-incompatibility in Petunia. New Phytol. 70:375-426. 102:443-447. Shewry, PR., Kreis, M., Burgess, S.R., Parmar, S., and Sharma, S., Singh, M.B., and Malik, C.P. 1981. Dark CO2 Miflin, B.J. 1983. The synthesis and deposition of the fixation during germination of Amaryllis vittata prolamin storage proteins (secalins) of rye. Planta pollen in suspension cultures. Indian ]. Expt. Biol. 159:439^55. 19:710-714. Shewry, PR., Pratt, H.M., Leggatt, M.M., and Miflin, B.J. 1979. Protein metabolism in developing endosperms Shayk, M., Kolattukudy, P.E., and Davis, R. 1977. of high-lysine and normal barley. Cereal Chem. Production of a novel extracellular cutinase by the 56:110-117. pollen and the chemical composition and ultrastructure of the stigma cuticle of nasturtium (Tropaeolum Shi, L., Mogensen, H.L., and Zhu, T. 1991. Dynamics of majus). Plant Physiol. 60:907-915. nuclear pore density and distribution patterns Sheffield, E., Cawood, A.H., Bell, P.R., and Dickinson, within developing pollen: implications for a funcH.G. 1979. The development of nuclear vacuoles tional relationship between the vegetative nucleus during meiosis in plants. Planta 146:597-601. and the generative cell. /. Cell Sci. 99:115-120. Sheldon, J.M., and Dickinson, H.G. 1983. Determination Shi, L., Zhu, T., Mogensen, H.L., and Keim, P. 1996. Sperm identification in maize by fluorescence in situ of patterning in the pollen wall of Lilium henryi. ]. hybridization. Plant Cell 8:815-821. Cell Sci. 63:191-208. Sheldon, J.M., and Dickinson, H.G. 1986. Pollen wall for- Shi, L., Zhu, T., Mogensen, H.L., and Smith, S.E. 1991. mation in Lilium: the effect of chaotropic agents, and Paternal plastid inheritance in alfalfa: plastid the organisation of the microtubular cytoskeleton nucleoid number within generative cells correlates during pattern development. Planta 168:11-23. poorly with plastid number and male plastid transmission strength. Curr. Genet. 19:399-401. Sheldon, J.M., and Hawes, C. 1988. The actin cytoskeleton during male meiosis in Lilium. Cell Biol. Int. Rep. Shiba, H., Hinata, K., Suzuki, A., and Isogai, A. 1995. 12:471^176. Breakdown of selfincompatibility in Brassica by the Shellhammer, ]., and Meinke, D. 1990. Arrested embryos antisense RNA of the SLG gene. Proc. Jap. Acad. 71B:81-83. from the biol auxotroph of Arabidopsis thaliana contain reduced levels of biotin. Plant Physiol. Shibata, D., Steczko, }., Dixon, J.E., Andrews, P C , Hermodson, M., and Axelrod, B. 1988. Primary 93:1162-1167. structure of soybean lipoxygenase L-2. /. Biol. Chem. Shen, J.B., and Hsu, F.C. 1992. Brassica anther-specific 263:6816-6821. genes: characterization and in situ localization of expression. Mol. Gen. Genet. 234:379-389. Shifriss, O. 1945. Male sterilities and albino seedlings in Shepard, ]., Boothroyd, E.R., and Stern, H. 1974. The effect cucurbits. A study in inbreeding. /. Hered. 36:47-52. of colchicine on synapsis and chiasma formation in Shii, C.T., Rabakoarihanta, A., Mok, M.C., and Mok, microsporocytes of Lilium. Chromosoma 44:423-437. D.W.S. 1982. Embryo development in reciprocal crosses of Phaseolus vulgaris L. and P. coccineus Lam. Sheridan, W.F. 1973. Nonaqueous isolation of nuclei from Theor. Appl. Genet. 62:59-64. lily pollen and an examination of their histones. Z. Pflanzenphysiol. 68:450-459. Shimamoto, K., Ackermann, M., and Dierks-Ventling, C. 1983. Expression of zein in long term endosperm Sheridan, W.F., and Clark, J.K. 1987. Maize embryogeny: a cultures of maize. Plant Physiol. 73:915-920. promising experimental system. Trend. Genet. 3:3-6. Sheridan, W.F., and Clark, J.K. 1993. Mutational analysis Shirsat, A., Wilford, N., Cray, R., and Boulter, D. 1989. Sequences responsible for the tissue specific proof morphogenesis of the maize embryo. Plant ]. moter activity of a pea legumin gene in tobacco. Mol. 3:347-358. Gen. Genet. 215:326-331. Sheridan, W.F., and Neuffer, M.G. 1980. Defective kernel
References
639
Shirsat, A.H., Meakin, P.J., and Gatehouse, J.A. 1990. Shotwell, M.A., Boyer, S.K., Chesnut, R.S., and Larkins, Sequences 5' to the conserved 28 bp leg box eleB.A. 1990. Analysis of seed storage protein genes of ment regulate the expression of pea seed storage oats. /. Biol. Chem. 265:9652-9658. protein gene legA. Plant Mol. Biol. 15:685-693. Shoup, J.R., Overton, J., and Ruddat, M. 1980. infrastructure and development of the sexine in the Shivanna, K.R. 1965. In vitro fertilization and seed formapollen wall of Silene alba (Caryophyllaceae). Bot. Gaz. tion in Petunia violacea Lindl. Phytomorphology 141:379-388. 15:183-185. Shivanna, K.R. 1982. Pollen-pistil interaction and control Shoup, J.R., Overton, J., and Ruddat, M. 1981. of fertilization. In Experimental Embryology of Vascular infrastructure and development of the nexine and Plants, ed. B.M. Johri, pp. 131-174. Berlin: Springerintine in the pollen wall of Silene alba Verlag. (Caryophyllaceae). Am. ]. Bot. 68:1090-1095. Shivanna, K.R., and Cresti, M. 1989. Effects of high Shufflebottom, D., Edwards, K., Schuch, W., and Bevan, M. 1993. Transcription of two members of a gene humidity and temperature stress on pollen memfamily encoding phenylalanine ammonia-Iyase brane integrity and pollen vigour in Nicotiana leads to remarkably different cell specificities and tabacum. Sex. Plant Reprod. 2:137-141. induction patterns. Plant J. 3:835-845. Shivanna, K.R., and Heslop-Harrison, J. 1981. Membrane state and pollen viability. Ann. Bot. 47:759-770. Shukla, A., and Sawhney, V.K. 1992. Cytokinins in a genie Shivanna, K.R., Heslop-Harrison, J., and Heslopmale sterile line of Brassica napus. Physiol. Plant. Harrison, Y. 1981. Heterostyly in Primula. 2. Sites of 85:23-29. pollen inhibition, and effects of pistil constituents on Shukla, A., and Sawhney, V.K. 1994. Abscisic acid: one of compatible and incompatible pollen-tube growth. the factors affecting male sterility in Brassica napus. Protoplasma 107:319-337. Physiol. Plant. 91:522-528. Shivanna, K.R., Heslop-Harrison, J., and Heslop- Shure, M., Wessler, S., and Fedoroff, N. 1983. Molecular Harrison, Y. 1983. Heterostyly in Primula. 3. Pollen identification and isolation of the waxy locus in water economy: a factor in the intramorph-incommaize. Cell 35:225-233. patibility response. Protoplasma 117:175-184. Shuttuck-Eidens, D.M., and Beachy, R.M. 1985. Shivanna, K.R., Heslop-Harrison, Y, and Heslop-Harrison, Degradation of p-conglycinin in early stages of soyJ. 1978. The pollen-stigma interaction: bud pollination bean embryogenesis. Plant Physiol. 78:895-898. in the Cruciferae. Ada Bot. Afeer//27:107-119. Siculella, L., and Palmer, J.D. 1988. Physical and gene Shivanna, K.R., Heslop-Harrison, Y, and Hesloporganization of mitochondrial DNA in fertile and Harrison, J. 1982. The pollen-stigma interaction in male sterile sunflower. CMS-associated alterations the grasses. 3. Features of the self-incompatibility in structure and transcription of the atpA gene. Nucl. response. Ada Bot. Neerl. 31:307-319. Acids Res. 16:3787-3799. Shivanna, K.R., Jaiswal, V.S., and Mohan Ram, H.Y 1974a. Sidorova, N.V. 1985. Characteristics of the ultrastructure Inhibition of gamete formation by cycloheximide in of isolated embryo sacs of tobacco. Doklady Akad. pollen tubes of Impatiens balsamina. Planta Nauk Ukr. SSR Ser. B 12:63-66. 117:173-177. Siebertz, B., Logemann, J., Willmitzer, L., and Schell, J. Shivanna, K.R., Jaiswal, V.S., and Mohan Ram, H.Y. 1974b. 1989. cis-Analysis of the wound-inducible promoter Effect of cycloheximide on cultured pollen grains of wunl in transgenic tobacco plants and histochemical Trigonella foenum-graecum. Plant Sci. Lett. 3:335-339. localization of its expression. Plant Cell 1:961-968. Shivanna, K.R., and Johri, B.M. 1985. The Angiosperm Silcock, D.J., Francis, D., Bryant, J.A., and Hughes, S.G. Pollen. Strudure and Function. New Delhi: Wiley 1990. Changes in nuclear DNA content, cell and Eastern Ltd. nuclear size, and frequency of cell division in the Shivanna, K.R., and Rangaswamy, N.S. 1969. Overcoming cotyledons of Brassica napus L. during embryogeneself-incompatibility in Petunia axillaris. I. Delayed sis. /. Expt. Bot. 41:401^07. pollination, pollination with stored pollen, and bud Silvanovich, A., Astwood, J., Zhang, L., Olsen, E., Kisil, F., pollination. Phytomorphology 19:372-380. Sehon, A., Mohapatra, S., and Hill, R. 1991. Nucleotide sequence analysis of three cDNAs codShivanna, K.R., and Rangaswamy, N.S. 1992. Pollen Biology. A Laboratory Manual. Berlin: Springer-Verlag. ing for Poa p IX isoallergens of Kentucky bluegrass Shivanna, K.R., and Sastri, D.C. 1981. Stigma-surface pollen. /. Biol. Chem. 266:1204-1210. esterase activity and stigma receptivity in some taxa Simon, A.E., Tenbarge, K.M., Scofield, S.R., Finkelstein, characterized by wet stigmas. Ann. Bot. 47:53-64. R.R., and Crouch, M.L. 1985. Nucleotide sequence of Shivanna, K.R., Xu, H., Taylor, P., and Knox, R.B. 1988. a cDNA clone of Brassica napus 12S storage protein Isolation of sperms from the pollen tubes of flowering shows homology with Iegumin from Pisum sativum. plants during fertilization. Plant Physiol. 87:647-650. Plant Mol. Biol. 5:191-201. Shoemaker, R.C., Christofferson, S.E., and Galbraith, D.W. Simon, R., Altschuler, Y, Rubin, R., and Galili, G. 1990. 1987. Storage protein accumulation patterns in Two closely related wheat storage proteins follow a somatic embryos of cotton (Gossypium hirsutum L.). markedly different subcellular route in Xenopus laePlant Cell Rep. 6:12-15. vis oocytes. Plant Cell 2:941-950. Shotwell, M.A., and Larkins, B.A. 1989. The biochemistry Simoncioli, C. 1974. Ultrastructural characteristics of and molecular biology of seed storage proteins. In "Diplotaxis erucoides (L.) DC." suspensor. Giorn. Bot. The Biochemistry of Plants. A Comprehensive Treatise, Hal. 108:175-189. Vol. 15, ed. A. Marcus, pp. 297-345. San Diego: Simpson, M.G., and Levin, G.A. 1994. Pollen ultrastrucAcademic Press. ture of the biovulate Euphorbiaceae. Int. J. Plant Sci. Shotwell, M.A., Afonso, C, Davies, E., Chesnut, R.S., and 155:313-341. Larkins, B.A. 1988. Molecular characterization of oat Sims, T.L. 1993. Genetic regulation of self-incompatibility. seed globulins. Plant Physiol. 87:698-704. Crit. Rev. Plant Sci. 12:129-167.
640
References
Singh, A., Ai, Y, and Kao, T.-H. 1991. Characterization of Temperature effects on endogenous indole-3-acetic ribonuclease activity of three S-allele-associated proacid levels in leaves and stamens of the normal and teins of Petunia inflata. Plant Physiol. 96:61-68. male sterile "stamenless-2' mutant of tomato (Lycopersicon esculentum Mill.). Plant Cell Environ. Singh, A., Evensen, K.B., and Kao, T.-H. 1992. Ethylene 15:373-377. synthesis and floral senescence following compatible and incompatible pollinations in Petunia inflata. Singh, S., and Walles, B. 1992. The ovarian transmitting Plant Physiol. 99:38-45. tissue in Lilium regale. Int. }. Plant Sci. 153:205-211. Singh, A., and Kao, T.-H. 1992. Gametophytic self-incom- Singh, S., and Walles, B. 1995. Ultrastructural differentiapatibility: biochemical, molecular genetic, and evotion of the ovarian transmitting tissue in Lilium lutionary aspects. Int. Rev. Cytol. 140:449-483. regale. Ann. Bot. 75:455-462. Singh, A., and Paolillo, D.J., Jr. 1990. Role of calcium in the Singh, S.P., and Hadley, H.H. 1961. Pollen abortion callose response of self-pollinated Brassica stigmas. in cytoplasmic male sterile Sorghum. Crop Sci. Am. J. Bot. 77:128-133. 1:430-432. Singh, A., Perdue, T.D., and Paolillo, D.J., Jr. 1989. Pollen- Singh, S.P., and Rhodes, A.M. 1961. A morphological and pistil interactions in Brassica oleracea: cell calcium in cytological study of male sterility in Cucurbita maxself and cross pollen grains. Protoplasma 151:57-61. ima. Proc. Am. Soc. Hort. Sci. 78:375-378. Singh, A.P., and Mogensen, H.L. 1975. Fine structure of Singleton, W.R., and Mangelsdorf, PC. 1940. Gametic the zygote and early embryo in Quercus gambelii. Am. lethals on the fourth chromosome of maize. Genetics J. Bot. 62:105-115. 25:366-390. Singh, A.P., and Mogensen, H.L. 1976. Fine structure of Sinha, S., Jha, K.K., and Roy, R.P. 1978. Segmentation patearly endosperm in Quercus gambelii. Cytologia tern of pollen in anther culture of Solanum surattense, 41:345-361. Luffa cylindrica and Luffa echinata. Phytomorphology Singh, M., Bhalla, P.L., and Malik, C.P. 1979. Peroxidase 28:43-49. localization during embryogenesis in Tropaeolum Sivaramakrishna, D. 1978. Size relationships of apical cell majiis. Phytomorphology 29:306-309. and basal cell in two-celled embryos in angiosperms. Can. ]. Bot. 56:1434-1438. Singh, M., and Brown, G.G. 1991. Suppression of cytoplasmic male sterility by nuclear genes alters expres- Sjodahl, S., Rodin, J., and Rask, L. 1991. Characterization sion of a novel mitochondrial gene region. Plant Cell of the 12S globulin complex of Brassica napus. 3:1349-1362. Evolutionary relationship to other 11-12S storage globulins. Eur. J. Biochem. 196:617-621. Singh, M.B., Bhalla, P.L., and Malik, C.P. 1980. Activity of some hydrolytic enzymes in autolysis of the embryo Sjodahl, S., Gustavsson, H.-O., Rodin, J., Lenman, M., suspensor in Tropaeolum majus L. Ann. Bot. Hoglund, A.-S., and Rask, L. 1993. Cruciferin gene 45:523-527. families are expressed coordinately but with tissueSingh, M.B., and Knox, R.B. 1984. Invertases of Liliwn specific differences during Brassica napus seed development. Plant Mol. Biol. 23:1165-1176. pollen. Characterization and activity during in vitro germination. Plant Physiol. 74:510-515. Sjodahl, S., Gustavsson, H.-O., Rodin, J., and Rask, L. Singh, M.B., and Knox, R.B. 1985. Gene controlling P1995. Deletion analysis of the Brassica napus crugalactosidase deficiency in pollen of oilseed rape. /. ciferin gene cru 1 promoter in transformed tobacco: Hered. 76:199-201. promoter activity during early and late stages of embryogenesis is influenced by as-acting elements Singh, M.B., and Malik, C.P. 1976. Histochemical studies in partially separate regions. Planta 197:264-271. on reserve substances and enzymes in female gametophyte of Zea mays. Ada Biol. Acad. Sci. Hung. Skokut, T.A., Manchester, J., and Schaefer, J. 1985. 27:231-244. Regeneration in alfalfa tissue culture. Stimulation of Singh, M.B., O'Neill, P.M., and Knox, R.B. 1985. Initiation somatic embryo production by amino acids and Nof postmeiotic p-galactosidase synthesis during 15 NMR determination of nitrogen utilization. Plant Physiol. 79:579-583. microsporogenesis in oilseed rape. Plant Physiol. 77:225-228. Skorupska, H.T., Desamero, N.V., and Palmer, R.G. 1993. Possible homeosis in flower development in cultiSingh, M.B., Hough, T., Theerakulpisut, P., Avjioglu, A., vated soybean [Glycine max (L.) Merr.]: morDavies, S., Smith, P.M., Taylor, P., Simpson, R.J., phoanatomy and genetics. /. Hered. 84:97-104. Ward, L.D., McCluskey, J., Puy, R., and Knox, R.B. 1991. Isolation of cDNA encoding a newly identified Skriver, K., Leah, R., Miiller-Uri, E, Olsen, F.-L., and Mundy, major allergenic protein of rye-grass pollen: intracelJ. 1992. Structure and expression of the barley lipid lular targeting to the amyloplast. Proc. Natl Acad. Sci. transfer protein gene Ltpl. Plant Mol. Biol. 18:585-589. USA 88:1384-1388. Skvarla, J.J., and Larson, D.A. 1966. Fine structural studies of Zea mays pollen I: cell membranes and exine Singh, N.K., and Shepherd, K.W. 1987. Solubility behaviour, synthesis, degradation and subcellular localizaontogeny. Am. }. Bot. 53:1112-1125. tion of a new class of disulfide-linked proteins in Skvarla, J.J., and Rowley, J.R. 1970. The pollen wall of Canna and its similarity to the germinal apertures of wheat endosperm. Aust. }. Plant Physiol. 14:245-252. other pollen. Am. }. Bot. 57:519-529. Singh, N.K., Donovan, G.R., Carpenter, H.C., Skerritt, J.H., and Langridge, P. 1993. Isolation and character- Sladky, Z., and Havel, L. 1976. The study of the conditions ization of wheat triticin cDNA revealing a unique for the fertilization in vitro in maize. Biol. Plant. lysine-rich repetitive domain. Plant Mol. Biol. 18:469-472. 22:227-237. Slay, R.M., Grimes, H.D., and Hodges, T.K. 1989. Plasma membrane proteins associated with undifferentiated Singh, S., and Sawhney, V.K. 1992. Cytokinins in a normal and embryonic Daucus carota tissue. Protoplasma line and the ogura (ogu) cytoplasmic male-sterile line 150:139-149. of rapeseed (Brassica napus). Plant Sci. 86:147-154. Singh, S., Sawhney, V.K., and Pearce, D.W. 1992. Slightom, J.L., Sun, S.M., and Hall, T.C. 1983. Complete
References
nucleotide sequence of a French bean storage protein gene: phaseolin. Proc. Nail Acad. Sci. USA 80:1987-1901. Slightom, J.L., Drong, R.F., Klassy, R.C., and Hoffman, L.M. 1985. Nucleotide sequences from phaseolin cDNA clones: the major storage proteins from Phaseolus vulgaris are encoded by two unique gene families. Nucl. Acids Res. 13:6483-6498. Slocombe, S.P., Cummins, I., Jarvis, R.P., and Murphy, D.J. 1992. Nucleotide sequence and temporal regulation of a seed-specific Brassica napus cDNA encoding a stearoyl-acyl carrier protein (ACP). Plant Mol. Biol. 20:151-155. Slocombe, S.P., Piffanelli, P., Fairbairn, D., Bowra, S., Hatzopoulos, P., Tsiantis, M., and Murphy, D.J. 1994. Temporal and tissue-specific regulation of a Brassica napus stearoyl-acyl carrier protein desaturase gene. Plant Physiol. 104:1167-1176. Smart, C.J., Moneger, F., and Leaver, CJ. 1994. Cell-specific regulation of gene expression in mitochondria during anther development in sunflower. Plant Cell 6:811-825. Smart, M.G., and O'Brien, T.P. 1983. The development of the wheat embryo in relation to the neighbouring tissues. Protoplasma 114:1-13. Smith, A.G., Gasser, C.G., Budelier, K.A., and Fraley, R.T. 1990. Identification and characterization of stamenand tapetum-specific genes from tomato. Mol. Gen. Genet. 222:9-16. Smith, A.M. 1988. Major differences in isoforms of starchbranching enzyme between developing embryos of round- and wrinkled-seeded peas (Pisum sativum L.). Planta 175:270-279. Smith, C.G. 1974. The ultrastructural development of spherosomes and oil bodies in the developing
641
tion of a novel gene in barley embryos. Plant Mol. Biol. 20:255-266. Smith, R.L., and Chowdhury, M.K.U. 1991. Characterization of pearl millet mitochondrial DNA fragments rearranged by reversion from cytoplasmic male sterility to fertility. Theor. Appl. Genet. 81:793-799. Smith, R.L., Chowdhury, M.K.U., and Pring, D.R. 1987. Mitochondrial DNA rearrangements in Pennisetum associated with reversion from cytoplasmic male sterility to fertility. Plant Mol. Biol. 9:277-286. Smith, S.M., and Street, H.E. 1974. The decline of embryogenic potential as callus and suspension cultures of carrot (Daucus carota L.) are serially subcultured. Ann. Bot. 38:223-241. Smyth, D.R. 1982. Proplastid DNA synthesis at pachytene in pollen mother cells of Lilium henryi. ]. Cell Sci. 56:293-302. Smyth, D.R., and Shaw, T. 1979. Cytoplasmic DNA synthesis at meiotic prophase in Lilium henryi. Aust. ]. Bot. 27:273-284. Smyth, D.R., and Stern, H. 1973. Repeated DNA synthesized during pachytene in Lilium henryi. Nature New Biol. 245:94-96. So, J.-S., and Larkins, B.A. 1991. Binding of an endosperm-specific nuclear protein to a maize betazein gene correlates with zein transcriptional activity. Plant Mol. Biol. 17:309-319. Soave, C, and Salamini, F. 1984. The role of structural and regulatory genes in the development of maize endosperm. Devel Genet. 5:1-25. Soave, C, Tardini, L., di Fonzo, N., and Salamini, F. 1981. Zein level in maize endosperm depends on a protein under control of the opaque-2 and opaque-6 loci. Cell 27:403-410. embryo of Crambe abyssinka. Planta 119:125-142. Sodmergen, Chen, G.H., Hu, Z.M., Guo, F.L., and Guan, Smith, C.R., Saunders, J.A., van Wert, S., Cheng, }., and X.L. 1995. Male gametophyte development in Matthews, B.F. 1994. Expression of GUS and CAT Plumbago zeylanica: cytoplasm localization and cell activities using electrotransformed pollen. Plant Sci. determination in the early generative cell. 104:49-58. Protoplasma 186:79-86. Smith, D.L. 1973. Nucleic acid, protein, and starch syn- Sodmergen, Luo, Y.Y., Kuroiwa, T., and Hu, S.Y. 1994. thesis in developing cotyledons of Pisum arvense L. Cytoplasmic DNA apportionment and plastid differAnn. Bot. 37:795-804. entiation during male gametophyte development in Smith, D.L., and Krikorian, A.D. 1988. Production of Pelargonium zonale. Sex. Plant Reprod. 7:51-56. somatic embryos from carrot tissues in hormone-free Sodmergen, Suzuki, T., Kawano, S., Nakamura, S., Tano, medium. Plant Sci. 58:103-110. S., and Kuroiwa, T. 1992. Behavior of organelle nuclei (nucleoids) in generative and vegetative cells Smith, D.L., and Krikorian, A.D. 1989. Release of somatic during maturation of pollen in Lilium longiflorum embryogenic potential from excised zygotic embryos of carrot and maintenance of proembryonic and Pelargonium zonale. Protoplasma 168:73-81. cultures in hormone-free medium. Am. ]. Bot. Somers, D.A., Rines, H.W., Gu, W., Kaeppler, H.F., and 76:1832-1843. Bushnell, W.R. 1992. Fertile, transgenic oat plants. Biotechnology 10:1589-1594. Smith, D.L., and Krikorian, A.D. 1990. Somatic embryogenesis of carrot in hormone-free medium: Somers, D.E., and Quail, P.H. 1995. Temporal and spatial external pH control over morphogenesis. Am. ]. Bot. expression patterns of PHYA and PHYB gene in 77:1634-1647. Arabidopsis. Plant J. 7:413-427. Smith, J.A., Krauss, M.R., Borkird, C, and Sung, Z.R. Sommer, H., Beltrdn, J.-P., Huijser, P., Pape, H., Lonnig, 1988. A nuclear protein associated with cell divisions W.-E., Saedler, H., and Schwarz-Sommer, Z. 1990. in plants. Planta 174:462-472. Deficiens, a homeotic gene involved in the control of Smith, J.G. 1973. Embryo development in Phaseolus vulflower morphogenesis in Antirrhinum majus: the progaris. II. Analysis of selected inorganic ions, ammotein shows homology to transcription factors. EMBO nia, organic acids, amino acids, and sugars in the /. 9:605-613. endosperm liquid. Plant Physiol. 51:454-458. Sommer, H., Nacken, W., Beltran, P., Huijser, P., Pape, H., Smith, L.G., Jackson, D., and Hake, S. 1995. Expression of Hansen, R., Flor, P., Saedler, H., and Schwarzknottedl marks shoot meristem formation during Sommer, Z. 1991. Properties of deficiens, a homeotic maize embryogenesis. Devel Genet. 16:344-348. gene involved in the control of flower morphogenesis Smith, L.M., Handley, }., Li, Y, Martin, H., Donovan, L., in Antirrhinum majus. Development Suppl. 1:169-175. and Bowles, D.J. 1992. Temporal and spatial regula- Sondheimer, E., and Linskens, H.F. 1974. Control of in
642
References
vitro germination and tube extension of Petunia Southworth, D., and Dickinson, D.B. 1975. |3-l,3-glucan hybrida pollen. Proc. Koninkl. Nederl. Akad. Wetensch. synthase from Lilium longiflorum pollen. Plant 77G116-124. Physiol. 56:83-87. Song,J., and Hedgcoth, C. 1994. A chimeric gene (orft.56) Southworth, D., and Dickinson, D.B. 1981. Ultrastructural is expressed as protein only in cytoplasmic malechanges in germinating lily pollen. Grana 20:29-35. sterile lines of wheat. Plant Mol. Biol. 26:535-539. Southworth, D., and Jernstedt, J.A. 1995. Pollen exine development precedes microtubule rearrangement Sood, P.P., Malik, C.P., and Tewari, H.B. 1968. A histoin Vigna unguiculata (Fabaceae): a model for pollen chemical study of the localization of succinicdehywall patterning. Protoplasma 187:79-87. drogenase in the germinating pollen grains of Portulaca grandiflora. Z. Biol. 116:215-220. Southworth, D., and Knox, R.B. 1989. Flowering plant sperm cells: isolation from pollen of Gerbera jamesonii Sood, R., Prabha, K., and Gupta, S.C. 1982. Is the Rejection (Asteraceae). Plant Sci. 60:273-277. reaction' inducing ability in sporophytic self-incompatibility systems restricted only to pollen and tape- Southworth, D., and Morningstar, P.A. 1992. Isolation of turn? Theor. Appl. Genet. 63:27-32. generative cells from pollen of Phoenix dactylifera. Sex. Plant Reprod. 5:270-274. Sood, R., Prabha, K., and Gupta, S.C. 1983. Overcoming self-incompatibility in Brassica campestris cv. Pusa Southworth, D., Platt-Aloia, K.A., and Thomson, W.W. Kalyani by bud and delayed pollinations. Beitr. Biol. 1988. Freeze fracture of sperm and vegetative cells in Pftanz. 58:59-65. Zea mays pollen. /. Ultrastr. Mol. Str. Res. 101:165-172. Sopory, S.K., and Maheshwari, S.C. 1976. Development of Southworth, D., Salvatici, P., and Cresti, M. 1994. Freeze fracture of membranes at the interface between vegpollen embryoids in anther cultures of Datura etative and generative cells in Amaryllis pollen. Int. J. innoxia. II. Effects of growth hormones. /. Expt. Bot. Plant Sci. 155:538-544. 27:58-68. Sorensen, M.B. 1992. Methylation of B-hordein genes in Southworth, D., Platt-Aloia, K.A., DeMason, D.A., and Thomson, W.W. 1989. Freeze-fracture of the generabarley endosperm is inversely correlated with gene tive cell of Phoenix dactylifera (Arecaceae). Sex. Plant activity and affected by the regulatory gene Lys3. Reprod. 2:270-276. Proc. Natl Acad. Sci. USA 89:4119-4123. Sorensen, M.B., Cameron-Mills, V, and Brandt, A. 1989. Southworth, D., Singh, M.B., Hough, T., Smart, I.J., Taylor, P., and Knox, R.B. 1988. Antibodies to pollen exines. Transcriptional and post-transcriptional regulation Planta 176:482-487. of gene expression in developing barley endosperm. Mol. Gen. Genet. 217:195-201. Souvr6, A., and Albertini, L. 1982. Etude des modifications de l'ultrastructure du tapis plasmodial du Sosnikhina, S.P., Fedotova, Y.S., Smirnov, V.G., Rhoeo discolor Hance au cours du developpement de Mikhailova, E.I., Kolomiets, O.L., and Bogdanov, Y.F. l'anthere, en relation avec les donnSes cytochim1992. Meiotic mutants of rye Secale cereale L. I. iques et autoradiographiques. Rev. Cytol. Biol. Veget.Synaptic mutant sy-1. Theor. Appl. Genet. 84:979-985. Bot. 5:151-169. Sossountzov, L., Ruiz-Avila, L., Vignols, F., Jolliot, A., Arondel, V., Tchang, F., Grosbois, M., Guerbette, F., Soyfer, V.N. 1980. Hereditary variability of plants under the action of exogenous DNA. Theor. Appl. Genet. Miginiac, E., Delseny, M., Puigdomenech, P., and 58:225-235. Kader, J.-C. 1991. Spatial and temporal expression of a maize lipid transfer protein gene. Plant Cell Spangenberg, G., Wang, Z.Y., Wu, X.L., Nagel, J., Iglesias, 3:923-933. V.A., and Potrykus, I. 1995a. Transgenic tall fescue (Festuca arundinacea) and red fescue (F. rubra) plants Soueges, R. 1919. Les premieres divisions de l'ceuf et les from microprojectile bombardment of embryogenic differenciations du suspenseur chez le Capsella bursasuspension cells. /. Plant Physiol. 145:693-701. pastoris Moench (1). Ann. Sci. Nat. Bot. Ser. X 1:1-28. Soueges, R. 1924. Developpement de l'embryon chez le Spangenberg, G., Wang, Z.-Y, Wu, X., Nagel, J., and Potrykus, I. 1995b. Transgenic perennial ryegrass Sagina procumbens L. Bull. Soc. Bot. Fr. 71:590-614. (Lolium perenne) plants from microprojectile bomSoueges, R. 1936. Embryog^nie des Saxifragac^es. bardment of embryogenic suspension cells. Plant Sci. Developpement de l'embryon chez le Saxifraga gran108:209-217. ulata L. Compt. Rend. Acad. Sci. Paris 202:240-242. Souer, E., van Houwelingen, A., Kloos, D., Mol, J., and Spanjers, A.W. 1978. Voltage variation in Lilium longifloKoes, R. 1996. The no apical meristem gene of Petunia rum pistils induced by pollination. Experientia is required for pattern formation in embryos and 34:36-37. flowers and is expressed at meristem and primordia Sparrow, A.H., Moses, M.J., and Steele, R. 1952. A cytoboundaries. Cell 85:159-170. logical and cytochemical approach to an understanding of radiation damage in dividing cells. Brit. Southworth, D. 1971. Incorporation of radioactive precurJ. Radiol. 25:182-188. sors into developing pollen walls. In Pollen: Development and Physiology, ed. J. Heslop-Harrison, Sparrow, A.H., Pond, V., and Kojan, S. 1955. pp. 115-120. London: Butterworth & Co. Microsporogenesis in excised anthers of Trillium erectum grown on sterile media. Am. J. Bot. 42:384-394. Southworth, D. 1975. Lectins stimulate pollen germinaSpassova, M., Moneger, F., Leaver, C.J., Petrov, P., tion. Nature 258:600-602. Atanassov, A., Nijkamp, H.J.J., and Hille, J. 1994. Southworth, D. 1983. Exine development in Gerbera Characterisation and expression of the mitochondrjamesonii (Asteraceae: Mutisieae). Am. J. Bot. ial genome of a new type of cytoplasmic male-sterile 70:1038-1047. sunflower. Plant Mol. Biol. 26:1819-1831. Southworth, D. 1988. Isolation of exines from gymSpena, A., and Schell, J. 1987. The expression of a heatnosperm pollen. Am. J. Bot. 75:15-21. inducible chimeric gene in transgenic tobacco plants. Southworth, D. 1990. Membranes of sperm and vegetaMol. Gen. Genet. 206:436^40. tive cells in pollen of Gerbera jamesonii. ]. Str. Biol. 103:97-103. Spena, A., Viotti, A., and Pirrotta, V. 1982. A homologous
References
643
repetitive block structure underlies the heterogene- Staiger, C.J., and Cande, W.Z. 1991. Microfilament distribution in maize meiotic mutants correlates with ity of heavy and light chain zein genes. EMBO /. microtubule organization. Plant Cell 3:637-644. 1:1589-1594. Spena, A., Estruch, J.J., Prinsen, Ev Nacken, W., van Staiger, C.J., Goodbody, K.C., Hussey, P.J., Valenta, R., Drabak, B.K., and Lloyd, C.W. 1993. The profilin Onckelen, H v and Sommer, H. 1992. Anther-specific multigene family of maize: differential expression of expression of the rolB gene of Agrobaderium rhizothree isoforms. Plant ]. 4:631-641. genes increases IAA content in anthers and alters anther development and whole flower growth. Staiger, D., and Apel, K. 1993. Molecular characterization Theor. Appl. Genet. 84:520-527. of two cDNAs from Sinapis alba L. expressed specifically at an early stage of tapetum development. Plant Spencer, ME., and Hodge, R. 1991. Cloning and sequencJ. 4:697-703. ing of the cDNA encoding the major albumin of Theobroma cacao. Identification of the protein as a Staiger, D., Kappeler, S., Muller, M., and Apel, K. 1994. member of the Kunitz protease inhibitor family. The proteins encoded by two tapetum-specific tranPlanta 183:528-535. scripts, Satap35 and S«tap44, from Sinapis alba L. are localized in the exine cell wall layer of developing Spencer, M.E., and Hodge, R. 1992. Cloning and sequencmicrospores. Planta 192:221-231. ing of a cDNA encoding the major storage proteins of Theobroma cacao. Identification of the proteins as Stalberg, K., Ellerstrom, M., Josefsson, L.-G., and Rask, L. 1993. Deletion analysis of a 2S seed storage protein members of the vicilin class of storage proteins. Planta 186:567-576. promoter of Brassica napus in transgenic tobacco. Plant Mol. Biol. 23:671-683. Speranza, A., and Calzoni, G.L. 1988. In vitro test of selfincompatibility in Mains domestica. Sex. Plant Reprod.Stanley, R.G., and Linskens, H.F. 1964. Enzyme activation 1:223-227. in germinating Petunia pollen. Nature 203:542-544. Speranza, A., and Calzoni, G.L. 1990. In vitro expression Stanley, R.G., and Linskens, H.F. 1965. Protein diffusion from germinating pollen. Physiol. Plant. 18:47-53. of self-incompatibility in Malus domestica mediated by stylar glycoproteins. Plant Physiol. Biochem. Stanley, R.G., and Linskens, H.F. 1967. Oxygen tension as a control mechanism in pollen tube rupture. Science 28:747-754. 157:833-834. Speranza, A., and Calzoni, G.L. 1992. Inhibition of pollentube growth and protein secretion by the monova- Stanley, R.G., and Linskens, H.F. 1974. Pollen: Biology, lent ionophore monensin. Sex. Plant Reprod. Biochemistry, Management. Berlin: Springer-Verlag. 5:232-238. Stapleton, A.E., and Bedinger, P.A. 1992. Immature maize Springer, B., Werr, W., Starlinger, P., Bennett, D.C., spikelets develop and produce pollen in culture. Zakolica, M., and Freeling, M. 1986. The shrunken Plant Cell Rep. 11:248-252. gene on chromosome 9 of Zea mays L. is expressed in Staswick, P., and Chrispeels, M.J. 1984. Expression of various plant tissues and encodes an anaerobic prolectin genes during seed development in normal and tein. Mol. Gen. Genet. 205:461^468. phytohemagglutinin-deficient cultivars of Phaseolus vulgaris. J. Mol. Appl. Genet. 2:525-535. Springer, P.S., McCombie, W.R., Sundaresan, V., and Martienssen, R.A. 1995. Gene trap tagging of PRO- Staswick, P.E., Hermodson, M.A., and Nielsen, N.C. 1984. LIFERA, an essential MCM2-3-5-like gene in The amino acid sequence of the A2Bla subunit of Arabidopsis. Science 268:877-880. glycinin. /. Biol. Chem. 259:13424-13430. Sree Ramulu, K., Bredemeijer, G.M.M., Dijkhuis, P., de Stauffer, C, Benito Moreno, R.M., and Heberle-Bors, E. Nettancourt, D., and Schibilla, H. 1979. Mentor 1991. Seed set after pollination with in-vitropollen effects on gametophytic incompatibility in matured, isolated pollen of Triticum aestivum. Theor. Nicotiana, Oenothera and Lycopersicum. Theor. Appl. Appl. Genet. 81:576-580. Genet. 54:215-218. Stax£n, I., Klimaszewska, K., and Bornman, C.H. 1994. Microtubular organization in protoplasts and cells of Sree Ramulu, K., Devreux, M., Ancora, G., and Laneri, U. somatic embryo-regenerating and non-regenerating 1976. Chimerism in Lycopersicum peruvianum plants cultures of Larix. Physiol. Plant. 91:680-686. regenerated from in vitro cultures of anthers and Stead, A.D., Roberts, I.N., and Dickinson, H.G. 1979. stem internodes. Z. Pflanzenziichtg 76:299-319. Pollen-pistil interaction in Brassica oleracea. Events Srivastava, H.K., Sarkissian, I.V., and Shands, H.L. 1969. prior to pollen germination. Planta 146:211-216. Mitochondrial complementation and cytoplasmic Stead, A.D., Roberts, I.N., and Dickinson, H.G. 1980. male sterility in wheat. Genetics 63:611-618. Pollen-stigma interaction in Brassica oleracea: the role Stacey, N.J., Roberts, K., and Knox, J.P. 1990. Patterns of of stigmatic proteins in pollen grain adhesion. /. Cell expression of the JIM4 arabinogalactan-protein epiSci. 42:417-423. tope in cell cultures and during somatic embryogenSteer, M.W. 1977a. Differentiation of the tapetum in Avena. esis in Daucus carota L. Planta 180:285-292. I. The cell surface. /. Cell Sci. 25:125-138. Staff, I.A., Taylor, P., Kenrick, J., and Knox, R.B. 1989. Ultrastructural analysis of plastids in angiosperm Steer, M.W. 1977b. Differentiation of the tapetum in Avena. II. The endoplasmic reticulum and Golgi pollen tubes. Sex. Plant Reprod. 2:70-76. apparatus. /. Cell Sci. 28:71-86. Staff, I.A., Taylor, P.E., Smith, P., Singh, M.B., and Knox, R.B. 1990. Cellular localization of water soluble, Steer, M.W., and Steer, J.M. 1989. Pollen tube tip growth. allergenic proteins in rye-grass (Lolium perenne) New Phytol. 111:323-358. pollen using monoclonal and specific IgE antibodies Stefaniak, B. 1984. Development of rye embryo. I. with immunogold probes. Histochem. J. 22:276-290. Preliminary investigations on ultrastructure of 7day-old proembryo. Ada Soc. Bot. Polon. 53:11-16. Staiger, C.J., and Cande, W.Z. 1990. Microtubule distribution in dv, a maize meiotic mutant defective in the Stefaniak, B. 1987. The in vitro development of isolated prophase to metaphase transition. Devel Biol. proembryos. Ada Soc. Bot. Polon. 56:37-^2. 138:231-242. Steffen, K. 1963. Male gametophyte. In Recent Advances in
644
References
the Embryology of Angiosperms, ed. P. Maheshwari, Stevens, V.A.M., and Murray, B.G. 1982. Studies on hetpp. 15-40. Delhi: International Society of Plant eromorphic self-incompatibility systems: physiologMorphologists. ical aspects of the incompatibility system of Primula obconica. Theor. Appl. Genet. 61:245-256. Steffensen, D.M. 1966. Synthesis of ribosomal RNA during growth and division in Lilium. Expt. Cell Res. Steward, EC. 1963. The control of growth in plant cells. 44:1-12. Sci. Am. 209 (4):104-113. Stein, J.C., and Nasrallah, J.B. 1993. A plant receptor-like Steward, EC, Mapes, M.O., and Mears, K. 1958. Growth gene, the S-Iocus receptor kinase of Brassica oleracea and organized development of cultured cells. II. L, encodes a functional serine/threonine kinase. Organization in cultures grown from freely susPlant Physiol. 101:1103-1106. pended cells. Am. J. Bot. 45:705-708. Stein, J.C., Dixit, R., Nasrallah, M.E., and Nasrallah, J.B. Steward, EC, Mapes, M.O., and Smith, J. 1958. Growth 1996. SRK, the stigma-specific S locus receptor and organized development of cultured cells. I. kinase of Brassica, is targeted to the plasma memGrowth and division of freely suspended cells. Am. brane in transgenic tobacco. Plant Cell 8:429-445. J. Bot. 45:693-703. Stein, J.C., Howlett, B., Boyes, D.C., Nasrallah, M.E., and Steward, EC, Mapes, M.O., Kent, A.E., and Holsten, R.D. 1964. Growth and development of cultured plant Nasrallah, J.B. 1991. Molecular cloning of a putative cells. Science 143:20-27. receptor protein kinase gene encoded at the selfincompatibility locus of Brassica oleracea. Proc. NatlStewart, J.M., and Hsu, CL. 1977. In ovulo embryo culture Acad. Sci. USA 88:8816-8820. and seedling development of cotton (Gossypium hirStejskal, J., and Griga, M. 1995. Comparative analysis of sutum L.). Planta 137:113-117. some isozymes and proteins in somatic and zygotic Stickland, R.G., and Wilson, K.E. 1983. Sugars and starch embryos of soybean (Glycine max [L.] Merr.). /. Plant in developing round and wrinkled pea seeds. Ann. Physiol. 146:497-502. Bot. 52:919-921. Stelly, D.M., and Palmer, R.G. 1982. Variable development Stieglitz, H. 1977. Role of P-l,3-glucanase in post-meiotic in anthers of partially male-sterile soybeans. /. Hered. microspore release. Devel Biol. 57:87-97. 73:101-108. Stieglitz, H., and Stern, H. 1973. Regulation of P-l,3-gluStenram, U., Heneen, W.K., and Skerritt, J.H. 1991. canase activity in developing anthers of Lilium. Devel Immunocytochemical localization of wheat storage Biol. 34:169-173. proteins in endosperm cells 30 days after anthesis. /. Stinissen, H.M., Peumans, W.J., and Carlier, A.R. 1983. Expt. Bot. 42:1347-1355. Occurrence and immunological relationships of lectins in gramineous species. Planta 159:105-111. Sterk, P., Booij, H., Schellekens, G.A., van Kammen, A., and de Vries, S.C. 1991. Cell-specific expression of Stinissen, H.M., Peumans, W.J., and Chrispeels, M.J. 1984. the carrot EP2 lipid transfer protein gene. Plant Cell Subcellular site of lectin synthesis in developing rice 3:907-921. embryos. EMBO J. 3:1979-1985. Sterling, C. 1955. Embryogeny in the lima bean. Bull. Stinissen, H.M., Peumans, W.J., and de Langhe, E. 1984. Torrey Bot. Cl. 82:325-338. Abscisic acid promotes lectin biosynthesis in developing and germinating rice embryos. Plant Cell Rep. Stern H. 1956. Sulfhydryl groups and cell division. Science 124:1292-1293. 3:55-59. Stern, H. 1958. Variations in sulfhydryl concentration dur- Stinson, J., and Mascarenhas, J.P. 1985. Onset of alcohol dehydrogenase synthesis during microsporogenesis ing microsporocyte meiosis in anthers of Lilium and in maize. Plant Physiol. 77:222-224. Trillium. J. Biophys. Biochem. Cytol. 4:157-161. Stern, H. 1961. Periodic induction of deoxyribonuclease Stinson, J.R., Eisenberg, A.J., Willing, P.R., Pe, M.E., Hanson, D.D., and Mascarenhas, J.P. 1987. Genes activity in relation to the mitotic cycle. /. Biophys. expressed in the male gametophyte of flowering Biochem. Cytol. 9:271-277. plants and their isolation. Plant Physiol. 83:442-447. Stern, H., and Hotta, Y. 1977. Biochemistry of meiosis. Phil. Trans. Roy. Soc. Lond. 277B:277-294. Stirn, S., and Jacobsen, H.-J. 1987. Marker proteins for embryogenic differentiation patterns in pea callus. Stern, H., and Hotta, Y. 1984. Chromosome organization in the regulation of meiotic prophase. Symp. Soc. Plant Cell Rep. 6:50-54. Expt. Biol. 38:161-175. Stoger, E., Benito Moreno, R.M., Ylstra, B., Vicente, O., Stern, H., and Kirk, PL. 1948. The oxygen consumption of and Heberle-Bors, E. 1992. Comparison of different the microspores of Trillium in relation to the mitotic techniques for gene transfer into mature and immature tobacco pollen. Transgen. Res. 1:71-78. cycle. /. Gen. Physiol. 31:243-248. Stern, H., and Timonen, S. 1954. The position of the cell Stoger, E., Fink, C, Pfosser, M., and Heberle-Bors, E. 1995. Plant transformation by particle bombardment of nucleus in the pathways of hydrogen transfer: embryogenic pollen. Plant Cell Rep. 14:273-278. cytochrome C, flavoproteins, glutathione, and ascorStorchova, H., Capkova V, and Tupy, J. 1994. A Nicotiana bic acid. /. Gen. Physiol. 38:41-52. tabacum mRNA encoding a 69-kDa glycoprotein Stern H., Westergaard, M., and von Wettstein, D. 1975. occurring abundantly in pollen tubes is transcribed Presynaptic events in meiocytes of Lilium longiflorum but not translated during pollen development in the and their relation to crossing-over: a preselection anthers. Planta 192:441^45. hypothesis. Proc. Natl Acad. Sci. USA 72:961-965. Stettler, R.F., Koster, R., and Steenackers, V. 1980. Stout, A.B. 1931. Pollen-tube behavior in Brassica pekinensis with reference to self-incompatibility in fertilizaInterspecific crossability studies in poplars. Theor. tion. Am. ]. Bot. 18:686-695. Appl. Genet. 58:273-282. Stevens, V.A.M., and Murray, B.G. 1981. Studies on het- Stout, A.B., and Chandler, C. 1933. Pollen-tube behavior in Hemerocallis with special reference to incompatieromorphic self-incompatibility systems: the cytobilities. Bull. Torrey Bot. Cl. 60:397-416. chemistry and ultrastructure of the tapetum of Stout, A.B., and Chandler, C. 1941. Change from selfPrimula obconica. J. Cell Sci. 50:419-431.
