The Nucleolus (Molecular Biology Intelligence Unit)

MOLECULAR BIOLOGY INTELLIGENCE UNIT

Mark O.J. Olson OLSON MBIU

The Nucleolus

The Nucleolus

MOLECULAR BIOLOGY INTELLIGENCE UNIT

The Nucleolus Mark O.J. Olson Professor and Chairman Department of Biochemistry The University of Mississippi Medical Center Jackson, Mississippi, U.S.A.

LANDES BIOSCIENCE / EUREKAH.COM GEORGETOWN, TEXAS U.S.A.

KLUWER ACADEMIC / PLENUM PUBLISHERS NEW YORK, NEW YORK U.S.A.

THE NUCLEOLUS Molecular Biology Intelligence Unit Landes Bioscience / Eurekah.com Kluwer Academic / Plenum Publishers Copyright ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system; for exclusive use by the Purchaser of the work. Printed in the U.S.A. Kluwer Academic / Plenum Publishers, 233 Spring Street, New York, New York, U.S.A. 10013 http://www.wkap.nl/ Please address all inquiries to the Publishers: Landes Bioscience / Eurekah.com, 810 South Church Street Georgetown, Texas, U.S.A. 78626 Phone: 512/ 863 7762; FAX: 512/ 863 0081 www.Eurekah.com www.landesbioscience.com The Nucleolus, edited by Mark O.J. Olson, Landes / Kluwer dual imprint / Landes series: Molecular Biology Intelligence Unit ISBN: 0-306-47873-0 While the authors, editors and publisher believe that drug selection and dosage and the specifications and usage of equipment and devices, as set forth in this book, are in accord with current recommendations and practice at the time of publication, they make no warranty, expressed or implied, with respect to material described in this book. In view of the ongoing research, equipment development, changes in governmental regulations and the rapid accumulation of information relating to the biomedical sciences, the reader is urged to carefully review and evaluate the information provided herein.

Library of Congress Cataloging-in-Publication Data The nucleolus / [edited by] Mark O.J. Olson. p. ; cm. -- (Molecular biology intelligence unit) Includes bibliographical references and index. ISBN 0-306-47873-0 1. Nucleolus. I. Olson, Mark O. J. II. Series: Molecular biology intelligence unit (Unnumbered) [DNLM: 1. Cell Nucleolus. 2. Nucleolus Organizer Region. QH 596 N9643 2004] QH596.N83 2004 571.6'6--dc22 2004001753

Dedication This book is dedicated to Joanne, whose support, patience and understanding made its publication possible and to Professors Harris Busch and Karel Smetana, who first introduced me to the nucleolus.

CONTENTS Preface ................................................................................................ xiii 1. Introduction ........................................................................................... 1 Mark O.J. Olson The Importance of the Nucleolus to the Cell ......................................... 1 The Complexity of Ribosome Biogenesis ............................................... 2 Advances in Our Understanding of the Nucleolus ................................. 4 A New Era for the Nucleolus ................................................................. 7 2. Nucleolar Ultrastructure in Vertebrates ................................................ 10 Wilhelm Mosgoeller Historical Note—Introduction ............................................................ 10 Nucleolar Internal Morphology ........................................................... 11 Nucleolar Appearance Reflects Activity ............................................... 12 Cytochemical Characterization ............................................................ 14 Structure Relates to Function? ............................................................. 15 Functional Interpretation of Nucleolar Components ........................... 17 Nucleolus in Human Sciences ............................................................. 18 3. The Nucleolar Ultrastructure in Yeast .................................................. 21 Isabelle Léger-Silvestre and Nicole Gas Cryomethods Provide Insights into the Yeast Nucleolar Organization ................................................................... 22 The Yeast Nucleolus Resembles the Higher Eukaryotic Nucleolus ...... 22 Dynamics of Assembly of the Yeast Nucleolus ..................................... 25 The Nucleolus an Integrated Part of the Nucleus ................................ 26 4. Dynamics of Nucleolar Components .................................................... 29 Thierry Cheutin, Tom Misteli and Miroslav Dundr Dynamics of Ribosomal RNA ............................................................. 30 Dynamics of the RNA Polymerase I Transcription Complex ............... 31 Dynamics of Components of Pre-Ribosomal RNA Processing and Ribosome Assembly .................................................................. 33 Dynamics of Various Nucleolar Proteins ............................................. 33 Trafficking to and through the Nucleolus ........................................... 35 Implications of Nucleolar Dynamics ................................................... 36 5. Behavior of the Nucleolus during Mitosis ............................................ 41 Danièle Hernandez-Verdun Disassembly of Nucleolus at the Beginning of Mitosis ......................... 43 Nucleolus during Mitosis .................................................................... 46 Assembly of the Nucleolus at Exit of Mitosis ....................................... 50 6. Structure and Organization of Vertebrate Ribosomal DNA ................. 58 James E. Sylvester, Iris L. Gonzalez and Edward B. Mougey rDNA Gene Structure and Variation ................................................... 60

7. The Structure of rDNA Chromatin ...................................................... 73 José M. Sogo and Fritz Thoma Approaching Chromatin Structure and Function with Psoralen Crosslinking and Restriction Enzymes ............................................. 73 Chromatin Structure of the rDNA Intergenic Spacer .......................... 74 Two Different Classes of Chromatin Coexist in the 35S Coding Region ................................................................................ 78 Regulation of rDNA Transcription ..................................................... 79 Replication of rDNA ........................................................................... 80 Inheritance and Establishment of Chromatin and Transcription during Replication ........................................................................... 81 HOT1 Dependent Recombination in rDNA ...................................... 83 8. Ribosomal DNA Transcription in Mammals ....................................... 88 Alice Cavanaugh, Iwona Hirschler-Laszkiewicz and Lawrence I. Rothblum General Background ............................................................................ 89 rDNA Transcription ........................................................................... 92 Formation of Preinitiation Complexes ............................................... 109 Regulation of rDNA Transcription ................................................... 110 9. Transcription of rDNA in the Yeast Saccharomyces cerevisiae ............. 128 Masayasu Nomura, Yasuhisa Nogi and Melanie Oakes rDNA Structure and Cis-Elements .................................................... 128 Pol I, Transcription Factors, and the Mechanism for rRNA Transcription Initiation ................................................................. 131 Regulation of rRNA Synthesis ........................................................... 138 Silencing of Pol II Transcription in rDNA ........................................ 142 Nucleolar Structures and rRNA Transcription .................................. 145 10. Three-Dimensional Organization of rDNA and Transcription .......... 154 Dominique Ploton, Marie-Françoise O’Donohue, Thierry Cheutin, Adrien Beorchia, Hervé Kaplan and Marc Thiry Organization of Actively Transcribed rDNA Genes ........................... 155 Relating rRNA Gene Transcription to Its in Situ Organization through a Three Dimensional Approach ....................................... 157 Relating 47 S Pre-rRNA Synthesis to Nucleolar Compartments at the Ultrastructural Level ............................................................ 159 In Situ Organization of a Transcriptionally Active rRNA Gene: Toward a Model ............................................................................ 160 Visualising Maturation of Pre-rRNA Transcripts............................... 162 Three-Dimensional Visualisation of Pre-rRNAs Synthesis and Processing ............................................................................... 162

11. Pre-Ribosomal RNA Processing in Multicellular Organisms .............. 170 Susan A. Gerbi and Anton V. Borovjagin Internal Modifications: 2'-O-Methylation and Pseudouridylation ..... 171 Addition of Ribosomal Proteins ........................................................ 171 Ribosomal RNA Processing ............................................................... 173 12. Pre-Ribosomal RNA Processing and Assembly in Saccharomyces cerevisiae: The Machine That Makes the Machine ............................. 199 Hendrik A. Raué Genetic Organization and Transcription of Yeast rRNA Genes ......... 200 Modification of Precursor-rRNA: How and Why? ............................ 200 Pre-rRNA Processing ......................................................................... 202 Cis-Acting Elements Required for Pre-rRNA Processing.................... 204 The Processing/Assembly Machinery and Its Components ................ 207 13. The snoRNPs and Related Machines: Ancient Devices That Mediate Maturation of rRNA and Other RNAs ........................ 223 Edouard Bertrand and Maurille J. Fournier Early History ..................................................................................... 224 More Recent Landmarks and Breakthroughs ..................................... 225 snoRNP Structure and Function ....................................................... 225 C/D snoRNPs ................................................................................... 225 H/ACA snoRNPs .............................................................................. 229 The MRP snoRNP ............................................................................ 230 What Roles Do the C/D and H/ACA snoRNPs Play in rRNA Cleavage? ........................................................................ 230 Modification of rRNA and U6 snRNA by snoRNPs ......................... 231 Effects of the Nm and Y Modifications ............................................. 232 Do Modifying snoRNPs Influence Other Aspects of Ribosome Synthesis? ................................................................. 232 When Do snoRNPs Act? ................................................................... 233 Related RNAs and RNPs ................................................................... 234 Archaeal Guide RNAs ....................................................................... 234 Small Cajal Body RNAs (scaRNAs) ................................................... 236 Telomerase RNA ............................................................................... 237 Tissue-Specific, Imprinted snoRNAs ................................................. 237 Just What Is a snoRNA, Anyway? ...................................................... 239 Biogenesis of snoRNPs ...................................................................... 239 Synthesis of snoRNAs ....................................................................... 239 snoRNP Assembly ............................................................................. 240 Trafficking and Localization .............................................................. 243 Exploiting snoRNPs .......................................................................... 245

14. Ribosomal Subunit Assembly ............................................................. 258 Jesús de la Cruz, Dieter Kressler and Patrick Linder Methods to Approach Ribosomal Subunit Assembly ......................... 259 Compilation of Trans-Acting Factors Involved in r-Subunit Assembly ................................................................... 263 From the Processosome to the Tollervey Model ................................ 271 The 66S Preribosomal Particles ......................................................... 276 Ribosomal Subunit Assembly in Higher Eukaryotes .......................... 279 15. Nuclear Export of Ribosomal Subunits .............................................. 286 Arlen W. Johnson Export of the Large Subunit .............................................................. 287 Export to the Cytoplasm ................................................................... 290 Export of 40S Subunits ..................................................................... 296 16. Proteomics of the Nucleolus ............................................................... 302 Yun Wah Lam, Archa H. Fox, Anthony K.L. Leung, Jens S. Andersen, Matthias Mann and Angus I. Lamond Isolation of Nucleoli .......................................................................... 303 Separation of Nucleolar Proteins for MS Analysis .............................. 305 MS Analysis and Protein Identification ............................................. 307 Validation of MS Data ...................................................................... 309 Presentation and Publication of Data ................................................ 310 17. Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus ........................................................................ 315 Thilo Sascha Lange The Relationship between Spliceosomal RNAs and the Nucleolus .... 316 Mechanisms of Nucleolar Trafficking of Spliceosomal RNAs ............ 320 18. Nontraditional Roles of the Nucleolus ............................................... 329 Mark O.J. Olson Viral Components in the Nucleolus .................................................. 329 Regulation of Tumor Suppressor and Oncogene Activities ................ 332 Cell Cycle Regulation in Yeast ........................................................... 333 Signal Recognition Particle Assembly ................................................ 334 Nucleolar Processing of RNA pol III Synthesized Transcripts ........... 334 Tissue-Specific Expression of Small Nucleolar RNAs ........................ 335 Yeast as a Model System for Aging .................................................... 336 Aging in Mammals ............................................................................ 336 Index .................................................................................................. 343

EDITOR Mark O.J. Olson Professor and Chairman Department of Biochemistry The University of Mississippi Medical Center Jackson, Mississippi, U.S.A. Chapters 1 and 18

CONTRIBUTORS Jens S. Andersen Protein Interaction Laboratory University of Southern Denmark Odense, Denmark Chapter 16

Jesús de la Cruz Department of Genetics Faculty of Biology University of Seville Sevilla, Spain Chapter 14

Adrien Beorchia DTI, UMR 6107 UFR de Sciences Reims, France Chapter 10

Miroslav Dundr National Cancer Institute National Institutes of Health Bethesda, Maryland, U.S.A. Chapter 4

Edouard Bertrand Institut de Génétique Moléculaire CNRS UMR 5535 Montpellier, France Chapter 13

Anton V. Borovjagin Division of Biology and Medicine Brown University Providence, Rhode Island, U.S.A. Chapter 11

Alice Cavanaugh Henry Hood Research Program Sigfried and Janet Weis Center for Research Geisinger Clinic Danville, Pennsylvania, U.S.A. Chapter 8

Maurille J. Fournier Department of Biochemistry and Molecular Biology Lederle Graduate Research Center University of Massachusetts Amherst, Massachusetts, U.S.A. Chapter 13

Archa H. Fox MSI/WTB Complex University of Dundee Dundee, U.K. Chapter 16

Nicole Gas LBME/IBCG du CNRS Université Paul Sabatier Toulouse, France Chapter 3

Thierry Cheutin National Cancer Institute National Institutes of Health Bethesda, Maryland, U.S.A. Chapters 4, 10

Susan A. Gerbi Division of Biology and Medicine Brown University Providence, Rhode Island, U.S.A. Chapter 11

Iris L. Gonzalez A.I. duPont Hospital for Children Wilmington, Delaware, U.S.A. Chapter 6

Angus I. Lamond MSI/WTB Complex University of Dundee Dundee, U.K. Chapter 16

Danièle Hernandez-Verdun Institut Jacques Monod CNRS, Universités Paris VI et Paris VII Paris, France Chapter 5

Thilo Sascha Lange Division of Biology and Medicine Brown University Providence, Rhode Island, U.S.A. Chapter 17

Iwona Hirschler-Laszkiewicz Henry Hood Research Program Sigfried and Janet Weis Center for Research Geisinger Clinic Danville, Pennsylvania, U.S.A.

Isabelle Léger-Silvestre LBME/IBCG du CNRS Université Paul Sabatier Toulouse, France Chapter 3

Chapter 8

Arlen W. Johnson Section of Molecular Genetics and Microbiology Institute for Cellular and Molecular Biology The University of Texas at Austin Austin, Texas, U.S.A. Chapter 15

Hervé Kaplan IFR53 Reims, France Chapter 10

Dieter Kressler Division of Biochemistry Biozentrum of the University of Basel Basel, Switzerland Chapter 14

Yun Wah Lam MSI/WTB Complex University of Dundee Dundee, U.K. Chapter 16

Anthony K.L. Leung MSI/WTB Complex University of Dundee Dundee, U.K. Chapter 16

Patrick Linder Département de Biochimie Médicale Centre Médical Universitaire University of Geneva Geneva, Switzerland Chapter 14

Matthias Mann Protein Interaction Laboratory University of Southern Denmark Odense, Denmark Chapter 16

Tom Misteli National Cancer Institute National Institutes of Health Bethesda, Maryland, U.S.A. Chapter 4

Wilhelm Mosgoeller Institute of Cancer Research University of Vienna Vienna, Austria Chapter 2

Edward B. Mougey Nemours Children's Clinic Jacksonville, Florida, U.S.A. Chapter 6

Yasuhisa Nogi Department of Molecular Biology Saitama Medical School Moroyama, Iruma-Gun Saitama, Japan Chapter 9

Masayasu Nomura Department of Biological Chemistry University of California-Irvine Irvine, California, U.S.A.

Hendrik A. Raué Department of Biochemistry and Molecular Biology IMBW, BioCentrum Amsterdam Vrije Universiteit Amsterdam, The Netherlands Chapter 12

Lawrence I. Rothblum Henry Hood Research Program Sigfried and Janet Weis Center for Research Geisinger Clinic Danville, Pennsylvania, U.S.A. Chapter 8

José M. Sogo Department of Biology Institute of Cell Biology ETH Honggerberg Zurich, Switzerland Chapter 7

Chapter 9

Marie-Françoise O’Donohue Unité Médian, CNRS UMR 6142 UFR de Pharmacie Reims, France Chapter 10

Melanie Oakes Department of Biological Chemistry University of California-Irvine Irvine, California, U.S.A.

James E. Sylvester Nemour's Children's Clinic Jacksonville, Florida, U.S.A. Chapter 6

Marc Thiry Laboratoire de Biologie Cellulaire et Tissulaire Université de Liège Liège, France Chapter 10

Chapter 9

Dominique Ploton Biologie Cellulaire UFR Médicine Reims, France

Fritz Thoma Department of Biology Institute of Cell Biology ETH Honggerberg Zurich, Switzerland

Chapter 10

Chapter 7

PREFACE

M

any topics in the sciences go through cycles of waxing and waning interest, with the nucleolus being no exception. The nucleolus was initially described in the first half of the 19th century and attracted considerable attention by the early cytologists. However, after the turn of that century the subject was relatively dormant until the 1960s, when its function as the primary factory for ribosome biogenesis was established. During the 70s and 80s the steady research activities of a relatively small number of laboratories laid the foundation for the extraordinary progress of the 90s and the early 21st century. Because of the marked accumulation of knowledge in the past decade, it seemed appropriate to compile a volume that summarizes the status of the field as of the middle of 2003. This book attempts to accomplish that at a high level of detail. For readers who want to study the topic in greater depth, numerous references are provided in the individual chapters. Because of space limitations, not every aspect of the nucleolus is covered. I apologize to the many researchers who have made significant contributions to nucleolar research but whose work is not discussed in this book. Over the past half-century, a few authors have assembled comprehensive volumes that covered the current understanding of the structure and function of nucleolus. The first of these was the classic volume by Busch and Smetana, (Busch H, Smetana, K. The Nucleolus. New York: Academic Press, 1970). Significantly, this book was published at the end of the decade in which it was established that the ribosomal RNA genes are located in the nucleolus organizer regions on chromosomes and that the nucleolus is the source of ribosomal RNA. In the next 10-15 years the field had matured to the point where much was learned about the locations and multiplicity of ribosomal RNA and protein genes, transcription and maturation of prerRNA, assembly of pre-ribosomal particles and regulation of ribosome biogenesis. Much of the information was collected by Jordan and Cullis from a symposium at the 200th meeting of the Society for Experimental Biology in 1980 in Oxford, UK, and compiled into a volume also entitled The Nucleolus (Jordan EG, Cullis CA. The Nucleolus. New York: Cambridge University Press, 1982). Even though it was published in 1985, the most recent comprehensive volume The Nucleolus and Ribosome Biogenesis (Hadjiolov AA. The Nucleolus and Ribosome Biogenesis. New York: Springer Verlag, 1985) remains an extremely useful source of information on this subject. With the rapid growth of new information, a few books have focused on specialized aspects of the nucleolus. A more recent book (Thiry M, Goessens G. The Nucleolus during the Cell Cycle. Austin: R.G. Landes Company, 1996) concentrates on the ultrastructural features of the nucleolus and also provides a concise historical timeline of discoveries concerning this sub-nuclear body.

Major advances in the area of rDNA transcription are found in a collection published in 1998 (Paule MR, ed., Transcription of Ribosomal RNA Genes by Eukaryotic RNA Polymerase I. Georgetown: R.G. Landes Company, 1998). Over the years, numerous reviews have provided timely coverage of progress in the field; most of these are cited in the chapters. I was first exposed to the nucleolus when I joined the faculty at Baylor College of Medicine. Coming from a classical biochemistry background with research experience in protein chemistry, I barely knew the nucleolus existed. Fortunately, I was enlightened on the subject by Harris Busch and Karel Smetana, who had just finished writing their book on the nucleolus. They had accomplished the daunting task of assembling most of the accumulated knowledge on the subject and concentrating it into one location. Over the next few years, I slowly absorbed not only the science of the nucleolus, but also the culture and personalities surrounding it. This has been an interesting and rewarding journey; but more importantly, it has been a pleasure to see the nucleolus transformed from a mysterious black box into a ribosome assembly factory whose machinery has been inventoried and whose manufacturing processes are generally understood. I am also honored to have the opportunity to be the editor of the current book. Because I was pointed down this path by the pioneers in archiving nucleolar information, I dedicate this book to Harris Busch and Karel Smetana. Mark O.J. Olson

CHAPTER 1

Introduction Mark O.J. Olson

The Importance of the Nucleolus to the Cell

T

he nucleolus is the most prominent structure in the nucleus of a cell observed by light or electron microscopy. In cells as in other aspects of life, the level of visibility usually depends on the importance of the job to be performed. Actively dividing cells have an enormous demand for proteins and that need is satisfied by the millions of ribosomes per cell in vertebrates.1 An equal number of new ribosomes must be produced in every generation; to accommodate these requirements, eukaryotes have evolved highly specialized and efficient factories for ribosome assembly in nucleoli. The size of a nucleolus is generally proportional to its rate of ribosomal subunit output; hence it has been described as “formed by the act of building a ribosome.”2 In fact, the nucleolus can occupy as much as 25% of the total nuclear volume in cells that are producing large amounts of protein.1 Thus, the prominence of the nucleolus is largely determined by the service it provides for the cell. The visibility of a structure can also be related to its ability to marshal and control available resources. In most eukaryotic cells, ribosome assembly facilities are concentrated in one or a limited number of regions, rather than being dispersed throughout the nucleus. For each nucleolus, a set of tandemly-repeated genes coding for preribosomal RNA (rDNA) serves as its foundation. All eukaryotic species have multiple copies of these and the number of them varies from less than 50 to several thousand per haploid; see Hadjiolov3 for a compilation of these from about 150 different species and Chapter 6 for details on their organization. These genes are usually located in the secondary constrictions of a few chromosomes of mitotic cells. Because they have the ability to initiate the formation of nucleoli during interphase, these segments of the chromosomes are called nucleolus organizer regions or NORs. It is theoretically possible for one nucleolus to arise from each NOR. For example, in human somatic cells there are five separate NORs, which would yield 10 nucleoli in diploid cells. However, nucleoli have a tendency to fuse as the cell progresses through interphase4,5 with the result that only one to four nucleoli are typically found in each mammalian cell. In yeast there is only one crescent-shaped nucleolus, which occupies one-third to one half of the volume of the nucleus.6 The evolutionary advantages of having the system of ribosome assembly concentrated in one or a few locations are not entirely clear; this issue is discussed briefly below and in Chapters 2, 9 and 14. Finally, an emerging factor in the prominence of the nucleolus is the presence of components associated with functions not related to ribosome biogenesis. The list of the non-traditional roles of the nucleolus is growing; of special note is the accumulating evidence that a relatively solid subnuclear compartment is important for temporary sequestration of regulatory factors and synthetic machinery and for assembly of some macromolecular complexes (see Chapters 17 and 18). These newly discovered tasks undoubtedly add extra capacity and mass to an already-busy subcellular factory.

The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

2

The Nucleolus

The Complexity of Ribosome Biogenesis At first glance, eukaryotic ribosome assembly appears to be a relatively simple process, which may be summarized as a series of a few major operations (Fig. 1). The initial step is the transcription of preribosomal RNA (pre-rRNA) from the multiple copies of rDNA. Before transcription is complete, proteins and small nucleolar RNAs (snoRNAs) associate with the pre-rRNA. This RNA is modified by pseudouridylation and methylation and eventually processed into the three major ribosomal RNAs. Ribosomal proteins and 5S rRNA are incorporated into the maturing preribosomal particles at various stages in the process. These preribosomal particles mature and are eventually exported to the cytoplasm. In actuality, ribosome biogenesis is very complex. The preribosomal RNA originates from transcription of the rDNA by RNA polymerase I, which takes place at the border between the fibrillar center and the dense fibrillar components (DFCs) of the nucleolus (see Chapters 2 and 3). The DFCs contain the nascent preribosomal particles, which after maturation eventually become the granular components (GCs). During and after transcription many manipulations are performed on the maturing preribosomal particles; these proceed through multiple identifiable stages of assembly and processing. Several unique intermediate particles have been isolated and characterized (see Chapters 12 and 14). The complexity is also compounded by the number of components involved in making the final products. At least one hundred different snoRNAs and 150 proteins (Chapters 13, 14 and 16) are involved in building the ribosomal subunits containing four species of RNA and some 80 unique ribosomal proteins. Why do we need such a complex system to assemble ribosomes? Clues to this question are found in the structure and function of the ribosome itself. The landmark X-ray crystallographic analyses of prokaryotic ribosomes reveal that it is a highly compact structure and that the rRNAs account for most of the mass.7 Although the ribosomal proteins make significant contributions to the volume of the ribosome, their primary role seems to be to stabilize the RNA core. High-resolution structures of higher eukaryotic ribosomes are not yet available; however, cryo-electron microscopy of yeast ribosomes at 15 Å resolution has shown that the cores of eukaryotic ribosomes are very similar to prokaryotic ones.8 Furthermore, studies on ribosome structure revealed that RNA rather than proteins in the large subunit carries out the peptidyl transferase reaction.9 In other words, the ribosome is a ribozyme. For the catalytic mechanism to operate properly, the RNA base pairs must be correctly located and the ribosomal proteins must be bound to the appropriate segments of the RNA. The length of the pre-rRNA (up to 13 kilobases) creates innumerable opportunities for improper folding and/or base pairing at incorrect locations. For example, there is a central pseudoknot in the small subunit that must be properly positioned for ribosome function and its formation is delayed until late in the process.10 In addition, a large number of post-transcriptional modifications must be placed at precise locations in the RNA sequence. The importance of these relatively subtle alterations in RNA structure is illustrated by the fact that the absence of a few pseudouridine residues at key locations greatly impairs ribosome function.11 The ribosome biogenesis machinery assures the fidelity of the assembly process by forcing it to follow a strict order of events. This maintains the RNA in a relatively loose structure to provide access to processing, modification and assembly factors until the desired structure of the final product is achieved. To this end, the assembly machinery has evolved to include numerous factors such as helicases and molecular chaperones to aid in the construction of the ribosomal subunits (see Chapters 14 and 16). Why do eukaryotes need the elaborate assembly system and prokaryotes do not? The classic work by Nomura and colleagues showed that E. coli ribosomal subunits can be on reconstituted from their constituent RNAs and proteins.12,13 These studies indicated that prokaryotic ribosome assembly is largely dependent on the ribosomal components themselves, not on exogenous factors. The higher level of complexity of eukaryotic ribosome assembly and the requirement for additional assembly factors are related to at least two issues. First, eukaryotic ribosomes are much larger than prokaryotic ones; the added proteins and longer RNAs increase the mass of yeast ribosomes by about 30% over E. coli ribosomes.14 Although the basic

Introduction

3

Figure 1. Major steps in eukaryotic ribosome biogenesis. The process starts with transcription of preribosomal RNA (pre-rRNA) from multiple copies of the genes for pre-rRNA (rDNA). Nonribosomal proteins (open circles) and small nucleolar RNAs (snoRNAs; open rectangles) associate with the nascent transcript. The pre-rRNA is methylated and pseudouridylated under the guidance of the snoRNAs. 5S rRNA, a component of the 60S subunit, is added to the maturing complex. The pre-rRNA undergoes a series of cleavages ultimately resulting in 18S, 5.8S and 28S (25S in yeast) rRNAs. The complex is split into the two precursor particles for the small (40S) and large (60S) ribosomal subunits. Ribosomal proteins (black circles) are added to the precursor complexes at various stages of assembly. The nearly mature subunits are exported to the cytoplasm through the nuclear pore complexes with the aid of adaptor proteins. The small and large subunits are eventually incorporated into ribosomes in the cytoplasm.

mechanisms of translation are generally the same in both classes of organisms, the difference in size appears to be related to ribosome function. In fact, there are significant differences in the mechanisms of initiation and elongation as well as in requirements for protein folding and transport in eukaryotic protein synthesis systems compared to prokaryotic ones. These differences are likely to be reflected in the variations in size and structure of the ribosomes to accommodate more complex translation factors (discussed in ref. 8). The longer RNAs in eukaryotes offer more opportunities for misfolding and the added proteins make assembly more difficult. In addition, eukaryotic ribosomes must be transported from the nucleus to the cytoplasm (see Chapter 15), whereas prokaryotes do not have to deal with this issue. Taken together, these observations support the idea that a larger and more complex ribosome requires a more complicated assembly system.

4

The Nucleolus

The second issue deals with the numbers of post-transcriptional modifications of ribosomal RNA in eukaryotes vs. prokaryotes and the mechanisms by which the modifications occur. Approximately 100 sites are methylated at the 2'-O position in vertebrates compared to around four in prokaryotes.15,16 Similarly, about 100 pseudouridines are found in mammalian rRNA, but only 11 in E.coli.17 Aside from the numbers, the most striking difference is how the rRNA is modified in prokaryotes versus eukaryotes. In prokaryotes, the modifications are introduced by a few enzymes that recognize specific sites. Eukaryotes have evolved a much more elaborate system, in which the sequence specificity resides not in the enzymes, but in numerous snoRNAs that are contained in larger RNP complexes (see Chapter 13). These snoRNAs direct the enzymes to the designated locations for modification by sequence-specific base pairing. Consequently, a large number of snoRNAs are required to carry out these tasks, which greatly increases the complexity of the system. Do cells need nucleoli? This is a provocative question for researchers who have devoted most of their careers to studies on the nucleolus. Intuitively, we would expect that organized ribosome assembly would have a selective advantage for the organism. In a centralized system of production, each of the multiple sites of assembly would have rapid access to the ribosomal precursor components as well as the factors for assembly.18 However, some cell lines are still viable after being engineered so that nucleoli are not clearly visible. For example, strains of yeast have been constructed in which the tandem array of rDNA has been deleted and replaced with multiple copies of rDNA on a plasmid; these strains are able to grow reasonably well without organized nucleoli.19 Other arrangements of rDNA that produce atypical nucleoli are discussed in Chapter 9. Nucleoli form because the numerous adjacent copies of rDNA concentrate the multiple ribosome assembly lines in a limited number of locations. The tandemly-repeated genes arose from gene duplication, which gave eukaryotic cells the capacity to produce large numbers of ribosomes and massive amounts of protein. As a consequence, the rDNA sequences have been placed in close proximity to one another within the NORs. Thus, it seems likely that nucleoli evolved as a result of gene duplication rather than out of a need for centralizing ribosome assembly. However, the fact that newly formed nucleoli tend to fuse during interphase suggests that there is some selective advantage to concentrating these cellular resources. The true benefits of a centralized ribosome biogenesis apparatus are not fully understood.

Advances in Our Understanding of the Nucleolus Two important discoveries during the twentieth century were crucial in establishing the nucleolus as the location of ribosome biogenesis. First, several laboratories determined that the nucleolus organizer regions contain the genes for ribosomal RNA.20-23 Second, the nucleolus was shown to be the site of synthesis of cytoplasmic ribosomal RNA.24-26 These findings laid the foundation for intensive research during the latter half of the twentieth century leading to our current detailed knowledge of the nucleolus. The unprecedented progress of the past decade and current interest are due to several developments. The first is the power of yeast genetics, which has produced rapid advances and, in some cases, breakthroughs in defining individual steps in the ribosome biogenesis pathway. Although yeast has served as an extremely useful model organism, there are significant differences in ribosome biogenesis between yeast and vertebrates. Therefore, separate chapters are provided for the two classes of organisms on several of the topics. Second, the developing field of proteomics, made possible by advances in mass spectrometry and the availability of nearly complete genomic information from several organisms, has facilitated the compilation of protein databases. In turn, this has made it feasible to determine the protein compositions of the nucleolus and many of its derived sub-particles in a few species. Finally, there has been the realization that the nucleolus is not simply a factory for ribosome assembly, but that it performs several other tasks, many of which are not clearly defined at this writing.

Introduction

5

This book is intended to serve as an update on the current status of our comprehension of the nucleolus; the primarily focus is on research performed within the last ten years. Earlier work and historical perspectives are covered in more detail in previous volumes.3,27-29 The following paragraphs summarize the highlights of the research described in the chapters of this book:

Nucleolar Ultrastructure Correlates with Cell Physiology and Biochemical Events Electron microscopic studies have been performed on vertebrate nucleoli for several decades (Chapter 2) and we now have an extensive database for correlation of specific morphologies with a variety of physiological conditions and disease states. As indicated above, yeasts have served as model systems for studying nucleolar mechanisms; however, the morphological details of yeast nucleoli have been neglected until recently. Technological advances in the past few years now make it possible to envision how mutations and their consequent biochemical alterations affect yeast nucleolar ultrastructure (Chapter 3). Finally, the developing technology of electron tomography enables the visualization of pre-rRNA transcription and maturation in three dimensions (Chapter 10) and provides us with snapshots of individual steps in ribosome biogenesis. These state-of-the art tools now make it feasible to relate nucleolar structure to function at a high level of resolution.

The Nucleolus Is Highly Dynamic The technique known as fluorescence recovery after photobleaching (FRAP) and related method have shown that nucleolar components rapidly exchange with the surrounding nucleoplasm (Chapter 4). The results of these studies have changed our concept of the nucleolus as a relatively static structure to one that is more like a “swarm of bees” in which the components are continuously buzzing in and out. This technology has cleared the path toward a more thorough understanding of how the nucleolus assembles, disassembles and communicates with the rest of the cell.

Nucleolar Components Distribute to a Variety of Cellular Locations during Mitosis One of the most remarkable features of the cell cycle in higher eukaryotes is that the nucleoli disappear during mitosis; disassembly begins in prophase and reassembly occurs in telophase (Chapter 5). These processes are controlled by cyclin-dependent protein kinases (CDKs), which act at several levels including regulating transcription by RNA polymerase I. During mitosis, the transcriptional apparatus remains associated with the chromosomal NORs. However, the processing apparatus distributes to several locations, depending on the stage of mitosis. It is noteworthy that partially processed preribosomal RNA is preserved during mitosis in association with much of the processing machinery. Thus, not only is transcription shut down during mitosis, but processing is also suppressed until it is time for nucleoli to reappear. The components from the pre-existing pre-rRNP complexes, which are contained in prenucleolar bodies during telophase, are transferred into the reassembling nascent nucleoli.

The Genes for Pre-rRNA Are Conserved in Organization and in Sequences of Critical Segments The extensive database of rDNA sequences has made it possible to analyze the evolution of these genes and to predict the secondary structures of the RNAs for which they code (Chapter 6). All of these genes have the same general organization. Although there is considerable sequence variability among the genes from different species and even among individuals within a species, the overall secondary structures of both the small and large rRNAs are conserved from eubacteria through higher eukaryotes. This conservation of secondary structure is undoubtedly dictated by the fact that ribosomes of all species act with a common mechanism of protein synthesis, with the RNA acting as a ribozyme.

6

The Nucleolus

There Are Two Different Classes of rDNA Chromatin One way of regulating expression of pre-rRNA is by controlling chromatin structure. Actively-transcribed rDNA genes do not contain nucleosomes, at least in their typical form, whereas inactive genes for pre-rRNA exhibit a traditional nucleosomal structure (Chapter 7). This is in contrast with the structure of chromatin actively transcribed by RNA polymerase II, in which nucleosomes are clearly detectable. This mechanism of programming chromatin for transcriptional potential is another unusual aspect of the nucleolus.

Transcription by RNA Polymerase I (Pol I) Requires Multiple Factors and Is Regulated at Several Different Levels The Pol I enzyme and associated factors are well characterized both in vertebrates and in yeast (Chapters 8 and 9). Besides the dozen or more subunits of Pol I, optimal rDNA transcription requires additional factors for initiation, elongation and termination. The mechanisms by which these stages of transcription occur have been studied in detail. Transcription of the rDNA is regulated essentially two ways: by changing the number of active rDNA genes via alterations in chromatin structure (Chapter 7) and by adjusting the rate of transcription. In the latter case, this is achieved by modifications of the associated transcription factors and possibly Pol I itself. New technology also has allowed us to visualize the transcriptional process in three dimensions (Chapter 10).

Pre-rRNA Maturation Is Precisely Directed by snoRNAs The pathways of pre-rRNA processing are now well understood, in both yeast and vertebrates (Chapters 11 and 12). The enzymatic machinery that carries out this is reasonably well defined in yeast, but incompletely characterized in higher eukaryotes. We owe much of our advanced understanding of this to the discovery and characterization of the one hundred plus small nucleolar RNAs (snoRNAs), which are covered in detail in Chapter 13. Segments of the snoRNAs form base pairs with specific sequences of pre-rRNA; along with their associated proteins, they serve as the guides for pre-rRNA cleavage as well as for its ribose methylation and pseudouridylation.

Ribosome Biogenesis Is Spatially and Temporally Coordinated and Utilizes Multiple Trans-Acting Factors Chapter 14 describes how ribosome subunit assembly proceeds through a series of discrete pre-ribosomal particles. Several of these have been isolated on the basis of unique components and their protein and RNA compositions have been determined. The non-ribosomal proteins contained in these particles serve as trans-acting factors that modify and process the pre-rRNA and facilitate the assembly of the pre-ribosomal particles. The latter class of proteins includes RNA helicases, nucleoside triphosphatases and molecular chaperones.

Ribosomal Subunits Are Transported to the Cytoplasm by Adapting the Nuclear Export System All eukaryotic cells have systems for export of components from the nucleus to the cytoplasm via the nuclear pore complex. The nuclear export of ribosomal subunits utilizes this same receptor-mediated process (Chapter 15). The export of the large ribosomal subunit is enabled by an adaptor protein that binds the nearly completed subunit to the export receptor. Export of the small ribosomal subunit is less well understood, but it also seems to utilize adaptor proteins.

Introduction

7

Virtually All of the Proteins in Nucleoli of Mammalian Cells Have Been Identified The availability of the sequences of genomes of several organisms together with advances in mass spectrometry have facilitated rapid progress in the field of proteomics. This technology has now been applied to the identification of nearly all proteins contained in vertebrate nucleoli (Chapter 16). The combined results of these analyses indicate that HeLa cell nucleoli contain over 350 proteins. Although a substantial proportion of these proteins are clearly related to ribosome biogenesis, many of them seem to be associated with non-traditional functions of the nucleolus. This inventory of nucleolar proteins will serve as one cornerstone for defining all functions of the nucleolus.

The Nucleolus Engages in Activities Not Related to Ribosome Biogenesis An increasing number of unexpected components have been found in the nucleolus over the past decade. Chapter 17 demonstrates that several small RNAs traffic through the nucleolus and these are modified in various ways on their way to their final destination in the cell. In addition to the small RNA modification, the nucleolus also performs a variety of non-traditional functions including interactions with viral components, regulation of tumor suppressor and oncogene activities, control of the cell cycle, signal recognition particle assembly and modulation of telomerase function (Chapter 18). These findings confirm that the nucleolus is not simply a factory for assembly of ribosomes, but that it also plays several other essential roles in the cell, many of which are not well understood.

A New Era for the Nucleolus The nucleolus is no longer the mysterious black box that it was a few decades ago. The ribosome assembly factory has been thoroughly dismantled, its machinery examined and its workings are now generally understood. We must now to pause and ask where nucleolar research will go from here. In spite of the achievements of the past few years, we don’t know everything there is to know about the nucleolus by any means. However, the challenges of the future will be different from those of the past half-century. Here are a few of the unresolved issues: 1) Although we have a nearly complete catalog of the nucleolar proteins, we don’t know precisely what many of them do. We can expect continued progress in determining the functions of individual proteins in the assembly process and detailed mechanisms of how these proteins work. 2) How much of what we have learned about ribosome biogenesis in yeast can now be applied to higher eukaryotes? Although the basic mechanisms of ribosome assembly are very similar in yeast and metazoans, there are some striking differences. For example, the behavior of nucleoli from the two classes of organisms during mitosis is markedly different and there are no orthologs for many of the proteins. Therefore, it will be of great interest to determine the similarities and differences between yeast and vertebrates. 3) The precise three-dimensional location of non-ribosomal factors in the intermediate pre-ribosomal particles will be essential for a complete understanding of how ribosomes are assembled. The technology is now available for isolation of the particles and for medium, if not high-resolution studies to determine the locations of the proteins and RNAs. 4) Our understanding of the regulation of ribosome biogenesis is in its infancy. Although much is known about the control of rDNA transcription and about the required sequence of events in processing and assembly, what really regulates these steps is far from clear. Furthermore, we have only a rudimentary knowledge of how the nucleolus communicates with the cytoplasm and the rest of the nucleoplasm. In the latter case, further investigation should clarify the intriguing relationship between the nucleolus and the Cajal bodies as well as other nucleoplasmic structures. 5) The newest frontier will be the elucidation of the non-traditional roles of the nucleolus. A few of these functions are reasonably well understood, but the advantages of their nucleolar location are not obvious. For others, the significance of the components being in the nucleolus is far

The Nucleolus

8

from clear. Many of these functions are related to diseases and aging, which should motivate researchers to put greater efforts into the non-conventional nucleolar functions. The above predictions of future research suggest trends in somewhat opposite directions: elucidation of structures and fundamental biochemical mechanisms as well as integration of all of the component parts of the nucleolus. Both of these approaches are needed to fully comprehend the nucleolus; however, we cannot predict what advances unforeseen technologies will bring. Undoubtedly, we can expect many more surprises from the nucleolus in the twenty-first century.

Acknowledgements The editor acknowledges the following individuals for their helpful discussions and critical reading of the manuscript: A. Szebeni, N. Huang, S. Negi, K. Hingorani, M. Dundr, D. Brown and M. Hebert. I am also most grateful to the authors of this book for their cooperation in exchanging information and sharing drafts of their manuscripts as well as for their enthusiasm for participating in this project. Finally, I thank Romie Brown for handling the correspondence associated with this book.

References 1. Alberts B, Johnson A, Lewis J et al. The Molecular Biology of the Cell. 4th ed. New York: Garland Publishing, 2002:331-351. 2. Melese T, Xue Z. The nucleolus: an organelle formed by the act of building a ribosome. Curr Opin Cell Biol 1995; 7(3):319-324. 3. Hadjiolov AA. The Nucleolus and Ribosome Biogenesis. New York: Springer Verlag, 1985. 4. Anastassova-Kristeva M. The nucleolar cycle in man. J Cell Sci 1977; 25:103-110. 5. Wachtler F, Schwarzacher HG, Smetana K. On the fusion of nucleoli in interphase. Eur J Cell Biol 1984; 34(1):190-192. 6. Warner JR. The nucleolus and ribosome formation. Curr Opin Cell Biol 1990; 2(3):521-527. 7. Ramakrishnan V, Moore PB. Atomic structures at last: the ribosome in 2000. Curr Opin Struct Biol 2001; 11(2):144-154. 8. Spahn CM, Beckmann R, Eswar N et al. Structure of the 80S ribosome from Saccharomyces cerevisiae-tRNA-ribosome and subunit-subunit interactions. Cell 2001; 107(3):373-386. 9. Nissen P, Hansen J, Ban N et al. The structural basis of ribosome activity in peptide bond synthesis. Science 2000; 289(5481):920-930. 10. Lafontaine DL, Tollervey D. The function and synthesis of ribosomes. Nat Rev Mol Cell Biol 2001; 2(7):514-520. 11. King TH, Liu B, McCully RR et al. Ribosome structure and activity are altered in cells lacking snoRNPs that form pseudouridines in the peptidyl transferase center. Molecular Cell 2003; 11(2):425-435. 12. Traub P, Nomura M. Structure and function of E. coli ribosomes. V. Reconstitution of functionally active 30S ribosomal particles from RNA and proteins. Proc Natl Acad Sci USA 1968; 59(3):777-784. 13. Nomura M, Erdmann VA. Reconstitution of 50S ribosomal subunits from dissociated molecular components. Nature 1970; 228(273):744-748. 14. Doudna JA, Rath VL. Structure and function of the eukaryotic ribosome: the next frontier. Cell 2002; 109(2):153-156. 15. Maden BEH. Eukaryotic rRNA methylation: the calm before the Sno storm. Trends Biochem Sci 1998; 23(11):447-450. 16. Ofengand J, Malhotra A, Remme J et al. Pseudouridines and pseudouridine synthases of the ribosome. Cold Spring Harbor Symp Quant Biol 2001; 66:147-159. 17. Ofengand J. Ribosomal RNA pseudouridines and pseudouridine synthases. FEBS Lett 2002; 514(1):17-25. 18. Olson MO, Hingorani K, Szebeni A. Conventional and nonconventional roles of the nucleolus. Int Rev Cytol 2002; 219:199-266. 19. Nierras CR, Liebman SW, Warner JR. Does Saccharomyces need an organized nucleolus? Chromosoma 1997; 105(7-8):444-451. 20. Brown DD, Gurdon JB. Absence of ribosomal RNA synthesis in the anucleolate mutant of Xenopus laevis. Proc Nat Acad Sci USA 1964; 51:139-146.

Introduction

9

21. Birnstiel, ML, Wallace H, Sirlin JL et al. Localization of the ribosomal DNA complements in the nucleolar organizer region of Xenopus laevis. Nat Cancer Inst Monogr 1966; 23:431-448. 22. Ritossa F, Atwood, K, Lindsley D et al. On the chromosomal distribution of DNA complementary to ribosomal and soluble RNA. Nat Cancer Inst Monogr 1966; 23:449-472. 23. Ritossa F, Spiegelman, S. Location of DNA complementary to rRNA in the nucleolus organizer region of Drosophila melanogaster. Proc Nat Acad Sci USA 1965; 53:737-745. 24. Perry RP, Hell A, Errera M. The role of the nucleolus in ribonucleic acid and protein synthesis. I. Incorporation of cytidine into normal and nucleolar inactivated HeLa cells. Biochim Biophys Acta 1961; 49:47-57. 25. Edstrom JE, Grampand N, Schor N. The intracellular distribution and heterogeneity of ribonucleic acid in starfish oocytes. J Biophys Biochem Cytol 1961; 11:549-557. 26. Perry RP. The cellular sites of synthesis of ribosomal and 4S RNA. Proc Nat Acad Sci USA 1962; 48:2179-2186. 27. Busch H, Smetana, K. The Nucleolus. New York: Academic Press, 1970. 28. Jordan EG, Cullis CA. The Nucleolus. New York: Cambridge University Press, 1982. 29. Thiry M, Goessens G, eds. The Nucleolus during the Cell Cycle. Austin: R.G. Landes Company, 1996.

10

The Nucleolus

CHAPTER 2

Nucleolar Ultrastructure in Vertebrates Wilhelm Mosgoeller

Abstract

I

n this chapter I would like to summarize some structural research on the “nucleus of the nucleus”—the nucleolus. It is important to note that structure analysis goes beyond the description of what a component looks like when observed through the microscope. It enables a understanding of functional relevance in so far as structures relate to functions. Because nucleolar activity indicates ribosome biogenesis and cellular activity I would like to give an overview on the possibilities and limitations of functional structure analysis related to ribosome biogenesis in the nucleolus.

Historical Note—Introduction

In the 18th century when Fontana described “un corps oviforme” in epidermal cells of an eel he knew little about the meaning of this intracellular structure1 (see Fig. 1). According to a cytological review of Montgomery2 the term nucleolus was introduced by Valentin in 1836.3 The world’s first psychoanalyst—Sigmund Freud—in his first neuroscientific studies worked on surviving nervous tissue and nerve cells of crawfish, and described plexiform bodies in the cell nucleus. These nucleoli of nerve cells were most helpful to him in so far as the movements and rotations of the nuclei and the “plexiform bodies” in the surviving tissue provided indication that the tissue is still alive.4 About 40 years later Heitz5 and Barbara McClintock6 described the chromosomal origin of the nucleolus. It forms after cell division at particular sites of some chromosomes—the secondary constriction. The part of the chromosomes were termed Nucleolus Organizer Region (NOR). Although Sigmund Freud knew little about the meaning of the “plexiform” nucleolus he quite rightly interpretated the changes of the internal structure and the irregular shapes of the nucleolus as signs of life in the fresh nervous tissue. He was not aware of the function of this internal structure. The 20th century saw the invention and development of the electron microscope, which can resolve and help to distinguish the internal structure and the different components of the nucleolus. The nucleolus is not an organelle but rather a domain within the nucleus. There is no membrane around it but there is still a visible border that allows distinction from the surrounding chromatin irrespective of the degree of chromatin condensation. The nucleolar body may contain inclusions of condensed chromatin; however, this may or may not be considered a typical nucleolar structure. In plant nucleoli there are chromatin inclusions with various degrees of condensation.7-11 Some are associated with small amounts of RNA polymerase I, indicating that these genes remain in a standby position or may have been involved in the transcription of ribosomal genes shortly before. In the nucleolus of vertebrate cells, chromatin inclusions can be observed; however, they carry no signs related to transcriptional activity.

The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Nucleolar Ultrastructure in Vertebrates

11

Figure 1. Presumably the first description of a nucleolus as an oval body (un corps oviforme) by Fontana.

Nucleolar Internal Morphology As shown in Figure 2 at the ultrastructural level typically three nucleolar components can be distinguished: • fibrillar center • dense fibrillar component • granular component

The morphology of fibrillar centers (FC) is made up by a meshwork of fine 4-5 nm thick fibrils giving a relatively low electron microscopic contrast in routine electron microscopic stainings. Typically their shape is roughly globular, with diameters ranging from about 50 nm to 1 µm. Number and size of FCs per nucleolus is variable, it changes with the cellular activity and the need for ribosome production. Cells with little activity like dormant lymphocytes, usually reveal only one small nucleolus with one FC in a central position Active cells may have several large nucleoli which contain several—sometimes up to one hundred—smaller FCs. FCs may be completely missing in particular cases, (see ref. 12 for review). The dense fibrillar component (DF) reveals also very fine (3-5 nm thick) but very densely packed fibrils, giving a high electron microscopic contrast. When FCs are present the DF usually surrounds them and forms a network that sometimes may reach far away from a FC. This is particularly true for activated states. The amount of DF roughly reflects the nucleolar engagement in ribosome biogenesis. In some cases this network occupies large areas of the nucleolus, occasionally interspersed with small FCs (so called “reticulate” or nucleoli with “nucleolonema”). Occasionally the term fibrillar complex is used to describe FCs and surrounding DF. The granular component (GC) appears to consist of small granules with a diameter of about 15 nm. Typically these granules form a mass around the fibrillar complexes in this way embedding the FCs and DF. Quite frequently a transition zone between DF and GC can be observed. The border to the surrounding chromatin and nucleoplasm is usually quite distinct.

Additional Aspects on Morphology Other components are interstices and chromatin inclusions. Interstices are regions with very low contrast, they vary in size and shape. In electron micrographs of thin sections they are irregularly distributed within all of the three main components described above. They have also been referred to as “vacuoles”. Most likely they represent invaginations of (non nucleolar) nucleoplasm of low density into the nucleolus. Chromatin surrounding the nucleolus is in some cell types stronger condensed (peri-nucleolar chromatin) and contains highly methylated DNA. Occasionally strands of condensed chromatin are running through the nucleolus ("intranucleolar chromatin").

12

The Nucleolus

Figure 2. Electron micrograph of a human lymphocyte nucleolus, the central round fibrillar center (FC) is surrounded by a shell of dense fibrillar component (DF). Both the FC and the DF are embedded in a mass of granular component (GC). This nucleolus is almost completely surrounded by condensed chromatin with some condensed chromatin inclusions close to the FC.

Sometimes the differentiation of nucleolar components can be difficult, it may require an “experienced eye”. The main criteria are the form of the ultra structural components and the internal structure (fibrillar or granular appearance), their density, and their relative contrast. This implies that in different cell types comparisons are sometimes difficult. In micrographs FCs are typically lightly stained compared to the DF, in rare cases not very different from the latter.

Nucleolar Appearance Reflects Activity At the present state most of the observations on NORs and nucleoli and the function of ribosomal DNA transcription and transcript processing allow us to achieve some general understanding of the ribosome biogenesis in the nucleolar components. However, the detailed role of many proteins involved is still enigmatic and will require additional intensive research. The individual nucleolar components have a different molecular composition corresponding to their function related to the process of ribosomal gene transcription, transcript processing, pre-ribosome assembly and ribosome export to the cytoplasm. However, despite the fact that the nucleolus is considered a paradigm for the study of the relationship of structure and function, some of these relationships are still a matter of debate. A standard approach to study structure-function relationships is the observation of structural changes in different states of activity. The investigation of dormant lymphocytes or monocytes which can easily be stimulated by incubation with phytohemagglutinin or interferon, respectively, provides a simple and easily reproducible cellular model.13,14

Low Activity and Ring Shaped Nucleoli Dormant lymphocytes as shown in Figure 2 have a so called ring shaped nucleolus, with a single round relatively large FC surrounded by a shell of DF. In electron microscopic sections this arrangement will reveal a “ring” around the fibrillar center. Chromatin inclusions with condensed DNA may occur in any of the three components. Quite frequently the nucleoli are

Nucleolar Ultrastructure in Vertebrates

13

Figure 3. Nucleolus of a lymphocyte from a human lymph node. As a first sign of stimulation the granular component is almost completely lost, while the DF starts to appear as irregularly shaped strands. The “rings” of DF become open and form strands of dense fibrillar component, the so called nucleolonema. Due to the loss of granular component the fibrillar center (FC) has direct contact with the nucleoplasm in the lower right side. Chr – chromatin.

surrounded by condensed chromatin, which may or may not be constituted from the nucleolar organizer region and contain the tandem repeats of ribosomal genes from one of the appropriate chromosomes.15

Cellular Activation Cell activation, and the demand for increased protein synthesis leads to a fast response, i.e., nucleolar metamorphosis and component rearrangement. Changes in metabolic state will stimulate the nucleolus, it then will change shape and internal organization. A first sign of activation is the loss of the granular component, because the reservoir of ribosomes will be recruited towards the cytoplasm (Fig. 3). While the granular component is lost and the dense fibrillar component starts to rearrange and increase the overall mass of fibrillar centers will roughly remain the same. Larger FCs will be divided by intrusions of DF,16 in this way the fibrillar centers become smaller; sometimes they are hardly seen. Typically the ring-shaped appearance is lost after about 12 hours. Then the nucleolus increases in size and the dominating structure is the strands of dense fibrillar component intermingled with granular component (Fig. 4). During this ultra structural reorganization additional chromosomes and NORs that have been “transcriptional silent” are recruited to participate in the formation of the nucleoli.17

High Activity Steady State If the stimulus persists the metamorphoses of nucleoli is paralleled by the transition from the small lymphocytes into large blast like cells. The nucleolus will grow larger and achieve a new steady state appearance at a higher production level. This steady state of enhanced activity is reflected in the appearance of a compact nucleolus. It consists of several smaller fibrillar centers, each of them is attached to or surrounded by a shell of dense fibrillar component; all of these fibrillar complexes are embedded in the granular component. Basically, the size of the nucleolus and/or the number of fibrillar complexes correlates with the cellular activity. Therefore, the different types of nucleoli are rarely found in one particular cell at the same time.13,18

14

The Nucleolus

Figure 4. Stimulated nucleolus of an activated lymphocyte in the process of component rearrangement. The nucleolus is growing as a whole, while the fibrillar centers become smaller. The DF forms irregular strands which intermingle with granular component. The border between nucleolus and nucleoplasm follows an irregular line.

During cell activation experiments in human lymphocytes the nucleolar changes can be observed within a few hours, during which several small micro nucleoli appear: it then takes another few hours until a nucleolonema type nucleolus is observed. After about 3 days the high level steady state is achieved in the cell population, which will reveal large compact nucleoli in most of the cells. About one or two weeks after the single initial stimulation, lymphocytes and monocytes will down regulate their activities and will return to an inactive state, which again is reflected by nucleolar changes. They become small and ring shaped again. The metamorphosis of the nucleoli is a morphological reflection of cell activation. All the described changes parallel the transformation of the whole cell, in particular the cell nucleus. Highly active, large nucleoli may correlate with cell division activity; however, they occur in non dividing cells and therefore, they do not necessarily reflect a high proliferation rate but rather a high protein demand of the cell for various reasons. Cells with a high division rate reveal a stimulated nucleolus because cell growth also demands an increased protein synthesis and hence more ribosomes from the nucleolus.

Cytochemical Characterization The ribosomal biogenesis within the nucleolus requires a specific molecular ambience for many different steps. Starting at the assembly of the transcription complexes, DNA transcription initiation is followed by elongation, transcript processing, including the proper folding on the transcripts and protein addition which already occurs while transcription is continued. The necessary proteins and enzymes needed to achieve this rather complex biochemical task to finally obtain a functional ribosome are described later in this book. In this chapter on structure I want to focus on cytochemical and immunohistochemical stainings as they contributed to the understanding of nucleolar structures and their possible role in ribosome biogenesis.

Silver Staining of Nucleolar Components

Since Schwarzacher and coworkers19 described the nature of silver precipitating proteins several methods of silver staining these NOR associated proteins have been developed and refined.20-22 Meanwhile some silver staining proteins have been identified.23,24 Although the individual nature of most of the argyrophilic proteins remained unclear for long time the staining method was used extensively to study nucleolar structures. The silver staining marks those

Nucleolar Ultrastructure in Vertebrates

15

Figure 5. The nucleolus of an exponentially growing Hela cell is large; it occupies a significant part of the nucleus. The high activity level of the cell can also be estimated by the fact that very little condensed chromatin is seen in the cell nucleus.

NORs on mitotic chromosomes which were active during the precedent interphase (Fig. 6). These proteins are attached to the ribosomal DNA of the active NORs. During interphase all three nucleolar components can be reactive with silver staining, though with a different sensitivity (Fig. 7).

DNA-Staining Because the silver staining reaction on nucleolar sections or on mitotic chromosomes correlates with the activity of ribosome biogenesis, and because the argyrophilic proteins are attached to the ribosomal DNA, to understand the functional organization it is important to know the distribution of decondensed (transcriptional active) DNA within the nucleolar components. The visualization of DNA can be achieved by several techniques. Feulgen like DNA-staining methods have been adapted for ultra structural requirements.25-27 Staining with DNA binding molecules like DNases or specific antibodies at the ultra structural level was followed by more refined or more sensitive techniques to locate not only DNA but specific ribosomal gene sequences by in situ hybridization on ultra structural sections. A relatively new method of DNA staining and distribution analysis in ultra structural sections utilizes the enzyme terminal transferase. The terminal transferase is a polymerase that can elongate blunt ends of DNA as they occur on the surface of resin embedded sections. The enzyme recognizes the DNA-“tails” and adds labeled nucleotides, which then can be visualized by immunogold techniques (Fig. 8).

Structure Relates to Function? There is a considerable body of evidence showing that the dense fibrillar component is not a homogeneous component in terms of molecular composition. Cytochemical silver staining revealed at least two different kinds of dense fibrillar component with differences in the molecular composition (Fig. 7). Feulgen-like DNA-staining has shown clusters of decondensed DNA.27 These clusters as well as transcription foci within the DF28 do not match the silver staining pattern, which indicates that during interphase the argyrophilic proteins are not all bound to DNA. In terms of molecular composition of the DF this implies a further sub-compartmentalization. The DF apparently has the molecular tools and compartments for several steps of ribosome biogenesis. The major part of dense fibrillar component appears to contain no DNA (Fig. 8). In previously published work we could

16

The Nucleolus Figure 6. Mitotic human chromosome spread according to Schwarzacher and Wolf,48 silver stained, and imaged by electron microscopy. The arrows point out argyrophilic proteins attached to the nucleolus organizer region. Their presence indicates that the ribosomal genes from this locus have been actively transcribed during the precedent interphase.

demonstrate focal accumulations of DNA associated proteins like polymerase-I upstream binding factor (UBF) and poly-ADP-ribose polymerase (PARP).29,30 Typically these accumulations appear in close spatial relationships to transcription sites as demonstrated by dual labeling experiments. The evidence supports the view that transcription can but does not have to be associated with the surface of fibrillar centers. Presumably the surface of the FC is a preferred site for transcription initiation, while the transcript elongation which occurs simultaneously with first processing steps gives rise to the formation of some dense fibrillar component. The granular component seems to be the place for ribosome maturation (RNA splicing and addition of ribonucleoproteins) and a storage place for pre-ribosomes. This is particular well demonstrated in the cell model with stimulated lymphocytes. One of the first reactions of the cellular activation is a rapid depletion of the granular component (compare Fig. 2 and Fig. 3). The granular appearance of the structures was responsible for the name; they are ribosomes or pre-ribosomes, most likely in the state of final processing and/or storage before export to the cytoplasm. It was shown that the fibrillar centers are actually located at the crossing points of intermediate filaments which gave rise to the hypothesis that they may also function as a structural core element.31 Following restriction enzyme treatment to cut DNA and electro elution of the cells, skeletal elements of the nucleus and nucleolar components remain as a network of intermediate filaments. The network is concentrated in the nucleolar region where fibrillar centers are attached. When cells are hypotonically pretreated which lets the granular component and the DF disperse, it then depends on the final hypotonic strength whether or not nucleolar components disintegrate leaving hypotony resistant structures. Very low ionic strength (about 30 mM) will completely disperse the granular component and major parts of the dense fibrillar component. Fibrillar centers belong to the hypotony resistant structures within the nucleolus.32 With in situ hybridization experiments we could show that the part of the ribosomal gene which contains the promoter and the upstream sequences of the gene remain attached to the surface of the fibrillar centers while the more downstream sequences of the ribosomal gene appears not attached to structural nucleolar components and therefore can relocate under these particular experimental conditions. These kinds of experiments led

Nucleolar Ultrastructure in Vertebrates

17

Figure 7. a) Shows the nucleolus of a human Sertoli cell after progressive silver staining. Sertoli cells have a typical component arrangement, which is easy to recognize under experimental conditions. The silver staining contrasts the fibrillar center (FC) and the dense fibrillar component (DF) quite intensively, while the granular component (GC) reveals less contrast. Already at low magnification it can be seen that the DF is not homogeneous in terms of stainability with silver. b) shows the FC and DF of another nucleolus in a human Sertoli cell at higher magnification and with less intense silver staining to allow for structure recognition “behind” the silver grains. The FC and the DF can be recognized. The FC is covered by evenly distributed silver grains. Those strands of DF close to the FC react with silver, other (more peripheral) parts do not stain. This staining pattern indicates a non homogeneous molecular composition of the DF.

to the hypothesis that ribosomal genes are indeed attached to nuclear skeletal elements, which the fibrillar center may be part of.

Functional Interpretation of Nucleolar Components Inactivation of rDNA transcription leads to a disintegration of the nucleolus. Depending on the circumstances there are different patterns of metamorphosis. It was mainly the work of Mirre and Stahl that described nucleolar inactivation and segregation during female oocyte maturation33-35 and male spermiogenesis.36,37 Experimental transcription inhibition, which can be achieved by low doses of actinomycin-D (AMD), causes a segregation of nucleolar components.38-41 Schöfer and coworkers38 have shown that even if the ribosomal gene transcription is completely inhibited by AMD treatment and the components of the nucleolus are segregated there are still some remains of nucleolar fibrillar centers and dense fibrillar components. Although the cell stimulation experiments have shown a strong correlation between the distribution pattern of FCs and the overall mass of DF, these results indicate that transcriptional activity is not necessarily reflected by the appearance of one particular component. Fibrillar centers may exist without transcription, and vice versa, transcription may occur without fibrillar centers. The presence of DF does not necessarily imply that there is transcription activity. In other words ribosomal gene transcription itself is not responsible for and is not organizing the formation of a component. In compact nucleoli of exponentially growing cells the actively transcribed ribosomal genes have been located to the peripheral zone of the FC,42 transcription more downstream (transcript elongation) appears likely to occur in the DF,29,30,43 which is also the site of simultaneous transcript processing (see also refs. 42 and 44 for review). During mitosis the ribosome biogenesis seizes. Therefore, the proteins attached to the NOR are a naturally occurring example of a FC-like protein assembly, which contains transcription associated proteins without transcriptional activity. In summary the observations support the hypothesis that fibrillar centers have also a “stand by” function and serve as protein storage for times with higher demand.44

18

The Nucleolus

Figure 8.Human sertoli cell after DNA staining with terminal transferase and immunogold labeling. Most of the labeling is seen over the condensed Chromatin (cCHR). The major part of the granular component carries no label, the same is true for the fibrillar center, only few grains are seen at the periphery. Within the dense fibrillar component (DF) small grain clusters suggest a local accumulation of DNA (arrows). In reference to in situ hybridization experiments most of this intranucleolar DNA contains the ribosomal genes.15,27,49

Nucleolus in Human Sciences In Tumor Diagnosis The experimental evidence suggests that the direct correlation between activity on the one hand and small FCs and increase of DF on the other hand is not entirely valid (e.g., experimental or natural transcription stop). However, in general terms a morphometric evaluation of nucleolar structures can provide valuable clues on cellular activity. As tumor cells tend to grow more aggressively compared to normal cells, the nucleoli of tumor cells tend to be larger to meet the increased demand for protein production. Consequently, morphometric analysis after silver staining provides clues about the growth rate of a tumor cell population. The technique works well under experimental culture conditions with homogeneous (monoclonal) cell lines. In pathologic tissue sections the differences can be seen, however in the hands of human pathologists the method failed to prove superiority to conventional evaluation of tumors.

In Experimental Neonatology Following experimental perinatal asphyxia (low oxygen supply to the pups while being born) in laboratory animals nucleolar transcription is down regulated.45 This is reflected in altered nucleolar structures recognizable at the light and electron microscopic resolution. At the ultra structural level the reservoir of the GC is depleted of ribosomes. The changes became more dramatic with longer duration of perinatal asphyxia. Severe acidosis may be responsible for decreased polymerase-I activity and for disintegration of nucleoli in neurons, which may persist for at least some days after the insult.46,47

Nucleolar Ultrastructure in Vertebrates

19

References 1. Fontana F. Traité sur le vénin de la vipere, avec des observations sur la structure primitive du corps animale. 1781. Florence. 2. Montgomery TH. Comparative cytological studies, with special regard to the morphology of the nucleolus. J Morph 1898; 15:265-282. 3. Valentin G. Repertorium für Anatomie und Physiologie 1. Verlag von Veit und Comp, 1836:1-293. 4. Freud S. Über den Bau der Nervenfasern und Nervenzellen beim Flußkrebs. 15-12-1881. Austrian Academy of Sciences. 5. Heitz E. Die Ursache gesetzmäßiger Zahl, Lage, Form, und Größe pflanzlicher Nucleolen. Planta 1931; 12:775-844. 6. McClintock B. The relation of a particular chromosomal element to the development of the nucleoli in Zea Mays. Z Zellf 1934; 21:294-328. 7. Moreno Diaz de la Espina S, Medina FJ, Risueño MC. Correlation of nucleolar activity and nucleolar vacuolation in plant cells. Eur J Cell Biol 1980; 22:724-729. 8. Risueño MC, Medina FJ, Moreno Diaz de la Espina S. Nucleolar fibrillar centres in plant meristematic cells: ultrastructure, cytochemistry and autoradiography. J Cell Sci 1982; 58:313-329. 9. Risueño MC, Medina FJ. The nucleolar structure in plant cells. Revis Biol Celular 1986; 7:1-154. 10. Sato S, Willson C, Dickinson HG. Origin of nucleolus-like bodies found in the nucleoplasm and cytoplasm of Vicia faba meristematic cells. Biol Cell 1988; 64:321-329. 11. Medina FJ, De la Espina SMD. On intranucleolar chromatin and fibrillar centers in higher plants. J Histochem Cytochem 1992; 40:1235-1236. 12. Thiry M, Goessens G. The nucleolus during the cell cycle. 1 ed. Heidelberg: Springer, 1996. 13. Wachtler F, Ellinger A, Schwarzacher HG. Nucleolar changes in human phytohaemagglutininstimulated lymphocytes. Cell Tissue Res 1980; 213:351-360. 14. Schedle A, Willheim M, Zeitelberger A et al. Nucleolar morphology and rDNA in situ hybridization in monocytes. Cell Tissue Res 1992; 269:473-480. 15. Wachtler F, Mosgoeller W, Schwarzacher HG. Electron microscopic in situ hybridization and autoradiography: Localization and transcription of rDNA in human lymphocyte nucleoli. Exp Cell Res 1990; 187:346-348. 16. Hozák P, Novak JT, Smetana K. Three-dimensional reconstruction of nucleolus-organizing regions in PHA-stimulated human lymphocytes. Biol Cell 1989; 66:225-233. 17. Wachtler F, Roubicek C, Schedle A et al. Nucleolus organizer regions in human lymphocytes as studied with premature chromosome condensation. Hum Genet 1990; 84:244-248. 18. Wachtler F, Schwarzacher HG, Smetana K. On the fusion of nucleoli in interphase. Eur J Cell Biol 1984; 34:190-192. 19. Schwarzacher HG, Mikelsaar AV, Schnedl W. The nature of the Ag-staining of nucleolus organizing regions. Electron- and light microscopic studies on human cells in interphase, mitosis and meiosis. Cytogenet Cell Genet 1978; 20:24-39. 20. Goodpasture C, Bloom SE. Visualization of nucleolar organizer regions in mammalian chromosomes using silver staining. Chromosoma 1975; 53:37-50. 21. Howell WM, Black DA. Controlled silver-staining of nucleolus organizer regions with protective colloidal developer: a 1 step method. Experientia 1980; 36:1014-1015. 22. Derenzini M, Trerè D. Importance of interphase nucleolar organizer regions in tumor pathology. Virchows Arch [B] 1991; 61:1-8. 23. Roussel P, Sirri V, Hernandez-Verdun D. Quantification of Ag-NOR proteins using Ag-NOR staining on western blots. J Histochem Cytochem 1994; 42:1513-1517. 24. Roussel P, Belenguer P, Amalric F et al. Nucleolin is an Ag-NOR protein—This property is determined by its amino-terminal domain independently of its phosphorylation state. Exp Cell Res 1992; 203:259-269. 25. Derenzini M, Hernandez-Verdun D, Bouteille M. Relative distribution of DNA and NOR-protein in nucleoli visualized by simultaneous Feulgen-like and Ag-NOR-staining procedures. Biol Cell 1981; 40:147-150. 26. Derenzini M, Hernandez-Verdun D, Bouteille M. Visualization in situ of extended DNA filaments in nucleolar chromatin of rat hepatocytes. Exp Cell Res 1982; 141:463-469. 27. Mosgoeller W, Schöfer C, Derenzini M et al. Distribution of DNA in human Sertoli cell nucleoli. J Histochem Cytochem 1993; 41:1487-1493. 28. Mosgoeller W, Schöfer C, Wesierska-Gadek J et al. Ribosomal gene transcription is organized in foci within nucleolar components. Histochem Cell Biol 1998; 109:111-118.

20

The Nucleolus

29. Mosgoeller W, Schöfer C, Wesierska-Gadek J et al. Ribosomal gene transcription is organized in foci within nucleolar components. Histochem Cell Biol 1998; 109:111-118. 30. Mosgoeller W, Steiner M, Hozák P et al. Nuclear architecture and ultrastructural distribution of poly(ADP-ribosyl)transferase, a multifunctional enzyme. J Cell Sci 1996; 109:409-418. 31. Hozák P, Cook PR, Schöfer C et al. Site of transcription of ribosomal RNA and intranucleolar structure in HeLa cells. J Cell Sci 1994; 107:639-648. 32. Mosgoeller W, Schofer C, Steiner M et al. Arrangement of ribosomal genes in nucleolar domains revealed by detection of “Christmas tree” components. Histochem Cell Biol 2001; 116:495-505. 33. Mirre C, Stahl A. Peripheral RNA synthesis of fibrillar center in nucleoli of japanese Quail oocytes and somatic cells. J Ultrastruct Res 1978; 64:377-387. 34. Mirre C, Stahl A. Ultrastucture and activity of the nucleolar organizer in the mouse oocyte during meiotic prophase. J Cell Sci 1978; 31:79-100. 35. Mirre C, Hartung M, Stahl A. Association of ribosomal genes in the fibrillar centers of the nucleolus: a factor influencing translocation and non-disjunction in the human meiotic oocyte. Proc Natl Acad Sci USA 1980; 77:6017-6021. 36. Arroua ML, Hartung M, Devictor M et al. Localisation of ribosomal genes by in-situ hybridization in the fibrillar centre of the nucleolus in the human spermatocyte. Biol Cell 1982; 44:337-340. 37. Stahl A, Wachtler F, Hartung M et al. Nucleoli, nucleolar chromosomes and ribosomal genes in the human spermatocyte. Chromosoma 1991; 101:231-244. 38. Schöfer C, Weipoltshammer K, Almeder M et al. Redistribution of ribosomal DNA after blocking of transcription induced by actinomycin D. Chromosome Res 1996; 4:384-391. 39. Crozet N. Effects of actinomycin D and cycloheximide on the nucleolar ultrastructure of porcine oocytes. Biol Cell 1983; 48:25-30. 40. Jordan EG, McGovern J. The quantitative relationship of the fibrillar centres and other nucleolar components and other nucleolar componentsto changes in growth conditions , serum deprivation and low doses of actinomycin D in cultures diploid human fibroblasts (strain MRC-5). J Cell Sci 1981; 523:373-389. 41. Connan G, Haguenau F, Rabotti GF. Transcription of pre-ribosomal RNA in RSV-transformed CEF presenting nucleolar lesions induced by actinomycin D and alpha- amanitin. Biol Cell 1980; 39:175-178. 42. Dundr M, Misteli T. Functional architecture in the cell nucleus. Biochem J 2001; 356:297-310. 43. Stanek D, Koberna K, Pliss A et al. Non-isotopic mapping of ribosomal RNA synthesis and processing in the nucleolus. Chromosoma 2001; 110:460-470. 44. Schwarzacher HG, Mosgoeller W. Ribosome biogenesis in man: Current views on nucleolar structures and function. Cytogenet Cell Genet 2000; 91:243-22. 45. Hoeger H, Labudova O, Mosgoeller W et al. Deficient transcription of subunit RPA 40 of RNA polymerase I and III in heart of rats with neonatal asphyxia. Life Sci 1998; 62:275-282. 46. Kastner P, Mosgoeller W, Fang-Kircher S et al. Deficient brain RNA polymerase and altered nucleolar structure persists until day 8 after perinatal asphyxia of the rat. Pediatr Res 2003; 53:62-71. 47. Mosgoeller W, Kastner P, Fang-Kircher S et al. Brain RNA polymerase and nucleolar structure in perinatal asphyxia of the rat. Exp Neurol 2000; 161:174-182. 48. Schwarzacher HG, Wolf U. Methods in human cytogenetics. 1 ed. Berlin,Heidelberg,NewYork: Springer, 1974: 49. Jiménez-García LF, Segura-Valdez MD, Ochs RL et al. Electron microscopic localization of ribosomal DNA in rat liver nucleoli by nonisotopic in situ hybridization. Exp Cell Res 1993; 207:220-225.

CHAPTER 3

The Nucleolar Ultrastructure in Yeast Isabelle Léger-Silvestre and Nicole Gas

Summary

T

he nucleolus is a highly dynamic compartment of the nucleus whose size, number and structure vary according to cell type and metabolic state. Despite this versatility, its morphological compartment is remarkably conserved throughout evolution. However, the assignment of precise functions to the identified morphological domains of the nucleolus is still debated. We present in this chapter the advantages that yeast offers as an experimental system to study the molecular determinants of the nucleolar structure. We develop how morphological analyses of cryofixed wild type and mutant cells combined with in situ approaches on conventional fixed yeast are used to investigate the nucleolar ultrastructure and functions. Dynamic of the assembly of the yeast nucleolus is discussed with respect to the role of the chromosomal context of rDNA and of the polymerase I. The yeast nucleolus is also presented as an integrated part of the nucleus and is analysed in relation with the nucleoplasm.

Introduction Due to its high density and refractive index, the nucleolus was one of the first subcellular compartments of the nucleus discovered by 18th-19th century microscopists. With the development of phase microscopy, there emerged a period when it became possible to consider the nucleolus as a specific organelle in living cells rather than a possible artefact that resulted from the fixation and staining procedure employed. With the advent of electron microscopy, major advances were possible in the description of the nucleolar ultrastructure. Combining microscopic approaches with genetics, biochemistry and drug treatments, it is now clear that the eukaryotic nucleolus is the structural framework around the chromosomal loci that contain the rRNA genes; it is the site where transcription of the preribosomal RNA (prerRNA) by RNA polymerase I (pol I) and its subsequent processing and assembly with ribosomal proteins take place to form 60S and 40S preribosomal particles. The nucleolus is a highly dynamic compartment of the nucleus whose size, number, and structure vary according to cell type and metabolic state. Despite this versatility, the fine structure of nucleoli of vertebrates can be described in terms of the distribution of three basic components: the fibrillar center (FC), the dense fibrillar component (DFC), and the granular component (GC).1 However, the exact significance of the nucleolar components in functional and molecular terms remains unclear and the precise function of these subcompartments is still controversial and subject to debate.2 Further studies are therefore required to clarify the relationships between the structure of the nucleolus and the ribosome biogenesis process and/or other nucleolar functions. For several reasons, yeasts are attractive models for further studies on the relationships between nucleolar structure and functions: 1) There is only one nucleolus per haploid cell.3 2) The nucleus of the yeast cell is actively engaged in the synthesis of ribosomal RNA.4 3) Yeasts are very well characterized models for genetic, biochemical and physiological manipulations. 4) Finally, all available data indicate that the major steps in ribosome synthesis are conserved throughout eukaryotes and The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

22

The Nucleolus

human homologues of most of the components involved in the ribosome biogenesis have been identified in yeast.5,6 In this chapter, we will present the main results of investigations in which electron microscopy was used not only to describe morphological structures of yeast nucleus but also to relate these structures to nuclear functions.

Cryomethods Provide Insights into the Yeast Nucleolar Organization Although yeast has a number of attributes that make it excellent for studying relationships between nucleolar structure and function, it also has a number of intrinsic disadvantages. The small size and tough cell wall of yeast considerably limit precise structural analysis of its nucleus and nucleolus. By electron microscopy (EM), two distinguishable regions in the yeast nucleus have been described: a region of low electron density composed of chromatin, which corresponds to the part of the nucleus which is stained by DAPI in growing cells7 and a crescent shaped electron-dense domain that occupies one-third to one-half of the nucleus. The latter structure has been referred to as the nucleolus on the basis of ultrastructural studies, immunocytochemical studies and in situ hybridization.7-10 In most of the conventional transmission electron micrographs of yeast, the nucleolus appears to lack the three basic components of vertebrate nucleoli. Regions of low electron density surrounded by a region of greater density have been observed in the nucleoli of chemically fixed Schizosaccharomyces pombe and Saccharomyces cerevisiae spheroplasts.9,11 Sillevis Smitt et al described in isolated nuclei of Saccharomyces carlsbergensis a reticular structure similar to the “nucleolonema” of some higher eukaryotic cells.12,13 Additional morphological details in the yeast nucleolus were observed when cells were fixed by the method of freeze-substitution,14-16 suggesting that precise structural analysis of the yeast nucleoli have been considerably limited, in part because of the difficulty of preserving the cell morphology by conventional treatments for electron microscopy. Cryofixation preserves cells structures because it avoids the freezing damage caused by ice crystals: the water of the specimen fixed by freezing is cooled so rapidly that it is vitrified (the water is immobilized before it has time to crystallize).17,18 This freezing technique also avoids chemical pretreatment that may result in morphology disruption because of its slow action and its selectivity for cellular components. One of the drawbacks of cryofixation is the limited depth (10-20µm) of a specimen that can vitrify during the cooling. However, taking advantage of the small size of the yeast cells, vitrification occurs up to a depth large enough to enclose a significant number of cells with morphology preservation of high quality. Combined with cryosubstitution (substitution of the ice of the frozen specimen by a chemical fixative at low temperature), cryofixation increases the potential for trapping native structures and distribution patterns.

The Yeast Nucleolus Resembles the Higher Eukaryotic Nucleolus General Morphology and Location of the Yeast Nucleolus Morphological analyses of cryofixed wild type and mutant cells combined with in situ hybridization and immunocytochemistry on conventional fixed yeast have been used to reinvestigate the yeast nucleolar ultrastructure.19-21 Nuclei of both cryofixed fission and budding yeasts show the characteristic bipartite appearance with a single and well-developed nucleolus corroborating previous ultrastructural analyses. Systematic morphological association between the nucleolus and the nuclear envelope has been reported in different yeast strains. In S. cerevisiae, the nucleolus is in close contact to the nuclear envelope.20 In S. pombe, the nucleolus displays more limited contact with the nuclear envelope, as described in higher eukaryotes.14,15,19

Ultrastructural Organization of the Yeast Nucleolus The morphological preservation of cryofixed samples allows us to gain greater insight into yeast subnucleolar organization. In both S. cerevisiae and S. pombe, the distribution of nucleolar components is reminiscent of the compartmentalization seen in the nucleolus of higher

The Nucleolar Ultrastructure in Yeast

23

Figure 1. Panel 1: Morphology of Saccharomyces pombe yeast after cryofixation and freeze-substitution. In the nucleus, one region of low electron density and a large electron-dense area referred to as the nucleolus are visible. In the nucleolus, three distinct morphological compartments are identified : zones of lower electron density (FC) are surrounded by a dense fibrillar component (DFC) that extends as a network throughout the nucleolar volume. A granular component is dispersed throughout the rest of the nucleolus. Reprinted, with permission, from Léger-Silvestre in Chromosoma (1997) 105:542-552. Panel 2: On this S. pombe section-plane, the continuity of the nucleolar pale fibrillar material (FCs) and the nucleoplasm is evident and the DNA-rich region protrudes into the nucleolus of S. pombe (arrow). Panel 3: The nucleoplasm in S. cerevisiae nucleus is heterogeneously labelled by osmium ammine. Its periphery is negative with only limited zones of labelling indicating DNA contacts with the nuclear envelope (arrows). Subnucleolar domains are also clearly labelled showing the presence of DNA in nucleolar foci corresponding to fibrillar centers. Panel 4: By immuno-gold labeling, RNA polymerase II is detected in the nucleoplasm of S. cerevisiae nucleus but is absent from its periphery. The nucleolus is not labeled.