References
645
sion and differentiation of pollen grains in Triticale incompatibility to self-compatibility accompanying anthers cultured in vitro. Sci. Sinica 17:47-54. change from diploidy to tetraploidy. Science 94:118. Strasburger, E. 1884. Neue Untersuchungen tiber den Sun, M., and Ganders, F.R. 1987. Microsporogenesis in Befruchtungsvorgang bei den Phanerogamen als Grnndlagemale-sterile and hermaphroditic plants of nine gynodioecious taxa of Hawaiian Bidens (Asteraceae). fur eine Theorie der Zeugung. Jena: Gustav Fischer. Am. J. Bot. 74:209-217. Street, H.E., and Withers, L.A. 1974. The anatomy of embryogenesis. In Tissue Culture and Plant Science Sun, S.M., Slightom, J.L., and Hall, T.C. 1981. Intervening sequences in a plant gene - comparison of the partial 1974, ed. H.E. Street, pp. 71-100. London: Academic sequence of cDNA and genomic DNA of French Press. bean phaseolin. Nature 289:37-41. Strokov, A.A., Bogdanov, Y.F., and Reznickova, S.A. 1973. A quantitative study of histones of meiocytes. II. Sun, S.M., Mutschler, M.A., Bliss, F.A., and Hall, T.C. 1978. Protein synthesis and accumulation in bean cotylePolyacrylamide gel electrophoresis of isolated hisdons during growth. Plant Physiol. 61:918-923. tones from Lilium microsporocytes. Chromosoma 43:247-260. Sunderland, N. 1971. Anther culture: a progress report. Stuart, D.A., and Strickland, S.G. 1984a. Somatic embryoSci. Prog. Oxford 59:527-549. genesis from cell cultures of Medicago saliva L. I. The Sunderland, N., Collins, G.B., and Dunwell, J.M. 1974. role of amino acid additions to the regeneration The role of nuclear fusion in pollen embryogenesis medium. Plant Sci. Lett. 34:165-174. of Datura innoxia Mill. Planta 117:227-241. Stuart, D.A., and Strickland, S.G. 1984b. Somatic embryo- Sunderland, N., and Dunwell, J.M. 1977. Anther and genesis from cell cultures of Medicago sativa L. II. The pollen culture. In Plant Tissue and Cell Culture, 2nd interaction of amino acids with ammonium. Plant ed., ed. H.E. Street, pp. 223-265. Berkeley: University Sci. Lett. 34:175-181. of California Press. Sturaro, M, Vernieri, P., Castiglioni, P., Binelli, G., and Sunderland, N., and Evans, L.J. 1980. Multicellular pollen Gavazzi, G. 1996. The rea (red embryonic axis) pheformation in cultured barley anthers. II. The A, B, and C pathways. /. Expt. Bot. 31:501-514. notype describes a new mutation affecting the response of maize embryos to abscisic acid and Sunderland, N., and Roberts, M. 1977. New approach to osmotic stress. /. Expt. Bot. 47:717-728. pollen culture. Nature 270:236-238. Sturm, A., Voelker, T.A., Herman, E.M., and Chrispeels, Sunderland, N., and Roberts, M. 1979. Cold-pretreatment MJ. 1988. Correct glycosylation, Golgi-processing, of excised flower buds in float culture of tobacco and targeting to protein bodies of the vacuolar proanthers. Ann. Bot. 43:405-414. tein phytohemagglutinin in transgenic tobacco. Sunderland, N., and Wicks, F.M. 1971. Embryoid formaPlanta 175:170-183. tion in pollen grains of Nicotiana tabacum. ]. Expt. Bot. 22:213-226. Sugimoto, T, Momma, M., Hashizume, K., and Saio, K. 1987. Components of storage protein in hypocotyl- Sunderland, N., and Wildon, D.C. 1979. A note on the preradicle axis of soybean (Glycine max cv. Enrei) seeds. treatment of excised flower buds in float culture of Agr. Biol. Chem. 51:1231-1238. Hyoscyamus anthers. Plant Sci. Lett. 15:169-175. Sugiyama, T., Rafalski, A., and Soil, D. 1986. The Sunderland, N., Roberts, M., Evans, L.J., and Wildon, D.C. nucleotide sequence of a wheat y-gliadin genomic 1979. Multicellular pollen formation in cultured barclone. Plant Sci. 44:205-209. ley anthers. I. Independent division of the generative and vegetative cells. /. Expt. Bot. 30:1133-1144. Sugiyama, T, Rafalski, A., Peterson, D., and Soil, D. 1985. A wheat HMW glutenin subunit gene reveals a Sung, Z.R. 1979. Relationship of indole-3-acetic acid and tryptophan concentrations in normal and 5-methylhighly repeated structure. Nucl. Acids Res. tryptophan-resistant cell lines of wild carrots. Planta 13:8729-8737. 145:339-345. Sullivan, T.D., and Kaneko, Y. 1995. The maize brittlel gene encodes amyloplast membrane polypeptides. Sung, Z.R., Lazar, G.B., and Dudits, D. 1981. CycloPlanta 196:477-484. heximide resistance in carrot culture: a differentiated function. Plant Physiol. 68:261-264. Sullivan, T.D., Strelow, L.I., Illingworth, C.A., Phillips, R.L., and Nelson, O.E., Jr. 1991. Analysis of maize Sung, Z.R., and Okimoto, R. 1981. Embryonic proteins in somatic embryos of carrot. Proc. Natl Acad. Sci. USA brittle-1 alleles and a defective suppressor-mutatorinduced mutable allele. Plant Cell 3:1337-1348. 78:3683-3687. Sumner, M.J. 1992. Embryology of Brassica campestris: the Sung, Z.R., and Okimoto, R. 1983. Coordinate gene entrance and discharge of the pollen tube in the synexpression during somatic embryogenesis in carrots. Proc. Natl Acad. Sci. USA 80:2661-2665. ergid and the formation of the zygote. Can. J. Bot. 70:1577-1590. Sung, Z.R., Smith, R., and Horowitz, J. 1979. Quantitative Sumner, M.J., and van Caeseele, L. 1989. The ultrastrucstudies of embryogenesis in normal and 5-methylture and cytochemistry of the egg apparatus of tryptophan-resistant cell lines of wild carrot. Planta Brassica campestris. Can. J. Bot. 67:177-190. 147:236-240. Sumner, M.J., and van Caeseele, L. 1990. The develop- Sung, Z.R., Belachew, A., Shunong, B., and Bertrandment of the central cell of Brassica campestris prior to Garcia, R. 1992. EMF, an Arabidopsis gene required fertilization. Can. ]. Bot. 68:2553-2563. for vegetative shoot development. Science Sumner-Smith, M., Rafalski, J.A., Sugiyama, T., Stoll, M., 258:1645-1647. and Soil, D. 1985. Conservation and variability of Suss, J., and Tupy, J. 1976. On the nature of RNA synthesized in pollen tubes of Nicotiana alata. Biol. Plant. wheat a/ P-gliadin genes. Nucl. Acids Res. 13:3905-3916. Sun, C.-S. 1978. Androgenesis of cereal crops. In 18:140-146. Proceedings of Symposium on Plant Tissue Culture, pp. Suss, J., and Tupy, J. 1978. tRNA synthesis in germinating pollen. Biol. Plant. 20:70-72. 117-123. Peking: Science Press. Sun, C.-S., Wang, C.-C, and Chu, C.-C. 1974. Cell divi- Suss, J., and Tupy, J. 1979. Poly(A)+RNA synthesis in ger-
646
References
minating pollen of Nicotiana tabacum L. Biol. Plant. O., Heberle-Bors, E., Ahorn, H., and Breitenbach, M. 21:365-371. 1995b. Isoforms of Bet v 1, the major birch pollen allergen, analyzed by liquid chromatography, mass Suss, J., and Tupy, J. 1982. Kinetics of uridine uptake and spectrometry, and cDNA cloning. /. Biol. Chem. incorporation into RNA in tobacco pollen culture. Biol. Plant. 24:72-79. 270:2607-2613. Sussex, I. 1975. Growth and metabolism of the embryo Syamasundar, J., and Panchaksharappa, M.G. 1975. A and attached seedling of the viviparous mangrove, cytochemical study of the hypertrophied synergid in Rhizophora mangle. Am. J. Bot. 62:948-953. Allium cepa L. Cytologia 40:371-376. Sussex, I., Clutter, M., Walbot, V., and Brady, T. 1973. Syamasundar, J., and Panchaksharappa, M.G. 1976. A hisBiosynthetic activity of the suspensor of Phaseolus tochemical study of some post-fertilization developcoccineus. Caryologia 25: Suppl. 261-272. mental stages in Dipcadi montanum Dalz. Cytologia 41:123-130. Sussmuth, J., Dressier, K., and Hess, D. 1991. Agrobacterium-mediated transfer of the GUS gene Szakacs, E., and Barnabas, B. 1988. Cytological aspects of into pollen of Petunia. Bot. Ada 104:72-76. in vitro androgenesis in wheat (Triticum aestivum L.) using fluorescent microscopy. Sex. Plant Reprod. Suzuki, G., Watanabe, M., Toriyama, K., Isogai, A., and 1:217-222. Hinata, K. 1995. Molecular cloning of members of the S-multigene family in self-incompatible Brasska Szakacs, E., and Barnabas, B. 1995. The effect of colchicine campestris L. Plant Cell Physiol. 36:1273-1280. treatment on microspore division and microsporederived embryo differentiation in wheat (Triticum Suzuki, M., Koide, Y., Hattori, T., Nakamura, K., and aestivum L.) anther culture. Euphytica 83:209-213. Asahi, T. 1995. Different sets of cis-elements contribute to the expression of a catalase gene from cas- Szymkowiak, E.J., and Sussex, I.M. 1992. The internal tor bean during seed formation and postembryonic meristem layer (L3) determines floral meristem size development in transgenic tobacco. Plant Cell and carpel number in tomato periclinal chimeras. Physiol. 36:1067-1074. Plant Cell 4:1089-1100. Suzuki, T., Nakamura, C., Mori, N., and Kaneda, C. 1995. Overexpression of mitochondrial genes in alloplas- Tabata, S., Sato, S., Watanabe, Y., Yamamoto, M., and mic common wheat with a cytoplasm of wheatgrass Hotta, Y. 1993. Evidence of meiosis-specific regula(Agropyron trichophorum) showing depressed vigor tion of gene expression in lily microsporocytes. Plant and male sterility. Plant Mol. Biol. 27:553-565. Sci. 89:31-41. Swain, E., Li, C.P., and Poulton, J.E. 1992a. Development Tagliasacchi, A.M., Forino, L.M.C., Cionini, P.G., of the potential for cyanogenesis in maturing black Cavallini, A., Durante, M., Cremonini, R., and cherry (Primus serotina Ehrh.) fruits. Plant Physiol. Avanzi, S. 1984. Different structure of polytene chro98:1423-1428. mosome of Phaseolus coccineus suspensors during early embryogenesis. 3. Chromosome pair VI. Swain, E., Li, C.P., and Poulton, J.E. 1992b. Tissue and subProtoplasma 122:98-107. cellular localization of enzymes catabolizing (R)amygdalin in mature Prunus serotina seeds. Plant Takahashi, M. 1987. Development of omniaperturate Physiol. 100:291-300. pollen in Trillium kamtschaticum (Liliaceae). Am. ]. Swaminathan, M.S., Chopra, V.L., and Bhaskaran, S. 1962. Bot. 74:1842-1852. Cytological aberrations observed in barley embryos Takahashi, M. 1989a. Development of the echinate pollen cultured in irradiated potato mash. Radiation Res. wall in Farfugium japonicum (Compositae: 16:182-188. Senecioneae). Bot. Mag. Tokyo 102:219-234. Swamy, B.G.L. 1979. Embryogenesis in Cheirostylis flabel- Takahashi, M. 1989b. Pattern determination of the exine in Caesalpinia japonica (Leguminosae: Caesallata. Phytomorphology 29:199-203. pinioideae). Am. J. Bot. 76:1615-1626. Swamy, B.G.L. 1980. Embryogenesis in Sagittaria sagittaeTakahashi, M. 1993. Exine initiation and substructure in folia. Phytomorphology 30:204-212. pollen of Caesalpinia japonica (Leguminosae: Swamy, B.G.L., and Ganapathy, P.M. 1957. A new type of Caesalpinioideae). Am. J. Bot. 80:192-197. endosperm haustorium in Nothapodytes foetida. Phytomorphology 7:331-336. Takahashi, M. 1994. Pollen development in a submerged plant, Ottelia alismoides (L.) Pers. (Hydrocharitaceae). Swamy, B.G.L., and Lakshmanan, K.K. 1962a. The origin /. Plant Res. 107:161-164. of epicotylary meristem and cotyledon in Halophila ovata Gaudich. Ann. Bot. 26:243-249. Takahashi, M. 1995. Three-dimensional aspects of exine initiation and development in Lilium longiflorum Swamy, B.G.L., and Lakshmanan, K.K. 1962b. (Liliaceae). Am. J. Bot. 82:847-854. Contributions to the embryology of the Najadaceae. /. Indian Bot. Soc. 41:247-267. Takahashi, M., and Kouchi, J. 1988. Ontogenetic developSwegle, M., Kramer, K.J., and Muthukrishnan, S. 1992. ment of spinous exine in Hibiscus syriacus Properties of barley seed chitinases and release of (Malvaceae). Am. }. Bot. 75:1549-1558. embryo-associated isoforms during early stages of Takahashi, M., and Skvarla, J.J. 1991. Exine pattern forimbibition. Plant Physiol. 99:1009-1014. mation by plasma membrane in Bougainvillea spectabilis Willd. (Nyctaginaceae). Am. J. Bot. Swift, H. 1950. The constancy of desoxyribose nucleic acid 78:1063-1069. in plant nuclei. Proc. Natl Acad. Sci. USA 36:643-654. Swoboda, I., Dang, T.C.H., Heberle-Bors, E., and Vicente, Takahashi, T, Naito, S., and Komeda, Y. 1992. The O. 1995a. Expression of Bet c 1, the major birch Arabidopsis HSP18.2 promoter/GUS gene fusion in pollen allergen, during anther development. transgenic Arabidopsis plants: a powerful tool for the isolation of regulatory mutants of the heat-shock Protoplasma 187:103-110. response. Plant J. 2:751-761. Swoboda, I., Jilek, A., Ferreira, R, Engel, E., HoffmannSommergruber, K., Scheiner, O., Kraft, D., Takaiwa, F., Kikuchi, S., and Oono, K. 1986. The strucBreiteneder, H., Pittenauer, E., Schmid, E., Vicente, ture of rice storage protein glutelin precursor
References
647
Tanaka, 1.1988. Isolation of generative cells and their prodeduced from cDNA. FEBS Lett. 206:33-35. toplasts from pollen of Lilium longiflorum. Takaiwa, R, Kikuchi, S., and Oono, K. 1987. A rice glutelin Protoplasma 142:68-73. gene family - a major type of glutelin mRNAs can be divided into two classes. Mol. Gen. Genet. 208:15-22. Tanaka, 1.1991. Microtubule-determined plastid distribution during microsporogenesis in Lilium longiflorum. Takaiwa, R, Kikuchi, S., and Oono, K. 1989. The comJ. Cell Sci. 99:21-31. plete nucleotide sequence of new type cDNA coding for rice storage protein glutelin. Nucl. Acids Tanaka, I. 1993. Development of male gametes in flowerRes. 17:3289. ing plants. /. Plant Res. 106:55-63. Takaiwa, R, and Oono, K. 1991. Genomic DNA sequences Tanaka, I., and Ito, M. 1980. Induction of typical cell division in isolated microspores of Lilium longiflorum of two new genes for new storage protein glutelin in and Tulipa gesneriana. Plant Sci. Lett. 17:279-285. rice. Jap. ]. Genet. 66:161-171. Takaiwa, P., Oono, K., and Kato, A. 1991. Analysis of the Tanaka, I., and Ito, M. 1981a. Studies on microspore development in liliaceous plants. III. Pollen tube develop5' flanking region responsible for the endospermment in lily pollens cultured from the uninucleate specific expression of a rice glutelin chimeric gene in microspore stage. Plant Cell Physiol. 22:149-153. transgenic tobacco. Plant Mol. Biol. 16:49-58. Takaiwa, R, Ebinuma, H., Kikuchi, S., and Oono, K. 1987. Tanaka, I., and Ito, M. 1981b. Control of division patterns Nucleotide sequence of a rice glutelin gene. FEBS in explanted microspores of Tulipa gesneriana. Lett. 221:43-47. Protoplasma 108:329-340. Takaiwa, R, Oono, K., Wing, D., and Kato, A. 1991. Tanaka, I., Kitazume, C, and Ito, M. 1987. The isolation and culture of lily pollen protoplasts. Plant Sci. Sequence of three members and expression of a new 50:205-211. major subfamily of glutelin genes from rice. Plant Mol. Biol. 17:875-885. Tanaka, I., Nakamura, S., and Miki-Hirosige, H. 1989. Structural features of isolated generative cells and Takats, S.T. 1959. Chromatin extrusion and DNA transfer their protoplasts from pollen of some liliaceous during microsporogenesis. Chromosoma 10:430-453. plants. Gamete Res. 24:361-374. Takats, S.T. 1962. An attempt to detect utilization of DNA breakdown products from the tapetum for DNA Tanaka, I., Taguchi, T, and Ito, M. 1979. Studies on microspore development in liliaceous plants. I. The synthesis in the microspores of Lilium longiflorum. duration of the cell cycle and developmental aspects Am. ]. Bot. 49:748-758. in lily microspores. Bot. Mag. Tokyo 92:291-298. Takats, S.T. 1965. Non-random nuclear orientation during DNA synthesis in Tradescantia pollen grains. /. Cell Tanaka, I., Taguchi, T., and Ito, M. 1980. Studies on Biol. 25:151-157. microspore development in liliaceous plants. II. The behavior of explanted microspores of the lily, Takats, S.T, and Wever, G.H. 1971. DNA polymerase and Lilium longiflorum. Plant Cell Physiol. 21:667-676. DNA nuclease activities in S-competent and Sincompetent nuclei from Tradescantia pollen grains. Tanaka, I., and Wakabayashi, T. 1992. Organization of the Expt. Cell Res. 69:25-28. actin and microtubule cytoskeleton preceding pollen germination. An analysis using cultured Takayama, S., Isogai, A., Tsukamoto, C, Ueda, Y., Hinata, pollen protoplasts of Lilium longiflorum. Planta K., Okazaki, K., and Suzuki, A. 1987. Sequences of S186:473-482. glycoproteins, products of the Brassica campestris self-incompatibility locus. Nature 326:102-105. Tanaka, T., Nishihara, M., Seki, M., Sakamoto, A., Tanaka, K., Irifune, K., and Morikawa, H. 1995. Successful Takegami, M.H., and Ito, M. 1975. Studies on the behavior expression in pollen of various plant species of in of meiotic protoplasts. III. Peatures of nuclear and vitro synthesized mRNA introduced by particle cell division for lily protoplasts in the cell-wall-free bombardment. Plant Mol. Biol. 28:337-341. state. Bot. Mag. Tokyo 88:187-196. Takegami, M.H., Yoshioka, M., Tanaka, I., and Ito, M. Tang, W., Diaz, P., Wilson, K.A., and Tan-Wilson, A.L. 1993. Spatial distribution of protein accumulation in 1981. Characteristics of isolated microsporocytes soybean seed. Phytochemistry 33:1289-1295. from liliaceous plants for studies of the meiotic cell Tang, X., Hepler, P.K., and Scordilis, S.P. 1989. cycle in vitro. Plant Cell Physiol. 22:1-10. Immunochemical and immunocytochemical identiTakei, Y., Yamauchi, D., and Minamikawa, T. 1989. fication of a myosin heavy chain polypeptide in Nucleotide sequence of the canavalin gene from Nicotiana pollen tubes. /. Cell Sci. 92:569-574. Canavalia gladiata seeds. Nucl. Acids Res. 17:4381. Takeno, K., Koshioka, M., Pharis, R.P., Rajasekaran, K., Tang, X., Lancelle, S.A., and Hepler, P.K. 1989. and Mullins, M.G. 1983. Endogenous gibberellin-like Fluorescence microscopic localization of actin in substances in somatic embryos of grape (Vitis pollen tubes: comparison of actin antibody and phalvinifera x Vitis rupestris) in relation to embryogenesis loidin staining. Cell Mot. Cytoskel. 12:216-224. and the chilling requirement for subsequent devel- Tanksley, S.D., Zamir, D., and Rick, CM. 1981. Evidence opment of mature embryos. Plant Physiol. for extensive overlap of sporophytic and gameto73:803-808. phytic gene expression in Lycopersicon esculentum. Science 213:453-455. Talbot, D.R., Adang, M.J., Slightom, J.L., and Hall, T.C. 1984. Size and organization of a multigene family Tano, S., and Takahashi, H. 1964. Nucleic acid synthesis in encoding phaseolin, the major seed storage protein growing pollen tubes. /. Biochem. 56:578-580. of Phaseolus vulgaris L. Mol. Gen. Genet. 198:42^9. Tantikanjana, T., Nasrallah, M.E., and Nasrallah, J.B. 1996. Tamura, Y, Kitano, H., Satoh, H., and Nagato, Y. 1992. A The Brassica S gene family: molecular characterizagene profoundly affecting shoot organization in the tion of the SLR2 gene. Sex. Plant Reprod. 9:107-116. early phase of rice development. Plant Sci. 82:91-99. Tantikanjana, T, Nasrallah, M.E., Stein, J.G, Chen, C.-H., and Nasrallah, J.B. 1993. An alternative transcript of Tan, L.W., and Jackson, J.R 1988. Stigma proteins of the two loci self-incompatible grass Phalaris coerulescens. the S locus glycoprotein gene in a class II pollenrecessive self-incompatibility haplotype of Brassica Sex. Plant Reprod. 1:25-27.
648
References
oleracea encodes a membrane-anchored protein. Rhododendron: quantitative cytology, three-dimenPlant Cell 5:657-666. sional reconstruction, isolation and detection using fluorescent probes. Sex. Plant Reprod. 2:254-264. Tara, C.P., and Namboodiri, A.N. 1974. Aberrant microsporogenesis and sterility in Impatiens sultani Taylor, P.E., Kenrick, J., Blomstedt, C.K., and Knox, R.B. (Balsaminaceae). Am. ]. Bot. 61:585-591. 1991. Sperm cells of the pollen tubes of Brassica: Tarasenko, L.V., and Bannikova, V.P. 1991. Ultrastructure ultrastructure and isolation. Sex. Plant Reprod. of pollen grain development in some angiosperm 4:226-234. species. Phytomorphology 41:11-20. Tebbutt, S.J., and Lonsdale, D.M. 1995. Deletion analysis Tautorus, T.E., Wang, H., Fowke, L.C., and Dunstan, D.I. of a tobacco pollen-specific polygalacturonase. Sex. 1992. Microtubule pattern and the occurrence of prePlant Reprod. 8:242-246. prophase bands in embryogenic cultures of black Tebbutt, S.J., Rogers, H.J., and Lonsdale, D.M. 1994. spruce (Picea mariana Mill.) and non-embryogenic Characterization of a tobacco gene encoding a cultures of jack pine (Pinus banksiana Lamb.). Plant pollen-specific polygalacturonase. Plant Mol. Bib/. Cell Rep. 11:419-423. 25:283-297. Taylor, D.C., Barton, D.L., Rioux, K.P., MacKenzie, S.L., Telmer, C.A., Newcomb, W., and Simmonds, D.H. 1993. Reed, D.W., Underhill, E.W., Pomeroy, M.K., and Microspore development in Brassica napus and the Weber, N. 1992. Biosynthesis of acyl lipids containeffect of high temperature on division in vivo and in ing very-long chain fatty acids in microsporevitro. Protoplasma 172:154-165. derived and zygotic embryos of Brassica napus L. cv. Telmer, C.A., Newcomb, W., and Simmonds, D.H. 1995. Reston. Plant Physiol. 99:1609-1618. Cellular changes during heat shock induction and embryo development of cultured microspores of Taylor, D.C., Ferrie, A.M.R., Keller, W.A., Giblin, E.M., Pass, E.W., and MacKenzie, S.L. 1993. Bioassembly Brassica napus cv. Topas. Protoplasma 185:106-112. of acyl lipids in microspore-derived embryos of Temple, M., Makaroff, C.A., Mutschler, M.A., and Earle, Brassica campeslris L. Plant Cell Rep. 12:375-384. E.D. 1992. Novel mitochondrial genomes in Brassica Taylor, D.C., Weber, N., Barton, D.L., Underhill, E.W., napus somatic hybrids. Curr. Genet. 22:243-249. Hogge, L.R., Weselake, R.J., and Pomeroy, M.K. 1991. Templeman, T.S., DeMaggio, A.E., and Stetler, D.A. 1987. Triglycerol bioassembly in microspore-derived Biochemistry of fern spore germination: globulin embryos of Brassica napus L. cv. Reston. Plant Physiol. storage proteins in Matteuccia struthiopteris L. Plant 97:65-79. Physiol. 85:343-349. Taylor, D.C., Weber, N., Underhill, E.W., Pomeroy, M.K., Templeman, T.S., Stein, D.B., and DeMaggio, A.E. 1988. A Keller, W.A., Scowcroft, W.R., Wilen, R.W., Moloney, fern spore storage protein is genetically similar to M.M., and Holbrook, L.A. 1990. Storage-protein regthe 1.7 S seed storage protein of Brassica napus. ulation and lipid accumulation in microspore Biochem. Genet. 26:595-603. embryos of Brassica napus L. Planta 181:18-26. Tepfer, S.S. 1953. Floral anatomy and ontogeny in Taylor, J.H. 1950. The duration of differentiation in Aquilegia formosa var. truncata and Ranunculus repens. excised anthers. Am. J. Bot. 37:137-143. Univ. Calif. Publ. Bot. 25:513-647. Taylor, J.H. 1953. Autoradiographic detection of incorpo- Tepfer, S.S., Greyson, R.I., Craig, W.R., and Hindman, J.L. ration of P32 into chromosomes during meiosis and 1963. In vitro culture of floral buds of Aquilegia. Am. mitosis. Expt. Cell Res. 4:164-173. J. Bot. 50:1035-1045. Taylor, J.H. 1958. Incorporation of phosphorus-32 into Terada, R., Nakayama, T., Iwabuchi, M., and Shimamoto, nucleic acids and proteins during microgametogenK. 1993. A wheat histone H3 promoter confers cell esis in Tulbaghia. Am. }. Bot. 45:123-131. division-dependent and -independent expression of the gus A gene in transgenic rice plants. Plant ]. Taylor, J.H. 1959. Autoradiographic studies of nucleic 3:241-252. acids and proteins during meiosis in Lilium longiflorum. Am. J. Bot. 46:477-484. Terada, R., Nakayama, T, Iwabuchi, M., and Shimamoto, K. 1995. A type I element composed of the hexamer Taylor, J.H., and McMaster, R.D. 1954. Autoradiographic (ACGTCA) and octamer (CGCGGATC) motifs plays and microphotometric studies of desoxyribose a role(s) in meristematic expression of a wheat hisnucleic acid during microgametogenesis in Lilium longiflorum. Chromosoma 6:489-521. tone H3 gene in transgenic rice plants. Plant Mol. Biol. 27:17-26. Taylor, J.R.N., Schiissler, L., and Liebenberg, N.v.d.W. 1985. Protein body formation in the starchy Terada, S. 1928. Embryological studies in Oryza sativa L. /. endosperm of developing Sorghum bicolor (L.) Coll. Agr. Hokkaido Imp. Univ. 19:245-260. Moench seeds. S. Afr. }. Bot. 51:35-40. Terasaka, O., and Niitsu, T. 1989. Peculiar spindle configuration in the pollen tube revealed by the anti-tubuTaylor, L.P., and Jorgensen, R. 1992. Conditional male ferlin immunofluorescence method. Bot. Mag. Tokyo tility in chalcone synthase-deficient Petunia. ]. Hered. 102:143-147. 83:11-17. Taylor, M.G., and Vasil, I.K. 1991. Histology of, and phys- Terasaka, O., and Niitsu, T. 1990. Unequal cell division and chromatin differentiation in pollen grain cells. II. ical factors affecting, transient GUS expression in Microtubule dynamics associated with the unequal pearl millet (Pennisetum glaucum (L.) R. Br.) embryos following microprojectile bombardment. Plant Cell cell division. Bot. Mag. Tokyo 103:133-142. Rep. 10:120-125. Terasaka, O., and Niitsu, T. 1994. Differential roles of microTaylor, M.G., Vasil, V., and Vasil, I.K. 1993. Enhanced GUS tubule and actin-myosin cytoskeleton in the growth gene expression in cereal /grass cell suspensions and of Pinus pollen tubes. Sex. Plant Reprod. 7:264-272. immature embryos using the maize ubiquitin-based Terasaka, O., and Niitsu, T. 1995. The mitotic apparatus during unequal microspore division observed by a plasmid pAHC25. Plant Cell Rep. 12:491-495. confocal laser scanning microscope. Protoplasma Taylor, P., Kenrick, J., Li, Y., Kaul, V., Gunning, B.E.S., and Knox, R.B. 1989. The male germ unit of 189:187-193.
References
649
1992. Accumulation of group 3 late embryogenesis Terasawa, M., Shinohara, A., Hotta, Y., Ogawa, H., and abundant proteins in Zea mays embryos. Role of Ogawa, T. 1995. Localization of RecA-like recombiabscisic acid and the viviparous-1 gene product. Plant nation proteins on chromosomes of the lily at variPhysiol. 99:607-614. ous meiotic stages. Genes Devel. 9:925-934. Terc£-Laforgue, T., and Pernollet, J.-C. 1982. fitude quan- Thomas, E., Konar, R.N., and Street, H.E. 1972. The fine structure of the embryogenic callus of Ranunculus titative et qualitative de 1'accumulation des gliadines sceleratus L. /. Cell Sci. 11:95-109. au cours du deVeloppement du caryopse de bl£ (Triticum aestivtim L.). Compt. Rend. Acad. Sci. Paris Thomas, M.K., and Dnyansagar, V.R. 1975. Carbohydrate Ser. Ill 294:529-534. metabolism in pollen of Petunia nydaginiflora Juss. during germination and tube growth. Indian J. Expt. Testillano, P.S., Gonzaiez-Melendi, P., Ahmadian, P., Biol. 13:268-271. Fad6n, Bv and Risueno, M.C. 1995. The immunolocalization of nuclear antigens during the pollen Thomas, M.S., and Flavell, R.B. 1990. Identification of an developmental program and the induction of pollen enhancer element for the endosperm-specific embryogenesis. Expt. Cell Res. 221:41-54. expression of high molecular weight glutenin. Plant Thangavelu, Mv Belostotsky, D., Bevan, M.W., Flavell, Cell 2:1171-1180. R.B., Rogers, H.J., and Lonsdale, D.M. 1993. Partial Thompson, A.J., Bown, D., Yaish, S., and Gatehouse, J.A. 1991. Differential expression of seed storage protein characterization of the Nicotiana tabacum actin gene family: evidence for pollen-specific expression of genes in the pea leg] subfamily; sequence of gene one of the gene family members. Mol. Gen. Genet. legK. Biochem. Physiol. Pflanz. 187:1-12. 240:290-295. Thompson, A.J., Evans, I.M., Boulter, D., Croy, R.R.D., and Gatehouse, J.A. 1989. Transcriptional and post-tranThangstad, O.P., Iversen, T.-H., Slupphaug, G., and Bones, scriptional regulation of seed storage-protein gene A. 1990. Immunocytochemical localization of expression in pea (Pisum sativum L.). Planta myrosinase in Brassica napus L. Planta 180:245-248. 179:279-287. Thanh, V.H., and Shibasaki, K. 1977. Beta-conglycinin from soybean proteins. Isolation and immunological Thompson, G.A., Boston, R.S., Lyznik, L.A., Hodges, and physicochemical properties of the monomeric T.K., and Larkins, B.A. 1990. Analysis of promoter forms. Biochim. Biophys. Ada 490:370-384. activity from an a-zein gene 5' flanking sequence in transient expression assays. Plant Mol. Biol. Theerakulpisut, P., Xu, H., Singh, M.B., Pettitt, J.M., and 15:755-764. Knox, R.B. 1991. Isolation and developmental expression of Bcpl, an anther-specific cDNA clone in Thompson, G.A., Scherer, D.E., Foxall-van Aken, S., Kenny, J.W., Young, H.L., Shintani, D.K., Kridl, J.C, Brassica campestris. Plant Cell 3:1073-1084. and Knauf, V.C. 1991. Primary structures of the preTheunis, C.H. 1990. Ultrastructural analysis of Spinacia cursor and mature forms of stearoyl-acyl carrier prooleracea sperm cells isolated from mature pollen tein desaturase from safflower embryos and grains. Protoplasma 158:176-181. requirement of ferredoxin for enzyme activity. Proc. Theunis, C.H., Cresti, M., and Milanesi, C. 1991. Studies Natl Acad. Sci. USA 88:2578-2582. of the mature pollen of Spinacia oleracea after freeze substitution and observed with confocal laser scan- Thompson, K.F., and Taylor, J.P. 1971. Self-compatibility in kale. Heredity 27:459-471. ning fluorescence microscopy. Bot. Ada 104:324-329. Theunis, C.H., McConchie, C.A., and Knox, R.B. 1985. Thompson, R.D., and Bartels, D. 1983. Hordein messenger RNA levels in wild type and mutant barley Three-dimensional reconstruction of the generative endosperm. Plant Sci. Lett. 29:295-304. cell and its wall connection in mature bicellular pollen of Rhododendron. Micron Micros. Ada Thompson, R.D., Bartels, D., and Harberd, N.P. 1985. 16:225-231. Nucleotide sequence of a gene from chromosome ID of wheat encoding HMW-glutenin subunit. Nucl. Theunis, C.H., Pierson, E.S., and Cresti, M. 1991. Isolation Acids Res. 13:6833-6846. of male and female gametes in higher plants. Sex. Thompson, R.D., Bartels, D., Harberd, N.P., and Flavell, Plant Reprod. 4:145-154. R.B. 1983. Characterization of the multigene family Theunis, C.H., Pierson, E.S., and Cresti, M. 1992. The coding for HMW glutenin subunits in wheat using microtubule cytoskeleton and the rounding of isocDNA clones. Theor. Appl. Genet. 67:87-96. lated generative cells of Nicotiana tabacum. Sex. Plant Thompson, R.D., Kemble, R.J., and Flavell, R.B. 1980. Reprod. 5:64-71. Variations in mitochondrial DNA organisation Theunis, C.H., and van Went, J.L. 1989. Isolation of sperm between normal and male-sterile cytoplasms of cells from mature pollen grains of Spinacia oleracea. maize. Nucl. Acids Res. 8:1999-2008. Sex. Plant Reprod. 2:97-102. Thie-baud, C.H., and Ruch, F. 1978. Cytophotometric Thompson, R.D., and Kirch, H.-H. 1992. The S locus of flowering plants: when self-rejection is self-interest. study of nuclear differentiation during pollen develTrend. Genet. 8:381-387. opment in Tradescantia paludosa. Histochemistry Thompson-Coffe, C, Findlay, T.A.S., Wagner, B.A. and 57:119-128. Orr, A.R. 1992. Changes in protein complement This, P., Goffner, D., Raynal, M., Chartier, Y., and Delseny, accompany early organogenesis in the maize ear. Int. M. 1988. Characterization of major storage proteins ]. Plant Sci. 153:31-39. of sunflower and their accumulation. Plant Physiol. Biochem. 26:125-132. Thomson, J.A., and Schroeder, H.E. 1978. Cotyledonary Thoma, S., Hecht, U., Kippers, A., Botella, ]., de Vries, S., storage proteins in Pisum sativum. II. Hereditary and Somerville, C. 1994. Tissue-specific expression variation in components of the legumin and vicilin of a gene encoding a cell wall-localized lipid transfractions. Aust. ]. Plant Physiol. 5:281-294. fer protein from Arabidopsis. Plant Physiol. Thorsness, M.K., Kandasamy, M.K., Nasrallah, M.E., and Nasrallah, J.B. 1991. A Brassica S-locus gene pro105:35-45. moter targets toxic gene expression and cell death to Thomann, E.B., Sollinger, }., White, C, and Rivin, CJ.