24

The Nucleolus

eukaryotes: regions of lower electron density resemble fibrillar centres surrounded by a dense fibrillar component that extends into a granular component (Fig. 1).19,20

Fibrillar Centers The appearance of the FCs is similar to the DNA- rich hemisphere of the nucleus (Fig. 1). In some section-planes, the continuity of the nucleolar pale fibrillar material (FCs) and the nucleoplasm is evident and the DNA-rich region protrudes into the nucleolus (Fig. 2, arrow). When yeast is treated with osmium ammine (a selective staining method to reveal both condensed and decondensed chromatin in situ22), subnucleolar domains are clearly labelled attesting the presence of DNA in nucleolar foci (Fig. 3). DAPI staining in the S. pombe nucleus also established the presence of DNA in limited zones of the nucleolus: one- half of the nucleus, corresponding to the DNA- rich region was intensely labelled and one or two fluorescent points were always visible in the nucleolar region.23 Taken together, these results suggest that FCs correspond to sections of chromatin protrusions inside the nucleolus. Detection by fluorescence in situ hybridization of rDNA on regions that protrude from the hemispherical chromosomal domain into the nucleolus of S. pombe10 and rDNA detection in the FCs at the EM level in both S.pombe19 and S.cerevisiae20 support the hypothesis that FCs correspond, at least in part, to chromosomal loci termed nucleolar organizer regions (NORs). The latter structures contain several hundred copies of rDNA genes arranged in tandem arrays.24,25

The Dense Fibrillar and Granular Components The fibrillar appearance of the DFC in close contact with granules suggests that it corresponds to nascent rRNAs that progressively mature and then condense into granular preribosomal particles. The codistribution in the DFC of polymerase I, rRNA transcripts and proteins involved in rRNA processing (Gar1p, Nop1p, Ssb1p),19-21 as well as the dynamic recruitment of the rRNA-processing machinery at the DFC during active transcription21 supports this hypothesis. Moreover, the observation of a DFC in mutant strains (see below) showed that this structure is independent of the polymerase engaged in rDNA transcription (pol I or pol II) and from the rDNA chromosomal environment suggesting that DFC is the result of active transcription and maturation, regardless of the structural context. Altogether, these results also suggest there is a spatial association of transcription sites and maturation sites in the yeast DFC. However, the transcription sites are not yet clearly identified in the yeast nucleolus . Indeed, rDNA detected by in situ hybridization at the EM level is restricted to the FCs sites and does not colocalize with pol I and nascent transcripts which are detected in the DFC throughout the nucleolus.19-21 Therefore, one may consider the DFC of the yeast nucleolus as a heterogeneous nucleolar subdomain with respect to molecular distribution: one part of the DFC surrounds the FCs where rDNA, pol I and newly-synthesized rRNA are detected and another part distributes throughout the nucleus as a fibrillar network where pol I and nascent transcripts are also detected, but rDNA is not. Whether rDNA is present in the nuclear part of DFC, but is not accessible has yet to be clarified. Further investigations are required to precisely define where the rDNA transcription actually takes place in the yeast nucleolus. The similar organization of the nucleolus in yeast and in higher eukaryotes shows that the location of nucleolar compartments is not related to the organization of the 5S rDNA since the 5S rRNA genes in S. cerevisiae are contained in the repeated rDNA segment, which is not the case in S. pombe or in vertebrates. Localization of 5S rRNA genes in situ has never been reported in yeast, but HA-tagged RNA pol III was detected in the DFC surrounding the FCs suggesting the 5S RNA genes are spatially distributed in S. cerevisiae at the same site the other rDNA genes.26

The Nucleolar Ultrastructure in Yeast

25

Dynamics of Assembly of the Yeast Nucleolus The permanence of the nucleolar subcompartments raises questions about the significance of the nucleolar organization and highlights the relevance of yeast as a model to study the relationships between nucleolar organization and functions. Yeast has proved to be an attractive experimental model to study the molecular determinants of the nucleolar structure. Is a precise spatial organization required for ribosome biogenesis to occur properly? Are the nucleolar subcompartments the direct consequence of the molecular mechanisms of ribosome biogenesis? Do some eukaryotic nucleolar features, such as transcription by pol I and the chromosomal context of rDNA, play a role in the organization of the nucleolus?

The Role of the Tandem Organization of Ribosomal Genes S. cerevisiae strains deleted of genomic rDNA allowed the study of the function of the tandem organization of ribosomal genes.21,27,28 In such strains, the 150 to 200 chromosomal rDNA repeats are largely deleted and yeast survival depends on rDNA transcription driven from plasmids. Functional synthesis of preribosomes occurs in these strains, which are characterized by the absence of a normal chromosomal context. Moreover, a nucleolar region was identified in these mutants indicating that rRNA genes clustering on chromosome XII is not necessary to form structures involved in ribosome biogenesis in S. cerevisiae. However, the nucleolar region is not crescent shaped and not confined to one side of the nucleus as in wild-type cells, suggesting that the tandem distribution of rDNA genes on one chromosome in yeast provides a spatial constraint for the formation of the nucleolus. The ultrastructural and functional identification of DFC and GC in this nucleolar structure suggests that the assembly of these nucleolar subcompartments is independent of rDNA chromosomal context in yeast. The absence of FCs indicates that ribosome biogenesis may occur while no FC is morphologically identified and supports the hypothesis discussed above that FCs correspond to the NORs.

The Role of RNA Polymerase I

The role of RNA pol I in nucleologenesis was also examined.21,29-31 In S. cerevisiae, the inhibition of ribosomal pol I transcription leads to a dynamic spatial rearrangement of the nucleolar region accompanied by dispersion of the rRNA-processing machinery. In a strain which bears a temperature-sensitive allele of RPA190 (rpa 190-2), an essential gene encoding the largest subunit A190 of pol I, the morphological consequences of rDNA transcription inhibition are similar to those observed in higher eukaryotes using actinomycin D, i.e., nucleolar segregation. The need for active pol I for nucleolus formation was also demonstrated in S. pombe thermosensitive pol I mutants.8 However, when rRNA genes are transcribed by pol II, ribosome biogenesis also occurs in identifiable nucleolar structures. The accurate recruitment of the rRNA processing machinery in these structures and the dispersion of pol I throughout the nucleus indicate that pol I accumulation is not required at site of nucleolar emergence, suggesting that pol I has no accessory role in the recruitment of the nucleolar proteins or in rRNA maturation. These results strongly suggest that the ribosomal transcripts rather than pol I are required to organize a nucleolar domain. However, pol I probably plays a role in the localization of the nucleolus in the nucleus. Indeed, when pol I transcribes rRNA genes, nucleolar structures are in contact with the nuclear envelope, regardless of the chromosomal context. In contrast, when pol II transcribes rRNA genes, or when transcription by pol I is inhibited, nucleolar structures do not establish any contact with the nuclear envelope. Taken together, these studies of mutants lead to the idea that the formation and spatial organization of the nucleolus in S. cerevisiae result from the recruitment of the transcription and maturation machineries around the NOR localized on chromosome XII. This organization clearly depends on the state of activity of the ribosomal genes.

26

The Nucleolus

The Nucleolus an Integrated Part of the Nucleus Yeast represents an attractive model for the study of ribosome synthesis. In the analysis of ribosome biogenesis, the nucleolus in yeast cannot be considered as an isolated sub-compartment of the nucleus, but it has to be analysed in relation to the nucleoplasm in which rRNA maturation and assembly of ribosomal proteins continue.6,32,33 In both S. cerevisiae and S. pombe, the nucleoplasm appears heterogeneous. Numerous small fibrils are visible, very small particles are aligned (means distributed on “a line”) and larger round particles are scattered in this chromatin-rich domain. Small fibrils might be equivalent to “perichromatin fibrils” observed in mammalian cells in close vicinity to the condensed chromatin; these fibrils could be nascent extranucleolar ribonucleoproteins (RNPs) referred to as the sites for splicing of premRNA.34 This hypothesis is not supported in yeast: indeed, using immuno-electron microscopy to detect the 2,2,7-trimethylguanosine cap structure of small nuclear RNAs and a protein component of U1 snRNP, Potashkin et al have suggested that functional domains involved in premRNA splicing are localized in the nucleolus rather than in the DNA-rich region.9 Moreover, in the absence of visible condensed chromatin, it is difficult to identify these fibrils as “perichromatin fibrils”. The small round particles scattered in the DNA-rich region morphologically resemble the preribosomal granules observed in the nucleolus but they may also be heterogeneous nuclear ribonucleoproteins (hnRNPs) or splicing complexes. Interchromatin granule clusters (IGC) of higher eukaryotic cells have never been described in yeast. Considering that these structures are dedicated to storage of maturation/splicing mRNA factors, their absence could be explained in S. cerevisiae by the fact that this yeast possesses many fewer intron-containing genes, resulting inwith a lower number of introns per gene than most higher eukaryotes. It is noteworthy that the outer zone of the nucleoplasm in S. cerevisiae is characterized by a relative morphological homogeneity when compared to the central zone.20 The absence of pol II at the nuclear periphery first suggested the presence of repressed chromatin in this region (Fig. 4). However, yeast treated with osmium ammine also shows a nonlabelled peripheral zone (Fig. 3) in contradistinction with the hypothesis that the yeast peripheral nuclear region could correspond to the heterochromatin of higher eukaryotes. Moreover, some specific rRNA intermediates accumulate at the nuclear periphery suggesting this part of the nucleoplasm could be dedicated to their maturation (N. Gas and S. Fakan unpublished data). These results show that the nucleoplasm is heterogeneous with respect to the spatial distribution of ribosome biogenesis steps. This heterogeneity is also illustrated by the difference in the distribution of the pre40S and pre60S particles in the nucleolus and the nucleoplasm suggesting that the two subunits exit the nucleus along distinct intranuclear pathways or with different kinetics.35 Indeed, kinetic analysis of rRNA export by labeling rRNA species and fractionating nucleus/cytoplasm showed that the small subunit appeared faster in the cytoplasm than the big subunit.36,37 The systematic localization of intermediates of maturation in the yeast nucleus should allow us to eventually build a spatio-temporal map of the post-transcriptional steps of ribosome production.

Conclusion Analyses of the yeast nucleolus in wild-type and mutant strains has helped in our understanding of its functional organization and assembly. However, these analyses clearly showed that the traditional concept of the relationship between structure and function is probably false for the nucleolus; e.g., a wild-type nucleolus does not seem to be required for ribosome biogenesis to occur properly. Furthermore, the DFC is heterogeneous with respect to the distribution of molecules and/or their accessibility. The analysis of relationships between structure and function is all the more difficult because recent developments suggest that the nucleolus is a plurifunctional nuclear domain.38-43 There is evidence that the nucleolus plays a role in the processing or export of a subset of mRNAs. Moreover, the RNA component of the SRP, a part of the translation machinery, appears to transit through the nucleolus. In addition, the RNA

The Nucleolar Ultrastructure in Yeast

27

subunit of telomerase is associated with the nucleolus of mammalian cells. A link between telomere-associated proteins and the yeast nucleolus was also described. U6 RNA, a component of spliceosome, also transits through the nucleolus. No nucleolar subdomain dedicated to these functions is characterized yet. However, very recently, Verheggen et al detected, under certain growth conditions, a “nucleolar body” that would represent an intermediate site in the localization of many small nucleolar RNAs involved in rRNA maturation.44,45 The advantages of the yeast model system clearly open further perspectives in understanding the multiple functions of nucleolus.

References 1. Jordan EG. Interpreting nucleolar structure: Where are the transcribing genes? J Cell Sci 1991; 98:437-42. 2. Raska I, Dundr M, Koberna K et al. Does the synthesis of ribosomal RNA take place within nucleolar fibrillar centers or dense fibrillar components? A critical appraisal. J Struct Biol 1995; 114:1-22. 3. Granot D, Snyder M. Segregation of the nucleolus during mitosis in budding and fission yeast. Cell Motil Cytoskeleton 1991; 20:47-54. 4. Woolford J, Warner J. The ribosome and its synthesis. In: JP Jr Broach, EW Jones, eds. The Molecular and Cellular Biology of the yeast Saccharomyces,. New-York: Cold Spring Harbor Laboratory Press, 1991; 587-626. 5. Tollervey D, Kiss T. Function and synthesis of small nucleolar RNAs. Curr Opin Cell Biol 1997; 9:337-42. 6. Kressler D, Linder P, de La Cruz J. Protein trans-acting factors involved in ribosome biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:7897-912. 7. Toda T, Yamamoto M, Yanagida M. Sequential alterations in the nuclear chromatin region during mitosis of the fission yeast Schizosaccharomyces pombe: video fluorescence microscopy of synchronously growing wild-type and cold-sensitive cdc mutants by using a DNA-binding fluorescent probe. J Cell Sci 1981; 52:271-87. 8. Hirano T, Konoha G, Toda T et al. Essential roles of the RNA polymerase I largest subunit and DNA topoisomerases in the formation of fission yeast nucleolus. J Cell Biol 1989; 108:243-53. 9. Potashkin JA, Derby RJ, Spector DL. Differential distribution of factors involved in premRNA processing in the yeast cell nucleus. Mol Cell Biol 1990; 10:3524-34. 10. Uzawa S, Yanagida M. Visualization of centromeric and nucleolar DNA in fission yeast by fluorescence in situ hybridization. J Cell Sci 1992; 101(Pt 2):267-75. 11. Hurt EC, McDowall A, Schimmang T. Nucleolar and nuclear envelope proteins of the yeast Saccharomyces cerevisiae. Eur J Cell Biol 1988; 46:554-63. 12. Sillevis Smitt WW, Vermeulen CA, Vlak JM et al. Electron microscopic autoradiographic study of RNA synthesis in yeast nucleus. Exp Cell Res 1972; 70:140-4. 13. Sillevis Smitt WW, Vlak JM, Molenaar I et al. Nucleolar function of the dense crescent in the yeast nucleus. A biochemical and ultrastructural study. Exp Cell Res 1973; 80:313-21. 14. Gallagher IM, Alfa CE, Hyams JS. p63cdc13, a B-type cyclin, is associated with both the nucleolar and chromatin domains of the fission yeast nucleus. Mol Biol Cell 1993; 4:1087-96. 15. Kanbe T, Kobayashi I, Tanaka K. Dynamics of cytoplasmic organelles in the cell cycle of the fission yeast Schizosaccharomyces pombe: three-dimensional reconstruction from serial sections. J Cell Sci 1989; 94(Pt 4):647-56. 16. Tanaka K, Kanbe T. Mitosis in the fission yeast Schizosaccharomyces pombe as revealed by freeze-substitution electron microscopy. J Cell Sci 1986; 80:253-68. 17. Dubochet J. High-pressure freezing for cryoelectron microscopy. Trends in Cell Biology 1995; 4:86-90. 18. Streinbrecht R, K Z. Cryotechniques in biological electron microscopy. Berlin, Heidelberg, New-York, London, Paris, Tokyo: Springer-Verlag: 1987. 19. Leger-Silvestre I, Noaillac-Depeyre J, Faubladier M et al. Structural and functional analysis of the nucleolus of the fission yeast Schizosaccharomyces pombe. Eur J Cell Biol 1997; 72:13-23. 20. Leger-Silvestre I, Trumtel S, Noaillac-Depeyre J et al. Functional compartmentalization of the nucleus in the budding yeast Saccharomyces cerevisiae. Chromosoma 1999; 108:103-13. 21. Trumtel S, Leger-Silvestre I, Gleizes PE et al. Assembly and functional organization of the nucleolus: ultrastructural analysis of Saccharomyces cerevisiae mutants. Mol Biol Cell 2000; 11:2175-89. 22. Biggiogera M, Courtens JL, Derenzini M et al. Osmium ammine: Review of current applications to visualize DNA in electron microscopy. Biol Cell 1996; 87:121-32.

28

The Nucleolus

23. McCully EK, Robinow CF. Mitosis in the fission yeast Schizosaccharomyces pombe: A comparative study with light and electron microscopy. J Cell Sci 1971; 9:475-507. 24. Goessens G. Nucleolar structure. Int Rev Cytol 1984; 87:107-58. 25. Shaw PJ, Jordan EG. The nucleolus. Annu Rev Cell Dev Biol 1995; 11:93-121. 26. Trumtel S. Etude de l’organisation fonctionnelle et de la dynamique du nucléole de la levure Saccharomyces cerevisiae. Université Paul Sabatier TOULOUSE III, Toulouse: 2001. 27. Chernoff YO, Vincent A, Liebman SW. Mutations in eukaryotic 18S ribosomal RNA affect translational fidelity and resistance to aminoglycoside antibiotics. EMBO J 1994; 13:906-13. 28. Nierras CR, Liebman SW, Warner JR. Does Saccharomyces need an organized nucleolus? Chromosoma 1997; 105:444-51. 29. Nogi Y, Yano R, Nomura M. Synthesis of large rRNAs by RNA polymerase II in mutants of Saccharomyces cerevisiae defective in RNA polymerase I. Proc Natl Acad Sci USA 1991; 88:3962-6. 30. Oakes M, Aris JP, Brockenbrough JS et al. Mutational analysis of the structure and localization of the nucleolus in the yeast Saccharomyces cerevisiae. J Cell Biol 1998; 143:23-34. 31. Oakes M, Nogi Y, Clark MW et al. Structural alterations of the nucleolus in mutants of Saccharomyces cerevisiae defective in RNA polymerase I. Mol Cell Biol 1993; 13:2441-55. 32. Udem SA, Warner JR. Ribosomal RNA synthesis in Saccharomyces cerevisiae. J Mol Biol 1972; 65:227-42. 33. Milkereit P, Gadal O, Podtelejnikov A et al. Maturation and intranuclear transport of preribosomes requires Noc proteins. Cell 2001; 105:499-509. 34. Fakan SaBW. Perichromatin fibrils are in situ forms of nascent transcripts. Trends in Cell Biology 1994; 4:86-90. 35. Gleizes PE, Noaillac-Depeyre J, Leger-Silvestre I et al. Ultrastructural localization of rRNA shows defective nuclear export of preribosomes in mutants of the Nup82p complex. J Cell Biol 2001; 155:923-36. 36. Udem SA, Warner JR. The cytoplasmic maturation of a ribosomal precursor ribonucleic acid in yeast. J Biol Chem 1973; 248:1412-6. 37. Trapman J, Planta RJ. Maturation of ribosomes in yeast. I Kinetic analysis by labelling of high molecular weight rRNA species. Biochim Biophys Acta 1976; 442:265-74. 38. Politz JC, Yarovoi S, Kilroy SM et al. Signal recognition particle components in the nucleolus. Proc Natl Acad Sci USA 2000; 97:55-60. 39. Pederson T. Proteomics of the nucleolus: more proteins, more functions? Trends Biochem Sci 2002; 27:111-2. 40. Pederson T. Viewing the ribosome and visiting the nucleolus at Lake Tahoe. RNA 2001; 7:1-4. 41. Pederson T, Politz JC. The nucleolus and the four ribonucleoproteins of translation. J Cell Biol 2000; 148:1091-5. 42. Pederson T. Movement and localization of RNA in the cell nucleus. FASEB J 1999; 13(Suppl 2):S238-42. 43. Pederson T. The plurifunctional nucleolus. Nucleic Acids Res 1998; 26:3871-6. 44. Verheggen C, Mouaikel J, Thiry M et al. Box C/D small nucleolar RNA trafficking involves small nucleolar RNP proteins, nucleolar factors and a novel nuclear domain. EMBO J 2001; 20:5480-90. 45. Verheggen C, Lafontaine DL, Samarsky D et al. Mammalian and yeast U3 snoRNPs are matured in specific and related nuclear compartments. EMBO J 2002; 21:2736-45.

CHAPTER 4

Dynamics of Nucleolar Components Thierry Cheutin, Tom Misteli and Miroslav Dundr

Introduction

T

he nucleolus is one of the best-characterized cellular organelles. Its large physical size facilitated its early discovery and led to its detailed morphological description.1-3 The essential role of the nucleolus in ribosome biogenesis has encouraged extensive functional studies and to date the nucleolus is the only nuclear sub-compartment with a well-defined function.4-6 Nucleoli are found in virtually all nucleated cells of most animals and tissues at most stages of development.7,8 The ubiquitous presence of nucleoli and the fact that nucleoli undergo little changes in shape or size when observed for extended periods of time in living cells by light microscopy have led to the intuitive assumption that nucleoli are stable, static structures. This view has recently been challenged and it has now become clear that the nucleolus is a highly dynamic intracellular organelle.2-6 Early indications that the nucleolus is not a static structure, but can be plastic, came from the observation that the morphological appearance of the nucleolus is often tissue-specific. In fact, clinicians have traditionally used the structural features of the nucleolus as an indicator for pathological changes of tissues. Furthermore, in most mammalian cell types the nucleolus increases in size as cells progress from G1 to G2, thus indicating that the nucleolus is structurally dynamic. The most dramatic manifestation of the dynamic nature of the nucleolus, however, occurs during mitosis, when it disassembles completely.9-11 As cells enter pro-metaphase ribosomal gene expression ceases. Concomitantly the nucleolus loses its structural integrity and its components disperse throughout the mitotic cell.5,11-13 Upon entry into telophase the nucleolus reforms around the transcriptionally reactivated ribosomal genes in the nucleolar organizing region (NOR). This correlation between transcriptional activity of ribosomal genes and structural integrity of the nucleolus is not limited to mitotic repression of rDNA transcription. In Xenopus oocytes nucleoli disassemble rapidly in the ovulated egg which is transcriptionally inactive.14 Similarly, treatment of cells with transcription inhibitors results in rapid reorganization of nucleolar structure.15-17 While these observations clearly demonstrate that the overall architecture of the nucleolus is dynamic, recent in vivo imaging approaches have revealed an additional, molecular, level of nucleolar dynamics. These experiments indicate that nucleolar resident proteins and RNA molecules are highly dynamic. They do not reside permanently in the nucleolus, but are steadily exchanged between the nucleolus and the surrounding nucleoplasm. Remarkably, the residence time of proteins and RNAs within the nucleus is on the order of tens of seconds and as a consequence the apparently stable global structure of the nucleolus reflects a highly dynamic steady-state of rRNA and proteins flowing through the nucleolus.18-20 The dynamic nature of nucleolar components at the molecular level has important consequences for the mechanisms which govern nucleolar architecture and function. This chapter aims to give an overview of the recent findings related to the newly discovered dynamic properties of nucleolar components and to discuss their implications. The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

30

The Nucleolus

Figure 1. Vectorial migration of rRNA as observed by pulse-chase labeling. Indicated times correspond to the duration of the chase. Black areas correspond to regions where label is detected.

Dynamics of Ribosomal RNA The main function of the nucleolus is to generate preribosomal particles for export to the cytoplasm. The biochemical steps of this process are well characterized and form a sequential series of events: rRNA transcription, rRNA processing, and assembly of preribosomal particles.21,22 Due to the central role of rRNA synthesis determination of nucleolar morphology, following the dynamics of rRNA transcription/processing is key to the understanding of the functional organization of the nucleolus. To uncover the dynamics of rRNA processing, 3H-UTP has been used to map rRNA synthesis and processing sites within the nucleolus using electron microscopy and autoradiography.23 These pulse-chase experiments allow to indirectly follow the dynamics of rRNA synthesis. More recently, this approach has been improved by replacing 3 H-UTP with Br-UTP which can be detected by light microscopy immuno-labeling. This new approach increases the spatial resolution and allows for more convenient observation by optical microscopy.24-31 These experiments demonstrate that nucleoli are dynamic structures, displaying a centrifugal organization.31 Upon labeling, incorporation of UTP first takes place in small foci distributed in the central parts of nucleoli (Fig. 1). Correlative electron microscopy studies identify these foci as dense fibrillar components (DFC) and their associated fibrillar centers (FC). These data concerning the incorporation of UTP are in agreement with other studies showing that

Dynamics of Nucleolar Components

31

the transcription machinery and early processing events occur in similar structures, and they demonstrate that rRNA transcription as well as the first step of rRNA processing occurs in the central parts of nucleoli.5,32-34 After a chase of 15-30 min, rRNAs are distributed in a wider area forming a network around unstained DFC + FC (Fig. 1). By electron microscopy, UTP is now found all over the DFC and the GC (Granular Component). These results strongly suggest that rRNA moves during its processing from the central nucleolar regions towards the external regions. The size of these foci progressively increases and they then fuse to form a network. Biochemical studies and localization of proteins involved in nucleolar functions indicate that during this stage rRNA undergoes processing events and begins to form complexes with ribosomal proteins.32,35 After 30-45 min of chase, rRNAs accumulate at the nucleolar periphery (Fig. 1). Electron microscopy indicates that the GC is now the only labeled compartment. During this stage, rRNAs continue their centrifugal migration and reach the nucleolar periphery. These results are consistent with previous studies showing that 45S rRNA is restricted to fibrillar components whereas 32S rRNA (pre-28S) is also found in granular components.36,37 After a chase for up to 60 min, rRNA molecules are also found in the cytoplasm indicating that pre-ribosomal particles move from the nucleolar periphery to the cytoplasm during this time. Indeed, labeled small ribosomal subunits can be seen in the cytoplasm as early as 30 min after UTP incorporation, whereas large ribosomal subunits need 45-60 min to be fully processed and observed in the cytoplasm.38 In these experiments, no particular paths or tracks have been observed between the nucleolar periphery and the nuclear envelope, suggesting that pre-ribosomal particles move largely by diffusion inside the nucleoplasm. Although pulse chase experiments with UTP have greatly improved our understanding of nucleolar rRNA dynamics, they have not allowed us to directly observe the behavior of rRNA in living cells. No stably associated rRNA-binding proteins suitable for use in GFP-fusion approaches to visualize rRNA in living cells has been reported. Similarly, experiments incorporating fluorescent nucleotides to probe rRNA dynamics in living cells have not yet been successfully implemented. Such approaches would be useful to visualize the vectorial migration of rRNA from central areas in nucleoli towards the periphery and to analyze this transport process in its molecular detail in living cells. This should soon be possible since ribosomal RNA has recently been visualized in living cells using microinjected fluorochrome-labeled antisense 2'-O-methyl oligoribonucleotides complementary to 28S rRNA.39 In conclusion, pulse chase experiments demonstrate that rRNA is synthesized in the central part of the nucleolus and moves centrifugally towards the nucleolar periphery during processing and pre-ribosomal particle formation. Further studies using living cells will be needed to visualize such movements and to directly study rRNA mobility in vivo. Nevertheless, one might speculate that rRNA mobility is relatively low in comparison to the nucleolar proteins involved in their processing, suggesting that rRNA can form a substrate on which nucleolar functions can take place.

Dynamics of the RNA Polymerase I Transcription Complex At the core of the nucleolus are the tandemly repeated ribosomal genes, which are exclusively transcribed by the RNA polymerase I (RNA pol I) transcription complex. In eukaryotes this multiprotein complex is made up of at least 12 subunits.40 In humans the RNA pol I complex consists of three basal components: the chromatin remodeling HMG box-containing upstream binding factor (UBF), the promoter-recognizing TATA-binding protein (TBP)-complex and the RNA polymerase I itself.41 UBF and the TBP-complex are responsible for the formation of a preinitiation complex on the ribosomal gene promoter.42 The latter complex is subsequently involved in recruitment of RNA pol I to the promoter to form the transcriptionally competent complex.41,43,44 The RNA pol I machinery has been extensively characterized biochemically.42-44 However, most of these experiments do not directly address the kinetic properties of the RNA pol I machinery and many studies have used in vitro systems, which do not necessary reflect all

32

The Nucleolus

Figure 2. Dynamics of RNA polymerase I. 1. RNA Pol I components rapidly exchange between the nucleoplasm and the nucleolus where they scan for rDNA promoters. 2. RNA pol I components have an equal chance to associate with an rDNA promoter or to leave the fibrillar center. 3. The incorporation of RNA pol I components into an elongating holoenzyme is inefficient. 4. Residence times of RNA pol I components indicate that RNA pol I machineries are not recycled but rather reassemble at each round of transcription.

features of cellular physiology. To address dynamic aspects of the RNA pol I transcription in living cells photobleaching microscopy has recently been used.45 These in vivo microscopy techniques allow measurement of dynamics of fluorescently tagged proteins in single living cells. In the most commonly used photobleaching method, Fluorescence Recovery After Photobleaching (FRAP), a small region of interest in a cell is rapidly bleached by a short laser pulse and the recovery of the fluorescent signal in the bleached area due to movement of unbleached molecules from the surrounding areas is quantitatively monitored. FRAP is a useful method to determine the mobility and binding properties of proteins in living cells.46,47 These types of experiments reveal that RNA polymerase I is a dynamic protein complex.45 The components of the preinitiation complex are rapidly and continuously exchanged from sites of ribosomal DNA transcription, suggesting that they only transiently associate with ribosomal genes. Similarly, the major fraction of components of the RNA pol I holoenzyme, including RPA194, the catalytic center of the polymerase, is not engaged in elongation at any given time but only a minor fraction of about 7-10% is associated with elongating polymerase at steady-state.45 Interestingly, this situation appears to be similar to that of the largest subunit of RNA polymerase II, possibly suggesting that some features of RNA polymerase I dynamics are conserved amongst all mammalian RNA polymerases.48,49 These experiments are also consistent with a model in which RNA pol I components are imported into the nucleolus as distinct subunits rather than as a completely preassembled holoenzyme. Application of computational kinetic modeling analysis to these in vivo microscopy data allowed determination of various quantitative properties of the RNA polymerase I machinery. The elongation time for the synthesis of a single rRNA transcript in vivo could be estimated to

Dynamics of Nucleolar Components

33

be about 140s, corresponding to a polymerization rate of ~95 nt/s for the ~13.3 kb long primate rDNA gene and an initiation interval of ~1.4s. Based on an occupancy of approximately 100-120 RNA pol I molecules on a single ribosomal gene,50 the probability of RNA pol I components that have entered the nucleolus to associate with the rDNA promoter can be calculated to be ~50% for all RNA pol I subunits. In striking contrast, the efficiency of incorporation of RNA pol I subunits into an elongation-competent complex was significantly lower, in the range of 1%-10% depending on the subunit.45 On the basis of these qualitative and quantitative observations several points regarding the assembly and elongation of RNA pol I in living cells can be made45 (Fig. 2). Several thousand molecules of each RNA pol I subunit pass through the nucleolus per second and collide, likely randomly, with promoter sequences where they are retained for short periods of time. The residence time for all RNA pol I components in the nucleolus is on the order of seconds if they are not successfully incorporated into an elongating RNA pol I complex. The continuous fast exchange of the vast majority of RNA pol I components between the nucleolus and the nucleoplasm suggests that RNA pol I machineries are not recycled after transcription termination but rather reassembled at each round of transcription.51 Different incorporation frequencies of pol I components suggest that RNA pol I subunits assemble in a sequential order. Consistent with biochemical data, the transcription factor TIF-IA/Rrn3 might act as a bridge between the preinitiation complex and RNA pol I52 since it has an incorporation frequency that is significantly lower than all other components of the elongation complex. RPA43, the subunit recruiting RNA pol I to the promoter through the interaction with transcription factor TIF-IA/ Rrn3,53 has a low incorporation frequency similar to the largest subunit of RNA pol I, RPA194. These observations suggest that the largest subunit RPA194 and the recruiting factors RPA43 and PAF53 join the preinitiation complex relatively early and then act as a nucleation center for assembly of the holoenzyme in a sequential manner possibly via metastable intermediates.

Dynamics of Components of Pre-Ribosomal RNA Processing and Ribosome Assembly Once the rRNA is fully synthesized by the RNA pol I complex, the transcript undergoes a series of processing, covalent nucleotide modification and folding steps, until it gives rise to the mature 18S, 5.8S and 28S rRNA which assemble into small and large ribosomal subunits. The maturation of pre-rRNA and assembly of ribosomal subunits involves a large number of factors including small nucleolar RNAs.5,54 Comparisons of the mobility of the factors involved in different steps of ribosome biogenesis has revealed distinct kinetic properties. The factors involved in various steps of pre-rRNA processing, including nucleolin, fibrillarin, protein B23 and Rpp29, exhibit fast FRAP recovery kinetics, suggesting that they are rapidly exchanged between the nucleolus and the surrounding nucleoplasm.18,19,55 When RNA pol I transcription is inhibited by actinomycin D, these factors reside in the nucleolus for a shorter period of time18,19 indicating that the pre-rRNA transcripts are likely responsible for retention of processing factors.18,19 Similarly rapid exchange occurs for fibrillarin in prenucleolar bodies at the end of mitosis where the protein presumably interacts with mitotically preserved pre-rRNA.13 In contrast, ribosomal proteins S5-GFP and L9-GFP exhibit significantly slower recovery kinetics and longer residence times in the nucleoli on the order of hundreds of seconds.19 This observation is consistent with the possibility that the slower kinetics of ribosomal proteins reflect the interaction of these proteins with assembling pre-ribosomal subunits in the nucleolus. These findings imply that rDNA transcription and pre-rRNA processing factors functionally cycle between the nucleolus and the nucleoplasm in contrast to ribosomal proteins which are exported in preribosomal subunits.

Dynamics of Various Nucleolar Proteins Apart from RNA pol I components and ribosomal processing factors the dynamic properties of several additional nucleolar components have recently been analyzed. The serine/threonine

34

The Nucleolus

protein phoshatase 1 (PP1) is involved in many different regulatory events and its substrate specificity is determined by targeting subunits by forming a complex with the PP1 catalytic subunit to bring the enzyme to specific cellular locations.56 Trinckle-Mulcahy et al57 studied the kinetic behavior of three isoforms of PP1, PP1α, PP1β, and PP1γ in living cells. PP1α and PP1β were primarily localized in the nucleoplasm, but PP1γ was found diffusely localized throughout the nucleoplasm and accumulated in the nucleolus. To study the interaction of these three PP1 isoforms with one of their targeting subunits, the authors chose the major PP1-interactive protein, nuclear inhibitor of PP1 (NIPP1). Interestingly, NIPP1 is localized primarily in the nucleoplasmic speckles where it colocalizes with pre-mRNAs splicing factors.57 Photobleaching experiments on PP1γ show that PP1γ is highly mobile and rapidly exchanges between the nucleolus and the nucleoplasm.57 Surprisingly, when NIPP1 is coexpressed with PP1γ in living cells, NIPP1 is able to relocalize PP1γ from the nucleolus to the nucleoplasmic speckles. These data indicate that PP1γ interacts with NIPP1 even though that predominant localization patterns of both proteins suggest that they are primarily segregated. Due to rapid exchange of PP1γ between the nucleolus and the nucleoplasm, PP1γ is continuously present in the nucleoplasm where it can functionally interact with NIPP. DNA topoisomerases I, IIα and IIβ are essential for DNA replication, chromosome condensation, segregation of daughter chromatids and they are stable structural components of the chromosomal scaffold.58,59 The major fraction of topoisomerases resides in the nucleolus. It has been documented that topo I and II play a role in maintenance of nucleolar structure and folding of active ribosomal genes. Photobleaching revealed that the nucleolar pools of topo I and both isoforms of topo II are less mobile than the pools present in the nucleoplasm.60,61 However, when cells are treated with camptothecin which stabilizes covalent topo I-DNA intermediates, GFP-topo I was relocated within 30 s from the nucleoli to radial substructures in the nucleoplasm. FRAP analysis performed on GFP-topo I in nucleoplasmic structures after camptothecin treatment revealed that the enzyme mobility was significantly slowed down. By contrast, GFP-topo I exchanged faster in the nucleolus than in the nucleoplasm after camptothecin treatment. A similar relocation from nucleoli to substructures in the nucleoplasm was observed for GFP-topo II isoforms when cells were treated with teniposide (VM26) which also stabilizes covalent catalytic DNA intermediates of topo II.61 These findings suggest that topoisomerases are very dynamic proteins that continuously search through the nucleus for binding sites on DNA. The nucleolar pools of topo I and II do not necessarily represent sites where topo I and II are most actively engaged in DNA catalysis and these observations indicate that topo I and II may be retained in the nucleolus due to protein-protein interactions rather than its binding to DNA.61 A proteomic analysis performed on purified human nucleoli has identified a large number of novel nucleolar proteins.62,63 One of these proteins is Paraspeckle Protein 1 (PSP1).64 PSP1 was found to accumulate in a new nucleoplasmic compartment, termed paraspeckles. Paraspeckles are discrete bodies, which are frequently localized adjacent to the splicing factor compartment. When PSP1 is detected by a specific antibody or expressed in cells as a GFP-fusion protein, it is not detectable in nucleoli. However, when transcription is inhibited by actinomycin D or DRB, PSP1 is dramatically redistributed to discrete caps at the periphery of the nucleolus.64 A similar redistribution was observed for the other known paraspeckle proteins PSP2 and p54/nrb. The paradox of the virtual absence of PSP1 in the nucleolus under normal conditions visualized by microscopy and its presence detected by mass spectroscopy was solved by photobleaching experiment showing that PSP1 continuously cycles between paraspeckles and nucleoli resulting in a steady-state accumulation in the nucleolus. These data powerfully demonstrate that the steady-state distribution of a protein in the nucleus, and other nuclear compartments, can be the result of its dynamic trafficking.

Dynamics of Nucleolar Components

35

Figure 3. Nucleolar morphology reflects the steady-state dynamics of its components. Nucleolar localization is due to dynamic exchange of proteins between the nucleoplasm and the nucleolus. Proteins that interact with relatively more immobile constituents in the nucleolus (rDNA, rRNA, proteins) are slowed down in their exchange and are thus effectively retained.

Trafficking to and through the Nucleolus Although it is well establish that nuclear proteins are transported to nuclei by means of a nuclear localization sequence (NLS), we still do not know how proteins are targeted to the nucleolus. Nucleolar localization signals have been described suggesting that targeting to the nucleolus is an active process. This hypothesis comes from studies where minimal sequences necessary for nucleolar localization have been identified using mutation analysis. Consequently, these sequences are defined as nucleolar localization signals (NoLS). For example, amino acids 949-1092 plus the NLS containing amino-acids 1358-1432 of the Werner syndrome protein, are sufficient to target this protein to nucleoli.65 Such nucleolar localization sequences have also been identified in p14ARF, MDM266 and in numerous viral proteins found in nucleoli.67,68 Furthermore, NoLS have been identified in ribosomal proteins, which are synthesized in the cytoplasm and move to the nucleolus where they bind rRNA and other ribosomal proteins to forms preribosomal particles. In accordance with the active targeting hypothesis two signals have been described for ribosomal protein L5: amino-acids 21-37 are sufficient for targeting to the nucleolus and amino-acids 101-111 constitute a nuclear export signal.69 Although NoLSs often contain sequences rich in basic amino acids, they are heterogeneous in size and sequence. The absence of a canonical NoLS sequence suggests that different mechanisms can transport proteins into the nucleolus. An alternative and simpler model arises from live-cell microscopy. These experiments show that many nucleolar proteins are not exclusively localized in the nucleolus, but are also present throughout the nucleoplasm and that most nucleolar proteins are highly mobile in the nucleoplasm as well as in nucleoli.20 These data demonstrate that nucleolar proteins have the capacity by virtue of their diffusional mobility to access most of the nucleus, thus implying that they may find their way to the nucleolus in the absence of specific targeting signals. Once in the nucleolus, they may interact with other nucleolar components, either proteins or RNAs, which will slow down their diffusion, effectively trapping them transiently in the nucleolus70 (Fig. 3). In this case, the identified NoLS does not act as a true targeting signal but merely mediates an interaction between these sequences and other nucleolar components leading to retention.

36

The Nucleolus

This model is consistent with the absence of known intra-nuclear transport mechanisms. Furthermore, such a passive retention model involving various types of protein-protein interactions is also consistent with the inability to define NoLSs from large proteomic analyses of more than 300 nucleolar proteins.62,63 Further evidence for passive targeting is the fact that numerous nucleolar proteins localize to multiple nuclear compartments.6 For example, many nucleolar proteins also accumulate in the Cajal body. If targeting were active, multiple pathways with multiple signals would be required, whereas a passive targeting process furnishes an easy explanation for the occurrence of proteins in multiple compartments. Several recent studies demonstrate that nucleolar proteins can shuttle between nuclear compartments. For example newly synthesized Sm proteins accumulate in nucleoli and Cajal bodies before reaching nuclear speckles.71 Similarly, the NHPX protein, which binds to RNA sequences found in snoRNAs containing box C/D and U4 colocalizes with fibrillarin in nucleoli and Cajal bodies in living cells as well as in fixed cells.72 Microinjection of EYFP-NHPX surprisingly showed that, in contrast to the Sm proteins, this protein traffics to speckles before going to nucleoli and Cajal bodies. These experiments suggest that proteins can follow a specific pathway between nuclear organelles to become fully active. In this regard, similar to the cytoplasm, the nucleus appears to be composed of numerous stations where specific processes occur. However, proteins not only populate the correct and specific nuclear location where they function, but some of them also shuttle between nuclear organelles as a consequence of their transient binding to partners and their ability to diffuse relatively freely throughout the nucleoplasm. A passive targeting process based on the dynamic properties of proteins is an elegant solution to explain how proteins accumulate in specific nuclear organelles such as the nucleolus.

Implications of Nucleolar Dynamics The dynamic nature of most nucleolar components reflects the function of the nucleolus as a synthesis and processing site. Ribosomal RNAs are generated in the interior of the nucleolus and must be vectorially transported through the organelle. Many nucleolar proteins are involved in processing of these moving pre-rRNAs and must thus transiently associate with the RNA molecules. All of these events are dynamic themselves and contribute to the dynamic turnover of proteins and RNA in the nucleolus. The dynamic properties of nucleolar proteins also have implications for their functional regulation via signaling pathways. Ribosomal gene expression is efficiently regulated by the ERK kinase pathway via phosphorylation of UBF.73 Similarly, a cycle of phosphorylation and dephosphorylation of Rrn3, presumably as a consequence of yet unidentified signaling events, is required for rDNA transcription.52 Since these RNA pol I subunits are only transiently associated with the ribosomal promoter, it is possible that the modifying kinases and phosphatases are not nucleolar and act on the nucleoplasmic pool of these proteins. The fact that overexpression of the PP1-regulatory subunit NIPP1 displaces PP1γ from the nucleolus supports this notion.57 Thus, each time a molecule dissociates from the ribosomal gene and is exchanged into the nucleoplasm it might become the substrate for regulatory kinases or phosphatases. Modification in the nucleoplasm might also apply to other RNA pol I components and rRNA processing factors which likely undergo cycles of modifications that regulate their association and dissociation with their target RNA.41,74 Individual nucleolar proteins and RNAs are clearly dynamic. However, the nucleus as a whole is structurally stable. How is this apparent paradox resolved? The answer is simply that the stable structure of the nucleolus represents the steady-state association and dissociation events of the individual dynamic nucleolar components (Fig. 3). It is likely that most nuclear proteins at some time traverse the nucleolus as they are diffusing relatively freely throughout the nucleus.20 Only molecules for which binding sites are present in the nucleolus will be retained for significant amounts of time and thus appear to be ‘resident’ proteins. Therefore,

Dynamics of Nucleolar Components

37

the structure of the nucleolus is largely a reflection of its function. Consistent with such a model in which the functional interactions of proteins define the morphological appearance of the organelle is the observation that nucleolar structure is strongly dependent on transcriptional activity.4,16,70 This self-organizing behavior of the nucleolus is further supported by the sequence of events involved in mitotic reassembly.11,13 As cells enter telophase, RNA polymerase I transcription resumes. The transcription machinery associates with the ribosomal genes and by synthesizing new pre-rRNA transcripts creates a seed for the association of the necessary processing factors. The more transcripts are generated, the more factors associate and the interphase appearance of the nucleolus is gradually restored. The fact that the presence of a processing target is critical for the proper localization of a protein to the nucleolus is dramatically illustrated by the observation that during reassembly the early processing factors such as fibrillarin accumulate prior to accumulation of late processing factors such as protein B23.13 These observations, together with the dynamic nature of molecular components of the nucleolus, are consistent with an important role for self-organization in nucleolar biogenesis.70 Despite the close correlation between transcriptional activity and nucleolar morphology, the presence of structural nucleolar elements can not be ruled out. An obvious structural component of the nucleolus, and likely a template in its self-organization, is the ribosomal DNA. The fibrillar centers are morphological structures maintained even after hypotonic shock or treatments which inhibit transcription.75 Nucleoli form around chromosome regions containing NORs and mini-nucleoli are formed upon introduction of rDNA sequences on plasmids.76 Therefore, rDNA sequences clearly serve as templates for nucleolus formation. Furthermore, in Xenopus laevis the nucleolar protein NO145 localizes in a shell surrounding the nucleolus consistent with its potential role as a nucleolar matrix component.77 In addition, in a proteomic analysis of nucleolar components, structural proteins such as actin, were detected.62,63 These observations suggest that the dynamic interplay of nucleolar components with each other and possibly with a yet-unidentified nucleolar skeleton is largely responsible for nucleolar architecture.

Conclusions The mammalian interphase nucleolus is highly dynamic. At a global level, the overall nucleolar structure is dynamic as it is lost during mitosis and is dependent on ongoing RNA pol I transcription. In addition, nucleolar components, with the exception of the rDNA, are not static residents of the nucleolus but are rapidly exchanged with the nucleoplasm. The morphological appearance of the nucleolus is very tightly linked to its transcriptional activity and rRNA processing functions. Therefore, the nucleolus is a prime example of a dynamic cellular structure formed by self-organization around a stable template. It will be critically important to uncover how common this principle of subcellular organization is. While the high level of dynamics in the nucleolus might be surprising at first, in hindsight it seems logical that the function of the nucleolus as a synthesis and processing site with many factors sequentially associating and dissociating with the rRNA as it is being processed gives rise to a highly dynamic structure. It now remains to be seen to what extent dynamics contribute to regulation of nucleolar functions and how signaling pathways affect the dynamic behavior of nucleolar components and in this way affect nucleolar structure and function. Much remains to be learned about this most prominent cellular organelle, but the realization that nucleolar components are dynamic has added a new dimension in the understanding of nucleolar structure and function.

Acknowledgements We thank Dr. Stan Gorski for critical reading of the manuscript. TM is a Keith R. Porter Fellow.

38

The Nucleolus

References 1. Harris H. The birth of the cell. New Haven: Yale University Press, 1999. 2. Scheer U, Benavente R. Functional and dynamic aspects of the mammalian nucleolus. Bioessays 1990; 12:14-21. 3. Carmo-Fonseca M, Mendes-Soares L, Campos I. To be or not to be in the nucleolus. Nat Cell Biol 2000; 2:E107-112. 4. Lewis JD, Tollervey D. Like attracts like: Getting RNA processing together in the nucleus. Science 2000; 288:1385-1389. 5. Olson MOJ, Dundr M, Szebeni A. The nucleolus: An old factory with unexpected capabilities. Trends Cell Biol 2000; 10:189-196. 6. Dundr M, Misteli T. Functional architecture in the cell nucleus. Biochem J 2001; 356:297-310. 7. Gurdon JB. Cytoplasmic regulation of RNA synthesis and nucleolus formation in developing embryos of Xenopus laevis. J Mol Biol 1965; 12:27-35. 8. Newport J, Kirschner M. A major developmental transition in early Xenopus embryos: I. characterization and timing of cellular changes at the midblastula stage. Cell 1982; 30:675-86. 9. Heitz E. Die Ursache der gesetzmaessigen Zahl, Lage, Form und Groesse planzlicher Nukleolen. Planta 1931; 12:775-884. 10. McClintock B. The relation of a particular chromosomal element to the development of nucleoli in Zea mays. Z Zellforschung 1934; 21:294-328. 11. Hernandez-Verdun D, Roussel P, Gebrane-Younes J. Emerging concepts of nucleolar assembly. J Cell Sci 2002; 115:2265-70. 12. Dundr M, Olson MOJ. Partially processed pre-rRNA is preserved in association with processing components in nucleolus-derived foci during mitosis. Mol Biol Cell 1998; 9:2407-22. 13. Dundr M, Misteli T, Olson MOJ. The dynamics of postmitotic reassembly of the nucleolus. J Cell Biol 2000; 150:433-46. 14. Gurdon JB. Changes insomatic cell nuclei inserted into growing and maturing amphibian oocytes. J Embryol Exp Morphol 1963; 20:401-414. 15. Benavente R, Rose K, Reimer G et al. Inhibition of nucleolar reformation after microinjection of antibodies to RNA polymerase I into mitotic cells. J Cell Biol 1987; 105:1483-91. 16. Oakes M, Nogi Y, Clark MW et al. Structural alterations of the nucleolus in mutants of Saccharomyces cerevisiae defective in RNA polymerase I. Mol Cell Biol 1993; 13:2441-2555. 17. Melese T, Xue Z. The nucleolus: An organelle formed by the act of building a ribosome. Curr Opi Cell Biol 1995; 7:319-324. 18. Phair RD, Misteli T. High mobility of proteins in the mammalian cell nucleus. Nature 2000; 404:604-609. 19. Chen D, Huang S. Nucleolar components involved in ribosome biogenesis cycle between the nucleolus and nucleoplasm in interphase cells. J Cell Biol 2001; 153:169-176. 20. Misteli T. Protein dynamics: Implications for nuclear architecture and gene expression. Science 2001; 291:843-847. 21. Hadjiolov AA. The Nucleolus and Ribosome Biogenesis. Springer Verlag, 1985. 22. Sollner-Webb B, Tyc K, Steitz J. Ribosomal RNA processing in Eucaryotes. In: Zimmerman R. DA, ed. Ribosomal RNA: Structure, evolution, processing and function in protein synthesis. Boca Raton: CRC Press, 1996:469-490. 23. Fakan S. High resolution autoradiography studies on chromatin functions. In: H B, ed. The cell nucleus. New York: Academic Press, 1978:3-53. 24. Dundr M, Raska I. Nonisotopic ultrastructural mapping of transcription sites within the nucleolus. Exp Cell Res 1993; 208:275-81. 25. Jackson DA, Hassan AB, Errington RJ et al. Visualization of focal sites of transcription within human nuclei. EMBO J 1993; 12:1059-65. 26. Wansink DG, Schul W, van der Kraan I et al. Fluorescent labeling of nascent RNA reveals transcription by RNA polymerase II in domains scattered throughout the nucleus. J Cell Biol 1993; 122:283-93. 27. Iborra FJ, Pombo A, McManus J et al. The topology of transcription by immobilized polymerases. Exp Cell Res 1996; 229:167-73. 28. Wei X, Somanathan S, Samarabandu J et al. Three-dimensional visualization of transcription sites and their association with splicing factor-rich nuclear speckles. J Cell Biol 1999; 146:543-58. 29. Pombo A, Jackson DA, Hollinshead M et al. Regional specialization in human nuclei: Visualization of discrete sites of transcription by RNA polymerase III. EMBO J 1999; 18:2241-53. 30. Koberna K, Stanek D, Malinsky J et al. Nuclear organization studied with the help of a hypotonic shift: Its use permits hydrophilic molecules to enter into living cells. Chromosoma 1999; 108:325-35.

Dynamics of Nucleolar Components

39

31. Thiry M, Cheutin T, O’Donohue MF et al. Dynamics and three-dimensional localization of ribosomal RNA within the nucleolus. RNA 2000; 6:1750-61. 32. Shaw PJ, Jordan EG. The nucleolus. Annu Rev Cell Dev Biol 1995; 11:93-121. 33. Scheer U, Hock R. Structure and function of the nucleolus. Curr Opin Cell Biol. 1999; 11:385-90. 34. Huang S. Building an efficient factory: Where is pre-rRNA synthesized in the nucleolus? J Cell Biol. 2002; 157:739-41. 35. Thiry M, Goessens G. The nucleolus during the cell cycle. Molecular Biology Intelligence Unit. New York: R.G.Landes Company/Springer, 1996. 36. Daskal Y, Prestayko A, Busch H. Ultrastructural and biochemical studies of the isolated fibrillar component of nucleoli from Novikoff hepatoma ascites cells. Exp Cell Res 1974; 88:1-14. 37. Royal A, Simard R. RNA synthesis in the ultrastructural and biochemical components of the nucleolus of Chinese hamster ovary cells. J Cell Biol 1975; 66:577-85. 38. Crocker J. Molecular and biochemical aspects of interphase nucleolar organizer regions. J Clin Pathol 1996; 49:M9-M11. 39. Molenaar C, Marras SA, Slats JC et al. Linear 2' O-Methyl RNA probes for the visualization of RNA in living cells. Nucleic Acids Res 2001; 29:E89-9. 40. Bischler N, Brino L, Carles C et al. Localization of the yeast RNA polymerase I-specific subunits. EMBO J 2002; 21:4136-44. 41. Moss T, Stefanovsky VY. At the center of eukaryotic life. Cell 2002; 109:545-8. 42. Paule M. Transcription of ribosomal genes by eukaryotic RNA polymerase I. New York: Springer Verlag, 1998. 43. Grummt I. Regulation of mammalian ribosomal gene transcription by RNA polymerase I. Prog Nucleic Acid Res Mol Biol 1999; 62:109-54. 44. Paule MR, White RJ. Survey and summary: Transcription by RNA polymerases I and III. Nucleic Acids Res 2000; 28:1283-98. 45. Dundr M, Hoffmann-Rohrer U, Hu Q et al. A kinetic framework for a mammalian RNA polymerase in vivo. Science 2002; 298:1623-6. 46. Lippincott-Schwartz J, Snapp E, Kenworthy A. Studying protein dynamics in living cells. Nat Rev Mol Cell Biol 2001; 2:444-56. 47. Phair RD, Misteli T. Kinetic modelling approaches to in vivo imaging. Nat Rev Mol Cell Biol 2001; 2:898-907. 48. Becker M, Baumann C, John S et al. Dynamic behavior of transcription factors on a natural promoter in living cells. EMBO Rep 2002; 3:1188-94. 49. Kimura H, Sugaya K, Cook PR. The transcription cycle of RNA polymerase II in living cells. J Cell Biol 2002; 159:777-82. 50. Miller OJ, Bakken AH. Morphological studies of transcription. Acta Endocrinol 1972; 168(Suppl):155-77. 51. Aprikian P. Moorefield B, Reeder RH. New model for the yeast RNA polymerase I transcription cycle. Mol Cell Biol 2001; 21:4847-55. 52. Cavanaugh AH, Hirschler-Laszkiewicz I, Hu Q et al. Rrn3 phosphorylation is a regulatory checkpoint for ribosome biogenesis. J Biol Chem 2002. 53. Peyroche G, Milkereit P, Bischler N et al. The recruitment of RNA polymerase I on rDNA is mediated by the interaction of the A43 subunit with Rrn3. EMBO J 2000; 19:5473-82. 54. Fatica A, Tollervey D. Making ribosomes. Curr Opin Cell Biol 2002; 14:313-318. 55. Snaar S, Wiesmeijer K, Jochemsen AG et al. Mutational analysis of fibrillarin and its mobility in living human cells. J Cell Biol 2000; 151:653-62. 56. Hubbard MJ, Cohen P. On target with a new mechanism for the regulation of protein phosphorylation. Trends Biochem Sci 1993; 18:172-7. 57. Trinkle-Mulcahy L, Sleeman JE, Lamond AI. Dynamic targeting of protein phosphatase 1 within the nuclei of living mammalian cells. J Cell Sci 2001; 114:4219-28. 58. Bakshi RP, Galande S, Muniyappa K. Functional and regulatory characteristics of eukaryotic type II DNA topoisomerase. Crit Rev Biochem Mol Biol 2001; 36:1-37. 59. Champoux J. DNA topoisomerase: Structure, function, and mechanism. Annu Rev Biochem 2001; 70:369-413. 60. Christensen MO, Barthelmes HU, Feineis S et al. Changes in mobility account for camptothecin-induced subnuclear relocation of topoisomerase I. J Biol Chem 2002; 277:15661-5. 61. Christensen MO, Larsen MK, Barthelmes HU et al. Dynamics of human DNA topoisomerases IIalpha and IIbeta in living cells. J Cell Biol 2002; 157:31-44.

40

The Nucleolus

62. Andersen JS, Lyon CE, Fox AH et al. Directed proteomic analysis of the human nucleolus. Curr Biol 2002; 12:1-11. 63. Scherl A, Coute Y, Deon C et al. Functional proteomic analysis of human nucleolus. Mol Biol Cell 2002; 13:4100-9. 64. Fox AH, Lam Y, Leung AK et al. Paraspeckles: A novel nuclear domain. Curr Biol 2002; 12:13-25. 65. von Kobbe C, Bohr V. A nucleolar targeting sequence in the Werner syndrome protein resides within residues 949-1092. J Cell Sci 2002; 115:3901-7. 66. Lohrum MA, Ashcroft M, Kubbutat MH et al. Identification of a cryptic nucleolar-localization signal in MDM2. Nat Cell Biol 2000; 2:179-81. 67. Catez F, Erard EM, Schaerer-Uthurralt N et al. Unique motif for nucleolar retention and nuclear export regulated by phosphorylation. Mol Cell Biol 2002; 22:1126-39. 68. Hiscox J. The nucleolus—A gateway to viral infection? Arch Virol 2002; 147:1077-89. 69. Rosorius O, Fries B, Stauber RH et al. Human ribosomal protein L5 contains defined nuclear localization and export signals. J Biol Chem 2000; 275:12061-8. 70. Misteli T. The concept of self-organization in cellular architecture. J Cell Biol 2001; 155:181-5. 71. Sleeman JE, Lamond AI. Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway. Curr Biol 1999; 9:1065-74. 72. Leung AK, Lamond AI. In vivo analysis of NHPX reveals a novel nucleolar localization pathway involving a transient accumulation in splicing speckles. J Cell Biol 2002; 157:615-29. 73. Pelletier G, Stefanovsky VY, Faubladier M et al. Competitive recruitment of CBP and Rb-HDAC regulates UBF acetylation and ribosomal transcription. Mol Cell 2000; 6:1059-66. 74. Sirri V, Hernandez-Verdun D, Roussel P. Cyclin-dependent kinases govern formation and maintenance of the nucleolus. J Cell Biol 2002; 156:969-81. 75. Mosgoeller W, Schofer C, Steiner M et al. Arrangement of ribosomal genes in nucleolar domains revealed by detection of “Christmas tree” components. Histochem Cell Biol 2001; 116:495-505. 76. Karpen GH, Schaefer JE, Laird CD. A Drosophila rRNA gene located in euchromatin is active in transcription and nucleolus formation. Genes Dev 1988; 2:1745-1763. 77. Kneissel S, Franke WW, Gall JG et al. A novel karyoskeletal protein: characterization of protein NO145, the major component of nucleolar cortical skeleton in Xenopus oocytes. Mol Biol Cell 2001; 12:3904-18.

CHAPTER 5

Behavior of the Nucleolus during Mitosis Danièle Hernandez-Verdun

Abstract

T

he nucleolus is the ribosome factory and also a multifunctional domain that plays an important role in nuclear organization and function. The nucleolus is assembled at the end of mitosis, is active during interphase, and dis-assembled in prophase. The nucleolar machineries of transcription and processing are inherited from parental to daughter cells through mitosis. The polymerase I transcription machinery is repressed during mitosis although assembled with ribosomal genes (rDNA). The repression of pol I transcription is achieved at the end of prophase and is maintained during mitosis through phosphorylation of transcription factors by cyclin-dependent kinase (CDK) 1. The nucleolar processing machineries relocalize from the nucleolus towards the periphery of all chromosomes in early prophase. The processing complexes remain associated with chromosomes until telophase and this chromosome association depends on CDK1 activity. In telophase, as a consequence of natural inhibition of CDK1 activity, pol I transcription is restored. The processing machineries are recruited to the sites of rDNA transcription after a temporary transit in foci assembled on the chromosome surface. These immobile foci called prenucleolar bodies (PNBs) could correspond to an assembly platform of processing complexes prior to recruitment to transcription sites. The contents of the PNBs are delivered to nucleoli with different kinetics and consequently PNBs have different life-times depending on their contents. Partly processed pre-rRNAs generated during prophase are inherited during mitosis and are found in PNBs. The possibility that the processing complexes forming PNBs can be nucleated by some pre-rRNAs passing through mitosis is proposed. Thereafter, nucleolar assembly is completed by cooperative interactions between chromosome territories. Thus the behavior of the nucleolus during mitosis illustrates the fact that the dynamics of nuclear organization at the beginning of interphase are integrated in a network of interactions and controls that is largely dependent on the coordination of mitotic events.

Introduction In higher eukaryotes, the nucleus is dis-assembled when chromosomes condense at the beginning of mitosis and re-assembled at the end of mitosis. During mitosis there is redistribution and/or inactivation of the nuclear machineries that will be further involved in re-building of nuclear functions. The behavior of these complexes during mitosis is crucial to ensure the establishment of nuclear functions in the two daughter cells. Consequently, this period of the cell cycle is particularly convenient to investigate how the complexes involved in nuclear functions are controlled at the time of repression or initial recruitment. The nucleolus is the factory in which ribosome subunits are synthesized and assembled before being exported to the cytoplasm (for a review see refs. 1-3). This nuclear domain results from an equilibrium between the level of ribosomal RNAs (rRNAs) synthesis, the efficiency of rRNA processing, the assembly of pre-rRNAs with ribosomal proteins, and finally the export of the ribosomal subunits. The steady state between transcription, processing and export of The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishing.

42

The Nucleolus

Figure 1. Nucleolus during mitosis. The nucleolar activity during interphase depends on RNA pol I transcription (Transcript) and rRNA specific processing machineries (Processing). At the G2/prophase transition, the rRNA processing machinery leaves the nucleolus and during prophase becomes distributed over the surface of all the condensed chromosomes. RNA pol I transcription is still active in early prophase and is arrested at the end of prophase. At prometaphase, metaphase and anaphase pol I transcription is repressed. The reactivation of pol I transcription in telophase is concomitant with the gathering of the rRNA processing machinery into PNBs at the chromosome periphery. The processing machineries then translocate from the PNBs to the transcription sites in the nucleolus (Nu) until early G1. The period during which the CDK1 activity controls repression and redistribution of the nucleolar machineries is indicated.

ribosomal subunits engenders a large nuclear space (a third of the total nuclear volume in yeast) with the highest cellular RNA concentration (80% of the total RNAs are synthesized in the nucleolus). During the cell cycle of higher eukaryotic cells, nucleoli assemble at the exit from mitosis in a manner concomitant with restoration of rDNA transcription and are functionally active throughout interphase (Fig. 1). Conversely, at the beginning of mitosis the nucleoli disassemble and are no longer observed throughout mitosis (Fig. 1). Nucleolar disassembly and assembly provide a good model to study the mitotic reorganization of interphasic nuclear machineries since the nucleolar machineries are specific of the different stages of ribosome biogenesis; they are abundant and nucleolar function requires cooperation between transcription and processing machineries (for a review see refs. 4-5). Consequently, the establishment of nucleolar functions at the end of mitosis depends on the activation, targeting and/or recruitment of the nucleolar machineries involved in transcription of the ribosomal genes (rDNA) and processing of rRNAs. In cycling cells, nucleolar assembly is generally initiated during telophase and continues for 1-2 hours into early G1 phase (Fig. 1). Nucleolar assembly at the exit of mitosis benefits from the machinery and complexes inherited from the previous cell cycle. Indeed the nucleolus dis-assembled at the beginning of mitosis and the nucleolar machineries are transmitted from parental to daughter cells.

Behavior of the Nucleolus during Mitosis

43

Furthermore, the formation of an active nucleolus is very important for nuclear architecture, functional compartmentalization of the nucleus and control of cell proliferation (for a review see refs. 6-12). Here, we report the general features of the nucleolus in open mitosis. In open mitosis the chromatin is condensed into chromosomes, the functional nucleus is dis-assembled and no transcriptional activity is detectable except at the transition between mitosis and interphase. This is completely different in yeast (Saccharomyces cerevisae) where during closed mitosis transcription continues and in particular an active nucleolus is maintained. This chapter is dedicated to the nucleolar dis-assembly and assembly in open mitosis. We discuss recent findings concerning the cell cycle controls on nucleolar assembly, the dynamics of delivery of the processing machinery and the role of pre-rRNAs in stabilizing the nucleolar machinery. These findings provide a more integrated view of the assembly process and its dynamics in which the localization of a component reflects its time of residency and its binding affinity.

Disassembly of Nucleolus at the Beginning of Mitosis In cycling cells, interphasic nucleoli are very dynamic structures varying greatly in size, number and position from one cell type to another. Nevertheless, three main specific components are identified in most eukaryotic nucleoli at electron microscopy resolution, the fibrillar centers (FC), the dense fibrillar component (DFC) and the granular component (GC). It has been demonstrated that these three main components correspond to different stages of ribosome biogenesis (see other chapters in this book). Briefly, the initiation of rDNA transcription occurs at the junction FC-DFC, early processing of the rRNAs in the DFC and late processing in the GC. This nucleolar organization is the consequence of the coordination between transcription of the rDNA and processing of the rRNAs.

Release of Nucleolar Processing Machinery At the end of G2, when the compaction of chromatin into chromosomes can be detected at the nuclear periphery while the nuclear envelope is still present, the shape of the nucleolus is modified. This is due to the progressive release of the processing machineries from the nucleolus. This release is visible in fixed cells using antibodies directed against DFC and GC.13 These nucleolar proteins are present in trabeculae emerging from the nucleolus and extending towards the periphery of the nucleus. As prophase progresses, the released nucleolar proteins form a network in the entire nuclear volume in the spaces between chromosomes. Indeed, while chromatin condenses into chromosomes during prophase, relocalization of the rRNA processing factors occurs and these nucleolar machineries translocate to the chromosome periphery forming a perichromosomal compartment (Figs. 1 and 2) (for a review see ref. 14). This behavior of the nucleolar processing proteins at the entry of mitosis is general since it was observed in several mammalian (human, marsupial, hamster, mouse) cell lines, in Xenopus and plant cells.15 The nucleolar proteins which relocate to the chromosome periphery are components of the DFC and GC of the active nucleolus. Fibrillarin in association with U3 snoRNAs (DFC) as well as protein B23 and Nop52 (GC) are representative of these nucleolar proteins.16-18 In addition to these prototypes, other nucleolar proteins behave similarly (for example nucleolin, PM-Scl 100 and Ki 67) supporting the hypothesis that proteins from the DFC and GC most frequently translocate from the nucleolus to the chromosome periphery during mitosis. The mechanism controlling this translocation is presently unknown as is the relationship between the nucleolar proteins and the chromosomes. It is noticeable that depending on the fixation procedure, the proteins can be solubilized in the cytoplasm and no longer detected at the chromosome periphery. Similarly, to preserve the organization of the perichromosomal compartment, cryofixation is more appropriate than chemical fixatives.19 Therefore, the location of the nucleolar proteins observed in fixed cells could reflect a concentration gradient of nucleolar proteins as proposed for nuclear constituents located around the chromosomes.

44

The Nucleolus

Figure 2. Distribution of an rRNA processing protein Nop52. Upper panel: in the two early G1 nuclei (DNA revealed by Dapi), Nop52 revealed by antibodies is detected in the nucleoli (Nu) and in PNBs (arrows). Lower panel: in metaphase cell, the chromosomes are condensed (DNA revealed by Dapi), and Nop52 revealed by antibodies is detected around all the chromosomes. Bar: 5µm. Courtesy of M. Savino (IJM, Paris, France).

However, the nucleolar complexes located at the chromosome periphery can be isolated with the chromosomes,13,20 suggesting a certain degree of association with chromosomes. In living cells, the nucleolar proteins tagged with GFP, fibrillarin-GFP and Nop52-GFP are concentrated around the chromosomes during mitosis and migrate with the chromosomes during anaphase.21 These coordinated movements of the nucleolar proteins and chromosomes indicate that they maintain their interactions during this process.

Repression of rDNA Transcription When cells enter into mitosis, transcription is detected (Fig. 3) by Br-UTP incorporation in prophase PtK1 cells.22 Such transcription is localized in the nucleolus and inhibited by low doses of actinomycin D, the signature of rDNA transcription mediated by RNA polymerase I (pol I). Quantitative analysis of tritiated uridine incorporation in human and hamster cells at different periods of mitosis, indicate that nucleolar transcription decreases by about 30% during early prophase compared to interphase and stopped in late prophase.23 Even if timing of pol I transcription arrest during the prophase remains to be precisely determined, it most probably occurs concomitantly or shortly before nuclear envelope breakdown. After breakdown, the nucleolus is no longer visible among condensed chromosomes. There are some exceptions to this mitotic nucleolar dis-assembly when a residual nucleolar body is still found in the cytoplasm of mitotic (hamster (CHO) cells and some plant cells) but this residual nucleolar body is no longer associated with the rDNA and is not at the site of transcription.24

Behavior of the Nucleolus during Mitosis

45

Figure 3. Repression of rRNA transcription during mitosis. Ribosomal DNA transcription was revealed in PtK1 cells by Br-UTP incorporation (images a-e). Images (a’-e’) present DNA staining with Dapi and allow identification of the mitotic stage. a) interphase cells. Several transcription beads are visible in the area corresponding to nucleoli. b) prophase cell; transcription is still visible. c and d) prometaphase and anaphase respectively; no labeling is observed. e) early telophase; transcription starts. In each set of chromosomes, two doublets of transcription per NORs are observed. In PtK1 cells there are two NORs but depending on the focus either one or two beads (arrows) are detected. Bar: 5 µm. From Gébrane-Younès J et al. J Cell Sci ©1997.

46

The Nucleolus

Partly Processed rRNA Generated at the Beginning of Mitosis In prophase, the rRNA processing machinery is released from the nucleolus before total repression of rDNA transcription. The rRNAs synthesized at this period accumulate as partially processed 45S pre-rRNAs detected by 5’ETS core probes.25 This confirms the previous observations of 45S and 32S rRNAs present in mitotic cells even in metaphase-arrested cells,26 and the hypothesis that RNA contained at the chromosome periphery is synthesized at the end of G2 or at the beginning of prophase.27 In prophase, these pre-rRNAs are localized in the nucleoli whereas after nucleolar dis-assembly they are at the chromosome periphery, and also in the cytoplasm from prophase to telophase (Fig. 4). After prophase, no significant signal for the presence of pre-RNAs has been found at the nucleolar organizer regions (NORs), i.e., clusters of rDNAs, supporting the hypothesis that these pre-rRNAs are released from the template.28-30 These partially processed pre-rRNAs are colocalized with rRNA processing proteins at the chromosome periphery and also in nucleolar derived foci (NDF) observed in the cytoplasm of mitotic cells.15,25,30 This colocalization is consistent with the fact that ribonucleoprotein complexes containing both pre-rRNAs and processing proteins are isolated from mitotic cells.31 In conclusion, nucleolar breakdown occurs in prophase concomitantly with chromatin compaction into mitotic chromosomes. It is a two step process involving first the translocation of the processing machineries around the chromosomes while transcription of rDNA is still active, followed by repression of rDNA transcription. Consequently, partly processed rRNAs are generated during prophase. The repression of rDNA transcription in late prophase depends on the CDK1-cyclin B pathway. Conversely, we presently do not know how dis-assembly of the processing machineries from the rDNA transcription sites in prophase is regulated and controlled.

Nucleolus during Mitosis Organization of the NOR and Ribosomal Genes during Mitosis Ribosomal DNAs are located in NORs of chromosomes, also called secondary constrictions. The number of NORs is species-specific, varying in mammals from one pair in Potorous tridactylis (PtK cells) to five pairs in humans, i.e. the acrocentric chromosomes 13, 14, 15, 21, 22. Electron microscopy shows that NORs are not a constriction but a region in which condensed chromatin in the chromosome axis is surrounded by less compact material.22,32-34 In this material, non-condensed chromatin (Fig. 5) associated with UBF (pol I transcription factor) was detected.22 Thus during mitosis, the NOR chromatin exhibits two configurations, only one of which is associated with the transcription factors and corresponds more likely to the competent rDNA (Fig. 6). It is presently not clear if the condensed pedicle corresponds to repressed rDNA copies. Alternatively, the axis could correspond to non transcribed spacer of the rDNA because in extracted chromosomes they are predominantly associated with the axis rather than the transcribed sequences.35

Pol I Transcription Machinery Remains Assembled in a Repressed Form during Mitosis From prometaphase to anaphase the rDNA transcription activity is abolished (Figs. 1 and 3). However, even in the absence of transcription, the components of the rDNA transcription machinery are colocalized with rDNA. In human cells there are 10 NOR-bearing chromosomes, but the rDNA transcription machinery (as observed for the pol I complex, SL1, UBF and TTF-1) remains associated in variable amounts with only six competent NORs, at least in the HeLa cell line used.36,37 It is interesting to note that each component varies in the same proportion in the different positive NORs suggesting that the complexes are maintained during mitosis. The amount of proteins associated with both chromatids of the same chromosome appears equivalent. Consequently, the separation of chromatids occurring during anaphase, leads to the equal partition of the rDNA transcription machinery and therefore between the two daughter cells.36

Behavior of the Nucleolus during Mitosis

47

Figure 4. Partially processed rRNA during mitosis. Unprocessed rRNA detected by the 5’ETS core (left row), fibrillarin detected by immunolabeling (middle row) and merge of the two labelings in one optical section at different periods of mitosis (right row). A) Prophase, B) Anaphase, C) Telophase. D) enlargement of the merge signals of A , E) enlargement of the merge signal of B. Bar :10 µm. Courtesy of T. Dousset (IJM, Paris, France).

48

The Nucleolus

Figure 5. Chromatin in NOR during mitosis. Ultrathin section of a NOR in prophase showing chromatin compaction. The DNA was contrasted by a specific staining after removing RNAs and blocking the protein contrast (NAMA-Ur staining). An axis of condensed chromatin forming a bridge between two parts of condensed chromosome is visible. Fine DNA fibers emerge from this axis and occupy the width of the chromosome (arrows). Bar: 0.1 µm. From Gébrane-Younès J et al. J Cell Sci © 1997.

Figure 6. Schematic representation of rDNA configuration during mitosis. In the NOR the rDNA copies (black) are either condensed as the adjacent parts of the chromosome (gray), or non-condensed and associated with the transcription factor UBF (circles). From Gébrane-Younès J et al. J Cell Sci © 1997.

Behavior of the Nucleolus during Mitosis

49

Figure 7. Transcription in mitotic cells by inhibition of CDK1. Ribosomal DNA transcription is restored in mitotic cells by in vivo inhibition of the CDK1-cyclin B activity by roscovitine. The transcription sites detected by BrUTP incorporation are observed in this metaphase HeLa cell on the DAPI-stained chromosomes aligned at the metaphase plate as shown by superimposition of the phase-contrast image. Bar: 10 ∝m. Bar: 10 µm. From Sirri et al. J Cell Biol © 2000.

During mitotic repression of rDNA transcription, the partners composing the basal pol I transcription machinery remain associated with rDNA chromatin as demonstrated by salt solubilization.37 It was proposed that inhibition of mitotic RNA pol I transcription was caused by phosphorylation of components of the rDNA transcription machinery directed by the CDK1-cyclin B.38,39 Indeed SL1 and TTF-1 are phosphorylated differently during mitosis than during interphase. The role of CDK1-cyclin B in mitotic repression of rDNA transcription was clearly demonstrated by inhibiting the CDK1-cyclin B kinase pathway in mitotic cells (Figs. 1 and 7). Indeed, inhibition of this pathway in mitotic cells induces resumption of rDNA transcription.39 However, the resulting transcripts are not processed since 47S-46S rRNAs accumulate and a nucleolus is not visible in these mitotic cells by phase contrast microscopy.

Distribution of the Processing Machinery around the Chromosome Periphery during Mitosis Is CDK1-Dependent It was pointed out long ago that the chromosome periphery is a particular compartment containing ribonucleoproteins (RNP) (27 and references therein). The ultrastructural organization of this compartment is complex, and is mainly constituted of a network of fibrils and granules.40 At the chromosome periphery the nucleolar processing complexes are regularly

50

The Nucleolus

distributed without visible foci using immunofluorescence labeling. The amount of complexes at the chromosome periphery progressively increases during anaphase and telophase. In addition, the same nucleolar complexes can accumulated in nucleolar derived foci (NDF) scattered throughout the cytoplasm during anaphase and telophase.30,41 This suggests a cell-cycle related behavior. Indeed, inhibition of CDK1 in mitotic-arrested cells induces relocation of the nucleolar processing machineries,42 demonstrating that location at the chromosome periphery is controlled during mitosis by the CDK1 pathway (Fig. 1). Inhibition of CDK1 induces the formation of foci corresponding to PNB and only a partial translocation of the processing machinery to transcription sites. Accordingly, there is no rRNA cleavage of the new transcripts synthesized in mitotic-arrested cells.42 Part of the early processing machinery is efficiently recruited on these pre-rRNAs indicating that the CDK1 pathway regulates this recruitment. As methylation and pseudouridylation of these large rRNAs are not documented, we do not know if the CDK1 pathway is also involved in the control of these rRNA modifications.

Assembly of the Nucleolus at Exit of Mitosis Nucleolar assembly occurs when cells exit from mitosis. Early reports suggested that nucleolar assembly depends on the activation of the pol I transcription machinery.43-45 This generates pre-rRNAs (47S in mammals), that recruit the rRNA-processing machinery. Noticeably, proteins and small nucleolar RNAs (snoRNAs) involved in rRNA processing are observed in PNBs, before localizing at sites containing newly transcribed rRNAs.46 It was proposed that PNBs are mobile nuclear bodies that participate in the delivery of the rRNA-processing complexes to sites of rDNA transcription.46

Activation of Transcription At telophase, the resumption of rDNA transcription occurs simultaneously in each pol I transcription machinery-associated NOR and the level of rDNA transcription activity seems to be directly related to the amount of pol I machinery present in the NOR.36 In human HeLa cells, the activation of the six competent NORs induces six individual foci of pol I transcription. From past electron microscopy investigations, we know that this initial transcription of the rDNA is located close to the reforming nuclear envelope.47,48 Recently in living cells, we also observed that the NORs, imaged by fibrillarin recruitment at the initial stage of nucleolar building, are distributed in the nucleus in six distinct foci moving slowly within the nuclear volume without apparent coordination.21 The CDK1-cyclin B kinase activity maintains mitotic repression of rDNA transcription. Indeed, inhibition of the CDK1-cyclin B kinase pathway in mitotic cells induces resumption of rDNA transcription. We propose that inactivation of the CDK1-cyclin B kinase occurring normally in telophase is sufficient to release mitotic repression of rDNA transcription while this not sufficient to restore proper nucleolar assembly.42

Relocation of the Processing Machinery In living cells, the processing nucleolar proteins tagged with GFP such as fibrillarin-GFP and Nop52-GFP, are concentrated around the chromosomes during mitosis and migrate with the chromosomes during anaphase.21 In telophase, the tagged proteins concentrate in many foci, corresponding to PNBs (Figs. 2 and 8). Several proteins of the rRNA processing machinery, the U3 and U14 snoRNAs and partially processed rRNAs are detected in PNBs.17,24,25,46,49-52 The PNBs display variable ultrastructures, reflecting different concentrations and associations of their components especially in the form of granules.21,53,54 We recently demonstrated that the PNBs form on the chromosome surface and remain associated with condensed chromatin.21 Strikingly, fibrillarin concentrates in PNBs and NORs (Fig. 9) when the decrease in CDK1 activity overcomes the mitotic repression of RNA pol I transcription,55 while PNBs-containing Nop52 appear later and are progressively recruited in NORs (Fig. 10). Indeed, PNBs have different lifetimes.21 Thus, it seems that recruitment of the processing

Behavior of the Nucleolus during Mitosis

51

Figure 8. 3-dimensional visualization of PNB distribution in early G1 nuclei. Distribution of B23-GFP in an early G1 nucleus. 3-dimensional (3-D) reconstruction of 91 sections, 0.25 µm thick, deconvoluted with Huygens. GFP is visible in the nucleoli and in the many PNBs distributed in the nucleoplasm. Courtesy of C. Chamot (IJM, Paris, France). (www.eurekah.com/abstract.php?chapid=1354&bookid=88&catid=54)

Figure 9. Dynamics of fibrillarin movement from telophase to early G1. Time-lapse sequences of fibrillarin-GFP from mitosis to early G1. The formation of PNBs starts in telophase. NORs become distinct 7 min after the beginning of telophase. The DFC-containing fibrillarin then expands progressively. Projection of 33 focal planes, acquisition frequency every 30 sec. From Savino et al. J Cell Biol © 2001. (www.eurekah.com/abstract.php?chapid=1354&bookid=88&catid=54)

52

The Nucleolus

Figure 10. Dynamics of Nop52 from metaphase to G1. Time-lapse sequences of Nop52-GFP corresponding to the projection of 15 focal planes collected every 15 sec. Nop52-GFP is found at the periphery of the chromosomes during metaphase, it follows the chromosome during anaphase reaching the poles after 4 min. Between 10 and 15 min, Nop52-GFP is redistributed and concentrates in PNBs. Nop52-GFP is then recruited in the nucleoli. At a high magnification, flow of the protein between PNB and the nucleolus is visible and is indicated by arrows. From Savino et al. J Cell Biol © 2001. (www.eurekah.com/ abstract.php?chapid=1354&bookid=88&catid=54)

machinery at the time of nucleolar assembly is a regulated process probably depending on cell cycle progression. Based on observations using fixed cells, the prediction was that the PNBs move for delivery of preassembled processing complexes to the site of rDNA transcription. PNB dynamics in living cells do not reveal such directed movement towards the nucleolus.21,52 Rather, the progressive delivery of PNB proteins to the nucleoli is ensured by directional flow between PNBs themselves and between PNBs and the nucleolus (Fig. 11).21 The processing machinery is first concentrated in the PNBs, and then released in a time-dependent manner. The role of this intricate delivery pathway remains an open question, in particular if it is cell cycle controlled. When the nucleolar function is established, the recruitment of processing proteins no longer depends on PNBs. Indeed, the dynamics of fibrillarin and other processing proteins analyzed by fluorescence recovery after photobleaching indicate rapid diffusion in the nucleoplasm and permanent recruitment in the nucleolus.56-58 However, the mobile fraction can be different in nucleoli and Cajal (Coiled) bodies,57 indicating that the residence time depends on specific interactions.59 Therefore the formation of PNBs at the mitosis/interphase transition suggests a steady state at this period of the cell cycle, favoring residence of processing factors close to the condensed chromatin either by self-assembly of processing factors or by specific interaction with pre-rRNAs. The role of proteins in PNB assembly was questioned using antibodies directed against PNB proteins. Immunodepletion of one protein did not preclude the recruitment of other proteins in PNBs assembled in Xenopus egg extracts except for fibrillarin.60 In accordance with this observation, blocking fibrillarin translocation during mitosis modified the PNB pathway,17 as opposed to antibodies directed against another PNB protein.50 Interestingly, the anti-fibrillarin antibodies had a negative effect on nucleolar assembly and RNA pol I transcription.