650
References
the pistil and pollen of transgenic Nicotiana. Devel carrot (Daucus carota L.). Protoplasma 153:24-29. Biol. 143:173-184. Timmers, A.C.J., Kieft, H., and Schel, J.H.N. 1995. An immunofluorescence study on calmodulin distribuThorsness, M.K., Kandasamy, M.K., Nasrallah, M.E., and tion during somatic and zygotic embryogenesis of Nasrallah, J.B. 1993. Genetic ablation of floral cells in Arabidopsis. Plant Cell 5:253-261. carrot (Daucus carota L.). Acta Bot. Neerl. 44:19-32. Thoyts, P.J.E., Millichip, M.I., Stobart, A.K., Griffiths, W.T., Timmers, A.C.J., Reiss, H.-D., Bohsung, J., Traxel, K., and Shewry, P.R., and Napier, J.A. 1995. Expression and Schel, J.H.N. 1996. Localization of calcium during in vitro targeting of a sunflower oleosin. Plant Mol. somatic embryogenesis of carrot (Daucus carota L.). Biol. 29:403^10. Protoplasma 190:107-118. Tian, H.-Q., and Yang, H.-Y. 1991. Ultrastructural obser- Tirlapur, U.K., and Cresti, M. 1992. Computer-assisted vations on parthenogenesis and antipodal apogamy video image analysis of spatial variations in memof Allium tuberosum Roxb. Acta Bot. Sinica brane-associated Ca2+ and calmodulin during pollen 33:819-824. hydration, germination and tip growth in Nicotiana tabacum L. Ann. Bot. 69:503-508. Tierney, M.L., Bray, E.A., Allen, R.D., Ma, Y, Drong, R.F., Slightom, J., and Beachy, R.N. 1987. Isolation and Tirlapur, U.K., Kranz, E., and Cresti, M. 1995. characterization of a genomic clone encoding the PCharacterisation of isolated egg cells, in vitro fusion subunit of (3-conglycinin. Planta 172:356-363. products and zygotes of Zea mays L. using the technique of image analysis and confocal laser scanning Tiezzi, A., Moscatelli, A., Cai, G., Bartalesi, A., and Cresti, microscopy. Zygote 3:57-64. M. 1992. An immunoreactive homolog of mammalian kinesin in Nicotiana tabacum pollen tubes. Cell Tirlapur, U.K., and Shiggaon, S.V. 1988. Distribution of Mot. Cytoskel. 21:132-137. Ca2+ and calmodulin in the papillate cells of the stigma surface, visualized by chlorotetracycline and Tiezzi, A., Moscatelli, A., Milanesi, C, Ciampolini, C, and fluorescing calmodulin-binding phenothiazines. Cresti, M. 1987. Taxol-induced structures derived Ann. Biol. India 4:49-53. from cytoskeletal elements of the Nicotiana pollen tube. /. Cell Sci. 88:657-661. Tirlapur, U.K., van Went, J.L., and Cresti, M. 1993. Tilton, V.R. 1980. The nucellar epidermis and micropyle of Visualization of membrane calcium and calmodulin in Ornithogalum caudatum (Liliaceae) with a review of embryo sacs in situ and isolated from Petunia hybrida L. these structures in other taxa. Can. }. Bot. and Nicotiana tabacum L. Ann. Bot. 71:161-167. 58:1872-1884. Tirlapur, U.K., and Willemse, M.T.M. 1992. Changes in Tilton, V.R. 1981a. Ovule development in Ornithogalum calcium and calmodulin levels during microsporocaudatum (Liliaceae) with a review of selected papers genesis, pollen development and germination in on angiosperm reproduction. II. Megasporogenesis. Gasteria verrucosa (Mill.) H. Duval. Sex. Plant Reprod. New Phytol. 88:459-476. 5:214-223. Tilton, V.R. 1981b. Ovule development in Ornithogalum Tirlapur, U.K., Cai, G., Faleri, C, Moscatelli, A., Scali, M., caudatum (Liliaceae) with a review of selected papers Del Casino, C, Tiezzi, A., and Cresti, M. 1995. on angiosperm reproduction. IV. Egg apparatus Confocal imaging and immunogold electron structure and function. New Phytol. 88:505-531. microscopy of changes in distribution of myosin Tilton, V.R. 1984. Stigma, style, and pollen recognition in during pollen hydration, germination and pollen Galinsoga quadriradiata Ruiz & Pav. (Compositae). tube growth in Nicotiana tabacum L. Eur. J. Cell Biol. Caryologia 37:423-433. 67:209-217. Tilton, V.R., and Homer, H.T., Jr. 1980. Stigma, style, and Tirlapur, U.K., Scali, M., Moscatelli, A., Del Casino, C, obturator of Ornithogalum caudatum (Liliaceae) and Cai, G., Tiezzi, A., and Cresti, M. 1994. Confocal their function in the reproductive process. Am. J. Bot. image analysis of spatial variations in immunocyto67:1113-1131. chemically identified calmodulin during pollen hydration, germination and pollen tube tip growth Tilton, V.R., and Lersten, N.R. 1981. Ovule development in Nicotiana tabacum L. Zygote 2:63-68. in Ornithogalum caudatum (Liliaceae) with a review of selected papers on angiosperm reproduction. III. Titz, W. 1965. Vergleichende Untersuchungen iiber den Nucellus and megagametophyte. New Phytol. Grad der somatischen Polyploidie an nahe ver88:477-504. wandten diploiden und polyploiden Sippen einschliefilich der Cytologie von Antipoden. Osterr. Tilton, V.R., and Mogensen, H.L. 1979. Ultrastructural Bot. Z. 112:101-172. aspects of the ovule of Agave parryi before fertilization. Phytomorphology 29:338-350. Tiwari, S.C. 1982. Callose in the walls of mature embryo Tilton, V.R., and Russell, S.H. 1984. Applications of in vitro sac of Torenia fournieri. Protoplasma 110:1-4. pollination/fertilization technology. BioScience Tiwari, S.C. 1983. The hypostase in Torenia fournieri Lind.: 34:239-242. a histochemical study of the cell walls. Ann. Bot. 51:17-26. Tilton, V.R., Wilcox, L.W., and Palmer, R.G. 1984. Postfertilization wandlabrinthe formation and func- Tiwari, S.C. 1989. Cytoskeleton during pollen develoption in the central cell of soybean, Glycine max (L.) ment in Tradescantia virginiana: a study employing Merr. (Leguminosae). Bot. Gaz. 145:334-339. chemical fixation, freeze-substitution, immunofluorescence, and colchicine administration. Can. }. Bot. Tilton, V.R., Wilcox, L.W., Palmer, R.G., and Albertsen, 67:1244-1253. M.C. 1984. Stigma, style, and obturator of soybean, Glycine max (L.) Merr. (Leguminosae) and their func- Tiwari, S.C. 1994. An intermediate-voltage electron tion in the reproductive process. Am. ]. Bot. microscopic study of freeze-substituted generative 71:676-686. cell in pear (Pyrus communis L.): features with releTimmers, A.C.J., de Vries, S.C, and Schel, J.H.N. 1989. vance to cell-cell communication between the two Distribution of membrane-bound calcium and acticells of a germinating pollen. Sex. Plant Reprod. vated calmodulin during somatic embryogenesis of 7:177-186.
References
651
Tiwari, S.C, and Gunning, B.E.S. 1986a. Cytoskeleton, cell Toriyama, K., Stein, J.C, Nasrallah, M.E., and Nasrallah, J.B. 1991a. Transformation of Brassica oleracea with an surface and the development of invasive plasmodial tapetum in Tradescantia virginiana L. Protoplasma S-locus gene from B. campestris changes the self133:89-99. incompatibility phenotype. Theor. Appl. Genet. 81:769-776. Tiwari, S.C, and Gunning, B.E.S. 1986b. An ultrastructural, cytochemical and immunofluorescence study Toriyama, K., Thorsness, M.K., Nasrallah, J.B., and of postmeiotic development of plasmodial tapetum Nasrallah, M.E. 1991b. A Brassica S locus gene promoter directs sporophytic expression in the anther in Tradescantia virginiana L. and its relevance to the pathway of sporopollenin secretion. Protoplasma tapetum of transgenic Arabidopsis. Devel Biol. 133:100-114. 143:427-431. Tiwari, S.C., and Gunning, B.E.S. 1986c. Colchicine Torrent, M., Poca, E., Campos, E., Ludevid, M.D., and Palau, J. 1986. In maize, glutelin-2 and low molecuinhibits plasmodium formation and disrupts pathlar weight zeins are synthesized by membraneways of sporopollenin secretion in the anther tapebound polyribosomes and translocated into tum of Tradescantia virginiana L. Protoplasma microsomal membranes. Plant Mol. Biol. 7:393-403. 133:115-128. Tiwari, S.C., and Gunning, B.E.S. 1986d. Development of Torres, M.-A., Rigau, J., Puigdomenech, P., and Stiefel, V. 1995. Specific distribution of mRNAs in maize growtapetum and microspores in Canna L.: an example of an invasive but non-syncytial tapetum. Ann. Bot. 57: ing pollen tubes observed by whole-mount in situ hybridization with non-radioactive probes. Plant J. 557-563. 8:317-321. Tiwari, S.C, and Polito, V.S. 1988a. Spatial and temporal organization of actin during hydration, activation, Torres-Ruiz, R.A., and Jiirgens, G. 1994. Mutations in the and germination of pollen in Pyrus communis L.: a PASS gene uncouple pattern formation and morphopopulation study. Protoplasma 147:5-15. genesis in Arabidopsis development. Development 120:2967-2978. Tiwari, S.C, and Polito, V.S. 1988b. Organization of the cytoskeleton in pollen tubes of Pyrus communis: a Touraev, A., Fink, C.S., Stoger, E., and Heberle-Bors, E. study employing conventional and freeze-substitu1995a. Pollen selection: a transgenic reconstruction electron microscopy, immunofluorescence, and tion approach. Proc. Natl Acad. Sci. USA rhodamine-phalloidin. Protoplasma 147:100-112. 92:12165-12169. Tiwari, S.C, and Polito, V.S. 1990a. An analysis of the role Touraev, A., Lezin, R, Heberle-Bors, E., and Vicente, O. 1995b. Maintenance of gametophytic development of actin during pollen activation leading to germinaafter symmetrical division in tobacco microspore tion in pear (Pyrus communis L.): treatment with culture. Sex. Plant Reprod. 8:70-76. cytochalasin D. Sex. Plant Reprod. 3:121-129. Tiwari, S.C, and Polito, V.S. 1990b. The initiation and orga- Traas, J., Bellini, C, Nacry, P., Kronenberger, J., Bouchez, D., and Caboche, M. 1995. Normal differentiation nization of microtubules in germinating pear (Pyrus communis L.) pollen. Eitr.}. Cell Biol. 53:384-389. patterns in plants lacking microtubular preprophase Tiwari, S.C, Polito, VS., and Webster, B.D. 1990. In dry bands. Nature 375:676-677. pear (Pyrus communis L.) pollen, membranes assume Traas, J.A., Burgain, S., and de Vaulx, R.D. 1989. The orgaa tightly packed multilamellate aspect that disapnization of the cytoskeleton during meiosis in eggpears rapidly upon hydration. Protoplasma plant [Solanum melongena (L.)]: microtubules and F-actin are both necessary for coordinated meiotic 153:157-168. division. /. Cell Sci. 92:541-550. Tobias, CM., Howlett, B., and Nasrallah, J.B. 1992. An Arabidopsis thaliana gene with sequence similarity to Trick, M. 1990. Genomic sequence of a Brassica S locusthe S-locus receptor kinase of Brassica oleracea. Plant related gene. Plant Mol. Biol. 15:203-205. Physiol. 99:284-290. Trick, M., and Flavell, R.B. 1989. A homozygous S genotype of Brassica oleracea expresses two S-like genes. Tohda, H. 1974. Development of the embryo of Pogonia Mol. Gen. Genet. 218:112-117. (Orchidaceae). Sci. Rep. Tohoku Univ. Ser. IV (Biol.) 37:89-93. Trick, M., and Heizmann, P. 1992. Sporophytic selfTong, Z., Wang, T, and Xu, Y. 1990. Evidence for involveincompatibility systems: Brassica S gene family. Int. Rev. Cytol. 140:485-524. ment of phytochrome regulation in male sterility of a mutant of Oryza sativa L. Photochem. Pliotobiol. Triplett, B.A., and Quatrano, R.S. 1982. Timing, localiza52:161-164. tion, and control of wheat germ agglutinin synthesis Toonen, M.A.J., Hendriks, T, Schmidt, E.D.L., Verhoeven, in developing wheat embryos. Devel Biol. H.A., van Kammen, A., and de Vries, S.C. 1994. 91:491^96. Description of somatic-embryo-forming single cells Trobner, W., Ramirez, L., Motte, P., Hue, I., Huijser, P., in carrot suspension cultures employing video cell Lonnig, W.-E., Saedler, H., Sommer, H., and tracking. Planta 194:565-572. Schwarz-Sommer, Z. 1992. GLOBOSA: a homeotic Topfer, R., Gronenborn, B., Schell, J., and Steinbiss, H.-H. gene which interacts with DEFIC1ENS in the control of Antirrhinum floral organogenesis. EMBO ]. 1989. Uptake and transient expression of chimeric 11:4693-4704. genes in seed-derived embryos. Plant Cell 1:133-139. Topfer, R., and Steinbiss, H.-H. 1985. Plant regeneration Trognitz, B.R., and Schmiediche, P.E. 1993. A new look at from cultured fertilized barley ovules. Plant Sci. incompatibility relationships in higher plants. Sex. Plant Reprod. 6:183-190. 41:49-54. Toriyama, K., Okada, T, Watanabe, M., Ide, T, Ashida, T, Trull, M.C, Holaway, B.L., Friedman, W.E., and Xu, H., and Singh, M.B. 1995. A cDNA clone encodMalmberg, R.L. 1991. Developmentally regulated ing an IgE-binding protein from Brassica anther has antigen associated with calcium crystals in tobacco 2+ significant sequence similarity to Ca -binding proanthers. Planta 186:13-16. teins. Plant Mol. Biol. 29:1157-1165. Tsaftaris, A.S., and Scandalios, J.G. 1986. Spatial pattern of
652 catalase (Cat!) gene activation in scutella during postgerminative development in maize. Proc. Natl Acad. Sci. USA 83:5549-5553. Tsai, C.-Y., and Nelson, O.E. 1966. Starch-deficient maize mutant lacking adenosine diphosphate glucose pyrophosphorylase activity. Science 151:341-343. Tsai, D.-S., Lee, H.-S., Post, L.C., Kreiling, K.M., and Kao, T.-H. 1992. Sequence of an S-protein of Lycopersicon peruvianwn and comparison with other solanaceous S-proteins. Sex. Plant Reprod. 5:256-263. Tsao, T.-H. 1949. A study of chemotropism of pollen tubes
References
self- and non-pollinated pistils of Nicotiana alata. Biol. Plant. 3:47-64.
Tupy, J. 1966. Synthesis of protein and RNA in pollen tubes stimulated with 2-thiouracil. Biol. Plant. 8:398-410. Tupy, J. 1977. RNA synthesis and polysome formation in pollen tubes. Biol. Plant. 19:300-308. Tupy, J. 1982. Alterations in polyadenylated RNA during pollen maturation and germination. Biol. Plant. 24:331-340. Tupy, J., Hrabetova\ E., and Balatkova, V. 1977. Evidence in vitro. Plant Physiol. 24:494-504. for ribosomal RNA synthesis in pollen tubes in culture. Biol. Plant. 19:226-230. Tsay, H.S., Yeh, C.C., and Hsu, J.Y. 1990. Embryogenesis and plant regeneration from anther culture of bam- Tupy, J., Hrabetova, E., and Capkova-Balatkovd, V. 1980. Interaction of microbial DNA with germinating boo (Simocalamus latiflora (Munro) McClure). Plant pollen of Nicotiana tabacum L. in culture. Physiol. Cell Rep. 9:349-351. Veget. 18:677-687. Tschermak-Woess, E. 1956. Notizen iiber die Riesenkerne und "Riesenchromosomen" in dem Antipoden von Tupy, J., and Rangaswamy, N.S. 1973. The investigation of Aconitum. Chromosoma 8:114-134. the effect of pollination on ribosomal RNA, transfer Tschermak-Woess, E. 1957. Uber Kernstrukturen in den RNA and DNA contents in styles of Nicotiana alata. endopolyploiden Antipoden von Clivia miniata. Biol. Plant. 15:95-101. Chromosoma 8:637-649. Tupy, J., and Rihovd, L. 1984. Changes and growth effect Tschermak-Woess, E. 1959. Endopolyploidie der of pH in pollen tube culture. /. Plant Physiol. 115:1-10. Narbenpapillen bei Spironema fragrans. Osterr. Bot. Z.Tupy, J., Ri'hov^, L., and Zarsky, V. 1991. Production of fer106:74-80. tile tobacco pollen from microspores in suspension culture and its storage for in situ pollination. Sex. Tschermak-Woess, E., and Enzenberg-Kunz, U. 1965. Die Plant Reprod. 4:284-287. Struktur der hoch endopolyploiden Kerne im Endosperm von Zea mays, das auffallende Verhalten Tupy, J., Stanley, R.G., and Linskens, H.F. 1965. ihrer Nucleolen und ihr Endopolyploidiegrad. Stimulation of pollen tube growth in vitro by Planta 64:149-169. thiouracil and other antimetabolites of nucleic acid bases. Ada Bot. Neerl. 14:148-154. Tsinger, N.V., and Petrovskaya-Baranova, T.P. 1961. The pollen grain wall - a living, physiologically active Tupy, J., Suss, J., and Ri'hovd, L. 1986. RNA synthesis and structure. Doklady Bot. Sci.. 138:104-107. ribosome status in pollen tube growth of Nicotiana Tsuchimoto, S., van der Krol, A., and Chua, N.-H. 1993. tabacum L.; effects of external pH. /. Plant Physiol. 123:467-476. Ectopic expression of pMADS3 in transgenic Petunia phenocopies the Petunia blind mutant. Plant Cell Tupy, J., Suss, J., Hrabetova\ E., and Rihova L. 1983. 5:843-853. Developmental changes in gene expression during Tsuchiya, T., Toriyama, K., Ejiri, S.-L, and Hinata, K. 1994. pollen differentiation and maturation in Nicotiana Molecular characterization of rice genes specifically tabacum L. Biol. Plant. 25:231-237. expressed in the anther tapetum. Plant Mol. Biol. Turala, K. 1958. Endomitosis in the tapetal cells of 26:1737-1746. Cucurbita pepo. Acta Biol. Cracow Ser. Bot. 1:25-34. Tsuchiya, T., Toriyama, K., Nasrallah, M.E., and Ejiri, S.-I. Turala, K. 1966. Strukturen endopolyploider Kerne im 1992. Isolation of genes abundantly expressed in rice Bereich der Samenanlage einiger Monocotylen. anthers at the microspore stage. Plant Mol. Biol. Osterr. Bot. Z. 113:529-541. 20:1189-1193. Turbin, N.V., Atrashenok, N.V., Palilova, A.M., and Lyul'kina, E.I. 1968. Mitochondrial structure in Tsuchiya, T., Toriyama, K., Yoshikawa, M., Ejiri, S.-I., and ovules of hybrid corn obtained by crossing sterile Hinata, K. 1995. Tapetum-specific expression of the lines with fertility-restoring lines. Doklady Biol. Sci. gene for an endo-B-l,3-glucanase causes male steril182:536-538. ity in transgenic tobacco. Plant Cell Physiol. 36:487^194. Turcich, M.P., Hamilton, D.A., and Mascarenhas, J.P. 1993. Isolation and characterization of pollen-specific Tsukaya, H., Takahashi, T., Naito, S., and Komeda, Y. maize genes with sequence homology to ragweed 1993. Floral organ-specific and constitutive expresallergens and pectate Iyases. Plant Mol. Biol. sion of an Arabidopsis thaliana heat-shock 23:1061-1065. HSP18.2::GUS fusion gene is retained even after homeotic conversion of flowers by mutation. Mol Turcich, M.P., and Mascarenhas, J.P. 1994. PREM-1, a putative maize retroelement has LTR (long terminal Gen. Genet. 237:26-32. repeat) sequences that are preferentially transcribed Tully, R.E., and Beevers, H. 1976. Protein bodies of castor in pollen. Sex. Plant Reprod. 7:2-11. bean endosperm. Isolation, fractionation, and the characterization of protein components. Plant Turcich, M.P., Bokhari-Riza, A., Hamilton, D.A., He, C, Physiol. 58:710-716. Messier, W., Stewart, C.-B., and Mascarenhas, J.P. 1996. PREM-2, a copia-type retroelement in maize is Turner, N.E., Thanh, V.H., and Nielsen, N.C. 1981. expressed preferentially in early microspores. Sex. Purification and characterization of mRNA from Plant Reprod. 9:65-74. soybean seeds. Identification of glycinin and (3-conglycinin precursors. /. Biol. Chem. 256:8756-8760. Turcich, M.P., Hamilton, D.A., Yu, X., and Mascarenhas, Tupy, J. 1959. Callose formation in pollen tubes and J.P. 1994. Characterization of a pollen-specific gene incompatibility. Biol. Plant. 1:192-198. from Tradescantia paludosa with an unusual cysteine grouping. Sex. Plant Reprod. 7:201-202. Tupy, J. 1961. Investigation of free amino-acids in cross-,
References
653
of seed oil bodies from diverse species. Plant Physiol. Turgut, K., Barsby, T., Craze, M., Freeman, J., Hodge, R., Paul, W., and Scott, R. 1994. The highly expressed 101:267-276. tapetum-specific A9 gene is not required for Tzen, J.T.C., and Huang, A.H.C. 1992. Surface structure male fertility in Brassica napus. Plant Mol. Biol. and properties of plant seed oil bodies. /. Cell Biol. 24:97-104. 117:327-335. Turley, R.B., and Trelease, R.N. 1990. Development and Tzen, J.T.C., Lai, Y.-K., Chan, K.-L., and Huang, A.H.C. 1990. Oleosin isoforms of high and low molecular regulation of three glyoxysomal enzymes during weights are present in the oil bodies of diverse seed cotton seed maturation and growth. Plant Mol. Biol. 14:137-146. species. Plant Physiol. 94:1282-1289. Turnham, E., and Northcote, D.H. 1982. The use of acetylCoA carboxylase activity and changes in wall com- Ueda, K., Miyamoto, Y., and Tanaka, 1.1990. Fusion studposition as measures of embryogenesis in tissue ies of pollen protoplasts and generative cell protocultures of oil palm (Elaeis guineensis). Biochem. J. plasts in Lilium longiflorum. Plant Sci. 72:259-266. 208:323-332. Ueda, K., and Tanaka, I. 1994. The basic proteins of male gametic nuclei isolated from pollen grains of Lilium Tuschall, D.M., and Hannah, L.C. 1982. Altered maize longiflorum. Planta 192:446-452. endosperm ADP-glucose pyrophosphorylases from revertants of a shrunken-2-dissociation allele. GeneticsUeda, K., and Tanaka, I. 1995a. The appearance of male gamete-specific histones gH2B and gH3 during 100:105-111. pollen development in Lilium longiflorum. Devel Biol. Twell, D. 1992. Use of a nuclear-targeted P-glucuronidase 169:210-217. fusion protein to demonstrate vegetative cell-specific gene expression in developing pollen. Plant ]. Ueda, K., and Tanaka, 1.1995b. Male gametic nucleus-spe2:887-892. cific H2B and H3 histones, designated gH2B and gH3, in Lilium longiflorum. Planta 197:289-295. Twell, D. 1994. The diversity and regulation of gene expression in the pathway of male gametophyte Ueda, T., and Messing, J. 1991. A homologous expression system for cloned zein genes. Theor. Appl. Genet. development. In Molecular and Cellular Aspects of 82:93-100. Plant Reproduction, ed. R.J. Scott and A.D. Stead, pp. 83-135. Cambridge: Cambridge University Press. Ueda, T., Waverczak, W., Ward, K., Sher, N., Ketudat, M., Schmidt, R.J., and Messing, J. 1992. Mutations of the Twell, D. 1995. Diphtheria toxin-mediated cell ablation in 22- and 27-kD zein promoters affect transactivation developing pollen: vegetative cell ablation blocks by the opaque-2 protein. Plant Cell 4:701-709. generative cell migration. Protoplasma 187:144-154. Twell, D., Yamaguchi, J., and McCormick, S. 1990. Pollen- Ueng, P., Galili, G., Sapanara, V., Goldsbrough, P.B., Dube, specific gene expression in transgenic plants: coordiP., Beachy, R.N., and Larkins, B.A. 1988. Expression nate regulation of two different tomato gene of a maize storage protein gene in Petunia plants is promoters during microsporogenesis. Development not restricted to seeds. Plant Physiol. 86:1281-1285. 109:705-713. Uknes, S., Dincher, S., Friedrich, L., Negrotto, D., Twell, D., Klein, T.M., Fromm, M.E., and McCormick, S. Williams, S., Thompson-Taylor, H., Potter, S., Ward, 1989a. Transient expression of chimeric genes delivE., and Ryals, J. 1993. Regulation of pathogenesisered into pollen by microprojectile bombardment. related protein-la gene expression in tobacco. Plant Plant Physiol. 91:1270-1274. Cell 5:159-169. Twell, D., Patel, S., Sorensen, A., Roberts, M., Scott, R., Ulrich, T.U., Wurtele, E.S., and Nikolau, B.J. 1990. Draper, J., and Foster, G. 1993. Activation and develSequence of EMB-1, an mRNA accumulating specifically in embryos of carrot. Nucl. Acids Res. 18:2826. opmental regulation of an Arabidopsis anther-specific promoter in microspores and pollen of Nicotiana Umbach, A.L., Lalonde, B.A., Kandasamy, M.K., Nasrallah, J.B., and Nasrallah, M.E. 1990. tabacum. Sex. Plant Reprod. 6:217-224. Immunodetection of protein glycoforms encoded by Twell, D., Wing, R., Yamaguchi, J., and McCormick, S. two independent genes of the self-incompatibility 1989b. Isolation and expression of an anther-spemultigene family of Brassica. Plant Physiol. cific gene from tomato. Mol. Gen. Genet. 93:739-747. 217:240-245. Twell, D., Yamaguchi, J., Wing, R.A., Ushiba, J., and Umbeck, PR, and Gengenbach, B.G. 1983. Reversion of male-sterile T-cytoplasm maize to male fertility in McCormick, S. 1991. Promoter analysis of genes that tissue culture. Crop Sci. 23:584-588. are coordinately expressed during pollen development reveals pollen-specific enhancer sequences Unger, E., Parsons, R.L., Schmidt, R.J., Bowen, B., and and shared regulatory elements. Genes Devel. Roth, B.A. 1993. Dominant negative mutants of 5:496-507. opaque2 suppress transactivation of a 22-kD zein promoter by opaque2 in maize endosperm cells. Tyagi, A.K., Rashid, A., and Maheshwari, S.C. 1979. High Plant Cell 5:831-841. frequency production of embryos in Datura innoxia from isolated pollen grains by combined cold treat- Ursin, V.M., Yamaguchi, J., and McCormick, S. 1989. Gametophytic and sporophytic expression of antherment and serial culture of anthers in liquid medium. specific genes in developing tomato anthers. Plant Protoplasma 99:11-17. Cell 1:727-736. Tyagi, A.K., Rashid, A., and Maheshwari, S.C. 1981. Promotive effect of polyvinylpolypyrrolidone on Usha, S.V. 1965. In vitro pollination in Antirrhinum majus pollen embryogenesis in Datura innoxia. Physiol. L. Cum Sci. 34:511-513. Plant. 53:405-406. Usmanov, P.D., and Miiller, A. 1973. Use of the embryo Tykarska, T. 1979. Rape embryogenesis. II. Development test for the analysis of embryonic lethals induced by of embryo proper. Ada Soc. Bot. Polon. 48:391-421. the irradiation of the pollen grains of Arabidopsis Tzen, J.T.C., Cao, Y.-Z., Laurent, P., Ratnayake, C, and thaliana (L.) Heynh. Sov. Genet. 6:894-902. Huang, A.H.C. 1993. Lipids, proteins, and structure Utsumi, S., Kitagawa, S., Katsube, T., Higasa, T, Kito, M.,
654
References
Takaiwa, R, and Ishige, T. 1994. Expression and van Eldik, G.J., Wingens, M., Ruiter, R.K., van Herpen, accumulation of normal and modified soybean M.M.A., Schrauwen, J.A.M., and Wullems, G.J. 1996. Molecular analysis of a pistil-specific gene expressed glycinins in potato tubers. Plant Sci. 102:181-188. in the stigma and stylar cortex of Solanum tuberosum. Uwate, W.J., and Lin, J. 1980. Cytological zonation of Plant Mol. Biol. 30:171-176. Primus avium L. pollen tubes in vivo. J. Ultrastr. Res. 71:173-184. van Engelen, F.A., de Jong, A.J., Meijer, E.A., Kuil, C.W., Uwate, W.J., and Lin, J. 1981a. Tissue development in the Meyboom, J.K., Dirkse, W.G., Booij, H., Hartog, M.V., Vandekerckhove, J., de Vries, S.C., and van stigma of Primus avium L. Ann. Bot. 47:41-51. Kammen, A. 1995. Purification, immunological charUwate, W.J., and Lin, J. 1981b. Development of the stigacterization and cDNA cloning of a 47 kDa glycomatic surface of Primus avium L., sweet cherry. Am. ]. protein secreted by carrot suspension cells. Plant Bot. 68:1165-1176. Mol. Biol. 27:901-910. Uwate, W.J., Lin, J., Ryugo, K., and Stallman, V. 1982. Cellular components of the midstylar transmitting van Engelen, F.A., Hartog, M.V., Thomas, T.L., Taylor, B., Sturn, A., van Kammen, A., and de Vries, S.C. 1993. tissue of Primus avium. Can. J. Bot. 60:98-104. The carrot secreted glycoprotein gene EP1 is expressed in the epidermis and has sequence homology to Valenta, R., Duchene, M., Pettenburger, Kv Sillaber, C, Brassica S-locus glycoproteins. Plant J. 4:855-862. Valent, P., Bettelheim, P., Breitenbach, M., Rumpold, H., Kraft, D., and Scheiner, O. 1991. Identification of van Engelen, F.A., Sterk, P., Booij, H., Cordewener, J.H.G., profilin as a novel pollen allergen; IgE autoreactivity Rook, W., van Kammen, A., and de Vries, S.C. 1991. in sensitized individuals. Science 253:557-560. Heterogeneity and cell type-specific localization of a cell wall glycoprotein from carrot suspension cells. Vallade, J. 1970. DeVeloppement embryonnaire chez un Plant Physiol. 96:705-712. Petunia hybrida hort. Compt. Rend. Acad. Sci. Paris 270D:1893-1896. van Geyt, J., D'Halluin, K., and Jacobs, M. 1985. Induction of nuclear and cell divisions in microspores of sugVallade, J. 1972. Structure et fonctionnement du arbeet {Beta vulgaris L.). Z. Pflanzenziichtg 95:325-335. me'risteme lors de la formation de la jeune racine primaire chez un Petunia hybrida hort. Compt. Rend. van Heel, W.A. 1981. A S.E.M.-investigation on the develAcad. Sci. Paris 274D:1027-1030. opment of free carpels. Blumea 27:499-522. Vallade, J. 1980. Donn£es cytologiques sur la proembryo- van Herpen, M.M.A., de Groot, P.F.M., Schrauwen, genese du Petunia; interet pour une interpretation J.A.M., van den Heuvel, K.J.P.T, Weterings, K.A.P., morphoge'ne'tique du de'veloppement embryonand Wullems, G.J. 1992. In-vitro culture of tobacco naire. Bull. Soc. Bot. Fr. Actual. Bot. 127:19-37. pollen: gene expression and protein synthesis. Sex. Plant Reprod. 5:304-309. Vallade, J., Cornu, A., Essad, S., and Alabouvette, J. 1978. Niveaux de DNA dans les noyaux zygotiques chez van Herpen, M.M.A., Reijnen, W.H., Schrauwen, J.A.M., le Petunia hybrida hort. Bull. Soc. Bot. Fr. Actual. Bot. de Groot, P.F.M., Jager, J.W.H., and Wullems, G.J. 1989. Heat-shock proteins and survival of germinat125:253-258. ing pollen of Lilium longiflorum and Nicotiana van Aelst, A.C., Theunis, C.H., and van Went, J.L. 1990. tabacum. J. Plant Physiol. 134:345-351. Freeze-fracture studies on isolated sperm cells of Spinacia oleracea L. Protoplasma 153:204-207. van Hoist, G.-J., and Clarke, A.E. 1986. Organ-specific arabinogalactan-proteins of Lycopersicon peruvianum van Aelst, A.C., and van Went, J.L. 1991. The ultrastruc(Mill) demonstrated by crossed electrophoresis. ture of mature Papaver dubium L. pollen grains. Ada Bot. Neerl. 40:319-328. Plant Physiol. 80:786-789. van Aelst, A.C., and van Went, J.L. 1992. Ultrastructural van Lammeren, A.A.M. 1986. A comparative ultrastrucimmuno-localization of pectins and glycoproteins in tural study of the megagametophytes in two strains Arabidopsis thaliana pollen grains. Protoplasma of Zea mays L. before and after fertilization. Agr. 168:14-19. Univ. Wageningen Papers 86-1:1-37. van Aelst, A.C., Mueller, T, Dueggelin, M., and van Lammeren, A.A.M. 1988a. Structure and function of Guggenheim, R. 1989. Three-dimensional observathe microtubular cytoskeleton during endosperm tions on freeze-fractured frozen hydrated Papaver development in wheat: an immunofluorescence dubium pollen with cryo-scanning electron study. Protoplasma 146:18-27. microscopy. Ada Bot. Neerl. 38:25-30. van Lammeren, A.A.M. 1988b. Observations on the strucvan Aelst, A.C., Pierson, E.S., van Went, J.L., and Cresti, tural development of immature maize embryos {Zea mays L.) during in vitro culture in the presence or M. 1993. Ultrastructural changes of Arabidopsis thaliana pollen during final maturation and rehydraabsence of 2,4-D. Ada Bot. Neerl. 37:49-61. tion. Zygote 1:173-179. van Lammeren, A.A.M., Bednara, J., and Willemse, M.T.M. 1989. Organization of the actin cytoskeleton van Boxtel, J., Berthouly, M., Carasco, C, Dufour, M., and during pollen development in Gasteria verrucosa Eskes, A. 1995. Transient expression of (3-glu(Mill.) H. Duval visualized with rhodamine-phalcuronidase following biolistic delivery of foreign loidin. Planta 178:531-539. DNA into coffee tissues. Plant Cell Rep. 14:748-752. van Driessche, E., Smets, G., Dejaegere, R., and Kanarek, van Lammeren, A.A.M., Keijzer, C.J., Willemse, M.T.M., L. 1981. The immuno-histochemical localization of and Kieft, H. 1985. Structure and function of the lectin in pea seeds (Pisum sativum L.). Planta microtubular cytoskeleton during pollen develop153:287-296. ment in Gasteria verrucosa (Mill.) H. Duval. Planta 165:1-11. van Eldik, G.J., Vriezen, W.H., Wingens, M., Ruiter, R.K., van Herpen, M.M.A., Schrauwen, J.A.M., and van Marrewijk, G.A.M., Bino, R.J., and Suurs, L.C.J.M. Willlems, G.J. 1995. A pistil-specific gene of Solanum 1986. Characterization of cytoplasmic male sterility tuberosum is predominantly expressed in the stylar in Petunia hybrida. I. Localization, composition and cortex. Sex. Plant Reprod. 8:173-179. activity of esterases. Euphytica 35:77-88.