Behavior of the Nucleolus during Mitosis

53

Figure 11. Dynamics of Nop52 from metaphase to G1. Schematic reconstruction of the nucleolus. Proteins at the periphery of chromosomes (dashed line) assemble in PNBs attached to chromosomes. Brownian movement of PNBs is observed, protein delivery between PNBs and contact of some PNBs with newly forming nucleolus (NOR). Several nucleoli are formed at this time and can fuse to one another. A certain lapse of time is required until the nucleolar components reorganize after fusion. From Savino et al. J Cell Biol © 2001.

Thus at early stages of processing, the PNB complex formation seems a crucial event for nucleolar assembly.

Role of Partly Processed rRNA in Assembly At the mitosis/interphase transition, the processing complexes forming PNBs can be nucleated by pre-rRNAs passing through mitosis.25,30,31,52 This scenario is based on the fact that pre-rRNAs are localized in PNBs,25 and the pre-RNAs as well as rRNA processing intermediates are immunopurified in mitotic processing complexes.31 This model of PNB formation by association around pre-rRNAs could explain the temporal order of nucleolar delivery of the processing machinery driven by pre-rRNA stability. Self-association of processing proteins in PNBs cannot be excluded because PNBs are generated without any rRNAs in reconstituted nuclei in Xenopus extracts.53,54,61 However, during de novo nucleolar assembly in Xenopus embryos prior to activation of zygotic transcription, the presence of pre-rRNAs was demonstrated in association with regrouping of PNBs around NORs.53,62 Similarly, partially processed pre-RNAs are present in the NDF in association with nucleolar processing components.30 An intringuing question is why the association in distinct bodies of the nucleolar processing machinery with pre-rRNA occurs at the end of mitosis since pre-rRNAs and released nucleolar processing machinery are present as of the beginning of mitosis. In conclusion, the recruitment of the nucleolar processing machinery during nuclear assembly in early G1 involves the formation of PNBs which is either an assembly or a storage platform of processing complexes associated with pre-rRNAs. The stability of these PNBs seems to be cell cycle regulated.

54

The Nucleolus

Association of Different NORs About one hour after the initial formation of six independent new nucleoli, fusion of these initial nucleoli is observed.21 Fusion appears to be a rapid process leading to reorganization of the nucleolus (Fig. 11) concomitant with general movement in the nucleus. These data indicate that during this period the position of the chromosome territories corresponding to NOR-bearing chromosomes is strongly modified. To explain this large-scale reorganization of the nuclear architecture, the hypothesis of interaction between the nucleolar products generated by ribosome biogenesis can be discarded for several reasons. First, in living cells we observed new nucleoli containing parts of the processing machinery and moving close to each other without fusion. Second, since association of the initial nucleoli takes place at a precise time during the cell cycle, this is more probably a regulated event rather than an event that depends on the amount of nucleolar activity which is variable in the different initial nucleoli. Finally, the most striking evidence is provided by the demonstration that even silent NORs are associated with nucleoli.63 It was proposed that regroupment of NORs more likely depends on the heterochromatin adjacent to rDNA.

Conclusions and Perspectives

As proposed by Mélèse,2 the nucleolus is a “functional domain formed by the act of building a ribosome.” The steady state between transcription, processing and export of ribosomal subunits engenders this large domain, the nucleolus present during interphase. During the time of nucleolar assembly or dis-assembly, this steady state is different since the pol I machinery remains assembled with rDNA in a repressed form whereas the processing machineries are delocalized to the chromosome periphery. Recent findings indicate that nucleolar assembly at exit from mitosis depends on cell cycle controls. Indeed, CDK1 activity represses pol I transcription during mitosis, and its inactivation releases this silencing. The formation of PNBs is also controlled by the CDK1 pathway, whereas the recruitment of the rRNA-processing machinery appears to depend on the activity of another CDK. The characterization of the CDK(s) controlling this process should be investigated in the future. Ribosome biogenesis involves the pol-I, pol-II and pol-III-dependent transcription pathways, the translocation of 5S RNAs and the ordered assembly of ribosomal proteins. Presently we do not know how these pathways are coordinated at the periods of assembly or dis-assembly of the nucleolus. This should therefore be an important goal of future research. The pre-rRNAs generated by pol I transcription are localized at the sites of active rDNA gene clusters. The binding affinity of the processing proteins for these pre-rRNAs can explain the compartmentalization of the processing machinery in the functional nucleolus. During nucleolar assembly, pre-rRNAs also appear to participate in compartmentalization of the processing machinery. Mitotic pre-rRNAs are involved in the reformation of the nucleolus after mitosis, and maternal pre-rRNAs in Xenopus embryos are involved in regrouping PNBs around rDNA. In both situations, the intriguing question is how the mitotic pre-rRNAs or the maternal pre-rRNAs regroup around rDNA genes. The presence of pre-rRNAs in PNBs and NDFs could also explain the formation of temporarily organized bodies. The stability of these rRNAs could determine their lifetime. Clearly, these questions must be addressed if we are to understand the role of stable rRNAs in the formation and/or maintenance of nucleolar structures. Another interesting and promising field of research is how the presence of a functional nucleolus contributes to the general nuclear architecture. It has been recently reported that the loci at the nucleolar periphery are significantly less mobile than the others and disruption of nucleoli increase the mobility of nucleolar-associated loci.12 This indicates that chromatin associated with the nucleolus is more restricted in its movements than non-associated chromatin. It is interesting to note that the nucleolus is the first active nuclear domain to be assembled at the end of mitosis; it creates a domain of sequestration or exclusion of molecules participating in cellular functions different from ribosome biogenesis. The exclusion of pol II from the nucleolus could

Behavior of the Nucleolus during Mitosis

55

explain that the nucleolus is a site of silencing for inserted pol II genes. However we do not know how this exclusion property is generated and when during assembly of nucleolar function it occurs.

Acknowledgments The author thanks A. L. Haenni for critical reading of the manuscript and M. Savino, T. Dousset, C. Chamot for providing their unpublished data. This work was supported in part by grants from the Centre National de la Recherche Scientifique and the Association pour la Recherche sur le Cancer (Contract 4290).

References 1. Hadjiolov AA. The nucleolus and ribosome biogenesis. Wien: Springer-Verlag, 1985:1-268. 2. Mélèse T, Xue Z. The nucleolus: an organelle formed by the act of building a ribosome. Curr. Opin Cell Biol 1995; 7:319-324. 3. Shaw PJ, Jordan EG. The nucleolus. Annu Rev Cell Dev Biol 1995; 11:93-121. 4. Scheer U, Thiry M, Goessens G. Structure, function and assembly of the nucleolus. Trends Cell Biol 1993; 3:236-241. 5. Scheer U, Weisenberger D. The nucleolus. Curr Opin Cell Biol 1994; 6:354-359. 6. Strouboulis J, Wolffe AP. Functional compartmentalization of the nucleus. J. Cell Sci. 1996; 109:1991-2000. 7. Lamond AI, Earnshaw WC. Structure and function in the nucleus. Science 1998; 280:547-553. 8. Pederson T. The plurifunctional nucleolus. Nucleic Acids Res 1998; 26:3871-3876. 9. Cockell MM, Gasser SM. The nucleolus: nucleolar space for rent. Current Biol 1999; 9:R575-R576. 10. Carmo-Fonseca M, Mendes-Soares L, Campos I. To be or not to be in the nucleolus. Nature Cell Biol 2000; 2:107-112. 11. Olson MOJ, Dundr M, Szebeni A. The nucleolus: an old factory with unexpected capabilities. Trends Cell Biol 2000; 10:189-196. 12. Chubb JR, Boyle S, Perry P et al. Chromatin motion is contrained by association with nuclear compartments in human cells. Curr Biol 2002; 12:439-445. 13. Gautier T, Robert-Nicoud M, Guilly M-N et al. Relocation of nucleolar proteins around chromosomes at mitosis- A study by confocal laser scanning microscopy. J Cell Sci 1992; 102:729-737. 14. Hernandez-Verdun D, Roussel P, Gautier T. Nucleolar proteins during mitosis. Chromosomes Today 1993; 11:79-90. 15. Medina FJ, Cerdido A, Fernandez-Gomez ME. Components of the nucleolar processing complex (pre-rRNA, fibrillarin, and nucleolin) colocalize during mitosis and are incorporated to daughter cell nucleoli. Exp Cell Res 1995; 221:111-125. 16. Ochs RL, Lischwe MA, Spohn WH et al. Fibrillarin: a new protein of the nucleolus identified by autoimmune sera. Biol Cell 1985; 54:123-134. 17. Fomproix N, Gébrane-Younes J, Hernandez-Verdun D. Effects of anti-fibrillarin antibodies on building of functional nucleoli at the end of mitosis. J Cell Sci 1998; 111:359-372. 18. Gautier T, Fomproix N, Masson C et al. Fate of specific nucleolar perichromosomal proteins during mitosis: Cellular distribution and association with U3 snoRNA. Biol Cell 1994; 82:81-93. 19. Gautier T, Masson C, Quintana C et al. The ultrastructure of the chromosome periphery in human cells. An in situ study using cryomethods in electron microscopy. Chromosoma 1992; 101:502-510. 20. Schubert I, Dolezel J, Houben A et al. Refined examination of plant metaphase chromosome structure at different levels made feasible by new isolation methods. Chromosoma 1993; 102:96-101. 21. Savino TM, Gébrane-Younès J, De Mey J et al. Nucleolar assembly of the rRNA processing machinery in living cells. J Cell Biol 2001; 153:1097-1110. 22. Gébrane-Younès J, Fomproix N, Hernandez-Verdun D. When rDNA transcription is arrested during mitosis, UBF is still associated with non-condensed rDNA. J Cell Sci 1997; 110:2429-2440. 23. Prescott DM, Bender MA. Synthesis of RNA and protein during mitosis in mammalian tissue culture cells. Exp Cell Res 1962; 26:260-268. 24. Azum-Gélade M-C, Noaillac-Depeyre J, Caizergues-Ferrer M et al. Cell cycle redistribution of U3 snRNA and fibrillarin. Presence in the cytoplasmic nucleolus remnant and in the prenucleolar bodies at telophase. J Cell Sci 1994; 107:463-475. 25. Dousset T, Wang C, Verheggen C et al. Initiation of nucleolar assembly is independent of RNA polmerase I transcription. Mol Biol Cell 2000; 11:2705-2717.

56

The Nucleolus

26. Fan H, Penman S. Regulation of synthesis and processing of nucleolar components in metaphase-arrested cells. J Mol Biol 1971; 59:27-42. 27. Moyne G, Garrido J. Ultrastructural evidence of mitotic perichromosomal ribonucleoproteins in hamster cells. Exp Cell Res 1976; 98:237-247. 28. Weisenberger D, Scheer U. A possible mechanism for the inhibition of ribosomal RNA gene transcription during mitosis. J Cell Biol 1995; 129:561-575. 29. Beven AF, Lee R, Razaz M et al. The organization of ribosomal RNA processing correlates with the distribution of nucleolar snRNAs. J Cell Sci 1996; 109:1241-1251. 30. Dundr M, Olson MOJ. Partially processed pre-rRNA is preserved in association with processing components in nucleolus derived foci during mitosis. Mol Biol Cell 1998; 9:2407-2422. 31. Pinol-Roma S. Association of nonribosomal nucleolar proteins in ribonucleoprotein complexes during interphase and mitosis. Mol Biol Cell 1999; 10:77-90. 32. Hsu TC, Brinkley BR, Arrighi FE. The structure and behaviour of the nucleolus organizers in mammalian cells. Chromosoma 1967; 23:137-153. 33. Ghosh S, Paweletz N. The nucleolar chromatin and the secondary constriction. Cell Biol Intern Rep 1990; 14:681-687. 34. Suja JA, Gébrane-Younès J, Géraud G et al. Relative distribution of rDNA and proteins of the RNA Polymerase I transcription machinery at chromosomal NORs. Chromosoma 1997; 105:459-469. 35. Bickmore W, Oghene K. Visualizing the spatial relationships between defined DNA sequences and the axial region of extracted metaphase chromosomes. Cell 1996; 84:95-104. 36. Roussel P, André C, Comai L et al. The rDNA transcription machinery is assembled during mitosis in active NORs and absent in inactive NORs. J Cell Biol 1996; 133:235-246. 37. Sirri V, Roussel P, Hernandez-Verdun D. The mitotically phosphorylated form of the transcription termination factor TTF-1 is associated with the repressed rDNA transcription machinery. J. Cell Sci 1999; 112:3259-3268. 38. Heix J, Vente A, Voit R et al. Mitotic silencing of human rRNA synthesis: inactivation of the promoter selectivity factor SL1 by cdc2/cyclin B-mediated phosphorylation. EMBO J 1998; 17:7373-7381. 39. Sirri V, Roussel P, Hernandez-Verdun D. In vivo release of mitotic silencing of ribosomal gene transcription does not give rise to precursor ribosomal RNA processing. J Cell Biol 2000; 148:259-270. 40. Hernandez-Verdun D, Gautier T. The chromosome periphery during mitosis. BioEssays 1994; 16:179-185. 41. Dundr M, Meier UT, Lewis N et al. A class of nonribosomal nucleolar components is located in chromosome periphery and in nucleolus-derived foci during anaphase and telophase. Chromosoma 1997; 105:407-417. 42. Sirri V, Hernandez-Verdun D, Roussel P. Cyclin-dependent kinases govern formation and maintenance of the nucleolus. J Cell Biol 2002; 156:969-981. 43. Benavente R. Postmitotic nuclear reorganization events analyzed in living cells. Chromosoma 1991; 100:215-220. 44. Thiry M, Goessens G. The nucleolus during the cell cycle. Heidelberg: Springer-Verlag, 1996:146. 45. Scheer U, Hock R. Structure and function of the nucleolus. Curr Opin Cell Biol 1999; 11:385-390. 46. Jiménez-Garcia LF, Segura-Valdez MdL, Ochs RL et al. Nucleologenesis: U3 snRNA-containing prenucleolar bodies move to sites of active Pre-rRNA transcription after mitosis. Mol Biol Cell 1994; 5:955-966. 47. Hernandez-Verdun D, Bourgeois CA, Bouteille M. Simultaneous nucleologenesis in daughter cells during late telophase. Bio Cell 1980; 37:1-4. 48. Bourgeois CA, Hubert J. Spatial relationship between the nucleolus and the nuclear envelope: structural aspects and functional significance. Int. Rev Cytol 1988; 111:1-52. 49. Ochs RL, Lischwe MA, Shen E et al. Nucleologenesis: composition and fate of prenucleolar bodies. Chromosoma 1985; 92:330-336. 50. Fomproix N, Hernandez-Verdun D. Effects of anti-PM-Scl 100 (Rrp6p exonuclease) antibodies on prenucleolar body dynamics at the end of mitosis. Exp Cell Res 1999; 251:452-464. 51. Savino TM, Bastos R, Jansen E et al. The nucleolar antigen Nop52, the human homologue of the yeast ribosomal RNA processing RRP1, is recruited at late stages of nucleologenesis. J Cell Sci 1999; 112:1889-1900. 52. Dundr M, Misteli T, Olson MOJ. The dynamics of postmitotic reassembly of the nucleolus. J Cell Biol 2000; 150:433-446.

Behavior of the Nucleolus during Mitosis

57

53. Verheggen C, Almouzni G, Hernandez-Verdun D. The ribosomal RNA processing machinery is recruited to the nucleolar domain before RNA polymerase I during Xenopus laevis development. J Cell Biol 2000; 149:293-305. 54. Verheggen C, Le Panse S, Almouzni G et al. Maintenance of nucleolar machineries and pre-rRNAs in remnant nucleolus of erythrocyte nuclei and remodeling in xenopus egg extracts. Exp Cell Res 2001; 269:23-34. 55. Clute P, Pines J. Temporal and spatial control of cyclin B1 destruction in metaphase. Nature cell Biol 1999; 1:82-87. 56. Phair RD, Misteli T. High mobility of proteins in the mammalian cell nucleus. Nature 2000; 404:604-609. 57. Snaar S, Wiesmeijer K, Jochemsen AG et al. Mutational analysis of fibrillarin and its mobility in living human cells. J Cell Biol 2000; 151:653-662. 58. Chen D, Huang S. Nucleolar components involved in ribosome biogenesis cycle between the nucleolus and nucleoplasm in interphase cells. J Cell Biol 2001; 153:169-176. 59. Misteli T. Protein dynamics: implications for nuclear architecture and gene expression. Science 2001; 291:843-847. 60. Bell P, Scheer U. Prenucleolar bodies contain coilin and are assembled in Xenopus egg extract depleted of specific nucleolar proteins and U3 RNA. J Cell Sci 1997; 110:43-54. 61. Bell P, Dabauvalle MC, Scheer U. In vitro assembly of prenucleolar bodies in Xenopus egg extract. J Cell Biol 1992; 118:1297-1304. 62. Verheggen C, Le Panse S, Almouzni G et al. Presence of pre-rRNAs before activation of polymerase I transcription in the building process of nucleoli during early development of Xenopus laevis. J Cell Biol 1998; 142:1167-1180. 63. Sullivan GJ, Bridger JM, Cuthbert AP et al. Human acrocentric chromosomes with transcriptionally silent nucleolar organizer regions associated with nucleoli. EMBO J 2001; 20:2867-2877.

58

The Nucleolus

CHAPTER 6

Structure and Organization of Vertebrate Ribosomal DNA James E. Sylvester, Iris L. Gonzalez and Edward B. Mougey

Summary

G

enes within rDNA clusters from vertebrates are tandemly repeated in a head to tail fashion and exist at multiple chromosomal locations. Each gene is comprised of a coding region for 18S, 5.8S, and 28S rRNA and an intergenic spacer (IGS). The coding region is transcribed as a long precursor 37S- 47S rRNA whose arrangement is 5’ETS (external transcribed spacer) – 18S – ITS1 (internal transcribed spacer) – 5.8S – ITS2 – 28S – 3’ETS. The mature 18S and 28S consist of highly conserved sequences alternating with variable (divergent or expansion) segments that have high G+C content similar to the transcribed spacers. The variable regions differ in size and sequence from organism to organism, within organisms, and within individuals of a species. The IGS is home to the gene promoter and terminator, the spacer promoter and terminator, and, in most cases, many SINE and LINE elements. Length variable (LV) regions exist upstream of the promoter, downstream of the terminator, and, at least in human, deep within the IGS. The LV regions flanking the gene appear to participate in gene transcription as enhancers and terminators respectively. The IGS also contains many sites of sequence motifs that can adopt an alternative structure such as Z-DNA, triple strand DNA, and bending DNA, however their role in rDNA is not understood. Although there is growing evidence concerning the role of DNA methylation in silencing rDNA transcription, the precise mechanism remains to be elucidated. Another repeated gene family for the fourth rRNA, namely 5S rRNA, resides at a separate chromosomal location and its product is imported into the nucleolus. The multi-copy, repeated nature, and multiple cluster arrangement of the ribosomal genes, which have regions that change at different rates, makes the evolution of these gene systems very complicated involving several distinct mechanisms of concerted evolution.

Introduction The last review concerning the structure and organization of ribosomal DNA (rDNA) was published in 1991 but it did not discuss much of the detailed molecular structure of the gene.1 These authors pointed out that molecular biologists could investigate features of rDNA across the entire range of evolution. While they noted trends, they also stated that the analyses were still tentative and raised more possibilities than they settled. We will discuss the wealth of information on vertebrate rDNAs that has been reported since that review and note some of the many issues that remain unsettled. We will present the molecular structure, genetic variation and comparative evolution of rDNA; we will also review known or potential function of particular segments of the gene. By necessity we will omit a detailed historical overview and instead direct the reader to survey the chapters in the 3-volume book The Cell Nucleus: rDNA2 and the original references contained therein. Finally, most of our discussion will focus on The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic /Plenum Publishers.

Structure and Organization of Vertebrate Ribosomal DNA

59

Figure 1. Schematic diagram of human rDNA. Up arrows denote EcoRI restriction sites which define A (7.3 kb), B (5.8 kb), C (11.8 kb), and D (approximately 20 kb). Large down arrows define the distal and proximal borders of rDNA with telomeric sequences shown as small boxes and centromeric sequences shown as diagonal stripes. The initiation site for transcription, i at +1, and the termination site, t, define the 13 kb primary transcript. The small down arrows point to various sections that are processed from the primary transcript. Inverted triangles denote size variable regions referred to in the text.

human rDNA since until recently it was the only vertebrate gene other than that of Xenopus whose complete sequence was known and on which most genetic studies have been performed. rDNA from other organisms will be compared and contrasted to human to illuminate conserved functional regions and point out important differences. What follows will serve as an introduction to the organization of human rDNA and will provide a foundation for a detailed discussion of the literature that will be presented later. The ribosomal gene is tandemly repeated in a head to tail fashion on the short arms (p-arms) of the five human acrocentric chromosomes (13, 14, 15, 21, and 22). These regions are commonly referred to as the nucleolar organizing regions (NORs) and are viewed cytogenetically as the secondary constrictions of the acrocentric chromosomes. Various techniques have established that there are between 200 to 400 copies per diploid cell or about 20 to 40 copies per chromosome, although unequal crossovers or uncommon amplification events can change copy number 2-3 fold in the cell or 10-fold within a chromosome without a phenotypic effect. The average size of the gene is 43-44 kb which equates to 15 microns and suggests a very compact secondary constriction. The gene repeat (see Fig. 1) includes a 13 kb transcribed region coding for a 45-47S rRNA precursor. The first nucleotide (+1, GenBank accession U13369) of this transcript begins the 5' external transcribed spacer (5’ETS) followed by the 18S coding region, the internal transcribed spacer I (ITSI), the 5.8S coding region, the internal transcribed spacer II (ITSII), the 28S coding region, and finally the 3' ETS. This precursor is processed to mature 18S, 5.8S, and 28S rRNAs for ribosome biogenesis. The coding region alternates with a 31-kb intergenic spacer (IGS) that contains the polymerase I specific rDNA gene promoter, a spacer promoter(s), enhancer repeats, multiple transposons, and many other sequence motifs (see below). Historically, four EcoRI sites have been used to define the A, B, C, and D fragments; subscript letters refer to other restriction enzyme sites within the EcoRI fragment that further subdivide the gene into workable (clonable) fragments (see Fig. 1). The telomere end junction (distal junction) between nonribosomal sequences and the beginning of the first repeat has been localized in human rDNA to be 3969 nucleotides (-3969) upstream of the transcription start site;3 the orientation of transcription is in the telomere to centromere direction.4 The

60

The Nucleolus

centromere end junction (proximal junction) was reported to be in an ITS1 region.5 The junctions appear to be identical for all arrays (see Fig. 1). The reader is directed to early papers and references therein describing rDNA molecular organization in various vertebrates: Xenopus,6-8 mouse,9,10 rat,11-13 and human.14 The number of rDNA-bearing chromosomes is highly variable among closely-related species; for instance, rDNA is found on 5 chromosomes in human, 2 in gorilla, 9 in orangutan, and only one in gibbon. This number can even vary between geographical isolates of the same species. Furthermore, rDNA is not always on acrocentrics, there are some primates with rDNA at a metacentric location. Regardless of location or number, all rDNAs undergo concerted evolution. It has been shown that if the amount of rDNA is doubled as in the case of a 14p+ chromosome, the relative amount of rRNA produced does not change (at least in cultured lymphoblasts).15 It is also known that normal development can occur in individuals who have lost up to 20% of their rDNA due to a Robertsonian translocation that eliminates the rDNA from two acrocentric chromosomes. Manipulating rDNA gene dosage and studying developmental ramifications is extremely difficult, although studies in chicken embryos have yielded interesting results. Chicken lines which have a large (P) and a small (p) nucleolus due to rDNA array size differences have been established and bred.16 Heterozygous +/p1 embryos contain about 66% of a normal (+/+) rDNA complement while p1/p1 contain only about 45%; this corresponds to approximately 290, 192, and 129 rDNA genes respectively.17 Using these and other established lines, it has been shown that at very early developmental times all lines produced adequate and somewhat equal amounts of rRNA. However, the rDNA-deficient embryo could only synthesize 58% of normal amounts of rRNA when demand was increased at the early gastrulation stage and thus development was arrested at this stage (50% lethality at 6 hours).18 Further study of commercial breeder lines indicates that there can be vast differences in rDNA copy number, the size of nucleoli, and the lengths of the IGS resulting in rDNA repeats from 11 kb to 50 kb.19-21 This raises the question of whether these inherited variations are selected for as these chickens are bred for commercial use; similar studies in other commercial animals such as pig, cow, etc., may help answer this question.

rDNA Gene Structure and Variation Transcribed Region Comparative sequencing of ribosomal genes from many organisms has revealed the basis for the size differences in the small (16S - 18S) and the large (23S - 28S) rRNAs. The rRNAs have a mosaic structure in that highly conserved “C” regions alternate with variable “V” regions (also called divergent, “D”, or expansion segments, ”ES”) (Fig. 2; (Table 1); and reviewed by Gerbi.22). The conserved regions contribute directly to catalytic function during translation,23 and are invariant in length and RNA secondary structure but may vary in nucleotide sequence.24 The variable regions vary considerably in size, sequence and structure and have no known function. Comparison of the 18S (1870 bp) coding sequences derived from mouse,25 rat,26 human27,28 and Xenopus (1826 bp) with those obtained from yeast (1650 bp) and a collection of prokaryotes, reveals the insertion/expansion of 10 variable regions in the vertebrate sequences (see Fig. 2). These comparisons together with biochemical probing of the rRNAs and compensatory mutation analysis have permitted the modeling not only of the secondary structure for the conserved regions which diverge only 1.5% between human and Xenopus and 0.1% between human and mouse, but also for the variable regions that, on average, diverge 6.4% and 1.3% respectively. Similar analysis of the 28S coding region of mouse (4712 bp),29 rat (4802 bp),30,31 human (5025 bp),32 pufferfish (4252 bp),33 Xenopus (4115 bp),34 yeast (3392 bp),35 and E. coli (2900 bp) reveals a great diversity in the size of this subunit rRNA and that this diversity is accounted for by the expansion and contraction of variable regions. See (Table 1). For example,

Structure and Organization of Vertebrate Ribosomal DNA

61

Figure 2. Structural conservation in SSU-rRNA (from Raue, et.al., 1988, with permission). The model is based on the structure of E. coli 16S rRNA. Regions conserved in all classes of ribosomes are indicated by the filled circles. Open circles indicate regions conserved in all classes except the mitochondrial one. Variable regions are depicted schematically and numbered from the 5'-end. Highly conserved nucleotides (present in >90% of all known SSU-rRNA sequences) are specifically indicated.

in E. coli13.5% of the 23S rRNA species is accounted for by variable regions compared to over 50% in human. As is the case for 18S, the conserved regions of vertebrate 28S rDNA form a shared structure while the “new”, GC-rich spacer-like variable sequences form self-contained highly structured domains that are presumably tolerated because they do not interrupt the structure/function relationship of the conserved domains of the molecule.36,37 Remarkably, the general secondary structures of both the small and large rRNA molecules have been conserved in the rRNAs of all eubacterial, archaebacterial, and eukaryotic organisms.24

The Nucleolus

62

Table 1. Location and size of variable regions in LSU-rRNA Number of Nucleotides Prokaryotes Region* V1 V2 (D1) V3 (D2) V4 (D3) V5 (D4) V6 (D5) V7 V8 (D6) V9 (D7a) V10 (D7b) V11 V12 V13 (D8) V14 V15 (D9) V16 (D10) V17 (D11) V18 (D12)

Coordinates

E. coli Eubacteria Chloro- Archaeplasts bacteria

131-176 271-365 533-560 636-655 845-847 931-933 1020-1029 1164-1185 1371-1373 1416-1419 1521-1542 1579-1586 1707-1751 1835-1905 2127-2161 2197-2226 2626-2629 2789-2810

46 95 28 20 3 3 10 22 3 4 22 8 45 71 35 30 4 22

31-48 31-33 23-53 90-126 121-140 128-156 25-28 24-28 79-86 19-20 19-21 44-56 2-4 1-7 2-3 2-4 5-20 3-6 10 10 8-15 15-17 16 (235§) 18-27 3 2-3 1-5 5-11 4-5 6-19 20-25 21-87 9-21 6-10 10-11 0-11 19-42 18-20 22-37 71 75-78 56 35 31-35 35-39 28-30 31-33 17-24 4 4 4-6 18-21 22 4-27

Eukaryotes Invertebrates 60-74 150-186 213-255 73-97 7-13 39-48 14-19 26-62 30-74 23-40 22-23 11-22 173-234 55 8-12 82-260 2-116 138-181

Vertebrates Plants‡ 62 163-260 527-873 107-130 13-30 36-42 16 43-197 37-38 40-79 23 14 352-745 55 28-61 89-99 3 154-226

64 150 222 68 10 40 18 27 24 21 23 11 168 55 10 84 1 126

*The D nomenclature given in parentheses refers to that used by Michot et al. (1984). †Coordinates are according to the E. coli 23S rRNA sequence. ‡Only one plant LSU-rRNA (rice) has so far been sequenced in its entirety. §This value applies if the sequence in question is not an intron (see text). ¶Only in C. ellipsoidae where no separate 4.5S rRNA is present. Reproduced from Raue, et al, 198824

Compensatory mutations and other nucleotide changes can also serve as data for long term evolutionary studies (see below). In contrast, the variable regions show a much higher rate of evolutionary change and because their sequences remain species specific, these segments can be used to study short-term evolution. They can also be used to measure gene variation within a given species and they are used extensively for purposes of taxonomic identification. Databases of rRNA sequences, alignments, phylogenies, and taxonomies are available on the World Wide Web.38,39,40 The presence of large variable regions in the ribosomal RNAs suggests the possibility of intra-species variability in the sequence of these regions. To address this possibility in human, sequence comparisons were made between 6 human clones using a 1429 bp BamHI restriction fragment of the 28S rRNA (see Fig. 1). This section contains two short and one longer variable region (V3, V4, and V5 corresponding to V6, V7, and V8 in Raue24) flanked by two conserved regions. Three of the six clones differed in the number of simple sequence repeats in V5 revealing a high degree of intraspecies sequence variation.32 Sequencing of 111 V8 segments from 9 human tissues and cell lines showed two hot spots resulting in 35 distinct variants.41 Subsequent characterization of the V5 regions within the rDNA of a single individual revealed variation at 4 specific sites and suggested that 32 variant sequences could theoretically exist in just V5 alone.42 Furthermore, rRNA expression was analyzed by RNase protection in many different tissues from a single individual (corresponding to a set of 10 separate rDNA clusters). It

Structure and Organization of Vertebrate Ribosomal DNA

63

was concluded that the relative contribution of each variant rRNA appeared to be the same in each of the tissues studied and that gene dosage and not regulated gene expression accounted for this pattern of gene expression.42 Since these variations have been found in rRNAs extracted from ribosomes and polyribosomes, it does not appear that they cause the inactivation of the translation complex.43 These analyses document that a huge amount of sequence variation exists in rDNA and thus the rRNA population is very heterogeneous. Although the true extent of intra-species and intra-individual variation is still largely unknown, it appears possible that no two rDNA transcribed sections are identical if one considers all the possible combinations and permutations. It remains to be determined how this diversity is generated under pressures to homogenize the gene family; and if there are novel functions for this variation. One way to look for putative variable region function is to assume that variable regions specifically interact with protein factors and design screens for these interactions. In such a screen using a human V5 RNA as probe, several known RNA binding proteins were identified along with numerous proteins of unknown function (Mougey and Sylvester, unpublished). The specificity and significance of these interactions have yet to be determined. The various spacers of the transcribed region account for almost half of its total length and they, along with the variable (expansion) segments are responsible for the notoriously high GC content of the gene in mammals. In fact, it has been speculated that there is a common evolutionary origin for these GC-rich segments.22,44 Although the sequences diverge quite rapidly among the species, the overall predicted secondary structures are fairly conserved in the mammals studied45-48 and show a remarkable resemblance to the structures seen in early electron micrographs of the 13 kb rRNA precursor that escaped denaturation due to their high G+C content.49 Variation within human transcribed spacer regions is largely due to alterations in the number of short simple repeat motifs or length variations in homopolymer tracts, both probably caused by slippage during DNA replication.50,51 Lastly, it should be mentioned that alternative 28S rDNA molecular structures do exist. For instance, in South American rodents of the genus Ctenomys, a 106-bp intron is located in the D629 divergent region.52 This intron is flanked by 9 bp direct repeats suggesting that it arose by insertion. In most tissues, the intron is removed post-transcriptionally via an unknown mechanism such that a 2600 base and an 1800 base rRNA product interact through extensive base pairing. Only in testis is the intron spliced-out such that a 4.4 kb product is produced.52 Whether other types of novel organization exist elsewhere and whether they offer a selective advantage, remains to be determined.

Intergenic Spacer (IGS) The majority of the early publications concerning the IGS (or nontranscribed spacer as it was called) involved identifying and characterizing the promoter and terminator regions and the protein factors that bind to them. Most systems were also found to harbor a spacer promoter(s), for example, in rat53 and in human (Mougey and Sylvester unpublished). They also contain a variable number of spacer repeats: 60/81 bp in Xenopus,54 135 bp in mouse,55-58 and rat,59,60 that have been shown to act as transcription enhancer elements in conjunction with the spacer promoter. Inbred strains of mice were found to carry 4-6 unique classes of this size-variable (135 bp)n region upstream of the transcription start site and individual chromosomes were found to carry specific classes implying little exchange between nonhomologous chromosomes.55 In the human, 72-93 bp repeats upstream of the promoter are found covering as much as 1000 bp or more.14 Although they have not yet been shown to function as enhancers, they consist of primarily (CTTT)n, which has the potential to form alternative (triple stranded) structures which could be involved in gene regulation. See chapter 7 by Rothblum for a detailed discussion of promoter/enhancer function. The variable number of these repeats in each system contributes to size variation in the IGS; in addition the repeat size can fluctuate due to slip-strand mispairing of internal sub-repeats during replication. Early hybridization studies indicated that sequences located within the human IGS upstream of the promoter and the region downstream of transcription termination contain

64

The Nucleolus

sequences that are found very frequently elsewhere in the genome. Eventually, these sequences were identified not to be intrinsically ribosomal but as retroposed elements known as SINEs (Alu, Alu2, B2, ID, etc.).59,61-64 A more detailed sequence comparison of 2.7 kb within three independent human clones that contain the first 4 Alu elements upstream of the promoter revealed that each Alu element was independently inserted into an A-rich section.65 In addition, it was found that equivalently positioned Alu elements are between 97-99.6% identical, indicating their concerted evolution, while only 73-80% identity exists between nonequivalently positioned Alus. Furthermore, detailed restriction analysis of genomic DNA does not show any evidence of unequal homologous exchanges between these nonhomologous Alu elements.65 A second publication described the results of Southern blots of various primate DNAs that were probed with rDNA-derived Alus 1 – 4, 6, 8, 9 and with the retroposed pseudogene derived from the cdc27hs gene (Fig. 3).66 It was found that this pseudogene is present in apes but not in Old World monkeys, indicating fixation in an ape ancestor. Five of the Alu elements are shared by the whole set of primates studied, indicating insertion and fixation prior to the split of apes and Old World monkeys. One Alu element is absent only from rhesus, while another is absent from both rhesus and gibbon.66 This supports the notion of waves of retroposition events occurring throughout evolutionary history and that parts of the IGS may have a junkyard character. Length variation in human rDNA (see Fig 1) was first identified via electron microscopic studies67 (and more recently by high resolution FISH on stretched DNA fibers.68) Subsequently, Southern blot studies using the 3' end of the 28S gene as probe were performed that identified a major size-variable region in rDNA that is located just downstream of the primary transcription termination site.63,64,69,70 Human genomic DNA contains 4 major BamHI fragment variants of 3.9 kb, 4.6 kb, 5.4 kb, and 6.2 kb in this region; the longest and shortest possibly variably present in an individual.63 Restriction and sequence analysis of human genomic clones revealed a 700 bp sequence that was tandemly repeated 1 to 5 times. The sequence contains a large number of simple sequences, stretches of poly-pyrimidine tracts (on the sense strand), and adjacent Alu elements (human)63,64 and B2 elements (rat).59 Analysis of this size-variable region allowed studies of the dynamics of inheritance of rDNA spacer length-variants within families. For example, a study of human rDNA of 51 individuals documented 8 structural variants of which 2 were common to all individuals and 6 appeared in different combinations and frequencies.71 Further analysis provided evidence that some of the structural variants were inherited as a single Mendelian locus suggesting that they formed clusters on single chromosomes.71 The size variation here was presumed to arise from unequal crossovers between homologous sequence repeats during meiosis. This event can be recapitulated during growth in bacteria of recombinant lambda virus clones that contain this size variable region from human rDNA: single cloned variants undergo recombination to generate smaller and larger variants.69 Since the variation occurs on multiple acrocentric chromosomes in humans it was taken as evidence that crossovers occur among nonhomologous chromosomes during meiosis although it could also be due to sister-chromatid exchanges or even intra-molecular looping. However, the variation appeared to be stable during mitosis and through multiple generations in cell culture.69 The inheritance pattern of this length variation was also studied in 121 individuals in multigenerational families.72 No direct evidence of recombination was detected indicating that all the patterns could be explained by normal meiotic segregation. However, other work has shown that children can inherit more copies of a variant than were present in the parents, suggesting that unequal exchange is indeed operative.73,74 A more recent publication that is discussed below examines the frequency of intra- versus inter-chromosomal exchanges.75 The polymerase I terminator element consists of an 18 bp (mouse)76 or 10 bp (human)77 conserved sequence motif, termed a Sal box due to the presence of an internal Sal I restriction site, and flanking pyrimidine rich sequences. Termination occurs 360 bp (human) and 565 bp (mouse) downstream of the 3’end of the mature 28S rRNA (3’ETS). The IGS immediately following this contains a variable number of terminator elements described above as the 700 bp

Structure and Organization of Vertebrate Ribosomal DNA

Figure 3. (Top) Base composition of the RNA-like strand of the entire rDNA repeat, averaged over an interval of 100 nucleotides. (Bottom) Detailed map, including locations of repeated IGS segments (R, LR, and 90-bp repeats), retroposons (Alu elements and pseudo-cdc-27 gene), and simple sequence repeats. Reproduced from Gonzalez and Sylvester, 1995, with permission).

65

66

The Nucleolus

IGS sequences, each of which contains a Sal box and pyrimidine-rich sequence. Most transcribing polymerases terminate at the first Sal box, however the remaining Sal boxes can serve to terminate polymerases that have managed to read-through previous terminators.78 Xenopus in contrast, does not contain a functional terminator at the 3' end of the 28S region.79 Instead, a single terminator is found 213 bp 5' of the pre-rRNA initiation site.79 Therefore polymerase I reads through the entire Xenopus spacer and this function is shown to enhance transcription from the downstream gene promoter.80 While lacking a terminator, this region of the Xenopus spacer does however contain two types of regulatory sequence arrays designated as regions ‘0’ (34 bp region) and region ‘1’ (101 bp).54 Transcription studies have found that these regions function as enhancers, although not as strongly as the 60/81 bp repetitive elements.81 A more detailed molecular mechanism for the termination event is described by Rothblum in Chapter 7. The complete sequence of the 43 kb human rDNA gene (GenBank Accession No. U13369) was reported in 1995.82 Although the Xenopus sequence and now the pufferfish rDNA sequence33,83 are available, comparisons among vertebrates can only be made for the transcribed section and not the IGS. A more detailed evolutionary comparison and analysis awaits the completion of the mouse genome, chimpanzee genome, and more distantly related primate spacer sequences. The molecular organization of human rDNA is shown in (Fig. 3). The preceding discussion has covered structural details with some reference to function for the transcription unit and the IGS regions that flank it. At this point a short discussion of the remaining (human) sequence motifs and a third size-variable region is presented. A long, 4.5 kb tandem repeat deep in the nontranscribed spacer was identified from cloned DNA and termed an “Xba element” even though it contained Alu elements.78 It was speculated that this region might house the origin of replication for rDNA and form the two ends of a recombinogenic rDNA unit. Analysis of the complete rDNA sequence82 mapped this DNA (“long repeats” LR1 and LR2, (Fig. 3)) to nucleotides 20,569-22,625 and 23,035-25,043 and showed that they were only 2 kb long. In addition, they had partial similarity to ESTs in the database that had 58-68% similarity to LINE elements, and thus could be retroposed; and that some rDNA genes contained three of these LRs. Other repeated motifs in the IGS identified by sequencing include a 320 bp fragment, three tandemly arranged 84 bp pyrimidine rich blocks, other 58 or 52 bp pyrimidine rich repeats, and assorted microsatellites.82 Alu element sequences contribute 5474 bp to the IGS and the pseudo-cdc27 gene (see above) contributes another 2327 bp, which combined account for 18% of total rDNA. Now that the human rDNA sequence is known, one can determine the exact position of various sequence motifs and correlate them with previously reported biochemical and biological characteristics in human, mouse, or rat systems. Please see (Fig. 3) for the following discussion. The promoter portion of rDNA was shown to be the most highly enriched region of rDNA in HeLa cell nuclear/nucleolar matrices by Southern blot hybridizations after restriction enzyme digestions of salt washed preparations and its enrichment was dependent upon active transcription.84 S1 nuclease sensitive sites were detected within polypyrimidine:polypurine sequences found in cloned IGS sequences approximately 1.2 kb downstream of 28S in human and within 4 kb up and downstream of coding sequences in rat rDNA that could be involved in transcription regulation or chromatin structure.85 The S1 nuclease sensitivity indicates that these regions form alternative structures such as triple strand DNA. Other work in human (Maguire and Sylvester, unpublished) shows that up to 1000 bases of polypyrimidine:polypurine sequences, (CTTT)n upstream of transcription initiation, are also S1 nuclease sensitive and that these sequences can bind nucleolin (shown by screening expression libraries with oligomers from this region). Rat rDNA was surveyed for nucleolar matrix attachment sequences, which were localized to specific restriction enzyme fragments containing polypyrimidine:polypurine tracts of the IGS that are within a few kb of the coding regions.86 The authors suggested that the attachment sites would permit the formation of loops that would contain the transcribed section and that the next step would be to identify the proteins involved. Potential Z-DNA forming sequences, poly (dG-dT):poly (dA-dC), that are resistant to exhaustive treatment by

Structure and Organization of Vertebrate Ribosomal DNA

67

DNase I, were found to reside in regions flanking the transcribed unit in mouse rDNA87 and in human rDNA.82 These were also proposed to participate in structuring the nucleolus perhaps by forming loops containing the transcription unit.87 To continue this perspective, the reader is directed to the discussion of transcription foci in the chapter on Nucleolar Ultrastructure by W. Mosgoeller and the chapter on rDNA Chromatin by Sogo and Thoma. In addition to the polypyrimidine:polypurine sequence motifs that have the potential to form alternative, triple-strand structures and alternating pyrimidine-purine sequences that can form Z-DNA structures, other motifs identified in the human gene include G-rich motifs that can form four-stranded G-quartets, and A+T-rich motifs with a 10 bp periodicity that can adopt a bent DNA structure.82 These motifs have been implicated in recombination, replication, and scaffold attachment in non-rDNA experimental systems. It seems likely that they could act in a similar manner for rDNA.88 It is becoming generally accepted that methylation of rDNA is associated with transcriptionally inactive rDNA genes. For the most part, the transcription unit itself is unmethylated under most circumstances, as is the region surrounding the promoter (see below). It is the methylation status of the upstream flanking region that contains the enhancer elements that appears to dictate gene activity (see earlier discussion of enhancers in rDNA.) There is experimental support for the general association of methylation status and transcription of rRNA. For example, cells with amplified rDNA do not increase rRNA production but this rDNA does bind 5-Me-C antibody.15 Also, reversal of methylation status with 5-aza-C incorporation increases rRNA production.89,90 Inactive NORs, as determined by lack of silver staining, are resistant to the methylation- sensitive restriction enzyme HpaII (CmpG) and will not be visualized by DNase I-directed in situ nick translation; whereas active AgNORs are digested by HpaII.91 It has been reported that inactive rDNA, but not active rDNA, can be crosslinked to histones with formaldehyde. Interestingly, it has been shown that the methylation status of a CpG site at –145 in rat rDNA might be diagnostic for gene activity.92 Subsequently, it was shown that methylation at nucleotide -133 in mouse rDNA prevented the binding of UBF.93 These experiments suggest that silencing of gene transcription may indeed occur by methylation of a promoter nucleotide(s) while propagation of gene inactivation requires methylation of the upstream IGS region.92,93 The situation concerning methylation, gene activity and nucleolar formation becomes more complex when one considers that silent human NORs can be found in essentially equal numbers with active (dominant) mouse NORs in nucleoli of hybrid cell lines.94 More interestingly, mutations in the protein ATRX (X-link α-thalassaemia mental retardation syndrome), a protein shown to localize to NORs and centromeres, alters methylation pattern, chromatin structure, and gene activity of some repeated genes, potentially providing a link among those activities.95 Changes in copy number and/or methylation patterns of rDNA in relation to aging have been studied in many systems over the last 25 years. An age related loss of hybridizable rDNA was reported for human brain,96 human heart,97 and various mouse tissues.98 An age-related increase in methylation of rDNA in mouse liver, brain, and spleen, mostly in the 5' spacer domain, accompanied inactivation of specific NORs as measured by Ag-staining.99 An individual specific age-related rate of decline of rRNA gene activity based on the number of Ag-stained NORs has been observed for human fibroblasts.100 Another study concerning the associations among aging, rDNA methylation, and rRNA expression was carried out with Werner syndrome (WS) (premature aging) cells. Cultures of WS fibroblasts obtained from donors aged 29 and 59 were compared to fibroblasts obtained from 1 day, 2 day, 84 year and 89 year old normal individuals. It was found that WS fibroblasts grew slowly and reached senescence after fewer cell doublings. The rDNA copy number did not change significantly for any cell during long-term growth. Although methylation in the coding region of rDNA increased in all cells as they progressed to senescence, it was more pronounced for WS cells, but steady-state levels of rRNA did not change appreciably.101

68

The Nucleolus

5S rRNA Genes The small, highly conserved 120 base 5S rRNA is encoded by a different multigene family than that coding for the 18S, 5.8S, and 28S rRNAs in vertebrates and, resides at a separate chromosome locus. The human 5S gene family consists of a 2331 bp unit that is tandemly repeated about 100 or so times on chromosome 1 band q42.102 A smaller 1.6 kb version of the gene has been identified which has a large deletion starting 12 nucleotides 3' from the coding region; there are about 5-10 copies per haploid genome.103 There also is a large number of variants or pseudogenes for 5S rDNA. The 5S gene is the same size, 2.3 kb, in hamster,104 but is predominantly 1.6 kb in rat105 and 1.6 – 1.7 kb in mouse.106 The genes are between 60%-70% GC, they have a pyrimidine stretch and an alternating pyrimidine/purine stretch of nucleotides that may vary in length. The human sequence has an oppositely oriented (with respect to transcription) Alu element commencing at the poly T termination site.102 There appears to be very little identity between the rodent and human spacer sequences; and although alignment is discontinuous, the mouse and the rat spacers are about 87% identical.105 The sequence variation in the 5S spacer within each species is less than 1%, thus this sequence can serve as an important tool for phylogenetic analysis. A 5S rRNA database website is available online through the World Wide Web at http://biobases.ibch.poznan.pl/5Sdata/.107

Evolutionary and Taxonomic Studies Using rDNA The divergence and variation among rDNA sequences has been exploited for both evolutionary studies and for taxonomic identification. On the evolutionary front, the most highly-conserved regions, especially in the 18S rRNA, have been used for deriving phylogenetic trees describing the deep branching of the basic divisions of living things, also of major groupings within phyla, and even for resolving the branching of species within orders. Variable region sequences (ITS 1 and V regions of 28S) have been used to establish phylogenies for closely-related organisms; the variation was sufficient to resolve the branching order of the higher primates.51 Detection of infectious agents in patient and environmental samples and their taxonomic identification often uses 18S or ITS1 and ITS2 sequences, as they are relatively easy to amplify by RT-PCR or PCR and resolve by either size or sequence analysis. Although the hundreds of rDNA copies contain intra-species variation, it was also shown that concerted evolution could largely homogenize the arrays present on both homologous and nonhomologous chromosomes.61,108 However, other work has shown that specific length-variants appear to be restricted to single chromosomes and can be inherited in Mendelian fashion.71,74 These findings opened the question of which exchanges are more frequent in concerted evolution: intra-chromosomal, or among homologous chromosomes, or among all rDNA-bearing chromosomes. Studying rDNA derived from single acrocentric chromosomes isolated in somatic cell hybrids addressed this problem. It was shown that all rDNA arrays have highly-conserved coding regions and adjacent regulatory regions, and remarkably-uniform nonribosomal sequences on the telomeric side of the array (for at least 8 kb) indicating frequent exchanges/correction of these regions on all chromosomes.75 However, the IGS contains segments that can be so divergent as to form “classes” and these classes do not localize to specific pairs of homologous chromosomes; further, a single rDNA array can contain 1 to 3 different classes. The chromosome distribution of these divergent IGS segments suggests that rDNA copies containing specific segment classes form sub-arrays, and indicates more frequent localized intra-array and less frequent inter-chromosomal homogenization of the IGS.75 In short, different parts of the rDNA homogenize at different rates and probably by different mechanisms.

Concluding Remarks We have attempted to provide a survey of the literature of vertebrate ribosomal genes by linking the results of the classic papers to the more recently performed molecular genetic studies. The vast extent of sequence heterogeneity of the gene family coupled to the central role rRNA plays in protein translation makes this a fascinating experimental model system. The

Structure and Organization of Vertebrate Ribosomal DNA

69

next few years should yield the results of rDNA sequence analysis from many more organisms and lead to an understanding of many evolutionary questions and the role of various sequence motifs within the gene. This information, when integrated with a more complete knowledge of the structure—function relationships among the components of the ribosome, and knowledge of the ever-expanding role of the nucleolus in the health and welfare of the cell should make the next decade of study very interesting indeed.

References 1. Srivastava AK, Schlessinger D. Structure and organization of ribosomal DNA. Biochimie 1991; 73(6):631-638. 2. Busch H, Rothblum L. The cell nucleus Volume X: rDNA, Part A. New York: Academic Press, Inc., 1982. 3. Sylvester JE, Petersen R, Schmickel RD. Human ribosomal DNA: novel sequence organization in a 4.5-kb region upstream of the promoter. Gene 1989; 84:193-196. 4. Worton RG, Sutherland J, Sylvester JE et al. Human ribosomal RNA genes: orientation of the tandem array and conservation of the 5' end. Science 1988; 239:69. 5. Sakai K, Ohta T, Minoshima S et al. Human ribosomal RNA gene cluster: identification of the proximal end containing a novel tandem repeat sequence. Genomics 1995; 26(3):521-526. 6. Wellauer PK, Dawid IB, Brown DD et al. The molecular basis for length heterogeneity in ribosomal DNA from Xenopus laevis. J Mol Biol 1976; 105:461-486. 7. Wellauer PK, Reeder RH, Dawid IB et al. Arrangement of length heterogeneity in repeating units of amplified and chromosomal ribosomal DNA from Xenopus laevis. J Mol Biol 1976; 105(4):487-505. 8. Boseley P, Moss T, Machler M et al. Sequence organization of the spacer DNA in a ribosomal gene unit of Xenopus laevis. Cell 1979; 17(1):19-31. 9. Grummt I, Gross HJ. Structural organization of mouse rDNA: comparison of transcribed and nontranscribed regions. Mol Gen Genet 1980; 177(2):223-229. 10. Kominami R, Urano Y, Mishima Y et al. Organization of ribosomal RNA gene repeats of the mouse. Nucleic Acids Res 1981; 9(14):3219-3233. 11. Chikaraishi DM, Buchanan L, Danna KJ et al. Genomic organization of rat rDNA. Nucleic Acids Res 1983; 11(18):6437-6452. 12. Mroczka DL, Cassidy B, Busch H et al. Characterization of rat ribosomal DNA. The highly repetitive sequences that flank the ribosomal RNA transcription unit are homologous and contain RNA polymerase III transcription initiation sites. J Mol Biol 1984; 174(1):141-162. 13. Yang-Yen HF, Subrahmanyam CS, Cassidy B et al. Characterization of rat ribosomal DNA II. identification of the highly repetitive DNA in the 3' nontranscribed spacer. J Mol Biol 1985; 184(3):389-398. 14. Sylvester JE, Whiteman DA, Podolsky et al. The human ribosomal RNA genes: structure and organization of the complete repeating unit. Hum Genet 1986; 73:193-198. 15. Tantravahi U, Breg WR, Wertelecki V et al. Evidence for methylation of inactive human rRNA genes in amplified regions. Hum Genet 1981; 56(3):315-320. 16. Delany ME, Muscarella DE, Bloom SE. Formation of nucleolar polymorphisms in trisomic chickens and subsequent microevolution of rRNA gene clusters in diploids. J Hered 1991; 82(3):213-220. 17. Delany ME, Muscarella DE, Bloom SE. Effects of rRNA gene copy number and nucleolar variation on early development: inhibition of gastrulation in rDNA-deficient chick embryos. J Hered 1994; 85(3):211-217. 18. Delany ME, Taylor Jr RL, Bloom SE. Teratogenic development in chicken embryos associated with a major deletion in the rRNA gene cluster. Develop Growth Differ 1995; 37:403-412. 19. Su MH, Delany ME. Ribosomal RNA gene copy number and nucleolar-size polymorphisms within and among chicken lines selected for enhanced growth. Poult Sci 1998; 77(12):1748-1754. 20. Delany ME, Krupkin AB. Molecular characterization of ribosomal gene variation within and among NORs segregating in specialized populations of chicken. Genome 1999; 42(1):60-71. 21. Delany ME. Patterns of ribosomal gene variation in elite commercial chicken pure line populations. Anim Genet 2000; 31(2):110-116. 22. Gerbi SA. Expansion Segments: Regions of Variable Size that Interrupt the Universal Core Secondary Structure of Ribosomal RNA. CRC Press, Inc., 1996. 23. Noller HF. Structure of ribosomal RNA. An Rev Biochem 1984; 53:119-162. 24. Raue HA, Klootwijk J, Musters W. Evolutionary conservation of structure and function of high molecular weight ribosomal RNA. Prog Biophys Mol Biol 1988; 51(2):77-129. 25. Raynal F, Michot B, Bachellerie JP. Complete nucleotide sequence of mouse 18 S rRNA gene: comparison with other available homologs. FEBS Lett 1984; 167(2):263-268.

70

The Nucleolus

26. Torczynski R, Bollon AP, Fuke M. The complete nucleotide sequence of the rat 18S ribosomal RNA gene and comparison with the respective yeast and frog genes. Nucleic Acids Res 1983; 11(14):4879-4890. 27. Gonzalez IL, Schmickel RD. The human 18S ribosomal RNA gene: Evolution and stability. Am J Human Genet 1986; 38:419-427. 28. McCallum FS, Maden BE. Human 18 S ribosomal RNA sequence inferred from DNA sequence. Variations in 18 S sequences and secondary modification patterns between vertebrates. Biochem J 1985; 232(3):725-733. 29. Hassouna N, Michot B, Bachellerie JP. The complete nucleotide sequence of mouse 28S rRNA gene. Implications for the process of size increase of the large subunit rRNA in higher eukaryotes. Nucleic Acids Res 1984; 12(8):3563-3583. 30. Hadjiolov AA, Georgiev OI, Nosikov VV et al. Primary and secondary structure of rat 28 S ribosomal RNA. Nucleic Acids Res 1984; 12(8):3677-3693. 31. Chan YL, Gutell R, Noller HF et al. The nucleotide sequence of a rat 18 S ribosomal ribonucleic acid gene and a proposal for the secondary structure of 18 S ribosomal ribonucleic acid. J Biol Chem 1984; 259(1):224-230. 32. Gonzalez IL, Gorski JL, Campen TJ et al. Variation among human 28S ribosomal RNA genes. Proc Natl Acad Sci USA 1985; 82(22):7666-7670. 33. Crollius HR, Jaillon O, Dasilva C et al. Characterization and repeat analysis of the compact genome of the freshwater pufferfish Tetraodon nigroviridis. Genome Res 2000; 10(7):939-949. 34. Labhart P, Reeder RH. DNA sequences for typical ribosomal gene spacers from Xenopus laevis and Xenopus borealis. Nucleic Acids Res 1987; 15(8):3623-3624. 35. Ajuh PM, Heeney PA, Maden BE. Xenopus borealis and Xenopus laevis 28S ribosomal DNA and the complete 40S ribosomal precursor RNA coding units of both species. Proc R Soc Lond B Biol Sci 1991; 245(1312):65-71. 36. Gorski JL, Gonzalez IL, Schmickel RD. The secondary structure of human 28S rRNA: The structure and evolution of a mosaic rRNA gene. J Mol Evol 1987; 24:236-251. 37. Nissen P, Hansen J, Ban N et al. The structural basis of ribosome activity in peptide bond synthesis. Science 2000; 289(5481):920-930. 38. Maidak BL, Cole JR, Lilburn TG et al. The RDP-II (Ribosomal Database Project). Nucleic Acids Res 2001; 29(1):173-174. 39. Wuyts J, De Rijk P, Van de PY et al. The European Large Subunit Ribosomal RNA Database. Nucleic Acids Res 2001; 29(1):175-177. 40. Wuyts J, Van de PY, Winkelmans T et al. The European database on small subunit ribosomal RNA. Nucleic Acids Res 2002; 30(1):183-185. 41. Leffers H, Andersen AH. The sequence of 28S ribosomal RNA varies within and between human cell lines. Nucleic Acids Research 1993; 21(6):1449-1455. 42. Kuo BA, Gonzalez IL, Gillespie DA et al. Human ribosomal RNA variants from a single individual and their expression in different tissues. Nucleic Acids Research 1996; 24:4817-4824. 43. Gonzalez I, Sylvester JE, Schmickel RD. Human 28S ribosomal RNA sequence heterogeneity. Nucl Acids Res 1988; 16:10313-10224. 44. Gray MW, Schnare MN. Evolution of rRNA Gene Organization. CRC Press, Inc., 1996. 45. Goldman WE, Goldberg G, Bowman LH et al. Mouse rDNA: sequences and evolutionary analysis of spacer and mature RNA regions. Mol Cell Biol 1983; 3(8):1488-1500. 46. Renalier MH, Mazan S, Joseph N et al. Structure of the 5'-external transcribed spacer of the human ribosomal RNA gene. FEBS Lett 1989; 249(2):279-284. 47. Bourbon H, Michot B, Hassouna N et al. Sequence and secondary structure of the 5' external transcribed spacer of mouse pre-rRNA. DNA 1988; 7(3):181-191. 48. Gonzalez I, Chambers C, Gorski JL et al. Sequence and structure correlation of human ribosomal transcribed spacers. J Mol Biol 1990; 212:27-35. 49. Wellauer PK, Dawid IB. Secondary structure maps of RNA: processing of HeLa ribosomal RNA. Proc Natl Acad Sci USA 1973; 70(10):2827-2831. 50. Maden BE, Dent CL, Farrell TE. Clones of human ribosomal DNA containing the complete 18 S-rRNA and 28 S-rRNA genes. Characterization, a detailed map of the human ribosomal transcription unit and diversity among clones. Biochem J 1987; 246(2):519-527. 51. Gonzalez IL, Sylvester JE, Smith TF et al. Ribosomal RNA gene sequences and hominoid phylogney. Mol Biol Evol 1990; 7:203-219. 52. Melen GJ, Pesce CG, Rossi MS et al. Novel processing in a mammalian nuclear 28S prerRNA: tissue-specific elimination of an ‘intron’ bearing a hidden break site. EMBO J 1999; 18(11):3107-3118. 53. Cassidy BG, Yang-Yen HF, Rothblum LI. Transcriptional role for the nontranscribed spacer of rat ribosomal DNA. Mol Cell Biol 1986; 6(8):2766-2773.

Structure and Organization of Vertebrate Ribosomal DNA

71

54. Moss T, Boseley PG, Birnstiel ML. More ribosomal spacer sequences from Xenopus laevis. Nucleic Acids Res 1980; 8(3):467-485. 55. Arnheim N, Treco D, Taylor B et al. Distribution of ribosomal gene length variants among mouse chromosomes. Proc Natl Acad Sci USA 1982; 79(15):4677-4680. 56. Kuehn M, Arnheim N. Nucleotide sequence of the genetically labile repeated elements 5' to the origin of mouse rRNA transcription. Nucleic Acids Res 1983; 11(1):211-224. 57. Kuhn A, Deppert U, Grummt I. A 140-base-pair repetitive element in the mouse rRNA gene spacer enhances transcription by RAN polymerase I in a cell-free system. Proc Natl Acad Sci USA 1990; 87:7527-7531. 58. Pikaard CS, Pape LK, Henderson SL et al. Enhancers for RNA polymerase I in mouse ribosomal DNA. Mol Cell Biol 1990; 10:4816-4825. 59. Yavachev LP, Georgiev OI, Braga EA et al. Nucleotide sequence analysis of the spacer regions flanking the rat rRNA transcription unit and identification of repetitive elements. Nucleic Acids Res 1986; 14(6):2799-2810. 60. Dixit A, Garg LC, Chao W et al. An enhancer element in the far upstream spacer region of rate ribosomal RNA gene. J Biol Chem 1987; 262:11616-11662. 61. Arnheim N, Krystal M, Schmickel R et al. Molecular evidence for genetic exchanges among ribosomal genes on nonhomologous chromosomes in man and apes. Proc Natl Acad Sci USA 1980; 77:7323-7327. 62. Higuchi R, Strang HD, Browne JK et al. Human ribosomal RNA spacer sequences are found interspersed elsewhere in the genome. Gene 1981; 15:177-186. 63. LaVolpe A, Simeone A, D’Esposito M et al. Molecular analysis of the heterogeneity region of the human ribosomal spacer. J Mol Biol 1985; 183:213-223. 64. Dickson KR, Braaten DC, Schlessinger D. Human ribosomal DNA: conserved sequence elements in a 4.3-kb region downstream from the transcription unit. Gene 1989; 84:197-200. 65. Gonzalez IL, Petersen R, Sylvester JE. Independent insertion of Alu elements in the human ribosomal spacer and their concerted evolution. Mol Biol Evol 1989; 6:413-423. 66. Gonzalez IL, Tugendreich S, Hieter P et al. Fixation times of retroposons in the ribosomal DNA spacer of human and other primates. Genomics 1993; 18:29-36. 67. Wellauer PK, Dawid IB. Isolation and sequence organization of human ribosomal DNA. J Mol Biol 1979; 128(3):289-303. 68. Schöfer C, Weipoltshammer K, Almeder M et al. Arrangement of individual human ribosomal DNA fragments on stretched DNA fibers. Histochem Cell Biol 1998; 110(2):201-205. 69. Erickson JM, Schmickel RD. A molecular basis for discrete size variation in human ribosomal DNA. Am J Hum Genet 1985; 37:311-325. 70. Braga EA, Avdonina TA, Zhurkin VB et al. Structural organization of rat ribosomal RNA genes: interspersed sequences and their putative role in the alignment of nucleosomes. Gene 1985; 36(3):249-262. 71. Garkavtsev IV, Tsvetkova TG, Yegolina NA et al. Variability of human rRNA genes: inheritance and nonrandom chromosomal distribution of structural variants of nontranscribed spacer sequences. Hum Genet 1988; 81(1):31-37. 72. Ranzani GN, Bernini LF, Crippa M. Inheritance of rDNA spacer length variants in man. Mol Gen Genet 1984; 196(1):141-145. 73. Schmickel RD, Gonzalez IL, Erickson JM. Nucleolus organizing genes on chromosome 21: recombination and nondisjunction. Ann N Y Acad Sci 1985; 450:121-131. 74. Kuick R, Asakawa J, Neel JV et al. Studies of the inheritance of human ribosomal DNA variants detected in two-dimensional separations of genomic restriction fragments. Genetics 1996; 144(1):307-316. 75. Gonzalez IL, Sylvester JE. Human rDNA: evolutionary patterns within the genes and tandem arrays derived from multiple chromosomes. Genomics 2001; 73(3):255-263. 76. Grummt I, Rosenbauer H, Niedermeyer I et al. A repeated 18bp sequence motif in the mouse rDNA spacer. Cell 1986; 45:837-846. 77. Bartsch I, Schoneberg C, Grummt I. Evolutionary changes of sequences and factors that direct transcription termination of human and mouse ribosomal genes. Mol Cell Biol 1987; 7:2521-2529. 78. Safrany G, Hidvegi EJ. New tandem repeat region in the nontranscribed spacer of human ribosomal RNA gene. Nucl Acids Res 1989; 17:3013-3022. 79. De Winter RF, Moss T. The ribosomal spacer in Xenopus laevis is transcribed as part of the primary ribosomal RNA. Nucleic Acids Res 1986; 14(15):6041-6051. 80. De Winter RF, Moss T. A complex array of sequences enhances ribosomal transcription in Xenopus Laevis. J Mol Biol 1987; 196:813-827.

72

The Nucleolus

81. Mougey EB, Pape LK, Sollner-Webb B. Virtually the entire Xenopus laevis rDNA multikilobase intergenic spacer serves to stimulate polymerase I transcription. J Biol Chem 1996; 271(43):27138-27145. 82. Gonzalez IL, Sylvester JE. Complete sequence of the 43 Kb human ribosomal DNA repeat: analysis of the intergenic spacer. Genomics 1995; 27:320-328. 83. Aparicio S, Chapman J, Stupka E et al. Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science 2002; 297(5585):1301-1310. 84. Keppel F. Transcribed human ribosomal RNA genes are attached to the nuclear matrix. J Mol Biol 1986; 187(1):15-21. 85. Financsek I, Tora L, Kelemen G et al. Supercoil induced S1 hypersensitive sites in the rat and human ribosomal RNA genes. Nucleic Acids Res 1986; 14(8):3263-3277. 86. Stephanova E, Stancheva R, Avramova Z. Binding of sequences from the 5'- and 3'-nontranscribed spacers of the rat rDNA locus to the nucleolar matrix. Chromosoma 1993; 102(4):287-295. 87. Thomas JR, Bolla RI, Rumbyrt JS et al. DNase I-resistant nontranscribed spacer segments of mouse ribosomal DNA contain poly(dG-dT).poly(dA-dC). Proc Natl Acad Sci USA 1985; 82(22):7595-7598. 88. Langst G, Schatz T, Langowski J et al. Structural analysis of mouse rDNA: coincidence between nuclease hypersensitive sites, DNA curvature and regulatory elements in the intergenic spacer. Nucleic Acids Res 1997; 25(3):511-517. 89. Ferraro M, Lavia P. Activation of human ribosomal genes by 5-azacytidine. Exp Cell Res 1983; 145(2):452-457. 90. Giancotti P, Grappelli C, Poggesi I et al. Persistence of increased levels of ribosomal gene activity in CHO-K1 cells treated in vitro with demethylating agents. Mutat Res 1995; 348(4):187-192. 91. Ferraro M, Prantera G. Human NORs show correlation between transcriptional activity, DNase I sensitivity, and hypomethylation. Cytogenet Cell Genet 1988; 47(1-2):58-61. 92. Stancheva I, Lucchini R, Koller T et al. Chromatin structure and methylation of rat rRNA genes studied by formaldehyde fixation and psoralen cross-linking. Nucleic Acids Res 1997; 25:1727-1735. 93. Santoro R, Grummt I. Molecular mechanisms mediating methylation-dependent silencing of ribosomal gene transcription. Mol Cell 2001; 8(3):719-725. 94. Sullivan GJ, Bridger JM, Cuthbert AP et al. Human acrocentric chromosomes with transcriptionally silent nucleolar organizer regions associate with nucleoli. EMBO J 2001; 20(11):2867-2874. 95. Gibbons RJ, McDowell TL, Raman S et al. Mutations in ATRX, encoding a SWI/SNF-like protein, cause diverse changes in the pattern of DNA methylation. Nat Genet 2000; 24(4):368-371. 96. Johnson R, Strehler BL. Loss of genes coding for ribosomal RNA in ageing brain cells. Nature 1972; 240(5381):412-414. 97. Johnson LK, Johnson RW, Strehler BL. Cardiac hypertrophy, aging and changes in cardiac ribosomal RNA gene dosage in man. J Mol Cell Cardiol 1975; 7(2):125-133. 98. Gaubatz JW, Cutler RG. Age-related differences in the number of ribosomal RNA genes of mouse tissues. Gerontology 1978; 24(3):179-207. 99. Swisshelm K, Disteche CM, Thorvaldsen J et al. Age-related increase in methylation of ribosomal genes and inactivation of chromosome-specific rRNA gene clusters in mouse. Mutat Res 1990; 237(3-4):131-146. 100. Thomas S, Mukherjee AB. A longitudinal study of human age-related ribosomal RNA gene activity as detected by silver-stained NORs. Mech Ageing Dev 1996; 92:101-109. 101. Machwe A, Orren DK, Bohr VA. Accelerated methylation of ribosomal RNA genes during the cellular senescence of Werner syndrome fibroblasts. FASEB J 2000; 14(12):1715-1724. 102. Little RD, Braaten DC. Genomic organization of human 5S rDNA and sequence of one tandem repeat. Genomics 1989; 4:376-383. 103. Sorensen PD, Frederiksen S. Characterization of human 5S rRNA genes. Nucleic Acids Res 1991; 19(15):4147-4151. 104. Hart RP, Folk WR. Structure and organization of a mammalian 5 S gene cluster. J Biol Chem 1982; 257(19):11706-11711. 105. Suzuki H, Sakurai S, Matsuda Y. Rat 5S rDNA spacer sequences and chromosomal assignment of the genes to the extreme terminal region of chromosome 19. Cytogenet Cell Genet 1996; 72(1):1-4. 106. Suzuki H, Moriwaki K, Sakurai S. Sequences and evolutionary analysis of mouse 5S rDNAs. Mol Biol Evol 1994; 11(4):704-710. 107. Szymanski M, Barciszewska MZ, Erdmann VA et al. 5S Ribosomal RNA Database. Nucleic Acids Res 2002; 30(1):176-178. 108. Naylor SL, Sakaguchi AY, Schmickel RD et al. Organization of rDNA spacer fragment variants among human acrocentric chromosomes in somatic cell hybrids. J Mol Appl Genet 1983; 2(2):137-146.

The Structure of rDNA Chromatin

73

CHAPTER 7

The Structure of rDNA Chromatin José M. Sogo and Fritz Thoma

Introduction

T

his chapter is focused on rDNA of yeast S. cerevisiae. The aim is to elucidate our current understanding of chromatin structure and its implications in transcription, replication, recombination and repair. Yeast rDNA is organized in single clusters of about 100 to 150 tamdemly repeated units (Fig. 1). Each unit consists of a non-transcribed spacer (NTS) and a rRNA precursor gene transcribed by RNA polymerase I (RNAP1). Only a fraction of the genes is transcribed, while the remaining fraction is silenced. Besides the promoter, the spacer contains an origin of replication, a transcriptional enhancer, replication fork barriers, and a 5S-rRNA gene. Due to the high copy number of rDNA, conventional biochemical techniques which are based on hybridization, were of limited use to investigate chromatin structure and function at individual units. Only alternative approaches, which allowed the identification and fractionation of active and inactive copies, established that rRNA genes lose nucleosomes when transcribed, while the spacer regions remain largely nucleosomal. These approaches included psoralen crosslinking of cellular DNA in combination with the use of restriction endonucleases to separate active from inactive chromatin. Tagging of individual rDNA copies with unique sequences and integration of reporter constructs were applied to investigate the influence of flanking rDNA environment on the expression and chromatin structure of the reporter construct.

Approaching Chromatin Structure and Function with Psoralen Crosslinking and Restriction Enzymes Psoralen is a drug which intercalates in double stranded DNA and generates covalent interstrand crosslinks upon irradiation with UV-A (predominantly 360nm). Psoralen crosslinking of DNA in cells or nuclei allows to discriminate between nucleosomal and non-nucleosomal chromatin. Psoralen does not intercalate in nucleosomal DNA,1 but it reacts with linker DNA between nucleosomes, nucleosome-free regions such as promoters, origins of replication or enhancers.2 Psoralen crosslinking is not significantly inhibited by elongating RNA- and DNA-polymerases3,4 nor does it affect salt dependent chromatin condensation into compact fibers in vitro.5 Consequently, nucleosomal DNA incorporates less psoralen than non nucleosomal DNA and migrates more rapidly in band shift assays than non-nucleosomal DNA6-8 (Fig. 2). Psoralen crosslinked DNA can be analyzed by electron microscopy under denaturing conditions. Nucleosomal DNA generates single stranded bubbles of the approximate size of nucleosomes. In yeast nucleosomal bubbles have a size of 146+/-32 bp.8 Non-nucleosomal DNA can not be denatured due to the density of crosslinking and appears as double stranded DNA (Fig. 2). Non-nucleosomal DNA was identified in the SV40 origin of replication,2 rDNA enhancers,9 and transcribed rDNA.10 Similar to DNA-crosslinking, cutting by some restriction enzymes is inhibited in nucleosomes, whereas non-nucleosomal DNA remains accessible. Thus restriction enzymes were The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

74

The Nucleolus

Figure 1. Structural organization of an rDNA repeat unit of yeast S. cerevisiae. a) Map of two adjacent rDNA repeats. The 35S and the 5S coding regions are represented by gray and black boxes, respectively. The transcription initiation site (5') and the 3' end of the 35S rRNA (arrowhead) are indicated. b) Enlargement of the intergenic spacer between two 35S coding regions. The 35S promoter (open box), the autonomously replicating sequence (ARS) element (black box) and the enhancer element (E, black box) are shown. The enhancer element with two replication fork barriers (RFB) and the non-transcribed spacer (NTS) 1 and 2 are also shown. c) Positioned nucleosomes (numbered circles 1, 3, 4 and 5) and the origin recognition complex (ORC) binding site (open circle) are indicated. The partially superimposed circles between the 5S and the 3' of the 35S coding region represent randomly distributed nucleosomes. The 5S sequence is packaged in a nucleosome with multiple positions. d) The HOT1 recombination elements are indicated.

applied in nuclei and isolated chromatin to separate accessible, transcribed chromatin from inactive, nucleosomal chromatin (Fig. 3). Since several restriction endonucleases can cleave psoralen crosslinked DNA, it is possible to combine both approaches.11

Chromatin Structure of the rDNA Intergenic Spacer In yeast S. cerevisiae, the ribosomal DNA contains two transcription units, namely the 35S and the 5S rRNA genes, and two regions which are not transcribed (non-transcribed spacer, NTS) (Fig. 1). The non-transcribed spacer 1 (NTS1) is located between the 5S gene and the 3' end of the 35S rRNA gene and accommodates the enhancer (E) and two replication fork barriers (RFB, see below). The non-transcribed spacer 2 (NTS2) extends from the 5S gene to the 5' end of the 35S transcription unit and contains the 35S promoter (P) and an autonomously replicating sequence (ARS) which constitutes a potential origin of replication. A hot spot of recombination (HOT1) was identified that requires sequences of the enhancer (E element) and the 35S promoter (I element).12,13 Several lines of evidence established that the non-transcribed spacers are packaged in nucleosomes. Psoralen cross-linking of rDNA in chromatin followed by electron microscopy revealed single stranded bubbles the size of nucleosomes.8 Chromatin analyses by micrococcal nuclease digestion (MNase) and mapping of the resulting cleavage sites by indirect end-labeling revealed footprints consistent with four positioned nucleosomes in NTS2 (N1, N3, N4, N5 in Fig. 1).14-16 These positions were confirmed by analysis of psoralen crosslinks using exonuclease.17 Whether the ARS region is folded in a nucleosome (N2) is unclear. Vogelauer and coworkers interpreted the MNase footprints in favor of a positioned nucleosome (N2).15 Our own MNase data recorded a footprint which was smaller and weaker than the footprints of the

The Structure of rDNA Chromatin

75

Figure 2. See figure legend on next page.

flanking nucleosomes (Meier and Thoma, unpublished results). DNA-repair analysis of UV-lesions by photolyase is an alternative approach to address DNA accessibility in chromatin in vivo.18 Using that approach, it was found that the ARS region was as slowly repaired as the flanking nucleosomes (N1 to N5), which indicates that a nucleosome or an other protein DNA complex inhibits repair (Meier and Thoma, unpublished results). On the other hand, exonuclease digestion of psoralen crosslinked DNA argued for the absence of nucleosomes in the 180 bp long ARS region.17 Since approximately 30% of the rARSs are activated as origins of replication,19-22 it is possible that ARS exists in two different states, either covered by a

76

The Nucleolus

Figure 2. Organization of transcribed and silent rRNA genes revealed by psoralen crosslinking. a) Cells or isolated nuclei are irradiated with UV light in the presence of psoralen. RNAP1 complexes are permeable to psoralen and consequently the rDNA from transcribed, non nucleosomal genes is rather uniformly crosslinked, whereas the silent DNA and NTS are packaged in nucleosomes and only linkers between nucleosomes are crosslinked. b) For nucleosome mapping (left panel), and in order to detect the crosslinked sites, the purified DNA was first cut with a restriction enzyme and the ends filled with Klenow DNA polymerase. Blunt-ends are the preferred substrates for λ-exonuclease. The exonuclease activity starting at the 5' end is blocked at the psoralen crosslink, leaving a single stranded tail with a free 3' end. Primer extension is performed until the psoralen adduct blocks elongation. The location of psoralen crosslinks, i.e., the position of the linkers, is deduced by the length of the extension products separated in alkaline agarose gels. In order to detect nucleosome positions, labeling was done with a single-stranded probe matching the primer used for primer extension.82 The autoradiogram represents an example of psoralen photoadittion sites in chromatin of the NTS2. Numbered boxes indicate positioned nucleosomes (see also Fig.1). c) For analysis of the transcribed and silent rRNA genes (right panel), the purified crosslinked DNA was digested with a restriction enzyme, fractionated in agarose gel, blotted and hybridized with a specific probe for rDNA. Each restriction fragment derived from the coding region is separated according to the amount of incorporated psoralen and resolved as a double band (lane 1). The slowly (s) migrated band contains heavily crosslinked DNA from the nucleosome-free transcribed genes. The fast (f) migrating band is formed by slightly crosslinked DNA fragments of the silent copies organized in nucleosomes (diagram and lane 1). Lane 2 is non-crosslinked DNA. In lane 3 (prior to psoralen crosslinking), the isolated nucleoli were submitted to a limited run-on in the presence of radioactive precursors. After purification of total nucleic acids, digestion with the appropriate restriction enzyme and fractionation on native agarose gel, the gel was exposed to X-ray film. The band represents genes engaged in transcription-elongation and migrates with the slowly migrating s-band (lane 1). The proportion of the transcribed genes with respect to the silent copies is deduced from the intensity of the s- and f-band, respectively. d) For EM analysis, cesium chloride-purified rDNA from psoralen-crosslinked cells was digested with the appropriate restriction enzyme and fractionated in a low melting native agarose gel. rDNA from the s- and f-band was eluted and prepared under denaturing conditions. s-band DNA appears double stranded and heavily crosslinked, whereas f-band DNA shows regularly spaced single-stranded bubbles of about 150 base-pairs in size, characteristic for nucleosomal organization. Modified figure taken from 10.

nucleosome and inactive in replication, or associated with the replication complex when engaged as an active origin of replication. Muller and coworkers used restriction enzymes on psoralen crosslinked nuclei to purify ARS containing DNA-fragments downstream of active and inactive genes. 2D gels showed that rARSes downstream from silenced genes were inactive, while rARSs downstream of transcribed rRNA genes are potentially active.11 Both classes of fragments gave similar results with the psoralen-primer extension technique arguing for the absence of nucleosomes in ARS. In addition, no significant changes were detected in the nucleosomal organization at the NTS2 throughout the cell cycle.23 It is possible that an origin recognition complex (ORC) is bound to the rARS during all stages of the cell cycle as observed in other ARSs.24 Apparently, the transition from pre-replication complex to post-replication complex does not affect nucleosome organization in the rARS region. The 35S rRNA promoter separates the non-transcribed spacer from the transcribed region. It consists of the upstream element (UE), which binds the upstream activation factor (UAF) and the core promoter, which binds the core factor (CF) and RNAP1 and initiates transcription (ref. 25 and references therein). Genomic footprinting assays with MNase and DNAseI established interactions of CF and UAF in the promoter region.15 Although in growing yeast cultures, where only about half of the rDNA is transcribed, no nucleosome footprint was detected indicating that at least some of the transcription factors might interact with active and inactive promoter elements and exclude nucleosome formation. Consistent with these data, we observed that UV-damage in the promoter region is resistant to repair by photolyase (Meier and Thoma, unpublished results). Not much is known on the chromatin structure of the 5S-gene. High resolution analyses of nucleosomal DNA obtained by DNase digestion suggested that a nucleosome can occupy

The Structure of rDNA Chromatin

77

Figure 3. Fractionation of transcriptionally active and silent rDNA copies for analysis of replicative intermediates and nucleosome positions at NTS. a) A part of the rDNA locus is represented schematically with transcribed (gray boxes with growing rRNA chains depicted as perpendicular bars) and silent (small circles corresponding to nucleosomes) rRNA genes. Transcriptionally active genes are accessible for the restriction enzyme (perpendicular arrows) whereas silent copies remain inaccessible. b) Subsequent re-digestion of the purified rDNA with a second restriction enzyme (arrowheads) generates four rARS-containing restriction fragments of different lengths (I, II, III, IV). c) Diagrams represent neutral/neutral 2D gel patterns generated by replication intermediates corresponding to restriction fragments containing silent rARS (fragments I and II) or active rARS (fragments III and IV). Note that the bubble arc is present only in the fragments with a transcriptionally active gene upstream from a given rARS (fragments III and IV). The position of the linear unreplicated monomers from each generated restriction fragment is represented as a prominent spot. d) Independent gel slices (circles in c) with accumulated unreplicated restriction fragments containing silent rARSs (I and II) or potentially active rARSs (III and IV) were removed, the rDNA was recovered and submitted to the the psoralen-exonuclease assay for nucleosome mapping (Fig. 2, for details see ref. 23).

multiple positions.26 It remains to be clarified what fraction of 5S-genes is folded in nucleosomes and how nucleosomes are related to transcription by RNA polymerase III (RNAP3). In contrast to NTS2, the NTS1 is not organized into positioned nucleosomes. The presence of nucleosomes was inferred from psoralen crosslinking,8,17 nuclease digestions,15 and DNA-repair (Meier and Thoma, unpublished results), but footprints which were consistent with discrete positions were not observed. The enhancer is a region of rDNA which plays a role in different functions and therefore might adopt different compositions and structures. It is involved in initiation and termination of rDNA transcription,27,28 and it contains two replication fork barriers that prevent a clash of transcription with replication.29,30 Moreover, it tends to be a hot spot for recombination31 and contributes to the (in)stability of rDNA repeats. Psoralen crosslinking revealed that inactive, nucleosomal genes were followed by nucleosomal enhancers, whereas non nucleosomal enhancers were found downstream of transcribed (non-nucleosomal) genes suggesting that enhancer proteins might act on the upstream promoters.9 Since open enhancers were also

78

The Nucleolus

detected in a RNAP1 deletion mutant, the open enhancer appears to be the cause rather than the consequence of transcriptional activity in the upstream gene.32 The Fob1 protein is required for the arrest of the replication forks.33 Fob1p weakly binds34 to the enhancer element at multiple sites.29,31 The enhancer element also contains a weak binding site for ARS-binding factor, Abf1.35 Abf1 is an abundant multifunctional protein, which activates different genes and binds to the mating type silencer and ARS regions.36 Abf1 is not essential for initiation of replication, but stimulates ARS1 activity.37,38 Since ARS1 is also nucleosome free,39 Abf1 might be involved in preventing nucleosome assembly.40,41 In summary, transcription, replication and recombination are observed simultaneously in the rDNA locus. The enhancer element in its active, non nucleosomal state, may play a coordinating role that regulates the density of the transcribing polymerases, prevents head-on collision between transcription and replication, promotes activation of rARSs as origins of replication, and regulates in the gain or loss of rDNA copies.

Two Different Classes of Chromatin Coexist in the 35S Coding Region The discovery of nucleosomes raised the question of whether elongating RNA-polymerases can transcribe through nucleosomes or whether nucleosomes are disrupted or displaced. While there is strong evidence that nucleosomes are lost in rDNA transcribed by RNAP1, it is believed that genes transcribed by RNAP2 maintain nucleosomes. Historically, rRNA genes were the first genes to be visualized by electron microscopy during transcription elongation.42 They appeared as “Christmas tree-like” structures with nascent transcripts extending from a chromatin axis. The high density of the polymerases prevented a detailed analysis of the transcribed region with respect to the presence or absence of nucleosomes. However, when nascent RNA was removed by RNase,43 short gaps of DNA were visible between polymerases consistent with the absence of nucleosomes. MNase digestion of cells with low and high RNAP1 activity revealed nucleosomal repeat patterns and smears which is consistent with the presence and absence of nucleosomes in inactive and active genes, respectively (for a detailed review and original references see ref. 10). Consistent with the enhanced accessibility to MNase, it was also found that restriction endonucleases efficiently cut in transcribed rDNA, while cutting in the silent copies was inhibited by nucleosomes.7,11,44,45 This observation allowed the separation of active from inactive rDNA chromatin (Fig. 3). The strongest support for the absence of nucleosomes in transcribed rDNA came from psoralen crosslinking studies. Psoralen crosslinking of chromatin followed by restriction endonuclease digestion of purified DNA and gel electrophoresis revealed a slowly (s) migrating band representing heavily crosslinked DNA and a fast (f ) migrating band containing slightly crosslinked DNA (Fig. 2). Moreover, nascent rRNA was found to be crosslinked to the “s band” proving that these DNA fragments originated from transcribed genes.7,8 Electron micrographs showed that the “s band” DNA was uniformly crosslinked while the “f band” DNA was organized in single stranded bubbles corresponding to nucleosomes. Treatment of “s band” DNA with exonuclease generated a smear, whereas the treatment of the “f band” DNA revealed a repeat pattern characteristic for nucleosomes. This coexistence of these two different classes of ribosomal chromatin was observed from yeast to human, plants and insects (ref. 10 and references therein; Sanz and Diez, personal communication). In contrast to the rDNA genes, it is generally accepted that genes transcribed by RNAP2 retain nucleosomes during transcription and that the passage of an RNA polymerase through a gene is coupled to mechanisms that propagate the breakdown of chromatin.46,47 Evidence for the presence of nucleosomes during transcription came from a study with SV40 minichromosomes transcribed in vivo by RNAP2 which showed that nascent RNA-transcript could be crosslinked to nucleosomal DNA by psoralen.3 Moreover, nuclease digestion and electron microscopy provided additional support for the presence of nucleosomes in genes transcribed by RNAP2.48 No loss of nucleosomes was measurable by psoralen crosslinking in a gene heavily transcribed from the yeast GAL1 promoter but the positions of nucleosomes were

The Structure of rDNA Chromatin

79

lost. Repositioning occurred rapidly after gene inactivation and argues for the presence of nucleosomes during transcription by RNAP2.16 The large size of eukaryotic RNA-polymerases and the growing size of the nascent RNA transcript make it difficult to understand how eukaryotic polymerases transcribe nucleosomes without major disruption. In vitro transcription of reconstituted nucleosome cores with RNAP2 results in a loss of histones H2A and H2B, while the position of the remaining nucleosome remained unaltered.49 In contrast to RNAP2, the histone octamer appeared to be transferred behind the polymerase when the nucleosome core was transcribed by a small SP6 polymerase.50 Thus, the different polymerases might use different mechanisms to cope with nucleosomes. Additional information comes from the recent discovery that elongation factors are associated with RNAP2 and contain histone acetyltransferase and nucleosome remodeling activities.46 Thus, an elongating RNA polymerase might modify nucleosomes and allow it to disrupt and reassemble the nucleosome in front and behind the polymerase, respectively. So far, no similar elongation factors have been reported for RNAP1. Thus, RNAP1 and RNAP2 may use different mechanisms to transcribe nucleosomal DNA. Rather than remodeling nucleosomes, RNAP1 apparently prefers to erase nucleosomes from its path. This mechanism might be a consequence of the high density of RNAP1 on the transcribed gene (about one RNAP1 every 130 bp) which does not leave enough space for reformation of a stable nucleosome. Although nucleosomes were not detected as a structural subunit of active rDNA genes, the fate of the histones and the mechanism by which nucleosomes are removed remain unclear. The state of histone acetylation is generally considered as a possible contribution to nucleosome stability. However, hyperacetylation alone does not explain nucleosome disruption in rDNA, since chromatin reconstituted with hyperacetylated histones is resistant to psoralen crosslinking.51 Several groups addressed the state of acetylation in rDNA chromatin. Chromatin immunoprecipitation experiments showed that acetylation of H3 and H4 was unaltered at the rDNA locus when the RPD3 gene coding for a histone deacetylase was deleted in yeast. Thus, the deacetylated histones may occur in a small subset of rRNA genes, which makes their detection very difficult.52 Mutskov and coworkers addressed histone acetylation by chromatin immunoprecipitation with antibodies against acetylated histones using rat cells that were treated with butyrate to inhibit deacetylation.53 While rDNA was precipitated with the UV-crosslinked chromatin, the fraction of rDNA in immunoprecipitated mononucleosomes was very small. This was taken as evidence for an association of acetylated histones with non-nucleosomal, active rDNA.53 It is, however, surprising to see that even the amount of rDNA in the total mononucleosome population is much smaller than expected. Thus, it remains to be investigated whether butyrate treatment might have activated most of the rDNA genes. Our own analysis by combined psoralen and formaldehyde crosslinking revealed that practically no peptides were crosslinked on rDNA coding fragments derived from the transcribed genes.54 Hence, the state of histone acetylation in active and inactive genes remains to be established. Open, non nucleosomal rDNA was observed not only in actively transcribed genes, but also in Friend cells blocked at mitosis where RNAP1 is immobilized in the rDNA. At least in mouse cells two states of chromatin are maintained independently of the transcriptional process and they are stably propagated through the cell cycle.7 Even when insect cells are treated with rRNA synthesis inhibitors like cycloheximide or actinomycin-D (C. Sanz and J. L. Diez, personal communication) the two classes of rDNA chromatin are detected.

Regulation of rDNA Transcription In yeast, ribosome biogenesis is tightly coupled to cell proliferation and adapts rapidly to changing growth conditions.55 Transcriptional activity of rDNA can be regulated in two ways. The fraction of active genes may be up- or downregulated in response to variations of environmental conditions as revealed by psoralen crosslinks.8 Alternatively, the rDNA enhancer promotes the rate of reinitiation on already active rRNA genes and not by activating silent transcription units.32 This double transcription regulation mechanism has been confirmed in rpd3∆

80

The Nucleolus

mutants growing in a post-log phase.52 Thus, yeast is able to activate and repress rDNA units and to modulate the number of polymerases per transcription unit in order to fulfill the requirement of ribosomes in different growth conditions. In higher eukaryotes, cell differentiation determines the fraction of transcribable rRNA genes and thereby the rate of rRNA synthesis in a particular cell type,56 (for details see ref. 10). Several recent studies elucidate the potential roles of histones and chromatin remodeling activities in activation and silencing of rDNA transcription. Histones H3 and H4 were copurified with the transcription factor UAF suggesting that histones are involved in setting up chromatin structure in the promoter region.57 Silencing of rDNA is frequently measured by insertion of a gene which is transcribed by RNAP2. Silencing of such a reporter gene requires the histone H3 and H4 tails and a complex containing Sir2 (a NAD dependent histone deacetylase),58 Net1, and Cdc14. The rDNA associated H4 was found to be hypoacetylated in a Sir2 dependent manner, suggesting that hypoacetylation of H3 and H4 might be involved in assembly of silent chromatin.59 In mammalian cells, NoRC, a nucleolar remodeling complex containing Snf2h, represses rDNA transcription by recruiting DNA methyltransferase and histone deacetylase activity to the rDNA promoter. Thus modifications that are characteristic for heterochromatin such as DNA methylation, histone hypoacetylation and methylation of the Lys9 residue of histone H360,61 are found in rDNA promoters. While the biochemical organization of rDNA becomes more and more clear, the structural information remains scarce. First, it is not clear whether histones H3 and H4 may play a structural role in the formation of a promoter specific chromatin structure that facilitates interactions between transcription factors. Moreover, it is not clear whether silencing of rDNA includes the formation of nucleosomes in the promoter and coding region only or whether it requires additional packaging in higher order structures. In this context, it is interesting to know that removal of UV lesions by nucleotide excision repair and by photolyase was found to be faster in the active rDNA genes and slower in the silenced genes which is consistent with a more open chromatin structure of the active genes. The inactive rDNA, however, was repaired slightly faster than the non transcribed nucleosomal GAL10 gene, demonstrating that silencing of rDNA did not induce additional compaction.45

Replication of rDNA Since rDNA is repetitive and only a fraction of the genes is transcribed, it is interesting to know whether replication reestablishes a default state, or whether and how the transcriptional state is inherited and propagated to the next generation. Three DNA elements play important roles in rDNA replication: the potential origin of replication (rARS), the replication fork barrier (RFB), and the enhancer. Replication of rDNA always initiates at an rARS and proceeds bidirectionally. Replication forks moving against the direction of transcription are arrested at the replication fork barrier (RFB) located at the 3' end of transcriptionally active units.62 Replication forks moving in the same direction as transcription are not blocked and merge with the forks stalled at the RFB. Thus, RFBs constitute the site of replication termination.19,20 RFBs have been identified in yeast and many other organisms including humans (for review see ref. 63). Most surprisingly, blockage of replication forks at the RFB was also detected in a strain in which the gene coding for the 135 Kd subunit of the RNAP1 was disrupted64 demonstrating that transcription elongation was not the cause for blocking replication.29,65 However, the presence of a subset of “open” nucleosome free enhancers detected in the same RNAP1 mutant strain9 suggested that enhancers remained occupied by factors responsible for transcription and RFB activity. Less than one third of the rARSs are used as replication origins in a given S-phase (reviewed in ref. 66). Deletion of FOB1, which codes for a factor required for RFB function, does not exhibit any obvious growth defect 33,67 nor does it affect transcription (see discussion in ref. 31) suggesting that the fraction of transcribed genes with open enhancers is similar in fob1∆ and wild type cells and that the initiation of replication is independent of FOB1. Initiation of

The Structure of rDNA Chromatin

81

replication always starts at rARSs downstream of transcribed genes, but not every rARS located downstream of a transcribed gene is used as an origin of replication11 suggesting a connection between origin activation and transcription. Each transcriptionally active gene is flanked at its 3' end by a non nucleosomal enhancer which contains a binding site for the ARS binding factor Abf1. Abf1 is known to be required for full ARS activity in other ARSs.37,38,68 Since the rARS does not contain an Abf1 binding site, rARS might be activated by Abf1 bound to the upstream enhancer. Since the RFB at the rRNA gene locus is a common feature of the eukaryotic cells,63 and since ribosomal origins of replication have been located in the intergenic spacers of metazoan (for review ref. 63) replication and transcription are likely to be coordinated in all eukaryotes.

Inheritance and Establishment of Chromatin and Transcription during Replication The consequence of origin activation downstream of transcribed genes is that most of the spacer and each coding region, silent or transcribed, is replicated by the rightward moving forks. Replication of DNA is intimately linked to recruitment of new histones and regeneration of chromatin structures. Electron microscopic analysis of replicative intermediates showed that nucleosome assembly on newly replicated rDNA occurs immediately after the passage of the replication machinery69 in a similar way as previously described by SV40 minichromosomes.2 Isolation of psoralen cross-linked replicative intermediates from preparative 2D gels 17 combined with exonuclease digestion or primer extension allowed to characterize nucleosome positions after replication. Nucleosomes in the NTSs appeared to be positioned only a few seconds after the passage of the fork (~600 bp behind the fork). This implies that the “old” segregated nucleosomes as well as the “new” recruited histone octamers were rapidly deposited at their most favored position. Similarly, EM analysis of psoralen crosslinked DNA revealed that replication of a silent (nucleosomal) gene gives rise to two silent (nucleosomal) copies. However, replication of non nucleosomal active genes generates coding regions packaged in nucleosomes. This indicates that the transcriptional state of chromatin is not inherited at the replication fork.9,70 Silencing of newly replicated rDNA units is transient. Regeneration of transcription is a post-replicative event involving the disruption of the nucleosomes by moving RNAP1,70 (see also Fig. 4). In contrast to the coding region, open promoters were detected immediately after the passage of the replication fork70 indicating that factors which prevent nucleosome formation must be recruited to the promoter during or immediately after replication. The fact that the initiation of replication was restricted to rARSs located downstream of transcriptionally active genes11 and that the subset of transcribed rRNA gene copies was not stably propagated during cell division9 implied that the rARSs to be activated as origins of replication are not inherited. This conclusion was recently confirmed by the visualization of origins of replication on single rDNA molecules.71 Moreover, Pasero and coworkers did not discard the possibility that the use of large populations of cells could mask the epigenetic memory of origin usage that persists only in a few generations. When SIR2 (NAD-dependent histone deacetylase58) was deleted, the number of activated origins increased two fold.71 Since the number of transcribed rDNA genes remained unchanged with respect to the wild type,52 the additional use of the rARSs in sir2 cells might be related to a lack of deacetylation of proteins involved in origin activation or histones involved in the chromatin organization.72 Nucleosome mapping at the NTS2 in the sir2 strain revealed that the position of the nucleosomes (Fig. 1) remains unaltered with respect to wild type cells, while linkers become more accessible to MNase.73 The restricted activation of the rARS and the location of the RFB downstream of transcribed genes, together with the establishment of potentially active promoters shortly after the passage of the replication fork demonstrate the tight link between transcription and replication at the rDNA locus.

82

The Nucleolus

Figure 4. Chromatin structure of rDNA replicative intermediates. Cesium chloride-purified rDNA from psoralen crosslinked cells was digested with PvuI, which cuts only once per repeat unit, and prepared under denaturing conditions for electron microscopy. a) Non-replicated rDNA unit. Indicated are the PvuI cutting site, the 5' end the 3' end of two contiguous 35S coding regions. The two coding regions appear as rather uniformly crosslinked duplex DNA that is typical for transcribed genes devoid of nucleosomes. The intergenic spacer between 3' and 5' end shows single-stranded bubbles representing nucleosomes. b) In this rDNA unit the origin of replication was active. The leftward-moving fork is arrested at the RFB in front of a non-replicated segment derived from a transcribed rRNA gene. The rightward-moving fork is replicating the downstream rRNA gene, which is also transcribed. The newly replicated strands behind the rightwards-moving fork are organized in single-stranded bubbles. The RFB and the putative 5' end of the two nascent 35S coding regions are indicated. c) When the rightward-moving fork in the rDNA repeat represented in b) reaches the PvuI cutting site, a replicating Y- shaped molecule is formed. In one of the newly replicated strands a segment of DNA is uniformly crosslinked (between the putative 5' end and the arrowhead) indicating that in this rRNA gene transcription has started. The arrested fork at the RFB in front of a transcribed rRNA gene is indicated. d) In this molecule, one of the two newly replicated rRNA genes appears uniformly crosslinked, indicating that it is transcribed. e) Model for rDNA chromatin replication. The origin of replication is (a’) activated downstream of a transcribed rRNA gene. The leftward-moving fork is arrested at the RFB always behind transcribed gene (b’). The rightward-moving fork, travelling co-directional to rRNA transcription (b’), merges with the stalled fork at the next RBF (c’), thus concluding replication. The rightward-moving fork replicates both the silent and transcribed rRNA gene copies. When the replication fork progresses through the transcribed gene, the newly replicated strands are immediately packaged in nucleosomes (b’). Nucleosomes at the NTS2 are rapidly positioned after the passage of the replication machinery. Frequently a putative transcription factor is bound to a promoter of one replicated gene that is detected as a short nucleosome-free region (b’, c’). Transcription initiation occurs and the moving polymerases open up the nucleosomal chromatin (c’, d’). Before replication is finished, eventually the promoter of the sibling rRNA gene is activated as well and transcription elongation occurs (d’).11, 70

The Structure of rDNA Chromatin

83

HOT1 Dependent Recombination in rDNA rDNA sequences involved in recombination (HOT1, see above) have been characterized when placed at ectopic sites between his4 repeats carrying a URA3 reporter gene or when inserted in plasmids. By using this recombination system, it was demonstrated that HOT1 activity is strictly dependent on promoter orientation, whereas the enhancer can function in either orientation 13. Furthermore, transcription by RNAP1 through the his4 sequences is necessary for HOT1-mediated recombination.13,74,75 At ectopic sites however, HOT1-stimulated recombination does not require the arrest of forks at the RFB.31 What happens with recombination at the rRNA gene locus? The observation that the FOB1 gene, is required for the stalling of the replication forks and for the enhanced activity of HOT1 recombination, suggests that the recombination is closely linked to the RFB.33 Furthermore, in yeast cells, the number of rDNA repeats is controlled by the RNAP1 transcription and FOB1 is required for the increase and loss of rDNA repeats.33 Mutational analysis by substitution of the RFB sequences and their flanking region with a URA3 gene fragment revealed that not only the RFBs but also their adjacent region (called EXP region) are required for FOB1-dependent rDNA expansion.76 Accordingly, Kobayashi et al proposed that a pausing replication machinery at the RFB induces double strand-breaks at the newly replicated lagging strand. Repair of the double strand break could be initiated by the strand invasion, and recombination with a rDNA unit either from the sister chromatid or with the parental unreplicated strand occurs. In the first case, the recombination will lead to a gain of rDNA units, and the second way results in a loss of rDNA repeats (for details see refs. 65, 76, 77). It is believed that the rDNA circles observed in aged cells are formed by a similar intrachromosomal recombination process between the rDNA repeats from the same chromatid.67 In fact, it has been shown that the increasing amount of stalled forks correlates with the accumulation of rDNA circles.71,78 So far, data supporting a double strand break at the RFB are lacking. Intuitively, one might expect that the length of single stranded DNA at the replication fork would affect the frequency of strand breaks. However, it was shown that the lagging strand of the stalled replication fork is completely processed and it is extended three bases further than the leading strand.30 Therefore, the induction of the break could likewise occur at the leading or lagging strand. It is conceivable that Fob1p acts as a nuclease possibly together with proteins of the replication fork such as helicases.79 An alternative model has been proposed which involves a Holliday junction at the arrested fork (reversed fork) as an intermediate DNA structure before the break takes place.67 Although Holliday junctions have been visualized in checkpoint-deficient yeast cells,69 no indication for reversed forks was found in wild type cells at rDNA molecules containing arrested forks.30 Assuming that a strand break occurs at the RFB, it is imaginable that the strand invasion of the free end is most efficient in “open” enhancers, since the DNA of non nucleosomal chromatin is more readily accessible than in nucleosomal regions (Fig. 5). Consequently, if the rRNA gene which is immediately downstream of the break is silent, the next ‘open’ enhancer (downstream of an active gene) could be located several rRNA gene copies downstream or upstream of the break (Fig. 5B). This is consistent with the observation that ECRs may carry several rDNA units,80 (Bouza and Sogo, unpublished data) and that the number of rDNA units in the largest ECRs roughly correlates with the average size of the yeast replicons having between three to five repeats.21,22 This is also supported by the observation that in old cells the fraction of ‘open’ (active) and nucleosomal repeats was similar in ECRs and genomic DNA (Bouza and Sogo, unpublished data). Thus, in old cells the RNA polymerase is not a limiting factor, and chromosomal and extrachromosomal copies appear to be equally well transcribed. Besides the possible engagement of the RFB and enhancer region in recombination, it is important to note that Holliday junctions were detected in the 35S coding region which led Zou and Rothstein to propose that recombination at the ribosomal locus is stimulated to repair replication related DNA lesions.81 Those lesions probably occur randomly along the entire rDNA locus and their repair is RFB independent. Replication dependent lesions, like the for-

The Nucleolus

84

Figure 5. Formation of ERC by intrachromosomal homologous recombination. a) A break is induced at one arm of the newly replicated strands stalled at the RFB. The resulting double strand break (DSB) localized at the nucleosome free enhancer region (arrowhead) is processed and recombined with the nucleosome-free enhancer from the transcriptionally active gene immediately downstream (arrowhead). The resulting circular molecule contains an individual rRNA gene (ECR-monomer). b) If the gene immediately downstream from the free end is silent, its nucleosomal enhancer does not favor the recombination process, and the free end is pared with its homologous sequences at the next nucleosome free enhancer (arrowheads). By this mechanism dimers or multimers are formed (for details see text).

mation of single stranded gaps, more frequently occur in strains with defective DNA-polymerases. Large single stranded gaps at the replication forks as well as hemi-replicated intermediates have been visualized in a DNA polymerase a-primase mutant.69 In conclusion, at the rDNA locus apparently two types of recombination events exist. One is important for the maintenance of the number of rDNA repeats and is RFB-dependent, and the other mainly repairs the rDNA damage occurring during replication. It will be of particular interest to determine the individual role of the enhancer and the promoter in rDNA recombination. A further aim is to elucidate the molecular mechanisms that link recombination, transcription and replication.

References 1. Hanson CV, Shen CK, Hearst JE. Cross-linking of DNA in situ as a probe for chromatin structure. Science 1976; 193(4247):62-4. 2. Sogo JM, Stahl H, Koller T et al. Structure of replicating simian virus 40 minichromosomes. The replication fork, core histone segregation and terminal structures. J Mol Biol 1986; 189(1):189-204. 3. De Bernardin W, Koller T, Sogo JM. Structure of in-vivo transcribing chromatin as studied in simian virus 40 minichromosomes. J Mol Biol 1986; 191(3):469-82. 4. Gruss C, Sogo JM. Chromatin replication. Bioessays 1992; 14(1):1-8. 5. Conconi A, Losa R, Koller T et al. Psoralen-crosslinking of soluble and of H1-depleted soluble rat liver chromatin. J Mol Biol 1984; 178(4):920-8. 6. Carlson JO, Pfenninger O, Sinden RR et al. New procedure using a psoralen derivative for analysis of nucleosome associated DNA sequences in chromatin of living cells. Nucleic Acids Res 1982; 10(6):2043-63. 7. Conconi A, Widmer RM, Koller T et al. Two different chromatin structures coexist in ribosomal RNA genes throughout the cell cycle. Cell 1989; 57(5):753-61.

The Structure of rDNA Chromatin

85

8. Dammann R, Lucchini R, Koller T et al. Chromatin structures and transcription of rDNA in yeast Saccharomyces cerevisiae. Nucleic Acids Res 1993; 21(10):2331-8. 9. Dammann R, Lucchini R, Koller T et al. Transcription in the yeast rRNA gene locus: distribution of the active gene copies and chromatin structure of their flanking regulatory sequences. Mol Cell Biol 1995; 15(10):5294-303. 10. Lucchini R, Sogo JM. The dynamic structure of ribosomal RNA gene chromatin. In: Paule MR, editor. Transcription of ribosomal RNA genes by Eukaryotic RNA polymerase I. Berlin, Heidelberg, Georgetown: Springer-Verlag and R.G. Landes Company, 1998:255-276. 11. Muller M, Lucchini R, Sogo JM. Replication of yeast rDNA initiates downstream of transcriptionally active genes. Mol Cell 2000; 5(5):767-777. 12. Keil RL, Roeder GS. Cis-acting, recombination-stimulating activity in a fragment of the ribosomal DNA of S. cerevisiae. Cell 1984; 39(2Pt1):377-386. 13. Voelkel-Meiman K, Keil RL, Roeder GS. Recombination-stimulating sequences in yeast ribosomal DNA correspond to sequences regulating transcription by RNA polymerase I. Cell 1987; 48(6):1071-1079. 14. Fritze CE, Verschueren K, Strich R et al. Direct evidence for SIR2 modulation of chromatin structure in yeast rDNA. EMBO J 1997; 16(21):6495-6509. 15. Vogelauer M, Cioci F, Camilloni G. DNA protein-interactions at the Saccharomyces cerevisiae 35 S rRNA promoter and in its surrounding region. J Mol Biol 1998; 275(2):197-209. 16. Cavalli. Transcription dependent chromatin transitions in the yeast Saccharomyces cerevisiae. ETH, 1994. 17. Lucchini R, Wellinger RE, Sogo JM. Nucleosome positioning at the replication fork. EMBO J 2001; 20(24):7294-7302. 18. Suter B, Livingstone-Zatchej M, Thoma F. Chromatin structure modulates DNA repair by photolyase in vivo. EMBO J 1997; 16(8):2150-2160. 19. Brewer BJ. When polymerases collide: replication and the transcriptional organization of the E. coli chromosome. Cell 1988; 53(5):679-686. 20. Linskens MH, Huberman JA. Organization of replication of ribosomal DNA in Saccharomyces cerevisiae. Mol Cell Biol 1988; 8(11):4927-4935. 21. Saffer LD, Miller OJ. Electron microscopic study of Saccharomyces cerevisiae rDNA chromatin replication. Mol Cell Biol 1986; 6(4):1148-1157. 22. Walmsley RM, Johnston LH, Williamson DH et al. Replicon size of yeast ribosomal DNA. Mol Gen Genet 1984; 195(1-2):260-266. 23. Muller M. Replication of the yeast rRNA gene locus: Initiation starts downstream of transcriptionally active gene copies. Dissertation, ETH-Zürich. 1999. 24. Diffley JF. Eukaryotic DNA replication. Curr Opin Cell Biol 1994; 6(3):368-72. 25. Nomura M. Transcription factors used by Saccharomyces cerevisiae RNA polymerase I and the mechanism of initiation. In: Paule MR, ed. Transcription of ribosomal RNA genes by Eukaryotic RNA polymerase I. Berlin, Heidelberg, Georgetown: Springer-Verlag and R.G. Landes Company, 1998:155-167. 26. Buttinelli M, Di Mauro E, Negri R. Multiple nucleosome positioning with unique rotational setting for the Saccharomyces cerevisiae 5S rRNA gene in vitro and in vivo. Proc Natl Acad Sci USA 1993; 90(20):9315-9. 27. Warner JR, Ikonomova R, Johnson SP et al. The role of the enhancer element in rRNA transcription in saccharomyces cerevisiae. In: Paule MR, ed. Transcription of ribosomal RNA genes by Eukaryotic RNA polymerase I. Berlin, Heidelberg, Georgetown: Springer-Verlag and R.G. Landes Company, 1998:243-248. 28. Reeder RH, Lang WH. Stopping RNA polymerase I. In: Paule MR, ed. Transcription of ribosomal RNA genes by Eukaryotic RNA polymerase I. Berlin, Heidelberg, Georgetown: Springer-Verlag and R.G. Landes Company, 1998:173-177. 29. Brewer BJ, Lockshon D, Fangman WL. The arrest of replication forks in the rDNA of yeast occurs independently of transcription. Cell 1992; 71(2):267-76. 30. Gruber M, Wellinger RE, Sogo JM. Architecture of the replication fork stalled at the 3' end of yeast ribosomal genes. Mol Cell Biol 2000; 20(15):5777-87. 31. Ward TR, Hoang ML, Prusty R et al. Ribosomal DNA replication fork barrier and HOT1 recombination hotspot: Shared sequences but independent activities. Mol Cell Biol 2000; 20(13):4948-57. 32. Banditt M, Koller T, Sogo JM. Transcriptional activity and chromatin structure of enhancer-deleted rRNA genes in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19(7):4953-60. 33. Kobayashi T, Horiuchi T. A yeast gene product, Fob1 protein, required for both replication fork blocking and recombinational hotspot activities. Genes Cells 1996; 1(5):465-74.

86

The Nucleolus

34. Gruber M. Molecular dissection of the yeast replication fork barrier:I: Architectire of the stalled replication fork, II: In vitro reconstitution of the replication fork barrier. Dissertation, ETH-Zürich. 1999. 35. Elion EA, Warner JR. An RNA polymerase I enhancer in Saccharomyces cerevisiae. Mol Cell Biol 1986; 6(6):2089-97. 36. Planta RJ, Goncalves PM, Mager WH. Global regulators of ribosome biosynthesis in yeast. Biochem Cell Biol 1995; 73(11-12):825-34. 37. Rhode PR, Elsasser S, Campbell JL. Role of multifunctional autonomously replicating sequence binding factor 1 in the initiation of DNA replication and transcriptional control in Saccharomyces cerevisiae. Mol Cell Biol 1992; 12(3):1064-77. 38. Marahrens Y, Stillman B. Replicator dominance in a eukaryotic chromosome. EMBO J 1994; 13(14):3395-400. 39. Thoma F, Bergman LW, Simpson RT. Nuclease digestion of circular TRP1ARS1 chromatin reveals positioned nucleosomes separated by nuclease-sensitive regions. J Mol Biol 1984; 177(4):715-33. 40. Lipford JR, Bell SP. Nucleosomes positioned by ORC facilitate the initiation of DNA replication. Mol Cell 2001; 7(1):21-30. 41. Lascaris RF, Groot E, Hoen PB et al. Different roles for abf1p and a T-rich promoter element in nucleosome organization of the yeast RPS28A gene. Nucleic Acids Res 2000; 28(6):1390-6. 42. Miller OL, Beatty BR. Visualization of nucleolar genes. Science 1969; 164:955-957. 43. Labhart P, Koller T. Structure of the active nucleolar chromatin of Xenopus laevis Oocytes. Cell 1982; 28(2):279-92. 44. Ness PJ, Labhart P, Banz E et al. Chromatin structure along the ribosomal DNA of Dictyostelium. Regional differences and changes accompanying cell differentiation. J Mol Biol 1983; 166(3):361-81. 45. Meier A, Livingstone-Zatchej M, Thoma F. Repair of active and silenced rDNA in yeast: the contributions of photolyase and transcription-couples nucleotide excision repair. J Biol Chem 2002; 277(14):11845-52. 46. Svejstrup JQ. Chromatin elongation factors. Curr Opin Genet Dev 2002; 12(2):156-61. 47. Orphanides G, Reinberg D. RNA polymerase II elongation through chromatin. Nature 2000; 407(6803):471-5. 48. Hill RS. A quantitative electron-microscope analysis of chromatin from Xenopus laevis lampbrush chromosomes. J Cell Sci 1979; 40:145-169. 49. Kireeva ML, Walter W, Tchernajenko V et al. Nucleosome remodeling induced by RNA polymerase II: loss of the H2A/H2B dimer during transcription. Mol Cell 2002; 9(3):541-52. 50. Studitsky VM, Clark DJ, Felsenfeld G. A histone octamer can step around a transcribing polymerase without leaving the template. Cell 1994; 76(2):371-382. 51. Nightingale KP, Wellinger RE, Sogo JM et al. Histone acetylation facilitates RNA polymerase II transcription of the Drosophila hsp26 gene in chromatin. EMBO J 1998; 17(10):2865-76. 52. Sandmeier JJ, French S, Osheim Y et al. RPD3 is required for the inactivation of yeast ribosomal DNA genes in stationary phase. EMBO J 2002; 21(18):4959-68. 53. Mutskov VJ, Russanova VR, Dimitrov SI et al. Histones associated with non-nucleosomal rat ribosomal genes are acetylated while those bound to nucleosome-organized gene copies are not. J Biol Chem 1996; 271(20):11852-7. 54. Stancheva I, Lucchini R, Koller T et al. Chromatin structure and methylation of rat rRNA genes studied by formaldehyde fixation and psoralen cross-linking. Nucleic Acids Res 1997; 25(9):1727-35. 55. Kief DR, Warner JR. Coordinate control of syntheses of ribosomal ribonucleic acid and ribosomal proteins during nutritional shift-up in Saccharomyces cerevisiae. Mol Cell Biol 1981; 1(11):1007-15. 56. Wiesendanger B. I. The proportion of non-nucleosomal, transcribable and nucleosomal, non-transcribable ribosomal RNA gene copies is cell type-specific. II. Replication fork barriers in Xenopus laevis and Xenopus borealis ribosomal DNA. Dissertation, ETH-Zürich. 1994. 57. Keener J, Josaitis CA, Dodd JA et al. Reconstitution of yeast RNA polymerase I transcription in vitro from purified components. TATA-binding protein is not required for basal transcription. J Biol Chem 1998; 273(50):33795-802. 58. Imai S, Armstrong CM, Kaeberlein M et al. Transcriptional silencing and longevity protein Sir2 is an NAD- dependent histone deacetylase. Nature 2000; 403(6771):795-800. 59. Hoppe GJ, Tanny JC, Rudner AD et al. Steps in assembly of silent chromatin in yeast: Sir3-independent binding of a Sir2/Sir4 complex to silencers and role for Sir2-dependent deacetylation. Mol Cell Biol 2002; 22(12):4167-80. 60. Zhou Y, Santoro R, Grummt I. The chromatin remodeling complex NoRC targets HDAC1 to the ribosomal gene promoter and represses RNA polymerase I transcription. EMBO J 2002; 21(17):4632-40.

The Structure of rDNA Chromatin

87

61. Santoro R, Li J, Grummt I. The nucleolar remodeling complex NoRC mediates heterochromatin formation and silencing of ribosomal gene transcription. Nat Genet 2002; doi:10.1038/ng1010. 62. Lucchini R, Sogo JM. Chromatin structure and transcriptional activity around the replication forks arrested at the 3' end of the yeast rRNA genes. Mol Cell Biol 1994; 14(1):318-26.. 63. López-Estraño C, Schvartzman JB, Hernandez P. The replication of ribosomal RNA genes in eukaryotes. Chromosomes Today 1997; 12:161-181. 64. Nogi Y, Vu L, Nomura M. An approach for isolation of mutants defective in 35S ribosomal RNA synthesis in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1991; 88(16):7026-30. 65. Kobayashi T, Heck DJ, Nomura M et al. Expansion and contraction of ribosomal DNA repeats in Saccharomyces cerevisiae: requirement of replication fork blocking (Fob1) protein and the role of RNA polymerase I. Genes Dev 1998; 12(24):3821-30. 66. Fangman WL, Brewer BJ. Activation of replication origins within yeast chromosomes. Annu Rev Cell Biol 1991; 7(375):375-402. 67. Defossez PA, Prusty R, Kaeberlein M et al. Elimination of replication block protein Fob1 extends the life span of yeast mother cells. Mol Cell 1999; 3(4):447-55. 68. Natale DA, Umek RM, Kowalski D. Ease of DNA unwinding is a conserved property of yeast replication origins. Nucleic Acids Res 1993; 21(3):555-60. 69. Sogo JM, Lopes M, Foiani M. Fork reversal and ssDNA accumulation at stalled replication forks owing to checkpoint defects. Science 2002; 297(5581):599-602. 70. Lucchini R, Sogo JM. Replication of transcriptionally active chromatin. Nature 1995; 374(6519):276-80. 71. Pasero P, Bensimon A, Schwob E. Single-molecule analysis reveals clustering and epigenetic regulation of replication origins at the yeast rDNA locus. Genes Dev 2002; 16(19):2479-84. 72. Ivessa AS, Zakian VA. To fire or not to fire: origin activation in Saccharomyces cerevisiae ribosomal DNA. Genes Dev 2002; 16(19):2459-64. 73. Cioci F, Vogelauer M, Camilloni G. Acetylation and accessibility of rDNA chromatin in Saccharomyces cerevisiae in (Delta)top1 and (Delta)sir2 mutants. J Mol Biol 2002; 322(1):41-52. 74. Stewart SE, Roeder GS. Transcription by RNA polymerase I stimulates mitotic recombination in Saccharomyces cerevisiae. Mol Cell Biol 1989; 9(8):3464-72. 75. Huang GS, Keil RL. Requirements for activity of the yeast mitotic recombination hotspot HOT1: RNA polymerase I and multiple cis-acting sequences. Genetics 1995; 141(3):845-55. 76. Kobayashi T, Nomura M, Horiuchi T. Identification of DNA cis elements essential for expansion of ribosomal DNA repeats in Saccharomyces cerevisiae. Mol Cell Biol 2001; 21(1):136-47. 77. Johzuka K, Horiuchi T. Replication fork block protein, Fob1, acts as an rDNA region specific recombinator in S. cerevisiae. Genes Cells 2002; 7(2):99-113. 78. Ivessa AS, Zhou JQ, Zakian VA. The Saccharomyces Pif1p DNA helicase and the highly related Rrm3p have opposite effects on replication fork progression in ribosomal DNA. Cell 2000; 100(4):479-89. 79. Dlakic M. A model of the replication fork blocking protein Fob1p based on the catalytic core domain of retroviral integrases. Protein Sci 2002; 11(5):1274-7. 80. Sinclair DA, Mills K, Guarente L. Accelerated aging and nucleolar fragmentation in yeast sgs1 mutants. Science 1997; 277(5330):1313-6. 81. Zou H, Rothstein R. Holliday junctions accumulate in replication mutants via a RecA homologindependent mechanism. Cell 1997; 90(1):87-96. 82. Wellinger RE, Lucchini R, Dammann R et al. In vivo mapping of nucleosomes using psoralen-DNA crosslinking and primer extension. Met Mol Biol 1999; 119:161-73.

88

The Nucleolus

CHAPTER 8

Ribosomal DNA Transcription in Mammals Alice Cavanaugh, Iwona Hirschler-Laszkiewicz and Lawrence I. Rothblum

Summary

T

he genes that code for 45S rRNA, the precursor of 18S, 5.8S and 28S rRNA, are transcribed by RNA polymerase I. In many eukaryotes, the genes are arranged as tandem repeats in discrete chromosomal clusters. Moreover, the transcription of these genes and the processing of the transcripts occur in a discrete subnuclear structure, the nucleolus. In vertebrates, at least two factors, SL1 and UBF, specific for transcription by RNA polymerase I, cooperate in the formation of the initiation complex. Interestingly, there are proteins analogous to SL1 in unicellular eukaryotes, but the requirement for a UBF-like factor appears to vary. Recent advances in our understanding of the rDNA transcription system and its regulation have demonstrated overlap with the other nuclear transcription systems (RNA polymerase II and III). This is exemplified by the utilization of TBP as a component of SL1, the role of antioncogenes, such as Rb in regulating rDNA transcription, and the involvement of TFIIH.

Introduction Protein synthesis is an essential process for all living cells. Cells must govern both the amounts of specific proteins synthesized (differentiation specific proteins) as well as the total protein (“housekeeping proteins”) synthesized in response to environmental signals and internal programming.1-5 In cycling cells, this coordination insures successful cell division and daughter cell survival. Alternatively, during terminal differentiation or in response to environmental stress, a cell may withdraw from the cell cycle. In many cases, this reduces the need for protein synthesis. The protein synthetic capacity of a cell is dictated by a number of processes such as mRNA availability, efficiency of translation, availability of translation factors and the number of ribosomes. The evidence accumulated to date indicates that overall the protein synthetic capacity is determined by the steady state number of ribosomes. This in turn is dictated by the relative rates of ribosome synthesis and degradation.1-6 Ribosome synthesis or biogenesis (Fig. 1) is a complex process dependent on the coordinated synthesis of approximately 85 ribosomal proteins, four ribosomal RNAs (rRNA), and their subsequent processing and assembly into mature ribosomes. In contrast, little is known about the process or regulation of ribosome degradation.1-7 In the majority of cells, ribosomes are relatively stable thus, their cellular content depends largely on the rate of ribosome biogenesis. Experimental evidence so far correlates regulation of ribosome biogenesis to altered rates of rRNA transcription rather than changes in rRNA processing or stability.2-4,7 Ribosomal DNA (rDNA) transcription is a major commitment for the cell since it accounts for approximately 40-60% of all cellular transcription and 80% of the steady-state cellular RNA content. The rate of rDNA transcription can vary over a wide range. For example, when Acanthamoeba castellanii encyst, rDNA transcription ceases and decreases from 75% of the total cellular transcription.3,8 Indeed rDNA transcription is regulated in The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Ribosomal DNA Transcription in Mammals

89

Figure 1. Summary of ribosome biogenesis and rDNA transcription.

response to different stages of development or cell cycle, nutritional state, and altered environmental or hormonal conditions.1,7,8-18 This illustrates that rDNA transcription, like the expression of cell-cycle specific genes, is a prime example of growth-regulated gene expression. Present data suggest that cells can utilize a diverse array of mechanisms to coordinate the rate of rDNA transcription with altered cellular requirements for protein synthesis. The relative importance of the various mechanisms to a specific stimulus has not been thoroughly investigated in any single cell. However, the data suggests that these mechanisms tend to be both cell type and stimulus dependent. In many cases the exact molecular mechanism(s) and signaling pathways involved in regulating rDNA transcription are not well understood. Since the regulation of rDNA transcription is a critical component of cellular homeostasis, it is important for us to understand and characterize the process.

General Background The Nucleolus The interphase nucleus contains varying numbers of nucleoli. In metazoans, the nucleolus is the site of 45S rRNA synthesis, i.e., transcription of the ribosomal genes (rDNA), rRNA processing and ribonucleoprotein (RNP) assembly.7,19 The only active genes in the metazoan nucleolus are the rRNA genes, and the only RNA polymerase is RNA polymerase I (RNA pol I). Indeed the ribosomal genes are the central elements of the nucleolus. In eukaryotic organisms, 5.8S, 18S and 28S rRNA are produced from a 35S-47S pre-rRNA that is transcribed by RNA polymerase I. Ribosomal genes (ribosomal DNA) are moderately repeated and organized in tandem arrays that are visualized on some chromosomes as secondary constrictions. It was noted that the number of secondary constrictions correlated with the number of nucleoli that formed and thus they were termed nucleolar organizer regions (NORs) by McClintock in 1934.20,21 The Xenopus genome contains 450 copies of the rDNA repeat in a single NOR, in Drosophila there is only 1 NOR on each X and Y chromosome. The 400 copies of the human rDNA repeat are distributed on the short arms of the five acrocentric chromosomes of the D and G groups (chromosomes 13-15,21,22). The ribosomal RNA genes appear to be the only genes found on the acrocentric p-arms. Sequences both distal and proximal to human NORs are comprised of satellite and telomeric DNA packaged as heterochromatin.22-25

90

The Nucleolus

Surrounding the NOR is a fine network of filaments which forms a scaffolding distinguishable in both organization and composition from that of the “nuclear matrix”. The scaffolding is thought to provide “structural support” or organization to the arrangement of transcriptionally active rDNA and/or the assembly and transport of ribosomal subunits. This is supported by the observation that the nucleolar scaffolding is absent from cells which are inactive in rRNA synthesis such as nucleated erthyrocytes and spermatocytes. Typically, mammalian nucleoli consist of three substructures named according to their appearance in transmission electron microscopy; i) Fibrillar Centers (FC); ii) Dense Fibrillar Component (DFC); and iii) Granular Component (GC).7,26 The FC are pale staining regions in the center of the nucleoli consisting of a fine fibril (4-8 nm thick) network which is relatively opaque in the electron microscope.26 The rDNA, RNA polymerase I, and other components of the rDNA transcription system such as UBF, SL1 and topoisomerase I have been localized to the periphery of this region.20,26-29 Thus, it is likely that the FC are the site where the primary rRNA transcript is generated. The DFC surrounds the FC and is characterized by densely packed fine fibrillae (3-5 nm thick), a high electron microscope contrast and a high content of a 34 kDa protein, fibrillarin.7,26 Fibrillarin is known to associate with ribonucleoprotein complexes required in the early stages of rRNA processing, such as the U3, U8 and U13 snoRNP (small nucleolar RNP).30 The GC is localized to the periphery of the nucleolus and consists of granular structures ranging in diameter from 10 to 15 nm, which are sometimes organized in short strings.26 The later stages of maturing ribosome precursor particles, before they are exported to the cytoplasm, have been localized to this region.26 The boundaries of these substructures are not always discreet, in fact three different patterns of compartmentalization have been described and these are used to classify nucleoli. The type of nucleoli identified depends on the rate of ribosome production. Typically cells with a high rate of ribosome production, such as nerve and Leydig cells, have large and complex nucleoli described as compact or reticulate. Alternatively, cells with a lower rate of ribosome biogenesis, such as monocytes and lymphocytes, exhibit ring-shaped small nucleoli and a single FC.26 A diploid human cell contains 10 NORs, thus it would be expected to have 10 nucleoli. However this is seldom the case. This discrepancy could be explained by two situations: i) not all NOR are active; or ii) more than one NOR can be included in a nucleolus. Both situations have been identified. For example, in some cells not all NOR’s are active.20,26-28 In human Hep-2 cells the transcription factor UBF associates with only six-to-eight of the possible ten NOR and in PtK1 cells UBF is found in 50% of the NOR’s.28 In each case, upon cell division there is an equal apportionment of UBF to the daughter cells. Alternatively, it has been shown that when lymphocytes are activated the previously inactive NOR fuse with the existing functional nucleoli.26 During mitosis, as the cells enter prophase, the nuclei and nucleoli undergo rapid changes. For example the nuclear envelope disintegrates, chromosomes condense and subsequently the spindle apparatus forms. In addition, the nucleoli disperse and disappear.26 At least a portion of some nucleolar components, such as RNA polymerase I, SL1, UBF, and topoisomerase I, remain associated with the NOR,7,20,27-29 while others are released, such as NO3831 and the snoRNP.30,32 Nucleolar reformation usually begins during telophase with the daughter nucleoli forming at the NOR. Complete restoration of nucleolar morphology requires both ribosomal chromatin and active rDNA transcription.20 Thus, those NOR containing the RNA polymerase I transcription apparatus are more quickly able to initiate rDNA transcription and contribute to nucleolar regeneration.20,27-29 In general, chromosomal DNA is organized in nucleosome structures. However, from electron microscopy, it has been suggested that the rRNA chromatin does not form a typical compact nucleosome structures. In fact, some reports suggest there are no nucleosomes on the transcribed rDNA.33,34 As one might expect, the nuclease digestion and psoralen cross-linking properties of the rDNA are atypical. This has also been examined using topoisomerase I digestion to examine the nucleoprotein structure of the rDNA. Topoisomerase I digestion sites were

Ribosomal DNA Transcription in Mammals

91

Figure 2. Schematic depiction of a mammalian ribosomal DNA repeat: The top portion of the cartoon depicts one and one-half ribosomal repeats in tandem, including the terminator Sal box, intergenic spacer, repetitive elements, enhancer region and the region transcribed to yield 45S rRNA. A section of the repeat is enlarged in the bottom portion of the cartoon. This section illustrates the placement of the spacer and 45S rDNA promoters, the proximal (To) and downstream promoter terminator elements (T1-7), the transcription initiation site (+6) and the external transcribed spacers (ETS) and the internal transcribed spacers (ITS).

found to be spaced with a periodicity of 200 bp and concentrated in the regions encoding the 18S, 5.8S and 28S rRNA.33 This pattern was due to binding of nuclear proteins to the rDNA and not dependent on the DNA sequence itself.33,35 UV laser-induced histone-DNA cross-links studies demonstrated that the rDNA coding sequence, spacer enhancer and spacer promoter were associated with histones in both transcriptionally active and inactive cells.36,37 Interestingly, a recent study38 suggested that the nucleosome structure might play a role in the regulation of initiation complex formation on the rDNA. That study demonstrated that histone octamers could compete with the transcription factors for the rDNA promoter, but only if the DNA was not first bound with an initiation complex.38 However, to date, the picture of the organization of the rDNA chromatin and its function(s) is unclear.

Synthesis and Assembly of Ribosomes The synthesis of ribosomes requires the coordinate effort of all three DNA-dependent RNA polymerases.7,19 RNA polymerase I, in the nucleolus, transcribes the rRNA gene that encodes the 45S precursor of the 18S, 5.8S and 28S rRNAs (Fig. 2). The 45S precursor rRNA is neither capped nor polyadenylated and can account for 30-60% of all nuclear RNA synthesis. To a lesser extent RNA polymerase I transcribes another transcript which originates from the spacer promoter located in the intergenic spacer. However, this second transcript is unstable and its function is yet to be established.4 RNA polymerase III, in the nucleus, transcribes the 5S RNA gene.7 RNA polymerase II, in the nucleoplasm, transcribes numerous genes encoding ribosome associated proteins (r-proteins). These mRNAs are transported to the cytoplasm, translated and the mature r-proteins returned to the nucleolus for assembly of the ribosome

92

The Nucleolus

components.7 Together, ribosome biogenesis can account for as much as 80% of the nuclear RNA synthetic capacity. Mammalian ribosomal subunits are assembled in discrete stages within the GC of the nucleolus. Initially the 45S precursor rRNA is processed via a complex series of specific exo- and endo-nucleolytic cleavages. The rRNA exons are not spliced together, thus the 45S precursor generates the 18S, 5.8S and 28S rRNAs. rRNA processing is directed by snoRNPs such as nucleoli U3 snRNP. U3 snRNP has been implicated in several steps, including the earliest step in rRNA processing, the cleavage at -650 in the 5' external transcribed spacer (ETS) (Fig. 2).7 The 18S, 28S and 5.8S rRNAs associate with the 5S rRNA and r-proteins to form a complex referred to as the 80S preribosome. The 80S preribosome is further processed to generate the 40S and 60S ribosomal subunits. Studies have shown that the order of r-protein addition in this process is essential for successful assembly of the ribosomal subunits. For example, a decrease in the cellular content of the r-proteins L13 or L16 can result in a deficiency of the 60S ribosomal subunit.7 The 40S and 60S ribosomal subunits are transported to the cytoplasm, where they the final stages of maturation occur. This involves the association with additional proteins, such as initiation factors. They are then able to participate in translation. The accumulation of mature ribosomes in the cell depends on the balance between the rate of subunit synthesis and the rate of degradation. Since mature ribosomes are fairly stable complexes, with half lives ranging from 4.5 days in rat liver to more than 10 days in cultured L cells, ribosome degradation is not considered to contribute significantly to the regulation of ribosome content.2 To date little is known about the signaling mechanism(s) involved in ribosomal degradation, although a recent publication implicated a role for ubiquitin in this process.39 Ubiquitin may function by binding to the ribosome, thereby stabilizing or protecting it from degradation. Subsequent removal of ubiquitin would signal the cell to degrade that ribosome.7,39 Interestingly, in the majority of cells, it is the rate of ribosome synthesis, i.e., rDNA transcription, that is the primary determinant of ribosome content.2

rDNA Transcription Essential components required for efficient rDNA transcription include the rRNA genes, RNA polymerase I, RNA polymerase I associated factors and a number of rDNA specific trans-acting factors such as SL1, the paralogue of TFIID, and UBF.18 In addition, other proteins have been reported to be components of the transcription initiation complex and may participate in the regulation of rDNA transcription.18 The specific contributions of these factors to the regulation of rDNA is poorly understood. However, they present interesting links between transcription by RNA polymerase II and RNA polymerase I.

The rRNA Genes There are approximately 150-200 copies of mammalian rRNA genes (rDNA) present per haploid genome. In general, the genes are distributed among several chromosomes and arranged in tandem, head to tail arrays with the coding regions of the primary transcript being separated by nontranscribed or intergenic spacer regions (Fig. 2). The length of the transcript generated from the rDNA varies from ~8 kb (yeast, Drosphila and Xenopus) to ~13 kb (mammals). For the most part, this increase in length reflects the lengths of the external and internal transcribed spacer regions.2,7 Although examination of the sequences of the rDNA promoters of different genera fails to demonstrate significant sequence identity, there is a high degree of conservation between the functional elements.3 In fact, the yeast, human, mouse, frog and rat rDNA promoters all share a similar molecular anatomy.4 In addition to the promoters, the vertebrate rDNA repeats contain terminator elements as well as additional transcription elements within the intergenic spacer. Surprisingly, the nontranscribed spacer of the yeast rDNA repeat also functions in termination and the anatomy of the yeast promoter is more similar to those of the vertebrate rRNA genes than is the Acanthamoeba rDNA promoter.

Ribosomal DNA Transcription in Mammals

93

Figure 3. Schematic depiction of the 45S rDNA promoters and the initiation complexes of the S. cerevisiae, A. castellanii, and vertebrate rDNA promoters.

rDNA Promoter The consensus of structure/function studies of the promoters of the eukaryotic ribosomal genes is that despite species-specific variations, the promoters contain two functionally homologous elements (Fig. 3, and reviewed in refs. 1-3,8). Eukaryotic rDNA promoters consist of two domains, a core promoter element (CPE) and an upstream promoter element (UPE) also referred to as the upstream control element (UCE). In yeast, the the 5' boundary of the core promoter extends to –42 and the 5' boundary of the upstream element lies between –150 and –101. In vertebrates, the core promoter element extends from ~-31 to +6, with respect to the transcription initiation site, and the upstream promoter element (UPE) extends from the CPE (-30) to ~ -167 (Fig. 3).3,4,6,7,18,40 The core promoter (-40 to +6) is required for the correct initiation of transcription. The CPE is necessary and sufficient for in vitro transcription, and is required but not sufficient for in vivo transcription. The second element, referred to as either the upstream promoter element (UPE) or the upstream control element (UCE), is not necessary for transcription initiation in vitro. However, its presence in cis results in elevated levels of transcription from the CPE under stringent conditions in vitro and it is required for transcription in vivo.1-4 This may reflect a requirement to compete with the endogenous full length promoters. However, studies in yeast suggest that the intertaction between the UPE and CPE is essential for transcription in vivo.41,42 a. The Core Promoter Element (CPE) appears to consist of functional sub-elements. Results from Muramatsu’s laboratory43 suggest that the region of the mouse promoter from -14 to -32 is sufficient to impart species-specific transcription. When we compared the rat spacer and 45S promoters, nucleotides -6 to -18 of both promoters were identical.44 That same element is found in the 45S and spacer promoters (i.e., pol I promoters located more than 600 bp upstream of the transcription start site) of mice and Chinese hamsters (Fig. 4). All vertebrate ribosomal genes have guanines at positions -16 and –7.1-3 Although mutations of the guanines have serious effects on the efficiency of rDNA transcription in vitro, these mutations, particularly –7, are much more striking in vivo than in vitro.1,44 Transcription from the CPE occurs without the formation of a stable committed template, and the experimental evidence suggests that the UPE is essential for the formation of the stable committed template complex in vitro.45,46

The Nucleolus

94

Figure 4. Duplication of part of the CPE in the UPE and in the spacer promoter.

b. The Upstream Promoter Element (UPE) is required for efficient transcription (reviewed in 1-3). The 5' boundary of this domain of the rat promoter lies between -147 and -127, depending on the assay used (promoter competition or promoter efficiency). This region is also required for the formation of one class of stable preinitiation complex.45 At least two factors interact with the region between -150 and -50 (Fig. 3). There is some evidence that the UPE and the CPE must be aligned stereochemically47 for efficient transcription in vivo or in vitro, suggesting an interaction between the protein-DNA complexes formed on the UPE and those that formed on the CPE. Complementary experiments from the laboratories of George Harauz and Bazett-Jones and Moss48,49 demonstrated that UBF could “coil” the promoter resulting in the juxtaposition of the SL1 binding sites in the UPE and CPE. Studies using deletion, point, and linker scanning mutants have demonstrated that the UPE is important for transcription but have failed, with one or two exceptions, to identify critical nucleotides within the UPE.

Experiments using distant-altering mutations demonstrated an interesting relationship between the UPE and CPE of the mouse rDNA promoter. Pape et al50 demonstrated that altering the spacing between the UPE and the CPE of the Xenopus rDNA promoter allowed that promoter to be transcribed efficiently by mouse extracts, violating the paradigm that rDNA transcription is species-specific. In addition, distant altering mutants of the rat rDNA promoter suggested that the distances between the UPE and CPE were critical for initiation. In fact, several sets of distance-altering mutations, demonstrated a sinusoidal relationship (Fig. 5) between the efficiency of transcription and the distance between the UPE and CPE. However, this response was not uniform across the entire UPE, suggesting that different segments of the UPE must have different functional roles, or are “neutral” with respect to their role in the structure of the preinitiation complex.51 To date, the results from published studies are consistent with a model in which the protein complexes that form on the UPE interact with and possibly stabilize those complexes bound to the CPE. However, the mechanism by which this is accomplished is unknown. For example, the model presented in (Fig. 3) contains two molecules of SL1 and one UBF dimer. While nucleotide analysis, footprinting experiments and mutagenesis experiments indicate that SL1 binds both to the UPE and to the CPE, there are no direct experiments analyzing the stoichiometry of the proteins in the preinitiation complex. At least two transcription factors have been shown to interact with the mammalian promoter elements, UBF and SL1.44,47,51,52 Interestingly, although the model by which transcription intiation is accomplished on vertebrate promoters is physically different than that proposed for Saccharomyces promoter, it is biochemically similar (discussed below).

Ribosomal DNA Transcription in Mammals

95

Figure 5. Relationship between transcription (Activity) and changes in distance when two cis-acting promoter elements must be on the same side of the helix.

Intergenic Spacer The intergenic spacer lies between the transcribed regions and is bound at both ends by transcription termination signals.2,4,7,18 In Xenopus laevis the intergenic spacer is punctuated by 2-7 spacer promoters and in turn these are separated by six to twelve 60 and 81 bp directly repeating elements.53 The spacer promoter is almost a perfect duplication of the rRNA promoter with as high as ~ 90% homology in the regions -145 to +4, and in the imperfect copy of a 42 bp sequence (active core), that localizes to the -72 to -114 region of the gene promoter.54 However, in rat and mouse the spacer and 45S promoters contain only one conserved block of 12-13 bp which includes the G’s at -7 and –16.3,54 The spacer promoter is transcribed by RNA polymerase I producing a transcript which terminates just upstream of the rRNA promoter, ~ -167 bp. In Xenopus, the spacer promoter may also enhance transcription from the gene promoter, possibly by delivering RNA polymerase I to the gene promoter.55 However, other studies contradict this observation.54 The intergenic spacer of Xenopus contains additional repetitive elements. The most notable of these are the 60 and 81 bp repetitive elements which are homologous to a portion of the 45S promoter. The 81 bp elements are identical to the 60 bp elements except they contain an additional 21 bp of unique sequence.53 The 3' end of the intergenic spacer (-2300 to -3950), i.e., the region near the 3' end of the 45S rRNA transcript, is a region that shows little homology with either the spacer promoter or the 60/81 bp elements. It consists of at least two repetitive elements (repeats 0 and 1) and some nonrepetitive elements.54 The intergenic spacers of Xenopus, yeast, Drosophila, mouse and rat, contain elements which enhance transcription from their “major” promoters.3,4,7 In Xenopus, the cis-acting 60 or 81 bp repeat elements, enhance transcription from both the 40S preribosomal RNA and the spacer promoters. In this case the rate of transcription has been shown to be directly proportional to the number of repeat elements and independent of their orientation or distance from the promoter.4,54 Such characteristics are typical of enhancers described in RNA polymerase II transcription. Other enhancer elements have been reported in the rat and mouse intergenic spacers, including the 130 bp element which comprises the variable region of the rDNA repeat47 and the 37 bp enhancer motif localized in the rat 174 bp nonrepetitive region which is able to enhance both RNA polymerase I and RNA polymerase II transcription.3 Additional 140 bp and 200 bp element have also been identified in rat and yeast, respectively.3,56 To date the mechanism by which these elements enhance rDNA transcription is not clear. The repeated

96

The Nucleolus

elements in the mammalian intergenic spacers and the 60/81 bp repeats of the Xenopus spacers have been shown to bind UBF and to act across species.56-58 However, it is not clear if UBF is the only factor that binds to the repeated elements and if UBF is solely responsible for enhancer activity.

Terminators At the 3' end of the primary transcript of mammalian 45S rRNA genes lie several copies of a 17 bp motif, referred to as the Sal box or terminator (Fig. 2). The Sal box functions as orientation dependent terminators of transcription.2,4,18,19,59,60 The 13 bp promoter proximal terminator (T0 ) located ~ -167 bp +1 is also a Sal box (Fig. 2). In both cases, the terminator elements act as binding sites for the 105 kDa RNA polymerase I transcription termination factor, TTF-1.59,61 TTF-1 binds to DNA in a polymerase specific but not a species-specific manner. This suggests that TTF-1 once bound to the terminator site acts by interacting with one of the unique subunits of RNA polymerase I.61 Interestingly, recent studies indicate that TTF-1 can associate with RNA polymerase I in the absence of DNA.62 In vertebrates, the process of transcription termination requires two steps: i) RNA polymerase I pausing and its subsequent release; and ii) release and processing of the 3' end of the prerRNA.4,19 Mammalian transcription termination requires TTF-1 for the pausing of RNA polymerase I ~11 bp upstream of the Sal box. Interestingly, the second step in termination requires a T-rich element upstream of the TTF-1 binding site and a releasing factor.59,63 In Xenopus a terminator factor63 that binds to the T3 box in the intergenic spacer has been identified as Rib2.63,64 Yeast rDNA repeats contain unique termination elements, which are comprised of two domains. The first domain consists of an 11 bp element, sometimes referred to as a REB1 element, which serves as the binding site for Reb1p.2,65,66 Transcription termination for yeast RNA polymerase I also requires about 46 bp of T-rich 5' flanking sequence. It has been suggested that the Reb1p-DNA complex comprises a pause element, while the 5' flanking sequence contains a release element. In contrast to the TTF-1-DNA complex, the Reb1p-DNA complex is not specific for polymerase I.66-68 The terminators may also serve secondary functions. The promoter proximal terminator element (T0 ) not only serves to terminate transcripts originating from the spacer promoter, but it may also activate transcription. It has been suggested that T0 may function to prevent promoter occlusion, a phenomemon whereby transcription through a promoter disrupts the preinitiation complex.3 Thus, in the absence of TTF-1, the transcription factors SL1 and/or UBF would be displaced from the rDNA by a polymerase that was transcribing through the promoter. Recent studies suggest that TTF-1 may contribute both to transcription activation and to the “silencing” of ribosomal genes. Silencing requires the TTF-dependent recruitment of a nucleolar remodeling complex (NoRC) and the subsequent recruitment of DNA methyltransferase and histone deacetylase (HDAC) and the resulting inactivation of the ribosomal chromatin.69,60,70-74 Transcription activation appears to be dependent on the reorganization of the rchromatin and involves repositioning nucleosomes in an ATP dependent fashion.35,69 However, the exact mechanism by which TTF-1 catalyzes these functions is unknown. It has also been proposed that TTF-1 may function in DNA replication, as it results in the arrest of replication fork movement and thus directs DNA replication in the same direction as transcription.70

Proteins Involved in rDNA Transcription RNA Polymerase I The core mammalian RNA polymerase I is a large, complex enzyme with a total approximate Mr of 500- 600,000. Historically, the number of subunits ranges from 11 subunits and 2-3 associated factors to only 2 large and 3-4 smaller subunits depending on the purification

Ribosomal DNA Transcription in Mammals

97

Figure 6. Evolutionary relationships of the high molecular weight subunits of the eukaryotic RNA polymerase complexes.

method implemented.75-79 Two recent studies, employing different purification schemes, report that mammalian RNA polymerase I is composed of at least 12 subunits with 2 or 3 associated factors (PAFs).80,81 In contrast, yeast RNA polymerase I has been subject to a detailed series of genetic and biochemical studies, and fourteen subunits have been identified and cloned.75,77,82 The subunits of yeast RNA polymerase I are classified into three groups*: i) four core subunits: β’-like (A190), β-like (A135) and two which are similar to the bacterial α subunits (AC40 and AC19); ii) five subunits common to all three RNA polymerases: ABC27, 23, 14.5, 10a, 10b; and iii) five specific subunits: A49, 43, 34.5, 14, 12.2.75 There is a large degree of sequence conservation between the homologous mammalian and yeast homologous RNA polymerase I subunits and also between the RNA polymerases I, II and III subunits themselves.75 Interestingly, the β and β’ subunits of yeast RNA polymerase I are more identical to the β and β’ subunits of rat RNA polymerase I than they are to the β and β’ subunits of yeast RNA polymerase II (Fig. 6). The majority of the yeast RNA polymerase I subunits are essential for growth especially the five that are shared between the three different polymerases and the two that are shared between RNA polymerases III and I.75 However, the A34.5 and A49 subunits are not strictly essential for cell growth. For example, mutations of A49 generate slow growing colonies with only reduced RNA polymerase I activity illustrating that it is important, but not essential for cell viability.75 However, the consideration of whether or not a subunit is essential for viability, should also be considered in terms of whether or not such mutants could coexist with wild-type *The yeast genes for the subunits of the RNA polymerases are referred to by a letter that refers to one of the three nuclear RNA polymerases in combination with the molecular weight of the protein. A, refers to a subunit of RNA polymerase I; B to a subunit of RNA polymerase II; and C refers to a subunit of RNA polymerase III. As some of the subunits are components of one or more polymerase, they can have multiple letters in their designations

98

The Nucleolus

constructs under selective conditions. Identification of the specific functions of the RNA polymerase I subunits has been limited and restricted mainly to the yeast system. Experimental evidence to date demonstrated that the A190 and A135 subunits cross-link to nascent chain RNA and contain putative Zn 2+ fingers. Other subunits containing putative zinc binding domains include A12.2, ABC10a and ABC10b. Independent studies have suggested that Zn2+ binding may be essential for activity and/or the structural integrity of RNA polymerase I.75 In order for functional RNA polymerase I to initiate transcription it must recognize and bind the transcription initiation site. The two highest molecular weight subunits and the ABC23 subunit have been implicated in this process. However, the domain(s) involved in this process are yet to be defined.75 Furthermore, the β subunits contains a putative nucleotide binding domain suggestive of a role in elongation.75,83 In addition, studies suggest that β’ may also play a role in elongation since resistance to α-amanitin, a drug which interferes with chain elongation, maps to the β’ subunits of RNA polymerase II.75,84

RNA Polymerase I Associated Factors In order for RNA polymerase I to engage in specific and productive transcription it must interact with the transcription factor-DNA complex that defines the committed template. Results from several laboratories and in different biological systems indicated that this interaction is mediated by one or more polymerase associated factors, that themselves were subject to regulation. To date a number of proteins such as TFIC, Factor C*, TIF-IA, TIF-IC and PAF5381,85-87 have been shown to closely interact with RNA polymerase I and be required for transcription. In addition, there is evidence suggesting that RNA polymerase I itself may interact with the transcription factor UBF. One study has demonstrated that UBF interacts with a 62kDa subunit of murine RNA polymerase I in vitro.88 However, RNA polymerase I purified by another group did not contain a 62 kDa subunit, and that laboratory reported an interaction between UBF and the 180, 114 and 44 kDa subunits of mouse RNA polymerase I, as well as with PAF53 in vitro.77,81 The association of RNA polymerase I with PAF53 has been confirmed.80,89 However, the interaction between RNA polymerase I and UBF has proven more problematic.80,90 The reasons for these disparities are unclear and further investigation is required. TFIC, TIF-IA and Factor C* Studies on the role(s) of the polymerase associated factors in rDNA transcription have been intimately related to studies on the regulation of rDNA transcription. The members of this interesting, and still poorly understood, class of transcription factors are present in very low levels in the cell, and associated with only a fraction of the RNA polymerase I molecules. Thus, biochemical studies that would lead to the differentiation of the various factors and an understanding of their roles in transcription have been very difficult. In the early 1970’s Feigelson and colleagues91 reported that cycloheximide caused a rapid cessation of nucleolar RNA synthesis (ribosomal DNA transcription) and concluded that a rapidly turning over protein was required for RNA polymerase I activity in vivo. Subsequent studies identified TIF-IA and Factor C* as those factors that were required for the complementation of extracts of quiescent or cycloheximide treated cells. In fact, at least three, and possibly more, polymerase associated proteins, TIF-IA, Factor C*, and TFIC,86,87,92 have been demonstrated to contribute to the regulation of rDNA transcription. TFIC was identified as that activity reconstituting transcription by extracts of glucocorticoid treated P1798 cells.86 This lymphosarcoma cell line exits the cell cycle in response to the synthetic glucocorticoid dexamethasone (DEX). Interestingly, TIF-IA, Factor C* and TFIC shared several properties, including a tight association with the core polymerase.87,92,93 Moreover, Factor C* and TFIC were demonstrated to be heat stable, i.e., heat-treated (45oC, 15 min) S100 extracts could reconstitute transcription when added to an extract from dexamethasone (DEX) treated P1798 cells or cycloheximide treated cells. Additional studies suggested that while not critical for the formation of a stable preinitiation complex, all three factors are required for the formation of the first phosphodiester bond of nascent prerRNA.94,95

Ribosomal DNA Transcription in Mammals

99

However, there were, and are, arguments against the possibility that TIF-IA, Factor C* and TFIC represent the same factor. For example, mouse TFIC activity copurifies with three polypetides present in a stoichiometric ratio of 1:1:1, with approximate molecular masses of 55, 50 and 42 kDa.93 In contrast, TIF-IA is a single 75 kDa polypeptide.94 There is also the question of functionality. For example, it has been reported that TIF-IA can be liberated from the initiating complex and recycled to facilitate transcription from other templates.94 In contrast, Factor C* or TFIC functions stoichiometrically in vitro,92,95 i.e., it can activate one round of transcription and is then “used up.” Interestingly, a factor, with properties similar to the mammalian polymerase associated factors, was identified in yeast. This protein, Rrn3p, has been shown to interact directly with RNA polymerase I, independent of DNA, and with CPF.96 Originally, it was suggested that Rrn3p stimulates the recruitment of the polymerase to the stable complex containing the rDNA promoter, and the Rrn6/7/11 and the Rrn5/9/10 complexes,96 and there was speculation that that TFIC, TIF-IA and Factor C may well represent the same biological component. In fact, subsequent studies demonstrated that there was a human homologue of Rrn397 and that human Rrn3 was the same as TIF-IA.98 Experiments reported by Cavanaugh et al99 demonstrated that Rrn3 and TFIC were different factors (described below), and more recently Hirschler-Laszkiewicz and her colleagues have demonstrated that Rrn3 functions stoichiometrically in transcription,100 eliminating one apparent difference between Rrn3/TIF-IA and Factor C*. Thus, it would seem that at least two of the polymerase associated factors, Rrn3/TIF-IA and TFIC function stoichiometrically. Both the yeast and mammalian forms of Rrn3 interact with the rpa43 subunit of RNA polymerase I,99 and with components of either the yeast CPF,101 or SL1.99,102 This has led to the hypothesis that Rrn3 plays a role in the recruitment of RNA polymerase I to the committed template.103 However, it is not clear if Rrn3 is essential for recruitment or if it facilitates recruitment. For example, Schnapp et al94 reported that “Preinitiation complexes can be assembled in the absence of TIF-IA, but formation of the first phosphodiester bonds of nascent rRNA is precluded.” They also reported that after initiation, TIF-IA is liberated from the initiation complex and facilitates transcription from templates bearing preinitiation complexes (PICs) which lack TIF-IA. Aprikian et al104 also found that RNA polymerase I could be recruited in the absence of Rrn3, but that the transcriptional activity of such PICs could not be rescued by the addition of purified yeast Rrn3. At face value, the two reports would appear to contradict one another. However, the experiments of Aprikian et al were carried out with immobilized templates, and included washes between the various steps before transcription and would have precluded the exchange or displacement of factors from one template to a second. In contrast, the experiments reported by Schnapp et al were done using templates in solution. Under those conditions, it is possible that Schnapp et al may have actually observed displacement or the formation of new PICs when Rrn3 was added to the reactions. Thus, the question of whether a PIC formed in the absence of Rrn3 can be converted to a functional PIC remains unanswered. However, both sets of experiments demonstrate that RNA polymerase I can be “recruited” to the committed template in the absence of Rrn3. Interestingly, both Milkereit and Tschochner and Hirschler-Laszkiewicz et al have demonstrated that Rrn3 dissociates from RNA polymerase I as the result of transcription.100,101 This then suggests that the “reassembly” of RNA polymerase I must occur before the polymerase can initiate transcription. The question then is whether the reassembly can occur on the committed template, i.e., recruitment in the absence of Rrn3, or if reassembly must occur before the polymerase is recruited. This suggests that there are interactions between RNA polymerase I and the assembled transcription factors in addition to those mediated by Rrn3. The question then is to identify the proteins that mediate those interactions and to determine the role, if any, they play in the formation of a transcriptionally competent preiniation complex.

100

The Nucleolus

TIF-IC TIF-IC was first identified as a 65kDa105 factor associated with RNA polymerase and is required for the assembly of the initiation complex, formation of first internucleotide bond and chain elongation.105 TIF-IC contributes to the chain elongation by stimulating elongation and suppressing RNA polymerase I pausing. TIF-IC also inhibits nonspecific initiation and supports the synthesis of full-length, run-off transcripts.105 However, as with several of the above RNA polymerase I associated activities, there are no antibodies to TIF-IC and this factor has not been cloned. This precludes a more complete analysis of its contribution to rDNA transcription. PAF 53 Recently, three proteins that copurify with RNA polymerase I in substoichiometric amounts were isolated and cloned from mouse cells, Polymerase Associated Factor 53 (PAF53), PAF51 and PAF49.81 All three factors are tightly associated with RNA polymerase I but dissociable under certain purification conditions, indicating that they are probably not core subunits of this enzyme. PAF53 and PAF51 are structurally related proteins since they are both recognized by anti-PAF53 antibodies. It is as yet unknown if PAF51 is a degradation product, an alternatively spliced isoform, or a post translational modification of PAF53. PAF49 however, is not detected by anti-PAF53 antibodies thus it appears to be a distinctively different protein.81 PAF53 is associated with RNA polymerase I purified from exponentially growing 3T3 cells, but not with RNA polymerase from quiescent NIH3T3 cells. In addition, antibodies to PAF53 block specific, but not random, transcription from the rDNA promoter. These observations suggest that PAF53 is not involved in template binding, nucleotide incorporation, polymerization activity or elongation. Instead they suggest that PAF53 is required for initiation of specific transcription from the rDNA promoter.81 In vitro studies indicate that PAF53 has the potential to interact with the transcription factor, UBF,81 suggesting a role for this factor in the recruitment of RNA polymerase I to the initiation complex. However, the mechanism by which PAF53 contributes to the regulation of rDNA transcription remains to be elucidated. Moreover, there is some question as to whether PAF53 is an essential component of the transcription apparatus. Firstly, PAF53 shares significant amino acid identity to the A49 subunit of yeast RNA polymerase I a subunit that is not essential for survival.106 Secondly, although Hanada et al81 demonstrated a decrease in the nucleolar content of PAF53 in serum starved cells, Seither et al107 reported that “the relative level of PAF53 was comparable in exponentially growing or growth-arrested cells, indicating that growth-dependent fluctuations in Pol I activity are not accompanied by alterations in the amount of PAF53.” There are counterarguments to both points. First, as mentioned previously, deletion of the A49 gene significantly inhibits cell growth, indicating an important if not necessary function. Moreover, when we examined the regulation of PAF53, we found a heterogeneous pattern of regulation of its level (Fig. 7). When NIH 3T6 cells were serum-starved, there was a significant decrease in the amount of PAF53 in the cells (Fig. 7 and ref. 108) that was not found when NIH 3T3 (Fig. 7, lanes 1 and 2), HEK 293 or CHO cells (data not shown) were serum starved. Further, we found that the levels of PAF53 in both NIH3T6 and H4-IIE-C3 cells correlated with the level of rDNA transcription in those cells.108 Clearly, more work remains before we understand the role of PAF53 in rDNA transcription.

rDNA Trans-Acting Factors There are at least two trans-acting factors required for efficient transcription of rDNA by RNA polymerase I. In mammals, they are referred to as SL1 (selectivity factor 1) and UBF (upstream binding factor). Studies in Acanthamoeba, described below, have unambiguously identified a multimeric complex, TIF-IB, with properties and a functional role similar to SL1. Interestingly, studies on transcription by yeast RNA polymerase I, have identified two complexes, described below, with properties similar to what might be considered to be a combination of SL1 and UBF. Briefly, SL1 is absolutely required for rDNA transcription in vitro.109,110

Ribosomal DNA Transcription in Mammals

101

Figure 7. The significantly decreased levels of PAF53 in serum starved 3T6 cells but not 3T3 cells indicates that PAF53 may be subject to different mechanisms of regulation in different cell lines and types.