References
655
van Overbeek, ]., Conklin, M.E., and Blakeslee, A.F. 1942. van Went, J.L., and Linskens, H.F. 1967. Die Entwicklung des sogenannten "Fadenapparates" im Embryosack Cultivation in vitro of small Datura embryos. Am. ]. von Petunia hybrida. Ziichter 37:51-56. Bot. 29:472-477. van Overbeek, ]., Siu, Rv and Haagen-Smit, A.J. 1944. van Went, J.L., and Willemse, M.T.M. 1984. Fertilization. In Embryology of Angiosperms, ed. B. M. Johri, pp. Factors affecting the growth of Datura embryos in 273-317. Berlin: Springer-Verlag. vitro. Am. J. Bot. 31:219-224. van Rensburg, J.G.J., and Robbertse, P.J. 1988. Seed devel- van der Donk, J.A.W.M. 1974a. Differential synthesis of RNA in self- and cross-pollinated styles of Petunia opment of Ornithogalum dubium, with special referhybrida L. Mol. Gen. Genet. 131:1-8. ence to fertilization and the egg apparatus. S. Afr. ]. Bot. 54:196-202. van der Donk, J.A.W.M. 1974b. Synthesis of RNA and protein as a function of time and type of pollentubevan Rooijen, G.J.H., and Moloney, MM. 1994. Plant seed style interaction in Petunia hybrida L. Mol. Gen. Genet. oil bodies as carriers for foreign proteins. 134:93-98. Biotechnology 13:72-77. van Rooijen, G.J.H., Terning, L.I., and Moloney, M.M. van der Donk, J.A.W.M. 1975. Recognition and gene expression during the incompatibility reaction in 1992. Nucleotide sequence of an Arabidovsis thaliana oleosin gene. Plant Mol. Biol. 18:1177-1179. Petunia hybrida L. Mol. Gen. Genet. 141:305-316. van Rooijen, G.J.H., Wilen, R.W., Holbrook, L.A., and van der Geest, A.H.M., Frisch, D.A., Kemp, J.D., and Hall, T.C. 1995. Cell ablation reveals that expression from Moloney, M.M. 1992. Regulation of accumulation of the phaseolin promoter is confined to embryogenemRNAs encoding a 20-kDa oil-body protein in sis and microsporogenesis. Plant Physiol. microspore-derived embryos of Brassica napus. Can. 109:1151-1158. ]. Bot. 70:503-508. van Tunen, A.J., Hartman, S.A., Mur, L.A., and Mol, van der Geest, A.H.M., Hall, G.E., Jr., Spiker, S., and Hall, J.N.M. 1989. Regulation of chalcone flavanone isoT.C. 1994. The p-phaseolin gene is flanked by matrix merase (CHI) gene expression in Petunia hybrida: the attachment regions. Plant J. 6:413-423. use of alternative promoters in corolla, anthers and van der Klei, H., van Damme, J., Casteels, P., and pollen. Plant Mol. Biol. 12:539-551. Krebbers, E. 1993. A fifth 2S albumin isoform is present in Arabidopsis thaliana. Plant Physiol. van Tunen, A.J., Mur, L.A., Brouns, G.S., Rienstra, J.-D., 101:1415-1416. Koes, R.E., and Mol, J.N.M. 1990. Pollen- and antherspecific chi promoters from Petunia: tandem promoter van der Kley, F.K. 1954. Male sterility and its importance in breeding heterosis varieties. Euphytica 3:117-124. regulation of the MA gene. Plant Cell 2:393^01. van Tunen, A.J., Mur, L.A., Recourt, K., Gerats, A.G.M., van der Krol, A.R., Brunelle, A., Tsuchimoto, S., and Chua, N.-H. 1993. Functional analysis of Petunia floand Mol, J.N.M. 1991. Regulation and manipulation ral homeotic MADS box gene pMADSl. Genes Devel. of flavonoid gene expression in anthers of Petunia: 7:1214-1228. the molecular basis of the Po mutation. Plant Cell van der Leede-Plegt, L.M., van den Ven, B.C.E., Bino, 3:39-48. R.J., van der Salm, T.P.M., and van Tunen, A.J. van Tuyl, J.M., Marcucci, M.C., and Visser, T. 1982. Pollen 1992. Introduction and differential use of various and pollination experiments. VII. The effect of pollen promoters in pollen grains of Nicotiana glutinosa treatment and application method on incompatibiland Lilium longiflorum. Plant Cell Rep. 11:20-24. ity and incongruity in Lilium. Euphytica 31:613-619. van Went, J., and Cresti, M. 1988a. Pre-fertilization degen- van der Maas, H.M., de Jong, E.R., van Aelst, A.C., eration of both synergids in Brassica campestris Verhoeven, H.A., Krens, F.A., and van Went, J.L. ovules. Sex. Plant Reprod. 1:208-216. 1994. Cytological characterization of isolated sperm cells of perennial ryegrass (Lolium perenne L.). van Went, J., and Cresti, M. 1988b. Cytokinesis in Protoplasma 178:48-56. microspore mother cells of Impatiens sultani. Sex. van der Maas, H.M., Zaal, M.A.C.M., de Jong, E.R., Krens, Plant Reprod. 1:228-233. F.A., and van Went, J.L. 1993a. Isolation of viable egg van Went, J., and Cresti, M. 1989. Cytoplasmic differenticells of perennial ryegrass (Lolium perenne L.). ation during tetrad formation and early microspore Protoplasma 173:86-89. development in Impatiens sultani. Protoplasma van der Maas, H.M., Zaal, M.A.C.M., de Jong, E.R., van 148:1-7. Went, J.L., and Krens, F.A. 1993b. Optimization of van Went, J.L. 1970a. The ultrastructure of the synergids isolation and storage of sperm cells from pollen of of Petunia. Acta Bot. Neerl. 19:121-132. perennial ryegrass {Lolium perenne L.). Sex. Plant van Went, J.L. 1970b. The ultrastructure of the egg and Reprod. 6:64-70. central cell of Petunia. Acta Bot. Neerl. 19:313-322. van Went, J.L. 1970c. The ultrastructure of the fertilized van der Meer, I.M., Stam, M.E., van Tunen, A.J., Mol, J.N.M., and Stuitje, A.R. 1992. Antisense inhibition of embryo sac of Petunia. Acta Bot. Neerl. 19:468-480. flavonoid biosynthesis in Petunia anthers results in van Went, J.L. 1981. Some cytological and ultrastructural male sterility. Plant Cell 4:253-262. aspects of male sterility in Impatiens. Acta Soc. Bot. van der Meer, Q.P., and van Bennekom, J.L. 1969. Effect of Polon. 50:248-252. van Went, J.L. 1984. Unequal distribution of plastids durtemperature on the occurrence of male sterility in ing generative cell formation in Impatiens. Theor. onion (Allium cepa L.). Euphytica 18:389-394. van der Meer, Q.P., and van Bennekom, J.L. 1973. Appl. Genet. 68:305-309. Gibberellic acid as a gametocide for the common van Went, J.L., and Gori, P. 1989. The ultrastructure of onion (Allium cepa L.). Euphytica 22: 239-243. Capparis spinosa pollen grains. /. Submicros. Cytol. Path. 21:149-156. van der Meer, Q.P., and van Dam, R. 1979. Gibberellic acid van Went, J.L., and Kwee, H.-S. 1990. Enzymatic isolation as a gametocide for cole crops. Euphytica 28:717-722. of living embryo sacs of Petunia. Sex. Plant Reprod. van der Plas, L.H.W., de Gucht, L.P.E., Bakels, R.H.A., and 3:257-262. Otto, B. 1987. Growth and respiratory characteristics
656
References
of batch and continuous cell suspension cultures Vasil'ev, A.E. 1970. Infrastructure of the stigmatoid cells derived from fertile and male sterile Petunia hybrida. in Lilium. Soviet Plant Physiol. 17:1035-1044. ]. Plant Physiol. 130:449^60. Vaughn, K.C., DeBonte, L.R., Wilson, K.G., and Schaffer, G.W. 1980. Organelle alteration as a mechanism for van der Pluijm, J., and Linskens, H. F. 1966. Feinstruktur maternal inheritance. Science 208:196-198. der Pollenschlauche im Griffel von Petunia. Ziichter 36:220-224. Vazart, B. 1970. Morphogenese du sporoderme et participation des mitochondries a la mise en place de la van der Pluijm, J.E. 1964. An electron microscopic investiprimexine dans le pollen de Linum usitatissimum L. gation of the filiform apparatus in the embryo sac of Torenia fournieri. In Pollen Physiology and Fertilization, Compt. Rend. Acad. Sci. Paris 270D:3210-3212. ed. H.F. Linskens, pp. 8-16. Amsterdam: North- Vazart, B. 1971. Infrastructure de microspores de Nicotiana Holland Publishing Co. tabacum L. susceptibles de se deVelopper en embryoi'des apres excision et mise en culture des antheres. Vancanneyt, G., Sonnewald, U., Hofgen, R., and Willmitzer, L. 1989. Expression of a patatin-like proCompt. Rend. Acad. Sci. Paris 272D:549-552. tein in the anthers of potato and sweet pepper flow- Vazart, B., and Vazart, J. 1965. Infrastructure de 1'ovule de ers. Plant Cell 1:533-540. lin, Linum usitatissimum L. Les cellules du sac embryVance, V.B., and Huang, A.H.C. 1987. The major protein onnaire. Compt. Rend. Acad. Sci. Paris 261:3447-3450. from lipid bodies of maize. Characterization and Vazart, B., and Vazart, J. 1966. Infrastructure du sac structure based on cDNA cloning. /. Biol. Chem. embryonnaire du lin (Linum usitatissimum L.). Rev. 262:11275-11279. Cytol. Biol. Veget. 29:251-266. Vance, V.B., and Huang, A.H.C. 1988. Expression of lipid Vazart, J. 1969. Organisation et ultrastructure du sac body protein gene during maize seed development. embryonnaire du lin (Linum usitatissimum L.). Rev. Spatial, temporal, and hormonal regulation. /. Biol. Cytol. Biol. Veget. 32:227-240. Chem. 263:1476-1481. Vedel, F., and Mathieu, C. 1983. Physical and gene mapping of chloroplast DNA from normal and cytoplasVanDerWoude, W.J., and Morre, D.J. 1968. Endoplasmic mic male sterile (radish cytoplasm) lines of Brassica reticulum-dictyosome-secretory vesicle associations in pollen tubes of Liliwn longiflorum Thunb. Proc. napus. Curr. Genet. 7:13-20. Indiana Acad. Sci. 77:164-170. Vedel, F., Mathieu, C, Lebacq, P., Ambard-Bretteville, F., VanDerWoude, W.J., Morre, D.J., and Bracker, C.E. 1971. Remy, R., Pelletier, G., Renard, M., and Rousselle, P. 1982. Comparative macromolecular analysis of the Isolation and characterization of secretory vesicles in cytoplasms of normal and cytoplasmic male sterile germinated pollen of Lilium longiflorum. J. Cell Sci. 8:331-351. Brassica napus. Theor. Appl. Genet. 62:255-262. Varagona, M.J., Schmidt, R.J., and Raikhel, N.V. 1991. Veen, H. 1963. The effect of various growth-regulators on Monocot regulatory protein opaque-2 is localized in embryos of Capsella bursa-pastoris growing in vitro. the nucleus of maize endosperm and transformed Acta Bot. Neerl. 12:129-171. tobacco plants. Plant Cell 3:105-113. Veit, B., Schmidt, R.J., Hake, S., and Yanofsky, M.F. 1993. Maize floral development: new genes and old Vasil, I.K. 1957. Effect of kinetin and gibberellic acid on mutants. Plant Cell 5:1205-1215. excised anthers of Allium cepa. Phytomorphology 7:138-149. Vellanoweth, R.L., and Okita, T.W. 1993. Analysis of Vasil, I.K. 1959. Cultivation of excised anthers in vitro nuclear proteins interacting with a wheat a/peffect of nucleic acids. /. Expt. Bot. 10:399-408. gliadin seed storage protein gene. Plant Mol. Biol. 22:25-41. Vasil, I.K. 1967. Physiology and cytology of anther development. Biol. Rev. 42:327-373. Venema, G., and Koopmans, A. 1962. A phase-contrast microscopic study of pollengrain germination, Vasil, I.K., and Johri, M.M. 1964. The style, stigma and nuclear movement and pollentube mitosis in pollen tube -1. Phytomorphology 14:352-369. Vasil, V, and Hildebrandt, A.C. 1965. Differentiation of Tradescantia virginiana. Cytologia 27:11-24. tobacco plants from single, isolated cells in micro- Venkateswarlu, J., and Rao, G.R. 1958. A contribution to cultures. Science 150:889-892. the life-history of Rubia cordifolia Linn. /. Indian Bot. Vasil, V., and Vasil, I.K. 1981. Somatic embryogenesis and Soc. 37:442-454. plant regeneration from suspension cultures of pearl Vennigerholz, F. 1992. The transmitting tissue in millet (Pennisetum amerkanum). Ann. Bot. 47:669-678. Brugmansia suaveolens: immunocytochemical localization of pectin in the style. Protoplasma 171:117-122. Vasil, V, and Vasil, I.K. 1982a. The ontogeny of somatic embryos of Pennisetum amerkanum (L.) K. Schum. I. Vergara, M.R., Biasini, G., Lo Schiavo, F., and Terzi, M. 1982. Isolation and characterization of carrot cell In cultured immature embryos. Bot. Gaz. mutants resistant to a-amanitin. Z. Pflanzenphysiol. 143:454-465. 107:313-319. Vasil, V., and Vasil, I.K. 1982b. Characterization of an embryogenic cell suspension culture derived from Vergara, R., Verde, F., Pitto, L., Lo Schiavo, F., and Terzi, cultured inflorescences of Pennisetum amerkanum M. 1990. Reversible variations in the methylation (pearl millet, Gramineae). Am. J. Bot. 69:1441-1449. pattern of carrot DNA during somatic embryogenesis. Plant Cell Rep. 8:697-700. Vasil, V, Castillo, A.M., Fromm, M.E., and Vasil, I.K. 1992. Herbicide resistant fertile transgenic wheat Vergne, P., and Dumas, C. 1988. Isolation of viable wheat plants obtained by microprojectile bombardment of male gametophytes of different stages of developregenerable embryogenic callus. Biotechnology ment and variations in their protein patterns. Plant 10:667-674. Physiol. 88:969-972. Vasil, V., Srivastava, V., Castillo, A.M., Fromm, M.E., and Vergne, P., Riccardi, F., Beckert, M., and Dumas, C. 1993. Vasil, I.K. 1993. Rapid production of transgenic Identification of a 32-kDa anther marker protein for wheat plants by direct bombardment of cultured androgenic response in maize, Zea mays L. Theor. immature embryos. Biotechnology 11:1553-1558. Appl. Genet. 86:843-850.
References
657
Verma, D.C., and Dougall, D.K. 1978. DNA, RNA and Populus alba and P. nigra. ]. Cell Sci. 87:483-490. protein content of tissue during growth and embryo- Viotti, A., Sala, E., Alberi, P., and Soave, C. 1978. genesis in wild-carrot suspension cultures. In Vitro Heterogeneity of zein synthesized in vitro. Plant Sci. 14:183-191. Lett. 13:365-375. Verma, S.C., Malik, R., and Dhir, I. 1977. Genetics of the Vishnyakova, M.A., and Willemse, M.T.M. 1994. Pollen-pistil interaction in wheat. Ada Bot. Neerl. incompatibility system in the crucifer Ernca sativa L. 43:51-64. Proc. Roy. Soc. Land. 196B:131-159. Vernon, D.M., and Meinke, D.W. 1994. Embryogenic Vithanage, H.I.M.V., Gleeson, P.A., and Clarke, A.E. 1980. The nature of callose produced during self-pollinatransformation of the suspensor in twin, a polyemtion in Secale cereale. Planta 148:498-509. bryonic mutant of Arabidopsis. Devel Biol. Vithanage, H.I.M.V., and Heslop-Harrison, J. 1979. The 165:566-573. pollen-stigma interaction: fate of fluorescentVernon, D.M., and Meinke, D.W. 1995. Late embryo-defeclabelled pollen-wall proteins on the stigma surface tive mutants of Arabidopsis. Devel Genet. 16:311-320. in rye (Secale cereale). Ann. Bot. 43:113-114. Verwoert, I.I.G.S., van der Linden, K.H., Walsh, M.C., Nijkamp, H.J.J., and Stuitje, A.R. 1995. Modification Vithanage, H.I.M.V., and Knox, R.B. 1976. Pollen-wall proteins: quantitative cytochemistry of the origins of of Brassica napus seed oil by expression of the intine and exine enzymes in Brassica oleracea. ]. Cell Escherichia coli fabH gene, encoding 3-ketoacyl-acyl Sci. 21:423-435. carrier protein synthase III. Plant Mol. Biol. Vithanage, H.I.M.V., and Knox, R.B. 1977. Development 27:875-886. and cytochemistry of stigma surface and response to Vicente, O., Benito Moreno, R.M., and Heberle-Bors, E. self and foreign pollination in Helianthus animus. 1991. Pollen cultures as a tool to study plant develPhytomorphology 27:168-179. opment. Cell Biol. Rev. 25:295-306. Vielle, J.P., Burson, B.L., Bashaw, E.C., and Hussey, M.A. Vithanage, H.I.M.V., and Knox, R.B. 1979. Pollen development and quantitative cytochemistry of exine and 1995. Early fertilization events in the sexual and intine enzymes in sunflower, Helianthus annuus L. aposporous egg apparatus of Pennisetum ciliare (L.) Ann. Bot. 44:95-106. Link. Plant ]. 8:309-316. Vijayaraghavan, M.R., and Bhat, U. 1983. Synergids Vithanage, H.I.M.V., and Knox, R.B. 1980. Periodicity of pollen development and quantitative cytochemistry before and after fertilization. Phytomorphology of exine and intine enzymes in the grasses Lolium 33:74-84. perenne L. and Phalaris tuberosa L. Ann. Bot. Vijayaraghavan, M.R., Jensen, W.A., and Ashton, M.E. 45:131-141. 1972. Synergids of Aquilegia formosa - their histochemistry and ultrastructure. Phytomorphology Vithanage, H.I.M.V, Howlett, B.J., Jobson, S., and Knox, R.B. 1982. Immunocytochemical localization of 22:144-159. water-soluble glycoproteins, including group 1 allerVijayaraghavan, M.R., and Prabhakar, K. 1984. The gen, in pollen of ryegrass, Lolium perenne, using ferendosperm. In Embryology of Angiosperms, ed. B.M. ritin-labelled antibody. Histochem. J. 14:949-966. Johri, pp. 319-376. Berlin: Springer-Verlag. Vijayaraghavan, M.R., and Shukla, A.K. 1977. Absence of Vivekananda, J., Drew, M.C., and Thomas, T.L. 1992. callose around the microspore tetrad and poorly Hormonal and environmental regulation of the cardeveloped exine in Pergularia daemia. Ann. Bot. rot lea-dass gene Dc3. Plant Physiol. 100:576-581. 41:923-926. Vodkin, L.O. 1981. Isolation and characterization of mesVilardell, J., Goday, A., Freire, M.A., Torrent, M., senger RNAs for seed lectin and Kunitz trypsin Martinez, M.C., Tome, J.M., and Pages, M. 1990. inhibitor in soybeans. Plant Physiol. 68:766-771. Gene sequence, developmental expression, and pro- Vodkin, L.O., and Raikhel, N.V. 1986. Soybean lectin and tein phosphorylation of RAB-17 in maize. Plant Mol. related proteins in seeds and roots of Le+ and Le" Biol. 14:423-432. soybean varieties. Plant Physiol. 81:558-565. Vilardell, J., Marti'nez-Zapater, J.M., Goday, A., Arenas, C., Vodkin, L.O., Rhodes, PR., and Goldberg, R.B. 1983. cA lectin gene insertion has the structural features of a and Pages, M. 1994. Regulation of the rafcl7 gene transposable element. Cell 34:1023-1031. promoter in transgenic Arabidopsis wild-type, ABAdeficient and ABA-insensitive mutants. Plant Mol. Voelker, T., Sturm, A., and Chrispeels, M.J. 1987. Differences in expression between two seed lectin Biol. 24:561-569. alleles obtained from normal and lectin-deficient Vilardell, J., Mundy, J., Stilling, B., Leroux, B., Pla, M., beans are maintained in transgenic tobacco. EMBO ]. Freyssinet, G., and Pages, M. 1991. Regulation of the maize rabl7 gene promoter in transgenic heterolo6:3571-3577. gous systems. Plant Mol. Biol. 17:985-993. Voelker, T.A., Herman, E.M., and Chrispeels, M.J. 1989. In vitro mutated phytohemagglutinin genes expressed Villanueva, V.R., Mathivet, V, and Sangwan, R.S. 1985. in tobacco seeds: role of glycans in protein targeting RNA, proteins and polyamines during gametoand stability. Plant Cell 1:95-104. phytic and androgenetic development in Nicotiana tabacum and Datura innoxia. Plant Growth Reg. Voelker, T.A., Staswick, P., and Chrispeels, M.J. 1986. 3:293-307. Molecular analysis of two phytohemagglutinin Villar, M., Gaget, M., Rougier, M., and Dumas, C. 1993. genes and their expression in Phaseolus vidgaris cv. Pollen-pistil interactions in Popnlus: p-galactosidase Pinto, a lectin-deficient cultivar of the bean. EMBO /. activity associated with pollen-tube growth in the 5:3075-3082. crosses P. nigra X P. nigra and P. nigra X P. alba. Sex. Vogt, T., Pollak, P., Tarlyn, N., and Taylor, L.P. 1994. Pollination- or wound-induced kaempferol accumuPlant Reprod. 6:249-256. lation in Petunia stigmas enhances seed production. Villar, M., Gaget, M., Said, C., Knox, R.B., and Dumas, C. Plant Cell 6:11-23. 1987. Incompatibility in Populus: structural and cytochemical characteristics of the receptive stigmas of von Allmen, J.-M., Rottmann, W.H., Gengenbach, B.G.,
658
References
Harvey, A.J., and Lonsdale, D.M. 1991. Transfer of Walker, J.C., and Zhang, R. 1990. Relationship of a putamethomyl and HmT-toxin sensitivity from Ttive receptor protein kinase from maize to the Scytoplasm maize to tobacco. Mol. Gen. Genet. locus glycoproteins of Brassica. Nature 345:743-746. 229:405-412. Walker, K.A., and Sato, S.J. 1981. Morphogenesis in callus tissue of Medicago sativa: the role of ammonium ion von Arnold, S., and Hakman, I. 1988. Regulation of somatic embryo development in Picea abies by in somatic embryogenesis. Plant Cell Tissue Organ abscisic acid (ABA). /. Plant Physiol. 132:164-169. Cult. 1:109-121. Vonder Haar, R.A., Allen, R.D., Cohen, E.A., Nessler, C.L., Wallace, J.C., Lopes, M.A., Paiva, E., and Larkins, B.A. and Thomas, T.L. 1988. Organization of the sun1990. New methods for extraction and quantitation flower US storage protein gene family. Gene of zeins reveal a high content of y-zein in modified 74:433-443. opaque-2 maize. Plant Physiol. 92:191-196. Voss, R., Schumann, K., and Nagl, W. 1992. Wallace, N.H., and Kriz, A.L. 1991. Nucleotide sequence of a cDNA clone corresponding to the maize Phytohemagglutinin gene expression during seed Globulin-2 gene. Plant Physiol. 95:973-975. development of the runner bean, Phaseolus coccineus. Plant Mol. Biol. 20:781-790. Walling, L., Drews, G.N., and Goldberg, R.B. 1986. Transcriptional and post-translational regulation of Vroemen, C, Langeveld, S., Mayer, U., Ripper, G., soybean seed protein mRNA levels. Proc. Natl Acad. Jiirgens, G., van Kammen, A., and de Vries, S.C. Sci. USA 83:2123-2127. 1996. Pattern formation in the Arabidopsis embryo revealed by position-specific lipid transfer protein Walsh, D.J., Matthews, J.A., Denmeade, R., and Walker, gene expression. Plant Cell 8:783-791. M.R. 1989. Cloning of cDNA coding for an allergen of cocksfoot grass (Dactylis glomerata) pollen. Int. Vujicic, R., Radojevic, L., and Neskovic, M. 1976. Orderly Arch. Allergy Appl. Immunol. 90:78-83. arrangement of ribosomes in the embryogenic callus tissue of Coryhis avellana L. /. Cell Biol. 69:686-692. Walter, C, Smith, D.R., Connett, M.B., Grace, L., and White, D.W.R. 1994. A biolistic approach for the transfer and expression of a gusA reporter gene in Wadsworth, G.J., and Scandalios, J.G. 1989. Differential embryogenic cultures of Pinus radiata. Plant Cell Rep. expression of the maize catalase genes during kernel 14:69-74. development: the role of steady-state mRNA levels. Devel Genet. 10:304-310. Walters, D.A., Vetsch, C.S., Potts, D.E., and Lundquist, R.C. 1992. Transformation and inheritance of a Wagner, G., and Hess, D. 1973. In yifro-Befruchtungen bei hygromycin phosphotransferase gene in maize Petunia hybrida. Z. Pflanzenphysiol. 69:262-269. plants. Plant Mol. Biol. 18:189-200. Wagner, V.T., Dumas, C, and Mogensen, H.L. 1989. Morphometric analysis of isolated Zea mays sperm. /. Walters, M.S. 1970. Evidence on the time of chromosome pairing from the preleptotene spiral stage in Lilium Cell Sci. 93:179-184. longiflorum "Croft". Chromosoma 29:375-418. Wagner, V.T., Kardolus, J.P., and van Went, J.L. 1989. Isolation of the lily embryo sac. Sex. Plant Reprod. Walters, M.S. 1985. Meiosis readiness in Lilium. Can. }. 2:219-224. Genet. Cytol. 27:33-38. Wagner, V.T., and Mogensen, H.L. 1988. The male germ Walthall, E.D., and Brady, T. 1986. The effect of the susunit in the pollen and pollen tubes of Petunia hybrida: pensor and gibberellic acid on Phaseolus vidgaris ultrastructural, quantitative and three-dimensional embryo protein synthesis. Cell Diffn 18:37-44. features. Protoplasma 143:101-110. Wan, Y., and Lemaux, P.G. 1994. Generation of large numbers of independently transformed fertile barley Wagner, V.T., Cresti, M., Salvatici, P., and Tiezzi, A. 1990. plants. Plant Physiol. 104:37-48. Changes in volume, surface area, and frequency of nuclear pores on the vegetative nucleus of tobacco Wan, Y., Sorensen, EX., and Liang, G.H. 1988. Genetic pollen in fresh, hydrated, and activated conditions. control of in vitro regeneration in alfalfa (Medicago Planta 181:304-309. sativa L.). Euphytica 39:3-9. Wagner, V.T., Song, Y.C., Matthys-Rochon, E., and Dumas, Wandelt, C, and Feix, G. 1989. Sequence of a 21 kd gene from maize containing an in-frame stop codon. Nucl. C. 1989. Observations on the isolated embryo sac of Acids Res. 17:2354. Zea mays L. Plant Sci. 59:127-132. Walbot, V. 1971. RNA metabolism during embryo devel- Wang, C, Rathore, K.S., and Robinson, K.R. 1989. The responses of pollen to applied electrical fields. Devel opment and germination of Phaseolus vulgaris. Devel Biol. 136:405-410. Biol. 26:369-379. Walbot, V. 1972. Rate of RNA synthesis and tRNA end- Wang, C.-C, Chu, C.-C, Sun, C.-S., Wu, S.-H., Yin, K.-C, and Hsu, C. 1973. The androgenesis in wheat labeling during early development of Phaseolus. (Triticum aestivum) anthers cultured in vitro. Sci. Planta 108:161-171. Sinica 16:218-222. Walbot, V. 1973. RNA metabolism in developing cotyledons of Phaseolus vulgaris. New Phytol. 72:479^183. Wang, C.-S., Walling, L.L., Eckard, K.J., and Lord, E.M. 1992a. Patterns of protein accumulation in developWalbot, V., and Dure, L., III. 1976. Developmental bioing anthers of Lilium longiflorum correlate with histochemistry of cotton seed embryogenesis and germilogical events. Am. J. Bot. 79:118-127. nation. VII. Characterization of the cotton genome. /. Mol. Biol. 101:503-536. Wang, C.-S., Walling, L.L., Eckard, K.J., and Lord, E.M. 1992b. Immunological characterization of a tapetal Walbot, V, Brady, T, Clutter, M., and Sussex, I. 1972. protein in developing anthers of Lilium longiflorum. Macromolecular synthesis during plant embryoPlant Physiol. 99:822-829. geny: rates of RNA synthesis in Phaseolus coccineus embryos and suspensors. Devel Biol. 29:104-111. Wang, C.-S., Walling, L.L., Eckard, K.J., and Lord, E.M. 1993. Characterization of an anther-specific glycoproWalburg, G., and Larkins, B.A. 1986. Isolation and charactein in Lilium longiflorum. Am. ]. Bot. 80:1155-1161. terization of cDNAs encoding oat 12S globulin mRNAs. Plant Mol. Biol. 6:161-169. Wang, F.-Y, Chen, K.-M., Wang, X.-Z., Zhou, W.-L., and
References
659
responsive promoter by the inverse polymerase Ni, J.-F. 1995. Two-dimensional gel electrophoresis chain reaction and its characterization in transgenic study on endosperm protein of the new line tobacco. Plant J. 3:191-201. obtained by introducing DNA of sorghum into wheat. Chinese ]. Bot. 7:2-8. Watanabe, M., Shiozawa, H., Isogai, A., Suzuki, A., Takeuchi, T., and Hinata, K. 1991. Existence of S-glyWang, H., and Cutler, A.J. 1995. Promoters from kinl and cor6.6, two Arabidopsis thaliana low-temperature- and coprotein-like proteins in anthers of self-incompatiABA-inducible genes, direct strong p-glucuronidase ble species of Brassica. Plant Cell Physiol. expression in guard cells, pollen and young devel32:1039-1047. oping seeds. Plant Mol. Biol. 28:619-634. Watanabe, M., Takasaki, T., Toriyama, K., Yamakawa, S., Isogai, A., Suzuki, A., and Hinata, K. 1994. A high Wang, H., Wu, H.-M., and Cheung, A.Y. 1993. degree of homology exists between the protein Development and pollination regulated accumulaencoded by SLG and the S receptor domain encoded tion and glycosylation of a stylar transmitting tissuespecific proline-rich protein. Plant Cell 5:1639-1650. by SRK in self-incompatible Brassica campestris L. Plant Cell Physiol. 35:1221-1229. Wang, H., Wu, H.-M., and Cheung, A.Y. 1996. Pollination induces mRNA poly(A) tail-shortening and cell Waterkeyn, L., and Bienfait, A. 1970. On a possible function of the callosic special wall in Ipomoea purpurea deterioration in flower transmitting tissue. Plant }. (L.) Roth. Grana 10:13-20. 9:715-727. Wang, H.L., Offler, C.E., and Patrick, J.W. 1995. The cellu- Watson, C.V., Nath, J., and Nanda, D. 1977. Possible mitolar pathway of photosynthate transfer in the develchondrial involvement in mechanism of male steriloping wheat grain. II. A structural analysis and ity in maize (Zea mays L.). Biochem. Genet. histochemical studies of the pathway from the crease 15:1113-1124. phloem to the endosperm cavity. Plant Cell Environ. Webb, M.C., and Gunning, B.E.S. 1990. Embryo sac devel18:373-388. opment in Arabidopsis thaliana. I. Megasporogenesis, including the microtubular cytoskeleton. Sex. Plant Wang, H.L., Patrick, J.W., Offler, C.E., and Wang, X.-D. 1995. The cellular pathway of photosynthate transfer Reprod. 3:244-256. in the developing wheat grain. III. A structural Webb, M.C., and Gunning, B.E.S. 1991. The microtubular analysis and physiological studies of the pathway cytoskeleton during development of the zygote, from the endosperm cavity to the starchy proembryo and free-nuclear endosperm in endosperm. Plant Cell Environ. 18:389^07. Arabidopsis thaliana (L.) Heynh. Planta 184:187-195. Wang, S.-Z., and Esen, A. 1986. Primary structure of a pro- Webb, M.C., and Gunning, B.E.S. 1994. Embryo sac develline-rich zein and its cDNA. Plant Physiol. 81:70-74. opment in Arabidopsis thaliana. II. The cytoskeleton Wang, T.L., Cook, S.K., Francis, R.J., Ambrose, M.J., and during megagametogenesis. Sex. Plant Reprod. Hedley, C.J. 1987a. An analysis of seed development 7:153-163. in Pisum sativum. VI. Abscisic acid accumulation. /. Weber, M. 1988. Formation of sperm cells in Galium mollugo Expt. Bot. 38:1921-1932. (Rubiaceae), Trichodiadema setuliferum (Aizoaceae), and Wang, T.L., Hadavizideh, A., Harwood, A., Welham, T.J., Avena saliva (Poaceae). Plant Syst. Evol. 161:53-64. Harwood, W.A., Faulks, R., and Hedley, C.L. 1990. Weber, M. 1989. Ultrastructural changes in maturing An analysis of seed development in Pisum sativum. pollen grains of Apium nodiflorum L. (Apiaceae), with XIII. The chemical induction of storage product special reference to the endoplasmic reticulum. mutants. Plant Breed. 105:311-320. Protoplasma 152:69-76. Wang, T.L., Smith, CM., Cook, S.K., Ambrose, M.J., and Weber, M. 1992. The formation of pollenkitt in Apium nodHedley, C.L. 1987b. An analysis of seed development iflorum (Apiaceae). Ann. Bot. 70:573-577. in Pisum sativum. III. The relationship between the r Weber, M. 1994. Stigma, style, and pollen tube pathway in locus, the water content and the osmotic potential of Smyrnium perfoliatum (Apiaceae). Int. ]. Plant Sci. seed tissues in vivo and in vitro. Ann. Bot. 59:73-80. 155:437-444. Wang, Y.-C, and Janick, J. 1986. In vitro production of Weber, M., and Frosch, A. 1995. The development of the jojoba liquid wax by zygotic and somatic embryos. /. transmitting tract in the pistil of Hacquetia epipactis Am. Soc. Hort. Sci. 111:798-806. (Apiaceae). Int. ]. Plant Sci. 156:615-621. Wang, Z.-Y., Wu, Z.-L., Xing, Y.-Y, Zheng, F.-G., Guo, X.- Wedderburn, F., and Richards, A.J. 1990. Variation in L., Zhang, W.-G., and Hong, M.-M. 1990. Nucleotide within-morph incompatibility inhibition sites in hetsequence of rice waxy gene. Nucl. Acids Res. 18:5898. eromorphic Primula L. Neiv Phytol. 116:149-162. Wang, Z.-Y, Zheng, F.-Q., Shen, G.-Z., Gao, J.-P., Snustad, Weeden, N.F., and Gottlieb, L.D. 1979. Distinguishing D.P., Li, M.-G., Zhang, J.-L., and Hong, M.-M. 1995. allozymes and isozymes of phosphoglucoisoThe amylose content in rice endosperm is related to merases by electrophoretic comparisons of pollen the post-transcriptional regulation of the waxy gene. and somatic tissues. Biochem. Genet. 17:287-296. Plant J. 7:613-622. Weeden, N.F., and Gottlieb, L.D. 1980. Isolation of cytoWarmke, H.E., and Lee, S.-L.J. 1977. Mitochondrial plasmic enzymes from pollen. Plant Physiol. 66:400-403. degeneration in Texas cytoplasmic male-sterile corn anthers. /. Hered. 68:213-222. Weeks, J.T., Anderson, O.D., and Blechl, A.E. 1993. Rapid production of multiple independent lines of fertile Warmke, H.E., and Lee, S.-L.J. 1978. Pollen abortion in T transgenic wheat (Triticum aestivum). Plant Physiol. cytoplasmic male-sterile corn (Zea mays): a sug102:1077-1084. gested mechanism. Science 200:561-563. Warmke, H.E., and Overman, M.A. 1972. Cytoplasmic Wehling, P., Hackauf, B., and Wricke, G. 1994. Phosphorylation of pollen proteins in relation to male sterility in Sorghum. I. Callose behavior in ferself-incompatibility in rye (Secale cereale L.). Sex. tile and sterile anthers. /. Hered. 63:102-108. Warner, S.A.J., Scott, R., and Draper, J. 1993. Isolation of Plant Reprod. 7:67-75. an asparagus intracellular PR gene (AoPRl) wound- Weigel, D., and Meyerowitz, E.M. 1993. Activation of flo-
660 ral homeotic genes in Arabidopsis. Science 261:1723-1726. Weigel, D., Alvarez,J., Smyth, D.R., Yanofsky, M.F., and Meyerowitz, E.M. 1992. LEAFY controls floral meristem identity in Arabidopsis. Cell 69:843-859. Weinand, U., Briischke, C, and Feix, G. 1979. Cloning of double stranded DNAs derived from polysomal mRNA of maize endosperm: isolation and characterisation of zein clones. Nucl. Acids Res. 6:2707-2715. Weisenseel, M.H., and Jaffe, L.F. 1976. The major growth current through lily pollen tubes enters as K+ and leaves as H+. Planta 133:1-7. Weisenseel, M.H., Nuccitelli, R., and Jaffe, L.F. 1975. Large electrical currents traverse growing pollen tubes. /. Cell Bioi 66:556-567. Weiss, C.A., Huang, H., and Ma, H. 1993. Immunolocalization of the G protein a subunit encoded by a GPA1 gene in Arabidopsis. Plant Cell 5:1513-1528. Weissinger, A.K., Timothy, D.H., Levings, C.S., HI, and Goodman, M.M. 1983. Patterns of mitochondrial DNA variation in indigenous maize races of Latin America. Genetics 104:365-379. Weissinger, A.K., Timothy, D.H., Levings, C.S., III, Hu, W.W.L., and Goodman, M.M. 1982. Unique plasmidlike mitochondrial DNAs from indigenous maize races of Latin America. Proc. Natl Acad. Sci. USA 79:1-5. Welk, M., Millington, W.F., and Rosen, W.G. 1965. Chemotropic activity and the pathway of pollen tube in lily. Am. J. Bot. 52:774-781. Welsh, J.R., and Klatt, A.R. 1971. Effects of temperature and photoperiod on spring wheat pollen viability. Crop Sci. 11:864-865. Wemmer, T., Kaufmann, H., Kirch, H.-H., Schneider, K., Lottspeich, F., and Thompson, R.D. 1994. The most abundant soluble basic protein of the stylar transmitting tract in potato (Solanum tuberosum L.) is an endochitinase. Planta 194:264-273. Wenck, A.R., Conger, B.V., Trigiano, R.N., and Sams, C.E. 1988. Inhibition of somatic embryogenesis in orchardgrass by endogenous cytokinins. Plant Physiol. 88:990-992. Weng, W.-M., Gao, X.-S., Zhuang, N.-L., Xu, M.-L., and Xue, Z.-T. 1995. The glycinin A3B4 mRNA from wild soybean Glycine soja Sieb. et. Zucc. Plant Physiol. 107:665-666. Wenzel, M., Gers-Barlag, H., Schimpl, A., and Riidiger, H. 1993. Time course of lectin and storage protein biosynthesis in developing pea {Pisum sativum) seeds. Biol. Chem. Hoppe-Seyler 374:887-894. Werr, W., Frommer, W.-B., Maas, C, and Starlinger, P. 1985. Structure of the sucrose synthase gene on chromosome 9 of Zea mays L. EMBO J. 4:1373-1380. Weselake, R.J., Pomeroy, M.K., Furukawa, T.L., Golden, J.L., Little, D.B., and Laroche, A. 1993. Developmental profile of diacylglycerol acyltransferase in maturing seeds of oilseed rape and safflower and microspore-derived cultures of oilseed rape. Plant Physiol. 102:565-571. Weselake, R.J., Taylor, D.C., Pomeroy, M.K., Lawson, S.L., and Underhill, E.W. 1991. Properties of diacylglycerol acyltransferase from microspore-derived embryos of Brassica napus. Phytochemistry 30:3533-3538. Wessler, S.R., and Varagona, MJ. 1985. Molecular basis of mutations at the waxy locus of maize: correlation with the fine structure genetic map. Proc. Natl Acad. Sci. USA 82:4177-4181.