In contrast, UBF is not absolutely required for specific initiation on the rDNA promoter in vitro, although its addition to UBF-depleted extracts increases the efficiency of in vitro transcription in a dose dependent manner.111-113 In addition, overexpression of UBF1 in cell lines or primary cultures of cardiomyocytes is sufficient to directly increase transcription of a reporter for rDNA transcription,114 as well as the endogenous rRNA genes.115 Factors Which Bind to the Core Promoter Element There is evidence that several of the rDNA transcription factors may interact with the core promoter element. However, the experimental evidence accumulated suggests rather strongly that the yeast CPF, and its paralogues in other systems, must be considered the primary factor that interacts with the CPE. Saccharomyces cerevisiae Core Factor Genetic studies in yeast demonstrated that TATA-binding protein (TBP), the highly studied component of TFIID, was involved in transcription by RNA polymerase I and III, before biochemical studies had demonstrated that it was a component of any of the factors which interacted with the core promoter elements of the rDNA.116-121 Subsequent biochemical and genetic experiments confirmed that TBP was a component of the human, mouse, yeast, and Acanthamoeba rDNA transcription systems. Biochemical and genetic studies in yeast provided evidence for two transcription factors, referred to as UAF (upstream activation factor) and CF (core factor). UAF (discussed below) is a multiprotein transcription factor which consists of at least five proteins. Both biochemical and genetic analyses confirm that CF is also a multiprotein transcription factor, and it consists of at least three proteins, Rrn6p, Rrn7p and Rrn11p.122,123 CF interacts with the core promoter element, but does not by itself form a stable DNA-protein complex. However, in the presence of UAF, which forms a stable complex with the upstream element,124 CF becomes committed to the template and directs the initiation of transcription. It is not clear, at this time, whether TBP is a component of CF as suggested by Lin et al125 or if TBP is a “bridge” between CF and UAF.126 As the topic of rDNA transcription in yeast is covered in elsewhere in this volume the reader is directed there for a more complete discussion of this topic. Acanthamoeba TIF-IB Studies on rDNA transcription have demonstrated that one protein, TIF-IB, is the TBP-containing transcription factor that binds the rDNA promoter to form the committed complex.110,127,128 While, TIF-IB has not been cloned, it has been purified to homogeneity

102

The Nucleolus

and its interactions with the rDNA promoter have been studied extensively. TIF-IB consists of TAFI s of 99,96,91,145 kDa as well as TATA-binding protein. Site-specific cross-linking experiments demonstrated that the TIF-IB contacts mapped from -19 to –66.110,128,129 Interestingly, TBP, as part of TIF-IB, only made contact with promoters derivatized between -38 and -43. This site is 22 bp upstream of the bend in the promoter induced by contact with TIF-IB, and consistent with the hypothesis that the DNA binding region of TBP may not be as involved in DNA-binding by TIF-IB as it is in TIF-ID.110 Subsequent studies on the interaction of TIF-IB and RNA polymerase I with the A. castellanii promoter demonstrated that TIF-IB could direct transcription from a core promoter terminated at -6. Additional cross-linking experiments demonstrated that, when assayed in combination, both TAFI 96 and the 133 kDa subunit of RNA polymerase I interacted with the region between –1 and -7. This region contains a conserved sequence which is present in a large number of rRNA promoters: n(g/r)(g/r)Gt(T/A)aTnTAgGG(a/g)gAn (A=+1). This leads to the hypothesis that the CPE of RNA polymerase I promoters contains both an upstream site that interacts with TIF-IB and an Inr-like element that strengthens the interactions of the core promoter factor with the promoter (ref. 130 and references therein). While partially purified TIF-IB can form a stable complex with the rDNA promoter, homogeneous TIF-IB cannot. Paule’s laboratory demonstrated that an additional factor, TIF-IE, is required along with homogeneous TIF-IB for the formation of a stable complex on the rDNA core promoter. They found that TIF-IE by itself, however, did not bind to the rDNA promoter and thus differs in its mechanism from mammalian upstream binding factor and yeast upstream activating factor, which carry out similar complex-stabilizing functions. In addition to its presence in impure TIF-IB, TIF-IE was found in highly purified fractions of polymerase I, with which it associates.131,132 SL1 The mammalian homologue of TIF-IB is referred to as SL1. Like TIF-IB, SL1 is a “basal” rDNA transcription initiation factor capable of directing multiple rounds of RNA polymerase I recruitment to the rDNA promoter. SL1 was first identified and its subunits cloned in humans.133 Subsequently, homologous proteins have been identified in rat (SL1),134 mouse (TIF-IB, factor D),127,135-137 and frog (Rib1).138 SL1 exists as a complex containing the TATA-binding protein (TBP) and at least three RNA polymerase I specific TBP associated factors (TAFI s).133,136 As mentioned, TBP is a the subunit common to the fundamental transcription factors for all three nuclear transcription systems. In every case, the functional regions of TBP are localized to the highly conserved C-terminal domain, which consists of two copies of an imperfect repeat of 61-62 amino acids. This region is sufficient for the correct assembly of SL1 and is necessary for transcriptional activity.139 In contrast to TBP, the RNA polymerase I TAFs exhibit no homology to the TAFs involved in transcription by RNA polymerase II or III.110,133 In addition, the molecular masses of the RNA polymerase I TAFs differ between species for example, the human TAFI s are 110, 63, and 48 kDa,133 and the mouse TAFIs 95, 68 and 48 kDa.135,137 TAFI 48 exhibits the highest degree of conservation among species and contains two stretches near the N-terminus which are imperfectly repeated at the C-terminal.133,137 The largest TAFI’s, mouse TAFI 95 and human TAFI110, are the least conserved, demonstrating only 66% identity at the amino acid level. A review of the literature suggests that the second largest TAFs also differ significantly. The published sequences for human TAFI63 contains a unique 40 amino acid N-terminal extension and mouse TAFI68 has 66 unique amino acids in its C-terminal region. Both proteins contain two putative Zn2+ fingers, although mTAFI68 may have a third Zn finger.133,137 To date, the 5' end of the cDNA for human TAFI68 has not been reported.133 However, we have run a Blast search of the human EST databank using the reported sequence for TAFI63, and completed the human sequence. Once this was done, it was found that the human

Ribosomal DNA Transcription in Mammals

103

Figure 8. Alignment of the human and mouse sequences of TAFI68 demonstrates that the two sequences are 73.7 % identical. The full height vertical lines reflect identical amino acids for both proteins.

sequence for “TAFI63” encoded a protein of ~68 kDa, and mouse and human TAFI68 are 73.7% identical (Fig. 8). The mechanism determining the association of TBP with the TAFI s rather than other TAFs to form TFIID, TFIIIB and SNAPc is not known. In vitro experiments demonstrated that when TBP is bound to any of the TAFII’s, it will no longer bind the TAFI’s, and vice versa.133 These studies suggest a mutually exclusive binding and that this binding specificity will direct the formation of the promoter- or polymerase-selective TBP-TAF complexes.133 SL1 activity could be reconstituted from the three human TAFI s.133 However, functional mouse SL1 could not be reconstituted from recombinant mouse TBP and TAFI s, although they did form a high molecular weight complex,136 and could complex with the human TAFI s. This observations suggests that SL1 may contain additional components, or that additional factors may be required to mediate the interaction between SL1 and RNA polymerase I. Formation of a stable SL1 complex involves multivalent contacts between TBP and the TAFs as well as between the individual TAFs.133,136 These contacts appear to be conserved, as the interactions between the mouse TAFs and human TBP appears to be the same as that observed with the human TAFs.133 How the SL1 complex interacts with the rDNA promoter and thus mediates rDNA transcription is currently under investigation. The original studies on human SL1 suggested that, by itself, SL1 was not a DNA-binding protein.140 Human SL1 did not footprint the human rDNA promoter. However, the addition of SL1 to a UBF footprinting assay resulted in a 5' extension of the UBF footprint.140-142 Rat SL1 was found to be sufficient to drive transcription from a promoter that extended from -37 to +164,51 but footprinted over the UPE of the rat rDNA promoter.51 It was noted that the position of that footprint was similar to the “extension” of the human UBF footprint by human SL1,51 and included the sequence TTCTACAT GGGGACCTCT that was a near duplication of the CPE. Mutagenesis of this sequence inhibited transcription in assays that focused on UPE function, indicating that the binding of SL1 to the UPE was required for UPE function. Interestingly, mouse SL1 yielded a “disperse” footprint, but that footprint included the CPE.140,143 This binding was abolished by a mutation at -16 with respect to +1. The same mutation results in a decrease in rDNA transcription, confirming that the binding of mouse SL1 to the CPE is required for promoter recognition and transcription initiation.143 Studies on the purified, recombinant hTAFIs suggest that hSL1 is a DNA-binding protein. One study reported that both hTAFI110 and hTAFI63 bound to the rDNA promoter,144 while a second paper demonstrated that human TAFI48 and TAFI63 (or mouse TAFI68) can bind to DNA.133,145 Although it has not been tested, experiments examining the interactions between the core promoter binding factors suggest an ordered strength of DNA-binding, mouse SL1>rat SL1>human SL1. This may explain the relative importance of UBF in these various transcrip-

104

The Nucleolus

tion systems. In this regard it should be noted that A. castellanii TIF-IB has an very strong affinity for its promoter (Kd of 30x10-9), and it is a matter of discussion if there is a UBF-like activity in that organism.146 The interaction between SL1 and UBF appears to be critical for UBF-dependent activation of transcription.142 It has been suggested that basal rDNA transcription requires SL1 and the CPE, while elevated levels of transcription also require UBF and the distal promoter elements.51 Coimmunoprecipitation studies demonstrated that SL1 can bind to UBF in the absence of DNA.90,147 UBF antibody depleted extracts of SL1 activity but not TFIIIB activity, demonstrating that this is a specific interaction.90,147 In vitro studies suggest that this interaction may be mediated by the SL1 components, TBP and TAFI48.144 However, the domains of the proteins involved are as yet undetermined. These studies suggest that SL1 serves to communicate between UBF and RNA polymerase I. The interaction between SL1 and the rDNA promoter is species specific, e.g., human SL1 is required for transcription from the human rDNA promoter.3,4,148,149 In contrast, UBF and RNA polymerase I are, at least to some degree, interchangeable between species.143,150 For example, extracts prepared from primate cells that actively transcribe the human rRNA promoter fail to initiate transcription from a rodent rRNA promoter, but will do so when supplemented with either mouse or rat SL1.4,51,143,151 However, this only extends so far, a similar study showed that frog and human extracts could not be “reprogrammed” to accurately transcribe one another’s genes.152 The subunit, or subunits, of SL1 responsible for reprogramming has not been identified.* UBF UBF has been cloned from humans,154 mice,155 rats150,156 and Xenopus,157,158 and a search of the gene bank yields UBF homologues from all vertebrates. A 125kDa protein has also been identified in Acanthamoeba that has some functional characteristics similar to UBF,3 but there does not appear to be a UBF homologue in yeast. The most striking characteristic of UBF is the presence of 4-7 domains of varying sequence identity to the DNA-binding domain of HMG1, the HMG boxes. The sequences of the vertebrate UBFs are highly conserved protein. Human and rat UBF1 are 97% identical, and there is only one, nonconservative amino acid change between the two.155 Even between mammals and Xenopus there is 73% conservation of the amino acids overall. This conservation becomes 90% when the N-terminal domains are compared.2 Purified UBF consists of two polypeptides, UBF1 and UBF2, the sizes of which vary depending upon the species.155,156 The human and rodent UBF isoforms are migrate in discontinuous SDS-PAGE as bands of 97 kDa (UBF1) and 94 kDa (UBF2), whereas in Xenopus laevis they are 85 and 83 kDa.19,157 The mouse UBF gene consists of 21 exons extending over 13 kb.155 Transcription of this gene generates a single transcript which results in the mRNA for UBF1 (764 amino acids), or, due to alternative splicing at exon 8, UBF2. The result of the splicing event is that UBF2 mRNA contains an in frame deletion of 37 amino acids in HMG box 2.155,156 In contrast, Xenopus UBF1 and UBF2 are generated by transcription from two different genes,157 and there is evidence for additional UBF genes or pseudogenes in the Xenopus genome.158 The xUBF1 gene encodes a protein which has 93% identity to xUBF2 and contains an insert of 22 unique amino acids between HMG box 3 and 4.3,157,158 Both isoforms of UBF can bind to the rDNA promoter, form homo or hetero dimers in solution and bind to synthetic DNA cruciforms with a similar affinity.44,109,159 However, UBF1 has been shown to be a more potent activator of transcription in vitro and in vivo. UBF2 is 1/ 3 to 1/10 as active as UBF1.111,159 This suggests that the activity difference is due to the alteration of HMG box 2 as found in UBF2. One study suggests that UBF2 may have a unique function in the formation of loops between the enhancers of the gene promoter.160 However, *N. B. It would appear that the rule of species specificity may not be not absolute. 1) Mouse extracts can initiate transcription on the Xenopus promoter but at +4.3,243 And 2) There is one report that the rat rDNA promoter can be transcribed by primate RNA polymerase I transcription apparatus.153

Ribosomal DNA Transcription in Mammals

105

Figure 9. Cartoon of the structure and putative phosphorylation sites of UBF1.

when COS cells overexpress p21 h-Ras they only express UBF1 and these cells are viable.159 This suggests that, at least in COS cells, UBF2 is not essential for cell viability.159 One report found that the ratio of UBF1/UBF2 in a cell reflected the growth state of the cell, i.e., the ratio of UBF2/UBF1 was approximately two in stationary 3T3 or MH134 cells 155 and the ratio approached 1 upon nutritional upshift. It has also been reported that the ratio of UBF1 to UBF2 changes with development. For example, during differentiation of F9 cells or mouse embryogenesis the ratio of UBF1 to UBF2 decreases both at the mRNA and protein level.159 Cumulatively, these data raise the possibility that UBF2 may have an as yet unrecognized function(s). While it is clear that the UBF gene is subject to both positive and negative regulation,10,12,161 the mechanism(s) by which UBF expression is regulated is unknown. UBF Structure The dominant structural elements of UBF are the HMG boxes which are similar to the DNA binding domain of the chromosomal high mobility group proteins 1 and 2 (HMG-1, HMG-2) (Fig. 9, ref. 113). Other proteins belonging to this family include T-lymphocyte receptor a-enhancer factor, sex determining region Y protein, mitochondrial transcription factor and the yeast mitochondrial nonhistone protein NHP6.3 HMG boxes are usually 80 amino acids long.4 The number “found” in UBF depends upon the stringency of homology to the consensus sequence used for classification. Thus, there are reports of four to six HMG boxes in UBF.88,113,155 For example, when the definition of a HMG box is applied stringently, mammalian UBF and Xenopus UBF have four and three HMG boxes, respectively.4,88,162,163 However, many papers cite six and five boxes respectively. Interestingly, each HMG box appears to play a specific role.158,162 An HMG box cannot be replaced with another box from the same protein. However, they can be replaced with the same HMG box from a distantly related species. Although some reports have suggested that the number and order of the HMG boxes is crucial for UBF to function in transcript,158,162 the finding that Xenopus UBF can activate transcription in vivo in mouse cells115 suggests that the observation may reflect other variables.

106

The Nucleolus

Functions other than DNA-binding have been assigned to the HMG boxes. Part of the UBF nucleolar localization signal is found in the NH2 -terminal region, which includes the dimerization domain164 and HMG box 1.165,166 In addition, nuclear transport, requires a short 24 amino acid sequence near HMG box “5” as well as the CO2H-terminus.165 Similar to other HMG-like proteins UBF contains a highly acidic CO2H-terminal domain and an NH2 –terminal dimerization domain. The acidic CO2H-terminal domain of UBF consists of a stretch of 89 amino acids of which 68% are Glu or Asp, 25% are serines, and 7% glycine.113 The acidic regions are interrupted by conserved serine-rich blocks.3 The NH2-terminal dimerization domain contains two short regions that are hypothesized to form amphipathic helices similar to a helix-loop-helix motif. The dimerization domain is also required for optimal DNA binding along with at least one HMG box,163,167 and additional HMG boxes appear to stabilize DNA-binding.167,168 Mechanism of UBF Action The action of UBF depends on the formation of homo- and/or hetero-dimers3,163,168 and its binding to DNA, via the minor groove.169 Various manuscripts have reported that UBF binds to the CPE, UPE, spacer promoters and the enhancer repeats in the intergenic spacer.31,44,53,113,150 In addition, Xenopus UBF can bind on each side of the promoter proximal terminator.3 Interestingly, footprinting analysis has demonstrated that the DNase footprint obtained with UBF depends on the rDNA promoter being footprinted, and is independent of the origin of the UBF used in the assay.19,44,113,150 As discussed above, this would suggest that rDNA promoters share underlying structural similarities despite their sequence differences. In general, UBF footprints the rDNA promoter in the UPE, from ~ -50 to ~ -130.113 However, UBF can also protect the CPE, from ~ -45 to ~ +20.113,150,166 As discussed previously, a mutation at either the guanine at -16 or -7 eliminates promoter activity,47,51,113 but did not affect UBF binding to the DNA.47 As mentioned, the interaction of UBF with SL1 results in an extension of the UBF footprint within the UPE.49,51,140-142 This is believed to be part of the mechanism by which UBF facilitates the generation of the preinitiation complex on the promoter. Initially it was reported that UBF bound predominantly to a GC rich consensus sequence.4 However, UBF may recognize a specific DNA structure, such as synthetic DNA cruciforms, four way junctions, rather than a sequence.2,159,165,168 The domains of UBF required for DNA binding and the DNA binding sequence recognized are controversial. UBF with every HMG box deleted, except HMG box 1, is able to bind DNA (as long as the dimerization domain is present). The addition of HMG boxes increases the strength of DNA binding.4,113,163 These results may be explained if UBF binding to DNA is the result of a summation of multiple HMG box-DNA contacts. This would also lessen the requirement that any single sequence (DNA recognition site) be stringently maintained. UBF binds to DNA and by inducing folding and bending shortens the DNA contour by ~190 bp.49,166 This generates a disk-like UBF-DNA complex which has been referred to as an enhancersome.2,49,167 The enhancersome contains a low-density protein core around which the DNA loops, probably by in-phase bending. UBF can force the DNA to generate a 3600 loop with a diameter of 19 nm.49 In this structure the HMG boxes would interact with the promoter in a collinear manner. This model is consistent with those reported by Xie et al.47 In that study, spacing changes of half a helical turn significantly decreased rDNA promoter activity, while a full turn only mildly affected promoter activity.47 Thus it would appear that UBF binds to the DNA and bends it. It is not clear if UBF binds cruciform DNA and then bends the remaining DNA to form an enhancersome or if the binding of UBF to the DNA then bends the DNA, facilitating the formation of a cruciform and increasing the stability of the DNA-protein complex. In either case, bending the rDNA promoter would make it possible for the two bound SL1’s to interact generating the preinitiation complex model illustrated in (Fig. 3).

Ribosomal DNA Transcription in Mammals

107

Regulation of UBF Specific Activity Cells can modify the specific activity of UBF by at least two different mechanisms: i) phosphorylation; and ii) sequestration. UBF is a phosphoprotein,14 especially the CO2 H-terminal tail which can be extensively phosphorylated.112 The ability of UBF to trans-activate the rDNA promoter is reduced when UBF is treated with phosphatase14 or when the phosphorylated CO2H-terminal tail is deleted.112 This suggests that phosphorylated UBF is the more active form of the transcription factor. UBF contains numerous consensus motifs for characterized kinases including CKII and MAPK. Interestingly, treatment of UBF with CKII in vitro increases UBF phosphorylation14,112,168 and enhances transcriptional activity.112 However, it is not known whether or not UBF is an endogenous substrate of CKII in vivo. In many models of growth the phosphorylation status of UBF has been demonstrated to correlate positively with the rate of rDNA transcription.14,112,170 These experiments are discussed in more detail in section 5.4.2. The activity of UBF can also be down-regulated by the interaction of UBF with the product of the retinoblastoma susceptibility gene, Rb 11011,171 or a second pocket protein, p130.172 Studies indicate that the ability of UBF to transactivate the rDNA promoter is severely compromised when UBF is sequestered, either directly or indirectly, by Rb. Moreover, the physiological relevance of this mechanism in the regulation of rDNA transcription has been demonstrated in vivo.11 These experiments are discussed in greater detail below Cellular Distribution In some cell types the localization of UBF can change during the cell cycle. For example, during early S phase there is an increased association of UBF, RNA polymerase I and SL1 within the nucleolus.27 The increase in UBF association with nucleoli may be due to an increase in the ability of UBF to compete with the histones for binding to the rDNA. Typically, UBF localizes to the FC and DFC of the nucleolus where it forms small bead like structures in a folded, filament pattern.28 This distribution is sustained during the G2 phase when the cells are actively transcribing the rDNA.29 At the end of G2, when rDNA transcription is “shut off ”, UBF, RNA polymerase I and SL1 accumulate in the mitotic NORs forming a few intensive spots on the chromosomes.27 Other Factors Ku/E1BF Ku/E1 BF was originally detected as an human autoantigen reacting with antibodies from patients with rheumatic disorders and has now been widely identified in a number of species.3 Ku/E1 BF exists as a heterodimer of two polypeptides, 70 and 86 kDa polypeptide.173-176 Ku/ E1 BF tends to bind DNA in a nonspecific manner, while recent studies have shown that it binds the rDNA promoter with high specificity.174,177,178 Interestingly, when Ku/E1 BF is added to cell-free transcription assays it can affect the rate of rDNA transcription (to be discussed below). Ku/E1 BF has been shown to be the DNA-binding component of the DNA-dependent protein kinase (DNA-dependent PK).178,179 DNA-dependent PK is a nuclear, serine/threonine protein kinase consisting of a 350 kDa catalytic subunit and Ku/E1 BF. The enzyme is most active when bound to DNA, a process dependent on Ku/E1 BF. To date DNA-dependent PK has been shown to be important in various cellular processes such as, cell signaling, DNA replication, RNA polymerase II transcription activation and DNA repair. CPBF Interestingly Ku/E1 BF has been shown to interact with another potential rDNA transcription factor, CPBF (core promoter-binding factor).3 CPBF is a rDNA binding protein, which has been isolated from both rat mammary adenocarcinoma ascites and HeLa cells. CPBF purifies as two polypeptides of 44 and 39 kDa.174 The 44 kDa peptide binds to Ku/E1 BF.175 Both

108

The Nucleolus

CPBF peptides specifically interact with the rDNA core promoter sequence resulting in trans-activation of the rDNA promoter in vitro.174 Moreover, CPBF and Ku/E1 BF function synergistically to enhance RNA polymerase I transcription.174 CPBF has been found to be the rat homologue of human USF which also consists of two peptides, 44 and 43 kDa.180 USF is a basic helix-loop-helix zipper, DNA binding protein which specifically binds E-boxes in genes transcribed by RNA polymerase II. Interestingly, USF and CPBF bind to the same E-box in the rat rRNA promoter suggesting a possible mechanism for their action on rDNA transcription. Oligonucleotides to the E-box sequence inhibit rDNA transcription possibly by preventing USF/CPBF binding to the DNA.181,182 Topoisomerases Topoisomerases are enzymes which modulate DNA topology by catalyzing cleavage-rejoining reactions of the phosphodiester bonds. There are two classes of topoisomerases, I and II. Topoisomerase II includes both a (170 kDa) and b (180 kDa) isoforms.33,183 Topoisomerase may act as a swivel, relieving torsonal stress generated during transcription. This would allow for a rotation of the transcribed DNA segments without having to turn any other part of the DNA or the transcription ensemble.184,185 Both topoisomerase I and II are nuclear enzymes. Topoisomerase I and II a are found in both the nucleoplasm and nucleolus. Topoisomerase II b is exclusively localized in the nucleolus.26,186-189 Topoisomerase I preferentially associates with actively transcribed regions of chromatin, and has also been implicated in the regulation of rDNA transcription.188,189 Topoisomerase I has been demonstrated to be required for rDNA transcription and replication in yeast.186,189 p16 p16 is an HMG-like, DNA-binding protein isolated from Novikoff hepatoma ascites and Hela cells.190 p16 binds to the oligo d(A)•d(T) tracts found both within the UPE (-620 to -417) and external transcribed spacer (+352 to +525) of the rat rRNA gene. p16 was demonstrated to stimulate rDNA transcription in a dose dependent and saturable fashion when either of those sites was in cis with the target promoter. To date, this factor has not been cloned, thus, further details on the nature and mechanism of the interaction between p16 and the rDNA transcription apparatus are not available. Perhaps one of the most unexpected series of recent advances in our knowledge of transcription by RNA polymerase I has been the finding that components of the RNA polymerase II transcription apparatus are utilized in transcription by RNA polymerase I. This includes components of the basic transcription apparatus as well as regulatory proteins. Included in the first category are components of TFIIH and TAF1, one of the components of TFIID. The second category includes proteins such as Rb and p130, described in a following section. TFIIH Transcription by RNA polymerase II requires five transcription factors (TFIIB, TFIID, TFIIE, TFIIF, and TFIIH) referred to as the general transcription factors (GTFs). Stable association of RNA polymerase II with promoter sequences requires TFIID (or TBP), TFIIB, and TFIIF.191-194 However, after RNAPII has stably associated with promoter sequences, two additional factors, TFIIE and TFIIH, are necessary for transcription. During the formation of a transcription initiation competent complex, the complex undergoes conformational changes resulting in the formation of an open complex. Open complex formation is followed by the formation of the first phosphodiester bond. RNAP can then enter into an abortive mode, producing catalytic amounts of short RNA molecules (up to nine nucleotides).195,196 Some RNA polymerase II molecules escape the abortive mode and enter into the productive cycle. In this case, the polymerase moves away from the promoter enabling a second polymerase molecule to enter into the transcription cycle. This step is defined as promoter clearance. Transcription by RNAPII requires the hydrolysis of the β•γ bond of ATP or dATP.194 ATP hydrolysis is required for the formation of a stable open complex and by a step subsequent to

Ribosomal DNA Transcription in Mammals

109

initiation of transcription, most likely promoter clearance. It has become evident that the ATP-dependent step is catalyzed by TFIIH. Bypassing the requirement for ATP hydrolysis also bypasses the requirement for TFIIH in the formation of the open promoter complex. TFIIH is a multifunctional RNA polymerase II general initiation factor that includes two DNA helicases encoded by the Xeroderma pigmentosum complementation group B (XPB) and D (XPD) genes, as well as a cyclin-dependent protein kinase encoded by the CDK7 gene. Previous studies have shown that the TFIIH XPB DNA helicase functions at multiple steps to promote efficient transcription initiation and promoter escape by RNA polymerase II. The TFIIH XPB DNA helicase catalyzes ATP(dATP)-dependent formation of the open complex before synthesis of the first phosphodiester bond of nascent transcripts, and it is required to suppress premature arrest of very early RNA polymerase II elongation intermediates at promoter-proximal sites ’10–12 bp downstream of the transcriptional start site before their escape from the promoter. Data from several laboratories suggest that TFIIH also functions in promoter escape and may also function in elongation.197,198 Iben et al199 reported a series of observations that demonstrate a role for TFIIH in rDNA transcription. 1) They found that pre-rRNA synthesis is impaired in TFIIH ts yeast strains. 2) They demonstrated that TFIIH is localized within the nucleolus and is associated with a subpopulation of RNA polymerase I. 3) They demonstrated that cell free transcription reactions were inactive in the absence of TFIIH. Interestingly, they found that TFIIH is required for productive but not abortive rDNA transcription, implying a postinitiation role in transcription. Thus, their results provided evidence that while TFIIH may play several roles in transcription initiation by RNA polymerase II, including promoter escape, it may also play a role in promoter escape by RNA polymerase I. Interestingly, the protein product of the Cockayne Syndrome B gene has been demonstrated to recruit TFIIH to the RNA polymerase II complex and has been demonstrated to play a role in transcription by RNA polymerase II in mammalian cells. Interestingly, disruption of the yeast RAD26 gene results in the inhibition of cell growth in the absence of a need for transcription coupled DNA repair200 suggesting a relationship between the CSB gene product and transcription, independent of its role in DNA repair. This was confirmed by Bradsher et al when they demonstrated that CSB played a role in rDNA transcription and may in fact mediate the interaction between TFIIH and RNA polymerase I.201 TAF1 TAF1 is one of the subunits of TFIID that has been shown to be in close proximity to the initiator elements of genes that lack TATA elements. Ts13 CHO cells express a temperature sensitive mutant of TAF1that causes the cells to arrest late in G1 at the nonpermissive temperature. Interestingly, ts13 cells do not show a global alteration in gene expression. Rather, the expression of a subset of genes is affected. Included in these are the cell cycle regulators, cyclin D1, A and E.202 Lin et al203 have demonstrated that TAF1can be localized to the nucleolus and is capable of binding to UBF. Moreover, they have demonstrated that the addition of TAF1 to in vitro transcription assays stimulates rDNA transcription. This evidence suggests an additional method for coordinating the expression of RNA polymerase I- and RNA polymerase II-transcribed genes with cell cycle progression.

Formation of Preinitiation Complexes The transcription cycle involves four distinct steps: i) initiation; ii) promoter clearance; iii) elongation; and iv) termination.4 Initiation involves the assembly of the preinitiation complex on the rDNA promoter, isomerization of the closed preinitiation complex to form an initiation competent open complex and, finally, generation of the first phosphodiester bond.105 Once the first bond is formed, and the promoter is cleared, an alteration occurs in RNA polymerase I conformation which commits the enzyme to undergo RNA chain elongation.204 Elongation involves catalyzing the processive addition of ribonucleotides to the 3' end of the growing

110

The Nucleolus

RNA chain until specific attenuation or termination signals are encountered. Lastly, transcription is terminated and the product of the polymerase released from the template.105 There are two different models for transcription initiation by RNA polymerase I. In the first model, which is supported by the preponderance of the biochemical experiments carried out to date, the transcription factors form a committed template which then recruits RNA polymerase I, resulting in the formation of the preinitiation complex. In this model, which we will call the “static” model, the regulation of rDNA transcription occurs at the level of formation of the preinitiation complex,205,206 i.e., the recruitment of RNA polymerase I to the committed template. This model is based on in vitro biochemical experiments, summarized below. These experiments demonstrated that once formed, the committed template complex is quite stable and the transcription factors remain in place on the promoter through multiple rounds of transcription. This model suggested that the number of complexes is determined by limiting the amounts or activity of one or more of the transcription factors.206,207 In the second model, the “dynamic” model, supported by both recent in vitro and in vivo experimental observations, the assembly of the committed template is transient, and may not survive more than one round of transcription.104,208 Moreover, while the static model would suggest that the rate of transcription will depend upon the number of committed templates available and the ability of RNA polymerase I to recognize the committed complex and initiate transcription, the dynamic model includes the possibility that the stability of the binding of the transcription factors may in fact reflect the required recruitment of RNA polymerase I. It is important to consider that much of the evidence in support of the static model comes from in vitro transcription experiments where the concentrations of the components of the transcription system are significantly lower than they would be in vivo. Thus, while the assays suggested that the binding of the transcription factors was relatively stable, the assays did not reflect the possibility that at higher concentrations, such as those obtained in vivo, one might in fact observe a rapid and continuing equilibrium between the bound and unbound factors. Whichever model is correct, if in fact the final truth does not lie elsewhere, it has become apparent that there are many steps in the initiation process that are subject to regulation and that the system being investigated may determine which regulatory step is dominant. Data consistent with the static model for initiation comes from studies from Acanthamoeba castellani. In this system, TIF (SL1) binds to the promoter, in the absence of either UBF (or a UBF-like factor) or RNA polymerase I, and causes a distinctive DNAse I footprint.204,205 TIF then recruits RNA polymerase I by protein-protein interactions. RNA polymerase I binding results in an extension of the TIF footprint to include the +1 site.164 Upon the addition of nucleotide triphosphates, elongation occurs and the RNA polymerase I footprint moves down the template leaving the original TIF footprint behind.206 In mammalian and Xenopus transcription systems, initiation is also believed to be a multistaged process. In mammals, initiation involves SL1 binding to the core promoter, a process which is facilitated by UBF, and possibly TIF-IC, to form a committed template.207 This complex is stable for a number of rounds of transcription and able to recruit competent RNA polymerase I to form the preinitiation complex. The result of these steps is a complex which, with the addition of ATP/CTP (mouse; GTP/CTP in human) and further NTPs, becomes an initiation competent complex.207,208 The complex is now ready for elongation during which time RNA polymerase I moves past the initiation complex and leaves the preinitiation complex intact.

Regulation of rDNA Transcription Given the large number of molecular signalling pathways that impinge on rDNA transcription, one would predict that different cells might demonstrate varied pathways for regulating this process. Moreover, one might expect that different molecular signals might target combinations of regulatory points. Potential sites for regulation of ribosome synthesis include transcription of the ribosomal precursor genes (45S and 5S), preribosomal splicing, and assembly

Ribosomal DNA Transcription in Mammals

111

of the ribosomal subunits, and transport from the nucleus to the cytoplasm. However, in the majority of cases ribosome synthesis has been shown to be regulated largely at the level of transcription of the ribosomal genes (rDNA). Theoretically, regulation of rDNA transcription can involve: i) changes in chromatin structure; ii) alterations in the amount, localization, or activity of RNA polymerase I; and/or iii) similar alterations in the associated transcription factors. Moreover, recent studies suggest that the rDNA transcription apparatus can assemble (or colocalize) on the rDNA, but not be actively transcribing, suggesting that there may also be mechanisms for inhibiting transcription.29,209-211 For quite some time a question central to the discussion of ribosomal DNA transcription has been the consideration of whether rRNA synthesis is modulated by varying the transcription rate of a set number of genes or by varying the number of active genes. Electron microscopic visualization of active ribosomal RNA genes (Miller spreads) has typically demonstrated genes heavily packed with elongating polymerase molecules and nascent rRNA transcripts, “Christmas trees”. Only on rare occasions have pictures been obtained that demonstrate differences in the polymerase packing ratio between transcription units. In addition, measurements of the structures of the ribosomal chromatin,212 contributed to a model that the genes could only exist in two states, open for transcription and closed. Although this model was not necessarily supported by the data,213 these observations led to the conclusion that once a gene has been activated (open for transcription) the rate of transcription is not limited by the transcription initiation rate.214 This model, referred to as the “binary” model by French et al was recently reexamined by those authors.215 They noted that the rate of rRNA synthesis was apparently the same in two yeast strains, one with ~140 copies of the ribosomal repeat and one with ~40 copies. This contradicted the “binary” model, which predicted that the yeast strain with more ribosomal genes (140 copies) would demonstrate a greater rate of rRNA synthesis, even if one assumed that only 50% (70 copies) of the genes were active . The authors measured the number of active genes per yeast nucleolus, and the number of polymerase molecules per gene. They found that when the gene number was reduced the polymerase density was increased, to as high as one polymerase every 41 nt. The authors concluded, “Our results show that rRNA synthesis in exponentially growing yeast cells is controlled by the ability of cells to load polymerases and not by the number of open genes.”

Chromatin The structure of ribosomal chromatin is discussed elsewhere in this volume, and the reader is directed to that chapter for an in depth discussion of this topic. Both UBF and TTF-1 may play roles in altering the structure of the chromatin. The association of UBF with chromatin in vitro results in the displacement of the linker histone H1, without affecting the core histones. The spacer promoter may also play a role in opening the enhancer chromatin to activating factors and thus be involved in an early stages of gene activation.2 To date, limited experimental techniques are available to fully examine the possibility that the epigenetic modification of the chromatin structure, of a subset of the ribosomal genes, might play a role in the regulation of transcription of these multicopy genes.

RNA Polymerase I Observations from several laboratories studying both vertebrate and invertebrate rDNA transcription suggested that only a fraction of the RNA polymerase I purified from cells is capable of initiating specific transcription. As stated above, this process has been ascribed to either the post-translational modification of RNA polymerase I or to the association of a polymerase associated factor with the core polymerase. Acanthamoeba castellanii encyst when starved, and their rate of rDNA transcription decreases concomitant with an increase in the content of a modified form of RNA polymerase I (PolA) and a decrease in PolB. PolA does not support specific transcription in vitro,79,85 while the other form, PolB, can initiate both specific and

112

The Nucleolus

nonspecific transcription.8,79,85,216 The difference between these two forms is unclear and has been be ascribed to a modification of one of the subunits of the enzyme (A. castellanii), and/or a change in its association with other RNA polymerase I associated factors. This phenomenon is similar to that observed with respect to the association of Rrn3 with yeast RNA polymerase I and may in fact reflect the inactivation of A. castellanii Rrn3 during encystment. Similarly, Hanada et al81 and Tower and Sollner-Webb85 demonstrated that mammalian RNA polymerase I exists in two forms only one of which was capable of specific transcription initiation. Hanada et al identified PAF53 as that factor associated with transcriptionally competent fraction of RNA polymerase I, and Brun et al92 referred to this factor as C*. This topic is discussed in greater detail below. Interestingly, several yeast RNA polymerase I subunits, including A190, A43, A34.5, ABC23 and AC19 and possibly C53,75 as well as the 194 kDa subunit of mammalian RNA polymerase I 80 are modified by phosphorylation. However, the roles of these various phosphorylations in RNA polymerase I activity have yet to be established. Although Fath et al have demonstrated that the association of yeast Rrn3 with RNA polymerase I reflects the phosphorylation status of the polymerase.217 Alternatively, as mentioned above, the difference between PolA and PolB in mammals may be due to changes in either their association with RNA polymerase I-associated factor(s) or to a change in that factor. These activities have been referred to as TFIC, Factor C*, TIF-IA, TIF-IC or PAF53.81,85-87 Specific examples of such modifications are discussed below. Interestingly, the results published by Fath et al,217 are an example of a link between the phosphorylation status of RNA polymerase I and the association of a polymerase associated factor with the polymerase.

RNA Polymerase I Associated Factors TFIC, TIF-IA and Factor C*

In the early 1970’s Feigelson and colleagues91 reported that cycloheximide caused a rapid cessation of nucleolar RNA synthesis (ribosomal DNA transcription) and concluded that a rapidly turning over protein was required for RNA polymerase I activity in vivo. Subsequent studies identified TIF-IA and Factor C* as those factors that were required for the complementation of extracts of quiescent or cycloheximide treated cells. A third factor, TFIC, was identified as the factor required to reconstitute transcription by extracts of dexamethasone treated P1798 cells.86 In fact, at least three, and possibly more, polymerase associated proteins, TIF-IA, Factor C*, and TFIC, have been demonstrated to contribute to the regulation of rDNA transcription. Interestingly, TIF-IA, Factor C* and TFIC shared several properties, including a tight association with the core polymerase. Moreover, both Factor C* and TFIC were demonstrated to be heat stable, i.e., heat-treated (45oC, 15 min) S100 extracts of control cells could reconstitute transcription when added to an extract from dexamethasone (DEX) treated P1798 cells or cycloheximide (CHX) treated cells (Fig. 10). However, there were some differences between the factors. Brun et al92 reported that the in vitro transcriptional capacity of a preincubated rDNA promoter complex becomes exhausted very rapidly upon initiation of transcription and that this was due to the rapid depletion of C* activity. They concluded that Factor C* was used stoichiometrically in the transcription process. In contrast, Schnapp et al 94 reported that after initiation, “TIF-IA (Rrn3) is liberated from the initiation complex and facilitates additional rounds of transcription.” Thus, although C* and TIF-IA seemed to be the same factor, there appeared to be at least one significant difference between the two. Finally, TIF-IA and TFIC were purified and found to consist of different polypeptides.93,94 However, the lack of immunologic and molecular tools precluded a definitive statement that TIF-IA and TFIC were the same or different proteins. Recently, genetic approaches to identifying the components of the yeast RNA polymerase I transcription apparatus led to the identification of a polymerase

Ribosomal DNA Transcription in Mammals

113

Figure 10. Rrn3, the factor that is inactivated when cells are treated with cycloheximide, is heat stable in whole cell extracts.

associated factor, Rrn3p, that defines that subset of the RNA polymerase I molecules capable of specific transcription.101 Subsequently, the human homologue of yeast Rrn3 was identified,97 and demonstrated to be the equivalent of TIF-IA.98 Rrn3 and TFIC Are Two Different Polymerase Associated Factors Using purified recombinant human Rrn3, Cavanaugh et al99 demonstrated that Rrn3/TIF-IA were in fact two different transcription factors. Firstly, they demonstrated that Rrn3 could complement inactive extracts from cycloheximide treated cells, but that it could not complement extracts from cells treated with dexamethasone (Fig. 11A). They then argued that if treatment with cycloheximide resulted in the inactivation of Rrn3, then those extracts should contain functional TFIC. Further, if treatment with dexamethasone inactivated TFIC, then extracts of dexamethasone treated cells should contain active Rrn3. Thus, mixing the two types of inactive extracts should, and did result in biochemical complementation, i.e., transcription. Finally, using partially purified TFIC, Cavanaugh (unpublished observation) demonstrated that partially purified TFIC did reconstitute transcription from dexamethasone treated cells (Fig. 11) as predicted by earlier reports.86,93,95 Phosphorylation Regulates the Activity of Rrn3 Although the mechanism by which the inhibition of protein synthesis by cycloheximide is linked to the decrease in nucleolar RNA synthesis is unknown, studies on the effect of cycloheximide on Rrn3 activity were undertaken with the goal of increasing our understanding of the regulation of Rrn3 activity. These studies demonstrated that, like yeast Rrn3, mammalian Rrn3 was a phosphoprotein.99 However, Cavanaugh et al found that mammalian Rrn3 must be phosphorylated in order to function. They demonstrated that the interaction between Rrn3 and TAFI68 did not require phosphorylation, but that Rrn3 must be phosphorylated in order to interact with rpa43 (RNA polymerase I) and function in transcription. Subsequently, Zhao et al demonstrated that at least one of the phosphorylation events that activate Rrn3 is carried out by ERK in a growth dependent manner.218 It should be noted that 1) mutation of the primary ERK phosphorylation site to Ala failed to completely inactivate Rrn3 and 2) Rrn3 may contain as many as 10 phosphotryptic peptides. Thus, it is possible that the complete story on Rrn3 phosphorylation remains to be written.

114

The Nucleolus

Figure 11. Rrn3 and TFIC are different polymerase associated activities.

PAF53 As mentioned above, PAF53 is one of a group of recently purified proteins that associate with RNA polymerase I. To date two lines of evidence support a role for PAF53 in the regulation of rDNA transcription. 1) There is a positive correlation between the accumulation of PAF53 in the nucleoli of 3T3 cells and the rate of rDNA transcription; and 2) PAF53 is isolated in a complex with Pol IB (the transcriptionally active form of RNA polymerase I), but not with Pol IA.81 It has been suggested that PAF53 mediates an interaction between RNA polymerase I and UBF.81 However, these results were obtained in vitro and await additional corroboration. Moreover, the possibility that either the cell content of PAF53 or its association with RNA polymerase I is subject to regulation requires additional study.

rDNA Trans-Acting Factors SL1 A priori one might predict that SL1 would be a primary target for regulation in that: i) it is the RNA polymerase I paralogue of TFIID; and ii) it is absolutely required for rDNA transcription. Despite this prediction, only a few years ago there was only one published report of a physiologically relevant alteration in the amount or activity of SL1. In that report, Zhai et al219 found that SV40 large T antigen can bind to SL1 and activate rDNA transcription.219 Subsequent experiments demonstrated that the binding of large T antigen resulted in the recruitment of a kinase activity that in turn phosphorylated and activated UBF.220 Thus, it was not apparent that T antigen was in fact targeting SL1. However, there are other reports that demonstrate the possibility of regulating rDNA transcription through the regulation of SL1. In one report Kuhn et al, demonstrated that the phosphorylation of SL1 by cdc2/cyclin B inactivates it. They noted that this would provide a mechanism for the mitotic silencing of the rDNA transcription apparatus.221

Ribosomal DNA Transcription in Mammals

115

In their studies of the regulation of rDNA transcription in differentiating U937 cells, Comai et al noted a significant repression of SL1 activity, without a significant change in SL1 content.222 In addition, Muth et al223 have reported that TTF may recruit a histone acetyltransferase activity to the promoter, and that the acetylation of TAFI68 enhances its ability to bind to the rDNA promoter. They suggested, “the reversible acetylation of TIF-IB/SL1 may be an effective means to regulate rDNA transcription in response to external signals.” The observation that SL1 can be regulated through acetylation/deacetylation may in fact be linked with an observation first made by Cavanaugh et al.11 We noted that during the differentiation of U937 cells Rb was recruited to the nucleolus. We found that Rb could repress UBF activity (described below). However, it is known that Rb can also recruit histone deacetylase. In fact, Pelletier et al found that the binding of Rb to UBF competed with the binding of CBP, suggesting Rb might interfere both with the interaction between UBF and SL1 and inhibit the activity of SL1 through the recruitment of a deacetylase.224 Thus, it is formally possible that the inhibition of rDNA transcription that is observed in vivo, results both from the ability of Rb to block the ability of UBF to activate transcription and from the Rb-dependent recruitment of a deacetylase that inactivates SL1.

UBF The hypothesis that regulating UBF activity in the cell might have an effect on rDNA transcription is controversial. For instance, there are believed to be 10,000-100,000 copies of UBF in the cell.28 This number, is in vast excess when compared to both the number of active ribosomal genes and to the estimated number of SL1 and RNA polymerase I complexes.1,3 This would suggest that UBF is not a rate-limiting component of the rDNA transcription apparatus. However, these estimates are based on the amount of UBF present in rapidly dividing, immortal cell lines. The cellular content of UBF in differentiated cells such as, adult hepatocytes, neonatal and adult cardiomyocytes is significantly lower than that observed in immortal cell lines (D. O’Mahony, R. Hannan, and L. Rothblum, unpublished observation). Moreover, the transfection and overexpression of UBF1 in neonatal cardiomyocytes is sufficient to stimulate transcription from a reporter construct for rDNA transcription in a dose dependent manner.114 Such observations have led groups to examine if UBF is a potential target for regulation during altered growth conditions. In theory, the cellular activity of UBF can be regulated by either altering the amount of UBF available to transactivate the rDNA promoter or by changing the activity of an individual molecule by posttranslational modifications such as phophorylation. In fact, both have of these mechanisms have been demonstrated to occur. Moreover, while these two mechanisms are not mutually exclusive, they appear to be dependent on both the cell type and stimulus being examined. Regulation of UBF Content Numerous studies have demonstrated a correlation between the cellular content of UBF and rDNA transcription. For example, the differentiation of L6 myoblasts into myotubes correlates with a decrease in UBF mRNA which precedes the decrease in UBF content and rDNA transcription.12 At the same time, myosin heavy chain protein accumulates, the mRNA level of myogenin increases, and transcription of the tubulin, r-protein L32, and 5S rRNA genes do not change.12 Thus, the observed decrease in UBF content during differentiation is not due to a general decrease in gene expression or translation.12 Serum starvation of cells, such as 3T6 cells, reduces rDNA transcription due to a decrease in the availability of the mitogenic factors found in serum that these cells require for growth. This decrease correlated with a decrease in the cellular content of UBF. Refeeding serum-starved 3T6 cells with serum restored UBF content, which preceded the elevation of rDNA transcription to levels observed in control cells. Accumulation of UBF protein was found to result from regulation at the level of transcription of the UBF gene, in a manner similar to that of c-myc.13

116

The Nucleolus

The regulation of rDNA transcription has been studied in LNCaP cells, an androgen dependent cell line. Nuclear run-on data demonstrated that DHT treatment of these cells increases rDNA transcription which correlated with an increase in UBF cellular content.225 In addition, extracts of prostate cells from orchiectomized rats showed a decrease in rDNA transcription and UBF protein. However, if the rats were treated with testosterone these levels did not decrease.225 Thus, androgens appear to, at least in part, stimulate rRNA synthesis by regulating the quantities UBF. A correlation between UBF content and the regulation of rDNA transcription has been extensively studied in primary cultures of neonatal cardiomyocytes. When neonatal cardiomyocytes are treated with various growth promoting stimuli such as adrenergic agents, they undergo hypertrophy. This is associated with an elevated protein synthetic capacity due to increased ribosome biogenesis, which is achieved by increasing rDNA transcription.9,170 A good correlation is observed between the degree to which cells grow (hypertrophy) in response to a growth stimulus and the degree to which rDNA transcription is increased. Phenylephrine affected neither the content of RNA polymerase I nor UBF phosphorylation. However, there were significant increases in the cellular contents of UBF mRNA and protein which correlated, both temporally and quantitatively, with changes in rDNA transcription.9 This correlation was confirmed by the observation that overexpressing UBF1, in the absence of hypertrophic stimuli, increased the activity of a cotransfected reporter for rDNA transcription.114 Regulation of UBF Phosphorylation Stimulation of neonatal cardiomyocytes with two other hypertrophic agents, phorbol 12-myristate 13-acetate (PMA) or endothelin-1 (ET-1) does not change the cellular content of UBF. Instead a significant increase in UBF phosphorylation was observed. This increase in UBF phosphorylation correlated, both temporally and quantitatively, with elevated rDNA transcription. The affects were not seem until 6-12 h after the onset of PMA or ET-1 treatment,170 suggesting that they did not result from the activation of protein kinase C. These findings emphasize that even within one cell type the mechanisms utilized to regulate rDNA transcription and UBF activity are stimuli specific. A correlation between the phosphorylation status of UBF and rDNA transcription has been observed in other cell culture systems. For example, the decreased rate of rDNA transcription that accompanies serum starvation of CHO cells correlates with a slow decrease in UBF phosphorylation, in the absence of changes in cellular content.14,112 The addition of serum restores both rDNA transcription and the degree of UBF phosphorylation. Pulse-chase experiments demonstrated that the decrease in UBF phosphorylation was due to a reduction in phosphorylation, and not the result of “active” dephosphorylation. Similarly, treatment of vascular smooth muscle cells (VSMC) with Angiotensin II (AII), a hypertrophic stimulus, rapidly (within 30 min) increases both rDNA transcription and UBF phosphorylation in the absence of changes in UBF content.226 The activity of CKII, an enzyme which phosphorylates UBF in vitro, is not altered in those cells, suggesting that either a serine kinase other than CKII is responsible for AII stimulation of UBF phosphorylation or the ability of CKII to specifically phosphorylate UBF was being regulated. Stefanovsky et al demonstrated that activation of the MAP kinase (ERK) leads to UBF phosphorylation and activation, demonstrating a direct link between growth factor activation and ribosomal DNA transcription.227 There are several reports demonstrating cell-cycle specific patterns of UBF activity that correlate with both the inhibition of rDNA transcription during mitosis and its activation during G1 and G2.228,229 Moreover, Voit et al229 demonstrated that Ser484 is a direct target for phosphorylation by a cyclin-dependent kinase 4 (cdk4)-cyclin D1- and cdk2-cyclin E. However, the mutation of Ser484 to Ala does not eradicate the ability of UBF to activate transcription, suggesting that other modifications are necessary to modulate UBF in a cell cycle specific manner.227 None of the above studies have established if there are qualitative changes in the specific serine residues phosphorylated, as they have examined the phosphorylation status of ectopically

Ribosomal DNA Transcription in Mammals

117

expressed UBF. In addition, while in vitro experiments have demonstrated that phosphorylated UBF is more transcriptionally active than dephosphorylated UBF, it remains to be determined whether alterations in the phosphorylation state are necessary or sufficient to effect changes in rDNA transcription rates in vivo. Future studies will have to define the sites phosphorylated and identify the enzymes responsible for phosphorylating authentic UBF in vivo. Sequestration of UBF Recent studies indicate that a direct measurement of the total cellular content of UBF or its degree of phosphorylation may not necessarily correlate with the amount of UBF available to transactivate the rDNA promoter or the specific activity of UBF. This conclusion stems from the observation that UBF can be sequestered into an inactive complex with the protein product of the Retinoblastoma susceptibility gene, Rb 110. Rb 110 functions as a tumor suppresser and is a negative regulator of growth,230 acting at the G1 checkpoint. It is the underphosphorylated (hypophosphorylated) form of Rb 110 which is the most active. This form predominates in quiescent cells, while the hyperphosphorylated form is prevalent in actively growing cells.230 The hypophosphoryated form of Rb predominates in Go and G1 phase cells. Hyperphosphorylated Rb predominates in the G2, M and S phases of the cell cycle.230 The initial observations of an interaction between UBF and Rb came from studies on the regulation of rDNA transcription in differentiating U937 cells.11 It was noted that when U937 cells differentiate, Rb accumulates in the nucleolus and rDNA transcription decreases in the absence of changes in UBF content. This study also demonstrated, using cell-free transcription assays, that the addition of Rb 110 to an extract containing limiting amounts of Rb resulted in inhibition of UBF-dependent rDNA transcription in vitro.11 Coimmunopreciptiation experiments demonstrated an increase in the association between Rb and UBF with differentiation.11 Moreover, affinity chromatography experiments demonstrated that this interaction was specific and required the A/B pocket of Rb. This was deduced since Rb 209, a biologically inactive form of Rb 110 containing a cysteine to phenylalanine mutation at amino acid 706 (in the A/B pocket), did not interact with UBF and did not inhibit rDNA transcription in vitro. In addition, the UBF-Rb interaction could be inhibited by a synthetic peptide that has been shown to interact with the A/B pocket and block the interaction of other proteins with the pocket.11,171 Subsequently, Hannan et al231 demonstrated that as 3T6 cells become confluent there is a marked reduction in the rate of rDNA transcription, that correlated with the accumulation of hypophosphorylated Rb in the nucleoli of the confluent cells. They then demonstrated that UBF interacted with the “active” or hypophosphorylated form of Rb and not with the “inactive” or hyperphosphorylated form. Moreover, they demonstrated that the overexpression of Rb inhibited rDNA transcription in vivo. Although there is agreement that Rb can inhibit rDNA transcription by binding to UBF. There is some discussion concerning the mechanism of action. Voit et al232 found that Rb inhibited UBF binding to the rDNA promoter, but not the ability of UBF to interact with SL1 or RNA polymerase I.232 Interestingly, in these experiments Rb209 was just as affective as Rb, and the authors concluded that the CO2H-terminal domain of Rb, and not the A/B pocket, was required for the interaction between Rb and UBF.232 In a subsequent study, Hannan et al found that Rb209, as found in H209 cells, did not coimmunoprecipitate with UBF. Moreover, they demonstrated that Rb and a second pocket protein, p130, did not interfere with DNA-binding by UBF, but did disrupt the interaction between UBF and SL1.172 Hannan et al also noted an inverse correlation between p130 levels and rDNA transcription in vivo, and were able to demonstrate that the overexpression of p130 resulted in the inhibition of rDNA transcription. These discrepancies suggest that further experiments are required in order to determine the exact mechanism involved in the regulation of rDNA transcription by the members of the pocket protein family.

118

The Nucleolus

Other Factors Ku/E1BF and CPBF In some reports Ku/E1 BF markedly inhibited rDNA transcription,233-235 while in others it stimulated transcription.173-176 The first finding corroborates in vivo studies which demonstrated that the expression of Ku/E1 BF correlates negatively with the proliferation state of the cell.3,236 However, the question remains as to how Ku/E1 BF acts as both a positive and negative regulator of rDNA transcription. It has been suggested that low concentrations of Ku/E1 BF have a positive effect on rDNA transcription whereas higher concentrations repress transcription.234,235,237 Anti-Ku antibodies can precipitate a repressor activity from HeLa cells, and stimulate rDNA transcription. The addition of UBF can also overcome Ku/E1 BF repression,175 thus suggesting that one mechanism by which UBF may enhance rDNA transcription is by releasing Ku/ E1 BF repression rather than by directly stimulating transcription.209 Since Ku/E1 BF interacts with the UBF and SL1 binding sites on the rDNA promoter it may compete with them for their DNA-binding sites on the rDNA promoter.175,237 In this model, a high concentration of Ku/E1 BF would titrate the UBF and/or SL1 binding sites and reduce rDNA transcription. One other study, which examined Ku/E1 BF in serum-starved rat NISI cells, suggested that there may be two forms of Ku/E1 BF; one which enhances (Ku/E1 BFc*) and one which inhibits transcription (Ku/E1 BFs : isolated from serum starved cells).234 The difference between these two forms is unclear and may involve alternative splicing or a post translational modification of Ku/E1 BF.234 It is likely that, in this case, the effects of Ku/E1 BF on rDNA transcription reflect the ratio of active (Ku/E1 BFc*) to repressive (Ku/E1 BFs ) forms in the cell. Ghoshal and Jacob have reported that heat shock (42o C, 3h) repressed rDNA transcription and leads to a reduction (90%) in E1 BF, demonstrating a correlation between the regulation of E1 BF and rDNA transcription.238 Interestingly, when Ku/E1 BF is complexed with DNA-dependent PK the complex represses rDNA transcription to a greater extent than Ku/E1 BF alone.178,179 It is possible that the enhanced repression may be due to DNA-dependent PK phosphorylation of certain components of the RNA polymerase I transcription complex. Cell extracts lacking CPBF are rDNA transcriptionally inactive and the subsequent addition of CPBF restores transcription in an dose dependent fashion.3 Interestingly, when the human homologue of CPBF, USF1 is overexpressed as a homodimer in CHO cells it represses rDNA transcription. However, when USF1 forms heterodimers with USF2, rDNA transcription is stimulated. It is possible that the form of dimer may affect the ability of USF to bind the E boxes of the rDNA promoter and thus alter transcription.3,181 Topoisomerases The effect of topoisomerase on rDNA transcription have been examined in a number of systems. Treatment of HeLa cells with topoisomerase I-specific inhibitors, such as camptothecin, rapidly inhibits 45S rRNA synthesis which is reversible with drug removal.239 Interestingly, topoisomerase I coprecipitates with TBP,240 a subunit of the transcription factor SL1, and has been reported to copurify with RNA polymerase I.241,242 These results suggest that it may be a component of an RNA polymerase I holoenzyme involved in the formation of the preinitiation complex and thus rDNA transcription. While that model would suggest that topoisomerase I plays a positive role in rDNA transcription, there is additional evidence that it may negatively regulate transcription. Topoisomerase II has been found to bind to the CPE of the rDNA promoter and inhibit transcription by preventing preinitiation complex formation.186 This process is counteracted by UBF. UBF may be competing with topoisomerase for the same DNA binding sites, thus if UBF is bound to the promoter Topo II is unable to repress transcription.186 The inhibition of topoisomerase I was also shown to generate a graded decrease (5' to 3') in the number of RNA polymerase I molecules associated withthe transcription unit. This has been interpreted as evidence that the inhibition of topoisomerase I results in the inhibit elongation.239 In addition, mutagenesis studies have demonstrated that both topoisomerase I and II are important for rRNA synthesis in S. cerivisiae.184

Ribosomal DNA Transcription in Mammals

119

References 1. Hannan RD, Taylor L, Cavanaugh AH et al. The role of UBF in regulating rDNA transcription. In: Paul M, ed. RNA Polymerase I: Transcription of Eukaryotic Ribosomal RNA. New York: Springer-Verlag, 1998:221-232. 2. Moss T, Stefanovsky VY. Promotion and regulation of ribosomal transcription eukaryotes by RNA polymerase I. Prog Nucleic Acid Res Mol Biol 1995; 50:25-66. 3. Paule MR. Transcription of ribosomal RNA by eukaryotic RNA polymerase I. In: Conaway RC, Conaway JW, ed. Textbook of Transcription: Mechanisms and regulation. NY: Raven Press,1994. 4. Reeder RH. Regulation of transcription by RNA polymerase I. In: McKnight SL, Yamamoto KR, eds. Textbook of Transcription Regulation. Cold Spring Harbor Lab Press, 1992. 5. Grummt I. Regulation of mammalian ribosomal gene transcription by RNA polymerase I. Prog Nucleic Acid Res Mol Biol 1999; 62:109–154. 6. Hannan RD, Rothblum LI. Regulation of ribosomal DNA transcription during neonatal cardiomyocyte hypertrophy. Cardiovas Res 1995; 30:501-510. 7. Larson DE, Zahradka P, Sells BH. Control points in eucaryotic ribosome biogenesis. Biochem Cell Biol 1991; 69:5-22. 8. Paule MR, Iida CT, Perna PJ et al. In vitro evidence that eukaryotic ribosomal RNA transcripiton is regulated by modification of RNA polymerase I. Nucleic Acids Res 1984; 12:8161-8180. 9. Hannan RD, Luyken J, Rothblum LI. Regulation of rDNA transcription factors during cardiomyocytes hypertrophy induced by adrenergic agents. J Biol Chem 1995; 270:8290-8297. 10. Hannan RD, Luyken J, Rothblum LI. Regulation of ribosomal DNA transcription during contraction-induced hypertrophy of neonatal cardiomyocytes. J Biol Chem 1996; 271:3213-3220. 11. Cavanaugh AH, Hemple WM, Taylor LJ et al. Activity of RNA polymerase I transcription factor UBF blocked by Rb gene product. Nature 1995; 374:177-180. 12. Larson D, Xie W, Glibetic M et al. Coordinated decreases in rRNA gene transcription factors and rRNA synthesis during muscle cell differentiation. Proc Natl Acad Sci 1993; 90:1-4. 13. Glibetic M, Taylor L, Larson D et al. The RNA polymerase I transcription factor UBF is the product of a primary response gene. J Biol Chem 1995; 270:4209-4212. 14. O’Mahony D, Xie W, Smith SD et al. Differential phosphorylation and localization of the transcription factor UBF in vivo in response to serum deprivation. J Biol Chem 1992; 267:35-38. 15. Antonetti DA, Kimball SR, Horetsky RL et al. Regulation of rDNA transcription by insulin in primary cultures of rat heptocytes. J Biol Chem 1993; 268:25277-25284. 16. Gokal PK, Cavanaugh AH, Thompson EA. The effects of cyclohexamide upon transcription of rRNA 5S RNA and tRNA genes. J Biol Chem 1986; 261:2536-2541. 17. Kermekchiev M, Muramatsu M. Presence of an inhibitor of RNA polymerase I mediated transcription in extracts from growth arrested mouse cells. Nucleic Acids Res 1993; 21:447-453. 18. Jacob ST. Regulation of ribosomal gene transcription. Biochem J 1995; 306:617-626. 19. Reeder RH. rRNA synthesis in the nucleolus. Trends Genet 1990; 6:390-395. 20. Roussel PC, Comai AL, Hernandez-Verdun D. The rDNA transcription machinery is assembled during mitosis in active NORs and absent in inactive NORs. J Cell Biol 1996; 133:235-246. 21. McClintock B. The relation of a particular chromosomal element to the development of the nucleoli in Zea mays. Z Zellforsch Mikrosk Anat 1934; 21:294-328. 22. Choo KH, Earle E, McQuillan C. A homologous subfamily of satellite III DNA on human chromosomes 14 and 22. Nucleic Acids Res 1990; 18:5641–5648. 23. Shiels C, Coutelle C, Huxley C. Contiguous arrays of satellites 1, 3, and beta form a 1.5-Mb domain on chromosome 22p. Genomics 1997; 44:35–44. 24. Tagarro I, Wiegant J, Raap AK et al. Assignment of human satellite 1 DNA as revealed by fluorescent in situ hybridization with oligonucleotides. Hum Genet 1994; 93:125–128. 25. Waye JS, Willard HF. Concerted evolution of alpha satellite DNA: Evidence for species specificity and a general lack of sequence conservation among alphoid sequences of higher primates. Chromosoma 1989; 98:273–279. 26. Schwarzacher HG, Wachtler F. The nucleolus. Anat Embryol 1993; 188:515-536. 27. Jordan PM, Mannervik LT, Carmo-Fonseca M. In vivo evidence that TATA-binding protein/ SL-1colocalizes with UBF and RNA polymerase I when rRNA synthesis is either active or inactive. J Cell Biol 1996; 133:225-234. 28. Roussel PC, Andre C, Masson C et al. Localization of the RNA polymerase I transcription factor hUBF during the cell cycle. J Cell Sci 1993; 104:327-337. 29. Junera HR, Masson C, Geraud et al. Involvement of in situ conformation of ribosomal genes and selective distribution of upstream binding factor in rRNA transcription. Mol Biol Cell 1997; 8:145-156. 30. Matera AG, Tycowski KT, Steitz JA et al. Organization of small nucleolar ribonucleoproteins snoRNPs by fluorescence in situ hybridization and immunocytochemistry. Mol Biol Cell 1994; 5:1289-1299.

120

The Nucleolus

31. Peter M, Nakagawa J, Doree M et al. Identification of major nucleolar proteins as candidate mitotic substrates of cdc2 kinase. Cell 1990; 60:791-801. 32. Weisenberger D, Scheer U. A possible mechanism for the inhibition of ribosomal RNA gene transcription during mitosis. J Cell Biol 1995; 129:561-575. 33. Culotta V, Sollner-Webb B. Sites of topoisomerase I action on X laevis ribosomal chromatin: Transcriptionally active rDNA has an ~200 bp repeating structure. Cell 1988; 52:585-597. 34. Labhart P, Koller T. Structure of the active nucleolar chromatin of Xenopus laevis oocytes. Cell 1982; 28:279-292. 35. Langst G, Schatz T, Langowski J et al. Structural analysis of mouse rDNA: Coincidence between nuclease hypersensitive sites DNA curvature and regulatory elements in the intergenic spacer. Nucleic Acids Res 1997; 25:511-517. 36. Mutskov VJ, Russanova VR, Dimitrov SI et al. Histones associated with nonnucleosomal rat ribosomal genes are acetylated while those bound to nucleosome-organized gene copies are not. J Biol Chem 1996; 271:11852-11857. 37. Dimitrov SI, Tateossyan HN, Stepanovsky VY et al. Binding of histones to Xenopus laevis ribosomal genes with different levels of expression. Eur J Biochem 1992; 204:977-981. 38. George P, Demeler B, Terpening C et al. Binding of the RNA polymerase I transcription complex to its promoter can modify positioning of downstream nucleosomes assembled in vitro. J Biol Chem 1993; 268:1947-1954. 39. Finley D, Bartel B, Varshavsky A. The tails of ubiquitin precursors are ribosomal proteins whose fusion to ubiquitin facilitates ribosome biogenesis. Nature 1989; 338:394-401. 40. Gong X, Radebaugh CA, Geiss GK et al. Site-directed photo-cross-linking of rRNA transcription initiation complexes. Mol Cell Biol 1995; 15:4956-63 41. Aprikian P, Moorefield B, Reeder RH. TATA binding protein can stimulate coredirected transcription by yeast RNA polymerase I. Mol Cell Biol 2000; 20:5269-5275. 42. Keener J, Josaitis CA, Dodd JA et al. Reconstitution of yeast RNA polymerase I transcription in vitro from purified components. TATA-binding protein is not required for basal transcription. J Biol Chem 1998; 273:33795-802. 43. Nagamine M, Kishimoto T, Aono J et al. Sequestration analysis for RNA polymerase I transcription factors with various deletion and point mutations reveals different functional regions of the mouse rRNA gene promoter. Mol Cell Biol 1987; 7:1486-1495. 44. Cassidy BG, Yang-Yen HF, Rothblum LI. Additional RNA polymerase I initiation site within the nontranscribed spacer region of the rat rRNA gene. Mol Cell Biol 1987; 7:2388-23896. 45. Cassidy B, Haglund R, Rothblum LI. Regions upstream from the core promoter of the rat ribosomal gene are required for the formation of a stable transcription initiation complex by RNA polymerase I in vitro. Biochem Biophys Acta 1987; 9:133-144. 46. Smith S, Hong S, Yang-Yen HF et al. Mechanisms of control of gene expression. In: Cullen B et al, ed. Textbook of UCLA Symposia on Molecular and Cellular Biology New Series. NY: 1989:95 47. Xie WQ, O’Mahony DJ, Smith SD et al. Complementary in vivo and in vitro analyses of the interactions between cis-acting elements of the rat rDNA promoter. Mol Cell Biochem 1991; 104:127-135. 48. Ridsdale RA, Semotok JL, Larson DE et al. Topology of recombinant rat upstream binding factor. Biochem Cell Biol 1998; 76:649-655. 49. Bazett-Jones DP, Leblanc B, Herfort M et al. Short-range DNA looping by the Xenopus HMG-box transcription factor xUBF. Science 1994; 264:1134-1137. 50. Pape L, Windle J, Sollner-Webb JB. Half helical turn spacing changes convert a frog into a mouse rDNA promoter: A distant upstream domain determines the helix face of the initiation site. Genes Dev 1990; 4:52-62. 51. Smith SD, Oriahi E, Lowe D et al. Characterization of factors that direct transcription of rat ribosomal DNA. Mol Cell Biol 1990; 10:3105-3116. 52. Smith SD, Oriahi E, Yang-Yen HF et al. Interaction of RNA polymerase I transcription factors with a promoter in the nontranscribed spacer of rat ribosomal DNA. Nucleic Acids Res 1990; 18:1677-1685. 53. Putnam CD, Pikaard CS. Cooperative binding of the Xenopus RNA polymerase I transcription factor xUBF to repetitive ribosomal gene enhancers. Mol Cell Biol 1992; 12:4970-4980. 54. Reeder R. Enhancers and ribosomal gene spacers. Cell 1984; 38:349-351. 55. De Winter RFJ, Moss T. Spacer promoters are essential for efficient enhancement of X laevis ribosomal transcription. Cell 1965; 44:313-318. 56. Pikaard CS, Pape LK, Henderson LS et al. Enhancers for RNA polymerase I in mouse ribosomal DNA. Mol Cell Biol 1990; 10:4816-4825. 57. Kuhn A, Deppert U, Grummt I. A 140-base-pair repetitive sequence element in the mouse rRNA gene spacer enhances transcription by RNA polymerase I in a cell-free system. Proc Natl Acad Sci USA 1990; 87:7527-31.

Ribosomal DNA Transcription in Mammals

121

58. Langst G, Schatz T, Langowski J et al. Structural analysis of mouse rDNA: Coincidence between nuclease hypersensitive sites, DNA curvature and regulatory elements in the intergenic spacer. Nucl Acids Res 1997; 25:511-517. 59. Mason SW, Sander EE, Grummt I. Identification of a transcript release activity acting on ternary transcription complexes containing murine RNA polymerase I. EMBO J 1997; 16:163-172. 60. Smid A, Finsterer M, Grummt I. Limited proteolysis unmasks specific DNA-binding of the murine RNA polymerase I-specific transcription termination factor TTFI. J Mol Biol 1992; 227:635-647. 61. Evers R, Smid A, Rudloff U et al. Different domains of the murine RNA polymerase I-specific termination factor mTTF-I serve distinct functions in transcription termination. EMBO J 1995; 14:1248-1256. 62. Hannan RD, Cavanaugh A, Hempel W et al. Identification of a mammalian RNA polymerase I holoenzyme Containing Components of the DNA Repair/replication System. Nucl Acids Res 1999; 27:3720-3727. 63. Labhart P, Reeder RH. A point mutation uncouples RNA 3'-end formation and termination during ribosomal gene transcription in Xenopus laevis. Genes Dev 1990; 4:269-276. 64. McStay B, Reeder RH. A DNA-binding protein is required for termination of transcription by RNA polymerase I in Xenopus laevis. Mol Cell Biol 1990; 10:2793-2800. 65. Lang WH, Morrow BE, Ju Q et al. A model for transcription termination by RNA polymerase I. Cell 1994; 79:527-534. 66. Lang WH, Reeder RH. The REB1 site is an essential component of a terminator for RNA polymerase I in Saccharomyces cerevisiae. Mol Cell Biol 1993; 13:649-658. 67. Lang WH, Reeder RH. Transcription termination of RNA polymerase I due to a T-rich element interacting with Reb1p. Proc Natl Acad Sci 1995; 92:9781-9785. 68. Jeong SW, Lang WH, Reeder RH. The release element of the yeast polymerase I transcription terminator can function independently of Reb1p. Mol Cell Biol 1995; 15:5929-5936. 69. Langst G, Blank TA, Becker PB et al. RNA polymerase I transcription on nucleosomal templates: The transcription termination factor TTF-I induces chromatin remodeling and relieves transcriptional repression. EMBO J 1997; 16:760-768. 70. Gerber JK, Gogel E, Berger C et al. Termination of mammalian rDNA replication: Polar arrest of replication fork movement by transcription termination factor TTF-I. Cell 1997; 90:559-567. 71. Zhou Y, Santoro R, Grummt I. The chromatin remodeling complex NoRC targets HDAC1 to the ribosomal gene promoter and represses RNA polymerase I transcription. EMBO J 2002; 17:4632-4640. 72. Santoro R, Li J, Grummt I. The nucleolar remodeling complex NoRC mediates heterochromatin formation and silencing of ribosomal gene transcription. Nat Genet 2002; 32:393-396. 73. Jansa P, Grummt I. Mechanism of transcription termination: PTRF interacts with the largest subunit of RNA polymerase I and dissociates paused transcription complexes from yeast and mouse. Mol Gen Genet 1999; 262:508-5014 74. Jansa P, Burek C, Sander EE et al. The transcript release factor PTRF augments ribosomal gene transcription by facilitating reinitiation of RNA polymerase I. Nucl Acids Res 2001; 29:423-429 75. Sentenac A, Riva M, Thuriaux P et al. Yeast RNA polymerase subunits and genes In: McKnight SL, Yamamoto KR, ed. Textbook of Transcription Regulation. Cold Spring Harbor Lab Press, 1992. 76. Goldberg MI, Perriard JC, Rutter WJ. Purification of rat liver and mouse ascites DNA-dependent RNA polymerase I. Biochemistry 1977; 16:1655-1665. 77. Song CZ, Hanada KI, Yano KI et al. High conservation of subunit composition of RNA polymerase IA between yeast and mouse and the molecular cloning of mouse RNA polymerase I 40-kDa subunit RPA40. J Biol Chem 1994; 269:26976-26981. 78. Schwartz LB, Roeder RG. Purification and subunit structure of deoxyribonucleic acid-dependent ribonucleic acid polymerase I from the mouse myeloma MOPC 315. J Biol Chem 1974; 249:5898-5906. 79. Matsui T, Onishi T, Muramatsu M. Nucleolar DNA-dependent RNA polymerase from rat liver 1 Purification and subunit structure. Eur J Biochem 1976; 71:351-360. 80. Hannan RD, Hempel WM, Cavanaugh AH et al. Affinity purification of mammalian RNA polymerase I Identification of an associated kinase. J Biol Chem 1998; 273:1257-1267. 81. Hanada KI, Song CZ, Yamamoto K et al. RNA polymerase I associated factor 53 binds to the nucleolar factor UBF and functions in specific rDNA transcription. EMBO J 1996; 15:2217-2226. 82. Gadal O, Mariotte-Labarre S, Chedin S et al. A34.5 a nonessential component of yeast RNA polymerase I cooperates with subunit A14 and DNA topoisomerase I to produce a functional rRNA synthesis machine. Mol Cell Biol 1997; 17:1787-1795. 83. Riva M, Schaffner AR, Sentenac A et al. Active site labeling of the RNA polymerases A, B and C from yeast. J Biol Chem 1987; 262:14377-14380. 84. Sawadogo M, Sentenac A. RNA polymerase B II and general transcirption factors. Annu Rev Biochem 1990; 59:711-754. 85. Tower J, Sollner-Webb B. Transcription of mouse rDNA is regulated by an activated subform of RNA polymerase I. Cell 1987; 50:873-883.

122

The Nucleolus

86. Cavanaugh AH, Thompson Jr EA. Hormonal regulation of transcription of rDNA: Glucocorticoid effects upon initiation and elongation in vitro. Nucl Acids Res 1985; 13:3357-3369. 87. Schnapp A, Pfleiderer C, Rosenbauer R et al. A growth-dependent transcription initiation factor TIF-IA interacting with RNA polymerase I regulates mouse ribosomal RNA synthesis. EMBO J 1990; 9:2857-2863. 88. Schnapp G, Santori F, Carles C et al. The HMG box-containing nucleolar transcription factor UBF interacts with a specific subunit of RNA polymerase I. EMBO J 1994; 13:190-199. 89. Seither P, Zatsepina O, Hoffmann M et al. Constitutive and strong association of PAF53 with RNA polymerase I. Chromosoma 1997; 106:216-225. 90. Hempel WM, Cavanaugh AH, Hannan RD et al. The species-specific RNA polymerase I transcription factor SL-1 binds to upstream binding factor. Mol Cell Biol 1996; 16:557-563. 91. Yu FL, Feigelson P. A proposed model for the glucocorticoidal regulation of rat hepatic ribosomal RNA synthesis. Biochem Biophys Res Comm 1973; 53:754-760. 92. Brun RP, Ryan K, Sollner-Webb B. Factor C* the specific initiation component of the mouse RNA polymerase I holoenzyme is inactivated early in the transcription process. Mol Cell Biol 1994; 14:5010-5021. 93. Mahajan PB, Thompson EA. Hormonal regulation of transcription of rDNA: Purification and characterization of the hormone-regulated transcription factor IC. J Biol Chem 1990; 265:16225-16233. 94. Schnapp A, Schnapp G, Erny B et al. Function of the growth-regulated transcription initiation factor TIF-IA in initiation complex formation at the murine ribosomal gene promoter. Mol Cell Biol 1993; 13:6723-6732. 95. Mahajan PB, Gokal PK, Thompson EA. Hormonal regulation of transcription of rDNA: The role of TFIC in formation of initiation complexes. J Biol Chem 1990; 265:16244-16247. 96. Yamamoto RT, Nogi Y, Dodd JA et al. Rrn3 gene of Saccharomyces cerevisiae encodes an essential RNA polymerase I transcription factor which interacts with the polymerase independently of DNA template. EMBO J 1996; 15:3964-3973. 97. Moorefield B, Greene EA, Reeder RH. RNA polymerase I transcription factor Rrn3 is functionally conserved between yeast and human. Proc Natl Acad Sci USA 2000; 97:4724-4729. 98. Bodem J, Dobreva G, Hoffmann-Rohrer U et al. TIF-IA, the factor mediating growth-dependent control of ribosomal RNA synthesis, is the mammalian homolog of yeast Rrn3p. EMBO Rep 2000; 1:171-175. 99. Cavanaugh AH, Hirschler-Laszkiewicz I, Hu Q et al. Rrn3 phosphorylation is a regulatory checkpoint for ribosome biogenesis. J Biol Chem 2002; 277:27423-27432. 100. Hirschler-Laszkiewicz I, Cavanaugh A, Mizra A et al. Rrn3 Becomes Inactivated in the Process of Ribosomal DNA Transcription. J Biol Chem, in press. 101. Milkereit HP, Tschochner H. A specialized form of RNA polymerase I, essential for initiation and growth-dependent regulation of rRNA synthesis, is disrupted during transcription. EMBO J 1998; 17:3692-3703. 102. Miller G, Panov KI, Friedrich JK et al. hRRN3 is essential in the SL1-mediated recruitment of RNA Polymerase I to rRNA gene promoters. EMBO J 2001; 20:1373-1382 103. Peyroche G, Milkereit P, Bischler N et al. The recruitment of RNA polymerase I on rDNA is mediated by the interaction of the A43 subunit with Rrn3. EMBO J 2000; 19:5473-5482. 104. Aprikian P, Moorefield B, Reeder RH. New model for the yeast RNA polymerase I transcription cycle. Mol Cell Biol 2001; 21:4847-4855. 105. Schnapp G, Schnapp A, Rosenbauer H et al. TIF-IC a factor involved in both transcription initiation and elongation of RNA polymerase I. EMBO J 1994; 13:4028-4035. 106. Liljelund P, Mariotte S, Buhler JM et al. Characterization and mutagenesis of the gene encoding the A49 subunit of RNA polymerase A in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1992; 89:9302-9305. 107. Seither P, Zatsepina O, Hoffmann M et al. Constitutive and strong association of PAF53 with RNA polymerase I. Chromosoma 1997; 106:216-225. 108. Hannan K, Rothblum LI, Jefferson L. Insulin and the regulation of rDNA transcription. Am J Physiol 1998; 275:C130-C138. 109. Jantzen HM, Chow AM, King DS et al. Multiple domains of the RNA polymerase I activator hUBF interact with the TATA-binding protein complex hSL-1 to mediate transcription. Genes Dev 1992; 6:1950-1963. 110. Radebaugh CA, Matthews JL, Geiss GK et al. TATA box-binding protein TBP is a constituent of the polymerase I-specific transcription initiation factor TIF-IB SL-1 bound to the rRNA promoter and shows differential sensitivity to TBP-directed reagents in Polymerase I II and III transcription factors. Mol Cell Biol 1994; 14:597-605. 111. Smith SD, O’Mahony DJ, Kinsella T et al. Transcription from the rat 45S ribosomal DNA promoter does not require the factor UBF. Gene Expr 1993; 3:229-236.

Ribosomal DNA Transcription in Mammals

123

112. Voit R, Schnapp A, Kuhn A et al. The nucleolar transcription factor mUBF is phosphorylated by casein kinase II in the C-terminal hyperacidic tail which is essential for transactivation. EMBO J 1992; 11:2211-2218. 113. Jantzen HM, Admon A, Bell SP et al. Nucleolar transcription factor hUBF contains a DNA-binding motif with homology to HMG proteins. Nature 1990; 344:830-836. 114. Hannan RD, Stefanovsky V, Taylor L et al. Overexpression of the transcription factor UBF1 is sufficient to increase ribosomal DNA transcription in neonatal cardiomyocytes: Implications for cardiac hypertrophy. Proc Natl Acad Sci 1996; 93:8750-8755. 115. Hannan R, Stefanovsky V, Arino T et al. Cellular Regulation of Ribosomal DNA Transcription: Both Rat and Xenopus UBF1 Stimulate rDNA Transcription in 3T3 Fibroblasts. Nucl Acids Res 1999; 27:1205-1213. 116. Siddiqi I, Keener J, Vu L et al. Role of TATA binding protein (TBP) in yeast ribosomal DNA transcription by RNA polymerase I: Defects in the dual functions of transcription factor UAF cannot be suppressed by TBP. Mol Cell Biol 2001; 21:2292-2297. 117. Colbert T, Hahn S. A yeast TFIIB-related factor involved in RNA polymerase III transcription. Genes Dev 1992; 61:1940-1949. 118. Huet J, Sentenac A. The TATA-binding protein participates in TFIIIB assembly on tRNA genes. Nucleic Acids Res 1992; 20:6451-6454. 119. Kassavetis GA, Joazeiro CA, Pisano M et al. The role of the TATA-binding protein in the assembly and function of the multisubunit yeast RNA polymerase III transcription factor TFIIIB. Cell 1992; 71:1055-1064. 120. Schultz MC, Reeder RH, Hahn S. Variants of the TATA-binding protein can distinguish subsets of RNA polymerase I II and III promoters. Cell 1992; 69:697-702. 121. Cormack BP, Struhl K. The TATA-binding protein is required for transcription by all three nuclear RNA polymerases in yeast cells. Cell 1992; 69:685-696. 122. Lalo D, Steffan JS, Dodd JA et al. RRN11 encodes the third subunit of the complex containing Rrn6p and Rrn7p that is essential for the initiation of rDNA transcription by yeast RNA polymerase I. J Biol Chem 1996; 271:21062-21067. 123. Keys DA, Vu L, Steffan JS et al. RRN6 and RRN7 encode subunits of a multiprotein complex essential for the initiation of rDNA transcription by RNA polymerase I in Saccharomyces cerevisiae. Genes Dev 1994; 8:2349-2362. 124. Keys DA, Lee BS, Dodd JA et al. Multiprotein transcription factor UAF interacts with the upstream element of the yeast RNA polymerase I promoter and forms a stable preinitiation complex. Genes Dev 1996; 10:887-903. 125. Lin CW, Moorefield B, Payne J et al. A novel 66-kilodalton protein complexes with Rrn6 Rrn7 and TATA-binding protein to promote polymerase I transcription initiation in Saccharomyces cerevisiae. Mol Cell Biol 1996; 16:6436-6443. 126. Steffan JS, Keys DA, Dodd JA et al. The role of TBP in rDNA transcription by RNA polymerase I in Saccharomyces cerevisiae: TBP is required for upstream activation factor-dependent recruitment of core factor. Genes Dev 1996; 10:2551-2563. 127. Schnapp A, Clos J, Hadelt W et al. Isolation and functional characterization of TIF-IB a factor that confers promoter specificity to mouse RNA polymerase I. Nucleic Acids Res 1990; 18:1385-1393. 128. Gong X, Radebaugh CA, Geiss GK et al. Site-directed photo-cross-linking of rRNA transcription initiation complexes. Mol Cell Biol 1995; 15:4956-4963. 129. Geiss G, Radebaugh C, Paule M. The Fundamental Ribosomal RNA Transcription Initiation Factor TIF-IB TIF-IB SL1 Factor D Binds to the rRNA Core Promoter Primarily by Minor Groove Contacts. J Biol Chem 1997; 272:29243-29254. 130. Perna P, Harris G, Iida C et al. The start site of the Acanthamoeba castellanii ribosomal RNA transcription unit. Gene Exp 1992; 2:71-78. 131. Radebaugh CA, Kubaska WM, Hoffman LH et al. A novel transcription initiation factor (TIF), TIF-IE, is required for homogeneous Acanthamoeba castellanii TIF-IB (SL1) to form a committed complex. J Biol Chem 1998; 273:27708-27715. 132. Al-Khouri AM, Paule MR. A novel RNA polymerase I transcription initiation factor, TIF-IE, commits rRNA genes by interaction with TIF-IB, not by DNA binding. Mol Cell Biol 2002; 22 :750-761. 133. Comai L, Zomerdijk JCBM, Beckmann H et al. Reconstitution of transcription factor SL-1: Exclusive binding of TBP by SL-1 or TFIID subunits. Science 1994; 266:1966-1972. 134. Comai L, Tanese N, Tjian R. The TATA-binding protein and associated factors are integral components of the RNA polymerase I transcription factor SL-1. Cell 1992; 68:965-976. 135. Eberhard D, Tora L, Egly JM et al. A TBP-containing multiprotein complex TIF-IB mediates transcription specifically of murine RNA polymerase I. Nucleic Acids Res 1993; 21:4180-4186.