References
West, M.A.L, Yee, K.M., Danao, J., Zimmerman, J.L., Fischer, R.L., Goldberg, R.B., and Harada, J.J. 1994. LEAFY COTYLEDON1 is an essential regulator of late embryogenesis and cotyledon identity in Arabidopsis. Plant Cell 6:1731-1745. Weterings, K., Reijnen, W., van Aarssen, R., Korstee, A., Spijkers, J., van Herpen, M., Schrauwen, J., and Wullems, G. 1992. Characterization of a pollen-specific cDNA clone from Nicotiana tabacum expressed during microgametogenesis and germination. Plant Mol. Biol. 18:1101-1111. Weterings, K., Reijnen, W., Wijn, G., van de Heuvel, K., Appeldoorn, N., de Kort, G., van Herpen, M., Schrauwen, J., and Wullems, G. 1995. Molecular characterization of the pollen-specific genomic clone NTPg303 and in situ localization of expression. Sex. Plant Reprod. 8:11-17. Wetherell, D.F., and Dougall, D.K. 1976. Sources of nitrogen supporting growth and embryogenesis in cultured wild carrot tissue. Physiol. Plant. 37:97-103. Wetherell, D.F., and Halperin, W. 1963. Embryos derived from callus tissue cultures of the wild carrot. Nature 200:1336-1337. Wetzel, C.L.R., and Jensen, W.A. 1992. Studies of pollen maturation in cotton: the storage reserve accumulation phase. Sex. Plant Reprod. 5:117-127. Wever, G.H., and Takats, S.T. 1971. Isolation and separation of S-competent and S-incompetent nuclei from Tradescantia pollen grains. Expt. Cell Res. 69:29-32. Weymann, K., Urban, K., Ellis, D.M., Novitzky, R., Dunder, E., Jayne, S., and Pace, G. 1993. Isolation of transgenic progeny of maize by embryo rescue under selective conditions. In Vitro Cell Devel Biol. 29P:33-37. Wheatley, W.G., Marsolais, A.A., and Kasha, K.J. 1986. Microspore growth and anther staging in barley anther culture. Plant Cell Rep. 5:47-49. Wheeler, C.T., and Boulter, D. 1967. Nucleic acids of developing seeds of Vicia faba L. J. Expt. Bot. 18:229-240. Whipple, A.P., and Mascarenhas, J.P. 1978. Lipid synthesis in germinating Tradescantia pollen. Phytochemistry 17:1273-1274. White, D.W.R., and Williams, E. 1976. Early seed development after crossing of Trifolium semipilosum and T. repens. New Zealand ]. Bot. 14:161-168. White, O.E. 1914. Studies of teratological phenomena in their relation to evolution and the problems of heredity. I. A study of certain floral abnormalities in Nicotiana and their bearing on theories of dominance. Am. J. Bot. 1:23-36. Whitmore, F.W., and Kriebel, H.B. 1987. Expression of a gene in Pinus strobus ovules associated with fertilization and early embryo development. Can. ]. For. Res. 17:408-412. Wiberg, E., Rahlen, L., Hellman, M., Tillberg, E., Glimelius, K., and Stymne, S. 1991. The microsporederived embryo of Brassica napus L. as a tool for studying embryo-specific lipid biogenesis and regulation of oil quality. Theor. Appl. Genet. 82:515-520. Wienand, U., Langridge, P., and Feix, G. 1981. Isolation and characterization of a genomic sequence of maize coding for a zein gene. Mol. Gen. Genet. 182:440-444. Wilde, H.D., Nelson, W.S., Booij, H., de Vries, S.C., and Thomas, T.L. 1988. Gene-expression programs in embryogenic and non-embryogenic carrot cultures. Planta 176:205-211.
References
661
1973. The activity of the nucleolus organising region Wilen, R.W., Mandel, R.M., Pharis, R.P., Holbrook, L.A., and the origin of cytoplasmic nucleoloids in meioand Moloney, M.M. 1990. Effects of abscisic acid and cytes of Lilium. Protoplasma 77:79-93. high osmoticum on storage protein gene expression in microspore embryos of Brassica napus. Plant Williams, E., and White, D.W.R. 1976. Early seed development after crossing of Trifolium ambiguum and T. Physiol. 94:875-881. repens. New Zealand J. Bot. 14:307-314. Wilen, R.W., van Rooijen, G.J.H., Pearce, D.W., Pharis, R.P., Holbrook, L.A., and Moloney, M.M. 1991. Effects of Williams, E.G. 1987. Interspecific hybridization in pasture jasmonic acid on embryo-specific processes in legumes. Plant Breed. Rev. 5:237-305. Brassica and Linum oilseeds. Plant Physiol. 95:399-405.Williams, E.G., Clarke, A.E., and Knox, R.B., ed. 1994. Genetic Control of Self-Incompatibility and Reproductive Wilkins, T.A., and Raikhel, N.V. 1989. Expression of rice Development in Flowering Plants. Dordrecht: Kluwer lectin is governed by two temporally and spatially Academic Publishers. regulated mRNAs in developing embryos. Plant Cell 1:541-549. Williams, E.G., and Heslop-Harrison, J. 1979. A comparison of RNA synthetic activity in the plasmodial and Willcox, M.C., Reed, S.C, Burns, J.A., and Wynne, J.C. secretory types of taperum during the meiotic inter1991. Effect of microspore stage and media on anther val. Phytomorphology 29:370-381. culture of peanut (Arachis hypogaea L.). Plant Cell Williams, E.G., Knox, R.B., and Rouse, J.L. 1982. Tissue Organ Cult. 24:25-28. Pollination sub-systems distinguished by pollen Willemse, M.T.M. 1972. Morphological and quantitative tube arrest after incompatible interspecific crosses in changes in the population of cell organelles during Rhododendron (Ericaceae). /. Cell Sci. 53:255-277. microsporogenesis of Gasteria verrncosa. Acta Bot. Williams, E.G., Knox, R.B., Kaul, V., and Rouse, J.L. 1984. Neerl. 21:17-31. Post-pollination callose development in ovules of Willemse, M.T.M. 1981. Polarity during megasporogeneRhododendron and Ledum (Ericaceae): zygote special sis and megagametogenesis. Phytomorphology wall. /. Cell Sci. 69:127-135. 31:124-134. Willemse, M.T.M., and Bednara, J. 1979. Polarity during Williams, E.G., Ramm-Anderson, S., Dumas, C., Mau, megasporogenesis in Gasteria verrucosa. PhytoS.L., and Clarke, A.E. 1982. The effect of isolated components of Primus avium L. styles on in vitro morphology 29:156-165. growth of pollen tubes. Planta 156:517-519. Willemse, M.T.M., and de Boer-de Jeu, M.J. 1981. Megasporogenesis and early megagametogenesis. Williamson, J.D., and Quatrano, R.S. 1988. ABA-regulation of two classes of embryo-specific sequences in Acta Soc. Bot. Polon. 50:111-120. mature wheat embryos. Plant Physiol. 86:208-215. Willemse, M.T.M., and Franssen-Verheijen, M.A.W. 1986. Stylar development in the open flower of Gasteria ver- Williamson, J.D., Quatrano, R.S., and Cuming, A.C. 1985. Em polypeptide and its messenger RNA levels are rncosa (Mill.) H. Duval. Acta Bot. Neerl. 35:297-309. modulated by abscisic acid during embryogenesis in Willemse, M.T.M., and Kapil, R.N. 1981a. Antipodals of wheat. Eur. ]. Biochem. 152:501-507. Gasteria verrucosa (Liliaceae) - An ultrastructural Williamson, J.D., and Scandalios, J.G. 1992. Differential study. Acta Bot. Neerl. 30:25-32. Willemse, M.T.M., and Kapil, R.N. 1981b. On some anomresponse of maize catalases to abscisic acid: Vpl alies during megasporogenesis in Gasteria verrucosa transcriptional activator is not required for abscisic acid-regulated Catl expression. Proc. Natl Acad. Sci. (Liliaceae). Acta Bot. Neerl. 30:439-447. USA 89:8842-8846. Willemse, M.T.M., Plyushch, T.A., and Reinders, M.C. 1995. In vitro micropylar penetration of the pollen Williamson, J.D., Galili, G., Larkins, B.A., and Gelvin, S.B. 1988. The synthesis of a 19 kilodalton zein protein in tube in the ovule of Gasteria verrucosa (Mill.) H. transgenic Petunia plants. Plant Physiol. 88:1002-1007. Duval and Lilium longiflorum Thunb.: conditions, Willing, R.P., Bashe, D., and Mascarenhas, J.P. 1988. An attraction and application. Plant Sci. 108:201-208. analysis of the quantity and diversity of messenger Willemse, M.T.M., and Reznickova, S.A. 1980. Formation RNAs from pollen and shoots of Zea mays. Theor. of pollen in the anther of Lilium. I. Development of Appl. Genet. 75:751-753. the pollen wall. Acta Bot. Neerl. 29:127-140. Willemse, M.T.M., and van Lammeren, A.A.M. 1988. Willing, R.P., and Mascarenhas, J.P. 1984. Analysis of the Structure and function of the microtubular complexity and diversity of mRNAs from pollen cytoskeleton during megasporogenesis and embryo and shoots of Tradescantia. Plant Physiol. 75:865-868. sac development in Gasteria verrucosa (Mill.) H. Willing, R.R., and Pryor, L.D. 1976. Interspecific hybridisDuval. Sex. Plant Reprod. 1:74-82. ation in poplar. Theor. Appl. Genet. 47:141—151. Willemse, M.T.M., and van Went, J.L. 1984. The female Wilms, H.J. 1980a. Ultrastructure of the stigma and style gametophyte. In Embryology of Angiosperms, ed. B.M. of spinach in relation to pollen germination and Johri, pp. 159-196. Berlin: Springer-Verlag. pollen tube growth. Acta Bot. Neerl. 29:33-47. Willemse, M.T.M., and Vletter, A. 1995. Appearance and Wilms, H.J. 1980b. Development and composition of the spinach ovule. Acta Bot. Neerl. 29:243-260. interaction of pollen and pistil pathway proteins in Gasteria verrucosa (Mill.) H. Duval. Sex. Plant Reprod. Wilms, H.J. 1981a. Ultrastructure of the developing 8:161-167. embryo sac of spinach. Acta Bot. Neerl. 30:75-99. Williams, B., and Tsang, A. 1991. A maize gene expressed Wilms, H.J. 1981b. Pollen tube penetration and fertilizaduring embryogenesis is abscisic acid-inducible and tion in spinach. Acta Bot. Neerl. 30:101-122. highly conserved. Plant Mol. Biol. 16:919-923. Wilms, H.J., Leferink-ten Klooster, H.B., and van Aelst, A.C. 1986. Isolation of spinach sperm cells: 1: Ultrastructure Williams, E., and Heslop-Harrison, J. 1975. Membraneand three-dimensional construction in the mature bounded cisternae associated with the chromosomes pollen grain. In Biotechnology and Ecology of Pollen, ed. at first meiotic metaphase in microsporocytes of D.L. Mulcahy, G.B. Mulcahy and E. Ottaviano, pp. Rhoeo spathacea. Protoplasma 86:285-289. 307-312. New York: Springer-Verlag. Williams, E., Heslop-Harrison, J., and Dickinson, H.G.
662
References
Wilson, D.R., and Larkins, B.A. 1984. Zein gene organiza- Woodard, J.W. 1958. Intracellular amounts of nucleic tion in maize and related grasses. /. Mol. Evol. acids and protein during pollen grain growth in 20:330-340. Tradescantia. J. Biophys. Biochem. Cytol. 4:383-390. Wilson, G.F., Rhodes, A.M., and Dickinson, D.B. 1973. Woodcock, C.L.F., and Bell, P.R. 1968a. The distribution of Some physiological effects of viviparous genes up, deoxyribonucleic acid in the female gametophyte of and vp5 on developing maize kernels. Plant Physiol. Myosurus minimus. Histochemie 12:289-301. 52:350-356. Woodcock, C.L.F., and Bell, P.R. 1968b. Features of the Wilson, H.J., Israel, H.W., and Steward, F.C. 1974. ultrastructure of the female gametophyte of Morphogenesis and the fine structure of cultured Myosurus minimus. ]. Ultrastr. Res. 22:546-563. carrot cells. /. Cell Sci. 15:57-73. Woodward, J.R., Bacic, A., Jahnen, W., and Clarke, A.E. 1989. N-linked glycan chains on S-allele-associated Wilson, H.M. 1977. Culture of whole barley spikes stimulates high frequencies of pollen calluses in individglycoproteins from Nicotiana alata. Plant Cell 1:511-514. ual anthers. Plant Sci. Lett. 9:233-238. Wilson, H.M., Mix, G., and Foroughi-Wehr, B. 1978. Early Woodward, J.R., Craik, D., Dell, A., Khoo, K.-H., Munro, microspore divisions and subsequent formation of S.L.A., Clarke, A.E., and Bacic, A. 1992. Structural microspore calluses at high frequency in anthers of analysis of the N-linked glycan chains from a stylar Hordeum vulgare L. /. Expt. Bot. 29:227-238. glycoprotein associated with expression of self-incompatibility in Nicotiana alata. Glycobiology 2:241-250. Wing, R.A., Yamaguchi, J., Larabell, S.K., Ursin, V.M., and McCormick, S. 1989. Molecular and genetic charac- Worrall, D., Hird, D.L., Hodge, R., Paul, W., Draper, J., terization of two pollen-expressed genes that have and Scott, R. 1992. Premature dissolution of the sequence similarity to pectate lyases of the plant microsporocyte callose wall causes male sterility in pathogen Erwinia. Plant Mol. Biol. 14:17-28. transgenic tobacco. Plant Cell 4:759-771. Wintz, H., Chen, H.-C, Sutton, C.A., Conley, C.A., Wright, D.J., and Boulter, D. 1972. The characterisation of Cobb, A., Ruth, D., and Hanson, M.R. 1995. vicilin during seed development in Vicia faba (L.). Expression of the CMS-associated urfS sequence in Planta 105:60-65. transgenic petunia and tobacco. Plant Mol. Biol. Wright, S.Y., Suner, M.-M., Bell, P.J., Vaudin, M., and 28:83-92. Greenland, A.J. 1993. Isolation and characterization of male flower cDNAs from maize. Plant ]. 3:41-49. Wise, R.R, Fliss, A.E., Pring, D.R., and Gengenbach, B.G. 1987. urfli-T of T cytoplasm maize mitochondria Wrobel, B., and Bednarska, E. 1994. Nuclear DNA content encodes a 13 kD polypeptide. Plant Mol. Biol. and ultrastructure of secretory cells of Vicia faba L. 9:121-126. stigma. Acta Soc. Bot. Polon. 63:139-145. Wise, R.R, Pring, D.R., and Gengenbach, B.G. 1987. Wu, F.S., and Murry, L.E. 1985a. Changes in protein and Mutation to male fertility and toxin insensitivity in amino acid content during anther development in Texas (T)-cytoplasm maize is associated with a fertile and cytoplasmic male sterile Petunia. Theor. frameshift in a mitochondrial open reading frame. Appl. Genet. 71:68-73. Proc. Natl Acad. Sci. USA 84:2858-2862. Wu, F.S., and Murry, L.E. 1985b. Proteolytic activity in Wobus, U., Baumlein, H., Bassiiner, R., Heim, U., Jung, R., anther extracts of fertile and cytoplasmic male sterMuntz, K., Saalbach, G., and Weschke, W. 1986. ile Petunia. Plant Physiol. 79:301-305. Characteristics of two types of legumin genes in the Wu, H.-M., Wang, H., and Cheung, A.Y. 1995. A pollen tube growth stimulatory glycoprotein is deglycosyfield bean {Vicia faba L. var. minor) genome as lated by pollen tubes and displays a glycosylation revealed by cDNA analysis. FEBS Lett. 210:74-80. gradient in the flower. Cell 82:395-403. Wochok, Z.S. 1973a. Microtubules and multivesicular bodies in cultured tissues of wild carrot: changes Wu, H.-M., Zou, J., May, B., Gu, Q., and Cheung, A.Y. during transition from the undifferentiated to the 1993. A tobacco gene family for flower cell wall proembryonic condition. Cytobios 7:87-95. teins with a proline-rich domain and a cysteine-rich domain. Proc. Natl Acad. Sci. USA 90:6829-6833. Wochok, Z.S. 1973b. DNA synthesis during development of adventive embryos of wild carrot. Biol. Plant. Wu, S.-C, Bogre, L., Vincze, E., Kiss, G.B., and Dudits, D. 15:107-111. 1988. Isolation of an alfalfa histone H3 gene: structure and expression. Plant Mol. Biol. 11:641-649. Wochok, Z.S., and Wetherell, D.F. 1972. Restoration of declining morphogenetic capacity in long term tis- Wu, X.-L., and Zhou, C. 1990. Nuclear divisions of isosue cultures of Daucus carota by kinetin. Experientia lated, in vitro cultured generative cells in Hemerocallis 28:104-105. minor Mill. Acta Bot. Sinica 32:577-581. Wodehouse, R.P. 1935. Pollen Grains, Their Structure, Wu, X.-L., and Zhou, C. 1991a. A comparative study on Identification and Significance in Science and Medicine. methods for isolation of generative cells in various angiosperm species. Acta Biol. Expt. Sinica 24:15-23. New York: McGraw-Hill Book Co. Woittiez, R.D., and Willemse, M.T.M. 1979. Sticking of Wu, X.-L., and Zhou, C. 1991b. Fusion experiments of isopollen on stigmas: the factors and a model. lated generative cells in several angiosperm species. Phytomorphology 29:57-63. Acta Bot. Sinica 33:897-904. Wong, K.C., Watanabe, M., and Hinata, K. 1994a. Protein Wu, Y., and Zhou, C. 1988. Observation on enzymatically profiles in pin and thrum floral organs of distylous isolated-embryo sacs in Adenophora axilliflora: fertilAverrhoa carambola L. Sex. Plant Reprod. 7:107-115. ization, zygote and endosperm. Acta Bot. Sinica Wong, K.C., Watanabe, M., and Hinata, K. 1994b. 30:210-212. Fluorescence and scanning electron microscopic Wu, Y, and Zhou, C. 1990. An ultrastructural study on study on self-incompatibility in distylous Averrhoa pollen protoplasts and pollen tubes germinated from them in Gladiolus gandavensis. Acta Bot. Sinica carambola L. Sex. Plant Reprod. 7:116-121. 32:493^198. Woodard, J.W. 1956. DNA in gametogenesis and embryogeny in Tradescantia. J. Biophys. Biochem. Cytol. Wu, Y., and Zhou, C. 1992. An ultrastructural study on triggering of cell division in young pollen protoplast 2:765-776.
References
663
during development. /. Expt. Bot. 46:687-694. culture of Hemerocaliis fulva L. Ada Bot. Sinica 34:20-25. Xu, N., Coulter, K.M., and Bewley, J.D. 1990. Abscisic acid and osmoticum prevent germination of developing Wu, Y., Zhou, C, and Koop, H.-U. 1993. Enzymatic isolaalfalfa embryos, but only osmoticum maintains the tion of viable nucelli at the megaspore mother cell synthesis of developmental proteins. Planta stage and in developing embryo sacs in Nicotiana 182:382-390. tabacum. Sex. Plant Reprod. 6:171-175. Wu, Y., Haberland, G., Zhou, C, and Koop, H.-U. 1992. Xu, N., Coulter, K.M., Krochko, J.E., and Bewley, J.D. 1991. Somatic embryogenesis, formation of morphoMorphological stages and storage protein accumulagenetic callus and normal development in zygotic tion in developing alfalfa (Medicago sativa L.) seeds. embryos of Arabidopsis thaliana in vitro. Protoplasma Seed Sci. Res. 1:119-125. 169:89-96. Xu, S.-X. (S.Y. Zee). 1991. Changes in microtubule cytoskeleton of generative cells during mitosis and Wulff, H.D. 1935. Galvanotropismus bei Pollencytokinesis. Chinese J. Bot. 3:77-80. schlauchen. Planta 24:602-608. Wunderlich, R. 1954. Uber das Antherentapetum mit Xu, S.-X. (S.Y. Zee), Zhu, C, and Hu, S.-Y. 1990. Changes besonderer Berucksichtigung seiner Kernzahl. in the organization of microtubules during generative cell division and sperm formation in Lilium. Ada Osterr. Z. Bot. 101:1-63. Bot. Sinica 32:821-826. Wurtele, E.S., and Nikolau, B.J. 1992. Differential accumulation of biotin enzymes during carrot somatic Xu, S.-X. (S.Y. Zee), Zhu, C, and Hu, S.-Y. 1993. embryogenesis. Plant Physiol. 99:1699-1703. Microtubule-cytoskeleton changes during generative cell division and sperm formation in the pollen Wurtele, E.S., Wang, H., Durgerian, S., Nikolau, B.J., and tubes of Zantedeschia aethiopica. Chinese J. Bot. Ulrich, T.H. 1993. Characterization of a gene that is 5:18-22. expressed early in somatic embryogenesis of Daucus carota. Plant Physiol. 102:303-312. Xu, X., and Li, B. 1994. Fertile transgenic indica rice plants obtained by electroporation of the seed embryo cells. Xiao, C.-M., and Mascarenhas, J.P. 1985. High temperaPlant Cell Rep. 13:237-242. ture-induced thermotolerance in pollen tubes of Xu, Z.H., Huang, B., and Sunderland, N. 1981. Culture of Tradescantia and heat-shock proteins. Plant Physiol. barley anthers in conditioned media. /. Expt. Bot. 78:887-890. 32:767-778. Xu, B., Grun, P., Kheyr-Pour, A., and Kao, T.-H. 1990a. Xue, J., Pihlgren, U., and Rask, L. 1993. Temporal, cell-speIdentification of pistil-specific proteins associated cific, and tissue-preferential expression of myrosiwith three self-incompatibility alleles in Solarium chanase genes during embryo and seedling coense. Sex. Plant Reprod. 3:54-60. development in Sinapis alba. Planta 191:95-101. Xu, B., Mu, J., Nevins, D.L., Grun, P., and Kao, T.-H. Xue, Y., Carpenter, R., Dickinson, H.G., and Coen, E.S. 1990b. Cloning and sequencing of cDNAs encoding 1996. Origin of allelic diversity in Antirrhinum S two self-incompatibility associated proteins in locus RNases. Plant Cell 8:805-814. Solatium chacoense. Mol. Gen. Genet. 224:341-346. Xue, Y., Collin, S., Davies, D.R., and Thomas, CM. 1994. Differential screening of mitochondrial cDNA Xu, R, Yao, Y.-X., Fu, C.-S., and Li, Z.-L. 1993. Studies on libraries from male-fertile and cytoplasmic malethe development of pollen wall in Pyrus betulaefolia sterile sugar-beet reveals genome rearrangements at Bge. (Rosaceae). Chinese ]. Bot. 5:41-46. atp6 and atpA loci. Plant Mol. Biol. 25:91-103. Xu, H., Davies, S.P., Kwan, B.Y.H., O'Brien, A.P., Singh, M., and Knox, R.B. 1993. Haploid and diploid Xue, Z.-T., Xu, M.-L., Shen, W., Zhuang, N.L., Hu, W.-M., expression of a Brassica campestris anther-specific and Shen, S.C. 1992. Characterization of a Gy4 gene promoter in Arabidopsis and tobacco. Mol. Gen. glycinin gene from soybean Glycine max cv. Forrest. Genet. 239:58-65. Plant Mol. Biol. 18:897-908. Xu, H., Knox, R.B., Taylor, P.E., and Singh, M.B. 1995a. XuHan, X., and van Lammeren, A.A.M. 1993. Bcpl, a gene required for male fertility in Arabidopsis. Microtubular configurations during the cellularizaProc. Natl Acad. Sci. USA 92:2106-2110. tion of coenocytic endosperm in Ranunculus sceleraXu, H., Theerakulpisut, P., Taylor, P.E., Knox, R.B., Singh, tus L. Sex. Plant Reprod. 6:127-132. M.B., and Bhalla, P.L. 1995b. Isolation of a gene pref- XuHan, X., and van Lammeren, A.A.M. 1994. erentially expressed in mature anthers of rice (Oryza Microtubular configurations during endosperm sativa L.). Protoplasma 187:127-131. development in Phaseolus vulgaris. Can. }. Bot. 72:1489-1495. Xu, N., and Bewley, J.D. 1991. Sensitivity to abscisic acid and osmoticum changes during embryogenesis of alfalfa (Medicago sativa).}. Expt. Bot. 42:821-826. Yabe, N., Takahashi, T., and Komeda, Y. 1994. Analysis of tissue-specific expression of Arabidopsis thaliana Xu, N., and Bewley, J.D. 1995a. Embryos of alfalfa HSP90-family gene HSP81. Plant Cell Plnjsiol. (Medicago sativa L.) can synthesize storage proteins 35:1207-1219. after the completion of germination, in response to abscisic acid and osmoticum, if maturation drying is Yadegari, R., de Paiva, G.R., Laux, T., Koltunow, A.M., prevented. Planta 196:469-476. Apuya, N., Zimmerman, J.L., Fischer, R.L., Harada, J.J., and Goldberg, R.B. 1994. Cell differentiation and Xu, N., and Bewley, J.D. 1995b. Temporal and nutritional morphogenesis are uncoupled in Arabidopsis raspfactors modulate responses to abscisic acid and berry embryos. Plant Cell 6:1713-1729. osmoticum in their regulation of storage protein synthesis in developing seeds of alfalfa (Medicago Yaklich, R.W., and Herman, E.M. 1995. Protein storage vacuoles of soybean aleurone cells accumulate a sativa L.). /. Expt. Bot. 46:675-686. unique glycoprotein. Plant Sci. 107:57-67. Xu, N., and Bewley, J.D. 1995c. The role of abscisic acid in germination, storage protein synthesis and desicca- Yaklich, R.W., Vigil, E.L., Erbe, E.F., and Wergin, W.P. 1992. The fine structure of aleurone cells in the soytion tolerance in alfalfa (Medicago sativa L.) seeds, as bean seed coat. Protoplasma 167:108-119. shown by inhibition of its synthesis by fluridone
664
References
Yakovlev, M.S., and Yoffe, M.D. 1957. On some peculiarimicroscopic study on in vitro parthenogenesis in ties in the embryogeny of Paeonia L. Phytomorphology sunflower. Sear. Plant Reprod. 2:154-166. 7:74-82. Yan, H., Yang, H.-Y., and Jensen, W.A. 1991a. Yamada, Y. 1965. Studies on the histological and cytologiUltrastructure of the micropyle and its relationship cal changes in the tissues of pistil after pollination. to pollen tube growth and synergid degeneration in Jap. J. Bot. 19:69-82. sunflower. Sex. Plant Reprod. 4:166-175. Yamagata, H., and Tanaka, K. 1986. The site of synthesis Yan, H., Yang, H.-Y, and Jensen, W.A. 1991b. and accumulation of rice storage proteins. Plant Cell Ultrastructure of the developing embryo sac of sunPhysiol. 27:135-145. flower (Helianthus annuus) before and after fertilizaYamagata, H., Sugimoto, T., Tanaka, K., and Kasai, Z. tion. Can. J. Bot. 69:191-202. 1982. Biosynthesis of storage proteins in developing Yan, L., Wang, X., Teng, X., Ma, Y, and Liu, G. 1986. Actin rice seeds. Plant Physiol. 70:1094-1100. and myosin in pollens and their role in the growth of Yamagata, H., Tamura, K., Tanaka, K., and Kasai, Z. 1986. pollen tubes. Chinese Sci. Bull. 31:267-272. Cell-free synthesis of rice prolamin. Plant Cell Yang, C, Xing, L., and Zhai, Z. 1992. Intermediate filaPhysiol. 27:1419-1422. ments in higher plant cells and their assembly in a cell-free system. Protoplasma 171:44-54. Yamaguchi, H., and Kakiuchi, H. 1983. Electrophoretic analysis of mitochondrial DNA from normal and Yang, H., Zhou, C, Cai, D., Yan, H., Wu, Y., and Chen, X. male sterile cytoplasms in rice. Jap. J. Genet. 1986. In vitro culture of unfertilized ovules in 58:607-611. Helianthus annuus L. In Haploids of Higher Plants In Yamaguchi-Shinozaki, K., Mundy, J., and Chua, N.-H. Vitro, ed. H. Hu and H. Yang, pp. 182-191. Beijing: 1989. Four tightly linked rab genes are differentially China Academic Publishers. expressed in rice. Plant Mol. Biol. 14:29-39. Yang, H.-Y., and Zhou, C. 1979. Experimental researches Yamaguchi-Shinozaki, K., Mino, M, Mundy, J., and Chua, on the two pathways of pollen development in N.-H. 1990. Analysis of an ABA-responsive rice gene Oryza sativa L. Ada Bot. Sinica 21:345-351. promoter in transgenic tobacco. Plant Mol. Biol. Yang, H.Y., and Zhou, C. 1982. In vitro induction of hap15:905-912. loid plants from unpollinated ovaries and ovules. Yamakawa, S., Shiba, H., Watanabe, M., Shiozawa, H., Theor. Appl. Genet. 63:97-104. Takayama, S., Hinata, K., Isogai, A., and Suzuki, A. Yang, H.-Y., and Zhou, C. 1984. Observations on megas1994. The sequences of S-glycoproteins involved in porogenesis and megagametophyte development in self-incompatibility of Brassica campestris and their Paulownia sp. and Sesamum indicum by enzymatic distribution among Brassicaceae. Biosci. Biotech. maceration technique. Acta Bot. Sinica 26:355-358. Biochem. Japan 58:921-925. Yang, H.-Y., and Zhou, C. 1989. Isolation of viable sperms Yamakawa, S., Watanabe, S., Hinata, K., Suzuki, A., and from pollen of Brassica napus, Zea mays and Secale Isogai, A. 1995. The sequences of S-receptor kinases cereale. Chinese J. Bot. 1:80-84. (SRK) involved in self-incompatibility and their Yang, L.J., Barratt, D.H.P., Domoney, C, Hedley, C.L., and homologies to S-locus glycoproteins of Brassica Wang, T.L. 1990. An analysis of seed development in campestris. Biosci. Biotech. Biochem. Japan 59:161-162. Pisum sativum. X. Expression of storage protein genes in cultured embryos. /. Expt. Bot. 41:283-288. Yamakawa, S., Watanabe, M., Isogai, A., Takayama, S., Satoh, S., Hinata, K., and Suzuki, A. 1993. The cDNA Yano, M., Isono, Y., Satoh, H., and Omura, T. 1984. Gene sequence of NS3-glycoprotein from Brassica analysis of sugary and shrunken mutants of rice, campestris and its homology to related proteins. Plant Oryza sativa L. Jap. J. Breed. 34:43-49. Cell Physiol. 34:173-175. Yano, M., Okuno, K., Kawakami, J., Satoh, H., and Omura, T. 1985. High amylose mutants of rice, Oryza Yamamoto, S., Nishihara, M., Morikawa, H., Yamauchi, D., and Minamikawa, T. 1995. Promoter analysis of sativa L. Theor. Appl. Genet. 69:253-257. seed storage protein genes from Canavalia gladiata Yanofsky, M.F., Ma, H., Bowman, J.L., Drews, G.N., D.C. Plant Mol. Biol. 27:729-741. Feldmann, K.A., and Meyerowitz, E.M. 1990. The Yamauchi, D., and Minamikawa, T. 1986. In vivo studies protein encoded by the Arabidopsis homeotic gene agamous resembles transcription factors. Nature on protein synthesis in developing seeds of 346:35-39. Canavalia gladiata D.C. Plant Cell Physiol. 27:1033-1041. Yao, J.-L., Wu, J.-H., Gleave, A.P., and Morris, B.A.M. 1996. Transformation of Citrus embryogenic cells Yamauchi, D., and Minamikawa, T. 1987. Synthesis of using particle bombardment and production of canavalin and concanavalin A in maturing Canavalia transgenic embryos. Plant Sci. 113:175-183. gladiata seeds. Plant Cell Physiol. 28:421^30. Yamauchi, D., and Minamikawa, T. 1990. Structure of the Yenofsky, R.L., Fine, M., and Liu, C. 1988. Isolation and gene encoding concanavalin A from Canavalia gladicharacterization of a soybean (Glycine max) lipoxygeata and its expression in Escherichia coli cells. FEBS nase-3 gene. Mol. Gen. Genet. 211:215-222. Lett. 260:127-130. Yeung, E.C. 1980. Embryogeny of Phaseolus: the role of the suspensor. Z. Pflanzenphysiol. 96:17-28. Yamauchi, D., Nakamura, K., Asahi, T., and Minamikawa, T. 1988. cDNAs for canavalin and concanavalin A Yeung, E.C, and Brown, D.C.W. 1982. The osmotic envifrom Canavalia gladiata seeds. Nucleotide sequence ronment of developing embryos of Phaseolus vulof cDNA for canavalin and RNA blot analysis of garis. Z. Pflanzenphysiol. 106:149-156. canavalin and concanavalin A mRNAs in develop- Yeung, E.C., and Cavey, M.J. 1988. Cellular endosperm ing seeds. Em. J. Biochem. 170:515-520. formation in Phaseolus vulgaris. I. Light and scanning electron microscopy. Can. J. Bot. 66:1209-1216. Yamaura, A. 1933. Karyologische und embryologische Studien iiber einige Bambws-Arten. Bot. Mag. Tokyo Yeung, E.C, and Clutter, M.E. 1978. Embryogeny of Phaseolus coccineus: growth and microanatomy. 47:551-555. Protoplasma 94:19^0. Yan, H., Yang, H.-Y, and Jensen, W.A. 1989. An electron
References
665
5:169-181. Yeung, E.C., and Clutter, M.E. 1979. Embryogeny of Phaseolus coccineus: the infrastructure and develop-Yu, H.-S., and Russell, S.D. 1993. Three-dimensional ultrament of the suspensor. Can. ]. Bot. 57:120-136. structure of generative cell mitosis in the pollen tube of Nicotiana tabacum. Eur. ]. Cell Biol. 61:338-348. Yeung, E.C., and Meinke, D.W. 1993. Embryogenesis in angiosperms: development of the suspensor. Plant Yu, H.-S., and Russell, S.D. 1994a. Populations of plastids Cell 5:1371-1381. and mitochondria during male reproductive cell maturation in Nicotiana tabacum L.: a cytological Yeung, E.C., and Sussex, I.M. 1979. Embryogeny of basis for occasional biparental inheritance. Planta Phaseolus coccineus: the suspensor and the growth of 193:115-122. the embryo-proper in vitro. Z. Pflanzenphysiol. 91:423-433. Yu, H.-S., and Russell, S.D. 1994b. Male reproductive cell development in Nicotiana tabacum: male germ unit Yeung, E.C., Thorpe, T.A., and Jensen, C.J. 1981. In vitro associations and quantitative cytology during sperm fertilization and embryo culture. In Plant Tissue maturation. Sex. Plant Reprod. 7:324-332. Culture. Methods and Applications in Agriculture, ed. T.A. Thorpe, pp. 253-271. New York: Academic Yuffd, A.M., de Garcia, E.G., and Nieto, M.S. 1994. Press. Comparative study of protein electrophoretic patYibrah, H.S., Manders, G., Clapham, D.H., and von terns during embryogenesis in Coffea arabica cv Arnold, S. 1994. Biological factors affecting transient Catimor. Plant Cell Rep. 13:197-202. transformation in embryogenic suspension cultures Yunes, J.A., Neto, G.C., da Silva, M.J., Leite, A., Ottoboni, of Picea abies. J. Plant Physiol. 144:472-478. L.M.M., and Arruda, P. 1994. The transcriptional Yin, Y, and Beachy, R.N. 1995. The regulatory regions of activator opaque2 recognizes two different target the rice tungro bacilliform virus promoter and intersequences in the 22-kD-like a-prolamin gene. Plant acting nuclear factors in rice (Oryza sativa L.). Plant ]. Cell 6:237-249. 7:969-980. Ylstra, B., Busscher, J., Franken, ]., Hollman, P.C.H., Mol, Zachgo, S., de Andrade Silva, E., Motte, P., Trobner, W., J.N.M., and van Tunen, A.J. 1994. Flavonols and ferSaedler, H., and Schwarz-Sommer, Z. 1995. tilization in Petunia hybrida: localization and mode of Functional analysis of the Antirrhinum floral homeotic DEF1CIENS gene in vivo and in vitro by action during pollen tube growth. Plant ]. 6:201-212. using a temperature-sensitive mutant. Development Ylstra, B., Touraev, A., Benito Moreno, R.M., Stoger, E., 121:2861-2875. van Tunen, A.J., Vicente, O., Mol, J.N.M., and Heberle-Bors, E. 1992. Flavonols stimulate develop- Zaki, M., and Dickinson, H. 1995. Modification of cell development in vitro: the effect of colchicine on ment, germination, and tube growth of tobacco anther and isolated microspore culture in Brassica pollen. Plant Physiol. 100:902-907. napus. Plant Cell Tissue Organ Cult. 40:255-270. Yokota, E., McDonald, A.R., Liu, B., Shimmen, T, and Palevitz, B.A. 1995. Localization of a 170 kDa myosin Zaki, M., and Kuijt, J. 1994. Ultrastructural studies on the heavy chain in plant cells. Protoplasma 185:178-187. embryo sac of Viscum minimum. II. Megagametogenesis. Can. ]. Bot. 72:1613-1628. Yoo, J., and Jung, G. 1995. DNA uptake by imbibition and expression of a foreign gene in rice. Physiol. Plant. Zaki, M., and Kuijt, J. 1995. Ultrastructural studies on the 94:453-459. embryo sac of Viscum minimum. I. MegasporoYou, R., and Jensen, W.A. 1985. Ultrastructural observagenesis. Protoplasma 185:93-105. tions of the mature megagametophyte and the fertil- Zaki, M.A.M. and Dickinson, H.G. 1990. Structural ization in wheat (Triticum aestivum). Can. ]. Bot. changes during the first divisions of embryos result63:163-178. ing from anther and free microspore culture in Brassica napus. Protoplasma 156:149-162. Youle, R.J., and Huang, A.H.C. 1981. Occurrence of low molecular weight and high cysteine containing albu- Zaki, M.A.M., and Dickinson, H.G. 1991. Microsporemin storage proteins in oilseeds of diverse species. derived embryos in Brassica: the significance of diviAm. J. Bot. 68:44-48. sion symmetry in pollen mitosis I to embryogenic development. Sex. Plant Reprod. 4:48-55. Young, E.G., and Hanson, M.R. 1987. A fused mitochondrial gene associated with cytoplasmic male sterility ZiSrsky, V., RihovA, L., and Tupy, J. 1987. Interference of pollen diffusible substances with peroxidase catis developmentally regulated. Cell 41:41-49. alyzed reaction. Plant Sci. 52:29-32. Young, L.C.T., and Stanley, R.G. 1963. Incorporation of tritiated nucleosides thymidine, uridine and cytidine Zarsky, V., Capkova, V., HrabetovS, E., and TupyJ. 1985. Protein changes during pollen development in in nuclei of germinating pine pollen. Nucleus Nicotiana tabacum L. Biol. Plant. 27:438-444. 6:83-90. Yu, H.-S., Hu, S.-Y, and Russell, S.D. 1992. Sperm cells in Zarsky, V., Garrido, D., Eller, N., Tupy, J., Vicente, O., Schoffl, F., and Heberle-Bors, E. 1995. The expression pollen tubes of Nicotiana tabacum L.: three-dimenof a small heat shock gene is activated during inducsional reconstruction, cytoplasmic diminution, and tion of tobacco pollen embryogenesis by starvation. quantitative cytology. Protoplasma 168:172-183. Plant Cell Environ. 18:139-147. Yu, H.-S., Hu, S.-Y, and Zhu, C. 1989. Ultrastructure of sperm cells and the male germ unit in pollen tubes Zarsky, V, Garrido, D., Rfhova, L., Tupy, J., Vicente, O., and Heberle-Bors. E. 1992. Derepression of the cell of Nicotiana tabacum. Protoplasma 152:29-36. cycle by starvation is involved in the induction of Yu, H.-S., Huang, B.-Q., and Russell, S.D. 1994. tobacco pollen embryogenesis. Sex. Plant Reprod. Transmission of male cytoplasm during fertilization 5:189-194. in Nicotiana tabacum. Sex. Plant Reprod. 7:313-323. Yu, H.-S., and Russell, S.D. 1992. Male cytoplasmic Zee, S.Y. 1992. Confocal laser scanning microscopy of diminution and male germ unit in young and microtubule organizational changes in isolated genmature pollen of Cymbidium goeringii: a 3-dimenerative cells of Allemanda neriifolia during mitosis. sional and quantitative study. Sex. Plant Reprod. Sex. Plant Reprod. 5:182-188.