124

The Nucleolus

136. Zomerdijk JCBM, Beckmann H, Comai L et al. Assembly of transcriptionally active RNA polymerase I initiation factor SL-1 from recombinant subunits. Science 1994; 266:2015-2018. 137. Helix J, Zomerdijk JCBM, Ravanpay A et al. Cloning of murine RNA polymerase I-specific TAF factors: Conserved interactions between subunits of the species-specific transcription factor TIF-IB/ SL-1. Proc Natl Acad Sci 1997; 94:1733-1738. 138. McStay B, Hu CH, Pikaard CS et al. xUBF and Rib1 are both required for formation of a stable polymerase I promoter complex with Xlaevis. EMBO J 1991; 10:2297-2303. 139. Rudloff U, Eberhard D, Grummt I. The conserved core domain of the human TATA binding protein is sufficient to assemble the multisubunit RNA polymerase I-specific transcription factor SL-1. Proc Natl Acad Sci 1994; 91:8229-8233. 140. Learned RM, Learned TK, Haltiner MM et al. Human rRNA transcription is modulated by the coordinant binding of two factors to an upstream control element. Cell 1986; 45:847-857. 141. Clos J, Buttgereit D, Grummt I. A purified transcription factor TIF-IB binds to essential sequences of the mouse rDNA promoter. Proc Natl Acad Sci 1986; 83:604-608. 142. Bell SP, Learned RM, Jantzen HM et al. Functional cooperativity between transcription factors UBF1 and SL-1 mediates human ribosomal RNA synthesis. Science 1988; 241:1192-1197. 143. Bell SP, Jantzen HM, Tjian R. Assembly of alternative multiprotein complexes directs rRNA promoter selectivity. Genes Dev 1990; 4:943-954. 144. Beckmann H, Chen JL, O’Brien T et al. Coactivator and promoter-selective properties of RNA polymerase I TAFs. Science 1995; 270:1506-1509. 145. Rudloff U, Eberhard D, Tora L et al. TBP-associated factors interact with DNA and govern species specificity of RNA polymerase I transcription. EMBO J 1994; 13:2611-2616. 146. Yang Q, Radebaugh CA, Kubaska W et al. Acanthamoeba castellanii contains a ribosomal RNA enhancer binding protein which stimulates TIF-IB binding and transcription under stringent conditions. Nucleic Acids Res 1995; 23:4345-4352. 147. Kwon H, Green MR. The RNA polymerase I transcription factor upstream binding factor interacts directly with the TATA box-binding protein. J Biol Chem 1994; 269:30140-30146. 148. Mishima Y, Financsek I, Kominami R et al. Fractionation and reconstitution of factors required for accurate transcription of mammalian ribosomal RNA genes: Identification of a species-dependent initiation factor. Nucleic Acids Res 1982; 10:6659-6670. 149. Grummt I, Roth E, Paule MR. Ribosomal RNA transcription in vitro is species specific. Nature 1982; 296:173-174. 150. Pikaard GS, Smith DS, Reeder RH et al. rUBF an RNA Polymerase I transcription factor from rats produces Dnase I footprints identical to those produced by xUBF its homolog from frogs. Mol Cell Biol 1990; 10:3810-3812. 151. Learned RM, Cordes S, Tjian R. Purification and characterization of a transcription factor that confers promoter specificity to human RNA polymerase I. Mol Cell Biol 1985; 5:1358-1369. 152. Bell SP, Pikaard CS, Reeder RH et al. Molecular mechanisms governing species-specific transcription of ribosomal RNA. Cell 1989; 59:489-497. 153. Ghosh AK, Niu H, Jacob ST. Rat ribosomal RNA gene can utilize primate RNA polymerase I transcription machinery: Lack of absolute species specificity in rDNA transcription. Biochem Biophys Res Commun 1996; 225:890-895. 154. Jantzen HM, Admon A, Bell SP et al. Nucleolar transcription factor hUBF contains a DNA-binding motif with homology to HMG proteins. Nature 1990; 344:830-836. 155. Hisatake K, Nishimura T, Maeda Y et al. Cloning and structural analysis of cDNA and the gene for mouse transcription factor UBF. Nucleic Acids Res 1991; 19:4631-4637. 156. O’Mahony DJ, Rothblum LI. Identification of two forms of the RNA polymerase I transcription factor UBF. Proc Natl Acad Sci 1991; 88:3180-3184. 157. Bachvarov D, Normandaeu M, Moss T. Heterogeneity in the Xenopus ribosomal transcription factor xUBF has a molecular basis distinct from that in mammals. FEBS 1991; 288:55-59. 158. Bachvarov D, Moss T. The RNA polymerase I transcription factor xUBF contains 5 tandemly repeated HMG homology boxes. Nucleic Acids Res 1991; 19:2331-2335. 159. Kuhn A, Voit R, Stefanovsky V et al. Functional differences between the two splice variants of the nucleolar transcription factor UBF: The second HMG box determines specificity of DNA binding and transcriptional activity. EMBO J 1994; 13:416-424. 160. McStay B, Sullivan GJ, Cairns C. The Xenopus RNA polymerase I transcription factor UBF has a role in transcriptional enhancement distinct from that at the promoter. EMBO J 1997; 16:396-405. 161. Nishimura T, Hanada KI, Maeda Y et al. Regulation of mouse UBF gene by multiple growth-related control elements. Biochem Biophysic Res Commun 1994; 205:1217-1225. 162. Cairns C, McStay B. HMG box 4 is the principal determinant of specifies specificity in the RNA polymerase I transcription factor UBF. Nucleic Acids Res 1995; 23:4583-4590. 163. Hu CH, McStay B, Jeong S-W et al. xUBF an RNA polymerase I transcription factor binds crossover DNA with low sequence specificity. Mol Cell Biol 1994; 14:2871-2882.

Ribosomal DNA Transcription in Mammals

125

164. McStay B, Frazier MW, Reeder RH. xUBF contains a novel dimerization domain essential for RNA polymerase I transcription. Genes Dev 1991; 5:1957-1968. 165. Maeda Y, Histake K, Kondo T et al. Mouse rRNA gene transcription factor mUBF requires both HMG-box1 and an acidic tail for nucleolar accumulation: Molecular analysis of the nucleolar targetting mechanism. EMBO J 1992; 11:3695-3704. 166. Leblanc B, Read C, Moss T. Recognition of the Xenopus ribosomal core promoter by the transcription factor xUBF involves multiple HMG box domains and leads to an xUBF interdomain interaction. EMBO J 1993; 12:513-525. 167. O’Mahony DJ, Smith SD, Xie W et al. Analysis of the phosphorylation DNA-binding and dimerization properties of the RNA polymerase I transcription factors UBF1 and UBF2. Nucleic Acids Res 1992; 20:1301-1308. 168. Copenhaver GP, Putnam CD, Denton ML et al. The RNA polymerase I transcription factor UBF is a sequence-tolerant HMG-box protein that can recognize structural nucleic acids. Nucleic Acids Res 1994; 22:2651-2657. 169. Putnam CD, Copenhaver GP, Denton ML et al. The RNA polymerase I transactivator upstream binding factor requires its dimerization domain and high-mobility-group HMG box 1 to bend wrap and positively supercoil enhancer DNA. Mol Cell Biochem 1994; 14:6476-6488. 170. Luyken J, Hannan RD, Cheung JY et al. Regulation of rDNA transcription during endothelin-1-induced hypertrophy of neonatal cardiomyocytes Hyperphosphorylation of upstream binding factor an rDNA transcription factor. Circ Res 1996; 78:354-361. 171. White RJ. Regulation of RNA polymerase I and III by the retinoblastoma protein: A mechanism for growth control? Trends in Biol Sci 1997; 22:77-80. 172. Hannan KM, Hannan RD, Smith SD et al. Rb and p130 regulate RNA polymerase I transcription: Rb disrupts the interaction between UBF and SL-1. Oncogene 2000; 19:4988-4999 173. Zhang J, Jacob ST. Purification and characterization of a novel factor which stimulates rat ribosomal gene transcription in vitro by interacting with enhancer and core promoter elements. Mol Cell Biol 1990; 10:5177-5186. 174. Niu H, Zhang J, Jacob St. E1BF/Ku interacts physically and functionally with the core promoter binding factor CPBP and promotes the basal transcription of rat and human ribosomal RNA genes. Gene Exp 1995; 4:111-124. 175. Kuhn A, Stefanovsky V, Grummt I. The nucleolar transcription activator UBF relieves Ku antigen-mediated repression of mouse ribosomal gene transcription. Nucleic Acids Res 1993; 21:2057-2063. 176. Hoff CM, Jacob ST. Characterization of the factor E1 BF from a rat hepatoma that modulates ribosomal RNA gene transcription and its relationship to the human Ku autoantigen. Biochem Biophys Res Commun 1993; 190:747-753. 177. Ghosh AK, Hoff CM, Jacob ST. Characterization of the 130-bp repeat enhancer element of the rat ribosomal gene: Functional interaction with transcription factor E1 BF. Gene 1993; 125:217-222. 178. Labhart P. DNA-dependent protein kinase specifically represses promoter-directed transcription initiation by RNA polymerase I. Proc Natl Acad Sci 1995; 92:2934-2938. 179. Kuhn A, Gottlieb TM, Jackson SP et al. DNA-dependent protein kinase: A potent inhibitor of transcription by RNA polymerase I. Genes Dev 1995; 9:193-203. 180. Liu Z, Jacob ST. Characterization of a protein that interacts with the rat ribosomal gene promoter and modulates RNA polymerase I transcription. J Biol Chem 1994; 269:16618-16626. 181. Datta PK, Ghosh AK, Jacob ST. The RNA polymerase I promoter-activating factor CPBP is functionally and immunologically related to the basic helix-loop-helix-zipper DNA-binding protein USF. J Biol Chem 1995; 270:8637-8641. 182. Ghosh AK, Datta PK, Jacob ST. The dual role of helix-loop-helix-zipper protein USF in ribosomal RNA gene transcription in vivo. Oncogene 1997; 14:589-594. 183. Wang JC. DNA topoisomerases. Ann Rev Biochem 1985; 54:665-697. 184. Brill SJ, DiNardo S, Voelkel-Meiman K et al. Need for DNA topoisomerase activity as a swivel for DNA replication for transcription of ribosomal RNA. Nature 1987; 326:414-416. 185. Zini N, Santi S, Ognibene A et al. Discrete localization of different DNA topoisomerases in HeLa and K562 cell nuclei and subnuclear fractions. Exp Cell Res 1994; 210:336-348. 186. Brou C, Kuhn A, Staub A et al. Sequence-specific transactivators counteract topoisomerase II-mediated inhibition of in vitro transcription by RNA polymerases I and II. Nucleic Acids Res 1993; 21:4011-4018. 187. Muller MT, Pfund WP, Mehta VB et al. Eukaryotic type I topoisomerase is enriched in the nucleolus and catalytically active on ribosomal DNA. EMBO J 1985; 4:1237-1243. 188. Fleischmann G, Pflugfelder G, Steiner EK et al. Drosophila DNA topoisomerase I is associated with transcriptionally active regions of the genome. Proc Natl Acad Sci 1984; 81:6958-6962.

126

The Nucleolus

189. Schultz MC, Brill SJ, Ju Q et al. Topoisomerases and yeast rRNA transcription: Negative supercoiling stimulates initiation and topoisomerase activity is required for elongation. Genes Dev 1992; 6:1332-1341. 190. Yang-Yen HF, Rothblum LI. Purification and characterization of a high-mobility-group-like DNA-binding protein that stimulates rRNA synthesis in vitro. Mol Cell Biol 1988; 8:3406-3414. 191. Conaway RC, Conaway JW. General Initiation Factors for RNA Polymerase II. Ann Rev Biochem 1993; 62:161-190. 192. Conaway JW, Conaway RC. Transcription elongation and human disease. Ann Rev Biochem 1999; 68:301-319. 193. Kugel JF, Goodrich J. A kinetic model for the early steps of RNA synthesis by human RNA polymerase II. J Biol Chem 2000; 275:40483-40491. 194. Dvir A, Conaway JW, Conaway RC. Mechanism of transcription initiation and promoter escape by RNA polymerase II. Curr Opin Genet Dev 2001; 11:209-214. 195. Dvir A. Promoter escape by RNA polymerase II. Biochim Biophys Acta 2002; 1577:208-223.. 196. Hsu LM. Promoter clearance and escape in prokaryotes. Biochim Biophys Acta 2002; 1577:191-207. 197. Kumar KP, Akoulitchev S, Reinberg D. Promoter-proximal stalling results from the inability to recruit transcription factor IIH to the transcription complex and is a regulated event. Proc Natl Acad Sci USA 1998; 95:9767–9772. 198. Spangler L, Wang X, Conaway JW et al. TFIIH action in transcription initiation and promoter escape requires distinct regions of downstream promoter DNA. Proc Natl Acad Sci USA 2001; 98:5544–5549. 199. Iben S, Tschochner H, Bier M et al. TFIIH plays an essential role in RNA polymerase I transcription. Cell 2002; 109:297-306. 200. Lee SK, Yu SL, Prakash L et al. Requirement for yeast RAD26, a homolog of the human CSB gene, in elongation by RNA polymerase II. Mol Cell Biol 2001; 21:8651-8656. 201. Bradsher J, Auriol J, Proietti de Santis L et al. CSB is a component of RNA pol I transcription. Mol Cell 2002; 10:819-829. 202. Wang EH, Zou S, Tjian R.TAFII250-dependent transcription of cyclin A is directed by ATF activator proteins. Genes Dev 1997; 11:2658-2669. 203. Lin CY, Tuan J, Scalia P et al. The Cell Cycle Regulatory Factor TAF1 Stimulates Ribosomal DNA Transcription by Binding to the Activator UBF. Curr Biol 2002; 12:2142-2146 204. Bateman E, Paule MR. Events during eukaryotic rRNA transcription initiation and elongation: Conversion from the closed to the open promoter complex requires nucleotide substrates. Mol Cell Biol 1988; 8:1940-1946. 205. Kownin P, Bateman E, Paule MR. Eukaryotic RNA polymerase I promoter binding is directed by protein contacts with transcription initiation factor and is DNA sequence-independent. Cell 1987; 50:693-699. 206. Kato H, Nagamine M, Kominami R et al. Formation of the transcription initiation complex on mammalian rDNA. Mol Cell Biol 1986; 6:3418-3427. 207. Schnapp A, Grummt I. Transcription complex formation at the mouse rDNA promoter involves the stepwise association of four transcription factors and RNA polymerase I. J Biol Chem 1991; 266:24588-24595. 208. Dundr M, Hoffmann-Rohrer U, Hu Q et al. A kinetic framework for a mammalian RNA polymerase in vivo. Science 2002; 298:1623-1626. 209. Kuhn A, Grummt I. Dual role of the nucleolar transcription factor UBF: Trans-activator and antirepressor. Proc Natl Acad Sci 1992; 89:7340-7344. 210. Derenzini M, Hernandez-Verdun D, Farabegoli F et al. Structure of ribosomal genes of mammalian cells in situ. Chromosoma 1987; 95:63-70. 211. Gebrane-Younes J, Fomproix N, Hernandez-Verdun D. When rDNA transcription is arrested during mitosis UBF is still associated with noncondensed rDNA. J Cell Science 1997; 110:2429-2440. 212. Conconi A, Sogo JM, Ryan CA. Ribosomal gene clusters are uniquely proportioned between open and closed chromatin structures in both tomato leaf cells and exponentially growing suspension cultures. Proc Natl Acad Sci USA 1992; 89:5256-5260. 213. Banditt M, Koller T, Sogo JM. Transcriptional activity and chromatin structure of enhancer-deleted rRNA genes in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:4953-4960. 214. Moss T, Stefanovsky VY. At the center of eukaryotic life. Cell 2002; 109:545-548. 215. French SL, Osheim Y, Cioci F et al. In Exponentially Growing Saccharomyces cerevisiae Cells, rRNA Synthesis Is Determined by the Summed RNA Polymerase I Loading Rate Rather than by the Number of Active Genes. Molecular and Cellular Biology 2002; 23:1558-1568. 216. Bateman E and Paule Mr. Regulation of eukaryotic ribosomal RNA transcription by RNA polymerase modification. Cell 1986; 47:445-450. 217. Fath S, Milkereit P, Peyroche G et al. Differential roles of phosphorylation in the formation of transcriptional active RNA polymerase I. Proc Natl Acad Sci USA 2001; 98:14334-14339.

Ribosomal DNA Transcription in Mammals

127

218. Zhao J, Yuan X, Frodin M et al. ERK-Dependent Phosphorylation of the Transcription Initiation Factor TIF-IA Is Required for RNA Polymerase I Transcription and Cell Growth. Molecular Cell 2003; 11:405–413. 219. Zhai W, Tuan JA, Comai L. SV40 large T antigen binds to the TBP-TAFI complex SL-1 and coactivates ribosomal RNA transcription. Genes Dev 1997; 11:1605-1617. 220. Zhai W, Comai LA. kinase activity associated with simian virus 40 large T antigen phosphorylates upstream binding factor (UBF) and promotes formation of a stable initiation complex between UBF and SL1. Mol Cell Biol 1999; 19:2791-2802. 221. Kuhn A, Vente A, Doree M et al. Mitotic phosphorylation of the TBP-containing factor SL1 represses ribosomal gene transcription. J Mol Biol 1998; 284:1-5. 222. Comai L, Song Y, Tan C et al. Inhibition of RNA polymerase I transcription in differentiated myeloid leukemia cells by inactivation of selectivity factor 1. Cell Growth Differ 2000; 11:63-70. 223. Muth V, Nadaud S, Grummt I et al. Acetylation of TAF(I)68, a subunit of TIF-IB/SL1, activates RNA polymerase I transcription. EMBO J 2001; 20:1353-1362. 224. Pelletier G, Stefanovsky VY, Faubladier M et al. Competitive recruitment of CBP and Rb-HDAC regulates UBF acetylation and ribosomal transcription. Mol Cell 2000; 6:1059-1066. 225. Kabler RL, Srinivasan A, Taylor LJ et al. Androgen regulation of ribosomal RNA synthesis in LNCaP cells and rat prostate. J Steroid Biochem Mol Biol 1996; 59:431-439. 226. Hershey JC, Hautmann M, Thompson MM et al. Angiotensin II-induced hypertrophy of rat vascular smooth muscle is associated with increased 18S rRNA synthesis and phosphorylation of the rRNA transcription factor upstream binding factor. J Biol Chem 1995; 270:25096-25101. 227. Stefanovsky VY, Pelletier G, Hannan R et al. An Immediate Response of Ribosomal Transcription to Growth Factor Stimulation in Mammals Is Mediated by ERK Phosphorylation of UBF. Mol Cell 2001; 8:1063–1073. 228. Klein J, Grummt I. Cell cycle-dependent regulation of RNA polymerase I transcription: The nucleolar transcription factor UBF is inactive in mitosis and early G1. Proc Natl Acad Sci USA 1999; 96:6096-6101. 229. Voit R, Hoffmann M, Grummt I. Phosphorylation by G1-specific cdk-cyclin complexes activates the nucleolar transcription factor UBF. EMBO J 1999; 181891-1899. 230. Levine AJ. The tumor suppressor genes. Ann Rev Biochem 1993; 62:623-651. 231. Hannan KM, Kennedy BK, Cavanaugh AH et al. RNA polymerase I transcription in confluent cells: Rb downregulates rDNA transcription during confluence-induced cell cycle arrest. Oncogene 2000; 19:3487-3497 232. Voit R, Schafer K, Grummt I. Mechanism of repression of RNA polymerase I transcription by the retinoblastoma protein. Mol Cell Biol 1997; 17:4230-4237. 233. Hoff CM, Ghosh AK, Prabhakar BS et al. Enhancer 1 binding factor a Ku-related protein is a positive regulator of RNA polymerase I transcription initiation. Proc Natl Acad Sci 1994; 91:762-766. 234. Niu H, Jacob ST. Enhancer 1 binding factor a Ku-related protein is a growth-regulated RNA polymerase I transcription factor: Association of a repressor activity with purified E1 BF from serum-deprived cells. Proc Natl Acad Sci 1994; 91:9101-9105. 235. Fewell JW, Kuff EL. Intracellular redistribution of Ku immunoreactivity in response to cell-cell contact and growth modulating components in the medium. J Cell Sci 1996; 109:1937-1946. 236. Datta PK, Budhiraja S, Reichel RR et al. Regulation of ribosomal RNA gene transcription during retinoic acid-induced differentiation of mouse teratocarcinoma cells. Exp Cell Res 1997; 231:198-205. 237. Michaelidis TM, Grummt I. Mechanism of inhibition of RNA polymerase I transcription by DNA-dependent protein kinase. Biol Chem 2002; 383:1683-1690 238. Ghoshal K, Jacob ST. Heat shock selectivity inhibits ribosomal RNA gene transcription and down-regulates E1BF/Ku in mouse lymphosarcoma cells. Biochem J 1996; 317:689-695. 239. Zhang H, Wang JC, Liu LF. Involvement of DNA topoisomerase I in transcription of human ribosomal RNA genes. Proc Natl Acad Sci 1988; 85:1060-1064. 240. Kretzschmar M, Meisterernst M, Roeder RG. Identification of human DNA topoisomerase I as a cofactor for activator-dependent transcription by RNA polymerase II. Proc Natl Acad Sci 1993; 90:11508-11512. 241. Rose KM, Szopa J, Han FS et al. Association of DNA topoisomerase I and RNA polymerase I: A possible role for topoisomerase I in ribosomal gene transcription. Chromosoma 1988; 96:411-416. 242. Rose KM. DNA topoisomerases as targets for chemotherapy. FASEB 1988; 2:2474-2478. 243. Culotta VC, Wilkinson JK, Sollner-Webb B. Mouse and frog violate the paradigm of species– specific transcription of ribosomal RNA genes. Proc Natl Acad Sci 1987; 84:7498-7502.

128

The Nucleolus

CHAPTER 9

Transcription of rDNA in the Yeast Saccharomyces cerevisiae Masayasu Nomura, Yasuhisa Nogi and Melanie Oakes

Introduction

T

he discovery of rDNA as a nucleolar organizer in metazoan systems, such as Xenopus laevis1,2 and Drosophila melanogaster3 initiated studies of rRNA synthesis, its process ing and ribosome assembly as the major function of the nucleolus in eukaryotes. Although earlier studies were mostly done in metozoan cells, the yeast Saccharomyces cerevisiae has now developed as an important model system because of the ability to combine powerful genetic and various modern molecular and biochemical approaches. In this chapter, we discuss the transcription of rDNA by RNA polymerase I (Pol I) and related subjects in the yeast S. cerevesiae system, and additionally will mention other systems when deemed pertinent to discussion. The synthesis of rRNA and ribosomes in yeast and its regulation have been previously reviewed,4-9 as have been the yeast Pol I enzyme10,11 and the Pol I transcription factors.12 The present review will also cover these earlier studies, but mainly focuses on more recent developments and unsolved current questions.

rDNA Structure and Cis-Elements In the yeast S. cerevisiae, approximately 150 rDNA repeats are tandemly arranged on chromosome XII. A single unit of rDNA contains two transcribed genes (Fig. 1A), the 35S rRNA gene that is transcribed by Pol I and the gene for 5S rRNA, which is transcribed by RNA polymerase III (Pol III). The 35S rRNA transcript is a precursor that is subsequently processed to generate the mature 18S, 5.8S and 25S rRNAs. The presence of the 5S gene within the rDNA repeat unit in S. cerevisiae (and some other organisms; see ref. 13) is in contrast to other well-studied higher eukaryotes, such as mammals, in that these latter organisms have 5S rRNA gene repeats separately from the nucleolar rDNA repeats. [In this chapter, we do not include discussion of 5S rRNA gene and its transcription by Pol III. Pol III transcription has been reviewed recently].14 In addition to the 35S and 5S rRNA genes, there are several cis elements within the rDNA repeat, mostly NTS1 and NTS2 (nontranscribed) regions that are unrelated to transcription, such as the origin of DNA replication (ARS) and the replication fork block (RFB) site. Some of these cis elements studied are shown in (Fig. 1A). During DNA replication, one in approximately five ARS sites is initiated for bi-directional replication,15,16 and recent studies indicate that these active ARSs are clustered and their activities are perhaps subject to epigenetic control.17 The RFB site is located near the end of the 35S rRNA gene and allows the progression of the replication fork in the direction of 35S rRNA transcription but not in the opposite direction.15,16,18,19 Although the exact mechanism is not known for the RFB activity, the gene FOB1 is required for this activity.20 The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

129

Figure 1. A) The structure of the S. cerevisiae rDNA and the model of rRNA gene transcription by Pol I. A single copy of rDNA repeats is shown. The 9.1 kb unit consists of the 35S rRNA coding region, the 5S rRNA coding region and two nontranscribed regions, NTS1 and NTS2. The NTS region is expanded and positions of terminators (T1 at +93 and T2 at +250; see text), enhancer, RFB (replication fork blocking, indicated by a symbol of two black triangles) region, EXP (the region required for FOB1-dependent rDNA repeat expansion24), CAR (cohesin associated region40) and ARS (autonomous replication sequence) are indicated. B) The structure of the S. cerevisiae rDNA and the model of rRNA gene transcription by Pol I. UAF is bound tightly to the upstream element of the promoter, forming a UAF-template committed complex. The UAF structure in the figure indicates its subunit composition but does not imply interactions among subunits or those between subunits and template, which are largely unknown. Arrows connecting UAF, TBP, CF, Rrn3p and Pol I show observed interactions between protein subunits in these transcription factors and Pol I. 3, 5, 6, 7, 9, 10 and 11 represent Rrn3p, Rrn5p, Rrn6p, Rrn7p, Rrn9p, Rrn10p and Rrn11p, respectively, and 30 represents Uaf30p.

130

The Nucleolus

Some of the cis elements within the NTS1 region are additionally involved in expansion and contraction of rDNA repeats. Although most eukaryotes have similar repeat structures, they have different sizes and numbers of repeats. The numbers of rDNA repeats can vary from less than 100 to more than 10,000.21,22 The repeat numbers appear to be maintained at an appropriate level for any given organism, however most organisms are able to vary the number of repeats in response to intracellular as well as extracellular conditions. For instance in yeast, the absence of a Pol I essential subunit results in a decrease in the number of chromosomal rDNA repeats to about half (~80) of the normal number (~150) and reintroduction of the missing subunit results in the restoration of the normal number of rDNA repeats.23 Using this system, it was demonstrated that FOB1 is required for the expansion and contraction of rDNA repeats. It has been suggested that repeat expansion and contraction involves recombinational events triggered by DNA breakages caused by FOB1-dependent pausing of the DNA replication machinery at RFB sites.23 Furthermore, analysis of yeast strains, in which the majority of rDNA repeats were deleted and two copies of rDNA covering the 5S-NTS2-35S regions and a single intact NTS1 region remained, indicated that, as expected, the RFB site is essential for FOB1-dependent rDNA repeat expansion. In addition to the RFB region, however, the adjacent ~400 bp region (“EXP” in Fig. 1A) in NTS1 was also found to be required for efficient repeat expansion but the “enhancer” region (see below) is not. 24 The discovery of FOB1-dependent rDNA repeat expansion and contraction has allowed construction of yeast strains, in which rDNA copy numbers are altered and the altered numbers maintained stably. Such strains have proved useful for studies of rDNA transcription by Pol I, as will be mentioned below. Within the rDNA repeats, cis elements that are necessary or that enhance the transcription of the 35S rRNA gene have been extensively studied. In all eukaryotes, the gene promoter for the large precursor RNA consists of two cis elements, the upstream element and the core (Fig. 1B). In the yeast system, in vitro studies have shown the former is required for high level transcription but is dispensable whereas the latter is essential for accurate transcription initiation.25-27 Examination of the promoter elements in rDNA transcription in vivo was done by mutational analysis in two different ways. One approach was to use a single rDNA copy on a CEN plasmid and introduce a tag at a suitable site of the 25S rRNA coding region to distinguish the plasmid derived transcripts from those encoded by chromosomal rDNA. Mutational analyses could then be carried out using this plasmid-encoded rDNA.28 Another approach was to use yeast strains (rdn∆∆) in which chromosomal rDNA repeats were deleted completely and rRNA synthesis was achieved by a plasmid (“Pol I helper plasmid”) carrying a single rDNA repeat transcribed by Pol I or a plasmid (“Pol II helper plasmid”) carrying, in addition to the 5S rRNA gene, a GAL7-35S rDNA fusion gene transcribed by RNA polymerase II (Pol II). The GAL7-35S rDNA fusion gene consists of the 35S rRNA coding region fused to the GAL7 promoter and is transcribed by RNA polymerase II (Pol II) in the presence of galactose, but not in glucose.29 Pol I helper plasmids carrying various mutations were introduced into a rdn∆∆ strain carrying a Pol II helper plasmid grown in the presence of galactose and effects of mutation on rDNA transcription by Pol I were analyzed in glucose media.30 Using these systems, mutational analyses essentially confirmed the results obtained using in vitro systems. Deletion analysis using the second approach also revealed that basal transcription from the core promoter, if it occurs in vivo, is not sufficient to allow cell growth.30 Another element, called the enhancer, which lies at the end of the 35S rRNA gene, was originally shown to greatly stimulate rDNA transcription in ectopic reporter systems.31,32 Utilizing rdn∆∆ strains, however, deletion of the enhancer from the Pol I helper plasmid did not cause any decrease in rRNA synthesis or cell growth.33 Additionally, expression from the reporter with or without the enhancer in rdn∆∆ strains was greater than that observed in wild type strains. This suggests that the ectopic reporter genes are poorly accessible to Pol I machinery in the nucleolus and that the enhancer is somehow able to improve accessibility.33 Yeast strains were also constructed, in which the enhancer element was deleted from all of the

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

131

chromosomal rDNA repeats. No difference was observed between such enhancer-less mutants and the control strains in rRNA synthesis or cell growth.33 Thus, the enhancer element is totally dispensable for Pol I transcription, and the mechanism of stimulation of Pol I transcription in ectopic systems by the enhancer is presently unknown. Regarding Pol I transcription termination sites, earlier studies gave some conflicting results (see, e.g., ref. 34). However, it now appears to be certain that there are two terminators.35 Approximately 90% of all transcripts terminate at the site located 93 nucleotides downstream of the 3'-end of the 25S rRNA (“+93 site”). Those transcripts, which read through the +93 site, terminate downstream at a fail-safe termination site located at 250 nucleotides downstream of the 3'-end of the 25S rRNA. The mechanism of termination at the +93 site was studied extensively.36-38 The terminator contains two elements; an upstream T-rich element encoding the last 10-12 nucleotides of the transcript terminated at this site and the 11 bp Reb1p binding site which is 12 bp downstream from the termination site. In vitro experiments suggest that Reb1p bound to the second element causes Pol I to pause and the T-rich element plays a role in release of transcripts. The Pol I terminator in the mouse system also appears to have two elements, one causes Pol I pause by binding the termination factor (TTF-1) and the other stimulates transcript release. This similarity has been emphasized as a possible universal mechanism for Pol I transcription termination (see a review by Reeder and Lang39). Regarding the two yeast terminators mentioned above, it should be noted that the enhancer element originally defined by Elion and Warner31,32 contains the Reb1p binding site essential for termination at the +93 site as well as the second fail-safe terminator. Thus, deletion of the enhancer region, i.e., deletion of the two terminators, does not cause any significant decrease in the synthesis of rRNA or in growth rate, as demonstrated in the enhancer deletion experiments mentioned above.33 It appears that these two terminators are not required for processing of rRNA to form functional ribosomes. How Pol I transcription terminates in these enhancer deletion strains is not known. There are other cis elements studied in connection with rDNA chromatin structures. For example, using chromatin immunoprecipitation (ChIP) analysis, Koshland and coworkers40 demonstrated that each 9-kb rDNA repeat has a single cohesin-associated region which is located in NTS2 near the 5S gene (“CAR” in Fig. 1A). Binding sites for proteins essential for silencing, such as Sir2p and Net1p discussed in later sections, were also studied. Although resolution was not very high in these studies, it was reported that Sir2p binding at NTS2 near the 5S gene and at the Pol I promoter region was significantly higher than at regions within the 25S rRNA gene.41 Undoubtedly, future studies along this line will contribute to our understanding of rDNA chromatin structures which must influence rDNA transcription and other nucleolar functions.

Pol I, Transcription Factors, and the Mechanism for rRNA Transcription Initiation RNA Polymerase I (Pol I) Pol I is the enzyme responsible for the synthesis of rRNA using rRNA genes as template. In fact, the sole essential function of Pol I in S. cerevisiae is the synthesis of 35S pre-rRNA. This was demonstrated by galactose-dependent suppression of lethal mutations in essential Pol I subunit genes by a multi-copy plasmid carrying the GAL7-35S rDNA fusion gene.29 Pol I contains 14 protein subunits (Table 1; for reviews, see Thuriaux and Sentenac;10 Carles and Riva11). Ten of them are either identical or closely homologous to subunits of Pol II and Pol III, and therefore, the structures of polymerase portions containing these 10 protein subunits (“core subunits”) are expected to be similar among the three eukaryotic polymerases, Pol I, Pol II and Pol III. Thus, the recent success in elucidation of atomic structure of the yeast Pol II enzyme containing these 10 core subunits (“Pol II ∆4/7” which is missing PolII-specific proteins Rpb4 and Rpb7) by Kornberg and coworkers42-44 has proved to be extremely useful in building structure models for Pol I and in interpreting various biochemical and genetic data obtained

132

The Nucleolus

Figure 2. An interaction diagram for the Pol I subunits. The interaction diagrams published for the 10 Pol II (∆4/7) subunits42,43 are used as the basis and are first modified by replacing Pol II subunits with corresponding Pol I homologs (see core subunits in Table. 1). Approximate positions of “Pol I specific subunits” A49, A34.5, A43 and A14 are similarly indicated using information described by Bischler et al.65 The schematic diagram corresponds to the top view of the atomic structure of Pol II (∆4/7).43 Downstream DNA bound in the cleft is also shown. The lines connecting 10 core subunit proteins are according to Cramer et al.42,43 based on the subunit contacts in the Pol II atomic structure. The lines connecting A43, A14 and ABC23 are based on the results obtained for yeast Pol I.71

in structural and functional analyses of Pol I. Table 1 lists Pol I protein subunits and their homologs present in yeast Pol II and E. coli RNA polymerase. Figure 2 is a schematic diagram of the yeast Pol I adapted from the results obtained in the crystallographic studies of Pol II as well as the results on Pol I specific subunits discussed below. The prokaryotic RNA polymerase α2ββ’ω subunit composition is conserved in the three eukaryotic polymerases as determined by primary sequence, functional and structural homologies. The largest and second largest subunits of eukaryotic RNA polymerases I, II and III are homologous to the bacterial β’ and β subunits, respectively.45,46 These two subunits (A190 and A135 for Pol I) have extensive surface area buried in their interface and the active center containing two metal ions consists of residues coming from these two subunits.43 The eukaryotic homologues of the α dimer subunit are polypeptide subunits which are heterodimers, AC40/ AC19 for Pol I and Pol III and Rpb3/Rpb11 for Pol II.47-50 Finally, the eukaryotic polypeptide subunit ABC23 which is shared amongst all eukaryotic RNA polymerases, has structural and functional homology to the prokaryotic ω subunit.51 Although all eukaryotic RNA polymerases have the bacterial subunit homologs and structural homology is extensive, the five subunits in eukaryotes do not appear to be able to form a functional enzyme. Eukaryotes need an additional four subunits which are shared between all the RNA polymerases, ABC27, ABC14.5, ABC10α and ABC10β.52-54 The common conserved subunits are essential and important for assembly of RNA polymerase(s).51,55-58 Another subunit, although enzyme specific, A12.2,

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

133

Table 1. Yeast Pol I subunits and their homologs in yeast Pol II and E. coli RNA polymerase Protein (Gene; Requirement)*

P and Zn**

Homologs Pol II E. coli

P(6), Zn Zn P (1-2) P (1-2)

Rpb1 Rpb2 Rpb3 Rpb11 Rpb6

Zn Zn Zn

Rpb5 Rpb8 Rbp12 Rpb10 Rpb9

Core subunits A190 (RPA190; E) A135 (RPA135; E) AC40 (RPC40; E) AC19 (RPC19; E) ABC23 (RPB6; E) ABC27 (RPB5; E) ABC14.5 (RPB8; E) ABC10a (RPC10; E) ABC10b (RPB10; E) A12.2 (RPA12; C)

β’ β a α ω

Pol I specific A49 (RPA49; C) A34.5 (RPA34; N) A43 (RPA43; E) A14 (RPA14; N)

P (2) P (4)

– – (Rpb7)+ (Rpb4)+

Subunits of Pol I are listed. Pol I specific subunits are designated by letter A followed by a number corresponding to the approximate size (in kilodalton) as estimated by its migration in SDS gel electrophoresis. AC40 and AC19 are shared with Pol III, and ABC23, ABC27, ABC14.5, ABC10α and ABC10β are shared with Pol II and Pol III. Homologs present in S. cerevisiae Pol II and E. coli RNA polymerase are also listed. * The name of genes are listed together with their requirement for growth. E, essential; C, conditionally essential; N, nonessential. ** P, phosphorylated subunits and the numbers of phosphorylated residues estimated10,165 are indicated in parenthesis. Zn, subunits binding zinc ions. + Rpb4 and A14 do not show any sequence similarity. However, based on a limited sequence similarity between A43 and Rpb7 and other features, the Rpb7-Rpb4 pair was recently proposed to be the possible Pol II counterpart of the A43-A14 pair.71

Rpb9 and C11, in Pol I, II and III, respectively,59-61 has structural and functional homologies among the RNA polymerases and thus is considered a common core subunit (see discussion in Van Mullem et al62). The Pol I structure containing these 10 core subunit proteins has a nonspecific enzymatic activity,55 although it is not active in specific transcription initiation. For the latter, A43 is essential as described below. Among the ten core subunit proteins, A12.2 is apparently not required for catalytic function. The gene for A12.2 is only conditionally essential, while the genes for all other core subunit proteins are essential (Table 1). Deletion of RPA12 shows a ts phenotype, which can be suppressed by a high dosage of RPA190.60 Therefore, A12.2 apparently plays a role in enzyme assembly, but not a significant role in the Pol I function. In addition to the above core subunit proteins, there are polymerase specific subunits which are different between the three polymerases. For Pol I there are four distinct subunits, A49, A43, A34.5 and A14. Only the gene (RPA43) for A43 is essential and the genes for the other three are either nonessential (RPA34 and RPA14) or conditionally essential (RPA49). Within this group of subunits, A49 and A34.5 constitute a subset and A14 and A43 constitute another

134

The Nucleolus

subset. A49 and A34.5 appear to be loosely associated with the enzyme and can be lost together by chromatography or electrophoresis, yielding the A* form of Pol I which lacks both subunits.63 In addition, Pol I isolated from rpa34 deletion mutants does not contain the A49 subunit.64 However, structural studies of Pol I have failed to give any evidence of a direct physical contact between A34.5 and A49.65 Although RPA34 is a nonessential gene, its deletion mutant is synthetically lethal in combination with cells lacking topoisomerase I.64 The same rpa34 mutation combined with rpa49 or rpa12 or rpa49 rpa12 double mutations do not give any effects on the phenotypes of these mutants. However, the rpa34 mutation in combination with a deletion of RPA14 is lethal.64 Cells lacking A49 are viable but they grow slowly66 and are cold-sensitive.67 Regarding the second subset, A14 and A43, they interact directly in the absence of other subunits forming a stable heterodimer. Cells with a deletion of A14 are viable, but growth is slow, especially at higher temperatures. The Pol I purified from this mutant strain is inactive in an in vitro transcription assay and lacks subunits A14, A43 and ABC23.68 This Pol I is inactive, but the addition of ABC23 restored the (nonspecific) enzymatic activity,55 showing that, while RPA43 is an essential gene,69 protein A43 is not required for the enzymatic activity. As will be mentioned below, A43 is essential for the interaction of Pol I with transcription factor Rrn3p.70 This explains why A43 is not required for nonspecific enzymatic activities, but is essential for transcription of rRNA genes and for growth. In addition to its role in the interaction with Rrn3p, A43 plays a role in the stabilization of subunits A14 and ABC23 within the Pol I structure. Pol I isolated from an rpa43 deletion mutant (growing by transcribing the GAL7-35S rDNA fusion gene on a plasmid by Pol II) was found to lack A14 completely and ABC23 partially, thus resembling the Pol I isolated from the rpa14 deletion mutant mentioned above.71 Several Pol I subunits contain Zn++ (Table 1) and Zn++ binding structures like zinc fingers appear to play a role in interactions of subunits or domains, as judged from the atomic structure of Pol II.43 Five Pol I subunits are phosphorylated in vivo (Table 1). Recent work has demonstrated the importance of phosphorylation of Pol I in the interaction with Rrn3p,72 as mentioned below. However, positions of these many phosphorylation sites and the significance of phosphorylation at these sites are largely unknown. The homologies among subunits of the RNA polymerases and collective homologies with the prokaryotic polymerase as well as the presence of common subunits in the polymerases predict that the structures of the polymerases will be similar. Initial direct analysis of the 3-dimensional structure of Pol I was done by electron microscopy of two-dimensional crystals, which revealed an irregularly shaped molecule approximately 11nm x 11nm x 15 nm in size at approximate resolution of 3 nm.73 Previous structural studies on the bacterial holoenzyme and the yeast Pol II ∆4/7 afforded an opportunity for comparison and not surprisingly, similarity between Pol I and these enzyme structures was observed. Subsequent immunoelectron microscopy of Pol I core subunits allowed the analysis of the spatial organization of the subunits on the three-dimensional model of the enzyme.74 As mentioned above, with the recent elucidation of the atomic structure of Pol II ∆4/7,42-44 an alignment became possible between the Pol II structure and a structural model of Pol I determined by cryo-electron microscopy.65 Various portions of the Pol II ∆4/7 atomic structure are described by domains or domain-like regions whose terms reflect their function.43 Among four mobile modules defined, the core module comprises approximately one-half of the enzyme mass. This module contains regions of Rpb1 (A190 for Pol I) and Rpb2 (A135) and together they constitute the active center. Other subunits within the core module are Rpb3 (AC40), Rpb10 (ABC10β), Rpb11 (AC19) and Rpb12 (ABC10α) and these four subunits together anchor the two largest subunits and have been suggested to be involved in Pol II assembly. The three additional modules are along the sides of the DNA binding cleft and are positioned in front of the active center. The “jaw-lobe” module comprises the “upper jaw” [regions of Rpb1 (A190) and Rpb9 (A12.2)] and the “lobe” of Rpb2 (A135). The “shelf ” module consists of the lower jaw [part of Rpb5 (ABC27)], the assembly domain of Rpb5 (ABC27) and the foot and cleft regions of Rpb1 (A190). The “clamp”

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

135

describes the final module and consists of regions of Rpb1 (A190), mostly its N-terminal region, together with the C-terminal region of Rpb2 (A135). A difference map obtained from the comparison of Pol I and Pol II ∆4/7 structures was interpreted to reflect densities due to Pol I specific proteins that were absent in Pol II ∆4/7.65 Subunit antibodies were then used to analyze the position of Pol I specific subunits on the atomic model. The stalk containing subunits A43 and A14 protrudes from a similar position as the carboxy-terminal domain of Rpb1 on the Pol II ∆4/7 structure. At the top of the clamp was subunit A49 and at the entrance of the DNA binding cleft was A34.5.65 For the atomic structure of Pol II, a strongly asymmetric charge distribution exists; the surface charge is almost entirely negative except for a uniformly positively charged lining of the cleft, the active center, and other parts, where DNA template and RNA transcript are suggested to have contacts.43 In analogy to the results obtained for the Pol II structure, it is expected that essentially straight promoter DNA binds to the positively charged cleft between the A190 and A135 subunits of Pol I, and makes a right angle bend at around the +1 start site near the active center. Similarly, the clamp is expected to swing over the cleft, trapping the template and transcript during transcription. The position of the stalk containing A43/A14 also appears to be appropriate in interacting, through Rrn3p, with transcription factors (CF and UAF to be discussed below) bound to the upstream portion of the promoter during the initiation step. The functions of A34.5 and A49 are not clear. Because A34.5 is localized at the end of the cleft, and its C-terminal 44 residues contain 50% lysine, it was suggested that A34.5 may stabilize the interaction of template DNA with Pol I.65 A49 is positioned on the clamp head domain and it was speculated to play a role in elongation step, specifically in connection with processivity.65 Interestingly, A49 deletion is lethal in combination with topoisomerase III deletion, but not with topoisomerase I deletion. In contrast, A34.5 deletion is lethal with topoisomerase I deletion, as mentioned above, but not with topoisomerase III deletion.67 Thus both A49 and A34.5 may play a role, jointly with topoisomerases III and I, respectively, in removing the topological stress created during transcription, but their roles are apparently distinct.

Pol I Transcription Factors and the Mechanism for Transcription Initiation and Termination For mammalian Pol I transcription systems, the identification and characterization of transcription factors were originally carried out by fractionating cell extracts that were active in specific transcription of rDNA template (for reviews see Grummt;75 Zomerdijk and Tjian;76 see also Chapter 8 in this volume). To complement and extend the biochemical approaches, genetic approaches were initiated using the yeast S. cerevisiae.77 Simultaneously, specific transcription of rDNA using yeast cell extracts was also developed,78-80 allowing a combination of genetic and biochemical approaches to characterize transcription factors and their functions. Mutants (rrn mutants for rRNA synthesis defective) were isolated from a yeast strain carrying the GAL7-35S rDNA fusion plasmid as those that could grow on galactose but not on glucose media.77 This genetic screen resulted in the isolation of many specific mutants defective in rRNA synthesis and the identification of twelve different genes. Of the twelve genes, five were found to encode essential and conditionally essential Pol I specific subunits (A190, A135, A49, A43 and A12.2) that were not shared by Pol II or Pol III. The remaining seven genes (RRN3, RRN5, RRN6, RRN7, RRN9, RRN10, RRN11) encoded factor proteins that were specific for Pol I transcription of rDNA. These genes were subsequently cloned and characterized and strains were constructed such that the gene was replaced with an engineered gene encoding the protein carrying an epitope tag. The purification of transcription factors carrying the epitope tagged gene product resulted in the identification of three Pol I transcription factors, upstream activation factor (UAF), core factor (CF) and Rrn3p (Fig. 1B; for a review, see Nomura12). In addition to these transcription factors, requirement for TBP in Pol I transcription was demonstrated earlier using unfractionated cell extracts from a strain with a mutational defect in TBP81 and also in vivo.82

136

The Nucleolus

Core factor consists of three proteins, Rrn6p, Rrn7p and Rrn11p.83-85 For basal level transcription in vitro using rDNA templates containing just the core promoter, only CF, Rrn3p and Pol I are required. TBP is not required for basal transcription. This conclusion was obtained in a system with fractionated extracts27,86 as well as in a system with purified components.87 Further studies on the interaction between these transcription factors and the core promoter revealed that Rrn3p interacts directly with Pol I, forming an active form of Pol I, Pol I-Rrn3p complex87-89 and thus plays an essential function in transcription, almost certainly in the step of Pol I recruitment.88 The recruitment of Pol I to the promoter by Rrn3p is probably mediated through its interaction with the A43 subunit of Pol I and with the Rrn6p subunit of CF.70 Like CF, UAF is also a multiprotein complex consisting of Rrn5p, Rrn9p, Rrn10p, Uaf30p and histones H3 and H4. The genes for Rrn5p, Rrn9p and Rrn10p were identified in the original genetic screen27 while histones H3 and H490 as well as Uaf30p91 were discovered in the purified UAF complex. In vitro transcription experiments using rDNA template with the complete rDNA promoter demonstrated that, for high levels of transcription, UAF and TBP are required in addition to CF, Rrn3p and Pol I. In the absence of the upstream element, UAF is not able to stimulate transcription. UAF binds tightly to the upstream element of template in the absence of any other components, committing the template to transcription. Conversely, no such template commitment takes place with extracts missing the intact UAF.27 TBP interacts with both UAF and CF; specifically, interactions of TBP with Rrn9p subunit of UAF and with the Rrn6p subunit of CF were demonstrated by in vitro experiments as well as the yeast two-hybrid system (Fig. 1B; refs. 84-86). Additional genetic experiments demonstrated that the interaction of TBP with Rrn9p is essential for TBP function in Pol I transcription of rDNA in vivo.92 From these results, the ability of UAF to stimulate transcription was concluded to be mediated by the tight binding of UAF to the upstream element which results in the recruitment of CF with the help of TBP followed by recruitment of the Pol I-Rrn3p complex.27,86,87 However, the original experiments27,86 did not give a definitive answer to the question of whether CF and TBP join the UAF-rDNA template complex, forming a stable complex in the absence of Pol I and Rrn3p. Later experiments utilizing an immobilized rDNA template confirmed formation of the UAF-rDNA complex in the absence of other factors, and further demonstrated that a stable pre-initiation complex containing CF and TBP can be obtained only in the presence of both Pol I and Rrn3p.93 The same conclusion was obtained using ChIP experiments; a promoter occupancy of CF was demonstrated in growing wild type yeast cells, but not in mutant cells defective in Pol I which are growing by transcribing the GAL7-35S rDNA fusion gene by Pol II (K. Johzuka and M. Nomura, unpublished). In vivo footprinting experiments using mutants defective in UAF, CF or Pol I also gave results which are consistent with the in vitro results mentioned above.94 The genetic screen that identified many RRN genes obviously cannot detect transcription factors such as TBP that are also essential for Pol II and/or Pol III transcription, or factors that would give only a stimulatory effect. In addition, although specific Pol I transcription has been demonstrated using all purified components, a possible presence of very small amounts of unidentified components in preparations may be difficult to eliminate. In fact, in addition to the transcription factors mentioned above, several other proteins have been reported to have an effect on rDNA transcription. An example is TFIIH, a Pol II basal transcription factor which also has a role in nucleotide excision repair. In mouse, TFIIH was localized to the nucleolus and found associated with subpopulations of Pol I.95 TFIIH was observed to be essential for rDNA transcription in vivo in yeast strains which carried temperature sensitive mutations of TFIIH subunits. Another protein with an effect on rDNA transcription is Net1p (also called Cfi1p) which originally was characterized by its roles in the RENT complex (regulator of nucleolar silencing and telophase) that controls mitotic exit and nucleolar silencing.96-98 Net1p was found to interact with Pol I and stimulate rDNA transcription in vitro as well as in vivo.99

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

137

Recently, a DNA polymerase was identified and characterized as having an essential function in rRNA synthesis.100 The polymerase, Pol5p, binds close to or at the enhancer region of the rDNA repeats and thus is localized to the nucleolus. Severe inhibition of rRNA synthesis was observed in a temperature sensitive mutant, pol5-3, after shifting mutant cells to the restrictive temperature. Since the temperature sensitivity cannot be rescued by the Pol II transcription of the GAL7-35S rDNA fusion gene, this DNA polymerase must have an essential function in addition to (or instead of ) its possible role in Pol I transcription of rDNA. Further studies are required for defining exact roles, direct or indirect, of these additional proteins in rDNA transcription. The general structure of rDNA promoters consisting of the upstream element and the core promoter is very similar in various eukaryote systems, although nucleotide sequences of the promoter are not conserved. Therefore, one might expect that a similar mechanism is involved, for example, in mediating the stimulatory activity of the upstream element in these systems, even though amino acid sequences in corresponding factors might not have been well conserved. However, when the four yeast transcription factors were characterized, the expected correspondence between the yeast system and mammalian systems was not obvious. In mammalian systems, two protein factors had been characterized, UBF as the upstream element binding factor, and SL1 as the essential transcription factor.76 UBF was originally considered as a factor mediating the stimulatory activity of the upstream element, and thus functionally equivalent to yeast UAF. However, UBF is a homodimer of a protein with six HMG boxes, which is present abundantly in the nucleolus and binds to DNA without primary sequence specificity. Thus, UBF is very different from UAF, which is a multiprotein complex that binds specifically to the upstream element, as mentioned above. The relationship between SL1, which contains TBP as a subunit, and CF, which does not contain TBP, was also not clear. However, recent developments have started to reveal strong resemblance in the mechanism of initiation between the yeast system and the mammalian systems, satisfying gradually the initial expectation of conservation of the mechanism of Pol I transcription. First, a human homolog of yeast Rrn3p was identified based on the sequence homology101 and mouse factor TIF-IA has also proved to be a mouse homolog of Rrn3p.102 In fact, the interaction of Rrn3p with the A43 subunit of Pol I is conserved in Schizosaccharomyces pombe103 and human cells.104 Interaction of human Rrn3p with SL1 through its interaction with the TAFI110 and TAFI63 subunits105 appear to correspond to the interaction of Rrn3p with CF through its interaction with the Rrn6p subunit. In addition, a Rrn7p homolog of S. pombe has been identified recently and a clear sequence conservation between this protein, S. cerevisiae Rrn7p, human TAFI63 and mouse TAFI68 has been recognized.106 These observations strongly indicate that mammalian SL1 is in fact the functional counterpart of yeast CF. Finally, an HMG-box protein of S. cerevisiae, Hmo1p, was recently identified as a multi-copy suppressor of rpa49 deletion mutants and its role in Pol I transcription was demonstrated by synthetic lethality of hmo1 deletion with rpa49 deletion and its suppression by transcription of the GAL7-35S rDNA fusion gene by Pol II.67 A comparison of various properties of Hmo1p and UBF, both HMG-box proteins, strongly suggests that they are functional homologs, binding to rDNA and forming special rDNA structures suitable for Pol I-dependent transcription.67 Considering these recent developments, it is possible that a functional counterpart of yeast UAF exists in other eukaryotes including mammals, but has not been discovered. As emphasized above, although most or all the essential components of the yeast Pol I transcription machinery may have been characterized, there are almost certainly additional undiscovered factors which specifically modulate activity of the Pol I machinery. Furthermore, as will be discussed in a later section, Pol I transcription in vivo uses rDNA chromatin with additional chromatin proteins, and not free rDNA, as a template. Isolation of rDNA chromatin in a native state and studying its transcription in vitro may be a challenge in future research.

138

The Nucleolus

Regulation of rRNA Synthesis Like the well studied E. coli cells, regulation of synthesis of rRNA in yeast cells is central to regulation of overall synthesis of ribosomes, and is intimately connected to cells’ growth state. Thus, like E. coli, ribosome content increases with growth rate and rRNA synthesis is subject to growth-rate-dependent control. Various changes in nutritional conditions as well as environmental insults, such as alterations of carbon sources or amino acid deprivation, are sensed by yeast cells, leading to prompt adjustment of the rate of rDNA transcription by Pol I and the rate of overall accumulation of ribosomes. The subject has been extensively reviewed.5,6,9,107 Despite its importance, however, detailed molecular mechanisms are still mostly unknown. Nevertheless, most of basic components required for Pol I transcription of rDNA are now elucidated as described in the previous sections and rapid future progress is now anticipated. Here, we will discuss only certain topics related to regulation, where significant progress has been made recently.

Two Different Modes of Regulation: Regulation by Changing the Number of Active rRNA Genes and That by Changing the Efficiency of Transcription of Individual Genes Oscar Miller’s pioneering work to visualize actively functioning rRNA genes directly by electron microscopy was the first to show that only a fraction of tandemly repeated rRNA genes is transcriptionally active and other genes are kept in an inactive state.108 Based on such EM studies from many different eukaryotic organisms showing active rRNA genes highly loaded with Pol I intermingled with inactive genes led to the proposal that in most cases the number of open genes is rate-limiting for rRNA synthesis (e.g., see Reeder107). According to this model (“once on-maximally on” model) the open genes are transcribed at a full efficiency, and regulation would occur by altering the number of active genes. For the yeast S. cerevisiae, Sogo and coworkers109 carried out psoralen cross-linking experiments and demonstrated that there are two different states of rRNA genes, an open or transcriptionally active state and a closed or transcriptionally inactive state, and that only a fraction of ~150 copies of the rRNA genes are active in rapidly growing cells. They reported that this active gene fraction decreases when cells enter stationary phase, but the degree of the decrease depends on growth media: in complex media, from ~50% to ~30%, and in minimal synthetic media, from ~30% to 0.109 Thus, the data published for synthetic media appeared to support the regulation model based on active gene copy number alteration. However, the data published for rich media indicated that decreases in both active gene copy number and the efficiency of individual active gene copies are used to decrease overall rate of rRNA gene transcription upon entry into quiescent states. Recent experiments using both psoralen cross-linking and EM-Miller chromatin spread analyses were done in complex media and confirmed qualitatively the conclusion described above for yeast cells grown in complex media.110 No similar studies were done using synthetic media. However, these experiments did not answer the question of whether the number of active genes is limiting rRNA synthesis rate in yeast cells growing under optimum growth conditions. Regarding the question of regulation by decreasing the numbers of active rRNA gene copies, an intriguing recent discovery is that Rpd3p histone deacetylase is required to decrease the number of active rRNA genes upon entering into stationary phase.110 In rich media, ~50% of the rRNA gene remained open, without any further decrease, upon entering stationary phase in rpd3∆ mutants. It was found that, in this mutant in stationary phase, the number of Pol I transcribing each open gene was significantly less than that in the control RPD3 cells, thus apparently compensating for the failure to reduce the number of active rRNA genes. One particular point the authors did not discuss in this work is that Rpd3p histone deacetylase is required to inactivate the open (active) RNA gene present in ~50% of the copies, but is apparently not required to maintain the inactive genes present in the other ~50% of the copies, implying that Rpd3p might be required for the process of the conversion of active to inactive states in response to a signal(s) created by approaching to the quiescent state, but not for the

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

139

maintenance of the inactive state itself. It is not known what really determines the number of active genes in rapidly growing yeast cells. Similarly, the significance of requirement for Rpd3 histone deacetylase in conversion of active to inactive states remains unexplained. As mentioned above, the model has been entertained which assumes that, in rapidly growing cells, the rate of rRNA synthesis is determined by the number of available active genes. According to this model, rapidly growing yeast cells are using each of ~50% (70-80 copies) of all the rRNA genes (~150 copies) at a maximum efficiency, and the model predicts that yeast cells require 70 to 80 rRNA genes to attain the maximum rRNA synthesis rate. Contrary to this prediction, it has recently been shown that decreasing the total numbers of rRNA genes to ~40 copies did not decrease the rate of rRNA synthesis or growth rate.23,111 EM Miller-spread analysis demonstrated that all or almost all of the ~40 copies are transcribed and approximately twice as many polymerases are engaged in transcription per gene in the ~40 copy strain compared to the control ~150 copy strains, making the total number of transcriptionally engaged polymerases approximately the same in the two strains.111 Thus, in exponentially growing yeast cells, rRNA synthesis is determined by the ability of cells to load and transcribe rDNA with Pol I and not by the number of genes open for transcription. In the wild type cells, about half of ~150 copies is available as active rRNA genes, and cells use all of these active genes equally with an efficiency that is only about half of the efficiency shown in the ~40 copy strain. These results support a model in which overall initiation rate, presumably reflecting the concentration of initiation competent Pol I, i.e., a Pol I-Rrn3p complex to be discussed below, determines the rRNA synthesis rate. The question of what determines the number of active genes (~70 to 80) in normal yeast cells carrying ~150 copies, is an unsolved question. As discussed below, one possibility is that rDNA chromatin structures are different between active and inactive rDNA repeats, and that the total amount of specific component(s), e.g., UAF, that is required for making rRNA genes active, is limiting, i.e., is present in an amount that is sufficient only for 70 to 80 rRNA genes in haploid nucleus. This question and the related question on the molecular mechanism for conversion between the active and inactive states of rRNA genes are the subjects to be studied in the future.

Rrn3p-Pol I Complex As the Active Form of Pol I Which Is Limited in Stationary Phase Earlier experiments using mammalian in vitro transcription systems showed that extracts from cells in quiescent states are inactive in Pol I transcription and that inactive extracts can be activated by the Pol I fraction from cells in growing cells. It was inferred that Pol I from growing cells is associated with a factor, called TIF-IA by Grummt and coworkers, which becomes inactive (or degraded) in quiescent states.112 However, TIF-IA was not completely purified and its molecular nature was not well characterized. In the yeast system, RRN3 was first identified as a gene essential for rDNA transcription and its product, Rrn3p, was shown to directly interact with Pol I, enabling the complex to be recruited to the promoter.87,88 Subsequently, it was found that in extracts from exponentially growing cells a small fraction of Pol I exists as an Rrn3p-Pol I complex, but such a complex does not exist in extracts from cells in stationary phase, explaining the difference between transcriptionally active and inactive extracts from growing and stationary phase cells, respectively.89 Because of the discovery of homologs in mammals,101 the question of the suspected identity of TIF-IA with the Rrn3p homolog was tested and confirmed.102 Thus, the mechanism of Pol I repression in quiescent states appears to be conserved from yeast to mammals and to take place at the step of the formation of an active form of Pol I, namely, an Rrn3p-Pol I complex. In fact, the interaction of Rrn3p (or it’s mammalian homolog) with Pol I was shown to take place through its interaction with subunit A43 (or its mammalian homolog) of Pol I both in yeast70 and in mammals.104 Furthermore, the mechanism of recruitment of the complex appears to involve an interaction of Rrn3p (or its mammalian homolog) with CF, specifically its subunit Rrn6p, in yeast70 or with transcription factor SL1, specifically its subunit TAFI110 and TAFI63, in mammals.105 Such conservation

140

The Nucleolus

underlies the importance of Rrn3p-Pol I interaction. However, there is an apparent discrepancy between the yeast system and mammalian system regarding the mechanism responsible for down-regulation of this interaction. It is known that both Pol I and Rrn3p are phosphorylated both in yeast and mammals. Studies carried out in the yeast system showed that dephosphorylation of Pol I by phosphatase inhibited formation of Rrn3p-Pol I complex in vitro, whereas both phosphorylated and nonphosphorylate forms of Rrn3p are able to associate with Pol I and to initiate transcription.72 In contrast, studies using extracts from cycloheximide treated mammalian cells suggested that a phosphorylated form of Rrn3p, and not unphosporylated Rrn3p, is the one which is capable of association with Pol I and thus initiating transcription.104 Regardless of this discrepancy, the current model for growth-rate control of Pol I in response to alteration of nutritional states, specifically entering or leaving quiescent states, involves changes in the amount of an initiation competent Rrn3p-Pol I complex. As described in the previous section, rRNA synthesis in exponentially growing yeast cells is limited by the ability of cells to load and transcribe rDNA with Pol I rather than by the number of an active form of the genes. It now appears certain that the amount of an initiation competent Rrn3p-Pol I complex is limiting rRNA synthesis in exponentially growing yeast cells.

The Rapamycin-Sensitive TOR-Signaling Pathway and Regulation of rDNA Transcription Rapamycin, originally identified as an anti-fungal and immunosuppressive compound, inhibits yeast cell growth and the inhibited cells display the state which resembles the quiescent state seen in stationary phase cells in many respects.113,114 Studies on the mechanism of rapamycin action in yeast led to the discovery of the TOR (target of rapamycin) signaling pathways as a central system conserved in all eukaryotes, which regulate many of the cellular activities coordinately in response to nutritional states (reviewed by Schmelzle and Hall115). TOR was originally identified by mutations in yeast that confer resistance to rapamycin,113 and is a phosphatidylinositol-kinase-related protein kinase. In yeast, there are two structurally very similar TORs, encoded by genes TOR1 and TOR2. Rapamycin first binds the FK506 binding protein or FKBP, and the rapamycin-FKBP complex inhibits TOR functions. Among functions of the yeast TOR, one is regulation of translation in general, and the other one is regulation of synthesis of ribosomes, which includes transcription of r-protein genes and transcriptional activity of Pol I and Pol III. Regulation of ribosome synthesis by TOR in yeast was shown by rapid and severe inhibition of rRNA synthesis and r-protein mRNA synthesis caused by treatment of yeast cells with rapamycin.116,117 The mechanism of this inhibition is not known, but a likely possibility may be through activation of a protein phosphatase(s). The yeast TOR signaling pathways are shown to involve a phosphatase switch composed of the type 2A-related phosphatase Sit4p (or the phosphatase catalytic subunit encoded redundantly by PPH21 and PPH22), Tap42p and Tip41p (ref. 118 and other references therein). Under favorable nutritional conditions, TOR stimulates formation of complexes of Tap42p with Sit4p (or the other type 2A phosphatase), thereby inactivating the phosphatases. This stimulation may involve a direct phosphorylation of Tap42p by TOR protein kinases.119 Upon rapamycin treatment or nutritional starvation, the phosphatases are dissociated from its inhibitor Tap42p, and act on various phosphorylated target proteins.115,118 As described above, we are now fairly certain that growth-rate-dependent control of rRNA synthesis takes place by altering the concentration of the Rrn3p-Pol I complex, and this complex formation is apparently dependent on (not well characterized) phosphorylation of Pol I and/or Rrn3p. Therefore, one may speculate that the activation of protein phosphatase by rapamycin, which must represent an intermediate step in signaling nutritional starvation in stationary phase, may lead to the eventual inhibition of Pol I by acting on Pol I and/or Rrn3p, perhaps by altering their phosphorylation states. However, there are no published experiments relevant to testing the validity of this model. In addition, the question of how TOR proteins in yeast cells sense nutritional states in order to send regulatory signals to various targets is currently unknown.

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

141

Pol I Repression Caused by Amino Acid Depletion In bacteria like E. coli, amino acid deprivation causes a severe inhibition of synthesis of both rRNA and tRNA, a response known as stringent control, and its mechanism has been studied extensively. In yeast, amino acid starvation also causes inhibition of rRNA synthesis (and r-protein mRNA synthesis), but not tRNA synthesis. In addition, like cycloheximide treatment, inhibition of protein synthesis generally causes repression of Pol I transcription (refs. 120, 121; regarding its possible biological significance, see Nomura9). Although the mechanism(s) involved in these cases is not well studied in yeast, the target of repression may likely be the step responsible for formation of the Rrn3p-Pol I complex as the above-mentioned recent study on cycloheximide-treated mammalian cells suggests.104 Amino acid deprivation causes upregulation of genes involved in amino acid biosynthesis. The molecular mechanism of this regulation, general amino acid control, has been well elucidated. In this case, amino acid deficiency is sensed through an increase of uncharged tRNA by eIF2 protein kinase encoded by GCN2, leading to the eventual activation of amino acid biosynthetic genes by an increased synthesis of transcription factor GCN4 (reviewed by Hinnebush122). Repression of Pol I transcription of rDNA caused by amino acid depletion also appears to take place through an increase of uncharged tRNA.123 However, it is apparently not through Gcn2 protein kinase, because gcn2 mutations do not interfere with the down-regulation of Pol I caused by histidine starvation (L. Vu and M. Nomura, unpublished experiments), as in the case of down-regulation of r-protein gene transcription.124 How amino acid deficiency is initially sensed and then leads to the eventual Pol I repression is unknown. In mammalian cells, responses to amino acid abundance/depletion is carried out through the TOR signaling pathway.125,126 The question of whether the TOR signaling pathway is involved in the Pol I repression caused by amino acid starvation in yeast has not been critically studied.

Pol I Repression Caused by Defects in Secretion Since ribosome synthesis is well coordinated with cellular growth, it may be expected that alterations in cellular functions essential for growth either by mutations or by environmental challenges affect synthesis of rRNA and/or r-proteins. For example, Warner and coworkers discovered that transcription of both rRNA and r-protein genes is repressed under conditions of inhibition of protein secretion pathways127 and that this down-regulation requires protein kinase C signaling pathway.128 Deletion of gene PKC1, which encodes protein kinase C, abolishes the repression of rDNA transcription by Pol I caused by secretion defects. It has been suggested that stress in the plasma membrane is sensed by protein kinase C, which initiates signaling that leads to down regulation of rRNA and r-protein synthesis.128 Since PKC1 deletion mutants are still sensitive to rapamycin,129 this signaling pathway leading to Pol I repression must be different from the TOR signaling pathway. Although several additional mutations that abolish this secretion-defects-induced repression of Pol I have been identified and involvement of a ribosome assembly machinery in this signaling pathway has been suggested,130,131 signals downstream of protein kinase C and how the signals eventually affect the Pol I machinery are currently unknown.

Cell Cycle and rRNA Synthesis In higher eukaryotic cells, bulk transcription activities including rRNA synthesis stop as cells enter mitosis and this is accompanied by the disappearance of the nucleolus, disassembly of nuclear envelope and chromosome condensation. In contrast, in yeast cells, which carry out mitosis without nuclear envelope disassembly, bulk RNA transcription apparently continues without disruption and nucleolar structures are also maintained during mitosis; a single nucleolus is kept after replication of rDNA in G2 and M, and it simply separates, presumably as chromosomal XII carrying rDNA separates, into the mother and daughter cells;132 see also Shou et al.96 Changes in macromolecular synthesis rates during cell cycle in yeast were extensively studied in the past (for a review of earlier studies, see Elliott and McLaughlin133). One particular

142

The Nucleolus

point debated was whether the rate of rRNA (and other RNA such as tRNA) synthesis increases two-fold as the genes duplicate. After some extensive analyses, it was concluded that the bulk of cellular RNA, rRNA and tRNA, as well as total cellular protein increases exponentially during cell cycle, and that both rRNA synthesis rates and r-protein synthesis rates per unit amount of cellular protein (or RNA) remain constant throughout all cell cycle stages;134,135 that is, rRNA synthesis rate per cell increases exponentially like bulk cellular protein and cellular RNA in exponentially growing yeast cells. As we have discussed above, we expect that rRNA synthesis rate should not be affected by a sudden change in the total number or rRNA genes during S phase, and that the control is probably exerted on the cell’s ability of loading of Pol I on rRNA genes. If this is in fact determined by the amount of the Rrn3p-Pol I complex, the amount of such rate-determining complex increases exponentially like bulk cellular protein. However, an analysis of the numbers of active relative to inactive rRNA genes and of Pol I density per active gene before and after duplication of rDNA in S phase, for example, using the EM-Miller chromatin spreading technique, have not been carried out to establish the above inference. In the earlier studies mentioned above, the absence of repression of Pol I during mitosis in yeast cells was conducted by pulse labeling of exponentially growing cells followed by fractionation of cells into groups of cells at different cell cycle stages. Direct analysis of cells arrested in mitosis with the microtubule-depolymerizing reagent nocodazole also confirmed this conclusion (K. Johzuka, P. Tongaonkar, and M. Nomura, unpublished experiments). It is known that rDNA repeats in mitotically arrested cells are condensed, but the degree of increase in the compaction ratio (defined as the predicted length of DNA in B form divided by the observed length of the chromosome segment occupied by the DNA) relative to the rDNA chromatin in interphase is small, being estimated to be ~2-fold or less.136 Thus, Pol I machinery is apparently able to transcribe rDNA template in such a condensed state. It should also be noted that increase in compaction ratios for mammalian DNA in mitotic chromosomes relative to interphase chromatins is 5 to 10-fold137 and is significantly higher than those seen for yeast nucleus. However, repression of Pol I during mitosis in mammalian cells is not because of a more condensed state of the chromosomes. Biochemical analysis indicated that repression of Pol I transcription may be achieved by inactivation of transcription factor SL1 by cdc2/cyclin-B-directed phosphorylation.138 Treatment of colchicine-arrested mitotic HeLa cells with an inhibitor of the cdc2/cyclin B led to derepression of Pol I transcription without exiting from mitosis,139 showing that Pol I transcription can take place on rDNA within condensed mitotic chromosomes in mammalian systems too. The biological significance of the absence of Pol I repression during mitosis in yeasts relative to its presence in higher eukaryotes is unknown, but might be related to the different modes to achieve mitosis, such as the breakdown of nuclear envelope in higher eukaryotes, but not in yeasts.

Silencing of Pol II Transcription in rDNA Silencing of genes brought near or within heterochromatic regions have been known for a long time in higher eukaryotes. Although there are no cytologically recognizable heterochromatin regions in yeast cells, similar silencing phenomena were discovered in three different kinds of chromosomal loci: the silent mating-type loci, the telomeric regions and the rDNA repeats (for reviews, see refs. 140-142). Silencing of Pol II genes in rDNA repeat was first discovered by analyses of retrotransposon TY1 integrated into rDNA repeats and of reporter genes artificially integrated into rDNA repeats.143,144 Independently, UAF-dependent silencing of transcription of chromosomal rRNA genes by Pol II was discovered in the course of studies on mutants defective in Pol I transcription factor UAF.145,146 We first discuss on this latter subject followed by discussion on the former in relation to active versus inactive rDNA repeats.

Silencing of Pol II Transcription of rDNA by UAF As discussed earlier in this article, three genes, RRN5, RRN9 and RRN10, encoding subunits of UAF were identified as “essential” genes uniquely required for Pol I transcription of

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

143

rRNA genes. Mutants with defects in these genes were isolated as those able to grow only in galactose, and not in glucose media by transcription of the GAL7-35S rDNA fusion gene by Pol II. An attempt to see whether low level “basal transcription” observed in vitro27,86,87 is able to support cell growth led to the finding that rrn9 (or rrn5 or rrn10) deletion strains are in fact able to grow, though extremely slowly at the beginning, but then give rise to faster-growing variants. Appearance of such variants was also observed by spreading galactose-dependent mutants, which carry the plasmid-encoded GAL7-35S rDNA fusion gene, on glucose plates. Variants able to form colonies were observed at frequencies of 10–4 to 10–5. Such variants were found to use Pol II to transcribe chromosomal rRNA genes using some cryptic promoters with several different transcription start sites (ranging from –9 to –95 with respect to Pol I start site as +1); deletion of genes for essential Pol I subunits or for essential factors such as CF did not show any negative effect on rRNA transcription or cell growth.145 It should be noted that in these UAF-defective mutants there is no Pol I transcription in vivo that would correspond to the basal transcription observed in vitro; only Pol II transcription of rRNA genes takes place in vivo. The switch to growth using the Pol II system (called the “PSW” state for polymerase switched) was shown to consist of two steps: a mutational alteration in UAF and an expansion of chromosomal rDNA repeats to levels (~400) several fold higher than the normal wild type level. The first step, a single mutation in a UAF subunit gene, RRN5, RRN9 or RRN10, is sufficient to eliminate all Pol I transcription and to allow Pol II transcription of rRNA genes from cryptic Pol II promoter(s). However, without the second step, repeat expansion, cells are unable to form colonies (called the “N-PSW” state for no growth, but polymerase switched). As mentioned above, these N-PSW cells can grow in galactose media by transcribing the plasmid-encoded GAL7-35S rDNA gene by Pol II and switches between the N-PSW to PSW state were studied by following the change in the galactose dependent phenotype. It was found that in the absence of selection, the two alternative states, N-PSW and PSW, are semi-stable and can be heritable not only in mitosis but also in meiosis. Association of these two phenotypes with two different kinds of chromosome XII, one carrying rDNA with a reduced repeat number (reduced to ~80, as in rpa135 mutants mentioned above) and the other carrying an expanded repeat number (~400) was demonstrated.145,146 Significantly, cells in the PSW state were found to have a round nucleolus localized away from the nuclear periphery, which is very different from the crescent nucleolus at the nuclear periphery seen in the wild type cells (Oakes et al;146 see below). UAF has a subunit encoded by the nonessential gene UAF30.91 Deletion of this gene decreases the cellular growth rate. Interestingly, the deletion mutants were found to use both Pol I and Pol II for rRNA transcription, indicating that the function of UAF to silence Pol II transcription is impaired even though rDNA transcription by Pol I is still occurring. A UAF complex isolated from the uaf30 deletion mutant was found to retain the in vitro Pol I activator function to a large extent. Therefore, Uaf30p appears to play a significant role in silencing of Pol II transcription of rRNA genes, but only a minor role, if any, in the activator function of UAF. The role of UAF in silencing Pol II transcription of chromosomal rRNA genes is unique. Mutations in subunits of Pol I, e.g., the A135 subunit, or other Pol I transcription factors such as CF and Rrn3p do not independently lead to Pol II transcription of rDNA. Significantly, mutations in SIR2, which disrupts silencing of reporter Pol II genes in rDNA, was also found not to disrupt silencing of Pol II transcription of chromosomal rRNA genes. However, sir2 mutations were found to increase the frequency of switch to the PSW state in the presence of a mutation in UAF by a factor of ~100, presumably stimulating the second step, i.e., rDNA repeat expansion through stimulation of unequal sister-chromatid recombination within rDNA repeats.146 Although mutations in UAF are unique in abolishing silencing, weak inhibition of silencing was also observed in wild type strains under some experimental conditions. For example, shifting yeast cell cultures grown at 25°C to 37°C allowed a weak, but definite, Pol II transcription of chromosomal rRNA genes displaying initiation sites identical to the sites seen for Pol II transcription in PSW strains (K. Eliason and M. Nomura, unpublished experiments).

144

The Nucleolus

There is a report related to the above-described silencing of Pol II transcription of chromosomal rRNA genes. Conrad-Webb and Butow147 reported Pol II transcription of rDNA in respiratory-deficient mitochondrial [rho°] strains of S. cerevisiae. In this case, however, transcription was reported to use mostly extra-chromosomal rDNA circles as template and to initiate at the same site (+1) as Pol I, and the relationship between Pol II transcription in [rho°] strains and that in UAF mutants is not clear.