666
References
Zee, S.Y., and Aziz-Un-Nisa. 1991. Mitosis and micro- Zhang, H.-Q., Croes, A.F., and Linskens, H.F. 1982. tubule organizational changes in isolated generative Protein synthesis in germinating pollen of Petunia: cells of Allemanda neriifolia. Sex. Plant Reprod. role of proline. Planta 154:199-203. 4:132-137. Zhang, H.-Q., Bohdanowicz, ]., Pierson, E.S., Li, Y.-Q., Zee, S.-Y., and O'Brien, T.P. 1971. Aleurone transfer cells Tiezzi, A., and Cresti, M. 1995. Microtubular organiand other structural features of the spikelet of millet. zation during asymmetrical division of the generaAust. ]. Biol. Sci. 24:391-395. tive cell in Gagea lutea. J. Plant Res. 108:269-276. Zee, S.-Y., and Siu, I.H.P. 1990. Studies on the ontogeny of Zhang, J.Z., Gomez-Pedrozo, M., Baden, C.S., and the pollinium of a massulate orchid (Peristylus spiHarada, J.J. 1993. Two classes of isocitrate lyase ranthes). Rev. Palaeobot. Palynol. 64:159-164. genes are expressed during late embryogeny and postgermination in Brassica napus L. Mol. Gen. Genet. Zee, S.-Y, Wu, S.C., and Yue, S.B. 1979. Morphological 238:177-184. and SDS-polyacrylamide gel electrophoretic studies of pro-embryoid formation in the petiole explants of Zhang, J.Z., Laudencia-Chingcuanco, D.L., Comai, L., Li, Chinese celery. Z. Pflanzenphysiol. 95:397-403. M., and Harada, J.J. 1994. Isocitrase lyase and malate synthase genes from Brassica napus L. are active in Zee, S.-Y, and Ye, X.L. 1995. Changes in the pattern of pollen. Plant Physiol. 104:857-864. organization of the microtubular cytoskeleton during megasporogenesis in Cymbidium sinense. Zhang, J.Z., Santes, CM., Engel, M.L., Gasser, C.S., and Protoplasma 185:170-177. Harada, J.J. 1996. DNA sequences that activate isocitrate lyase gene expression during late embryogeneZeijlemaker, F.C.J. 1956. Growth of pollen tubes in vitro sis and during postgerminative growth. Plant and their reaction on potential differences. Acta Bot. Physiol. 110:1069-1079. Neerl. 5:179-186. Zenkteler, M. 1962. Microsporogenesis and tapetal devel- Zhang, S., Chen, L., Qu, R., Marmey, P., Beachy, R., and opment in normal and male-sterile carrots (Daucus Fauquet, C. 1996. Regeneration of fertile transgenic indica (group 1) rice plants following microprojectile carota). Am. ]. Bot. 49:341-348. transformation of embryogenic suspension culture Zenkteler, M. 1967. Test-tube fertilization of ovules in cells. Plant Cell Rep. 15:465-469. Melandrium album Mill, with pollen grains of several species of the Caryophyllaceae family. Experientia Zhang, W., McElroy, D., and Wu, R. 1991. Analysis of rice 23:775. Actl 5' region activity in transgenic rice plants. Plant Cell 3:1155-1165. Zenkteler, M. 1990. In vitro fertilization and wide hybridization in higher plants. Crit. Rev. Plant Sci. Zhang, X.S., and O'Neill, S.D. 1993. Ovary and gameto9:267-297. phyte development are coordinately regulated by auxin and ethylene following pollination. Plant Cell Zenkteler, M., Maheswaran, G., and Williams, E.G. 1987. 5:403-418. In vitro placental pollination in Brassica campestris Zhao, J.-P., Simmonds, D.H., and Newcomb, W. 1996. and B. napus. ]. Plant Physiol. 128:245-250. Induction of embryogenesis with colchicine instead Zenkteler, M., Misiura, E., and Guzowska, 1.1975. Studies of heat in microspores of Brassica napus L. cv. Topas. on obtaining hybrid embryos in test tubes. In Form, Planta 198:433-439. Structure and Function in Plants, ed. H.Y. Mohan Ram, J.J. Shaw and C.K. Shaw, pp. 180-187. Meerut: Zhao, Y, Leisy, D.J., and Okita, T.W. 1994. Tissue-specific expression and temporal regulation of the rice Sarita Prakashan. glutelin Gt3 gene are conferred by at least two spaZenkteler, M., and Nitzsche, W. 1984. Wide hybridization tially separated ris-regulatory elements. Plant Mol. experiments in cereals. Theor. Appl. Genet. 68:311-315. Biol. 25:429^136. Zenkteler, M., and Nitzsche, W. 1985. In vitro culture of ovules of Triticum aestiviim at early stages of embryo- Zheng, Y, He, M., Hao, S., and Huang, B. 1992. The ultragenesis. Plant Cell Rep. 4:168-171. structural evidence on the origin of protein bodies in the rough endoplasmic reticulum of developing Zhan, X.-Y, Wu, H.-M., and Cheung, A.Y. 1996. Nuclear cotyledons of soybean. Ann. Bot. 69:377-383. male sterility induced by pollen-specific expression Zheng, Y.-Z., He, M.-Y, Hu, A.-L., and Hao, S. 1990. of a ribonuclease. Sex. Plant Reprod. 9:35-43. Patterns of the transition of vacuoles into protein Zhang, R, and Boston, R.S. 1992. Increases in binding probodies in developing cotyledon cells of soybean. tein (BiP) accompany changes in protein body morActa Bot. Sinica 32:97-102. phology in three high-lysine mutants of maize. Zheng, Z., Kawagoe, Y, Xiao, S., Li, Z., Okita, T, Hau, Protoplasma 171:142-152. T.L., Lin, A., and Murai, N. 1993. 5' distal and proxiZhang, G., Gifford, D.J., and Cass, D.D. 1993. RNA and mal cis-acting regulator elements are required for protein synthesis in sperm cells isolated from Zea mays L. pollen. Sex. Plant Reprod. 6:239-243. developmental control of a rice seed storage protein glutelin gene. Plant}. 4:357-366. Zhang, G., Campenot, M.K., McGann, L.E., and Cass, D.D. 1992a. Flow cytometric characteristics of sperm Zheng, Z., Sumi, K., Tanaka, K., and Murai, N. 1995. The bean seed storage protein fJ-phaseolin is synthecells isolated from pollen of Zea mays L. Plant Physiol. sized, processed, and accumulated in the vacuolar 99:54-59. type-II protein bodies of transgenic rice endosperm. Zhang, G., Williams, CM., Campenot, M.K., McGann, Plant Physiol. 109:777-786. L.E., and Cass, D.D. 1992b. Improvement of longevity and viability of sperm cells isolated from Zhong, H., Bolyard, M.G., Srinivasan, C, and Sticklen, pollen of Zea mays L. Plant Physiol. 100:47-53. M.B. 1993. Transgenic plants of turfgrass (Agrostis palustris Huds.) from microprojectile bombardment Zhang, G., Williams, CM., Campenot, M.K., McGann, of embryogenic callus. Plant Cell Rep. 13:1-6. L.E., Cutler, A.J., and Cass, D.D. 1995. Effects of calcium, magnesium, potassium and boron on sperm Zhong, H., Zhang, S., Warkentin, D., Sun, B., Wu, T., Wu, R., and Sticklen, M.B. 1996. Analysis of the funccells isolated from pollen of Zea mays L. Sex. Plant tional activity of the 1.4-kb 5-region of the rice actin Reprod. 8:113-122.
References
667
embryogenesis and analysis of peroxidase in cul1 gene in stable transgenic plants of maize (Zea mays tured lettuce (Lactuca sativa L.) cotyledons. Ann. Bot. L.). Plant Sci. 116:73-84. 69:97-100. Zhou, C. 1987a. A study of fertilization events in living embryo sacs isolated from sunflower ovules. Plant Zhu, C, and Liu, B. 1990. Microtubules in generative and Sci. 52:147-151. sperm cells of Amaryllis pollen tubes. Chinese J. Bot. 2:1-6. Zhou, C. 1987b. Cell-biological studies on artificially isolated generative cells from angiosperm pollen. Ada Zhu, C, Hu, S., Xu, L., Li, X., and Shen, J. 1980. Ultrastructure of sperm cell in mature pollen grain Bot. Sinica 29:117-122. of wheat. Sri. Sinica 23:371-379. Zhou, C. 1988. Isolation and purification of generative cells from fresh pollen of Viciafaba L. Plant Cell Rep. Zhu, J.-K., Damsz, B., Kononowicz, A.K., Bressan, R.A., 7:107-110. and Hasegawa, P.M. 1994. A higher plant extracellular vitronectin-like adhesion protein is related to the Zhou, C. 1989a. A study on isolation and culture of pollen translational elongation factor-la. Plant Cell protoplasts. Plant Sci. 59:101-108. 6:393^04. Zhou, C. 1989b. Cell divisions in pollen protoplast culture Zhu, Q., Doemer, P.W., and Lamb, C.J. 1993. Stress inducof Hemerocallis fulva L. Plant Sci. 62:229-235. tion and developmental regulation of a rice chitinase Zhou, C, and Wu, X.-L. 1990. Mass isolation and purifipromoter in transgenic tobacco. Plant J. 3:203-212. cation of generative cells from pollen grains. Ada Zhu, T., Ma, Z.-L., and Li, W.-T. 1990. Ultrastructural Bot. Sinica 32:404-406. localization of adenosine triphosphatase activity in Zhou, C, and Yang, H.-Y. 1980. Anther culture and androstigma cells of Populus lasiocarpa and its changes durgenesis of Hordeum vulgare L. Ada Bot. Sinica 22:211-215. ing development. Ada Bot. Sinica 32:91-96. Zhou, C, and Yang, H.-Y. 1982. Enzymatic isolation of Zhu, X, Mogensen, H.L., and Smith, S.E. 1990. Generative cell composition and its relation to male plastid embryo sacs in angiosperms: isolation and microinheritance patterns in Medicago sativa. Protoplasma scopical observation on fixed materials. Ada Bot. Sinica 24:403-407. 158:66-72. Zhou, C, and Yang, H.Y. 1985. Observations on enzymat- Zhu, T., Mogensen, H.L., and Smith, S.E. 1991. Quantitative cytology of the alfalfa generative cell ically isolated, living and fixed embryo sacs in sevand its relation to male plastid inheritance patterns eral angiosperm species. Planta 165:225-231. in three genotypes. Theor. Appl. Genet. 81:21-26. Zhou, C, and Yang, H.Y. 1991. Microtubule changes during the development of generative cells in Zhu, T., Mogensen, H.L., and Smith, S.E. 1992. Heritable paternal cytoplasmic organelles in alfalfa sperm Hippeastrum vittatum pollen. Sex. Plant Reprod. 4:293-297. cells: ultrastructural reconstruction and quantitative cytology. Eur. J. Cell Biol. 59:211-218. Zhou, C, Yang, H.-Y, and Xu, S.-X. 1990. Fluorescence microscopic observations on actin filament distribu- Zhu, T., Mogensen, H.L., and Smith, S.E. 1993. tion in corn pollen and Gladiolus pollen protoplasts. Quantitative, three-dimensional analysis of alfalfa Ada Bot. Sinica 32:657-662. egg cells in two genotypes: implications for biparental plastid inheritance. Planta 190:143-150. Zhou, C, Zee, S.Y., and Yang, H.Y. 1990. Microtubule organization of in situ and isolated generative cells Zhu, Z.-P., Shen, R.-J., and Tang, X.-H. 1980a. Studies on in Zephyranthes grandiflora Lindl. Sex. Plant Reprod. the developmental biology of embryogenesis in higher plants. II. Biochemical changes during 3:213-218. embryogeny in Oryza sativa L. Ada Phytophysiol. Zhou, C, Orndorff, K., Allen, R.D., and DeMaggio, A.E. Sinica 6:141-148. 1986a. Direct observations on generative cells isolated from pollen grains of Haemanthus katherinae Zhu, Z.-P, Shen, R.-J., and Tang, X.-H. 1980b. Studies on Baker. Plant Cell Rep. 5:306-309. the developmental biology of embryogenesis in higher plants. III. Kinetic changes of nucleic acids Zhou, C, Orndorff, K., Daghlian, C.P., and DeMaggio, and protein during embryogenesis of wheat A.E. 1988. Isolated generative cells in some (Triticum vulgaris L). Ada Bot. Sinica 22:122-126. angiosperms: a further study. Sex. Plant Reprod. 1:97-102. Zhu, Z.-P, Shen, R.-J., and Tang, X.-H. 1989. The biosynZhou, C, Yang, H., Tian, H., Liu, Z., and Yan, H. 1986b. In thesis of storage protein and changes of free amino acids content during seed development of rice. vitro culture of unpollinated ovaries in Oryza sativa L. In Haploids of Higher Plants In Vitro, ed. H. Hu and Chinese]. Bot. 1:123-130. H. Yang, pp. 165-181. Beijing: China Academic Zhu, Z.-Q., Sun, J.-S., and Wang, J.-J. 1978. Cytological Publishers. investigation on androgenesis of Triticum aestivum. Ada Bot. Sinica 20:6-12. Zhou, G.Y., Weng, J., Gong, Z., Zhen, Y., Yang, W., Shen, W., Wang, Z., Tao, Q., Huang, J., Qian, S., Lin, G., Zhukova, G.Y., and Sokolovskaya, T.B. 1977. Ultrastructure of antipodals of Aconitum napellus L. Ying, M., Xue, D., Hong, A., Xu, Y, Chen, S., and Duan, X. 1988. Molecular breeding of agriculture. A (Ranunculaceae) embryo sac before fertilization. Bot. Zh. 62:1600-1611. technique for introducing exogenous DNA into Ziebur, N.K., Brink, R.A., Graf, L.H., and Stahmann, M.A. plants after self pollination. Sri. Agr. Sinica 21:1-6. Zhou, J.-Y. 1980. Pollen dimorphism and its relation to 1950. The effect of casein hydrolysate on the growth the formation of pollen embryos in anther culture in vitro of immature Hordeum embryos. Am. ]. Bot. of wheat (Triticum aestivum). Ada Bot. Sinica 37:144-148. 22:117-121. Zimmerman, J.L. 1993. Somatic embryogenesis: a model Zhou, X., and Fan, Y.-L. 1993. The endosperm-specific for early development of higher plants. Plant Cell expression of a rice prolamin chimeric gene in trans5:1411-1423. genic tobacco plants. Transgen. Res. 2:141-146. Zimmerman, J.L., Apuya, N., Darwish, K., and O'Carroll, Zhou, X., Han, Y, Yang, W., and Xi, T. 1992. Somatic C. 1989. Novel regulation of heat shock genes during
668
References
carrot somatic embryo development. Plant Cell interaction in Brassica. III. Hydration of the pollen 1:1137-1146. grains. /. Cell Sri. 76:321-336. Zonia, L.E., and Tupy, J. 1995. Lithium treatment of Zubkova, M., and Sladky, Z. 1975. The possibility of Nicotiana tabacum microspores blocks polar nuclear obtaining seeds following placental pollination in migration, disrupts the partitioning of membranevitro. Biol. Plant. 17:276-280. associated Ca2+, and induces symmetrical mitosis. zur Nieden, U., Manteuffel, R., Weber, E., and Neumann, Sex. Plant Reprod. 8:152-160. D. 1984. Dictyosomes participate in the intracellular Zou,J., Abrams, G.D., Barton, D.L., Taylor, D.C., Pomeroy, pathway of storage proteins in developing Viciafaba cotyledons. Eur. J. Cell Biol. 34:9-17. M.K., and Abrams, S.R. 1995. Induction of lipid and oleosin biosynthesis by (+)-abscisic acid and zur Nieden, U., Neumann, D., Bucka, A., and Nover, L. its metabolites in microspore-derived embryos 1995. Tissue-specific localization of heat-stress of Brassica napus L. cv. Reston. Plant Physiol. proteins during embryo development. Planta 108:563-571. 196:530-538. Zou, J.-T., Zhan, X.-Y., Wu, H.-M., Wang, H., and Cheung, zur Nieden, U., Neumann, D., Manteuffel, R., and Weber, A.Y. 1994. Characterization of a rice pollen-specific E. 1982. Electron microscopic immunocytochemical gene and its expression. Am. ]. Bot. 81:552-561. localization of storage proteins in Vicia faba seeds. Eur. J. Cell Biol. 26:228-233. Zuberi, M.I., and Dickinson, H.G. 1985. Pollen-stigma
Index A-B chromosome interchange 400 ABA-deficient (aba) mutant 433 ABA-INSENSITIVE (AB1) gene, (abi) mutant 433, 455 Abelmoschus esculentus 408 aberrant testa shape (ats) mutant 156 abscisic acid (ABA) 9,10,114,117,134,135, 381, 387, 388, 400-402, 406, 411, 426-433, 436, 451-453, 463, 464, 473, 482, 489, 492, 493, 495, 520, 521 Abutilon megapotamicum 71 Acacia retinodes 84,185, 253 ACC (see 1-aminocyclopropane-l-carboxylic acid) Acetabularia mediterranea 352 acetohydroxyacid synthase 520 acetolactate synthase 40 acetyl choline esterase 188 acetyl-coA carboxylase 488 acetylesterase 380 Achillea millefolium 170 acid phosphatase 34, 37, 44, 70-72, 92, 93,137,194, 219, 230, 247, 380, 520 Aconitum sp. 170 A. napellus 170 A. vitlparia 170,171 acridine orange 260 actin, actin gene 45, 47, 78,102,105,107,109,117, 214-217, 227, 229, 234-238, 240, 241, 295, 307, 326, 406, 426, 495 actin-binding factor 118 actin corona 308 actin-depolymerizing factor (ADF) 109, 111 actinomycin-D 53, 54, 221, 224, 225, 260, 517 activator (ac), mobile or transposable element 345, 369 actomyosin 215, 237, 238, 241 acyl-ACP thioesterase 451, 462 acyl carrier protein (ACP), ACP gene 104,116,197, 200, 451, 462, 464 adenine phosphoribosyl transferase 134 adenosine triphosphatase (ATPase) 56,142,176,187 S-adenosylmethionine decarboxylase 136, 491 adenosine diphosphate-glucose pyrophosphorylase 335, 336 gene 117 Fj-adenosine triphosphate synthase a-subunit gene (ATP-A) 41,146,147,149,150 P-subunit 99 subunit-6 gene (ATP-6) 107, 147, 148-150 subunit-9 gene (ATP-9) 41,107,144,145 adenylate cyclase 187, 270 ADF-like proteins 216 adventive embryogenesis 3,14,176 Aesculus sp. 502 A. californicus 329
A. hippocastanum 513 A. woerlitzensis 329 African blood lily (see Haemanthus katherinae) African violet (see Saintpaulia ionantha) AGAMOUS (AG) gene, ag mutant 20-25,153, 156 AGAMOUS-LIKE (AGL) gene 20, 22, 23, 25, 425 Agapanthus umbellatus 51, 214, 240 Agave americana 96 A. attenuata 99 A. parryi 164,166 Agrobacterium sp. 13, 20, 525-528, 530 A. tumefaciens 10, 20, 39,117, 227, 351, 398, 413, 414, 454, 479, 497, 525-528, 531 A. rhizogenes 122, 132, 426, 496, 531 agroinfection 527 Agropyron trichophorum 149 Agrostemma githago 74 Agrostis palustris 530 AINTEGUMENTA (ANT) gene, ant mutant 156,174 alanine amino transferase 479 Albizzia sp. 503 albumin 336, 337, 341, 438, 441, 444, 450, 459, 460-462 alcohol dehydrogenase (ADH), ADH gene 19, 91-93,104, 107,116,117, 219, 334, 372, 426 aldehyde dehydrogenase 138 aldolase 77 Alectra vogelii 367 aleurone cells 323, 324, 327, 333-335, 339, 340, 342, 347, 348, 350, 352, 353, 355, 384, 400, 401 aleurone layer (see aleurone cells) alfalfa 80, 200, 454, 460, 473, 478, 484, 495 alfin 441, 449 Alisma lanceolatum 378, 379 A. plantago-aquatka 378, 379 alkaline phosphatase 40, 380 Allamanda neriifolia 85, 88 Allanblackia parviflora 329 allergenic proteins (see allergens) allergens 72, 73, 83,106, 108,109, 117-118 Allium sp. 168 A. ammophilum 168 A. angulosum 168 A. cepa 58, 81,133,134,160,168 A. chamaemoly 86 A. nutans 168 A. pidchellwn 168, 170 A. schoenoprasum mitochondrial transcripts of 150 A. tuberosum 164 A. ursinum 170 Alocasia macrorrhiza 88 Aloe ciliaris 214
669
670 A.jucunda 80 A. secundiflora 80 A. vera 128 Alopecurus myosuroides 126 A. pratensis 236, 298 Alstroemeria sp. 235 Alyssum maritimum 378 oc-amanitin -resistant cells 497 Amaranthus hypochondriacus 445, 450 Amaryllis sp. 70 A. belladonna 83, 235, 241 A. vittata 88,187,190, 214 Ambrosia sp. 72 A. artemisiifolia (see also ragweed) 70, 72,108,109 A. elatior 70 A. psilostachya 70 A. tenuifolia 70, 205 A. trifida 70,118, 203 1-aminocyclopropane-l-carboxylic acid (ACC) oxidase 207 synthase 207 Ammophila arenaria 170 amygdalin hydrolase 412, 413 amylase (see a -amylase) a-amylase, a-amylase gene 70,137, 219, 340, 401, 527 a-amylase inhibitor 436 amylose extender (ae) mutant 92, 336 Ananas sp. 253 androecium 19 androgenesis 317, 500 Aneilema sp. 182,186,188 Anemarrhena asphodeloides 35 Anemone sp. 502 A. canadensis 507 Angiospermae (see angiosperms) angiosperms 1-9,11,13,14,17, 25, 29, 36, 41, 42, 48, 56, 60-62, 64, 65, 70, 73, 81, 82, 86-91,100,104,118, 119,120-123,129,151,152,154,158-160,162, 163,166,168,172-174,176-178,181-183,185, 195, 214, 222, 226, 229, 232, 234, 240, 244, 245, 251, 268, 292-299, 301, 308, 309, 311, 313, 317, 321, 322, 326, 328, 337, 361-363, 367, 371, 376, 377, 379, 393, 394, 395, 407, 408, 411, 432, 437, 441, 459, 464, 467, 470, 494, 500, 501, 521, 531 Anoda crispa 72 anther 2, 4-8,10,13,14, 42,49, 51, 53, 63, 72, 73, 75, 77, 86, 90, 91, 94-96, 99,101,102,104-108,111-114,116, 120-138,142,145,146, 150,151,181,182,192, 202, 211, 244, 245, 247-249, 257, 258, 264, 277, 279, 284, 289, 472, 501-510, 512, 519, 526, 528 allometric relations in 25-26 cell lineages of 28 culture 4,10, 48, 57-60, 75, 501-522, 526, 528 determination 17-25 development 25-41 wall 5, 8, 26-28, 30-33, 49, 80,102,112,123,124,129, 130 anther locule (see microsporangium) anther-specific genes 29-30, 41, 60, 90, 104, 111, 115, 116
Index anthocyanin, anthocyanin genes 340, 368, 400, 401, 434, 526 Anthophyta 1 anthranilate synthetase, anthranilate synthetase gene 200, 497 Anthurium sp. 67 antipodals 2, 3,162,163,172-175,177, 317 structure of 169-171 Antirrhinum sp. 24, 25 Antirrhinum majus (see also snapdragon) 8, 19, 20, 37, 122, 123,132,153,172,198, 291, 303, 314 floral homeotic mutants in 24-25 antisense construct (see antisense RNA) antisense RNA 11,12, 20, 21, 39^1,106, 111, 122,132, 200, 282, 288, 395, 449, 494 APETALA (AP) gene, ap mutant 20-25,153,156 aphidicolin 446 Apium nodiflorum 35, 83 apocytochrome-6 gene (COB) 107,146,149 apomict 322, 408 apomictic species (see apomict) apomixis 176-178 apple 223, 290, 291, 329 Aptenia cordifolia 188 Aquilegia formosa 18,153,166 A. vulgaris 166 Arabidopsis 7, 8, 25, 40, 41 63,107,116,117,123,127-129, 131,134,135,150,153-156,160,164,167, 169-172,174,192,197, 200, 202, 204-206, 227, 228, 276, 278, 283, 284, 305, 322, 352, 358, 359, 361, 362, 364, 370, 372, 376, 382, 384, 390, 398, 399, 402, 419^26, 429, 432, 433, 442, 450, 452, 453, 459, 461^63, 505, 527 embryo-lethal mutants in 413-415 floral homeotic mutants in 19-24 transgenic 21, 23, 24,112,114,116,117,132,156,197, 200, 228, 283, 369, 426, 429, 434, 456, 459 A. thaliana 7 arabinogalactan 9, 83,102,110,187,191,194-196, 200, 208, 232, 311, 478, 484, 486 Arabis cypria 271 Arachis sp. 502 A. hypogaea 442, 506 Aranda sp. 25 arcelin 436 archegonium 163, 407 archesporial cell (see archesporial initial) archesporial initial 5, 26, 62,123,156-159,177 archesporium (see archesporial initial) Argemone mexicana 157,173, 314 A. ochroleuca 314 arginine decarboxylase 491 Armeria maritima 249 Arum italicum 35 A. maculatum 170, 326 Asclepias exaltata 304 ascorbate oxidase 110, 111, 227 ascorbic acid 77,175, 413 asexual embryogenesis (see adventive embryogenesis) asexual reproduction 3, 176
Index asparaginase gene 426 Asparagus sp. 502 A. officinalis 116, 531 Aspergillus oryzae 41 Asphodelus tenuifolius 325 Aster tripolium 214, 225 H+-ATPase, H+-ATPase gene 116, 227, 239 Atropa belladonna 512 autoradiography 6,11, 29, 31, 35-38, 44, 48, 51-53, 76, 86, 94-96, 99,100,170, 202, 222, 232, 239, 326, 333, 379, 380, 516 auxin 3,10,18,132,133, 207, 329, 364, 381, 385, 420, 426, 468, 471, 472, 475-478, 482, 484, 486^88, 492, 493, 497, 504, 506, 507, 527 role in somatic embryogenesis 480-481 auxin transport 364, 365 auxotroph 414, 417 Avena coleoptile test 480 Avena sativa (see also oat) 32, 33, 87, 298, 337 avenin 337, 339, 350, 355 Averrhoa carambola 250, 251 avocado (see Persea americana) B-A translocation 92, 415 Bacillus amyloliquefaciens 41 B. thuringiensis 529 bamboo (see Simocalamus latiflora) BAR gene 528-530 barley 86, 87,117,122,126,128,129,131,134-137, 205, 296, 299, 300, 305, 307, 309-312, 317, 323-326, 329, 331, 333, 336, 339-342, 347, 348, 355, 385-387, 389, 392, 407,425, 427, 430, 432, 435,458, 484, 486, 488, 503, 506, 508, 509, 511, 515, 518, 520, 528-530 hordein genes of 352-353 storage protein gene expression in 347 BARNASE gene 41, 42,132, 204 BARSTAR gene 132 basic region/helix-loop-helix protein 455 bean 30,117,146,151,197, 440, 442, 443, 454, 455, 529 beet root 80, 304, 317 BELL (BEL-1) locus, bel-1 mutant 156,174 l,2-benz-isoxazol-3-acetic acid 480 benzylaminopurine 451, 507, 529 Berberis vulgaris 66 Bertholletia excelsa (see also Brazil nut) 450 Beta sp. 503 B. vulgaris (see also beet root) 80,128,130, 149,164,166, 167,169,170, 294, 305, 507 Betula pendula 493 B. verrucosa (see also birch) 106,117 Bidens cervicata 124 biolistic approaches (see microprojectile bombardment) biolistic methods (see microprojectile bombardment) Bipolaris (Helminthosporium) maydis 138 birch 106,117 boron in pollen germination 210 Bougainvillea spectabilis 65 Brassica sp. 264, 267, 272, 273, 275, 276, 281, 284, 288, 292, 298
671 B. alboglabra 487 B. campestris 40, 41, 65, 92,104,106,112,127,129,132, 148,164,166,167,169,172, 226, 258, 261, 266, 274-277, 279, 280, 282, 294, 297, 300, 301, 305, 308, 315, 358, 359, 380, 462, 464, 502, 504, 520 B. juncea 364, 390, 502, 529 B. napus (see also rapeseed) 21, 25, 35, 40, 41, 64, 68, 71, 72, 79, 81, 82, 87,104-107,110-112,114,116, 118,125,131-135,147,148,197, 204, 227, 264, 275-284, 294, 298, 300, 311, 363, 373, 378, 387, 412, 422, 423, 425, 428, 438, 441, 445, 446, 449-453, 459-464, 482, 495, 502, 503, 506, 509, 510, 512, 515, 518-520, 528 cybrids of 147,148 ogura (ogu) cms line of 133,134 transgenic 40, 41, 283, 423, 452 B. nigra 502 B. oleracea 34, 35, 64, 71, 72,102,104,116,127,134,137, 185,188, 201-203, 205, 211, 254, 260-264, 266, 273-284, 294, 297, 300 B. pekinensis 502 B. rapa 118 Brazil nut 450, 460, 462 BRITTLE (BT) gene, bt mutant 332, 335, 336 broad bean 150, 458 Brugmansia (Datura) suaveolens 192,194, 257 bud pollination 261, 263 Bupleurum dianthifolium 160 Burmannia pusilla 367, 368 Butomus umbellatus 35, 382 Ca2+ 81, 138,191, 202, 206, 231, 242, 254, 260, 270, 287, 290-292, 302-304, 306, 308, 316, 391, 433, 507 as second messenger 290 -binding proteins 118 gradient in pollen tube 238-240 in pollen mitosis 77 in pollen germination 210, 217-218, in pollen recognition 34 in pollen tube entry 168 in somatic embryogenesis 478 -modulated phosphatase 433 cabbage (see Brassica oleracea) cacao 253, 481 Caesalpinia japonica 65 caffeine 86 Cajanus sp. 502 C. cajan 127,131 calcium-dependent calmodulin-independent protein kinase (CDPK) gene 106,109 Calcium Green-1 290, 291 Calendula officinalis 166 callase, callase gene 40, 63,125,128,131 callose 35, 42, 46, 61-65, 67, 74, 78, 84, 86, 89,125-128, 131,158,159,177, 220, 228, 229, 232, 254-256, 264, 270, 271, 298, 307, 358, 476 callus-specific proteins 488 calmodulin 109,191, 217, 239, 240, 316, 477, 478 Calocedrus decurrens 458 Caltha palustris 170
672 Camellia japonica 224, 234 C. sasanqua 234 C. sinensis 234 CaMV (see cauliflower mosaic virus) canal cells 192,195-197 secretion products of 195-197 canary grass (see Phalaris tuberosa) Canavalia sp. 441 C. ensiformis 460 C. gladiata 445, 446, 449, 460, 461 canavalin, canavalin gene 441, 449, 460, 461 Canna sp. 34 C. generalis 43, 65 Cannabis sativa (see also hemp) 68 Capparis spinosa 83, 241 Capsella bursa-pastoris 156,158,160,164-166,168-170, 174, 306, 307, 309, 358, 359, 361-365, 371, 372, 374, 376, 377, 382, 383, 385, 386, 390, 391, 393, 398, 442 C. grandiflora 257 Capsicum sp. 502 C. annuum (see also pepper) 31,128,130,133,134, 506, 510 caraway 482 Cardamine pratensis 257 Carica sp. 502 C. papaya (see also papaya) 529, 531 carnation (see Dianthus caryophyllus) carotene 340, 400 carpel (see also pistil) 2, 19-25, 122, 123, 133, 152, 156, 161, 192, 258 determination 153-154 polypeptides of 19 carrot 3, 10, 99,126, 341, 352, 364, 370, 374, 425, 427, 468-472, 474-499 somatic embryogenesis in 470-472 transgenic cell line of 491, 496 Carthamus tinctorius 462 Carum carvi (see also caraway) 482 cassava (see Manihot esculenta) Castilleia sp. 81, 83 castor bean 339-341 CAT (see chloramphenicol acetyltransferase) catalase, catalase gene 107, 219, 333, 334, 401, 412, 431, 435 Catananche caerulea 298 cauliflower mosaic virus (CaMV) promoter 21, 39,117, 132, 352-354, 369, 401, 426, 457, 462, 527-529 cDNA clones 20, 25, 29, 30, 39, 40, 56, 63, 73,105,106, 108-112,116-118,155,197-199, 262, 263, 273, 274, 276, 277, 279, 280, 284-291, 323, 331, 335, 336, 342, 345-347, 350, 351, 353, 355, 395, 401, 403, 405, 407, 411, 425, 427-431, 435, 446, 447, 449, 450, 453-455, 458-460, 462, 463, 491-493, 495, 521 cDNA library 8, 9, 11,12, 90, 285, 395, 411, 439 from anthers 29, 30, 39, 40,118 from carrot cells 494 from egg cells 174 from embryos 411, 428-430
Index from flowers 20, 25, 40, 198, 200 from microsporocytes 55-56 from mitochondria 140,149 from pistil 197, 198 from pollen 12, 40,104, 111, 117 from pollen embryoids 521 from pollen tubes 111, 227 from proembryogenic masses 494 from seedlings 422 from somatic embryos 495 from stamens 37 from stigma 197, 273, 276, 277, 279 from style 284, 288 cDNA probes (see cDNA clones) cell-cycle control protein kinase gene (CDC-2) 423, 426 cell-free translation 11,19, 221, 222, 225, 260, 287, 338, 342, 343, 347, 405-407, 409, 411, 446, 447, 487-489, 491, 495, 517, 518 cellulase 11, 208, 219, 232, 300 Cenchrus ciliaris 530 central cell 2,162,163,165-167,170,172,174-176, 293, 300, 306, 308, 309, 311, 312, 315, 316, 321-325, 361, 382, 383 structure of 168-169 Cerastium arvense 314 Ceratozamia sp. 470 Ceras canadensis 488 chalaza 6,154,157-163,165,166,170,171, 173-176, 304, 306-308, 312, 325, 326, 328, 330, 332, 333, 358, 359, 361, 366, 368, 376-378, 382-384, 397 chalazal end (see chalaza) chalazal half (see chalaza) chalazal part (see chalaza) chalazal pole (see chalaza) chalazal proliferating tissue 383 chalazogamy 304 chalcone flavanone isomerase, chalcone flavanone isomerase gene 41,106,115, 227, 426 chalcone synthase, chalcone synthase gene 30, 41,121, 132 Chamaenerion angustifolium 156 Chara 238 chasmogamous flowers 246 Cheiranthus cheiri 261 chemotropism 302, 303 chick pea (see Cicer arietinum) chicory 166, 489 chimeric gene 8,10,12, 21, 39, 41,112,115,116,122,125, 132,148, 350, 352, 353, 355, 403, 415, 423, 455, 459, 461-464, 496, 527, 529, 530 Chinese celery (see Brassica alboglabra) chitinase, chitinase gene 191,199, 347, 353, 484 chloramphenicol 221, 224 chloramphenicol acetyltransferase (CAT), CAT gene 12, 13,116, 352, 353 2-chloroethyl-trimethyl ammonium chloride (CCC) 134 chlorophyll a/b-binding protein (CAB) gene 116, 424 Chlorophytum comosum 80, 84 C. elatum 84 chloroplast DNA
Index role in male sterility 150-151 chlorotetracycline 217, 239, 240 chlorpromazine 217 Chorisia chodatii 253 C. speciosa 253 chromosome puffs 379-381 Chrysanthemum alpinum 170 Cicer arietinum 428, 445, 527 Cichorium intybus 166, 476, 489-491 Cirsium arvense 123 C. oleraceum 124 C. palustre 127 C. spinosissimum 127 cis-acting elements 56,112, 114-116, 282, 353, 354, 439, 450, 454, 457, 461, 464 cis-acting sequences (see cis-acting elements) ris-regulatory domain (see cis-acting elements) Cistanche tubulosa 367 citrate synthase 412 Citrullus lanatus 185,189 Citrus sp. 253, 330, 502, 503 C. aurantifolia 124 C. aumntium 484, 486 C. jambhiri 374 C. /imon 34, 68,190,195 C. reticulata X C. paradisi 531 C. smewsis 476, 481, 487 Clarkia dudleyana 92 C. unguiculata 219 clathrin-coated vesicles 242 CLAV54TA3 (OV-3) gene 420 cleft coleoptile 397 cleistogamous flowers 129, 246 cleistogamy 246 Clivia miniata 170, 223, 235, 303 C. nobilis 302 cms 121,122, 125,127,128, 130,131,133-151, 510 CO 2 effect on incompatibility 258 Cobaea scandens 69 Cochlearia danica 384 coconut (see Cocos nucifera) coconut milk 329, 389, 390, 468, 473-475, 501, 503 coconut water (see coconut milk) Cocos sp. 502 C. nucifera 329, 385, 511 Coffea sp. 503, 531 C. arabica 489 coffee (see Coffea arabica) cofilin 111 Coix lacryma-jobi 316, 337, 346, 355 coixin 337, 355 colcemid (see colchicine) colchicine 50, 54, 66, 67, 75, 78, 79, 85,114, 217, 236, 237, 241, 503, 512 coleoptile 366, 367, 383, 397, 398, 415-417, 424, 425, 435, 436 coleorhiza 366, 367, 424, 425, 435, 436 Collomia grandiflora 246 Colocasia esculenta 386
673 Commelina sp. 182, 186, 188 C. erecta 186 complementary DNA (see cDNA) concanavalin-A (con-A), con-A gene 187, 201, 220, 265, 266, 268, 441, 449, 460, 461 conglycinin, conglycinin gene 403, 441, 444, 445, 448, 449, 456, 457 p-conglycinin (see conglycinin) connective 17,18, 25, 28, 30 Convallaria majalis 80, 83 convicilin, convicilin gene 441, 444, 447, 458 cordycepin 224 Cordylanthus sp. 81, 83 coriander (see Coriandrum sativum) Coriandrum sp. 503 C. sativum 341 coronatine 128 Corydalis cava 170 C. nobilis 170 Corylus avellana 211, 238, 476 Cosmos bipinnatus 44, 65, 66, 70-72, 74, 75, 203, 257, 264 cotton 45, 81,115,151,190, 206, 296, 306-309, 311, 373, 382, 386, 387, 393, 405-407, 409, 427-432, 447, 450, 451, 453, 460, 493, 527, 529 cotyledon 1, 2, 6, 9, 277, 362-370, 372-375, 377, 380, 381, 383, 384, 386, 387, 398, 402, 405-409, 411-414, 417, 419, 420, 422, 424, 426-428, 434, 436-438, 440-453, 455, 460, 462, 464, 470-472, 475, 477, 481, 488, 510, 528, 529 4-coumarate: coenzyme A ligase gene 29, 30, 116,197 Crambe abyssinica 442 Craterostigma plantagineum 116, 427 Crepis capillaris 168, 305 C.foetida257 C. tectorum 164,166,167,169 Crinum defixum 187,190 Crocus chrysanthus 74, 186, 205, 271 C. longiflorus 95 C. vernus 70, 71 cross-incompatibility 14, 244, 245, 269, 271, 316 Crotalaria retusa 192 cruciferin, cruciferin gene 363, 434, 441, 445, 449-452, 459, 520 Cryptomeria japonica 72,108 cucumber 18, 110,153 Cucumis sp. 393 C. melo 59,126,127 C. sativus (see also cucumber) 18, 329, 389, 390 Cucurbita sp. (see also pumpkin) 92, 442, 446 C. maxima 125, 131, 389, 390 C. moschata 223, 389, 390 C. pepo 36, 63,110, 317 cuticle 10, 71, 73, 74, 172, 183, 185-189, 205, 206, 246, 254, 271, 374, 425, 484 cuticle-pellicle 249 cutin (see cuticle) cutinase 71, 72, 74, 205, 219, 224, 269 Cycfls sp. 470 cycloheximide 50, 221, 223, 224, 231, 251 -resistant cells 488, 497