Silencing of Reporter Pol II Genes in rDNA After the initial discovery of silencing of Pol II reporter genes in rDNA, several genes required for efficient silencing were identified. Among the three SIR genes, SIR2, SIR3, and SIR4, which are required for silencing at both the mating type loci and telomere regions, only SIR2 is required for silencing at rDNA. 143,144 Subsequently, SIR2 was identified as NAD-dependent protein deacetylase,148,149 explaining the previous observations that histones in the silenced chromatins at mating type loci and telomere regions are hypoacetylated.150,151 The importance of chromatin structures in silencing of reporter genes was also supported by negative effects on silencing caused by mutations affecting rDNA chromatins, specifically, top1 mutations, deletions of HTA1 and HTB1 which reduce histone H2A-H2B levels, and cac1 mutations which affect nucleosome assembly during DNA replication.144,152 In addition, the requirement for RAD6/UBC2 and SET1 in rDNA silencing144,153 (and in silencing at mating type loci and telomere regions) combined with the identification of SET1 encoded protein as histone H3 (Lys4) methylase154 also supported the importance of special features of rDNA chromatin for silencing of reporter genes. It was also shown that ubiquitination of histone H2B at Lys123 by the RAD6/UBC2-encoded protein is essential for the histone H3 methylation by Set1p.155 Interestingly, silencing mediated by SET1 was reported to be exerted by a Sir2p-independent mechanism.153 Thus, rDNA “heterochromatin” structures responsible for reporter Pol II silencing appear to be complex, partly sharing structural components with telomeres and mating type loci and partly containing components unique to rDNA. One of the unique components identified is Net1p (also called Cfi1p), which forms a complex called RENT with Sir2p, Cdc14p and other not-well-characterized proteins.96-98 Net1p is required for silencing by tethering Sir2p to rDNA; net1 mutations abolish the association of Sir2p with rDNA, while sir2 mutations do not affect the association of Net1p with rDNA, as analyzed by ChIP.97 In contrast to the Sir2p containing complex at rDNA, Sir2p exerts its silencing function at mating type loci and telomeres by forming a complex that includes Sir3p and Sir4p.151,156-158 Studies of silencing at mating type loci and telomeres revealed that the regions showing general silencing are covered by silencing proteins such as Sir2p, Sir3p and Sir4p which interact with nucleosomes, and that these proteins are recruited to the regions first by interacting with specific DNA binding proteins that bind to chromosome ends or to specific regulatory DNA sequences called silencers, followed by spreading of silencing heterochromatic structures along chromosomes.142,151 In the case of rDNA silencing proteins, both Net1p and Sir2p are bound to most, if not all, regions within the 9.1 kb unit,97 although the NTS1 region appears to be more enriched with Sir2p.41,159 However, the question of how these silencing proteins are recruited to rDNA is unknown. One possibility is that they are recruited to rDNA through the interaction of Net1p with Pol I. As mentioned earlier, Net1p interacts with Pol I and stimulates Pol I transcription both in vitro and in vivo.99 However, this possibility has not been rigorously tested experimentally. The generally accepted model for gene silencing invokes a higher order of repressive chromatin structures containing silencing proteins and variously modified histones which prevent general access of transcriptional machinery, recombination machinery or some artificially introduced structural probes such as dam methyltransferase. Silencing at rDNA is unique because it is the site of active transcription by Pol I. Two possibilities were considered earlier.142,143,144,159 First, rDNA chromatin prevents general access of the Pol II machinery and other macromolecules, as mentioned above, but somehow allows an access of Pol I transcrip-

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

145

tion machinery. The second model proposes that, since only about half of ~150 rDNA repeats in normal yeast cells are transcriptionally active,109,110 the silencing of reporter genes takes place only in the subset of repeats that are not actively transcribed by Pol I. In this model, it was assumed that some special rDNA chromatin structures act in a repressive way on both Pol I and Pol II machineries (called “concerted silencing model”), and that each of the rDNA repeats alternates rapidly between inactive (or closed) and active (or open) states, as suggested from psoralen cross-linking experiments.160 The relatively weak silencing of reporter genes in rDNA was explained based on the assumption that the activity of reporter genes represents an average of the silenced one and the nonsilenced one.142,143 This concerted silencing model was originally supported by psoralen cross-linking experiments; sir2 mutations which abolish silencing, increased the fraction of rDNA repeats that is in the active form, i.e., increased transcription of reporter Pol II genes and apparently Pol I activity.143 However, more comprehensive later analysis failed to reproduce the earlier results, concluding that sir2 mutations do not affect the number of active rRNA genes nor overall rRNA synthesis rate.110 The two models mentioned above were suggested based on the original work on silencing of reporter genes artificially integrated in rDNA repeats and do not adequately explain the silencing of Pol II transcription of chromosomal rRNA genes by UAF described earlier. In this case, UAF, which is an essential Pol I transcription factor, functions to silence Pol II transcription. In normal cells, Pol I is on and Pol II is off. In UAF defective mutants, Pol I is off and Pol II is on. That is, Pol I and Pol II transcription of native rRNA genes are in a reciprocal relationship. In addition, Net1p, which is required for silencing of reporter genes, stimulates Pol I transcription both in vivo and in vitro,99 as mentioned earlier. In fact, requirement for the intact Pol I machinery in silencing of reporter genes integrated in rDNA repeats was recently demonstrated.161,162 It was also demonstrated that silencing of reporter genes integrated into the coding region is much stronger in a mutant with a reduced copy number (~40 copies or ~25 copies), all of which are transcriptionally active111 as mentioned earlier, than in control strains, which have about half of the rDNA repeats as active and the other half as inactive.161 These observations do not support concerted silencing model or independent silencing model. Instead, they support the alternative model (called “reciprocal silencing model”), which proposes that rDNA chromatin structures that favor Pol I transcription are repressive to Pol II transcription, and that reporter Pol II genes integrated into Pol-I-transcription-inactive rDNA copies escape silencing.162 It should be noted that silencing of Pol II transcription of chromosomal rRNA genes requires the intact UAF, but does not apparently require the intact Pol I or the intact CF. Nor does this silencing require other silencing proteins such as Sir2p. Perhaps, UAF, which binds specifically to the upstream element of the Pol I promoter, may nucleate a special chromatin structure in the promoter region that prevents the access of Pol II to the cryptic Pol II promoters for chromosomal rRNA gene transcription, but fail to prevent the access of Pol II to promoters of reporter genes inserted in other regions. The local special chromatin structure at the promoter might then initiate spreading of a structure which is repressive to Pol II transcription at other rDNA regions. According to this hypothesis, proteins such as Sir2p, that are required for silencing of reporter Pol II genes at other rDNA sites, but not for silencing of Pol II transcription of chromosomal rRNA genes, are required for the spreading to cover the entire rDNA. Recent work on silencing of the non-rDNA regions adjacent to rDNA indicated that the direction of spreading is controlled by the direction of Pol I transcription, giving a strong support to the above hypothesis.161

Nucleolar Structures and rRNA Transcription The nucleolus does not have a membrane, but occupies a region which is functionally separated from the rest of nuclear regions. A single rDNA repeat placed in ectopic chromosomal sites is usually not transcribed to any significant extent unless the normal nucleolar structure is disrupted by mutations (ref. 33; K. Johzuka, K. Eliason and M. Nomura, unpublished experiments). The Pol I transcription machinery (and other proteins related to rRNA

146

The Nucleolus

processing/ribosome assembly) is sequestered in the nucleolus and is not available to such ectopic genes. The nucleolus in S. cerevisiae is a crescent-shaped structure that makes extensive contact with the nuclear envelope. Analysis of yeast mutants (rdn∆∆ mutants), in which chromosomal rDNA repeats are completely deleted, has provided some information on factors responsible for this unique nucleolar morphology and localization. The growth of such rdn∆∆ mutants is supported by a multi-copy plasmid carrying a single rDNA repeat on a Pol I helper plasmid or a Pol II helper plasmid carrying the GAL7-35S rDNA as explained earlier.30,163 It was found that rdn∆∆ strains carrying a Pol I helper plasmid contains a fragmented nucleolus (“mininucleoli”) distributed throughout the nucleus, primarily localized at the nuclear periphery. In contrast, rdn∆∆ strains carrying a Pol II helper plasmid were found to have a single (and occasionally two, rarely more) rounded nucleolus that often lacked extensive contact with the nucleolar envelope. The nucleolar morphology and localization in this case is similar to those observed in PSW strains mentioned earlier. From these observations, it appears that the nature of transcriptional machinery is important in determining the localization of the nucleolus. Pol I transcription takes place mostly at the nuclear periphery, while Pol II transcription appears to take place at sites away from the nuclear envelope. However, in the absence of template rRNA genes, the Pol I machinery fails to localize at the nuclear periphery. In rdn∆∆ strains carrying a Pol II helper plasmid, Pol I and Pol I transcription factors diffuse throughout the nucleus, while proteins such as Nop1p involved in rRNA processing/ribosome assembly are localized to the round, interior nucleolus.163 Thus, there are at least two different kinds of nucleolar proteins. Protein components of the Pol I machinery, together with rDNA template, may play a primary role in determining nucleolar localization, whereas others, such as those involved in rRNA processing and ribosome assembly, apparently do not. The difference in the nucleolar localization between the two instances, one which uses Pol I for rRNA transcription and the other which uses Pol II for rRNA transcription, as described above, may suggest a possible explanation for silencing of reporter Pol II genes integrated into chromosomal rDNA repeats. According to the reciprocal silencing model, which is consistent with most of the observations as mentioned in the previous section, Pol II genes in active (open) rDNA copies are silenced and those in inactive (closed) rDNA copies are not silenced. It is conceivable that the Pol II machinery is not freely diffusible, being unable to penetrate into the inside of the nucleolus. Perhaps, ~150 rDNA repeats in normal strains have a flexibility in their localization. Those active copies transcribed by Pol I may be “inside” the nucleolus at the nuclear periphery, and hence reporter genes integrated in those active repeats may not be accessible to the Pol II machinery, i.e., are silenced. The inactive rDNA repeats are perhaps localized at more “outside” of the nucleolus without a constrain imposed by Pol I transcription, and reporter genes in those copies may be more accessible to the Pol II machinery. Such a model and the model discussed in the previous section, which assumes special rDNA chromatin structures as the basis for a reciprocal relationship in gene expression between Pol I and Pol II, are not mutually exclusive.

Concluding Remarks The unique subnuclear localization of the nucleolus discussed in the preceding section must be related to evolution of the Pol I machinery (for further discussion, see Nomura164). We can imagine that ribosome production and regulation of rRNA gene transcription are important and unique, and hence, the Pol I machinery has evolved presumably for the efficiency and the unique regulation. Having the Pol I machinery in a separate compartment may have been advantageous, especially because of complex processes required for making ribosomes and of high demand for ribosome production under rapid growth conditions. To keep many copies of rRNA genes in a separate compartment, tandemly repeated structures may have been convenient. In addition, repeated structures could have been an advantage for sequence homogenization and adaptation, in terms of evolution as well as short-term adaptation as exemplified by

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

147

repeat expansion which allows growth using the Pol II system in response to mutational inactivation of UAF. The presence of a sub-nuclear compartment may have also become useful in evolving some other nucleolar functions, such as control of cell cycle by sequestering certain crucial regulatory molecules like Cdc14p. The chromosomal rDNA repeats and the Pol I machinery constitute the central components in organizing the nucleolus, as described above. With the powerful genetics, and availability of various genetically engineered yeast strains, the yeast system promises further contribution to the general questions related to not only the mechanism of Pol I transcription and its regulation, but also other important functions of the nucleolus.

Acknowledgements The work carried out in the Nomura laboratory has been supported by a grant from NIH GM-35949. We thank S. VanAmburg for help in preparation of the manuscript.

References 1. Brown DD, Gurdon JB. Absence of ribosomal RNA synthesis in the anucleolate mutant of Xenopus laevis. Proc Natl Acad Sci USA 1964; 51:139-146. 2. Wallace H, Birnsteil ML. Ribosomal cistrons and the nucleolar organizer. Biochim Biophys Acta 1996; 114:296-310. 3. Ritossa RM, Spiegelman S. Localization of DNA complementary to ribosomal RNA in the nucleolus organizer region of Drosophila melanogaster. Proc Natl Acad Sci USA 1965; 53:737-745. 4. Warner JR. The yeast ribosome: Structure, function and synthesis. In: Strathern JN, ed. Molecular Biology of the Yeast Saccharomyces: Metabolism and Gene Expression. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1982. 5. Warner JR. Synthesis of ribosomes in Saccharomyces cerevisiae. Microbiol Rev 1989; 53:256-271. 6. Warner JR. The economics of ribosome biosynthesis in yeast. Trends Biochem Sci 1999; 24:437-440. 7. Woolford Jr JL, Warner JR. The ribosome and its synthesis. In: Broach JR, Pringle JR, Jones EW, eds. The Molecular Biology of the Yeast Saccharomyces: Genome Dynamics, Protein Synthesis, and Energetics. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1991:587-626. 8. Planta RJ. Regulation of ribosome synthesis in yeast. Yeast 1997; 13:1505-1518. 9. Nomura M. Regulation of ribosome biosynthesis in Escherichia coli and Saccharomyces cerevisiae: Diversity and common principles. J Bacteriol 1999; 181:6857-6864. 10. Thuriaux P, Sentenac A. Yeast nuclear RNA polymerases. In: Jones EW, ed. The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression. Vol. II. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1992:1-48. 11. Carles C, Riva M. Yeast RNA polymerase I subunits and genes. In: Paule MR, ed. Transcription of Ribosomal RNA Genes by Eukaryotic RNA Polymerase I. Austin: RG Landes Co., 1998:9-38. 12. Nomura M. Transcription factors used by Saccharomyces cerevisiae RNA polymerase I and the Mechanism of initiation. In: Paule MR, ed. Transcription of Ribosomal RNA Genes by Eukaryotic RNA Polymerase I. Austin: RG Landes Co., 1998:156-172. 13. Drouin G, de Sá MM. The concerted evolution of 5S ribosomal genes linked to the repeat units of other multigene families. Mol Biol Evol 1995; 12:481-493. 14. Geiduschek EP, Kassavetis GA. The RNA polymerase III transcription apparatus. J Mol Biol 2001; 310:1-26. 15. Brewer BJ, Fangman WL. A replication fork barrier at the 3' end of yeast ribosomal RNA genes. Cell 1988; 55:637-643. 16. Linskens MHK, Huberman JA. Organization of replication of ribosomal DNA in Saccharomyces cerevisiae. Mol Cell Biol 1988; 8:4927-4935. 17. Pasero P, Bensimon A, Schwob E. Single-molecule analysis reveals clustering and epigenetic regulation of replication origins at the yeast rDNA locus. Genes Dev 2002; 16:2479-2484. 18. Brewer BJ, Lockshon D, Fangman W. The arrest of replication forks in the rDNA of yeast occurs independently of transcription. Cell 1992; 71:267-276. 19. Kobayashi T, Hidaka M, Nishizawa M et al. Identification of a site required for DNA replication fork blocking activity in the rRNA gene cluster in Saccharomyces cerevisiae. Mol & Gen Genet 1992; 233:355-362. 20. Kobayashi T, Horiuchi T. A yeast gene product, Fob1 protein, required for both replication fork blocking and recombinational hotspot activities. Genes Cells 1996; 1:465-474. 21. Long EO, Dawid IB. Repeated genes in eukaryotes. Annu Rev Biochem 1980; 49:727-764.

148

The Nucleolus

22. Hadjiolov AA. The nucleolus and ribosome biogenesis. Cell Biology Monographs. Wien: Springer-Verlag, 1985:12. 23. Kobayashi T, Heck DJ, Nomura M et al. Expansion and contraction of ribosomal DNA repeats in Saccharomyces cerevisiae: Requirement of replication fork blocking (Fob1) protein and the role of RNA polymerase I. Genes Dev 1998; 12:3821-3830. 24. Kobayashi T, Nomura M, Horiuchi T. Identification of DNA cis elements essential for expansion of ribosomal DNA repeats in Saccharomyces cerevisiae. Mol Cell Biol 2001; 21:136-147. 25. Choe SY, Schultz MC, Reeder RH. In vitro definition of the yeast RNA polymerase I promoter. Nucleic Acids Res 1992; 20:279-285. 26. Kulkens T, Riggs DL, Heck JD et al. The yeast RNA polymerase I promoter: ribosomal DNA sequences involved in transcription initiation and complex formation in vitro. Nucleic Acids Res 1991; 19:5363-5370. 27. Keys DA, Lee BS, Dodd JA et al. Multiprotein transcription factor UAF interacts with the upstream element of the yeast RNA polymerase I promoter and forms a stable preinitiation complex. Genes Dev 1996; 10:887-903. 28. Musters W, Knol J, Maas P et al. Linker scanning of the yeast RNA polymerase I promoter. Nucleic Acids Res 1989; 17:9661-9678. 29. Nogi Y, Yano R, Nomura M. Synthesis of large rRNAs by RNA polymerase II in mutants of Saccharomyces cerevisiae defective in RNA polymerase I. Proc Natl Acad Sci USA 1991; 88:3962-3966. 30. Wai HH, Vu L, Oakes M et al. Complete deletion of yeast chromosomal rDNA repeats and integration of a new rDNA repeat: Use of rDNA deletion strains for functional analysis of rDNA promoter elements in vivo. Nucleic Acids Res 2000; 28:3524-3534. 31. Elion EA, Warner JR. The major promoter element of rRNA transcription in yeast lies 2 kb upstream. Cell 1984; 39:663-673. 32. Elion EA, Warner JR. An RNA polymerase I enhancer in Saccharomyces cerevisiae. Mol Cell Biol 1986; 6:2089-2097. 33. Wai HH, Johzuka K, Vu L et al. Yeast RNA polymerase I enhancer is dispensable for transcription of the chromosomal rRNA gene and cell growth, and its apparent transcription enhancement from ectopic promoters requires fob1 protein. Mol Cell Biol 2001; 21:5541-5553. 34. van der Sande CAFM, Kulkens T, Kramer AB et al. Termination of transcription by yeast RNA polymerase I. Nucleic Acids Res 1989; 17:9127-9146. 35. Reeder RH, Guevara P, Roan JG. Saccharomyces cerevisiae RNA polymerase I terminates transcription at the Reb1 terminator in vivo. Mol Cell Biol 1999; 19:7369-7376. 36. Lang WH, Reeder RH The REB1 site is an essential component of a terminator for RNA polymerase I in Saccharomyces cerevisiae. Mol Cell Biol 1993; 13:649-658. 37. Lang WH, Reeder RH. Transcription termination of RNA polymerase I due to a T-rich element interactive with Reb1p. Proc Natl Acad Sci USA 1995; 92:9781-9785. 38. Lang WH, Morrow BE, Ju Q et al. A model for transcription termination by RNA polymerase I. Cell 1994; 79:527-534. 39. Reeder RH, Lang WH. Terminating transcription in eukaryotes: Lessons learned from RNA polymerase I. Trends Biochem Sci 1997; 22:473-477. 40. Laloraya S, Guacci V, Koshland D. Chromosomal addresses of the cohesin component Mcd1p. J Cell Biol 2000; 151:1047-1056. 41. Gotta M, Strahl-Bolsinger S, Renauld H et al. Localization of Sir2p: The nucleolus as a compartment for silent information regulators. EMBO J 1997; 16:3243-3255. 42. Cramer P, Bushnell DA, Fu J et al. Architecture of RNA polymerase II and implications for the transcription mechanism. Science 2000; 288:640-649. 43. Cramer P, Bushnell DA, Kornberg RD. Structural basis of transcription: RNA polymerase II at 2.8 ångstrom resolution. Science 2001; 292:1863-1876. 44. Gnatt AL, Cramer P, Fu J et al. Structural basis of transcription: an RNA polymerase II elongation complex at 3.3 å resolution. Science 2001; 292:1876-1882. 45. Mémet S, Gouy M, Marck C et al. RPA190, the gene coding for the largest subunit of yeast RNA polymerase A. J Biol Chem 1988; 263:2830-2839. 46. Yano R, Nomura M. Suppressor analysis of temperaturesensitive mutations of the largest subunit of RNA polymerase I in Saccharomyces cerevisiae: A suppressor gene encodes the second largest subunit of RNA polymerase I. Mol Cell Biol 1991; 11:754-764. 47. Mann C, Buhler J-M, Treich I et al. RPC40, a unique gene for a subunit shared between yeast RNA polymerases A and C. Cell 1987; 48:627-637. 48. Dequard-Chablat M, Riva M, Carles C et al. RPC19, the gene for a subunit common to yeast RNA polymerases A (I) and C (III). J Biol Chem 1991; 266:15300-15307.

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

149

49. Lalo D, Carles C, Sentenac A et al. Interactions between three common subunits of yeast RNA polymerase I and III. Proc Natl Acad Sci USA 1993; 90:5524-5528. 50. Ulmasov T, Larkin RM, Guilfoyle TJ. Association between 36- and 13.6-kDA α-like subunits of Arabidopsis thaliana RNA polymerase II. J Biol Chem 1996; 271:5085-5094. 51. Minakhin L, Bhagat S, Brunning A et al. Bacterial RNA polymerase subunit ω and eukaryotic RNA polymerase subunit RPB6 are sequence, structural, and functional homologs and promote RNA polymerase assembly. Proc Natl Acad Sci 2001; 98:892-897. 52. Woychik NA, Young RA. RNA polymerase II subunit RPB10 is essential for yeast cell viability. J Biol Chem 1990; 265:17816-17819. 53. Woychik NA, Liao SM, Kolodziej PA et al. Subunits shared by eukaryotic nuclear RNA polymerases. Genes Dev 1990; 4:313-323. 54. Carles C, Treich I, Bouet F et al. Two additional common subunits, ABC10 alpha and ABC10 beta, are shared by yeast RNA polymerases. J Biol Chem 1991; 266:24092-24096. 55. Lanzendörfer M, Smid A, Klinger C et al. A shared subunit belongs to the eukaryotic core polymerase. Genes Dev 1997; 11:1037-1047. 56. Rubbi L, LabarreMariotte S, Chedin S et al. Functional characterization of ABC10α, an essential polypeptide shared by all three forms of eukaryotic DNA-dependent RNA polymerases. J Biol Chem 1999; 274, 31485-31492. 57. Voutsina A, Riva M, Carles C et al. Sequence divergence of the RNA polymerase shared subunit ABC14.5 (Rpb8) selectively affects RNA polymerase III assembly in Saccharomyces cerevisiae. Nucl Acids Res 1999; 27:1047-1055. 58. Briand J-F, Navarro F, Rematier P et al. Partners of Rpb8p, a small subunit shared by yeast RNA polymerases I, II, and III. Mol Cell Biol 2001; 21:6056-6065. 59. Woychik NA, Lane WS, Young RA. Yeast RNA polymerase II subunit RPB9 is essential for growth at temperature extremes. J Biol Chem 1991; 266:19053-19055. 60. Nogi Y, Yano R, Dodd J et al. Gene RRN4 in Saccharomyces cerevisiae encodes the A12.2 subunit of RNA polymerase I and is essential only at high temperatures. Mol Cell Biol 1993; 13:114-122. 61. Chedin S, Riva M, Schultz P et al. The RNA cleavage activity of RNA polymerase III is mediated by an essential TFIIS-like subunit and is important for transcription termination. Genes Dev 1998; 12:3857-3871. 62. Van Mullem V, Landrieux E, Vandenhaute J et al. Rpa12p, a conserved RNA polymerase I subunit with two functional domains. Mol Microbiol 2002; 43:1105-1113. 63. Huet J, Buhler JM, Sentenac A et al. Dissociation of two polypeptide chains from yeast RNA polymerase A. Proc Natl Acad Sci USA 1975; 72:3034-3038. 64. Gadal O, Mariotte-Labarre S, Chedin S et al. A34.5, a nonessential component of yeast RNA polymerase I, cooperates with subunit A14 and DNA topoisomerase I to produce a functional rRNA synthesis machine. Mol Cell Biol 1997; 17:1787-1795. 65. Bischler N, Brino L, Carles C et al. Localization of the yeast RNA polymerase I-specific subunits. EMBO J 2002; 21:4136-4144. 66. Liljelund P, Mariotte S, Buhler JM et al. Characterization and mutagenesis of the gene encoding the A49 subunit of RNA polymerase A in Saccharomyces cerevisiae. Proc. Natl Acad Sci 1992; 89:9302-9305. 67. Gadal O, Labarre S, Boschiero C et al. Hmo1, an HMG-box protein, belongs to the yeast ribosomal DNA transcription system. EMBO J 2002; 21:5498-5507. 68. Smid A, Riva M, Bouet F et al. The association of three subunits with yeast RNA polymerase is stabilized by A14. J Biol Chem 1995; 270:13534-13540. 69. Thuriaux P, Mariotte S, Buhler JM et al. Gene RPA43 in Saccharomyces cerevisiae encodes an essential subunit of RNA polymerase I. J Biol Chem 1995; 270:24252-24257. 70. Peyroche G, Milkereit P, Bischler N et al. The recruitment of RNA polymerase I on rDNA is mediated by the interaction of the A43 subunit with Rrn3. EMBO J 2000; 19:5473-5482. 71. Peyroche G, Levillain E, Siaut M et al. The A14-A43 heterodimer subunit in yeast RNA pol I and their relationship to Rpb4-Rpb7 pol II subunits. Proc Natl Acad Sci USA 2002; 99:14670-14675. 72. Fath S, Milkereit P, Peyroche G et al. Differential roles of phosphorylation in the formation of transcriptional active RNA polymerase I. Proc Natl Acad Sci USA 2001; 98:14334-14339. 73. Schultz P, Célia H, Riva M et al. Three-dimensional model of yeast RNA polymerase I determined by electron microscopy of two-dimensional crystals. EMBO J 1993; 12:2601-2607. 74. Klinger C, Huet J, Song D et al. Localization of yeast RNA polymerase I core subunits by immunoelectron microscopy. EMBO J 1996; 15:4643-4653. 75. Grummt I. Regulation of mammalian ribosomal gene transcription by RNA polymerase I. Prog Nucleic Acid Res Mol Biol 1999; 62:109-154.

150

The Nucleolus

76. Zomerdijk JCBM, Tjian R. Initiation of transcription on human rRNA genes. In: Paule MR, ed. Transcription of Ribosomal RNA Genes by Eukaryotic RNA Polymerase I. Austin: RG Landes Co., 1998:121-134. 77. Nogi Y, Vu L, Nomura M. An approach for isolation of mutants defective in 35S ribosomal RNA synthesis in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1991; 88:7026-7030. 78. Riggs, DL, Nomura M. Specific transcription of Saccharomyces cerevisiae 35S rDNA by RNA polymerase I in vitro. J Biol Chem 1990; 265:7596-7603. 79. Lue NF, Kornberg RD. Accurately initiated, enhancer-dependent transcription by RNA polymerase I in yeast extracts. J Biol Chem 1990; 265:18091-18094. 80. Schultz MC, Choe SY, Reeder RH. Specific initiation by RNA polymerase I in a whole-cell extract from yeast. Proc Natl Acad Sci USA 1991; 88:1004-1008. 81. Schultz MC, Reeder RH, Hahn S. Variants of the TATA-binding protein can distinguish subsets of RNA polymerase I, II and III promoters. Cell 1992; 69:697-702. 82. Cormack BP, Struhl K. The TATA-binding protein is required for transcription by all three nuclear RNA polymerases in yeast cells. Cell 1992; 69:685-696. 83. Keys DA, Vu L, Steffan JS et al. RRN6 and RRN7 encode subunits of a multiprotein complex essential for the initiation of rDNA transcription by RNA polymerase I in Saccharomyces cerevisiae. Genes Dev 1994; 8:2349-2362. 84. Lalo D, Steffan JS, Dodd JA et al. RRN11 encodes the third subunit of the complex containing Rrn6p and Rrn7p that is essential for the initiation of rDNA transcription by yeast RNA polymerase I. J Biol Chem 1996; 271:21062-21067. 85. Lin CW, Moorefield B, Payne J et al. A novel 66-kilodalton protein complexes with Rrn6p, Rrn7p, and TATA-binding protein to promote polymerase I transcription initiation in Saccharomyces cerevisiae. Mol Cell Biol 1996; 16:6436-6443. 86. Steffan JS, Keys DA, Dodd JA et al. The role of TBP in rDNA transcription by RNA polymerase I in Saccharomyces cerevisiae: TBP is required for upstream-activation-factor-dependent recruitment of core factor. Genes Dev 1996; 10:2551-2563. 87. Keener J, Josaitis CA, Dodd JA et al. Reconstitution of yeast RNA polymerase I transcription in vitro from purified components. J Biol Chem 1998; 273:33795-33802. 88. Yamamoto RT, Nogi Y, Dodd JA et al. RRN3 gene of Saccharomyces cerevisiae encodes an essential RNA polymerase I transcription factor which interacts with the polymerase independently of DNA template. EMBO J 1996; 15:3964-3973. 89. Milkereit P, Tschochner H. A specialized form of RNA polymerase I, essential for initiation and growth-dependent regulation of rRNA synthesis, is disrupted during transcription. EMBO J 1998; 17:3692-3703. 90. Keener J, Dodd JA, Lalo D et al. Histones H3 and H4 are components of upstream activation factor required for the high-level transcription of yeast rDNA by RNA polymerase I. Proc Natl Acad Sci USA 1997; 94:13458-13462. 91. Siddiqi IN, Dodd JA, Vu L et al. Transcription of chromosomal rRNA genes by both RNA polymerase I and II in yeast uaf30 mutants lacking the 30 kDa subunit of transcription factor UAF. EMBO J 2001; 20:4512-4521. 92. Steffan JS, Keys DA, Vu L et al. Interaction of TATA-binding protein with upstream activation factor is required for activated transcription of ribosomal DNA by RNA polymerase I in Saccharomyces cerevisiae in vivo. Mol Cell Biol 1998; 18:3752-3761. 93. Aprikian P, Moorefield B, Reeder RH. New model for the yeast RNA polymerase I transcription cycle. Mol Cell Biol 2001; 21:4847-4855. 94. Bordi L, Cioci F, Camilloni G. In vivo binding and hierarchy of assembly of the yeast RNA polymerase I transcription factors. Mol Biol Cell 2001; 12:753-760. 95. Iben S, Tschochner H, Bier M. TFIIH plays an essential role in RNA polymerase I transcription. Cell 2002; 109:297-306. 96. Shou W, Seol JH, Shevchenko A et al. Exit from mitosis is triggered by Tem1-dependent release of the protein phosphatase Cdc14 from nucleolar RENT complex. Cell 1999; 97:233-244. 97. Straight AF, Shou W, Dowd GJ et al. Net1, a Sir2-associated nucleolar protein required for rDNA silencing and nucleolar integrity. Cell 1999; 97:245-256. 98. Visintin R, Hwang ES, Amon A. Cfi1 prevents premature exit from mitosis by anchoring Cdc14 phosphatase in the nucleolus. Nature 1999; 398:818-823. 99. Shou W, Sakamoto KM, Keener J et al. Net1 stimulates RNA polymerase I transcription and regulates nucleolar structure independently of controlling mitotic exit. Mol Cell 2001; 8:45-55. 100. Shimizu K, Kawasaki Y, Hiraga S-I et al. The fifth essential DNA polymerase phi in Saccharomyces cerevisiae is localized to the nucleolus and plays an important role in synthesis of rRNA. Proc Natl Acad Sci USA 2002; 99:9133-9138.

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

151

101. Moorefield B, Greene EA, Reeder RH. RNA polymerase I transcription factor Rrn3 is functionally conserved between yeast and human. Proc Natl Acad Sci USA 2000; 97:4724-4729. 102. Bodem J, Dobreva G, Hoffmann-Rohrer U et al. TIF-IA, the factor mediating growth-dependent control of ribosomal RNA synthesis, is the mammalian homolog of yeast Rrn3p. EMBO Rep 2000; 1:171-175. 103. Imazawa Y, Hisatake K, Nakagawa K et al. The fission yeast RPA21 subunit of RNA polymerase I: an evolutionarily conserved subunit interacting with ribosomal DNA (rDNA) transcription factor Rrn3p for recruitment to rDNA promoter. Genes Genet Syst 2002; 77:147-157. 104. Cavanaugh AH, Hirschler-Laszkiewicz I, Hu Q et al. Rrn3 phosphorylation is a regulatory checkpoint for ribosome biogenesis. J Biol Chem 2002; 277:27423-27432. 105. Miller G, Panov KI, Friedrich JK et al. hRRN3 is essential in the SL1-mediated recruitment of RNA polymerase I to rRNA gene promoters. EMBO J 2001; 20:1373-1382. 106. Boukhgalter B, Liu M, Guo A et al. Characterization of a fission yeast subunit of an RNA polymerase I essential transcription initiation factor, SpRrn7h/TAFI68, that bridges yeast and mammals: Association with SpRrn11h and the core ribosomal RNA gene promoter. Gene 2002; 291:187-201. 107. Reeder RH. Regulation of RNA polymerase I transcription in yeast and vertebrates. Prog Nucleic Acid Res Mol Biol 1999; 62:293-327. 108. McKnight SL, Miller Jr OL. Ultrastructural patterns of RNA synthesis during early embryogenesis of Drosophila melanogaster. Cell 1976; 8:305-319. 109. Dammann R, Lucchini R, Koller T et al. Chromatin structures and transcription of rDNA in yeast Saccharomyces cerevisiae. Nucleic Acids Res 1993; 21:2331-2338. 110. Sandmeier JJ, French S, Osheim Y et al. RPD3 is required for the inactivation of yeast ribosomal DNA genes in stationary phase. EMBO J 2002; 21:4959-4968. 111. French SL, Osheim YN, Cioci F et al. In exponentially growing Saccharomyces cerevisiae cells, ribosomal RNA synthesis is determined by the summed RNA polymerase I loading rate rather than by the number of active genes. Mol Cell Biol 2003; 23:1558-1568. 112. Buttgereit D, Pflugfelder G, Grummt I. Growth-dependent regulation of rRNA synthesis is mediated by a transcription initiation factor (TIF-IA). Nucleic Acids Res 1985; 13:8165-8180. 113. Heitman J, Movva NR, Hall MN. Targets for cell cycle arrest by the immunosuppressant rapamycin in yeast. Science 1991; 253:905-909. 114. Barbet NC, Schneider U, Helliwell SB et al. TOR controls translation initiation and early G1 progression in yeast. Mol Biol Cell 1996; 7:25-42. 115. Schmelzle T, Hall MN. TOR, a central controller of cell growth. Cell 2000; 103:253-262. 116. Zaragoza D, Ghavidel A, Heitman J et al. Rapamycin induces the G0 program of transcriptional repression in yeast by interfering with the TOR signaling pathway. Mol Cell Biol 1998; 18:4463-4470. 117. Powers T, Walter P. Regulation of ribosome biogenesis by the rapamycin-sensitive TOR-signaling pathway in Saccharomyces cerevisiae. Mol Biol Cell 1999; 10:987-1000. 118. Jacinto E, Guo B, Arndt KT et al. TIP41 interacts with TAP42 and negatively regulates the TOR signaling pathway. Mol Cell 2001; 8:1017-1026. 119. Jiang Y, Broach JR. Tor proteins and protein phosphatase 2A reciprocally regulate Tap42 in controlling cell growth in yeast. EMBO J 1999; 18:2782-2792. 120. Shulman RW, Sripati CE, Warner JR. Noncoordinated transcription in the absence of protein synthesis in yeast. J Biol Chem 1977; 252:1344-1349. 121. Warner JR, Gorenstein C. Yeast has a true stringent response. Nature 1978; 275:338-339. 122. Hinnebusch AG. Translational regulation of yeast GNC4. J Biol Chem 1997; 272:21661-21664. 123. McLaughlin CS, Magee PT, Hartwell LH. Role of isoleucyl-transfer ribonucleic acid synthetase in ribonucleic acid synthesis and enzyme repression in yeast. J Bacteriol 1969; 100:579-584. 124. Moehle CM, Hinnebusch AG. Association of RAP1 binding sites with stringent control of ribosomal protein gene transcription in Saccharomyces cerevisiae. Mol Cell Biol 1991; 11:2723-2735. 125. Iiboshi Y, Papst PJ, Kawasome H et al. Amino acid-dependent control of p70s6k. Involvement of tRNA aminoacylation in the regulation. J Biol Chem 1999; 274:1092-1099. 126. Dennis PB, Jaeschke A, Saitoh M et al. Mammalian TOR: A homeostatic ATP sensor. Science 2001; 294:1102-1105. 127. Mizuta K, Warner JR. Continued functioning of the secretory pathway is essential for ribosome synthesis. Mol Cell Biol 1994; 14:2493-2502. 128. Nierras CR, Warner JR. Protein kinase C enables the regulatory circuit that connects membrane synthesis to ribosome synthesis in Saccharomyces cerevisiae. J Biol Chem 1999; 274:13235-13241. 129. Krause SA, Gray JV. The protein kinase C pathway is required for viability in quiescence in Saccharomyces cerevisiae. Curr Biol 2002; 12:588-593.

152

The Nucleolus

130. Tsuno A, Miyoshi K, Tsujii R et al. RRS1, a conserved essential gene, encodes a novel regulatory protein required for ribosome biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 2000; 20:2066-2074. 131. Miyoshi K, Tsujii R, Yoshida H et al. Normal assembly of 60 S ribosomal subunits is required for the signaling in response to a secretory defect in Saccharomyces cerevisiae. J Biol Chem 2002; 277:18334-18339. 132. Granot D, Snyder M. Segregation of the nucleolus during mitosis in budding and fission yeast. Cell Motil Cytoskel 1991; 20:47-54. 133. Elliott SG, McLaughlin CS. The yeast cell cycle: Coordination of growth and division rates. Prog Nucleic Acid Res Mol Biol 1983; 28:143-176. 134. Elliott SG, McLaughlin CS. Regulation of RNA synthesis in yeast III. Molec Gen Genet 1979; 169:237-243. 135. Elliott SG, Warner JR, McLaughlin CS. Synthesis of ribosomal proteins during the cell cycle of the yeast Saccharomyces cerevisiae. J Bacteriol 1979; 137:1048-1050. 136. Guacci V, Hogan E, Koshland D. Chromosome condensation and sister chromatid pairing in budding yeast. J Cell Biol 1994; 125:517-530. 137. Lawrence JB, Villnave CA, Singer RH. Sensitive, high-resolution chromatin and chromosome mapping in situ: Presence and orientation of two closely integrated copies of EBV in a lymphoma line. Cell 1988; 52:51-61. 138. Heix J, Vente A, Voit R et al. Mitotic silencing of human rRNA synthesis: Inactivation of the promoter selectivity factor SL1 by cdc2/cyclin B-mediated phosphorylation. EMBO J 1998; 17:7373-7381. 139. Sirri V, Roussel P, Hernandez-Verdun D. In vivo release of mitotic silencing of ribosomal gene transcription does not give rise to precursor ribosomal RNA processing. J Cell Biol 2000; 148:259-270. 140. Freeman-Cook L, Kamakaka R, Pillus L. Chromatin contributions to epigenetic transcriptional states in yeast. In: Elgin SCR, Workman JL, eds. Chromatin Structure and Gene Expression, 2nd Ed. New York: Oxford University Press, 2000:203-227. 141. Gasser SM, Cockell MM. The molecular biology of the SIR proteins. Gene 2001; 279:1-16. 142. Moazed D. Common themes in mechanisms of gene silencing. Mol Cell 2001; 8:489-498. 143. Smith JS, Boeke JD. An unusual form of transcriptional silencing in yeast ribosomal DNA. Genes Dev 1997; 11:241-254. 144. Bryk M, Banerjee M, Murphy M et al. Transcriptional silencing of Ty1 elements in the RDN1 locus of yeast. Genes Dev 1997; 11:255-269. 145. Vu L, Siddiqi I, Lee BS et al. RNA polymerase switch in transcription of yeast rDNA: role of transcription factor UAF (upstream activation factor) in silencing rDNA transcription by RNA polymerase II. Proc Natl Acad Sci USA 1999; 96:4390-4395. 146. Oakes M, Siddiqi I, Vu L et al. Transcription factor UAF, expansion and contraction of ribosomal DNA (rDNA) repeats, and RNA polymerase switch in transcription of yeast rDNA. Mol Cell Biol 1999; 19:8559-8569. 147. Conrad-Webb H, Butow RA. A polymerase switch in the synthesis of rRNA in Saccharomyces cerevisiae. Mol Cell Biol 1995; 15:2420-2428. 148. Imai S-I, Armstrong CM, Kaeberlein M et al. Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 2000; 403:795-800. 149. Landry J, Sutton A, Tafrov ST et al. The silencing protein SIR2 and its homologs are NAD-dependent protein deacetylases. Proc Natl Acad Sci USA 2000; 97:5807-5811. 150. Braunstein M, Rose AB, Holmes SG et al. Transcriptional silencing in yeast is associated with reduced nucleosome acetylation. Genes Dev 1993; 7:592-604. 151. Hecht A, Strahl-Bolsinger S, Grunstein M. Spreading of transcriptional repressor SIR3 from telomeric heterochromatin. Nature 1996; 383:92-96. 152. Smith JS, Caputo E, Boeke JD. A genetic screen for ribosomal DNA silencing defects identifies multiple DNA replication and chromatin-modulating factors. Mol Cell Biol 1999; 19:3184-3197. 153. Bryk M, Briggs SD, Strahl BD et al. Evidence that Set1, a factor required for methylation of histone H3, regulates rDNA silencing in S cerevisiae by a Sir2-independent mechanism. Curr Biol 2002; 12:165-170. 154. Briggs SD, Bryk M, Strahl BD et al. Histone H3 lysine 4 methylation is mediated by Set1 and required for cell growth and rDNA silencing in Saccharomyces cerevisiae. Genes Dev 2001; 15:3286-3295. 155. Sun Z-W, Allis CD. Ubiquitination of histone H2B regulates H3 methylation and gene silencing in yeast. Nature 2002; 418:104-108.

Transcription of rDNA in the Yeast Saccharomyces cerevisiae

153

156. Hecht A, Laroche T, Strahl-Bolsinger S et al. Histone H3 and H4 N-termini interact with SIR3 and SIR4 proteins: A molecular model for the formation of heterochromatin in yeast. Cell 1995; 80:583-592. 157. Moazed D, Kistler A, Axelrod A et al. Silent information regulator protein complexes in Saccharomyces cerevisiae: A SIR2/SIR4 complex and evidence for a regulatory domain in SIR4 that inhibits its interaction with SIR3. Proc Natl Acad Sci USA 1997; 94:2186-2191. 158. Ghidelli S, Donze D, Dhillon N et al. Sir2p exists in two nucleosome-binding complexes with distinct deacetylase activities. EMBO J 2001; 20:4522-4535. 159. Fritze CE, Verschueren K, Strich R et al. Direct evidence for SIR2 modulation of chromatin structure in yeast rDNA. EMBO J 1997; 16:6495-6509. 160. Dammann R, Lucchini R, Koller T et al. Transcription in the yeast rRNA gene locus: Distribution of the active gene copies and chromatin structure of their flanking regulatory sequences. Mol Cell Biol 1995; 15:5294-5303. 161. Buck SW, Sandmeier JJ, Smith JS. RNA polymerase I propagates unidirectional spreading of rDNA silent chromatin. Cell 2002; 111:1003-1014. 162. Cioci F, Vu L, Eliason K et al. Silencing in yeast rDNA chromatin: Reciprocal relationship in gene expression between RNA polymerase I and II. Mol Cell 2003; in press. 163. Oakes M, Aris JP, Brockenbrough JS et al. Mutational analysis of the structure and localization of the nucleolus in the yeast Saccharomyces cerevisiae. J Cell Biol 1998; 143:23-34. 164. Nomura M. Ribosomal RNA genes, RNA polymerases, nucleolar structures, and synthesis of rRNA in the yeast Saccharomyces cerevisiae. Cold Spring Harbor Symp Quant Biol 2001; 66:555-565. 165. Bréant B, Buhler JM, Sentenac A et al. On the phosphorylation of yeast RNA polymerase A and B. Eur J Biochem 1983; 130:247-251.

154

The Nucleolus

CHAPTER 10

Three-Dimensional Organization of rDNA and Transcription Dominique Ploton, Marie-Françoise O’Donohue, Thierry Cheutin, Adrien Beorchia, Hervé Kaplan and Marc Thiry

Abstract

I

n the nucleolus, relating ultrastructural features observed by electron microscopy to biochemical events of biogenesis and step-wise maturation of pre-rRNAs has led for many years to contradictory results. Technical limitations were partly responsible for this failure. In this chapter, we describe new methodologies that have recently been developed to overcome these problems. By giving a three-dimensional view of these cellular events, the results obtained bring new insights to rRNA gene transcription and pre-rRNA synthesis and maturation.

Introduction

The nucleolus was first described in the early 19th century as a nuclear sub-compartment.1 A few decades ago, it was further characterized as the nuclear compartment in which synthesis and maturation of cytoplasmic rRNAs mainly take place.2,3 This process entails transcription of rDNA genes and gradual maturation of the primary ribosomal transcripts (pre-rRNAs, 45-47 S), which contain long external and internal transcribed spacers. The primary transcripts are processed through a complex pathway of endonucleolytic cleavages, exonucleolytic digestions and covalent modifications to give 18 S, 5.8 S, and 28 S rRNAs, which are finally associated with ribosomal and non-ribosomal proteins in addition to the separately transcribed 5 S rRNAs.4 In this way, the large and small pre-ribosomal particles are formed and subsequently released from the nucleolus into the cytoplasm, where they assembled into fully functional ribosomes. Pre-rRNAs are synthesised by RNA polymerase I, a complex enzyme entirely devoted to the rRNA gene transcription, and its associated machinery.5 The rRNA genes of higher eukaryotes occur in multiple, tandemly repeated copies, that are separated by (non-transcribed) intergenic “spacers”.6 However, only a subset of all rRNA genes is normally active and their number is tightly up- or down-regulated depending on cellular activity. At the ultrastructural level, three major substructures are observed in the nucleoli of mammalian cells: the fibrillar center (FC), the dense fibrillar component (DFC) and the granular component (GC) (Fig. 1).7 These nucleolar sub-compartments are not only characteristic of a cell type, but their number and organization are directly connected to cell activity. During the past decade, many researchers have attempted to correlate these morphologically distinct structures with biochemical features of the synthesis of rRNAs and stepwise maturation of ribosomes described at the molecular level.8-11 However, despite an accumulation of experimental data performed by many different groups, the conclusions remained contradictory. The main reasons lay both in the spatial complexity undergone by the transcribing rDNA gene, and in the speed of the transcriptional process. Recently developed approaches allowed us to tackle this biological problem and brought some clues toward its elucidation. The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Three-Dimensional Organization of rDNA and Transcription

155

Figure 1. Identification of the nucleolar compartments at the ultrastructural level. After embedding, ultrathin sections (80 nm) of KB cells were counterstained with lead and uranyl acetate and observed with an electron microscope at 100 kV. The nucleolus appears as a denser area within the nucleus. It is composed of three distinct compartments: the fibrillar centers (FC), the dense fibrillar component (DFC), and the granular component (GC). The bar is 500 nm.

Organization of Actively Transcribed rDNA Genes Our present knowledge of the molecular organization of transcribed rDNA genes was mainly established using chromatin spreads.12 After a hypotonic treatment, each transcribing gene de-condenses and appears as a so-called “Christmas tree”, ~5 µm in length. It consists of 135 to 180 fibrils of growing pre-rRNAs, each of which is connected to the rDNA axis by a RNA polymerase I molecule.13 On their free 5’ end, the nascent transcripts carry a 25 nm thick granule called the “terminal ball”, which contains the processing machinery complex.14 Considering the length of an extended, transcribed rRNA gene relative to the size of a nucleolus, it is obvious that these structures are compacted in the nucleoli. However, localisation of the “Christmas trees” relative to the nucleolar compartments within the nucleolus has proved to be a complex problem. RNA polymerase I molecules have been clearly detected in the FC.15,16 Pre-rRNA synthesis, investigated by incorporation of [3H]-UTP or BrUTP, occurs in the DFC,17-22 although the results of some authors indicate a low level of synthesis in the FC.18,21,23-27 With regard to the rDNA transcribed sequences, the situation remains contradictory, by using in situ hybridisation rDNA was found either in the DFC,28-31 or in the FC.32-34 Therefore, although it is clear that transcription occurs in the fibrillar regions of the nucleolus, it is unclear whether it is within the DFC, or at the border of the DFC and the FC, or within the FC. The data obtained so far are contradictory mainly because the speed of rRNA synthesis is very high35 (25-50 nucleotides/second). Thus, the localisation of incorporated BrUTP (or [3H]-UTP) is probably representative of both incorporation and accumulation sites. In order to determine in which compartment pre-rRNA transcripts are synthesised, it is necessary to develop experimental procedures in which transcription is slowed down. In addition, conventional microscopy approaches give two-dimensional pictures that only account for a partial view of the labelling. Indeed, immunocytochemistry only detects particles located at the surface of the ultra-thin sections. On the contrary, the electron micrographs display a projection of the ultrastructural components contained in the preparation thickness, that although reduced is not negligible. As a result, the number of labelled molecules is underestimated, while their positioning with regard to the ultrastructure can be very ambiguous, specially when the particles are located at the border between two different compartments (Fig. 2). As a consequence, the precision of the localisation of a given labelling increases with the diameter of the FCs, which means that cell lines displaying larger FCs should be preferred for these studies. Moreover, classical electron microscopy brings very useful information to localise a

156

The Nucleolus

Figure 2. Schematic representation of the localisation at the ultrastructural level of labelled particles with regard to the fibrillar components of the nucleolus. A fibrillar center (FC) (hatched lines), 500 nm in diameter, is surrounded by the dense fibrillar component (DFC) (dotted area). Ultrathin sections (#1-3), performed within the fibrillar components of a nucleolus, represent volumes whose thickness, although reduced, is not negligible (~100 nm). Since the micrographs obtained by electron microscopy are projections, the area occupied by the DFC can be overestimated, depending on the position at which the sectioning has been performed (for example, compare transversal sections and projections obtained from section 1 to those corresponding to sections 2 and 3). When labelling is performed after embedding, the detected particles (black dots) are only those located at the surface of the preparation, as seen on transversal sections. Here, three cases are represented: immunodetection of a protein located in the FC (left lane), at the border of the FC and the DFC (middle lane), or in the DFC (right lane). On sections 2 and 3, particles positioned on the FC (left lane) or at the border between FC and DFC (middle lane) are in fact predominantly seen on the DFC (section 2), or even entirely located in this compartment (section 3). Consequently, it is important to choose a cellular model in which the FCs are large in order to get a more accurate localisation.

labelling with regard to the ultrastructural components, but this technique cannot describe the three-dimensional organization of rDNA genes in situ. Here, we describe various approaches that have been developed recently to overcome these technical problems and tentatively answer the biological questions of both the functional organization of rRNA gene transcription in vivo and the precise three-dimensional localisation of pre-rRNA synthesis and processing.

Three-Dimensional Organization of rDNA and Transcription

157

Figure 3. Study by electron tomography of A549 cells immunolabelled with anti-RNA polymerase I antibodies. (A): a 500 nm-thick section, observed using a STEM working at 250 kV, shows several independent clusters, 270 nm in diameter. Human A549 cells were chosen because their FCs, which are small and homogenous in size, are fully contained in the section thickness. (B-D): different projections of the tomogram were calculated after three-dimensional reconstruction of the cluster framed in (A). At +15° (D), five 60-nm coils are evidenced, as indicated by brackets (#1-5). The circle shows the area where the coils are fused together. The arrows point to 20 nm thick twines (B-D). (E): a stereo-pair of the tomogram, presented in the same orientation as in (D), was calculated with a surfacic rendering mode. (F-I): four successive 30 nm-thick sections were achieved with a coronal orientation within the tomogram presented in (D). Asterisks (G, H) indicate the internal part of the cluster, devoid of labelling, and arrows (G-I) refer to twines. The bar is 200 nm (A) or 100 nm (B-I). Reprinted with permission from: Cheutin et al, J. Cell Sci. (2002) 115, 3297-3307. The Company of Biologists Limited. Movies corresponding to this figure are available at: http://www.eurekah.com/abstract.php?chapid=1296&bookid=88&catid=54.

Relating rRNA Gene Transcription to Its in Situ Organization through a Three Dimensional Approach Electron tomography is the only technique that can reveal the distribution of labelled particles within a cellular volume at the electron microscope level.36,37 In a recent study, the three-dimensional organization of active rDNA genes was detected prior to embedding through RNA polymerase I molecules, by using specific antibodies revealed by fluoronanogold (Nanoprobes), an electron-dense probe combined with a fluorescent dye.38 Sections whose thickness (500 nm) was sufficient to completely visualise the transcription sites were prepared and observed in a Philips medium-voltage CM30 electron microscope working at 250 kV in the STEM mode.39 Each section was tilted every 2° from -50° to +50°, thus allowing fifty-one

158

The Nucleolus

Figure 4. Organization of the coils. (A, B): two successive 1.5 nm-thick sections were performed parallel to the long axis of coil #2. The positions of seven twines, orthogonal to the long axis of coil #2, are indicated by arrows. (C, D): two successive 1.5 nm-thick sections performed perpendicularly to the long axis of coils #1-3. Twines (circles) constitute closed or partially open circles. The bar is 100 nm. (E): stereo-pair showing the twines. The volume of the cluster was cut in half along a plane parallel to the long axis of coil #2. To simplify the visualisation, the rear part was eliminated. Several open and closed circles corresponding to bent twines are evidenced. Reprinted with permission from: Cheutin et al, J. Cell Sci. (2002) 115, 3297-3307. The Company of Biologists Limited. Movies corresponding to this figure are available at: http:// www.eurekah.com/abstract.php?chapid=1296&bookid=88&catid=54.

projections to be recorded. After image alignment by using a sinogram technique,40 three-dimensional reconstructions were performed by using an extended field-additive algorithm-reconstruction technique.36 As shown in Figure 3, RNA polymerase I molecules formed several discrete clusters, organized as spheroids ∼270 nm in diameter. Each cluster contained ∼150 individualized silver particles and displayed a similar volumic organization. Thanks to a rotation at +15° (Fig. 3, D and E; Movie 1, available at: http://www.eurekah.com/ abstract.php?chapid=1296&bookid=88&catid=54 ), it was clearly seen that the particles formed three curved coils (#1-3) and two shorter ones (#4-5). These coils, 60 nm in diameter, shared a common origin (circles on D and I), but displayed separate extremities (brackets on D). By computing coronal sections (Fig. 3, F-I), the central region of the cluster appeared devoid of particles (asterisks). The particles forming the coils were frequently aligned to form 20 nm-thick twines with a free terminal end (arrows on G-I; Movie 2, available at: http://www.eurekah.com/ abstract.php?chapid=1296&bookid=88&catid=54 ). When sections were performed perpendicularly to the axis of the coil, it appeared that the twines formed full or partially open circles, 60 nm in diameter (Fig. 4, circles on C and D; Movies 3 and 4, available at: http:// www.eurekah.com/abstract.php?chapid=1296&bookid=88&catid=54 ). Finally, counterstained ultrathin sections, immunolabelled according to the same experimental protocol as that used in STEM, were observed by conventional electron microscopy. The micrographs obtained unambiguously demonstrated that RNA polymerase I molecules are mainly localised within the FC (Fig. 5), as shown previously.15,16,41

Three-Dimensional Organization of rDNA and Transcription

159

Figure 5. Ultrastructural localisation of RNA polymerase I within A549 cells. Anti-RNA polymerase I antibodies were revealed with fluoronanogold (Nanoprobres), followed by silver enhancement. After embedding, ultrathin sections (80 nm) were counterstained and observed with an electron microscope at 100 kV. The main nucleolar components are identified (FC, DFC and GC). A high density of particles is observed within the fibrillar components of the nucleolus. Noticeably, the diameter of the FCs is about 250-300 nm, which corresponds to the size of the clusters observed on Figure 3. The bar is 500 nm. Reprinted with permission from: Cheutin et al, J. Cell Sci. (2002) 115, 3297-3307. The Company of Biologists Limited.

Relating 47 S Pre-rRNA Synthesis to Nucleolar Compartments at the Ultrastructural Level The experimental data obtained by electron tomography showed the detailed organization of RNA polymerase I in situ, but this needed to be related to incorporation data, in order to localise both the active rRNA genes and their primary 45 S/47 S transcripts. Since the speed of rRNA synthesis in vivo is very high35 (25-50 nucleotides/second), the localisation of incorporated BrUTP (or [3H]-UTP) could be representative of both incorporation and accumulation sites. In order to solve this problem, we used isolated nucleoli42, which exhibit decreased transcriptional activity compared to intact cells.2,35 To reduce localisation problems (Fig. 2), Ehrlich ascite tumour cells, displaying large FCs, were used. The nucleoli were incubated in the presence of BrUTP for increasing lengths of time, then the distribution of Br-RNAs was visualised after immunogold labelling with 10 nm gold particles. Quantification of labelling over a 25 min period revealed that its density was progressively increased on both the FC and the DFC as a function of time (Fig. 6). In order to determine whether pre-rRNAs moved toward new compartments during the elongation process, the distribution of labelling was analysed in isolated nucleoli submitted to a 10 minute pulse with BrUTP, followed by an optional 20 minute chase with UTP (Fig. 7). In both cases, labelling was consistently present on the fibrillar components. However, the labelling density on the FC was significantly higher than that on the DFC in the pulse experiment, while it was higher on the DFC relative to the FC after the chase. This demonstrates that the FC is the site of primary BrUTP incorporation and that the pre-rRNAs enter within the surrounding DFC during the elongation process. Then, isolated nucleoli were incubated with BrUTP after a transient inhibition of elongation by cordycepin. This inhibitor leads to premature transcription termination and release of incomplete transcripts from their templates,43,44 but after the release of inhibition, RNA polymerase I molecules

160

The Nucleolus

Figure 6. Quantification of BrUTP particles incorporated in nascent pre-rRNA transcripts. After an ultrastructural localisation of nascent rRNA molecules within the nucleolus, densities (gold particles/µ2) on the FC, DFC, GC and resin (R) were calculated in nucleoli isolated from Ehrlich ascites tumour cells and incubated with BrUTP for 1 to 25 min. This cell line was selected because it contains large FCs relative to the DFC. Results represent mean values ± SEM. 7, 14, 13 and 12 random micrographs were analysed and 51, 120, 211 and 274 gold particles were counted, respectively. Student’s t-test for nucleolar components versus resin (+ P3' exonucleases. RNA 2000; 6:1698-1703. 49. Allmang C, Kufel J, Chanfreau G et al. Functions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J 1999; 18:5399-5410. 50. Mitchell P, Petfalski E, Tollervey D. The 3' end of yeast 5.8S rRNA is generated by an exonuclease processing mechanism. Genes Devel 1996; 10:502-513. 51. Faber AW, van Dijk M, Raué HA et al. Ngl2p is a Ccr4p-like RNA nuclease essential for the final step in 3'-end processing of 5.8S rRNA in Saccharomyces cerevisiae. RNA 2002; 8:1095-1101. 52. Allmang C, Henry Y, Wood H et al. Recognition of cleavage site A2 in the yeast pre-rRNA. RNA 1996; 2:51-62. 53. Torchet C, Hermann-Le Denmat S. Bypassing the rRNA processing endonucleolytic cleavage at site A2 in Saccharomyces cerevisiae. RNA 2000; 6:1498-1508. 54. Bousquet-Antonelli C, Vanrobays E, Gélugne J-P et al. Rrp8p is a yeast nuclear protein functionally linked to Gar1p and involved in pre-rRNA cleavage at site A2. RNA 2000; 6:826-843. 55. Henry Y, Wood H, Morrissey JP et al. The 5' end of yeast 5.8S rRNA is generated by exonucleases from an upstream cleavage site. EMBO J 1994; 13:2452-2463. 56. Chu S, Archer RH, Zengel JM et al. The RNA of RNase MRP is required for normal processing of ribosomal RNA. Proc Natl Acad Sci USA 1994; 91:659-663. 57. Dichtl B, Tollervey D. Pop3p is essential for the activity of the RNase MRP and RNase P ribonucleoproteins in vivo. EMBO J 1997; 16:417-429. 58. Eppens EA, Faber AW, Rondaij M et al. Deletions in the S1 domain of Rrp5p cause processing at a novel site in ITS1 of yeast pre-rRNA that depends on Rex4p. Nucl Acids Res 2002; 30:4222-4231. 59. Briggs MW, Burkard KTD, Butler JC. Rrp6p, the yeast homologue of the human PM-Scl 100-kDa autoantigen, is essential for efficient 5.8 S rRNA 3' end formation. J Biol Chem 1998; 273:13255-13263. 60. Nissan TA, Baßler J, Petfalski E et al. 60S Pre-ribosome formation viewed from assembly in the nucleus until export to the cytoplasm. EMBO J 2002; 21:5539-5547.

Pre-Ribosomal RNA Processing and Assembly in Saccharomyces cerevisiae

219

61. Hsu CL, Stevens A. Yeast cells lacking 5'-3' exoribonuclease 1 contain mRNA species that are poly(A) deficient and partially lack the 5' cap structure. Mol Cell Biol 1993; 13:4826-4835. 62. Johnson AW. Rat1p and Xrn1p are functionally interchangeable exoribonucleases that are restricted to and required in the nucleus and cytoplasm, respectively. Mol Cell Biol 1997; 17:6122-6130. 63. Kressler D, Linder P, De La Cruz J. Protein trans-acting factors involved in ribosome biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:7897-7912. 64. Venema J, Planta RJ, Raué HA. In vivo mutational analysis of ribosomal RNA in Saccharomyces cerevisiae. In: Martin R, ed. Protein Synthesis: Methods and Protocols. Vol 77. Totowa, NJ: Humana Press, 1998:257-270. 65. Beltrame M, Tollervey D. Base pairing between U3 and the pre-ribosomal RNA is required for 18S rRNA synthesis. EMBO J 1995; 14:4350-4356. 66. Van Nues RW, Venema J, Rientjes JMJ et al. Processing of eukaryotic pre-rRNA: The role of the transcribed spacers. Biochem Cell Biol 1995; 73:789-811. 67. Venema J, Henry Y, Tollervey D. Two distinct recognition signals define the site of endonucleolytic cleavage at the 5'-end of yeast 18S rRNA. EMBO J 1995; 14:4883-4892. 68. Côté CA, Peculis BA. Role of the ITS2-proximal stem and evidence for indirect recognition of processing sites in pre-rRNA processing in yeast. Nucl Acids Res 2001; 29:2106-2116. 69. Van Beekvelt CA, De Graaff-Vincent M, Faber AW et al. All three functional domains of the large ribosomal subunit protein L25 are required for both early and late pre-rRNA processing steps in Saccharomyces cerevisiae. Nucl Acids Res 2001; 29:5001-5008. 70. Côté CA, Greer CA, Peculis BA. Dynamic conformational model for the role of ITS2 in pre-rRNA processing in yeast. RNA 2002; 8:786-797. 71. Hughes JMX. Functional base-pairing interaction between highly conserved elements of U3 small nucleolar RNA and the small ribosomal subunit RNA. J Mol Biol 1996; 259:645-654. 72. Sharma K, Tollervey D. Base pairing between U3 small nucleolar RNA and the 5' end of 18S rRNA is required for pre-rRNA processing. Mol Cell Biol 1999; 19:6012-6019. 73. Lindahl L, Archer R, Zengel JM. Alternate pathways for processing in the Internal Transcribed Spacer 1 in pre-rRNA of Saccharomyces cerevisiae. Nucl Acids Res 1994; 22:5399-5407. 74. Van Nues RW, Rientjes JMJ, Van der Sande CAFM et al. Separate structural elements within Internal Transcribed Spacer I of Saccharomyces cerevisiae precursor ribosomal RNA direct the formation of 17S and 26S rRNA. Nucl Acids Res 1994; 22:912-919. 75. Van Beekvelt CA, Jeeninga RE, Van ‘t Riet J et al. Identification of cis-acting elements involved in 3’-end formation of Saccharomyces cerevisiae 18S rRNA. RNA 2001; 7:896-903. 76. Yeh L-C, Thweatt R, Lee JC. Internal Transcribed Spacer 1 of the yeast precursor ribosomal RNA. Higher order structure and common structural motifs. Biochemistry 1990; 29:5911-5918. 77. Lygerou Z, Allmang C, Tollervey D et al. Accurate processing of a eukaryotic precursor ribosomal RNA by ribonuclease MRP in vitro. Science 1996; 272:268-270. 78. Van Nues RW, Rientjes JMJ, Morré SA et al. Evolutionary conserved structural elements are critical for processing of Internal Transcribed Spacer 2 from Saccharomyces cerevisiae. J Mol Biol 1995; 250:24-36. 79. Peculis BA, Greer CL. The structure of the ITS2-proximal stem is required for pre-rRNA processing in yeast. RNA 1998; 4:1610-1622. 80. Yeh L-CC, Lee JC. Structural analysis of the Internal Transcribed Spacer 2 of the precursor ribosomal RNA from Saccharomyces cerevisiae. J Mol Biol 1990; 211:699-712. 81. Joseph N, Krauskopf E, Vera MI et al. Ribosomal internal transcribed spacer 2 (ITS2) exhibits a common core of secondary structure in vertebrates and yeast. Nucl Acids Res 1999; 27:4533-4540. 82. Allmang C, Tollervey D. The role of the 3' external transcribed spacer in yeast pre-rRNA processing. J Mol Biol 1998; 278:67-78. 83. Gerbi SA. Expansion segments: Regions of variable size that interrupt the universal core secondary structure of ribosomal RNA. In: Zimmermann RA, Dahlberg AE, eds. Ribosomal RNA: Structure, Evolution, Processing and Function in Protein Biosynthesis. Boca Raton: CRC Press, 1996:71-87. 84. Jeeninga RE, Van Delft Y, DeGraaff-Vincent M et al. Variable regions V13 and V3 of Saccharomyces cerevisiae contain structural features essential for normal biogenesis and stability of 5.8S and 25S rRNA. RNA 1997; 3:476-488. 85. Van Nues RW, Venema J, Planta RJ et al. Variable region V1 of Saccharomyces cerevisiae 18S rRNA participates in biogenesis and function of the small ribosomal subunit. Chromosoma 1997; 105:523-531. 86. Van Beekvelt CA, Kooi EA, De Graaff-Vincent M et al. Domain III of Saccharomyces cerevisiae 25S ribosomal RNA: Its role in binding of ribosomal protein L25 and 60S subunit formation. J Mol Biol 2000; 296:7-17.

220

The Nucleolus

87. Jansen R, Tollervey D, Hurt EC. A U3 snoRNP protein with homology to splicing factor PRP4 and G-beta-domains is required for ribosomal RNA processing. EMBO J 1993; 12:2549-2558. 88. Wehner KA, Gallagher JEG, Baserga SJ. Components of an interdependent unit within the SSU processome regulate and mediate its activity. Mol Cell Biol 2002; 22:7258-7267. 89. Kaser A, Bogengruber E, Hallegger M et al. Brix from Xenopus laevis and Brx1p from yeast define a new family of proteins involved in the biogenesis of large ribosomal subunits. Biol Chem 2001; 382:1637-1647. 90. Pestov DG, Stockelman MG, Strezoska Z et al. ERB1, the yeast homolog of mammalian BOP1, is an essential gene required for maturation of the 25S and 5.8S ribosomal RNAs. Nucl Acids Res 2001; 29:3621-3630. 91. Strezoska Z, Pestov DG, Lau LF. Functional Inactivation of the mouse nucleolar protein Bop1 inhibits multiple steps in pre-rRNA processing and blocks cell cycle progression. J Biol Chem 2002; 277:29617-29625. 92. Rigaut G, Shevchenko A, Rutz B et al. A generic protein purification method for protein complex characterization and proteome exploration. Nature Biotechnol 1999; 17:1030-1032. 93. Baßler J, Grandi P, Gadal O et al. Identification of a 60S preribosomal particle that is closely linked to nuclear export. Mol Cell 2001; 8:517-529. 94. Saveanu C, Bienvenu D, Namane A et al. Nog2p, a putative GTPase associated with pre-60S subunits and required for late 60S maturation steps. EMBO J 2001; 20:6475-6484. 95. Fatica A, Tollervey D. Making ribosomes. Curr Op Cell Biol 2002; 14:313-318. 96. Intinea RVA, Good L, Nazar RN. Essential structural features in the Schizosaccharomyces pombe pre-rRNA 5' external transcribed spacer. J Mol Biol 1999; 286:695-708. 97. Lalev AI, Abeyrathne PD, Nazar RN. Ribosomal RNA maturation in Schizosaccharomyces pombe is dependent on a large ribonucleoprotein complex of the internal transcribed spacer 1. J Mol Biol 2000; 302:65-77. 98. Lalev AI, Nazar RN. A chaperone for ribosome maturation. J Biol Chem 2001; 276:16655-16659. 99. Kass S, Sollner-Webb B. The first pre-rRNA-processing event occurs in a large complex: analysis by gel retardation, sedimentation, and UV cross-linking. Mol Cell Biol 1990; 10:4920-4931. 100. Mougey EB, O’Reilly M, Osheim Y et al. The terminal balls characteristic of eukaryotic rRNA transcription units in chromatin spreads are rRNA processing complexes. Genes Dev 1993; 7:1609-1619. 101. Kruiswijk T, Planta RJ, Krop JM. The course of the assembly of ribosomal subunits in yeast. Biochimica Biophysica Acta 1978; 517:378-389. 102. Gelperin D, Horton L, Beckman J et al. Bms1p, a novel GTP-binding protein, and the related Tsr1p are required for distinct steps of 40S ribosome biogenesis in yeast. RNA 2001; 7:1268-1283. 103. Schäfer T, Strauss D, Petfalski E et al. The path from nucleolar 90S to cytoplasmic 40S pre-ribosomes. EMBO J 2003; 22:1370-1380. 104. Vanrobays E, Gelugne J-P, Gleizes P-E et al. Late cytoplasmic maturation of the small ribosomal subunit requires RIO proteins in Saccharomyces cerevisiae. Mol Cell Biol 2003; 23:2083-2095. 105. Angermayr M, Bandlow W. Rio1, an extraordinary novel protein kinase. FEBS Lett 2002; 524:31-36. 106. Ito T, Chiba T, Ozawa R et al. A comprehensive two-hybrid analysis to explore the yeast protein interactome. Proc Natl Acad Sci USA. 2001; 98:4569-4574. 106a.Geerlings T, Faber AW, Bister MD et al. Rio2p, an evolutionarily conserved low-abundant protein kinase essential for processing of 20S pre-rRNA in Saccharomyces cerevisiae. J Biol Chem 2003; 278:22537-2254. 107. Wehner KA, Baserga SJ. The sigma(70)-like motif. A eukaryotic RNA binding domain unique to a superfamily of proteins required for ribosome biogenesis. Mol Cell 2002; 9:329-339. 108. Senger B, Lafontaine DLJ, Graindorge J-S et al. The nucle(ol)ar Tif6p and Efl1p are required for a late cytoplasmic step of ribosome synthesis. Mol Cell 2001; 8:1363-1373. 109. Dechampesme A-M, Koroleva O, Leger-Silvestre I et al. Assembly of 5S ribosomal RNA is required at a specific step of the pre-rRNA processing pathway. J Cell Biol 1999; 145:1369-1380. 110. Lafontaine DL, Tollervey D. The function and synthesis of ribosomes. Nat Rev Mol Cell Biol 2001; 2:514-520. 111. Liang W-Q, Clark JA, Fournier MJ. The rRNA-processing function of yeast small nucleolar RNA can be rescued by a conserved RNA helicase-like protein. Mol Cell Biol 1997; 17:4124-4132. 112. Van Hoof A, Lennertz P, Parker R. Three conserved members of the RNase D family have unique and overlapping functions in the processing of 5S, 5.8S, U4, U5, RNase MRP and RNase P RNAs in yeast. EMBO J 2000; 19:1357-1365. 113. Van Hoof A, Parker R. The exosome: A proteasome for RNA? Cell 1999; 99:347-350. 114. De La Cruz J, Kressler D, Linder P. Unwinding RNA in Saccharomyces cerevisiae: DEAD-box proteins and related families. Trends Biochem Sci 1999; 24:192-198.

Pre-Ribosomal RNA Processing and Assembly in Saccharomyces cerevisiae

221

115. Daugeron MC, Linder P. Characterization and mutational analysis of yeast Dbp8p, a putative RNA helicase involved in ribosome biogenesis. Nucl Acids Res 2001; 29:1144-1155. 116. Colley A, Beggs JD, Tollervey D et al. Dhr1p, a putative DEAH-Box RNA helicase, is associated with the box C+D snoRNP U3. Mol Cell Biol 2000; 20:7238-7246. 117. Venema J, Bousquet-Antonelli C, Gelugne J-P et al. Rok1p is a putative RNA helicase required for rRNA processing. Mol Cell Biol 1997; 17:3398-3407. 118. He F, Jacobson A. Identification of a novel component of the nonsense-mediated mRNA decay pathway by use of an interacting protein screen. Genes Dev 1995; 9:437-454. 119. Liang S, Hitomi M, Hu YH et al. A DEAD-box-family protein is required for nucleocytoplasmic transport of yeast mRNA. Mol Cell Biol 1996; 16:5139-5146. 120. Adams CC, Jakovljevic J, Roman J et al. Saccharomyces cerevisiae nucleolar protein Nop7p is necessary for biogenesis of 60S ribosomal subunits. RNA 2002; 8:150-165. 121. De La Cruz J, Kressler D, Tollervey D et al. Dob1p (Mtr4p) is a putative ATP-dependent RNA helicase required for the 3' end formation of 5.8S rRNA in Saccharomyces cerevisiae. EMBO J 1998; 17:1128-1140. 122. Kressler D, Doère M, Rojo M et al. Synthetic lethality with conditional DBP6 alleles identifies Rsa1p, a nucleoplasmic protein involved in the assembly of 60S ribosomal subunits. Mol Cell Biol. 1999; 19:8633-8645. 123. Rout MP, Aitchison JD, Suprapto A et al. The yeast nuclear pore complex: composition, architecture, and transport mechanism. J Cell Biol 2000; 148:635-652. 124. Lafontaine D, Vandenhaute J, Tollervey D. The 18S rRNA dimethylase Dim1p is required for pre-ribosomal RNA processing in yeast. Genes Devel 1995; 9:2470-2481. 125. Kressler D, Rojo M, Linder P et al. Spb1p is a putative methyltransferase required for 60S ribosomal subunit biogenesis in Saccharomyces cerevisiae. Nucl Acids Res 1999; 27:4598-4608. 126. Pintard L, Kressler D, Lapeyre B. Spb1p is a yeast nucleolar protein associated with Nop1p and Nop58p that is able to bind S-adenosyl-L-methionine in vitro. Mol Cell Biol 2000; 20:1370-1381. 127. Sprang SR. G protein mechanisms: Insights from structural analysis. Annu Rev Biochem 1997; 66:639-678. 128. Wegierski T, Billy E, Nasr F et al. Bms1p, a G-domain-containing protein, associates with Rcl1p and is required for 18S rRNA biogenesis in yeast. RNA 2001; 7:1254-1267. 129. Billy E, Wegierski T, Nasr F et al. Rcl1p, the yeast protein similar to the RNA 3'-phosphate cyclase, associates with U3 snoRNP and is required for 18S rRNA biogenesis. EMBO J 2000; 19:2115-2126. 130. Dougan DA, Mogk A, Zeth K et al. AAA+ proteins and substrate recognition, it all depends on their partner in crime. FEBS Lett 2002; 529:6-10. 131. Basu U, Si K, Warner JR et al. The Saccharomyces cerevisiae TIF6 gene encoding translation initiation factor 6 is required for 60S ribosomal subunit biogenesis. Mol Cell Biol 2001; 21:1453-1462. 132. Schaper S, Fromont-Racine M, Linder P et al. A yeast homolog of chromatin assembly factor 1 is involved in early ribosome assembly. Curr Biol 2001; 11:1885-1890. 133. Sasaki T, Toh-e A, Kikuchi K. Yeast Krr1p physically and functionally interacts with a novel essential Kri1p, and both proteins are required for 40S ribosome biogenesis in the nucleolus. Mol Cell Biol 2000; 20:7971-7979. 134. Baudin-Baillieu A, Tollervey D, Cullin C et al. Functional analysis of Rrp7p, an essential yeast protein involved in pre-rRNA processing and ribosome assembly. Mol Cell Biol 1997; 17:5023-5032. 135. Iouk TL, Aitchison JD, Maguire S et al. Rrb1p, a yeast nuclear WD-repeat protein involved in the regulation of ribosome biosynthesis. Mol Cell Biol 2001; 21:1260-1271. 136. Finley D, Bartel B, Varshavsky A. The tails of ubiquitin precursors are ribosomal proteins whose fusion to ubiquitin facilitates ribosome biogenesis. Nature 1989; 338:394-401. 137. Monia BP, Ecker DJ, Finley D et al. A human ubiquitin carboxyl extension protein functions in yeast. J Biol Chem 1990; 265:19356-19361. 138. Nishi R, Hashimoto H, Kidou S et al. Isolation and characterization of a rice cDNA which encodes a ubiquitin protein and a 52 amino acid extension protein. Plant Mol Biol 1993; 22:159-161. 139. Barrio R, Del Arco A, Cabrera HL et al. Structure and expression of the Drosophila ubiquitin-80-amino-acid fusion-protein gene. Biochem J 1994; 302:237-244. 140. Stage-Zimmermann T, Schmidt U, Silver PA. Factors affecting nuclear export of the 60S ribosomal subunit in vivo. Mol Biol Cell.2000; 11:3777-3789. 141. Liu PC, Thiele DJ. Novel stress-responsive genes EMG1 and NOP14 encode conserved, interacting proteins required for 40S ribosome biogenesis. Mol Biol Cell 2001; 12:3644-3657. 142. Milkereit P, Strauss D, Bassler J et al. A Noc complex specifically involved in the formation and nuclear export of ribosomal 40 S subunits. J Biol Chem 2003; 278:4072-4081.

222

The Nucleolus

143. Ho JH-N, Johnson AW. NMD3 encodes an essential cytoplasmic protein required for stable 60S ribosomal subunits in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:2389-2399. 144. Gadal O, Strauß D, Kessl J et al. Nuclear export of 60S ribosomal subunits depends on Xpo1p and requires a Nuclear Export Sequence-containing factor, Nmd3p, that associates with the large subunit protein Rpl10p. Mol Cell Biol 2001; 21:3405-3415. 145. Suzuki N, Noguchi E, Nakashima N et al. The Saccharomyces cerevisiae small GTPase, Gsp1p/ Ran, is involved in 3' processing of 7S-to-5.8S rRNA and in degradation of the excised 5'-A0 fragment of 35S Pre-rRNA, both of which are carried out by the exosome. Genetics 2001; 158:613-625. 146. Dunbar DA, Dragon F, Lee SJ et al. A nucleolar protein related to ribosomal protein L7 is required for an early step in large ribosomal subunit biogenesis. Proc Natl Acad Sci USA 2000; 97:13027-13032. 147. Morita D, Miyoshi K, Matsui Y et al. Rpf2p, an evolutionarily conserved protein, Interacts with ribosomal protein L11 and is essential for the processing of 27SB pre-rRNA to 25S rRNA and the 60S ribosomal subunit assembly in Saccharomyces cerevisiae. J Biol Chem 2002; 277:28780-28786. 148. Venema J, Tollervey D. RRP5 is required for formation of both 18S and 5.8S rRNA in yeast. EMBO J 1996; 15:5701-5714. 149. Torchet C, Jacq C, Hermann-le Denmat S. Two mutant forms of the S1/TPR-containing protein Rrp5p affect the 18S rRNA synthesis in Saccharomyces cerevisiae. RNA 1998; 4:1636-1652. 150. Eppens EA, Rensen S, Granneman S et al. The roles of Rrp5p in the synthesis of yeast 18S and 5.8S rRNA can be functionally and physically separated. RNA 1999; 5:779-793. 151. Torchet C, Hermann-Le Denmat S. High dosage of the small nucleolar RNA snR10 specifically suppresses defects of a yeast rrp5 mutant. Mol. Genet. Genomics 2002; 268:70-80. 152. Ramakrishnan V. Ribosome structure and the mechanism of translation. Cell 2002; 108:557-572. 153. Du YC, Stillman B. Yph1p, an ORC-interacting protein: potential links between cell proliferation control, DNA replication and ribosome biogenesis. Cell 2002; 109:835-848. 154. Lerch-Gaggl A, Haque J, Li J et al. Pescadillo is essential for nucleolar assembly, ribosome biogenesis, and mammalian cell proliferation. J Biol Chem 2002; 277:45347-45355. 155. Hodges PE, McKee AH, Davis BP et al. The Yeast Proteome Database (YPD): a model for the organization and presentation of genome-wide functional data. Nucl Acids Res 1999; 27:69-73. 156. Angermayr M, Roidl A, Bandlow W. Yeast Rio1p is the founding member of a novel subfamily of protein serine kinases involved in the control of cell cycle progression. Mol Microbiol 2002; 44:309-324. 157. Watkins NJ, Segault V, Charpentier B et al. A common core RNP structure shared between the small nucleolar box C/D RNPs and the spliceosomal U4 snRNP. Cell 2000; 103:457-466.

CHAPTER 13

The snoRNPs and Related Machines: Ancient Devices That Mediate Maturation of rRNA and Other RNAs Edouard Bertrand and Maurille J. Fournier

Abstract

I

t has been known for several years that eukaryotic cells contain large populations of small nucleolar RNA-protein complexes, called snoRNPs, and that these complexes mediate the formation of modified nucleotides in rRNA and facilitate cleavage of rRNA precursors. Most snoRNPs fall into two large classes named for the snoRNA component, i.e., box C/D and H/ACA snoRNPs, and most snoRNPs in each family participate in nucleotide modification synthesis of 2’-O-methylated nucleotides in the case of the C/D snoRNPs and pseudouridine in the case of H/ACA snoRNPs. In a remarkable departure from previously described reaction schemes, the modifications are targeted by guide motifs in the snoRNA and the reaction is catalyzed by a core snoRNP protein. Recent results make clear that the period of major discovery is still very much in progress. Novel snoRNA-like guide RNAs have been identified for splicing snRNAs and mRNAs, and a set of snRNA-specific guides has been discovered to reside in the Cajal bodies; the latter species are called scaRNAs and include unusual structural variants of the canonical snoRNAs. Studies of snoRNP biogenesis have characterized the major steps involved in snoRNA maturation, early events in snoRNP assembly, and shown that trafficking of new snoRNPs involves transit through Cajal bodies in animals and structurally related nucleolar bodies in yeast. Archaeal organisms have been determined to contain C/D and H/ACA guide RNAs (called sRNAs) and corresponding core proteins, indicating that the snoRNP machinery is of ancient origin; notably, Archaea contain guide RNAs for tRNA as well as rRNA. Opening the way for in-depth structure and function studies, guided modification has been achieved with archaeal and yeast cell-free systems. These and other major advances are reviewed in the present chapter.

Introduction Nucleoli contain scores of small stable RNAs known as snoRNAs, an acronym for small nucleolar RNAs, and these RNAs exist as snoRNA:protein complexes called snoRNPs (‘snorps’, as reviewed recently).1,2 The snoRNPs function in the maturation of ribosomal RNA (and other RNAs), by: 1) creating two types of modified nucleotides, i.e., 2’-O-methylated nucleotides (Nm) and pseudouridine (Ψ), and, 2) mediating endonucleolytic cleavages of pre-rRNA. Nearly all snoRNPs fall into two large families that are defined by pairs of conserved ‘box’ elements in the snoRNA component; the families are the box C/D and box H/ACA snoRNPs. Most box C/D snoRNPs create Nm modifications and most box H/ACA snoRNPs convert uridine to Ψ. In both types of modification, the target nucleotide is selected through base pairing of the snoRNA to the substrate (guide function), and an integral snoRNP protein catalyzes the reaction. The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

224

The Nucleolus

Cleavage of pre-rRNA requires a small number of box C/D and box H/ACA snoRNPs, and an additional snoRNP, called MRP, which does not belong to either of the box C/D or H/ACA families. The MRP snoRNP catalyzes an endonucleolytic cleavage, whereas the functions of the other processing snoRNPs are still obscure, although they also bind to rRNA. The snoRNP machines are of ancient origin, as clear orthologs are present in Protists and, strikingly, among the Archaea. The archaeal snoRNA-like RNAs are called sRNAs and the corresponding RNPs are known as sRNPs.3,4 In addition to guiding nucleotide modification of rRNA, snoRNAs have been shown to target Nm modification in vertebrate U6 splicing snRNAs, and candidate guide snoRNAs have been discovered that are predicted to modify mRNA, in human brain and trypanosome transcripts.5,6 In a startling recent development, a novel subset of snoRNA-like guide RNAs has been found to reside in the Cajal (coiled) bodies of animal cells; Cajal bodies occur in the nucleoplasm, often in close association with nucleoli.7 The new RNAs, designated scaRNAs (small Cajal body RNAs), are predicted to modify the U1, U2, U4 and U5 snRNAs, which occur transiently in this same compartment.8 In the archaeal kingdom, many candidate Nm guide sRNAs have been identified that are specific for tRNAs, a situation not yet known to occur in eukaryotes.9-11 Several important new possibilities are suggested by these latter results: 1) a variety of non-rRNAs known or suspected to pass through the nucleolus could be substrates for snoRNPs; 2) the range of nuclear substrates could include many more mRNAs than have been implicated to date—as well as other RNAs, for example, the recently discovered tiny, non-coding RNAs thought to occur in vast numbers in eukaryotes;12 3) additional snoRNP-like machines might be discovered to function outside of the nucleolus, and; 4) the function of the snoRNPs may not be limited to modification and processing, but could also include chaperone-like functions in RNA folding or RNP assembly, or quality control. In the present chapter, we will review the current state of knowledge about the snoRNAs and snoRNPs. Readers are also referred to other chapters in this volume on processing of rRNA (Chapters 11,12) and trafficking of other RNAs through the nucleolus (Chapter 17). In addition, several excellent reviews on the present themes are available.1-3,13-18 Literature citations are through November 2002.