674 suppression of self-incompatibility by 260 Cymbidium sinense 156 Cyphomandra betacea 298 Cypripedium californicum 45 C. fasciculatum 78 C. insigne 358 Cyrtandra pendula 84 cysteine proteinase 155 Cylinus hypocistis 155,166 cytochalasin 214-216, 231, 234, 236, 237, 241 cytochrome oxidase 137,138 cytochrome oxidase-c subunit genes (COX-1, COX-U) 41, 143-145,147,149,150 cytochrome P450 monooxygenase 155 cytokinesis-defective (cyd) mutant 417 cytokinin 3,134, 381, 385, 386, 391, 471, 472, 474, 506, 507, 527 role in somatic embryogenesis 481 cytoplasmic male sterility (see cms) cytotoxic gene (see diphtheria toxin chain-A gene) Dactylis glomerala 118,129, 474, 489 DAPI (see 4'6-diamidino-2-phenylindole) date palm 86, 488 Datura sp. 502 D. innoxia 86, 385, 501-504, 507-510, 512, 514, 516, 518, 520, 527, 528 D. metel 102, 507, 508, 510, 511, 514 D. stramonium 182, 329, 358, 385, 389, 390 D. tatula 389, 390 Daucus aureus 164 D. carota (see also carrot) 3,130,164, 312, 425 mitochondrial transcripts of 150 D. muricatus 164, 312, 313 deaminases 77 defective endosperm-B30 (de*-B30) mutant 344, 346 defective in pollen-pistil interactions (pop) mutant 204 defective kernel (dek) mutant 332, 333, 415, 416 defective seed (de-17) mutant 332 defense-related genes 434-438 DEFIC1ENS (DEF) gene, def mutant 20, 24, 25,123 dehydrins 426, 428 Dendrathema grandiflora 531 Dendrobium sp. 156,157 deoxyadenosine 49, 50 2-deoxy-2-glucose 254 desiccation tolerance gene expression related to 426-434 4'6-diamidino-2-phenylindole (DAPI) 81,101,173, 310 7,8-diaminopelargonic acid aminotransferase 415 Dianthus caryophyllus 208 D. serotinns 314 Dichanthium aristatum 133 (2,4-dichlorophenoxy) propanoic acid 472 2,4-dichlorophenoxyacetic acid (2,4-D) 18,140, 316, 385, 468, 471^75, 478-483, 488, 491, 493, 495, 497, 498, 501, 507, 529 dicots 1, 8, 9,13, 29, 32, 52, 61, 114, 115,183,194, 325, 353, 362, 363, 366, 367, 383, 393, 415, 427-430, 469, 470, 502, 503, 525, 528-531
Index Dicotyledonae (see dicots) dicotyledons (see dicots) Dicranostigma franchetianum 314 dicyclohexylcarbodiimide (DCCD) 142 Digitalis sp. 502, 503 D. lanata 482, 484, 489, 491, 493 a-difluoromethylarginine 491 difluoromethylorni thine 491 digitonin 138 2,4-dinitrophenol 221 Dipcadi montanum 157 diphtheria toxin chain-A (DT-A) gene 12, 84,116,122, 204, 283, 459 Diplotaxis erucoides 378, 382 D. tenuifolia 205 dissociation (ds) transposable element 335 distyly 250, 251 features of 245-247 DNA polymerase 77, 86 DNA replication (see DNA synthesis) DNA sequences 108-112,140-151, 273-275, 277, 279, 350-356, 423, 427, 429, 430, 453^64 DNA sequencing 12, 39 DNA synthesis 159,173 during haploid pollen mitosis 76-77 during microsporocyte meiosis 43, 47-52 during pollen development 76-77 during pollen embryogenesis 516 in embryos 373-374 in the developing endosperm 326-327 in the generative cell 86 in the suspensor 379-380 in the vegetative cell 86 DNA-unwinding protein ("U-protein") 54 DNA uptake 526-527 DNase 48, 77,173 Doritis pulcherrima 45, 47 dormancy-related genes 432-434 dot-blot hybridization 12, 29, 30,107,132, 263, 347, 446-450
double fertilization 2, 6,181, 208, 293, 298, 299, 301, 309, 311, 314, 317, 321, 322, 356, 399 nuclear fusions (karyogamy) during 308-311 Downingia bacigalupi 375 D. pulchella 375 Drosophila melanogaster 7,107, 228 dynein 238 early methionine-labeled (EM) protein, EM gene 401, 402, 427, 429-431, 463, 493, 494 Echinochloa utilis 324 Echinocystis macrocarpa 329 egg 2-6, 9, 81,152,162,163,165-169,172-178, 243, 293, 297-301, 305-317, 357-359, 361, 382, 393, 395, 396, 398, 399, 407, 408, 498, 501, 505, 507, 526, 527, 531 macromolecules of 173-174 eggplant (see Solanum melongena) Eichhornia paniculata 249 electroporation 13,114, 352, 401, 525, 526, 528, 530
Index elongation factor 1 (EF1, EF-loc) 57, 344, 494 Elymus rectisetus 177 E. scabrus 177 embellum 304 embryo 1-3, 5, 6, 9,10,13,14, 40, 111, 166,168,176,177, 245, 269, 271, 311-315, 317, 321-324, 327, 329-334, 336, 337, 340, 347, 350, 352, 353, 355-358, 369-382, 394, 395, 428, 437, 467-474, 477, 478, 486, 487, 492, 493, 495, 497, 498, 525, 527-531 biotin autotroph 414 nutrition of the 382-384 types 361-363, 367-368 embryo axis 367, 369, 374, 411, 412, 416, 417, 419, 424, 431, 436-438, 444, 445, 450, 452, 461, 464, 529 embryo culture 329, 331, 384-394, 416-417, 420, 432, 446, 451-453 embryo development (see embryogenesis) embryo-lethal genes, embryo-lethal mutants 398, 413-417, 420 embryo mother cell (see embryoid mother cell) embryo pattern mutants 417-422 embryo sac 2-6, 9,152,156,159,162-164,166-178,181, 209, 217, 252, 253, 257, 293, 297, 299, 301, 304-309, 311, 313-317, 321, 323, 325-330, 361, 376, 381-384, 389, 397, 399, 407, 527 culture of 392-393 isolation of 172-173 embryo-specific genes, embryo-specific proteins 9, 404, 408, 425, 435 EMBRYO-SPECIFIC (EMB) locus, emb mutant 417, 418 embryo storage lipids genes of 462-464 origin of 443 synthesis of 445-446 transcripts of 450-451 embryo storage proteins genes of 451-462 mRNA changes associated with 446-450 origin of 441-443 regulation of synthesis of 443-445, 451-453 embryo transformation 527-531 embryogenesis (see also pollen embryogenesis, somatic embryogenesis) 6, 7, 9,10,13,14,173,176, 244, 314, 322, 325, 357, 358, 362, 379, 384, 386-388, 393, 394, 395, 467, 501, 525, 527 cell lineage during 358, 365, 368-370 cellular aspects of 371-374 comparative 363-367 gene expression patterns during 404-413 genes involved in 413-422 genetic control of 395—404 tissue differentiation during 374-376 embryogenic callus (see embryogenic cells) embryogenic cells 3,10,13,14, 401, 469, 471, 472, 474-480, 482, 484, 485, 487, 489, 491^93, 499, 525, 526, 528, 530, 531 embryogenic pollen 10, 501-504, 506, 508, 509, 515-520, 528 embryogenic pollen grains (see embryogenic pollen)
675 embryoid (see also pollen embryoid) 3, 4,10, 467-484, 486-512, 514-518, 520-522, 525, 528, 530, 531 embryoid mother cell 475-477, 479, 511, 514, 516 EMBRYOLESS (EML-1) gene, eml-1 mutant 417 embryology experimental 14 molecular 13,14 embryonic axis (see embryo axis) embryonic genes, embryonic proteins 487-490, 522 endo-p-l,4-glucanase 198 endochitinase 116,198, 484, 486 endocytosis 242 endomitosis 36,177,185 endopolyploid (see endopolyploidy) endopolyploidy 504, 512 endoreduplication 6,168,170,185, 326, 327, 332, 356, 409 in embryo cells 373-374 in suspensor cells 379-380 in tapetal cells 31, 36 endosperm 2, 5-7,13,14, 40,104,163,168,170,171,177, 253, 269, 293, 311, 314, 315, 321-356, 360, 367, 368, 376, 378, 379, 384, 386, 389, 390, 393, 396^01, 415-417, 419, 425, 437, 454, 458, 460, 464, 470, 477, 497, 527, 529, 530 cellular 6, 322, 323, 325, 328, 329, 333, 337 chemical composition of 329 development 322-326 free-nuclear, nuclear 6, 322, 323-325, 328, 329, 333 haustoria 328 helobial 322, 325, 328 mutants 92, 332-336 role in embryo nutrition 382-383 storage reserves of 9,10, 40, 321-323, 325, 327, 331-356, 440, 441 endosperm balance number 330-331 endosperm box 349-351, 353, 354 endosperm culture 342, 345 endosperm gene expression 333-334 endosperm nucleus 2, 6,168, 321-323, 325, 326, 328, 330, 333, 356, 382, 399, 527 endothecial cells (see endothecium) endothecium 27-31, 60,112,129 endothelium 154,163,175, 330 Endymion non-scriptus 81, 88, 214, 236 enoyl-ACP-reductase 451, 462 enoyl-coenzyme-A (coA) hydratase 412 5-enolpyruvylshikimate-3-phosphate synthase 116 Ephedra sp. 311 E. trifurca 309, 311 epiblast 366, 367 epicotyl 367, 368 Epidendrum scutella 164,166,169,170, 307, 359 Epilobium sp. 80 male sterility in 150 E. angustifolium 229 E. hirsutum 150 E. montanum 150 E. palustre 160, 307 epiphysis 363 episomal DNA143
676 epithelium 477 Eranthis hiemalis 398 Eriocaulon robusto-brownianum 368 E. xeranthemum 368 Eruca saliva 157, 257, 379, 391 Erwinia sp.109 £. chrysanthemi 108 Erysimum capitulum 271 Escherichia coli 55, 141,142, 278, 281, 290, 345, 354 Eschscholzia californica 314 esterase 34, 70, 72, 93, 136,194, 201, 205, 219, 247, 259, 479, 487, 520 ethrel 128,135, 507, 510 ethylene 11, 207, 482, 492, 507 ethylene-response-binding protein 156 ettin (ett) mutant 192 eucalypt (see Eucalyptus rhodantha) Eucalyptus globulus 529 E. rhodantha 211 Euphorbia dulcis 81, 298 E. helioscopia 169, 382 exine 2, 5, 8, 31, 33-35, 39, 61-63, 70-72, 89, 91,117,118, 121,126,128,129, 201, 203, 204, 218, 246, 247, 249, 257, 263, 264, 501, 507, 510, 511, 515 inclusions of 68-69 proteins of 34, 73-74 structure and formation of 64-68 exocytosis 188,189, 242, 476 exotoxin-A 122 extensin 102, 198, 200, 232 extensinlike protein (see extensin) extracellular proteins in somatic embryogenesis 483-486 fackel (fk) mutant 419, 420 Fagus sylvatica 442 Farsetia hamiltonii 157 fasciated mutant 153-154 FASS (FS) gene,/s mutant 419-421 Fe2+/ascorbate-dependent oxidase 198 female gametogenesis (see megagametogenesis) female gametophyte (see embryo sac) female germ unit 2 female sporogenesis (see megasporogenesis) ferritin 185 fertility restorer genes 122,138-140,142,143,146-149 fertilization 2, 3, 6,13, 14, 75, 82,121,162,163,166-170, 172-174,176, 181, 197, 198, 244, 245, 252, 254, 257, 269, 271, 293, 294, 298-300, 303, 305-308, 311-317, 322, 329, 332, 333, 357-359, 361, 369, 382, 383, 396, 398, 399, 407, 408, 413, 414, 498, 526, 527 fertilization-independent endosperm (fie) mutant 322 Festuca arundinacea 530 F. rubra 530 Feulgen reaction 5, 36,173, 326, 379 fiddlehead (fdh) mutant 202 filament 2,17, 18, 25, 31, 470 filiform apparatus 5,166-168,172,175, 304-307, 314, 360 filter hybridization 12
Index first haploid microspore mitosis (see first haploid mitosis) first haploid mitosis 30, 36, 39, 61, 75, 86, 89, 92, 94, 95, 97,100-102,105-107,110,112, 114, 311, 506, 510 514-516, 519, 522 DNA synthesis during 76-77 ultrastructural changes during 78-82 first haploid pollen mitosis (see first haploid mitosis) FITC (see fluorescein isothiocyanate) flax (see Linum usitatissimum) FLORAL BINDING PROTEIN (FBP) gene 154 floral meristem 2, 4, 7, 8,17-25, 28, 58, 60,152-154 floral iflo) mutant 20 floret (see flower) floury (fl) mutant 332, 344, 346, 416 flower 1, 2, 4, 5, 7, 8, 17-25, 31, 37, 40, 48, 56-60, 63, 68, 104, 120-123,128,132-134, 147,149, 151-154, 166,181,182,188,190-192,194, 196-198, 200, 201, 204, 207, 208, 244-247, 249, 251-253, 258, 259, 261-264, 266, 268, 272, 283, 287, 288, 292, 305, 306, 313, 317, 398, 407, 503, 508, 510, 526, 527 flowering plants 1, 4, 6, 10, 13, 42, 75,152, 245, 525 fluridone 400, 432, 452 fluorescein isothiocyanate (FITC) 72,101, 231 fluphenazine 217 Forsythia intermedia 188, 247, 251 Fragaria sp. 503 Freesia sp. 503 Fritillaria imperialis 80 fruit 1, 2, 6 ripening 10,11 fruit fly (see Drosophila melanogaster) functional megaspore 157-163, 173,177 fungal endopolygalacturonase 379 fungal toxins effect on mitochondria 138,141,142 funiculus 154, 175,176, 304, 328, 384, 428 Fusarium sp. 71 FUSCA (FUS) gene,fus mutant 434 G-box 454, 455 G-box/C-box 455, 463 Gagea lutea 88 galactomannan 334 P-galactosidase 92,104 Galanthus nivalis 241 Galium mollugo 87, 298 gametocides 128,132,135 gametoclonal variation 504 Gasteria sp. 253 G. verrucosa 34, 35, 45, 46, 66, 80,156, 157,160,170-172, 192, 206, 217, 304 Gaudinia fragilis 256 GC box (see G-box/C-box) gene action 7,10, 67, 90-92, 94, 97, 100,102,104, 106, 111, 112,114-117,120,121,123,130,133-135,137,151, 159, 223-224, 243, 248, 249, 257, 258, 298, 395, 415, 420, 422, 443, 486, 487, 489, 492, 497, 514
Index during anther dehiscence 129 during endosperm ontogeny 331-334 during meiosis 125-128 in radiation-induced mutants 397-399 in spontaneous mutants 399-404 in wide hybrids 396-397 postmeiotic 128-129 gene activation (see gene action) gene activity (see gene action) gene-cytoplasmic male steriles 121,122 gene expression 7-12,14, 56, 62, 84, 90-92, 94,111-115, 117-120,153,154,197, 221, 225, 226, 228, 245, 253, 254, 259, 260, 271, 276-278, 280, 322, 334, 340, 354-356, 363, 364, 367, 369, 374, 392, 395, 440, 471, 487, 500, 516 during anther development 28-30, 36-41 during embryo storage protein synthesis 443-453 during embryogenesis 404-413, 422-426 during endosperm storage protein synthesis 342-350 during female gametogenesis 173-174 during pollen development 104-108 during pollen embryogenesis 520-521 during somatic embryogenesis 492-498 in self-incompatible plants 281-284, 287-288, 290 in the floral meristem 20-25 in the tapetum 36-41 gene transfer 10,12 General Control Nonderepressible motif (GCN4) 349, 353 generative cell 2, 3, 5, 6, 61, 62, 73, 75, 78-86, 91, 94-97, 99-101,104,106,107,112,114,119,181, 212, 214, 215, 217, 221, 222, 224, 229, 235 238, 240, 241, 295, 296, 298, 299, 301, 307, 311, 312, 399, 507, 510, 516-520, 521 division of 87-89 embryogenic division of 511-515 isolation of 89 nuclear pores in 86-87 generative pole microtubule system 78-79 geneticin resistance 530 genie male steriles 121,122,124,125,131,134-136 genomic library 12, 28,107,109, 277, 402 Gentiana acaulis 35 Gerbera sp. 503 G. hybrida 200 G. jamesonii 300, 317 germin 486 germless mutant 415 Gibasis karwinskyana 35 G. venustula 35, 65 gibberellic acid (GA) 18,19, 58,133,134,137, 306, 329, 340, 381, 385, 386, 391, 393, 433, 481, 482 gibberellin (see gibberellic acid) Ginkgo biloba 458, 504 ginseng (see Panax ginseng) Gladiolus sp. 503 Gladiolus gandavensis 70,185,187,196, 201, 205, 216, 300, 301
677 gliadin 337, 338, 342, 348, 349, 351, 353, 355 GLOBOSA (GLO) gene, glo mutant 24, 25,123 globulin-like protein 348 globulins, globulin genes 336, 337, 339, 341, 342, 350, 352, 355, 400, 402, 407, 441, 444, 450, 457, 458-460 a-globulins (see globulins) glucan hydrolases 259 (3-glucan synthetase 234 (l,3)-p-glucan (see also callose) 35, 232, 234, 242 (l,3)-p-glucanase (see also callase) 40,198 glucose dehydrogenase 430 glucose-6-phosphatase 194 p-glucuronidase (GUS), GUS gene 12,13, 28, 30, 39-41, 56,111-117,132, 227, 282, 283, 284, 289, 352, 354, 369, 401, 423, 426, 429, 452, 454, 455, 459-461, 463, 464, 479, 493, 496, 497, 520, 526-529, 531 glutamate dehydrogenase 259 glutamic dehydrogenase 487 glutamic-oxaloacetic transaminase 92, 93 glutamine synthetase 117, 227 glutathione 48, 77 glutelin 336, 337, 339, 341, 342, 344, 350, 351, 355 glutenin 337, 348, 349, 353, 354, 529 glyceraldehyde-phosphate dehydrogenase, glyceraldehyde-phosphate dehydrogenase gene 77, 219, 334 Glycine max (see also soybean) 56,127,128,156,160,164, 166,167,169,187,193, 305, 325, 358, 359, 373, 378, 382, 383, 387, 424, 441, 442, 446, 463, 531 G. soja 456 glycinin, glycinin gene 374, 403, 441, 444, 445, 448, 449, 456, 457 glycocalyx 65 glycosylrransferases 442 Gnetum sp. 311 Gnetutn gnemon 309 GNOM (GN) gene, gn mutant, GN protein 419, 420, 422 Gossypium sp. 502 G. anomalum 122 G. hirsutum (see also cotton) 45, 82, 83,105,109,122,161, 163,166,169,183,193, 206, 213, 294, 305, 325, 358, 359, 361, 369, 372, 373, 387, 412, 424, 428, 445, 446, 463, 531 G. thurberi 122 grape 482, 484, 485 GTP-binding protein (G-protein) 384, 492 gurke (gu) mutant 419, 420 GUS (see P-glucuronidase) gymnosperms 2,14, 33,108, 222, 226, 228, 309, 311, 337, 363, 407, 467, 481, 494, 495 pollen embryogenesis in 504 somatic embryogenesis in 470 transformation in 531 gynoecium 19,153, 207 gynogenesis 317 Hacquetia epipactis 188 Haemanthus katherinae 80, 85, 325, 326, 383 haploid embryogenesis 500
678 haploid microspore mitosis (see first haploid mitosis) haploid mitosis (see first haploid mitosis) Haynaldia villosa 338 hazel (see Corylus avellana) heat-shock proteins (HSP), HSP genes 56, 102-104,107, 108,117, 200, 228, 346, 411, 412, 423, 445, 494, 495, 498, 518-520, 522 Hebe salicifolia 127 H. townsoni 126 Hela cells 352 helianthinin, helianthinin gene 441, 450, 460, 461 Helianthus sp. 502, 503 H. annuiis (see also sunflower) 30, 65, 66, 68, 72, 104, 111, 128,130,156,164, 166, 169, 170, 172, 205, 306, 307, 325, 358, 359, 371, 412, 423, 441, 445, 452, 453, 460, 463 Helleborus foetidus 33, 34, 36, 65, 68, 78, 80, 214, 241, 298 Hemerocallis sp. 253 7-7. fulva 208, 219, 301 7-7. minor 301 hemp 68,126 henbane (see Hyoscyamus niger) heterostyly 245, 248, 251 Hevea sp. 502 Hevea brasiliensis 126, 482, 491 hexokinase 334 hexose phosphate isomerase 334 Hibiscus aculeatus 330, 397 H. costatus 330, 358, 397 H. furcellatus 330, 358, 397 H. rosa-sinensis 72 Hieracium 177 Hippeastrum belladonna 53, 80,100 H. vittatum 85, 298, 301 Hippuris vulgaris 67 histone 489 H2 100,101, H3 gene 28, 29, 37,174, 424, 429, 495 H4 gene 424 of pollen 100,101 synthesis during meiosis 53 homeobox domain (see homeobox gene) homeobox gene 367, 423, 495, 496 homeotic mutants 7, 8,123,153 in Antirrhinum majus 24-25 in Arabidopsis 19-24 hordein, hordein gene 337, 341, 342, 347, 351, 352, 355, 458 hordenin 337 Hordeum bulbosum 322, 373, 408 H. distichum 170, 373, 408, 412, 512 H. vulgare (see also barley) 86, 92,128,160,164,170,171, 185,294,316, 322,324, 337, 343, 348,358, 371-373, 387, 393, 396-398, 408, 463, 505, 512, 513 horse chestnut (see Aesculus woerlitzensis) Hosta sp. 80 Hyacinthoides non-scripta 45 Hyacinthus orientalis 84, 86, 88, 95, 99, 214, 241 hydrolases from pollen 219, 224
Index 3-hydroxy-3-methylglutaryI coenzyme A reductase (HMC) gene 116,197 hydroxyacyl-coA dehydrogenase 412 hydroxyproline 102,194, 227, 423, 424 hydroxyproline-rich protein (see hydroxyproline) hygromycin resistance 527, 530 Hymenocallis occidentalis 80 Hyoscyamus sp. 502 H. niger 29, 37, 38, 80, 95, 96, 99,100, 222, 502, 506-509, 511, 512, 514, 516, 517, 519 Hyphaene natalensis 329 hypocotyl 2,13, 362, 363, 365-370, 372, 375, 385, 414, 419, 426, 450 hypophysis 363, 364, 366, 369, 371, 376, 420, 421, 468, 470-472, 474, 475, 477, 483 hypoploids 92-94 hypostase 175, 382 Iberis amara 34 7. semperflorens 72 7. sempervirens 72, 74, 264 immunoglobulin heavy-chain-binding protein (BiP) 337, 346 Impatiens balsamina 84 7. glandulifera 80, 309 7. palmata 191 7. roylei 328 7. sultani 45, 126 7. walleriana 80, 235 in situ accumulation (see in situ hybridization) in situ hybridization 11, 20-22, 24, 26, 28-30, 37, 40, 44, 95,112,146,158,173,174,197, 226, 263, 281, 288, 309, 344, 359, 363, 379, 414, 423, 424, 429, 437, 446, 449, 493, 494, 506 in situ localization (see in situ hybridization) incongruity 271 indeterminate gametophyte (ig) mutant 331 indoleacetic acid (IAA) 18,133, 208, 259, 306, 364, 381, 385, 390, 393, 468, 474, 480, 481, 497, 501, 507 indolebutyric acid (IBA) 507 integument 154,156,161,167,172,174-177, 304, 323, 328, 330, 361, 382, 383 integumentary tapetum (see endothelium) intermediate filaments 78 interspecific incompatibility 74 intine 61, 63, 67, 74, 80, 83, 84,109,110,117, 211, 218, 220, 264, 513, 515 development of 69-70 proteins of 34, 71-73 intraovarian pollination 313-314 intraspecific incompatibility 74 invertase 219, 224, 332 in vitro fertilization 300, 315-317 in vitro translation (see cell-free translation) ionophore A23187 239, 507 Ipomoea sp. 503 7. purpurea 67, 378 7. trifida 291 7ns pallida 130
Index I. pseudacorus 235, 237 isocitrate lyase (ICL), ICL gene 107,116, 412, 422, 423 isoflavone reductase gene 200 N6(A-isopentenyl) adenine (2iP) 471, 481, 507 isoperoxidase 487 isopropyl-N-phenylcarbamate 85 isozymes 92, 93,136, 333-335, 372, 412, 446, 487-488, 520 Jacranda sp. 502 Japanese cedar (see Cryptomeria japonica) Japanese pear (see Pyrus serotina) ]asione montana 169, 358, 359 jasmonate (see jasmonic acid) jasmonic acid 135, 432, 452, 493, 520 Juglans regia 329, 531
for embryo growth 391 in pollen 225, 239 traffic in embryoids 478 kaempferol 190,191, 204 kafirin 337, 355 kanamycin resistance 526, 527, 531 Kentucky blue grass (see Poa pratensis) kernel mutants 331-333, 344 KEULE (KEU) gene 419 kinesin 238 kinetin 18,19, 58, 59,134,153, 386, 391, 393, 471^73, 501, 507 KNOLLE (KN) gene 419, 420, 422 KNOPF (KNF) gene 419 KNOTTED-1 (KN-1) gene 367, 420, 495 Kunitz trypsin inhibitor (KTI), KTI gene 108, 363-364, 375, 403, 404, 437 lactic dehydrogenase 37 Lactuca saliva (see also lettuce) 317, 362, 373 Lagenaria vulgaris 157 Lagerstroemia speciosa 358 Larix sp. 470 Larix X eurolepis 477 late embryogenesis abundant (LEA) genes, LEA proteins 9,116, 347, 388, 402, 414, 426, 434, 436, 453, 489, 492-494 ABA effects on 431^32 from embryos of dicots 427—429 from embryos of monocots 429-431 Lathyrus odomtus 125, 126 L. sativus 531 LEA genes, LEA proteins (see late embryogenesis abundant) LEAFY COTYLEDON (LEC) gene, lee mutant 434 LEAFY (LFY) gene, Ify mutant 23, 24 lectin, lectin gene 187, 201, 220, 265, 403, 404, 435-438, 441^43, 462 Ledebouria socialis 35, 68 Ledum groenlandicum 358 legumin, legumin gene 381, 383, 441, 443, 444, 446-452, 457-459 legumin-box 457-460, 462
679 leguminlike proteins (see legumin) lemon (see Citrus limon) Lens culinare 28 leptotene protein (L-protein) 49 lettuce 317, 488 leucine aminopeptidase 198, 219 leucine zipper DNA-binding protein (bZIP) (see leucine zipper protein) leucine zipper protein 56, 345, 346, 350, 354, 429, 430, 431, 455 LEUN1G (LUG) gene, lug mutant 20, 21 Lilium hybrid 55 Lilium sp. 37, 44, 48, 50, 52, 54, 56, 63, 64, 67, 68, 77, 88, 253, 302 L. auratum 219, 236, 302 L. candidum 53, 58, 80, 95,156,157,160 L. davidii 185,190 L. formosanum 399 L. henryi 36, 37, 45, 47, 48, 51, 66, 67, 96,101 L. hybrida 35, 45 L. japonicum 302 L. leucanthum 302 L. longiflorum (see also lily) 25, 34, 35-37, 43-45, 47-49, 51-54, 56, 57, 59, 64-70, 75-77, 80, 89, 94-96, 98,100-102,104, 111, 114,117,156,159,160, 168,190,192,195,196, 214-217, 221, 223, 225, 227-230, 232-240, 242, 253, 255, 259, 260, 264, 265, 268, 296, 302, 306, 315 L. martagon 80 L. regale 169,185,190,195,196, 228, 302 L. rubrum 241 L. speciosum 48-51, 53, 54 L. superbum 302 L. tigrinum 48 lily 25, 26, 37, 44, 49-52, 54-56, 76, 97, 111, 210, 211, 214, 218, 220-222, 224, 225, 233, 234, 239, 240, 303, 317 Limnophyton obtusifolium 361 Limonium sp. 249 L. vulgare 250 Linaria bipartita 328 Linum sp. 246, 247 L. austriacum 331 L. cateri 67 L. catharticum 164,169, 359 L. grandiflorum 69,182,187, 218, 246, 247, 249-251 L. perenne 331 L. pubescens 246 L. usitatissimum 66,131,164,166,169, 305, 307, 359, 387, 390, 391 lipid body membrane protein 492 lipid transfer protein, lipid transfer protein gene 10, 39, 40,197, 200, 352, 355, 414, 425, 434, 484, 486 lipoxygenase 137, 446, 452 Litchi sp. 502 Lobelia erinus 81, 224 Lolium perenne (see also ryegrass) 71, 72, 80, 81, 300, 530 mitochondrial transcripts of 150 Lonicera japonica 45
600 Lotus corniculatus 259 transgenic 117, 227 L-phosphinothricin (L-PPT), L-PPT gene 528-530 luciferase gene 352, 436 Luculia gratissima 250 Luffa sp. 502 L. cylindrica 237, 510, 512 L echinata 512 Lupinus sp. 481 L. albus 442 L, angustifolius 446 L. arboreus 426 L. luteus 329, 389, 411 L. mutabilis 127,129 Lychnis alba 205, 220 Lycium barbarum 508 L. chinense 508 Lycopersicon sp. 270, 502 L. esculentum (see also tomato) 18, 33, 36, 43, 45, 92,122, 128-130,150,153,185,193, 270, 412, 505 anther cDNA clones of 30,104 L. peruvianum 153, 154,188,189,192,194, 213, 229, 235, 255, 261, 262 267, 270, 289, 290, 329, 330, 505 L. pimpinellifolium 329, 330 Lythrum sp. 248 L. junceum 247, 248, 250 L. salicaria 250 Macadamia integrifolia 169 macerozyme 300 MADS protein, MADS-box gene 24, 25,154, 425 Magnolia denudata 45, 47 M. salicifolia 458 M. soulangeana 45 M. tripetala 45, 47 maize 19, 25, 28, 30, 36, 37, 40, 47, 76, 91, 92, 94, 97-99, 105-110,112,114-116,122-128,131,133-136, 147,151,154,159,171-174, 216, 218, 219, 226-228, 232, 278, 299-302, 304, 309, 314-317, 323, 324, 326, 327, 329-336, 338, 340, 341-346, 348, 351-353, 355, 367, 369, 385, 386, 389, 392, 399-402, 409-411, 420, 423-426, 428-433, 445, 450-452, 460, 463, 476, 477, 486-488, 492, 495, 503, 505, 508, 511, 526-530 cms-C 122,137-140,143 cms-S 122,137-140,142-144 cms-T 122,136-143 embryo-lethal mutants in 415-417 storage protein gene expression in 342-347 transgenic 117, 426, 530 zein genes of 350-352 malate dehydrogenase 219, 412 malate synthase (MS), MS gene 107,116, 412, 422, 423 male gametogenesis (see microgametogenesis) male gametophyte 2,14, 60, 68, 89,178, 265, 283, 293, 399, 526 male germ unit 2, 296-299, 317 male sporogenesis (see microsporogenesis) male sterile lines (see male sterility)
Index male sterile mutants (see male sterility) male steriles (see male sterility) male sterility 8, 41,120-132, 244 amino acid changes associated with 135-136 effects of photoperiod on 133 effects of temperature on 133 hormonal effects on 133-135 role of mitochondrial genome in 8, 137-151 role of mitochondrial proteins in 140-142, 146 protein changes associated with 136 Malus sp. 503 M. communis 193 M. domestica (see also apple) 223, 225, 231, 234, 235, 265, 268 Malvaviscus arboreus 70, 72 mandelnitrile lyase 412, 413 Manihot esculenta 127, 531 mannopine synthase gene 117 Matteuccia struthiopteris 441 matteuccin 441 medicagin 441, 449 Medicago sativa (see also alfalfa) 80, 87,109, 131, 158,174, 195, 299, 312, 378, 383, 387, 388, 397, 441, 445, 449, 452, 475, 477, 479, 482, 491, 494, 495 somatic embryogenesis in 472 M. scutellata 378, 397 megagametogenesis 7, 9,152,154,155,162,172-174, 177, 178, 293, 525 megagametophyte (see also embryo sac) 152,154,158, 162, 168, 170, 172, 173, 177, 178, 407 nutrition of 174-176 megagametophytic tissue 450 megaspore 2,158-163,176 megaspore mother cell (see megasporocyte) megasporocyte 2,156-162,174,177 megasporogenesis 5,13,152, 154,155,157-160,162,174, 177,178 meiosis 2, 8, 61-67, 76, 90-92, 97,105-107,123-128, 130, 131,133-136,158-160,162,174,177, 257, 284, 478, 513, 527 endonuclease activity during 50 in microsporocytes 17, 28, 29, 32-60 molecular biology of 47-57 organelle dedifferentiation during 44-45,159-160 meiosis-specific genes 55-57 meiotic division (see meiosis) meiotic histone 53 meiotic mutants 47, 159 Melandrium album 314 M. rubrum 314 meiotin-1 53-54 melon (see Cucumis melo) Mercurialis annua 134 mericarp 471 meromyosin 235 mesogamy 304 metallothionein-like protein 411 methomyl 138,141,142 6-methylaminopurine 386
Index methylglyoxal bis(guanylhydrazone) 491 6-methylpurine 260 5-methyltryptophan -resistant cells 480-481, 497 MICKEY (MIC) gene 419 microfilaments (see also actin) 45, 47, 78, 79, 83, 85, 86, 107,127, 128,157, 172, 214-216, 229, 234-237, 241, 242, 295, 296, 307, 326, 515 microgametogenesis 7, 8, 62, 89, 90,100,101,119,120, 293, 525 microinjection 3,13, 526, 528 microprojectile bombardment 10,13, 112,117, 227, 340, 352,525,526, 528-531 micropylar end (see micropyle) micropyle 6, 154,156, 160-163, 165-168,172-176, 304, 306, 312, 325, 328, 330, 359, 361, 364-366, 368, 375, 377, 380, 382-384, 469 microsporangium 17, 20, 28, 31-34, 42, 62, 67, 68, 72, 77, 83, 92,123-125,127-131, 283, 501, 508, 510 microspore 2,17, 26, 28, 31, 32, 34, 35, 38-42, 44, 45, 47, 48, 50, 59-72, 75-81, 86, 89, 92, 94, 95, 97, 99,102, 104,106,107,110,112,114-116,123,125-132, 134,136,137,146, 159, 257, 283, 284, 301, 392, 500, 504, 506, 509, 510, 512, 516, 518-521, 526, 528 microspore mother cell (see microsporocyte) microsporocyte 2,17, 26-29, 32-39, 41-67, 75, 76, 90-92, 97, 104,106, 107,123-128,130, 131,133-137,146, 159, 257, 284 culture of 58 microsporogenesis 5,13,17, 25-27, 31, 32, 34, 37, 41, 58, 59,104,107,108, 112,120-123,128,129,132,133, 151,159 ultrastructural changes during 42-47 microtubules (see also tubulin) 32, 34, 35, 45^7, 66, 69, 78-80, 83-85, 87-89,107,127,128, 156,157, 171, 172, 216, 217, 229, 235-238, 241, 294, 295, 297, 306, 307, 325, 326, 358, 361, 477, 515 preprophase band of 45, 78, 79, 158, 325, 361 middle layers 27, 28, 31, 32, 39,124,137 miniature seed (mn) mutant 332 minichromosome maintenance (MCM-1) gene 20, 24 Mirabilis jalapa 81, 309 mitochondria in male sterility 137-151 inheritance of 80-81 mitomycin-C 50 mitotic cyclin gene 426 molecular hybridization 11, 374, 379 cDNA-poly(A)-RNA hybridization 98-99, 405, 407 cDNA-RNA reassociation 447 DNA-DNA hybridization 48 DNA-rRNA hybridization 380, 487 RNA-DNA hybridization 104, 405-406 solution hybridization 453 monensin 231, 338, 443 Monocotyledonae (see monocots) monocotyledons (see monocots) monocots 1, 9, 10, 32, 52, 61,114,183,194, 325, 327, 353,
681 362-364, 366, 367, 393, 415, 429, 470, 502, 503, 525, 528, 530 monopteros (mp) mutant 419-421 Monotropa hypopiiys 359 M. uniflora 80, 358, 359, 362, 367, 368 Monotropsis odorata 207 morphactin 123 mRNA (see also poly(A)-RNA, stored mRNA) 6, 8, 9-12, 20, 22, 23, 26, 29, 30, 37, 39, 40, 49, 52-56, 95, 97-99, 102,104,106,107,109-111,117, 118,132, 145,173,174,198, 200, 206, 221, 224-226, 243, 263, 273, 278, 279, 284, 285, 288, 334, 338, 339, 342-350, 352, 359, 376, 389, 402^09, 411, 412, 414, 422-433, 435-437, 442, 446^54, 456, 457, 479, 487, 489, 491^95, 509, 515, 517, 518, 520, 521 mucronate (me) mutant 344, 346 multigene family 105, 276, 277, 280, 285, 350, 351, 353, 355, 427, 431, 437, 450, 453, 454, 456, 458, 462 Musa sp. 530 Muscari racemosum 86 mutase 334 mutator (mu) transposon 333, 415 MYB motif 352 MYC motif 352 myosin 83, 215, 237, 238, 240, 241 myosinlike proteins (see myosin) Myosurus minimus 164,173 myrosinase, myrosinase gene 438 NADH dehydrogenase 520 NADH dehydrogenase subunit-3 (NADS) gene 145 NADPH-dependent aldose reductase 430 Najas marina 240 naphthaleneacetic acid (NAA) 58, 386, 420, 451, 472, 473, 481, 507 napin, napin gene 122, 363, 441, 445, 449-452, 459, 460, 495, 520 narbonin 449, 460 Narcissus biflorus 510 N. pseudonarcissus 214, 215, 217, 220, 235, 241, 302, 303 neomycin phosphotransferase (NPTII) 527, 529-531 Nerium oleander 375 Nicotiana sp. 259, 272, 273, 286, 288, 292, 330, 362, 502 N. alata 19, 81,191,192, 194,195,197-199, 210, 214, 219, 223, 225, 232, 234, 235, 237, 238, 245, 259, 261, 263, 267, 281, 282, 284-289, 302 N. debneyi 19 N. forgetiana 245 N. gossei 227, 526 N. glutinosa 114, 117, 508 N. langsdorfii 19, 288 N. plumbaginifolia 116, 288, 354, 489, 493 transgenic 458 N. rustica 114,117, 314, 359, 398, 503 N. sylvestris 99,188,192, 206, 207, 225, 234, 236, 237, 511, 512 N. tabacum (see also tobacco) 3,18,19, 52, 63, 75, 80, 81, 86-88, 95, 97, 102, 104, 105, 108, 114, 117, 122,
682 153,164,166,167,172,185,187,192,193,195, 198, 200, 210, 214, 215, 217, 219, 223, 225-228, 232, 233, 235-238, 240, 282, 286, 288, 294, 296, 298, 299, 305, 308, 312, 314, 358, 361, 412, 423, 461, 502-504, 506, 508-512, 519, 520, 528 transgenic 283, 458, 459 nicotinamide adenine dinucleotide (NAD +) 138 nifedipine 217, 239 Nigellla damascena 36 nigericin 443 Nitella 238 nitrate reductase, nitrate reductase gene 487, 520 NO APICAL MERISTEM (NAM) gene, nam mutant 420 nodulation (NOD) genes 484 nonembryogenic callus (see nonembryogenic cells) nonembryogenic cells 3, 10, 476, 477, 480, 482, 484, 485, 487, 491, 492, 495, 496 nopaline synthase (NOS) gene 39, 227, 353, 354, 401, 527 norflurazon 63 Northern blot 12, 20, 29, 37, 57,104-107,110,112,114, 140,198, 226, 227, 262, 263, 273, 279, 285, 287, 288, 348, 402, 422, 423, 429, 431, 435, 446, 447, 449, 451, 493, 521 Nothapodytes foetida 328 NPTII (see neomycin phosphotransferase) nucellar cap 382 nucellar cells (see nucellus) nucellar tissues (see nucellus) nucellus 2, 3, 9,154,156-158,161-163,168,170-172,174, 175, 177, 253, 304, 323, 324, 328, 330, 333, 361, 382, 383, 393, 470, 474, 481, 486 nuclear male sterility 121 nuclease-C 81 nucleases, from pollen 219 nucleoside diphosphate kinase 224 nucleoside diphosphate sugar transferase 335 Nymphaea alba 45 oat 32, 326, 337, 339, 340, 342, 350, 355, 385, 386, 530 storage protein gene expression in 350 obturator 304 Oenothera sp. 503 0. biennis 158, 329, 396 O. hookeri 80,128,131 O. lamarckiana 160,164, 329, 396 O. longiflora 303 O. murkata, 329, 396 O. organensis 104-106,108,110, 206, 218, 219, 226, 227, 253, 264, 265 okadaic acid 279 Olea europaea 32, 33, 35, 44, 69,185,193 oleosin, oleosin gene 40, 111, 341, 425, 434, 443, 445, 450-452, 462-464, 489, 492, 520 oligomycin 221, 260 olive (see Olea europaea) oncus (see Zwischenkorper) onion (see Allium cepa) Onoclea sensibilis 427 OPAQUE (OP) gene, op mutant, OP protein 332, 344-347
Index OP-2 MODIFIER gene 346 Ophiocephalus sp. 81, 83 Ophrys lutea 82 orbicules 33-35 orchard grass (see Dactylis glomemta) Orchis maculata 158 ornithine decarboxylase, ornithine decarboxylase gene 136, 491 Ornithogalum caudatum 157,161, 166,185,195, 304, 306 O. virens 45, 88, 224 Orobanche aegyptiaca 367 Orthocarpus sp. 81, 83 Oryza sativa (see also rice) 25,104,125,164,166,169,174, 305, 327, 337, 341, 358, 371, 372, 387, 408, 424, 501, 503 oryzenin 337 oryzin 337 osmotin gene 116, 227 ovary 2, 7,14,104,152,181,192,196, 205, 207-209, 250, 251, 253, 257-259, 268, 270, 282-284, 287, 301-304, 313-315, 317, 332, 376, 408, 471 culture of 392-393 ovulata (ovu) mutant 24 ovule 2, 4, 7, 9,13,14,18,104,122,133,152, 153,159,161, 163, 172-177,181,191,192, 207-209, 251, 253, 257, 282, 293, 303-307, 313-317, 322, 323, 328, 330-333, 348, 359, 365, 376, 379, 381, 382, 384, 387-389, 395-398, 402, 407, 408, 428, 432, 467, 474, 488 ATPase activity in 176 culture of 392-393 development of 154-158 ovule-mediated transformation 526-527 Oxalis sp. 68, 248 P-particles 230, 233 P-pollen 510 pachytene DNA (P-DNA) 50, 51 pachytene-small-nuclear DNA (psn-DNA) 51 pachytene-small-nuclear RNA (psn-RNA) 51 Paeonia sp. 70, 313, 363, 510 P. anomala 363 P. lactiflora 114,117 P. moutan 363 P. tenuifolia 37, 52, 53, 95, 99 P. vnttmanniana 363 palindrome 114, 426 palindromic sequence (see palindrome) Panax ginseng 527 Papaver sp. 272, 273, 292, 503 P. dubium 87 P. nudicaule 314, 359 P. rhoeas 170, 253, 260, 265, 268, 290, 291, 313, 314 P. somniferum 314, 325, 393, 508 papaya 529 transgenic 529 Paphiopedilum spicerianum 160 parietal cell 5, 26, 27,131,156,157 Parietaria officinalis 83 Parkinsonia aculeata 80
Index parsley 29, 30,197 Partheniutn argentatum 257 parthenogenesis 176, 408 particle acceleration (see microprojectile bombardment) particle bombardment (see microprojectile bombardment) Paspalum orbiculare 304 patatin 31,117 pathogenesis-related proteins 198, 200, 347 pattern-forming genes 417-422 PCR (see polymerase chain reaction) pea 34,117,122,123,126, 200, 383-385, 402, 403, 405, 407, 409, 417, 424, 428, 432, 437, 440, 442^44, 446-449, 451, 453, 458, 459, 481, 489 disease-resistant response proteins of 420 peanut (see Arachis hypogaea) pear (see Pyrus communis) pearl millet 123, 528 pectate lyase 108, 226 pectin esterase 110, 111, 226 pectinase 108,110, 208, 219, 232, 300 Pelargonium sp. 503 P. hortorum 80 P. zonale 80, 82,165 pellicle 74,187-188, 201, 270 Peltophorum pterocarpum 506 Pennisetum americanum (see also pearl millet) 123,125, 126, 473 male sterility in 150 P. ciliare 170 P. glaucum 167, 303 male sterility in 150 P. typhoides 205, 206 P. typhoideum 230 pepper 31,122 perianth 20, 58 pericarp 174, 412 periclinal chimeras 153 peroxidase 93,194,219, 225, 251, 381,482,483,486,487, 520 role in self-incompatibility 259 peroxidase-labeled con-A 279, 485 Persea americana 189,193 petal 2, 7,17-25,104,106,122,123,153,154, 202, 261, 268 Petroselinum crispum (see also parsley) 29,116 Petunia sp. 10, 81,116,117,144-146,151, 204, 267, 273, 288, 289, 292, 420, 502, 516, 526 somatic hybrids in 144,145 transgenic 117,121,145,154, 227 P. axillaris 59, 261, 267, 315 P. hybrida 8, 25, 34, 36, 41, 59,102,106,115,117, 125,130, 135-137,144,145,154,164,166,169,183, 186-192, 206, 218, 219, 223-225, 227, 229, 233-235, 237, 255, 258, 259-261, 263, 265, 267, 288, 289, 296, 298, 299, 303, 306-308, 315, 359, 371, 373, 426, 456, 508, 526 po mutation in 115 transgenic 40,132, 204, 352, 426, 457 P. inflata 105,106, 111, 198, 207, 227, 261, 265, 288-290 transgenic 111, 289
683 P. parodii 136,144,145 P. violacea 314 Phalaenopsis sp. 45, 47,155, 207 P. equestris 78, 79 Phalaris coerulescens 268, 291 P. minor 201 P. tuberosa 70, 75, 203 Pharbitis sp. 503 P. nil 188, 191 phaseolin, phaseolin gene 117,122, 352, 429, 436, 441, 442, 448, 453-456, 462 (5-phaseolin (see phaseolin) Phaseolus sp. 377, 379, 396, 502 P. acutifolius 192, 396 P. atropurpureus 128,130 P. coccineus 377-382, 391, 392, 396, 408, 411, 436 P. htnatus 396 P. vulgaris (see also bean) 30,117,127,146, 325, 326, 352, 372-374, 378-381, 388, 396, 404, 408, 409, 411, 413, 436, 441, 442, 445, 446, 448, 452, 455, 458 Phaseolus vulgaris ABAI-3-like factor (PVALF) gene 455 phenylalanine ammonia lyase (PAL), PAL gene 30,117, 137,197 Phleum nodosum 127 Phoenix dactylifera (see also date palm) 86 phosphoglucoisomerase 92, 93, 219 phosphoglucomutase 93 phosphorylation activity 47,138, 204, 221, 260, 270, 278, 279, 287, 290, 431, 433, 434 photosystem II quinone-binding protein (PSB-A) gene 107, 424 phragmoplast 45, 47, 61, 79, 80, 89, 325, 326, 515 Phyllosticta maydis 138 Physalis ixocarpa 257 phytase 224 phytic acid 224, 234, 258 phytochrome, phytochrome gene 117,133, 426 phytohaemagglutinin (PHA) 220, 404, 436, 437 Picen sp. 470 P. abies 117, 227, 228, 445, 470, 482, 484, 491, 531 P. glauca 442, 445, 446, 450, 477, 484, 494, 531 P. glauca/P. engelmatmii 445, 450, 491 P. marina 477, 531 Picris echioides 166 pigeon pea (see Cajanus cajan) pin-formed (pin-1) mutant 364 pin morph 182,185,197, 218, 246, 247, 249-251 Finns sp. 470 P. banksiana 33, 66, 68, 75 P. caribaea 484, 486 P. densiflora 235, 236 P. monticola 445 P. pinaster 227, 228 P. ponderosa 222 P. radiata 531 P. strobus 407 P. taeda 222, 226, 228 P. thunbergii 116, 235, 236, 411 pistil 2, 7, 19, 192, 200, 202, 245, 251, 252, 257-259, 261,
684
Index
263, 265-269, 271, 272, 277, 279, 281, 283, 285, 510, 513, 514, 516, 521 pollen embryogenesis 4,10,13,14, 317, 357, 500-516, 302, 303, 314, 315 abortion 19 520, 521 pistil-specific genes 197-199 genotypic effects on 505-506 PISTILLATA (PI) gene, pi mutant 20, 21, 23-25,123, 153 pollen embryoid 474, 503, 507, 515, 517, 518, 520, 528, 531 Pisum sp. 502 transformation mediated by 528 P. arvense 373 pollen germination 2, 8, 63, 69, 73, 74, 83, 97,102,108, 109,119,121,188,195, 202-206, 208, 210-217, P. sativum (see also pea) 34, 80,122,127,159,160, 235, 373, 231, 234, 239-241, 245, 247, 249-255, 258, 260, 378, 411, 412, 423, 428, 441, 442, 443, 446, 458, 264, 265, 266, 268, 270, 271, 283, 299, 301, 302, 462, 527 314, 315, 399, 516 placenta 153,154,175, 304, 306, 307, 315, 328, 332 biochemistry of 220-226 placental pollination 315 efflux of metabolites during 217-220 Plantago lanceolata 134 gene expression during 226-228 male sterility in 150 in heterostylous plants 249-250 plasmatubules 206, 207, 234 pollen grain (see pollen) plastids pollen isozymes 92 inheritance of 80-82, 229, 294, 311 plastocyanin gene 200 pollen-mediated transformation 526-527 PLENA (PLE) gene, pie mutant 24,153 pollen mother cell (see microsporocyte) pollen recognition 8,14, 34, 36, 63, 64, 68, 69, 73-75,102, Plumbagella sp. 306 P. micrantha 168, 306, 308 109, 111, 181, 187,188,191, 201-208 Plumbago sp. 306 during cross-incompatibility 269-271 P. capensis 168 during self-incompatibility 251-268, 279-292 P. zeylanka 80,164-166,168, 169,171-173, 294, 296-301, gene for 277-279 Pollen RetroElement Maize (PREM) 109,110 306-308, 310, 312 pollen-specific genes 8,12, 30, 90,118,119, 226, 227, 232, plumular pole (see plumule) plumule 362, 367, 368, 400 290 Poa annua 68 early 105-106 P. pratensis 118, 385 expression of 105-108, 112 Podocarpus macrophyllus 75 isolation of 104-105 polar fusion nucleus 2,162,163, 293, 298, 299, 309, 311, late 105-106, 111, 117 promoter analysis of 112-117 312, 321, 322, 333 structure of 108-112 polarity 9, 359, 395, 419 of egg 165,173-174 pollen-stigma interaction (see pollen recognition) of embryoids 478-479, 515 pollen tube 5, 7-9,14, 61, 69, 73-75, 82-89,102,105,108, of embryos 364, 370, 420, 422 109, 119,121,168,181, 185, 187,191,192, of megaspores 160-161 194-198, 200-211, 245, 247, 249-260, 264, 265, of microspore nucleus 78 267-271, 280, 282-284, 286-296, 298-302, of organelles 80 311-315, 399, 408 of pollen 66, 78 calcium gradient in 238-240 of somatic embryos 478-479 cytoplasmic streaming of 234 of zygote 361 cytoskeleton of 234-238 pollen 2-10, 12-14,17, 28-30, 33-41, 57, 59, 61, 62, 65, gene expression during growth of 226-228 growth 240-242 68-75, 78, 80-87, 89-123,125-135,140,142, 146, 151,181,182,185,187,188,191,195,198, growth in heterostylous plants 249-251 201-210, 229-231, 234, 235, 238, 241-261, guidance into embryo sac 302-308 263-272, 276, 277, 279-284, 286-292, 294, 296, structure of 228-231 298-303, 313-315, 332, 333, 396, 399, 400, 413, transformation mediated by 526, 527, 531 415, 500-505, 511, 513-515, 518-521 wall, structure of 231-234 abortion 106, 111, 120,122, 128-132,135,137,150 pollen wall 2, 5, 45, 61, 62, 83, 84, 87, 89, 93,108,110,117, cytoskeleton of 214-217 125, 128,131,132, 136, 203, 515 dimorphism 91, 509 adaptive significance of 73-75 embryogenic competence of 509-510 assembly of 63-70 germination pores (apertures, germ pores, germinaof hypoploids 93 tion sites) of 63-67, 69, 70, 72, 73, 76, 80, 83, 94, proteins of 70-74 211, 214-217, 220, 264 pollenkitt 35, 36, 41, 68, 70, 257 pollen callus 501-512 pollination 4, 8,14, 73-75, 91,108,177,181, 182, 187, 188, pollen culture 191,196-198, 200-209, 217, 219, 225, 244-255, 257-261, 263-265, 268-272, 276, 277, 283, 286, for embryogenic induction 501-504, 506, 508, 509,
Index 287, 299, 303, 305, 306, 313-315, 317, 322, 330, 332, 334, 341, 342, 346, 369, 396, 400, 417, 423, 526, 527, 531 poly(A)-RNA 11,12, 20, 26, 28-30, 37, 40, 49, 50, 52, 55, 57, 97, 98,158,174,197,198, 221-223, 225, 273, 284, 343, 368, 403, 407, 411, 422, 428-430, 446, 493, 494, 521 polyamines 136, 210, 379 in pollen embryogenesis 519-520 in somatic embryogenesis 490-492 polyembryony 311, 315 polygalacturonase 11,108-110,115, 226 Polygonum capitatum 176 P. divarication 163 polymerase chain reaction (PCR) 9,12,13,107,118,174, 277, 279 polynucleotide kinase 50 polynucleotide ligase 50, 77 polyploid (see polyploidy) polyploidy 36,170, 326, 356, 420 polyspermy 316, 317 polytene (see polyteny) polytenic (see polyteny) polyteny 31, 36,168,170, 328, 379-381 Poncirus sp. 502 Pontederia cordata 247, 249, 250 P. rotundifolia 247 P. sagittata 247, 250, 251 ponticulus 304 poplars 74 Populus sp. 270, 317, 503 P. alba (see also poplars) 70, 74, 75, 270 P. deltoides (see also poplars) 70, 74, 75, 92, 270, 295, 298, 305, 311 P. lasiocarpa 188 P. maximowiczii 92 P. nigra 92 P. tremula 67, 68 P. tremuloides 270 P. trichocarpa 270 porogamy 304 postament 175,176 potato 25, 30, 31, 107, 116, 197, 270, 506 transgenic 117, 456 precocious germination 386-389, 406, 428, 432, 434, 451, 452, 473, 474, 482 Primula sp. 250, 502 P. acaidis 251 P. obconica 247, 250 P. officinalis 202 P. sinensis 250 P. veris 250 P. vulgaris 185,193, 247, 250 pro-orbicules 33 pro-Ubisch bodies 32 Proboscidea louisianica 166, 305 proembryo 313, 322, 329, 362, 363, 365-371, 375-377, 383, 384, 386, 393, 397-399, 415, 417-^19, 424, 530 culture of 389-392
685 proembryogenic mass 471, 475, 476, 478-481, 484, 486-488, 492-194 profilin 102,106,109,117-118 prolamins 336-339, 341, 342, 350, 352, 355 prolifera (prl) mutant 174 proline-hydroxyproline-rich proteins 198 promoter analysis 90 of embryo storage protein genes 453-464 of endosperm storage protein genes 350-356 of pollen-specific genes 112-117 of self-incompatibility genes 281-284 prophase chromosome reduction (see reductional grouping) protease 70, 387, 487 protein kinase, protein kinase gene 54, 109, 111, 227, 240, 260, 270, 277-279, 287, 327, 429, 434, 482, 484, 492, 495 protein phosphatase 482 protein receptor kinase (see protein kinase) protein synthesis and metabolism during endosperm development 333-334 during meiosis 53-55 during pollen development 99-104 during pollen embryogenesis 518-520 during pollen germination and pollen tube growth 223-226 following cross- and self-pollinations 260 for vesicle production 231 in developing embryos 411-413 in suspensor 380-381 in tapetum 37 protoplasts 3, 4,10,13,14, 34, 65, 78,125, 345, 353, 401, 402, 435, 454, 455, 470, 475, 477, 478, 499, 525 from embryo sac cells 172 from microspores 301 from microsporocytes 59 from pollen 70, 75, 85, 89,100,101,114, 214-217, 235, 236 fusion of 137,144,148-150 tapetal 34 prunasin hydrolase 412, 413 prunin 450 Prunus amygdalus 450, 451 P. armeniaca 267 P. avium 80,188,193, 229, 230, 235, 267, 268 P. cerasws 267 P. persica 267, 329 P. serotina 412 PSB-A gene (see photosystem II quinone-binding protein) Pseudomonas aeruginosa 122 P. syringae 128 Pseudotsuga sp. 470 P. menziesii 450, 458 Psidium sp. 503 Psophocarpus sp. 502 Pulmonaria affinis 250 pumpkin 110 pyrophosphorylase 334, 335 Pyrus communis 84, 211, 212, 214-217, 219, 232, 235, 236, 241
686 P. serotina 193,194, 268, 290, 291 pyruvate decarboxylase 107 pyruvate orthophosphate dikinase (PPDK) 334, 346
Index
355, 367, 402, 407, 409, 411, 417, 419, 425, 426, 431, 432, 435, 454, 477, 487, 489, 493, 501, 503, 511, 512, 527, 529, 530 anther development in 26-28 quercetin 251 glutelin genes of 355-356 Quercus gambelii 164,166,176, 305, 308, 325, 358-361, 371 male sterility in 148 quiescent center 376, 424, 425 storage protein gene expression in 350 Quinchamalium chilense 175 transgenic 28,117,197, 336, 350, 355, 426, 527 rice tungro bacilliform virus (RTBV) 117,197 Ricinus communis (see also castor bean) 339 RAB gene (see responsive to ABA gene) radicle 365, 367, 368, 372, 385, 386, 424, 435, 436, 451, 478 RNA helicases 111 RNA polymerase 53, 409, 497 radicular pole (see radicle) RNA synthesis and metabolism radish 35, 409, 411, 431, 459 ragweed 72,118 during meiosis 44, 52-53 Rana catesbeiana 501 during pollen development 94-97 R. pipiens 501 during pollen embryogenesis 516-518 Ranunculus macranthus 80 during pollen germination and pollen tube growth R. repens 153, 222-223 R. sceleratus 170, 323, 325, 474, 477 during somatic embryogenesis 487 rapeseed 21,148,151, 427, 440, 445, 459, 462 following cross- and self-pollinations 259 Raphanus sp. 147,148 in developing embryos 408—410 R. caudatus 384 in generative cell 94-95 R. landra 384 in suspensor 379-381 R. raphanistrum 193,205 in tapetum 37 R. sativus (see also radish) 35, 36, 68, 73, 74,147,148,190, in vegetative cell 94-95 205, 254, 261, 264, 384, 428, 446, 449, 450, 459, RNase 41, 51, 70,117,132, 195, 268, 286, 288-292 502 ROL gene (see root loci gene) ogura (ogu) cms line of 147 root 2, 3,13, 28,102,104, 117, 362-367, 369-371, 374-376, reassociation protein ("recA-protein") 54-56 385-387, 397, 398, 400, 414, 417,419-421, 423-425, receptacle (seefloralmeristem) 431, 433, 435, 454, 459, 468-472, 475, 476, 479, reduced endosperm (re) mutant 332 501, 515, 527, 531 reductional grouping 478 root loci (ROL) gene 122, 132, 426, 496, 497 ROOTLESS (RTL) gene 419 responsive to ABA (RAB) gene 402, 427, 428, 431, 493 restitution nucleus 36,177 Rottboellia exaltata 133 restriction fragment length polymorphism (RFLP) 277, rubber (see Hevea brasiliensis) Rubia cordifolia 175 312, 414 RFLP (see restriction fragment length polymorphism) Rubisco (see ribulose-l,5-bisphosphate carboxylase) Rhinanthus minor 328 rugosus (r) locus 402, 403 Rhizobium leguminosarum 484 Rumex acetosa 25 Rhizophora mangle 375, 402 run-off transcription 39, 342, 422, 436, 446, 448, 451, 461 Rhododendron sp. 80, 271, 299, 358 Ruscus aculeatus 191, 202 R./orh(nei207 rutin 251 RY-repeats 457, 463 R. laetum 84, 86, 229, 294-296, 298 R. macgregoriae 84, 294-296, 298, 300 rye 48, 87,159, 203, 260, 329, 338, 340-342, 355, 503, 504, Rho family GTPases 235 527 ryegrass 71,117,118 Rhoeo discolor 52, 53 R. spathacea 35-37, 44, 53 Ribes rubrum 52 S-associated proteins (see S-locus-specific glycoproteins) ribitol dehydrogenase 430 S-family proteins (see S-locus-specific glycoproteins) ribosomal DNA (rDNA) 44,170, 487 S-gene (see sterility gene) ribosomal RNA (rRNA), rRNA gene 41, 44, 52, 174, 222, S-LOCUS ANTHER (SLA) gene 279 223, 286, 373, 379, 380, 408, 409, 487 S-LOCUS RECEPTOR KINASE (SRK) gene 277-281, 291 ribosome S-LOCUS-RELATED (SLR) gene 276, 277, 280, 281, 283, 284, 484 changes during meiosis 44, 52 ribulose-l,5-bisphosphate carboxylase (Rubisco) 107, S-locus-specific glycoproteins (SLG), SLG genes 108,122, 151, 413 261, 262, 264-268, 273-282, 284, 285, 289, 291, large subunit (RBC-L) gene 57,107, 424 484 small subunit (RBC-S) gene 424 S-RNase (see stylar RNase) rice 25, 27, 29, 37, 40, 41, 95,105,109,115-117,123,125, Saccharum bengalense 264 126,133, 227, 317, 324, 326, 329, 334, 336, 39-342, S. officinarum (see also sugarcane) 126
Index safflower (see Carthamus tinctorius) Sagina procumbens 362 Sagittark sagittaefolia 366 Saintpaulia ionantha 176 Salix hiitnilis 68 Sambucus sp. 502 S. nigra 87 Saponaria officinalis 64 Saxifraga granulata 362 Scilla bifolia 170 S. peruviana 81 S. sibirica 169 scutellum 366, 367, 369, 371, 383, 386, 387, 397, 416, 417, 419, 423-425, 430, 431, 435, 436, 445, 473, 477, 527, 529, 530 SDS-PAGE 11,12, 43, 53, 70, 99,100,102,103,140,150, 219, 225, 226, 237, 261, 262, 268, 284, 301, 341, 343, 405, 443, 485, 488, 490, 518 Secale cereale (see also rye) 48, 58, 59,126, 130,185,189, 202, 203, 211, 222, 230, 237, 298, 300, 322, 337, 371, 389, 396 secalin 337, 338, 341, 342, 355 Sechium edule 389 secreted proteins role in somatic embryogenesis 482-486 seed 1-3, 6, 9,14,19,120,121,135,152,176,177, 244, 245, 260, 261, 263, 265, 271, 313-315, 323, 325, 328, 329, 331, 334, 341, 352-354, 356, 367, 368, 373, 382-384, 386-389, 393-399, 402-405, 409, 411-415, 422, 423, 426-429, 432-434, 436, 437, 440-442, 444, 446, 449, 450, 452, 454-456, 458, 459-462, 464, 468, 525 synthetic 3, 499 seed embryo 369, 370, 384-386, 399, 405-407, 482, 484, 489, 490, 492-495, 498, 520, 527-529 selective gene amplification 326-327, 348, 373-374, 380, 446, 486 self-incompatibility 6, 8,14, 74, 75, 121,195,197,198, 201, 244, 245, 248, 249, 251-254, 257-263, 265, 268, 269, 271-273, 276, 277, 279-284, 286-292, 315, 484 acquisition of 261, 263 heteromorphic 245-251 homomorphic 251-269 relation to host-pathogen responses 292 sepal 2, 7,17-22, 24, 25,122,123,153,154, 202 sepaloidea (sep) mutant 24, 123 Sequoia sp. 470 serum response factor (SRF) gene 20, 24 Sesamum indicnm 270 S. mulayanum 270 Setaria httescens 324 sexual reproduction 1,17,176, 201, 208, 293, 531 -SH compounds 47, 48, 77 shed pollen culture 503-504, 506 shikimate dehydrogenase 93 shoot (see shoot apex) shoot apex 2-5, 7,13, 24, 99,104, 364-370, 372, 374, 375, 385, 397, 400, 416, 417, 419, 420, 423-425, 449, 468, 472, 473, 479, 501, 527, 529 shoot apical meristem (see shoot apex)
687 shoot bud (see shoot apex) SHOOT MERISTEMLESS (STM) gene 419, 420 shoot pole (see shoot apex) short integuments (sin) mutant 156,174 shrunken endosperm expressing xenia (sex) mutant 333 shrunken endosperm (se and seg) mutant 333 SHRUNKEN (SH) gene, s/i mutant 332, 335, 336, 401 signal transduction 102, 111, 201, 204, 208, 210, 217, 227, 240, 260, 270, 277, 279, 281, 287, 290, 292, 322, 384, 400, 401, 429, 430, 434, 482, 507 Silene alba 67 S. latifolia 25 S. pendula 64-66, 68 S. schafta 314 S. tatarka 314 Simmondsia chinensis 446 Simocalamus latiflora 506 Sinapis alba 25, 37, 39, 373, 411, 438, 442, 445, 452, 502 slot-blot assay (see dot-blot hybridization) slot-blot hybridization (see dot-blot hybridization) SMALL POLLEN (SP-1) gene 91 Smyrnium perfoliatum 188 snapdragon 8, 25 Solanum sp. 269, 270, 510 S. acaule 150 S. carolinense 507 S. chacoense 80, 267, 270, 289 S. melongena 45 S. nigrum 164,166,167,169 S. phureja 257 S. pinnatisectum 257 S. soukupii 270 S. stentomum 257 S. surattense 510, 512 S. tuberosum (see also potato) 25, 29, 31, 80,126,150,188, 198, 199, 263, 267, 289, 290 S. verrucosum 270 somatic embryo (see somatic embryogenesis) somatic embryo-mediated transformation 530-531 somatic embryogenesis 3,10,13,14, 357, 364, 370, 385, 399, 467-470, 499, 501, 522, 525, 527, 530, 531 biochemistry of 486-492 cell biology and physiology of 475-486 cell lineage during 469 direct 474 from callus 471-474 from protoplasts 475 gene expression during 492-498 recurrent 474-475 transgenic 531 somatoplastic sterility 330 Sophora flavescens 379 sorghum 65, 336, 339-342, 355, 411, 527, 529 Sorghum bicolor (see also sorghum) 65, 68, 103,122,125, 130,133,147, 228, 316, 324, 337, 339 S. vulgare 130,131,135,136 Southern blot 12,13, 39, 105,144, 273, 276, 285, 286, 350, 374, 453, 526-528 Southern hybridization (see Southern blot) soybean 56,117,123,128, 131,193, 227, 323, 364, 374, 375,
688 382, 384, 387, 403-405, 407, 409, 424, 429, 432, 437, 438, 440, 442, 444-449, 452, 453, 456, 457, 462, 481, 488, 489, 495, 496, 529 transgenic 529 soybean embryo factor-3 (SEF-3) 457 Spathoglottis plicata 368
Index of endosperm 350-356 of somatic embryos 494-495 storage protein manipulation 529 storage proteins (see also embryo storage proteins) endosperm - storage reserves of 6, 321-323, 325, 327, 363, 372-374, 381, 383, 384, 388, 399, 400, 402-404, 411, 414, 422, 429, 432-434, 436, 437, 489, 490 of embryos 441-464, 522 of endosperm 334-342 stored mRNA in the egg 407-408 of embryos 411 of pollen grains 97, 221-223 sequence complexity of 97-99
sperm 2, 3, 5, 6, 61, 62, 80-82, 87, 89,100,109,119,152, 163, 166,168,177,178,181, 209, 229, 241, 243, 255, 293-296, 303-305, 307-317, 333, 357, 399, 408, 526 isolation of 299-301 sperm cells (see sperm) sperm dimorphism 296, 297, 299, 309, 317 sperm recognition 311 spikelet culture 503 spinach 164, 300, 312 Striga gesnerioides 367, 368 Spinacia oleracea (see also spinach) 164,166,167, 169,193, stylar RNase {S-RNase) gene, protein 286-290 stylar secretion 193-197 294, 298, 305, 309 Spironema fragrans 185 style 2, 7-9, 91,104,108,129,152,175, 181-183,185,187, Spodoptera frugiperda 454 188,190-202, 204-209, 217, 218, 232, 240, 242, 243, 245-253, 255, 258-265, 267-270, 272, 273, sporogenous cells 5, 26-28, 31, 33, 34, 36, 40, 47, 53, 58, 281-288, 291, 300-305, 308, 313, 314, 527 121-124,128-131,135,136,151,156,157, 526 sporogenous tissues (see sporogenous cells) genes expressed in 197-201 sporopollenin 5, 33, 41, 61, 63, 65, 73, 89,128,131-133, succinic dehydrogenase 137, 224, 372 sucrose synthase 334-336 136, 211, 515 synthesis of 67-68 SUCROSE SYNTHASE (SUS) gene from maize 335 stamen 2, 7,17-25, 37, 91,121-123,133,134,136,137, sucrose synthase gene from potato 107 sucrose synthetase 412 151-153, 245-249, 258 sugarcane 126, 530 polypeptides of 19 sugary {su) mutant 332, 336 starch-branching enzyme 403, 424 sunflower 30, 41,105,106, 111, 134, 137,149, 317, 352, stearoyl-acyl carrier protein (ACP) desaturase, stearoyl369, 411, 428, 429, 441, 450, 454, 461 ACP desaturase gene 40,116, 451, 462 SUPERMAN (SUP) gene, sup mutant 20, 23,153,156 Stellaria media 157,169,173, 325, 361, 377 sterility gene (S-gene) 7, 8, 245, 248, 253, 254, 257, 258, superoxide dismutase 107,116, 435 suppressor mutator (spm) mobile element 345 261, 266-268, 271-273, 276, 281, 287-290 suspensor 6, 312, 313, 331, 362-366, 371, 372, 376, 383, stigma 2, 7-9,14, 34, 36, 73, 74,102,104, 121, 129,152, 393, 397-399, 408, 411, 414, 421, 425, 426, 469, 478 181-183,185,187-192,194,195,197,198, 201-209, 218, 237, 242-244, 246-256, 258, nuclear cytology of 379-380 260-271, 272-274, 276-284, 287, 290, 291, 302, physiology and biochemistry of 380-382 role in the growth of cultured embryos 389, 391 303, 313, 414 anatomy of 182-187 ultrastructure of 377-379 senescence of 191-192 SUSPENSOR (SUS) gene, sus mutation 398 stigma-specific genes 197-201 suspensorlike structure 469, 511 sweet cherry (see Prunus avium) stigmatic exudate 182,186,187,193, 202, 205, 209, 249 synaptonema (see synaptonemal complex) composition of 190-191 synaptonemal complex 43, 49, 50, 52, 54, 56,158 secretion of 188-190 synergids 2, 3, 5,162,163,169,170, 172-175,177,178, types of 182-183 297, 311, 314, 315, 317, 360 stigmatic papillae 73, 74,156,182,183,185-191,197, 201-206, 247, 250, 254, 255, 258, 260, 264, 271, role in fertilization 304-309 272, 279-284, 287 structure of 166-168 Stipa elmeri 164,166,169 syntaxin 422 stomium 28, 31, 60,129 transcript accumulation in 30 Tabebuia caraiba 253 storage compounds (see storage proteins) T. ochracea 253 storage lipids 443, 445-446, 450-451, 462-464, 520 Tagetes erecta 122 storage materials (see storage proteins) tapetal cells (see tapetum) storage products (see storage proteins) taperum 5, 8, 26-29, 31-33, 42, 53, 58, 60, 63, 67, 68, 72, storage protein genes 74, 75, 77, 99,102,112,116,117,124,125,133, expression of 342-350, 520 135-137,142,151, 257, 263, 264, 283, 284, 501, of embryos 453-462 508, 509
Index amoeboid (periplasmodial) 34, 35, 37,130 cytology of 36-37 DNA accumulation in 36 DNA transfer from 34 gene expression in 37-41 glandular (secretory) 34, 35, 37 role in pollen sterility 129-132 structure-function relationships of 31-36 ultrastructure of 32 tapetum-specific genes 37-41,122, 125 taro (see Colocasia esculenta) tassel seed (ts) mutant 19 TATA-box 114, 289, 349, 351, 352, 427, 455-459, 461, 463 TATA element (see TATA-box) TATA sequence (see TATA-box) T-DNA (see transferred DNA) temperature-sensitive cell lines 497-498 test-tube fertilization 314-315 tetrad 31, 32, 34, 35, 37, 38, 40, 44-^6, 53, 57, 61-69, 71, 75, 78, 89, 100,102,104,106, 123,125-128,130, 131, 133,134,137,158-162,177, 506 Theobroma cacao (see also cacao) 253, 257, 438, 446, 460 thiamine pyrophosphatase 37 thiol endopeptidase 29 thiolase 412 thionin 30, 198, 339, 341, 342 y-thionin (see thionin) thioredoxins 111, 291 2-thiouracil 210 thrum morph 182,187, 218, 246, 247, 249-251 thymidine kinase 77 Tillandsia caput-medusae 80, 82 tissue culture (see also under anther, embryo culture, ovary, ovule, pollen culture, zygote) 3, 4, 10, 13, 14, 18,19, 25, 57-60, 75,134-136,140, 141,144, 153,172, 314-317, 321, 329, 338, 345, 364, 370, 381, 400, 402, 407, 408, 414, 525, 527-530 tobacco 3, 4, 8,10, 18, 25, 29-31, 3 9 ^ 1 , 63, 77, 78, 86, 88, 97,102, 104,106-108, 111, 112,114-116, 132,134, 141,149,151, 153,176, 197,198, 200, 208, 222, 223, 227, 230, 236, 239, 281, 282, 303, 304, 312, 317, 352, 353, 355, 383, 432, 437, 454-457, 459, 461, 462, 468, 503, 505, 506, 508-511, 513, 514,516, 518, 520, 525, 528 cDNA clones from 29-31 cybrid (somatic hybrid) 137, 149 transgenic 21, 30, 3 9 ^ 2 , 75, 84, 111, 112,114-117, 125,126,132,145,197,198, 200, 204, 227, 228, 281-283, 288, 290, 352-355, 403, 404, 423, 426, 427, 429, 431, 433, 436, 452, 454-464, 493, 525 tomato 11,18, 25, 31, 37, 39, 93, 103,105,106, 108-110, 112,113, 115,117, 122,123,131, 133,134,136, 153,197,198, 200, 226, 290 GA-deficient mutant of 134 stamenless-2 (sl-2) mutant of 18,133, 136,137,153 transgenic 112, 113, 198, 423 Torenia fournieri 163, 166,167, 308 totipotency 4, 468, 470, 498, 521 Tradescantia sp. 86, 182 T. bracteata 34, 65, 69, 78, 84, 86
689 T. paludosa 37, 45, 48, 53, 58, 76, 79, 80, 86, 94-99,104-106, 109, 115, 173, 214, 221-225, 234, 309, 373 T. reflexa 84, 213 T. virginiana 34, 35, 66, 86, 88, 89, 214, 230, 231, 235, 237, 239, 241, 295, 296 trans-acting factors 9,113,114, 280, 334, 344, 349, 426, 431, 439 transcription factor 20, 23, 40, 111, 114,116,156, 345, 346, 350, 354, 355, 401, 423, 425, 429, 433, 434, 455, 496 transcription regulator (see transcription factor) transfer cell 69, 83, 84,167,169,175,186,195, 234, 323, 324, 335, 378, 382, 384 transferase 336 transferred DNA (T-DNA) 13, 20,122,127, 398, 413, 426, 434, 454, 479, 497, 526-528 transient assay 112, 113, 115-117, 353-355, 454, 455, 458 transmitting tissue 185,188,192,193,198-200, 204-207, 232, 259, 270, 280-284, 287, 288, 301, 303 transposon mutagenesis 7, 8, 24, 92,122,150, 333, 345, 369, 400, 415, 417 transposon tagging (see transposon mutagenesis) 2,4,5-trichlorophenoxyacetic acid 472 2,4,6-trichlorophenoxyacetic acid (2,4,6-T) 480 Trichodiadema setuliferum 87, 298 trichome 185, 203, 206 Trifolium sp. 257 T. ambiguum 330, 396 T. medium 396 T. pratense 192,194, 489 T. repens 330, 396, 477 T. rubens 489 T. semipilosum 396 Trigonella foenum-graecum 223 Trillium sp. 48 T. erectum 44, 47, 48, 53, 54, 58, 76, 77, 94 triosephosphate isomerase 92, 93 Tripsacum laxum 126 tristyly 249 features of 247-248 Triticale 170, 309, 322, 512 triticin 337, 348, 353 Triticum X Hordeum 530 Triticum sp. 322, 337 T. aestivum (see also wheat) 9, 26, 45, 51, 65, 66, 68, 76, 80, 149, 164,165,167, 169,170, 192, 214, 300, 308, 316, 322, 325, 337, 349, 361, 364, 366, 367, 372, 387, 390, 393, 396, 435, 452, 521 T. durum 337, 399, 445, 446 T. timopheevi 149 T. vulgare 337, 408 Tropaeolum majus 186, 219, 224, 378, 380, 381 tryphine 35, 36, 68, 70, 74, 201, 204, 257, 264 trypsin inhibitors 30 tubulin (a, P, and y) 45, 57, 88, 99,107,116, 156, 217, 227, 235-237, 406, 495 Tulbaghia violacea 36, 48, 52, 76, 80, 81, 94, 99 tulip (see Tulipa gesneriana) Tulipa sp. 33, 35 T. gesneriana 53, 78, 234 tumor-inducing (Ti) plasmid 13, 20, 39
690 tunicamycin 260, 442, 483, 484 tunicate (tu) mutant 19 twin (twn) mutant 398 Typha latifolia 212, 214, 221, 223, 225
Index
cms in 149 gliadin and glutenin genes of 353-354 histone H3 gene from 28 ribosomal RNA probe of 174 storage protein gene expression in 347-350 ubiquitin 102,116,195, 460, 528 tapetal disintegration in 130 Ubisch bodies 33 transformation of 526-530 Ulmus sp. 503 transgenic 530 uniconazol 433 wheat germ agglutinin (WGA) 435, 436 unilateral incompatibility 269-271 wide hybridization 329-331, 395-397 UNIVERSAL FLORAL ORGANS (UFO) gene, ufo mutant wild soybean (see Glycine soja) 20, 23, 24 WUSCHEL (WUS) gene 419 uridine diphosphoglucose dehydrogenase 224 X-bodies 309 Vanda sp. 174, 361, 368 Xenopus sp. 352 X. laevis 259, 265, 338, 442 V. sanderiana 373 vegetative cell 2, 5, 6, 61, 62, 70, 75, 78-84, 92, 94-97, xyloglucan 334 99-102,112, 114,119,181, 202, 211, 212, 214-217, xyloglucan endotransglycosylase 484 220-222, 229, 235, 238, 240, 241, 255, 295-301, yeast 174, 349, 351, 353, 354, 458 307-310, 507, 516-519, 521 DNA metabolism of 86 cell division cycle phosphoprotein (p34cdc2) 495 embryogenic division of 510-515 expression of protein in 141 isolation of 89 secretory protein 422 nuclear pores in 86-87, 214 verapamil 217 Z-Iayer (see Zwischenkorper) Verbascum phlomoides 189 Zamia sp. 470 Viciafaba (see also broad bean) 150,172,185,187, 206, 303, Zantedeschia aethiopica 88 323, 372, 373, 378, 381-383, 408, 413, 441, 442, Zea mays (see also maize) 19, 26, 32, 36, 45, 52, 53, 65, 66, 444, 446, 449, 452, 458, 459, 462 68, 92, 99,102-104,107,122,130,156,157,159, 160,163,165,166,169,170,174,183,185, 216, V. narbonensis 449, 460 vicilin, vicilin gene 381, 383,441,443,444, 446-452,458-460 219, 221, 294, 300, 305, 316, 323, 337, 340, 344, 345, 359-361, 367, 369, 387, 392, 393, 411, 412, vicilin-box 458, 461 Vigna sp. 502 423, 424, 445, 452, 515 cms-C 122,130 V. radiata 411 cms-S 122,130 V. sinensis 323, 378, 382, 383 cms-T 122, 130 V. unguiculata 66,192, 411, 445, 527 inflorescence culture of 59 V. vexillata 65, 66 zeatin 134, 329, 391, 471, 481, 482, 507 virulence (VIR) gene 13 zeatin riboside 134, 391, 507 Viscum minimum 156,164,167,169 Vitis sp. 502, 531 zein, zein genes 337-339, 341-343, 350, 352, 355, 460 V. berlandieh X V. rupestris 531 characterization of 350-352 V. rupestris 531 expression of 342-347, 351-352 Zepkyranthes sp. 393 V. vinifera (see also grape) 188, 373, 491, 531 Z. Candida 301 vitronectin 200, 303 Z. grandiflora 85, 236 vitronectinlike protein (see vitronectin) Z. rosea 157 VIVIPAROUS (VP) genes, vp mutants, VP protein 399-402, 433, 455 zinc finger domains 111 Zwischenkorper 63, 211 vivipary 399, 400, 402, 432 zygote 1-3, 6, 9, 82,163,168,173, 253, 293, 309, 310-316, 322, 329, 357, 362-364, 366, 367, 372-374, 377, walnut (see Juglans regia) 382-384, 389, 391, 394, 396-399, 407, 420, watermelon (see Citrullus lanatus) 467-469, 471, 478, 498, 525-527, 531 WAXY (WX) gene, wx mutant 92,117, 332, 335, 530 culture of 392-393 expression of 336 ultrastructure of 358-361 Western blot 12, 238, 273, 423, 485 wheat 9, 29, 56,102,128,135-138,148,150,151, 296, 298, zygotene DNA (zyg-DNA) 49-51 zygotene RNA (zyg-RNA) 49-50 305, 315, 317, 323, 324, 326, 329, 334, 336-340, zygotic embryogenesis 10,13,14, 357, 467, 469, 470, 478 342, 347, 348, 352, 354, 355, 382, 385, 387, 389, zygotic embryos 467^71, 478, 494, 495, 501, 520, 521, 401, 411, 427-432, 435, 436, 453, 473, 482, 486, 531 488, 493, 503, 506, 509-512, 516