Early History

The term snoRNA was coined in 1981,19,20 over 10 years after the first small nuclear RNAs were detected and the first nucleolar species was defined.21,22 The U3 species was the first snoRNA described and also the first to be examined in detail. It was observed on fractionation of small RNA prepared from rat cell nuclei by gel electrophoresis.23,24 Other RNA bands observed in the same analysis included the yet-to-be-recognized splicing snRNAs, and because all of the species detected were rich in uridine, all were given the U-designations in use today,19,20 We now understand that U3 was the most abundant snoRNA in those preparations and that all eukaryotic cells contain scores of snoRNAs, as indicated in early reviews.20,25 U3 is required for the first endonucleolytic cleavages of precursor rRNA26 and appears to be ubiquitous among eukaryotes. Sequence analysis of the first few U3 snoRNAs isolated revealed several conserved ‘box’ sequences (A-D), and related secondary structures.27-33 Box C/D elements were determined to be present in other small nuclear RNAs discovered some years later (1989) and this common feature was the origin of the box C/D nomenclature subsequently used to define one of the two large families of snoRNAs.34 The first efforts to describe the population of nuclear small RNAs in yeast were reported in 1983 and 1988, using different gel electrophoresis fractionation strategies.35-38 Subsequent characterization revealed the presence of splicing snRNAs and snoRNAs. Remarkably, virtually all of the several dozens of non-splicing RNAs detected39 have turned out to be snoRNAs. The large family of box H/ACA snoRNAs was identified in yeast (in 1996), from comparative sequencing of nuclear small RNAs, and identified in humans soon thereafter.39,40 Most snoRNAs

The snoRNPs and Related Machines

225

described are from human and yeast cells, however, snoRNAs have been shown to occur in a variety of other metazoan organisms (e.g., mouse, rat, Xenopus and plants), and in protists as well, early in the history of snoRNA research in some cases. The cell systems featured most heavily in biosynthesis and function studies have been yeast, human, mouse and Xenopus. Excellent progress has also come from characterization of orthologous guide RNAs in archaeal cells.9,41,42

More Recent Landmarks and Breakthroughs Other major developments in the field include: 1) the discovery that snoRNAs can be encoded within introns of protein genes;43 2) evidence that snoRNAs are required for cleavage of pre-rRNA;26,44 3) discovery of the guide functions in nucleotide modification;45-47 4) revelation that the C/D and H/ACA box elements are sufficient for localization to the nucleolus;48-62 5) identification of the core proteins in the C/D and H/ACA families of snoRNPs;63-72 6) the finding that at least some snoRNAs localize to the nucleolus by way of the Cajal bodies;49,58,60,73 7) identification of brain-specific guide snoRNAs predicted to methylate mRNA, and a putative mRNA-specific Ψ guide for Trypanosome mRNA;5,6 8) discovery of snoRNP orthologs in the Archaea;9,41,42,74-82 9) good success in reconstituting core snoRNP complexes;76,77,83-85 10) development of in vitro guided Nm and Ψ modification systems;77,86-88 11) describing the proteins present in the MRP and U3 processing snoRNPs;89,90 12) discovery of snoRNA-like guide RNAs that reside in mammalian Cajal bodies (called scaRNAs), which are expected to guide methylation and pseudouridylation of splicing snRNAs that also occur at this location,8,86 and; 13) excellent progress in defining key steps in snoRNP assembly and trafficking in vivo.58-60,62,91-94 These and other developments are reviewed in the sections that follow, with emphasis on recent findings.

snoRNP Structure and Function Individual snoRNPs consist of a single snoRNA molecule, a small set of stably associated, family-specific ‘core’ proteins and, in some, perhaps all cases, less tightly associated proteins. Additional proteins are known to be unique to particular processing snoRNPs involved in cleavage of pre-rRNA. In no case yet has a snoRNP been described that contains more than one snoRNA, however, remarkable exceptions to the typical snoRNA structure have been found among the snoRNA-like RNAs identified in the Cajal bodies (scaRNAs8,86). This population includes hybrid Nm and Ψ guide RNAs, with targeting motifs for both types of modification, and covalently linked ‘twin’ snoRNAs that target a single type of modification (see below). It seems possible that novel RNA and RNP variants could also occur among the population of more conventional snoRNPs as well; variants are also known to occur in the Archaea (see below). Important progress has been made in characterizing snoRNP proteins, including: the core proteins of the C/D and H/ACA snoRNPs, and the full repertoire of proteins in two specialized snoRNPs involved in rRNA processing, the U3 and MRP snoRNPs.

C/D snoRNPs Box C/D snoRNAs. The snoRNAs in the box C/D family contain at least one set of box C (PuUGAUGA) and box D (UCUGA) elements and commonly contain a second, degenerate pair designated boxes C’ and D’ (Fig. 1).95 Most C/D snoRNAs and snoRNPs function in ribose methylation of rRNA, and quite likely other RNAs in the nucleolar compartment (see below and Chapter 17). A few others are involved in processing (cleavage) of rRNA, and some of these participate in both processing and modification. The canonical C/D boxes are usually near the 5’ and 3’ ends of the mature snoRNA sequence, respectively (the U3 processing snoRNA is an exception), and form a characteristic helix-asymmetric bulge-helix structure.72,96 This motif, which was first identified in the structure of human U4 snRNA, also occurs in rRNA. The classic C/D elements influence several overlapping aspects of snoRNA and snoRNP synthesis and function, including: 1) binding of core snoRNP proteins;15,31,72,76,77,85,100,101

226

The Nucleolus

Figure 1. Structures of the box C/D, box H/ACA and MRP snoRNAs. Consensus structures are shown for the three known classes of small nucleolar RNAs. (A) The snoRNAs in the C/D family have a characteristic C/D folding motif (kink turn, flanked by stem I and stem II) that is usually near the 5’ and 3’ ends (enlarged), and frequently, a second set of similar, but less conserved C and D elements called boxes C’ and D’. Most C/D snoRNAs target 2’-O-methylation of rRNA or other RNAs, using a long (9-21 nt) guide sequence located upstream of box D or D’; methylation occurs at a substrate site 5 nts upstream of box D/D’, in the region of complementarity. A few C/D snoRNAs function in cleavage of pre-rRNA, most likely by facilitating substrate folding. The C/D snoRNPs have four common core proteins (see also Table 1) and one of these (NHX/Snu13p) binds to the universal C/D motif, and to the same motif in U4 snRNA. Methylation is catalyzed by the fibrillarin/Nop1p core protein. (B) Members of the H/ACA snoRNA family have distinguishing stem-bulge-stem folding domains followed by an ANANNA element in a ‘hinge’ region (i.e., H box), and ACA box near the 3’ end. Most H/ACA snoRNAs target pseudouridine formation, using two short (3-10 nt) guide sequences located in the bulge region (enlarged). The uridine to be modified occupies Figure legend continued on next page

The snoRNPs and Related Machines

227

Figure 1 legend continued from last page. a ‘pocket’ between the regions of base pairing and is located 14-16 nts from the trailing H or ACA box. A few H/ACA snoRNAs participate in pre-rRNA cleavage. The H/ACA snoRNPs contain four core proteins and the dyskerin/Cbf5p core protein catalyzes the isomerization reaction. The mammalian Cajal bodies contain variants of C/D and H/ACA snoRNAs, called scaRNAs, with guide sequences for splicing snRNAs. Other snoRNA-like RNAs contain guide sequences for brain mRNAs and spliced leaders of trypanosomes. Both types of guide RNAs occur in Archaeal organisms as well, with specificity for rRNA and tRNA. (C) The MRP snoRNA is structurally related to the RNA component of RNAse P. The upper portion of the RNA structure shown is conserved among eukaryotes, whereas the lower portion is variable. The nine proteins listed occur in the yeast MRP complex, and eight of these also occur in yeast RNase P.

2) providing metabolic stability;49,56,57,102-104 3) defining the 5’ and 3’ ends of the mature snoRNA;49,105,106 4) hypermethylation of monomethylated 5’ caps, where these occur;57,59,107 5) localizing the newly synthesized snoRNA to the nucleolus and Cajal bodies,48-62 and; 6) 2’-O-methylation activity in the case of the Nm guide snoRNAs.45,95 These functions are almost certainly related to binding of the core proteins, and perhaps other, unidentified shared proteins. The only other RNA structural feature known to be common, but not universal to all C/D snoRNAs is the presence of one or two guide sequences that target methylation (Fig. 1). These are located on the 5’-side of box D or D’ (1-2 nts upstream), and consist of long (9-21 nt) sequences that are complementary to the substrate region to be modified.45,95,108-110 Methylation occurs at a substrate site located five (and sometimes six) nucleotides upstream of box D or D’, typically 4-5 nucleotides within the region of complementarity. Some C/D processing snoRNAs also contain guide elements, for example the U14 species, which is universal among eukaryotes, is needed for both processing and methylation of 18S rRNA.111,112 The fact that two guide motifs have been identified in some snoRNAs and only one in others suggests that new, non-rRNA substrates may be found, or that non-guiding elements may be intermediates or stable end-products in snoRNA evolution. In principle, the antisense sequences could also be involved in chaperone-like functions (see below). C/D snoRNP proteins. Four core proteins are stably associated with the C/D snoRNAs (Table 1) and these are believed to be universal among eukaryotes.63-72 Clear orthologs of these proteins also occur among archaeal organisms as well.9,41,74-77,81 The eukaryotic core proteins were first identified as snoRNP components in yeast and humans, and are usually referred to in other organisms (including the Archaea) by the human designations. The yeast (and human) proteins include: Snu13p (NHPX, also 15.5 kDa); Nop1p (fibrillarin); Nop56p (hNop56p), and; Nop58p (hNop58p). Interestingly, Snu13p/NHPX also occurs in U4 snRNP complexes and in both types of RNPs interacts directly with a K-turn folding motif.61,72,96-98 Nop56p and Nop58p are closely related, exhibiting 41% identify in yeast and 37% identity in humans, and the single archaeal variant (Nop56/58p) is closely related to both, suggesting all have a common ancestor.67,70,75 Another distinction for the Archaea is that the counterpart of Snu13p/NHPX is a ribosomal protein (aL7a).3,76,77 Nop1p/fibrillarin is widely accepted to catalyze the methylation reaction. Although this has not yet been demonstrated directly, the case is compelling. Key evidence includes: 1) the presence of four short sequence elements conserved among several types of methylases that utilize S-adenosylmethionine as the methyl donor;113,114 2) results showing that point mutations in the motif-containing region of yeast Nop1p can block formation of Nm modifications in rRNA in a global manner,114,115 and; 3) remarkable similarity of the 3-D structures of an archaeal ortholog of Nop1p (from M. jannaschii) and several methylases, over most of the length of the protein.114 Notably, the human autoimmune disease, scleroderma, is characterized by autoantibodies targeted to fibrillarin.116 Information about the assembly and interactions of the core proteins is provided below, in a section devoted to snoRNP assembly. Properties of the C/D snoRNPs. Beyond the four core proteins, the actual numbers and types of proteins in the natural C/D snoRNPs are still obscure. It seems likely that the subset of

228

The Nucleolus

Table 1. Core snoRNP proteins

processing and modifying snoRNPs have different compositions, reflecting their different, specialized functions, however, information on this point is also sparse. Nor is it known to what extent the individual processing or modifying snoRNPs are related at the protein level. Most information on these issues comes from characterization of the U3 box C/D snoRNP, and the MRP snoRNP (see below), which play highly specialized roles early in rRNA processing.

The snoRNPs and Related Machines

229

Early genetic and immunological studies of U3 snoRNPs identified nearly a dozen U3-associated proteins unique to the yeast or human particle, as cited elsewhere.90 Those findings were extended very substantially by recent direct analysis of yeast U3 complexes, isolated by an affinity purification procedure.90 More than 30 proteins associated with the snoRNP complex were detected and identified by microsequencing. In addition to the four C/D core proteins, 23 of 24 other proteins thought to be specific were found to be essential for growth and 18S rRNA production. It remains to be seen if the list of proteins is complete; three proteins that had been identified by others as U3-associated were not observed. Conversely, some proteins may not be integral snoRNP proteins, but are part of the pre-rRNP maturation complex. Key fractionation steps included enrichment of C/D snoRNPs by selection of a tagged C/D core protein, followed by selection of a tagged U3-specific protein.90

H/ACA snoRNPs Box H/ACA snoRNAs. Similar to the C/D snoRNAs, a few H/ACA snoRNAs are involved in processing of rRNA, however, most guide Ψ formation, and cases exist where a snoRNA has both functions. The RNAs in this family are distinguished by: 1) two types of short box elements that can vary in sequence, but are positionally conserved in the secondary structure, and; 2) a consensus secondary structure consisting of two characteristic helix-bulge-helix domains separated by a hinge region Figure 1.39,40 The signature H (hinge) and ACA boxes occur downstream of the distinguishing folding domains. Because of the variable nature of the box elements, secondary structure analysis is an important aid in identifying snoRNAs in this class. Another complication in identifying these RNAs is the fact that the secondary structures are not limited to the consensus features and other folding domains can be present as well; this is a substantial complication for the yeast snoRNAs, which are typically much larger than those from vertebrates and other sources. The H/ACA boxes influence synthesis and function of both the snoRNAs and corresponding snoRNPs, as is the case for the C/D boxes. In particular, the H/ACA boxes are, variously, involved in: 1) providing metabolic stability;39,40 2) defining the ends of the mature snoRNA (ACA);39,40 3) localization to the nucleolus and Cajal bodies,53,54 and; 4) pseudouridylation, in the case of the Ψ guide RNAs.46,117 These effects are believed to be a consequence of protein binding, in particular the core proteins, but perhaps others are involved as well. The Ψ guide snoRNAs contain pairs of short guide sequences (3-10 nts) in the bulge region of one or both folded domains.47 Site selection involves: 1) base pairing of the two guide sequences with substrate nucleotides that flank the uridine to be isomerized, and; 2) a distance measurement (14-16 nts) between the target uridine and corresponding H or ACA box element.46,47 As for the methylation guide snoRNAs, a single Ψ guide snoRNA can use one or both of the guide domains to target modification. Time may reveal the actual modification capacity to be larger than currently understood, and that the snoRNAs have additional functions. H/ACA snoRNP proteins. Four core proteins have also been identified for this family of snoRNPs and, these too are thought to be common to all family members (Table 1). The yeast (and human) proteins include: Cbf5p (dyskerin), Gar1p (hGar1p), Nhp2p (hNhp2p), and Nop10p (hNop10p).118-120 Catalysis of Ψ formation is almost certainly mediated by Cbf5p/ dyskerin based on: 1) the presence of three signature sequence elements conserved among known Ψ synthases,121 and; 2) global disruption of Ψ synthesis in yeast rRNA when point mutations were introduced into these elements in Cbf5p.122 Properties of the H/ACA snoRNPs. Among the H/ACA snoRNP proteins, Gar1 and Nhp2p are known to interact directly with H/ACA snoRNAs.83,123 The Nhp2p protein has the interesting property of being related to the C/D snoRNP protein Snu13p/15.5 kDa, which binds to the C/D motif.72 The sequences of the two yeast proteins are 38% identical and 61% similar, which could reflect the presence of common binding domains. No K-turn has yet been identified in any H/ACA snoRNA, however, similar motifs occur in Archaeal H/ACA sRNAs (see

230

The Nucleolus

below), and are recognized by the archaeal L7 protein, which also binds to the K-turn of C/D sRNAs.3,4,124 Interactions between the H/ACA proteins and snoRNA are described below. The first attempt to isolate and characterize an individual H/ACA snoRNP was made with a yeast complex required for rRNA processing (snR30); the resulting particle contained seven stably associated proteins.125 Examining the gel pattern retrospectively, it appears that all four of the known core proteins were present, plus three others. Additional proteins may exist in the natural snoRNP, which has a sedimentation coefficient of about 12S.119 Electron micrographs of the isolated complex revealed a V-shaped particle, which is consistent with the bipartite consensus structure of the H/ACA snoRNAs. The dimensions of the particle were estimated at 15 nm long and 12 nm wide.119

The MRP snoRNP Complexes containing this RNA were first reported to function in processing of RNA primers involved in mitochondrial DNA replication, and this is the origin of the MRP designation (mitochondrial RNA-processing). Subsequently, the RNA was determined to occur in the nucleolus 126-128 and to be required for cleavage of pre-rRNA, in the ITS-1 segment, upstream of 5.8S rRNA44,129-131 (also see Chapters 11,12). The MRP snoRNA lacks the box elements that distinguish the two major classes of snoRNAs, however, its secondary structure is remarkably similar to that of RNase P RNA, inferring that it may be a ribozyme.132-134 This view was strengthened with the discovery that many of the proteins in the yeast MRP and RNase P enzyme complexes are the same, and that an isolated yeast MRP particle is able to cleave pre-rRNA in vitro, in a site-specific manner.135 Although ribozyme activity has not been demonstrated directly, nine proteins have been identified in the yeast RNase P complex and eight of these are present in yeast MRP as well.89,136-143 Interestingly, RNase P also occurs in the nucleolus, where it mediates processing of tRNA.128,144-148 The MRP snoRNA has a K-turn,98 but the signficance of this is not yet known. The MRP snoRNP has been linked to cell cycle regulation,149 and implicated as the causal agent in cartilage-hair hypoplasia.150,151 Autoantigens from certain scleroderma patients recognize MRP snoRNP proteins.152,153

What Roles Do the C/D and H/ACA snoRNPs Play in rRNA Cleavage? It’s clear that several snoRNAs in addition to MRP are required for cleavage of rRNA transcripts, and most have been shown to interact directly with rRNA through base pairing (see Chapters 11,12). The roles of these other snoRNPs are still obscure, but possible functions include: 1) recruiting a nuclease and perhaps other factors to a cleavage site, and; 2) serving as a chaperone to organize pre-rRNA for cleavage, by either a self-cleaving mechanism or a trans-acting nuclease that does not associate with a snoRNP. The processing functions of the snoRNPs were identified first, in both animal and yeast cells, and all but one of the four yeast processing snoRNAs (U3, U14, snR30, snR10) are essential for growth; the exception is snR10.154 In addition to MRP, only two other processing snoRNPs are believed to be universal among eukaryotes, the U3 and U14 complexes (both are C/D species), and it remains to be seen if these occur among the Archaea. U3 and U14 are both required for cleavages that occur early in processing, and certain cleavages require both snoRNPs (and others), arguing that one or both play organizational rather than catalytic roles;111,155,156 as reviewed elsewhere.154 Two other processing snoRNPs, U8 and U22, are common to vertebrates,157 but homologs have not yet been found in yeast. The U8 snoRNP is required for cleavages of precursor 5.8S-28S rRNA and activity appears to involve regulation of pre-rRNA folding, through binding with the U8 snoRNA.158 The U3 snoRNP is believed to be the first to bind to a nascent pre-rRNA transcript, near the 5’ end, and is thought to play a critical organizational role in forming an active processing complex90,159,160 (see also Chapters 11,12). The snoRNA is more abundant and considerably

The snoRNPs and Related Machines

231

larger than most other snoRNAs and has an atypical structure; in this last regard, box C is located deep in the sequence, rather than at the 5’ end as with other C/D snoRNAs.31,49,52,53,55 As noted above, the U3 snoRNP contains many more proteins than other characterized snoRNPs and most appear to be unique to the U3 complex.90 Progress with in vitro processing systems that feature snoRNPs is still at an early stage. Key advances include demonstrations showing that processing is disrupted in animal cell extracts when several individual snoRNAs (U3, U8, U13, U14, U17/E1, E2, E3) are depleted by RNase H or immunoprecipitation,26,156 and that processing by the yeast MRP snoRNP can be achieved in a simple reaction mix containing in vitro transcribed yeast rRNA and affinity purified MRP complex.135

Modification of rRNA and U6 snRNA by snoRNPs The possibility that snoRNAs may be involved in targeting nucleotide modification was raised when long sequence elements complementary to rRNA were identified in several C/D snoRNAs.25,108,161 Chronologically, the methylation guide function of the C/D snoRNAs was discovered initially, 45,109,162-164 and then, the Ψ guide function of the H/ACA snoRNAs.46,47,162,165 It seems that the populations of the large C/D and H/ACA families are roughly comparable in size and that nearly all participate in modification. In yeast, nearly all guide snoRNAs examined - and thus, snoRNPs, are dispensable; the exceptions are dual function snoRNAs that participate in both processing and modification, i.e., the box C/D species U14 and the H/ACA species snR10. There is no evidence that snoRNAs play a role in methylation of RNA bases, which is the third major category of nucleotide modification in rRNA and other RNAs. Several factors argue that most Nm and Ψ modifications in rRNA are formed by snoRNPs. These include: 1) the solid impression that the snoRNA populations in yeast and humans are sufficiently large to accommodate the number of modifications known to occur—knowing that many snoRNAs target two modifications, and; 2) identification of a set of C/D guide snoRNAs in yeast by a computational approach, that accounts for nearly all Nm modifications in rRNA.110 The number of Nm modifications in yeast and human rRNA are 55 and 105 respectively, and the number of Ψ modifications are 44 and 91.166-168 For yeast, 20 guide snoRNAs have been identified experimentally that are able to serve 27 sites of Ψ modification.169 While in principle, all Nm and Ψ modifications in cytoplasmic rRNA could be formed by snoRNPs, exceptions can also reasonably be expected. In a surprising exception at the time, snoRNPs were shown to also be capable of methylating the human and S. pombe U6 splicing snRNAs.170-172 The evidence included identification of snoRNAs with U6 guide sequences170-172 and demonstrating that one such snoRNA services both U6 and rRNA from the same guide element.170 In addition, modification was observed when a U6 sequence was expressed in animal cells, as part of an rRNA minigene, and in yeast cells when an animal C/D guide RNA was targeted to yeast U6.171 Important progress has occurred recently in developing cell-free systems to characterize the RNA-guided modification processes. The first breakthrough was establishing a successful Nm modification system from reconstituted archaeal components.4,77 The complex was assembled at elevated temperature, from an in vitro transcribed guide RNA and recombinant C/D core proteins, in the presence of a fragment of rRNA. The substrate rRNA was methylated at the expected site with excellent efficiency, also at elevated temperature. Using an alternative approach, successful in vitro Nm and Ψ modification systems have been established with affinity-enriched natural snoRNPs from yeast. Success was achieved initially for Nm modification of an rRNA fragment with a preparation of bulk C/D snoRNPs,88 and then for formation of Ψ in a fragment of rRNA with bulk H/ACA snoRNPs.87 In both cases, modification occurred at the expected natural site. These early achievements demonstrate that characterization of the guided modification processes can now be pursued with both natural and reconstituted RNPs, important advances in the field.

232

The Nucleolus

Effects of the Nm and Ψ Modifications Addition of Nm and Ψ modifications is known to alter the folding properties of RNA, in addition to the sequence. As a generalization, both types of modification enhance structural stability in a localized way, by reducing conformational flexibility.173 Methylation blocks H-bonding potential of the ribose and also protects against hydrolysis of the internucleotide bond. Hydrogen bonding potential is also altered in Ψ formation, due to gain of an additional H-donor. Thus, depending on location, individual modifications or clusters of modifications have the potential to affect a wide variety of activities of the product RNA. Changes can be expected in rates and order of RNA folding, conformational stability of individual folding domains, and in the activity of the final RNA. Modifications in rRNA can hypothetically affect any stage of ribosome synthesis, ribosome activity, and turnover of the ribosome. Similar types of effects can be expected for other RNAs containing these modifications. The availability of ribosome crystal structures has made it possible to place the known nucleotide modifications in rRNA in a three-dimensional context.174-176 The 3-D maps show that ribosome regions known or predicted to be important for function are rich in modifications. In yeast, substantial numbers of modifications also occur in regions with no known direct function in translation, and the same situation is predicted for the prokaryotic archaeon Pyrococcus horikoshii.4,174 Taken together, the patterns are consistent with modifications affecting different aspects of ribosome synthesis and activity. Many individual Nm or Ψ modifications have been blocked in yeast by depleting cells of individual guide snoRNAs, and with no strong affect on growth. In one study, depletion of multiple Ψs in the reaction center region of the large subunit showed synergistic, negative effects on growth, and blocking formation of a single Ψ in the A-loop caused the rate of in vivo translation to drop by 20%.177 Results from the various depletion analyses and previous studies in E. coli 176 argue that most single modifications have small, positive effects, but the full repertoire is highly beneficial. That view is supported by findings showing that disrupting Nm or Ψ modification in a global manner, with enzyme point mutations, have very strong, negative effects on cell growth rate.115,122

Do Modifying snoRNPs Influence Other Aspects of Ribosome Synthesis? At the simplest level, the snoRNPs that mediate modification can be considered nano-scale machines that create modified nucleosides and have no other effect on the target RNA. However, it is also reasonable to consider that modifying snoRNPs might affect RNA folding or RNP assembly as well, either directly or indirectly (Fig. 2). This notion was strengthened substantially following discovery of the first several C/D snoRNAs with antisense sequences, which were later determined to guide modification. It was suggested that such snoRNAs might function as chaperones in ribosome synthesis, to bring order to the complicated processes of RNA folding and assembly of ribosomal subunits.25,108,161 Such effects could be relevant for only a few snoRNPs or perhaps involve a large number. This hypothesis remains very interesting and attractive, however, firm, supporting evidence is still lacking. Continuing in this vein, a chaperone function by a snoRNP could, in principle, be the dominant basis for selecting and maintaining a modifying snoRNP in a cell, in a situation where the resulting modification may or may not provide any benefit. A conserved snoRNP (U14) that mediates both processing and methylation has been implicated in rRNA folding, through potentially concomitant binding at two distant segments in 18S rRNA.178 However, it is too soon to know if this or any other modifying snoRNP has effects beyond its modification function. For snoRNPs that mediate processing, but not modification, the evidence for roles in rRNA folding is stronger, especially for the U3 and U8 species (see also Chapters 11,12).158,179-185

The snoRNPs and Related Machines

233

Figure 2. Do modifying snoRNPs influence rRNA folding and ribosome assembly? In principle, binding and release of modifying snoRNPs can promote or block specific pre-rRNA folding events, and fine-tune protein binding processes. Some snoRNPs or modifications could also function in quality control of ribosome synthesis.

When Do snoRNPs Act? Early studies of rRNA modification show that unprocessed precursors have a high content of 2'-O-methylations, indicating that most modification reactions occur before cleavage; this situation was observed for yeast186-188 and animal cells189-192 as reviewed previously.168 However, it is not known if modification occurs during or after transcription. The details of this situation promise to be very interesting, as large numbers of ribosomal and non-ribosomal proteins also associate with pre-rRNA at an early stage, and binding and release of snoRNPs must be coordinated with the rRNA folding and assembly events. Good progress is being made in characterizing pre-rRNP intermediates from yeast160,193-197 and from animal cells.198 Results from protein composition analyses with yeast complexes suggest that assembly of the precursors to the small and large subunits occurs in rather distinct stages, prior to processing. From the patterns, it appears that a quite-mature form of the small subunit is created while assembly of the large subunit is still at a very early stage,160 as reviewed elsewhere.199 Thus far, only the U3 snoRNP has been detected in the pre-ribosomal RNP complexes analyzed, in 90S and 60S yeast complexes (with variable yields in the latter), which either reflects low abundance or absence of the other snoRNPs from these intermediates.160,194,197 Detection of other snoRNPs may be a problem in any case, if the half-life of a snoRNP:rRNA complex is short, or the interactions are not biochemically stable. Because the total number of snoRNPs involved in modifying and processing rRNA is very substantial—probably at least 60-70 for yeast and twice as many for humans, it would seem simplest mechanistically, for most modifying snoRNPs to bind and act early during transcription, while ribosome assembly is at a very early stage. Post-transcriptional binding scenarios can be imagined as well, where snoRNP binding sites in pre-rRNA are still exposed or, less attractive, made accessible through large scale remodeling of the pre-rRNP complexes (Fig. 3). The case for snoRNPs acting early is stronger yet when the relative volumes of the snoRNPs and a fully condensed ribosomal subunit are considered. Estimates of snoRNP size are still at an early stage, but the aforementioned electron microscopic analysis of the snR30 (processing) and snR42 (modifying) snoRNPs from yeast indicate these complexes have substantial volume. The dimensions of the bipartite structures translate into a volume that is roughly 15% of the 80S ribosome (6 x 105 Å3 vs. 4.3 x 106Å3, based on different methods).119,200 Even if a generic modifying snoRNP is somewhat smaller, as expected, it seems likely that snoRNPs complete their tasks before the pre-rRNP substrate is very much condensed, either during or after transcription; condensation of the individual precursor rRNPs could occur after the snoRNPs have acted. In this context, snoRNPs or the resulting modified nucleotides could play roles in staging assembly events, and a subset of snoRNPs or rRNA modifications might

234

The Nucleolus

Figure 3. When do snoRNPs act? Binding of processing and modifying snoRNPs is coordinated with rRNA folding, binding of ribosomal and non-ribosomal proteins and formation of precursor ribosomal subunit complexes. The U3 processing snoRNP binds near the 5’ end of pre-rRNA during transcription, and plays a key organizational role in forming a pre-rRNP complex. The MRP snoRNP cleaves pre-rRNA directly. Other processing snoRNPs might have chaperone-like roles in folding of pre-rRNA, perhaps for self-cleavage or to provide platforms for assembly of cleaving machinery. Most, and possibly all of the nucleotide modification reactions mediated by snoRNPs are thought to occur before pre-rRNA cleavages commence. Subunit assembly may take place in distinct stages, with small subunit complexes developing before large subunit rRNPs. Most likely, the scores of modifying snoRNPs act before the individual pre-subunit complexes become highly condensed. These snoRNPs may bind, act and be released rapidly, or a time-lag may exist between binding, and the reaction and release processes. Three disparate schemes depicting when modifying snoRNPs might act are shown; each assumes that snoRNPs bind during transcription, but binding at later times is possible: a) binding and reaction occur as soon as rRNA binding sequences become available; b) snoRNPs act first on small subunit pre-rRNPs, before or after transcription is completed, and then on large subunit pre-rRNPs, and; c) snoRNPs act simultaneously on small and large subunit pre-rRNPs, after transcription is complete. The snoRNPs are large relative to the mature subunit complexes, arguing that most act early in the assembly processes. Other, more complex mechanisms are also possible.

function in quality control. Some protein-only rRNA modifying enzymes in yeast and E. coli have already been reported to have essential, non-modifying functions, i.e., the protein, but not the modification is important; the vital functions have not yet been defined201-203(and T. Mason person. commun).

Related RNAs and RNPs A growing number of snoRNA-related RNAs exist that have striking similarities and striking differences to the snoRNAs that function in eukaryotic rRNA synthesis. The related species include: 1) C/D and H/ACA guide RNAs for rRNA and tRNA in the Archaeal kingdom,9,41,42 2) familiar and new guide RNA forms that reside in Cajal bodies,8 3) telomerase RNA which has features of H/ACA snoRNAs,204 and; 4) guide RNAs predicted to target mRNAs.5,6

Archaeal Guide RNAs Hints of the presence of box C/D snoRNA homologs in Archaea have been available since the discoveries that rRNA of some Archaeal species contain numerous Nm modifications205,206 and that archaeal genomes have clear orthologs of fibrillarin and Nop56p/Nop58p.9,68,74,207,208 More recently, modification guide RNAs were described in Archaea. Immunoprecipitation and cloning of archaeal RNA associated with fibrillarin and Nop56p/58p orthologs,9 as well as computer-assisted genomic searches,9,41,42 revealed the existence of scores of box C/D-like small RNAs. Evidence that these RNAs are actual guide RNAs came initially from documentation of Nm modifications at the expected sites in natural rRNA, and subsequently in tRNA as well9,10,41 and reviewed elsewhere.3,4,11 Because Archaea are thought to lack nucleoli or an equivalent structure, these RNAs are not formally snoRNAs and were thus called sRNAs.9

The snoRNPs and Related Machines

235

Box C/D sRNAs appear to be widespread among archaeal species, having been detected in the two phyla, Crenarchaeota and Euryarchaeota, and in a variety of species within these phyla.9,11 It thus appears that the box C/D RNA family is extremely ancient, predating the appearence of splicing snRNAs, and may be as ancient as the tRNAs and RNAse P; in addition, this RNA family appears to be very highly conserved. The sequences of the box elements and the general fold of the RNA are nearly identical between Archaea and Eukarya, and archaeal box C/D sRNAs are assembled properly and can guide methylation when introduced into Xenopus oocytes, despite separation of approximately 2 billions years between the two organisms 209. However, systematic comparison of the C/D sRNAs and snoRNAs has also revealed some characteristic features for each family.3,9,11,41,42 First, the eukaryotic RNAs are, in general, larger than their archaeal counterparts, with sizes of about 75 and 100 nucleotides in yeast and mammals, respectively, compared to an average of 50-60 nts for Archaea. Second, only about 20% of eukaryotic snoRNAs guide ribose methylation from both boxes D and D’, and in these cases the substrates are often different rRNA molecules. In contrast, the majority of archaeal RNAs appear to direct rRNA methylation from both box D and box D’, and this often occurs at nearby sites in the same rRNA molecule. Interestingly, the number of C/D sRNAs among different species increases with optimal growth temperature,9 consistent with the notion that C/D RNPs might act as chaperones to help RNA folding and that the general role of the modifications is to stabilize active rRNA structures. Additional support for a stabilizing role for the modifications (or sRNPs) comes from a study showing that the content of Nm nucleotides in the 16S rRNA of one archaeal species (Sulfolobus solfactaricus) increases with growth temperature.206 Two other aspects of the archaeal C/D guide RNAs are particularly interesting. Strikingly, the C/D guide motif can occur within tRNA introns,3,9,11,41,42 demonstrating that the link between snoRNA expression and introns is very ancient. In addition, the occurrence of archaeal C/D guides for tRNA indicates that guide RNAs target more than one type of RNA in ancient organisms. The fact that no tRNA modification guides have yet been found in eukaryotes suggests that the repertoire of RNAs altered by guided modification may have been larger in the past. Remarkably, nucleotide modification by an intronic sRNA (sR50) appears to occur in cis,10 meaning that the sRNA sequences target 2'-O-methylation of the pre-tRNA while embedded within it! This unusual situation suggests that some C/D RNAs may have evolved from tRNA introns, perhaps as a machine that modifies tRNA or helps tRNA folding. In another interesting case, a C/D RNA was found in the non-coding region of precursor rRNA,210 suggesting that it may act in cis in rRNA synthesis. Thus, box C/D RNAs could have evolved from spacer sequences in ribosomal genes, consistent with the fact that archaeal C/D sRNAs utilize a ribosomal protein, L7a, as a core component of the RNP.76,77 Very recently, archaeal orthologs of H/ACA snoRNAs were also described.42 It seems likely that these assemble with proteins similar to the eukaryotic counterparts, as orthologs of Nop10p, Gar1p and Cbf5p also occur in the Archaea.82 The content of pseudouridines in various archaeal rRNAs is low, and much less abundant than in eukaryotes.176,211 Consistent with this situation, only a few archaeal H/ACA sRNAs have been identified thus far (four, in Archaeoglobus fulgidus).42 These RNAs appear to be true modification guides, as well, since Ψ occurs at the predicted positions in rRNA. The structures of the H/ACA sRNAs can differ significantly from those of the canonical eukaryotic guides, in some cases containing a single hairpin ending with the ACA triplet and no box H domain, or three guide-like motifs.42 It should be noted that the single-hairpin cases are not unprecedented as similar guide RNAs also occur in Trypanosomes, an early-branching protist.212 In an interesting development, the archaeal L7 protein, which binds to the K-turn in C/D sRNAs, has been shown to bind a similar motif in H/ ACA sRNAs, located a short distance above the pseudouridylation pocket.124 Taken together, the discoveries that both the C/D and H/ACA RNAs predate the split between Archaea and Eukarya, and that both show intimate relationships with tRNA and rRNA maturation, suggests that they may have originated from extremely ancient organisms, and may have facilitated RNA biogenesis as the RNA World evolved.

236

The Nucleolus

Figure 4. Structures of snoRNA-related RNAs. (A) Schematic of U85, the founding member of the scaRNA family. (B) Secondary structure of vertebrate telomerase RNA. Conserved regions are in grey, and conserved nucleotides appear as white dots. The pseudoknot region is critical for reverse transcriptase binding, and also provides the enzyme template sequence. (C) Brain-specific snoRNAs. The genomic organization of the human 15q11q13 locus is depicted (top). IC: Imprinting center. Paternally transcribed genes are depicted as a black box, maternal ones with a white box. Each thin blue bar corresponds to a snoRNA, and the total number of repetitions is indicated. Bottom: potential base-pairing between HBII-52 and HT2c mRNA. Edited nucleotides are identified with a blue arrowhead, and the base-pair at the site of 2’-O-methylation is shaded in grey.

Small Cajal Body RNAs (scaRNAs) Completion of the human genome sequence has greatly facilitated ‘RNomics’ studies aimed at characterizing the non-coding RNAs in cells. One of the outcomes of these studies has been the description of a novel family of modification guide RNAs.8 The first member of this family, U85,86 has several unique features (Fig. 4): 1) it is significantly longer than other vertebrate snoRNAs (330 nts versus an average of 100); 2) it contains the motifs characteristic of both the box C/D and H/ACA families—its structure corresponds to an H/ACA snoRNA embedded within a C/D snoRNA(!), and; 3) most importantly, U85 possesses guide sequences that are not complementary to rRNA, but to U5 snRNA. As predicted, an in vitro experiment showed that U85 can, indeed, target Ψ formation in U5.86 It was later determined that U85 is the founding member of a novel family of guide RNAs involved in nucleotide modification of the snRNAs transcribed by RNA polymerase II.8 While some of these RNA are not chimerae, and have consensus structures of C/D or H/ACA snoRNAs, they have a second important property in common: they are absent from nucleoli, but concentrated in Cajal bodies. Hence, they were renamed scaRNAs for small Cajal body specific RNAs. The location of these RNAs argues that they mediate their functions in the Cajal bodies as well. Indeed, Cajal bodies have been known for several years to accumulate high levels of

The snoRNPs and Related Machines

237

snRNAs.213-217 More recent studies showed that snRNPs localize in Cajal bodies in the process of re-entering the nucleus following cytoplasmic assembly, and before distributing to speckles.218 This trafficking pattern is in line with recent data suggesting that snRNAs undergo nucleotide modification following nuclear re-import.219 The U6 snRNA is an exception to this scheme of snRNA modification, as it is not exported to the cytoplasm, but instead localizes transiently to nucleoli.56,220 Accordingly, the aforementioned modification guides for this RNA are bona-fide snoRNAs, with a predominant nucleolar localization at steady-state.170,171 The splicing snRNAs of S. cerevisiae constitute a second exception to this rule. Cajal bodies do not exist in this organism and modification of the snRNAs appears to be catalyzed by protein-only enzymes.221 Furthermore, the snRNAs of this yeast have been suggested to transit through the nucleolus to undergo maturation, instead of passing through the cytoplasm.59 While some modifications could be guided by snoRNAs, none have been reported to date.

Telomerase RNA As noted above, the RNA component of vertebrate telomerase RNP belongs to the family of H/ACA snoRNAs.204 This RNA provides docking sites for the reverse transcriptase and its substrate sequence. It is localized in the nucleolus53,204,222 and can be immunoprecipitated with antibodies against the core proteins of the H/ACA snoRNPs.84,223,224 These snoRNA properties appear to be peculiar to vertebrate telomerase, since yeast telomerase RNA belongs to the family of snRNAs and binds Sm proteins.225 Vertebrate telomerase RNA does not appear to catalyze nucleotide modification of rRNA or other RNA, and its secondary structure is slightly divergent from the classical H/ACA snoRNAs226,227 (Fig. 4). These discoveries have a number of important implications. First, they show that the C/D or H/ACA motifs can be utilized as stability and localization elements, distinct from any role in RNA maturation. Second, they indicate that even though the RNP is localized in the nucleolus at steady-state, it likely performs its functions elsewhere in the nucleus, since telomeres are believed to be elongated in the nucleoplasm. Third, these findings also suggest that localization of the RNA is regulated, possibly during the cell cycle. Consistent with this possibility, a recent localization analysis of telomerase reverse transcriptase in human nuclei showed that in primary cells the enzyme is localized in nucleoli except at the time of telomere replication.228 In addition, in transformed cells, or in primary cells expressing large T antigen, the reverse transcriptase is delocalized to the nucleoplasm, but can become re-routed to nucleoli in response to double-strand DNA breaks, most likely to block telomere elongation at these sites.228 While the absence of detailed information on the localization of telomerase RNA is clearly a gap that needs to be filled, the results of these early studies reveal a number of new possibilities for snoRNA functions in the nucleus. Mutations in human dyskerin and in telomerase RNA itself have been correlated with a disease known as Dyskeratosis congenita.223,229,230 Symptoms of the disease include premature aging of rapidly dividing cells, early death from bone marrow failure, increased susceptibility to cancer and other problems. The actual basis of the disease is not yet known, but in principle could reflect defects in ribosome biogenesis, in telomerase function, or both.223,229-235 Results from one study indicate that loss of telomerase function is not the primary defect in mice deficient in dyskerin, as cellular changes associated with the disease are observed well before defects in telomerase function are detected.235 These results argue that not only impaired telomerase function, but also impaired formation of Ψ in rRNA may cause the disorder, although other effects related to dyskerin function may also be possible.

Tissue-Specific, Imprinted snoRNAs Another startling outcome of recent RNomics approaches came from characterization of brain-specific snoRNAs.5,236 While many of the snoRNAs that were cloned by this approach fall into the canonical families of modification guide snoRNAs,237 some clearly do not guide

238

The Nucleolus

modification of either rRNA or snRNA, as they lack sequence complementarity with those RNAs. Further analyses revealed that a few have a unique feature that is in contrast to all snoRNAs characterized to date: these RNAs are expressed only in a few tissues, and mostly in the brain or particular areas of the brain.5 For instance, one mouse RNA (MBI-36) is mostly expressed in the chorioid plexus. To date, brain-specific snoRNAs have been characterized mainly at two loci in mammals: 14q32 and 15q11q13 in human, and at syntenic loci in mice and rats.5,236 The 15q11q13 locus is especially interesting because the snoRNA genes are located in a region that is often mutated in Prader-Willi and Angelman syndromes, disorders that are characterized by developmental, behavioral and mental problems. This region expresses 6 brain-specific box C/D snoRNAs that are conserved between mammals,5 and two of them are repeated 27 (HBII-85) and 47 times (HBII-52). All these snoRNAs are expressed from the introns of a gigantic primary transcript (Fig. 4), which spans more than 460 kb! A similar genomic organization is also found at the second locus. At the rat locus syntenic to 14q32, a gigantic primary transcript is produced that contains an intron-exon unit that is repeated about a hundred times, and the intron contains a box C/D snoRNA (RBII-36).236 The syntenic mouse and human loci also contain repeated box C/D snoRNAs, but these are quite divergent from RBII-36.238 Interestingly, the putative guide-sequence of RBII-36 is not very well conserved between the repeats, suggesting lack of strong selective pressure on these elements. This may be related to another unique characteristic of these snoRNAs, the imprinted state of their genes. Transcription of the snoRNAs at 14q32 occurs only from the maternal allele, while that of 15q11q13 occurs from the paternal one.5,238,239 Non-coding RNAs are often found in imprinted regions,240,241 and in some cases have been demonstrated to play a direct role in the establishment or maintenance of the imprinted state of neighboring protein genes.242-244 Thus, this situation suggests that brain-specific snoRNAs may play a role in the imprinting process. Large amounts of RBII-36 snoRNA precursors accumulate at the transcription site (J. Cavaillé, personal communication), and it’s possible that these transcripts provide docking sites for chromatin remodeling factors that could repress expression of the neighboring genes. Remarkably, some of these brain-specific snoRNAs have the potential to target methylation to certain mRNAs that are also expressed in a tissue specific manner. Indeed, the guide sequence of one (MBII-52) is complementary to 18 nucleotides in 5-HT2c mRNA (Fig. 4)5, an isoform of the serotonin receptor mRNA that is highly expressed in the chorioid plexus. While this correlation may be fortuitous, several facts suggest that the complementarity is functionally significant.5 First, the guide sequence of the snoRNA and its putative target sequence in HT2c mRNA is conserved between human and mouse. Second, this sequence is located just a few nucleotides upstream of an alternative splice site in the HT2c mRNA, in a region that is heavily edited in vivo. In fact, the nucleotide that is potentially methylated corresponds precisely to one of the edited nucleotides. Since 2'-O-methyl groups have been shown to efficiently inhibit editing in vitro,245 these results suggest a possible role for 2'-O-methylation in mRNA regulation. Finally, the HT2c pre-mRNA itself contains an H/ACA snoRNA in its first intron, and has a long 5’UTR that might function as an internal ribosome entry site. This unusual structure suggests that this mRNA could follow a special trafficking pathway in neural cells. However, despite these fascinating observations, it is not yet known if the HT2c mRNA is methylated in vivo. At least one of the brain specific snoRNAs (RBII-36) accumulates in nucleoli, based on in situ hybridization results.236 Importantly, nucleolar localization does not preclude a role for these RNAs in the nucleoplasm, as exemplified by the behavior of telomerase RNA.222,246 Indeed, studies of the mobility of both fibrillarin and Snu13p/15.5kD within live cell nuclei showed that despite nucleolar localization at steady-state, these box C/D core proteins exchange rapidly between the nucleolus and the nucleoplasm.61,247 The mean residency time in nucleoli is only about one minute, suggesting that the snoRNP is able to explore the entire volume of the nucleus in relatively short times.

The snoRNPs and Related Machines

239

Just What Is a snoRNA, Anyway? The occurrence of these different types of box C/D and H/ACA RNAs challenges the present system of defining snoRNAs and predicting roles of snoRNA-like RNAs in the cell. In addition, the existence of these RNAs clearly demonstrates that while they may have originated from chaperone or nucleotide modification devices acting in cis, they have taken up a number of new functions during evolution. Indeed, the C/D and H/ACA RNA folds are simple and could provide powerful means to process, stabilize and localize RNAs, distinct from their functions in nucleotide modification. It seems all but certain that future studies will reveal an even greater diversity of roles and functions for these classes of small (and not-so-small) RNAs.

Biogenesis of snoRNPs Production of snoRNPs is a complex matter, involving intricate control of many synthesis and assembly reactions, and trafficking of proteins and nascent RNP complexes through different sub-cellular compartments, as reviewed recently1 and see below. The snoRNAs are transcribed in the nucleoplasm, typically from coding sequences within introns, and the proteins are imported from the cytoplasm. One or more snoRNP core proteins are believed to bind to the pre-snoRNA molecule, followed by nucleolytic processing by a variety of enzymes. At least some and possibly all of the nascent snoRNPs then move to specialized nuclear bodies, i.e., the Cajal bodies in animal cells and functionally related nucleolar bodies (NBs) in yeast. The scaRNPs remain in the Cajal bodies, to function there, and the nucleolar species move on to that structure.8,86 At nuclear locations not yet defined, snoRNAs also undergoes nucleotide modification, and the final assembly steps of the mature snoRNP take place. Residency in the nucleolus may be temporary as well, as results suggest that snoRNPs may cycle between this structure and the Cajal bodies, as demand for nucleolar function waxes and wanes. Some of these processes are tightly coupled and, not surprisingly, production of snoRNPs is linked to production of ribosomes, as reviewed previously.1-3,17,154,248-250

Synthesis of snoRNAs The coding sequences for snoRNAs occur in a variety of genomic arrangements, and the types and relative abundance varies among organisms, as reviewed earlier.1,3,249 In vertebrates, most snoRNAs are derived from coding units that are embedded within introns of protein genes; early lists appear elsewhere.25,40,43,45,108,164,237,251-254 However, other snoRNAs are generated from independent genes, such as U3; pseudogenes for U3 also exist.255 Processing snoRNAs are encoded in both arrangements, whereas the guide snoRNAs appear to be largely, if not exclusively specified by intronic coding units. Notably, some intronic snoRNAs are encoded in mRNA-like transcripts that are not translatable.157,256-261 The aforementioned class of brain-specific snoRNAs characterizes yet another arrangement, where intronic coding units are organized in large, tandem repeats. The situation is different in yeast, where a small minority of snoRNAs is encoded in introns, and the others are transcribed from genes that produce snoRNAs only.25,254 This last class consists of both monocistronic and polycistronic coding schemes, and includes multimeric precursors containing from two to seven snoRNAs, as cited for representative cases.37,110,169,262 Plants have all of these arrangements and polycistronic coding units seem to be common.263-270 Remarkably, one plant transcription unit encodes a tRNA in addition to several copies of a single snoRNA.271 Among the Archaea, most coding units for the known and predicted C/D guide sRNAs occur in intergenic regions, and a few overlap with an ORF, raising questions about how the latter snoRNAs would be produced3,9,11,41,42 As noted above, coding sequences for guide sRNAs also occur in tRNA introns and, in transcribed spacers of rRNA; the latter encoded as half-molecules, which are joined after cleavage of the pre-rRNA.10,210 If the coding units for the C/D RNAs arose from a single ancestral source, ancient coding units for pre-tRNA and

240

The Nucleolus

rRNA must be viewed as strong, logical contenders. Speculating about the origin of the H/ ACA snoRNAs is more difficult, because of a much smaller genomic database. Following transcription, precursor snoRNAs undergo nucleolytic processing by a variety of endo- and exonucleases.272-277 Processing of intronic snoRNAs appears to occur in most cases after the intron has been excised and debranched, with 5’ and 3’ exonucleases carrying out the final trimming reactions.274,278,279 However, production can also occur when debranching is blocked, mediated by endonucleolytic cutting followed by exonucleolytic trimming. In one gene system, alternative pathways exist, in which processing yields either mRNA or snoRNA.277 For the C/D snoRNAs in mammals, the distance of the intron from the branch point (∼70 nts) is important, implying that interference can occur between formation of the splicing and snoRNA maturation complexes.106 Besides end-processing, snoRNAs also undergo nucleotide modification, however, the relationship of processing and modification is still obscure.25,35,280 Moreover, the identities of the factors that catalyze modification of the snoRNAs themselves are not known. Are they protein-only factors or do dedicated RNPs perform these tasks? Does this occur in the nucleolus or in another nuclear compartment such as the Cajal bodies? For precursors generated from mono- or polycistronic operons, processing involves the concerted action of endonucleases, which liberate smaller precursors and these, in turn, are trimmed as needed by exonucleases. Exceptions occur for a subset of snoRNAs with monomethylated caps at the 5’ end, such as U3; in these cases the cap becomes trimethylated and trimming is limited to the 3’ end.60,104,107,281 Yeast U3 is unusual in that it contains an intron that is also removed during maturation.282 It seems clear that binding of proteins to the precursor snoRNAs is essential for proper processing and for metabolic stability. For the C/D and H/ACA snoRNAs, these properties are believed to result from binding of core proteins to the corresponding motifs, but other, non-snoRNP proteins may also be involved, for example, in RNP assembly, nuclease recruitment and quality control (see below). Excellent progress has been made in defining processing pathways for yeast snoRNAs and the events and activities seem largely conserved in vertebrates. Key steps and enzymes involved include: 1) the endonuclease Rnt1p, a homolog of the bacterial RNase III enzyme, which cleaves hairpin duplexes that precede or follow a snoRNA;272,283,284 2) two 5’-> 3’ exonucleases, Rat1p and Xrn1p,274,279 and; 3) 3’-> 5’ exonucleases that are part of the exosome, including Rrp6p.275,276 In some cases, trimming of 3’ trailer sequences is initiated not by Rnt1p, but by an endonucleolytic cleaving activity that also cleaves pre-mRNA, at the site where poly-A formation takes place.277,285

snoRNP Assembly Core proteins. Assembly of box C/D snoRNPs occurs as a series of ordered events. The first protein to bind is Snu13p/NHPX, which recognizes specifically and directly the K-turn formed by the non-canonical GA.AG base-pairs in the box C/D motif.72 The crystal structure of human NHPX complexed with the U4 snRNA 5' stem-loop shows that its binding induces a 150° bend in the phosphodiester backbone.96 Results from in vitro binding studies with snoRNAs demonstrate that this protein is absolutely required for assembling the other C/D core proteins, fibrillarin, Nop56p, and Nop58p, suggesting that the RNA conformational change induced by Snu13p/NHPX binding is essential to create the other binding sites. A recent study using crosslinking agents inserted at specific locations showed that the U.U pair in internal stem II directly binds snoRNP proteins; the U in box C contacts Nop58p, and the one in box D is in contact with fibrillarin.286 Importantly, the C’/D’ motif contained by most box C/D snoRNAs is much less conserved than the box C/D motif,95 and is thus unlikely to bind NHPX independently in a stable manner. The cross-linking results further showed that the C/D motif in the RNP complex was in proximity to fibrillarin and Nop58p, while the C’/ D’ motif associates with fibrillarin and Nop56p. Thus, the eukaryotic box C/D snoRNPs could adopt a pseudo-symmetrical structure, with NHPX, fibrillarin and Nop58p at the box C/D motif, and fibrillarin and Nop56p at the box C’/D’ motif. To compensate for the absence of

The snoRNPs and Related Machines

241

Snu13p/NHPX at the box C’/D’ motif, binding of fibrillarin and Nop56p could be stabilized by protein-protein interaction with partners located at the box C/D motif. This model is consistent with effects of point mutations in stem II of the C/D motif, which suggest that Nop56p binds together with fibrillarin and independent of Nop58p.101 The results are also in agreement with in vivo studies of box C/D snoRNP assembly in yeast, which showed that depletion of fibrillarin prevented binding of Nop56p, but not Nop58p.94 The box C/D motif thus appears as a bipartite structure, with the K-turn formed by the GA.AG base pairs binding NHPX, and specific, conserved nucleotides in stem II assembling with fibrillarin and Nop58p. The box C’/D’ motif likely provides a secondary site to assemble with Nop56p and a second molecule of fibrillarin. A pseudo-symmetrical structure for the C/ D snoRNA would be consistent with both the structure of the archaeal box C/D sRNAs, which contain a much better conserved C’/D’ motif and have a single homolog of Nop56p and Nop58p,9,41 and with the fact that both the box C/D and C’/D’ motifs can direct methylation of target sequences (but see below). Distinct from the assembly situation for the animal snoRNPs, an in vitro study of archaeal sRNP complex formation revealed a simpler pattern of protein binding (E.S. Maxwell, pers. commun.). The smaller set of core proteins bound to both the C/D and C’/D’ motifs in a symmetrical manner, with all three proteins present at both sites. In vitro methylation assays showed that each individual RNP domain is active with adjoining guide sequences, but maximal activity required the presence of both the C/D and C’/D’ RNP complexes. The asymmetric protein binding patterns observed for the eukaryotic snoRNP could reflect increased specialization of the larger set of binding proteins, and possible divergence of the structures of the C/D and C’/D’ binding domains. It seems possible that some eukaryotic guide RNAs might have the same symmetry as archaeal sRNAs, in particular snoRNAs that use both the C/D and C’/D’ motifs for methylation reactions. An important crystal structure has been reported for an archaeal fibrillarin-Nop56/58 complex (with S-adenosyl methionine). The results show direct interaction of the proteins in a pre-formed heterodimer, and association of these complexes to form a dimeric heterodimer. The heterodimers are linked through interaction of extended coiled-coil elements at the carboxy end of Nop56/58, and fibrillarin protein interacts with Nop56/58 at its amino end. In the homodimeric heterodimer the fibrillarin molecules are located at opposite ends of the extended 4-protein complex. This structure provides a model for placement of methylase proteins at both the C/D and C’/D’ motifs.287 Box H/ACA snoRNPs contain one protein related to NHPX, i.e., Nhp2p.119,120 This protein also binds RNA, however, the binding site in the snoRNP is not yet defined precisely. Results from in vivo studies in yeast have established that Nhp2p, Nop10p, and Cbf5p are all required for metabolic stability of the snoRNA, in contrast to Gar1p, which is required for snoRNP function, but not stability. This situation suggests that Nhp2p, Nop10p and Cbf5p may form a complex that collectively generates RNA binding specificity. As noted above, the only homolog of Nhp2p in the Archaea is the ribosomal protein L7, and recent work suggests that it is part of the archaeal box H/ACA snoRNPs.124 The archaeal RNAs differ from the eukaryotic counterparts by having a kink-turn motif placed 5-11 nt above each pseudouridylation pocket, which provides the docking site for L7. Thus, similar to the case for the box C/D snoRNAs, L7 may play a key role in nucleating snoRNP assembly of archaeal box H/ACA snoRNAs. In eukaryotes, the lack of stringent specificity of Nhp2p binding may be compensated by a more strict arrangement of boxes H and ACA. Consistent with this possibility, the RNA folding models and electron microscopy images of a yeast box H/ ACA snoRNP (snR30) showed a bipartite structure, which is presumed to correspond to proteins bound to the two hairpins in the RNA.119 In Archaea (and Trypanosomes), this bipartite structure is not conserved, and a variety of arrangements that contain from 1 to 3 hairpins have been found.42

242

The Nucleolus

SnoRNP assembly factors. A number of proteins have been shown recently to function as snoRNP assembly factors. Interestingly, several of these are involved in production of both C/ D and H/ACA snoRNPs.1 The most conserved are the p50/p55 proteins (Tip48, Tip49 in humans, Rvb1p, Rvb2p in yeast), which are inter-related and known to have DNA helicase activity. These proteins are found in Archaea, yeast, and humans, and they have been linked to a number of processes, including DNA repair and transcription, as cited elsewhere.75,288 While p50/p55 are not present in the mature snoRNP and are not localized in nucleoli, results from a genetic depletion study in yeast demonstrate that they are required for the biogenesis of both C/D and H/ACA snoRNAs.288 Importantly, the levels of the mRNAs that encode snoRNAs within introns are not affected at the time of snoRNA loss, pointing to a post-transcriptional role for these proteins. In vitro studies demonstrated that p50 and p55 assemble onto a model C/D snoRNA.75,85 Complex formation requires Snu13p/15.5 kDa binding, specific nucleotides in the box C/D motif, and appears to occur when Nop58p joins the snoRNP complex.101 Both p50 and p55 possess ATPase consensus sequences, and mutations in these elements lead to both temperature sensitive growth, and a concomitant mis-localization of C/D snoRNAs at the non-permissive temperature. Several possible roles can be envisioned for these proteins, for example, they could chaperone snoRNP assembly, carry assembled snoRNPs to nucleoli, or alternatively, prevent nucleolar localization until needed. Another important protein recently suggested to function in the assembly of both box H/ ACA and C/D snoRNPs is SMN, the survival of motor neuron protein.289-291 SMN is an essential protein that is conserved in many species, including fission yeast, invertebrates and vertebrates, but which is absent from S. cerevisiae.292 SMN is an oligomeric protein that has been shown to be essential for cytoplasmic assembly of splicing snRNPs, where it functions as a chaperone to mediate assembly of the heptameric ring of Sm proteins around the Sm binding site in the snRNA.293 The SMN protein is also present in the nucleus in structures called gems, which often overlap with Cajal bodies.294,295 Recently, it was shown that SMN binds in vitro and in vivo with both fibrillarin and Gar1p, through domains containing RG repeats.289,290 In addition, a dominant-negative mutant of SMN was shown to block the nucleolar accumulation of human U3 snoRNA,289,290 providing good support to the idea that SMN plays a role in snoRNP biogenesis, possibly by providing a platform to assemble snoRNP proteins onto the RNA. Two other proteins, Naf1p and Shq1p, are also essential for the biogenesis of yeast H/ACA snoRNPs.91-93 However, in contrast to p50/p55 and SMN, these proteins do not affect accumulation of C/D snoRNAs (or other stable RNAs). Naf1p and Shq1p form a complex and they interact in vitro and in two-hybrid assays with at least two protein components of the H/ ACA snoRNPs, i.e., Nhp2p and Cbf5p. Naf1p also binds in vivo to H/ACA snoRNP proteins and several H/ACA snoRNAs. However, only small amounts of these components were recovered in co-immunoprecipitation assays, suggesting that they are not part of the mature snoRNP. In agreement, the major portions of these proteins localize to the nucleoplasm, and small amounts detected in the nucleolus are excluded from the dense fibrillar compartment, the site where mature H/ACA snoRNPs accumulate and are thought to function. Naf1p has the interesting property of binding to the phosphorylated CTD of RNA polymerase II and binds RNA through a domain similar to that of Nrd1p, an RNA binding protein required for 3’-end formation of snRNAs and snoRNAs in yeast.285 These data suggest a model in which Naf1p and Shq1p recruit Nhp2p and Cbf5p to the transcribing RNA polymerases, bind newly synthesized RNA and assist snoRNP assembly. A role in the transport of the snoRNP toward the nucleolus is also possible, given that a small fraction of these proteins is nucleolar. It is also of interest that Shq1p interacts with Rnt1p in two-hybrid assays, suggesting a possible role in coupling snoRNA processing with RNP assembly. Moreover, Rnt1p interacts with Gar1p in vivo and in vitro, and is required for its nuclear import296 providing further evidence of possible links between RNA processing, RNP assembly

The snoRNPs and Related Machines

243

and transport. Clearly, these intricate, coupled processes need to be dissected in detail. It should be noted that Naf1p and Shq1p have homologs in many species, including mammals, raising the possibility that the roles of these proteins in H/ACA snoRNP biogenesis is widespread in metazoans.

Trafficking and Localization Localization pathways. Early studies of snoRNA trafficking showed that the box C/D and H/ACA motifs are sufficient to target a reporter RNA to the nucleoli of yeast, Xenopus oocytes, and mammalian cells.48,49,53,54 More recently, the roles of individual proteins in snoRNA trafficking have been analyzed in genetic studies in yeast.58 Depletion of each of the C/D core snoRNP proteins leads to a partial mislocalization of snoRNAs to the nucleoplasm. This finding suggests that the putative nucleolar localization signal is split and that the snoRNP is stabilized in the nucleolus by multiple weak interactions, consistent with the absence of highly efficient nucleolar localization signals in the snoRNP proteins themselves. However, other interpretations are also possible, for example, quality control mechanisms could block mis-assembled snoRNPs from entering the nucleolus. Interestingly, in vertebrates, reporter RNAs containing C/D or H/ACA motifs are also localized to Cajal bodies,48,49,53,54 a structure that is biochemically and spatially related to nucleoli.297,298 Time-course experiments following injection in Xenopus oocytes showed that box C/D snoRNAs first localize in Cajal bodies, and only later to nucleoli.73 This important result demonstrated that snoRNAs follow a specific intranuclear route. This was the second evidence for trafficking pathways within the nucleus, following the demonstration that splicing snRNAs localize in Cajal bodies after entering the nucleus, and before they become distributed in speckles.218 The fact that the trafficking of these two RNA species flows through the same compartment suggests an important role for Cajal bodies in sorting molecules within the nucleus. In the case of the H/ACA snoRNAs, it is not clear if localization in Cajal bodies reflects a localization pathway, as they seem to reach nucleoli and Cajal bodies at about the same time.53 In this case, the flow of RNA molecules may be too fast to be resolved. It is also possible that these RNAs follow a more complex route, or simply do not follow a single pathway, but have two destinations. Cajal bodies and snoRNP assembly. The transient localization of C/D snoRNAs to Cajal bodies suggests that some steps of snoRNA maturation take place there. A recent study that focused on U3 snoRNA maturation showed that this is likely to be the case.60 Similar to the situation in yeast, mammalian U3 is synthesized as a precursor that contains a 3' extension of a few nucleotides and a monomethylated cap. Interestingly, this precursor is assembled with Snu13p/15.5 kDa, but not with fibrillarin or Nop58p. Furthermore, this precursor is present at the snoRNA transcription site and within Cajal bodies, but is excluded from nucleoli. In contrast, mature U3 RNA, which bears a trimethyl cap structure, is detected both in Cajal bodies and in nucleoli. Since all C/D core snoRNP proteins are present in Cajal bodies, these data suggest that U3 snoRNPs are preferentially assembled in Cajal bodies, consistent with the concentration of SMN in this structure and its possible role as a chaperone in snoRNP assembly (Fig. 5). A role for Cajal bodies in box C/D snoRNP assembly received further support from the finding that Snu13p/15.5kD and fibrillarin follow very distinct localization pathways.61 Indeed, when entering the nucleus, fibrillarin localizes rapidly to Cajal bodies and to nucleoli, while Snu13p/15.5kD is first distributed in speckles and Cajal bodies, and is detected in nucleoli only after a lag of about 30 minutes. This situation may indicate that Snu13p/15.5kD first binds intronic snoRNA precursors in speckles, and then accompanies them toward nucleoli. In contrast, fibrillarin may join pre-assembled complexes at the level of Cajal bodies and may thus reach the nucleoli more rapidly.60,61 Finally, characterization of the enzyme responsible for cap hypermethylation also supports a role for the Cajal body in snoRNP maturation and assembly. The yeast enzyme, Tgs1p, was

244

The Nucleolus

Figure 5. A model for the trafficking and biogenesis of box C/D snoRNAs in yeast and human cells. See text.

shown to hypermethylate the caps of snRNAs, and box C/D and H/ACA snoRNAs, where these exist.59 The human genome encodes only one homolog of the enzyme, and the protein has a localization pattern similar to that of SMN: it is present both in the cytoplasm and in Cajal bodies,60 where it could modify the cap structure of U3. Importantly, the enzyme does not bind RNA directly, but docks on the basic C-termini of Cbf5p (dyskerin), Nop56p and Nop58p,59 providing another line of evidence that snoRNP assembly takes place in Cajal bodies. Yeast nucleolar bodies. S. cerevisiae does not have Cajal bodies, and some evidence suggests that in this organism the nucleolus, or a specific nucleolar sub-domain, may provide some of its

The snoRNPs and Related Machines

245

functions. First, Tgs1p, the cap hypermethylase, localizes in the nucleolus.59 Remarkably, it has been shown that when yeast are grown on solid medium, a fraction of the cells concentrate Tgs1p in a dot-like structure that is within the nucleolar territory;60 however, factors actively involved in rRNA biogenesis such as Snu13p are excluded from this structure. Second, U3 precursors that are not assembled with fibrillarin, Nop56p, and Nop58p distribute throughout the nucleus including nucleoli, and can also be found in the Tgs1p-enriched nucleolar domain.60 Over-expression of certain box C/D snoRNAs also triggers its assembly.58 These results indicate that this domain is not likely a storage area for excess Tgs1p enzyme, but rather, is involved in the biogenesis of small RNAs. Electron-microscopy images show that when induced by snoRNA over-expression, the Tgs1p-rich domain has a round shape, is connected to the dense fibrillar component of the nucleoli, and strikingly, has a similar appearance.58 Consistent with these properties this structure has been named the ‘nucleolar body’. Because the human homolog of Tgs1p localizes in Cajal bodies,60 where it is likely involved in U3 biogenesis, it appears that the nucleolar body is related to Cajal bodies, and may share with it at least certain functions. It is also remarkable that in many cells Cajal bodies are often physically linked with nucleoli, and that in certain cell lines they are even located within nucleoli, similar to the location of the yeast nucleolar body.298 While it appears that the nucleolar body forms only in certain growth conditions, its existence nevertheless shows that the nucleolar activities involved in the biogenesis of small RNAs can self-assemble and segregate from those involved in ribosome production. These properties are likely to be ancient, and they may have been at the basis of the formation of modern Cajal bodies as an independent structure.

Exploiting snoRNPs Two major properties of snoRNAs have been harnessed for novel in vivo applications. One is using snoRNA localization elements to deliver new RNA sequences to the nucleolus. The other is guiding Nm and Ψ modifications to new sites in rRNA by expressing snoRNAs with novel guide sequences. The number of reports is still small, however good success has been achieved and additional applications seem certain to follow. Transport of RNA aptamers and ribozymes. RNA sequences have been directed to the nucleolus (and Cajal) bodies by incorporating the experimental sequence into an internal position in a natural snoRNA or by engineering smaller snoRNA variants in which the localization signals have been added to the ends of the RNA of interest. One study designed to explicitly test this potential demonstrated that expressed polylinker sequences could be targeted to the nucleolus by simply inserting the experimental coding sequence between the terminal box C and box D elements.49 More recently, snoRNAs have been used as vehicles to deliver a protein-binding signal to the nucleoli of mammalian cells299,300 and a hammerhead ribozyme to the nucleoli of yeast and human cells.301,302 The protein binding sites are from the HIV genome and known to bind the Rev and Tat proteins. These proteins are required for virus production and expression of the binding elements in a C/D snoRNA context altered snoRNA localization (Rev) or virus yield (Tat) in cultured cells. The results demonstrate the feasibility of localizing aptamers by this approach. The ability to deliver a ribozyme to the nucleolus was first demonstrated in yeast, with remarkable results.301 A hammerhead ribozyme was carried to the nucleolus by the U3 snoRNA and cleaved another U3 variant with nearly perfect efficiency (>95%). The hybrid snoRNA-ribozyme, called a ‘snorbozyme’, was shown to localize to the nucleolus. Building on this strategy, a hammerhead snorbozyme specific for HIV RNA was expressed in virus-infected cells, shown to localize to nucleoli and Cajal bodies, and to inhibit virus infection.302 Targeting nucleotide modification to novel RNA sites. This capability was established when the guide functions were identified45-47,109,162 and has since been demonstrated in two contexts. One is to first deplete a guide snoRNA for a natural modification and then introduce the modification at the same or nearby sites, by expressing a new guide RNA.45,162 The other

246

The Nucleolus

application is functional mapping, where creating modifications at novel sites in an RNA impairs activity.303 Nm modifications have been introduced at many new sites,45,109,162,303 whereas success in targeting Ψ modifications to new rRNA sites has been achieved in only a few cases.46,117 Engineering new Ψ guide snoRNAs has been more difficult, most likely due to more stringent structural requirements for achieving good activity. Creating new Ψ guides involves inserting two guide sequences into a highly folded secondary structure domain versus placing a single Nm guide sequence in a RNA region that appears to have little or no secondary folding. The most striking example of a positional effect seen with a natural modification, is for a Ψ in the yeast large subunit rRNA that occurs in the A-loop of the peptidyltransferase center. This loop interacts with the –CCA end of aminoacyl-tRNA and the Ψ modification is conserved among eukaryotes.175 Blocking Ψ formation caused a 20% loss in the rate of in vivo protein synthesis activity. Remarkably, shifting the modification to an adjacent uridine caused a severe slow-growth phenotype (R. McCully, T. King and M.J.F., unpublished). Excellent success has been obtained in creating interference function maps of yeast rRNA by conditional expression of novel Nm guide snoRNAs (from an inducible promoter). Modifications have been targeted to individual pre-selected sites in 18S and 25S rRNAs, and, with a library strategy, to all sites in an 800-nucleotide segment of 25S rRNA that encompasses the peptidyl transferase center (PTC)303 (and B. Liu and MJF, unpublished). Transformants are screened for growth defects, and the guide snoRNAs of interest identified by sequencing the novel snoRNA gene. Strong growth defects have been observed for several rRNA sites, whereas most new guide snoRNAs have no effect, consistent with the introduction of point mutations. The sensitive nucleotides include several sites known or predicted to be important for ribosome activity. Interestingly, the most toxic snoRNAs have been determined to interfere with ribosome biogenesis or ribosome activity (B. Liu and MJF, unpublished). Taken together, these early results indicate that snoRNPs can be effectively exploited to manipulate RNA levels and activity in vivo. As our understanding of snoRNP trafficking and substrates advances, the potential for creating powerful and important new tools increases as well.

Acknowledgments We are grateful to: Wayne Decatur, Ben Liu, Denis Lafontaine and E.S. Maxwell for critically reading the manuscript, W. Decateur and B. Liu for important help with literature citations, and to W. Decatur for excellent help in preparing figures. We also thank E.S. Maxwell for advice and new results, and other colleagues who also shared unpublished information. This work was supported by grants from the AFM, CNRS (ACI), EMBO YIP program (to EB), and the U.S. National Institutes of Health (GM3951 to MJF).

References 1. Filipowicz W, Pogacic V. Biogenesis of small nucleolar ribonucleoproteins. Curr Opin Cell Biol 2002; 14:319-327. 2. Kiss T. Small nucleolar RNA-guided post-transcriptional modification of cellular RNAs. EMBO J 2001; 20:3617-3622. 3. Bachellerie JP, Cavaille J, Huttenhofer A. The expanding snoRNA world. Biochimie 2002; 84:775-790. 4. Omer AD, Ziesche S, Decatur WA et al. RNA-modifying machines in Archaea. Molecular Microbiol 2003; 48:617-629. 5. Cavaille J, Buiting K, Kiefmann M et al. Identification of brain-specific and imprinted small nucleolar RNA genes exhibiting an unusual genomic organization. Proc Natl Acad Sci USA 2000; 97:14311-14316. 6. Liang XH, Xu YX, Michaeli S. The spliced leader-associated RNA is a trypanosome-specific sn(o) RNA that has the potential to guide pseudouridine formation on the SL RNA RNA 2002; 8:237-246. 7. Gall JG. A role for Cajal bodies in assembly of the nuclear transcription machinery. FEBS Lett 2001; 498:164-167.

The snoRNPs and Related Machines

247

8. Darzacq X, Jady BE, Verheggen C et al. Cajal body-specific small nuclear RNAs: a novel class of 2'-O-methylation and pseudouridylation guide RNAs. EMBO J 2002; 21:2746-2756. 9. Omer AD, Lowe TM, Russell AG et al. Homologs of small nucleolar RNAs in Archaea. Science 2000; 288:517-522. 10. Clouet d’Orval B, Bortolin ML, Gaspin C et al. Box C/D RNA guides for the ribose methylation of archaeal tRNAs. The tRNATrp intron guides the formation of two ribose-methylated nucleosides in the mature tRNATrp. Nucleic Acids Res 2001; 29:4518-4529. 11. Dennis PP, Omer A, Lowe T. A guided tour: Small RNA function in Archaea Mol Microbiol 2001; 40:509-519. 12. Pasquinelli AE. MicroRNAs: deviants no longer. Trends Genet 2002; 18:171-173. 13. Decatur WA, Fournier MJ. RNA-guided nucleotide modification of ribosomal and other RNAs. J Biol Chem 2003; 278:695-698. 14. Kiss T. Small nucleolar RNAs: An abundant group of noncoding RNAs with diverse cellular functions. Cell 2002; 109:145-148. 15. Peculis BA. RNA-binding proteins: If it looks like a sn(o)RNA. Curr Biol 2000;10:R916-918. 16. Terns MP, Terns RM. Small nucleolar RNAs: versatile trans-acting molecules of ancient evolutionary origin. Gene Expr 2002; 10:17-39. 17. Perumal K, Reddy R. The 3' end formation in small RNAs. Gene Expr 2002; 10:59-78. 18. Carmo-Fonseca M. New clues to the function of the Cajal body. EMBO Rep 2002; 3:726-727. 19. Reddy R, Busch H. Small nuclear RNAs: RNA sequences, structure, and modifications. In: Birnstiel ML, ed. Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. New York: Springer-Verlag, 1988:1-37. 20. Busch H, Reddy R, Rothblum L et al. SnRNAs, SnRNPs, and RNA processing. Annu Rev Biochem 1982; 51:617-654. 21. Weinberg RA, Penman S. Small molecular weight monodisperse nuclear RNA. J Mol Biol 1968; 38:289-304. 22. Prestayko AW, Tonato M, Busch H. Low molecular weight RNA associated with 28S nucleolar RNA. J Mol Biol 1970; 47:505-515. 23. Reddy R, Henning D, Busch H. Nucleotide sequence of nucleolar U3B RNA. J Biol Chem 1979; 254:11097-11105. 24. Reddy R, Sitz TO, Ro-Choi TS et al. Two-dimensional polyacrylamide gel electrophoresis separation of low molecular weight nuclear RNA. Biochem Biophys Res Commun 1974; 56:1017-1022. 25. Maxwell ES, Fournier MJ. The small nucleolar RNAs. Annu Rev Biochem 1995; 64:897-934. 26. Kass S, Tyc K, Steitz JA et al. The U3 small nucleolar ribonucleoprotein functions in the first step of preribosomal RNA processing. Cell 1990; 60:897-908. 27. Wise JA, Weiner AM. Dictyostelium small nuclear RNA D2 is homologous to rat nucleolar RNA U3 and is encoded by a dispersed multigene family. Cell 1980; 22:109-118. 28. Parker KA, Steitz JA. Structural analysis of the human U3 ribonucleoprotein particle reveal a conserved sequence available for base pairing with pre-rRNA. Mol Cell Biol 1987; 7:2899-2913. 29. Porter GL, Brennwald PJ, Holm KA et al. The sequence of U3 from Schizosaccharomyces pombe suggests structural divergence of this snRNA between metazoans and unicellular eukaryotes. Nucleic Acids Res 1988; 16:10131-10152. 30. Jeppesen C, Stebbins-Boaz B, Gerbi SA. Nucleotide sequence determination and secondary structure of Xenopus U3 snRNA. Nucleic Acids Res. 1988; 16:2127-2148. 31. Mereau A, Fournier R, Gregoire A et al. An in vivo and in vitro structure-function analysis of the Saccharomyces cerevisiae U3A snoRNP: Protein-RNA contacts and base-pair interaction with the pre-ribosomal RNA. J Mol Biol 1997; 273:552-571. 32. Samarsky DA, Fournier MJ. Functional mapping of the U3 small nucleolar RNA from the yeast Saccharomyces cerevisiae. Mol Cell Biol 1998; 18:3431-3444. 33. Wormsley S, Samarsky DA, Fournier MJ et al. An unexpected, conserved element of the U3 snoRNA is required for Mpp10p association. RNA. 2001; 7:904-919. 34. Tyc K, Steitz JA. U3, U8 and U13 comprise a new class of mammalian snRNPs localized in the cell nucleolus. EMBO J 1989; 8:3113-3119. 35. Wise JA, Tollervey D, Maloney D et al. Yeast contains small nuclear RNAs encoded by single copy genes. Cell 1983; 35:743-751. 36. Riedel N, Wise JA, Swerdlow H et al. Small nuclear RNAs from Saccharomyces cerevisiae: unexpected diversity in abundance, size, and molecular complexity. Proc Natl Acad Sci USA 1986; 83:8097-8101. 37. Zagorski J, Tollervey D, Fournier MJ. Characterization of an SNR gene locus in Saccharomyces cerevisiae that specifies both dispensable and essential small nuclear RNAs. Mol Cell Biol 1988; 8:3282-3290.

248

The Nucleolus

38. Thompson JR, Zagorski J, Woolford JL et al. Sequence and genetic analysis of a dispensable 189 nucleotide snRNA from Saccharomyces cerevisiae. Nucleic Acids Res 1988; 16:5587-5601. 39. Balakin AG, Smith L, Fournier MJ. The RNA world of the nucleolus: Two major families of small RNAs defined by different box elements with related functions. Cell 1996; 86:823-834. 40. Ganot P, Caizergues-Ferrer M, Kiss T. The family of box ACA small nucleolar RNAs is defined by an evolutionarily conserved secondary structure and ubiquitous sequence elements essential for RNA accumulation. Genes Dev 1997; 11:941-956. 41. Gaspin C, Cavaille J, Erauso G et al. Archaeal homologs of eukaryotic methylation guide small nucleolar RNAs: Lessons from the Pyrococcus genomes. J Mol Biol 2000; 297:895-906. 42. Tang TH, Bachellerie JP, Rozhdestvensky T et al. Identification of 86 candidates for small non-messenger RNAs from the archaeon Archaeoglobus fulgidus. Proc Natl Acad Sci USA 2002; 99:7536-7541. 43. Leverette RD, Andrews MT, Maxwell ES. Mouse U14 snRNA is a processed intron of the cognate hsc70 heat shock pre-messenger RNA. Cell 1992; 71:1215-1221. 44. Chu S, Archer RH, Zengel JM et al. The RNA of RNase MRP is required for normal processing of ribosomal RNA. Proc Natl Acad Sci USA 1994; 91:659-663. 45. Kiss-Laszlo Z, Henry Y, Bachellerie JP et al. Site-specific ribose methylation of preribosomal RNA: a novel function for small nucleolar RNAs. Cell 1996; 85:1077-1088. 46. Ni J, Tien AL, Fournier MJ. Small nucleolar RNAs direct site-specific synthesis of pseudouridine in ribosomal RNA. Cell 1997; 89:565-573. 47. Ganot P, Bortolin ML, Kiss T. Site-specific pseudouridine formation in preribosomal RNA is guided by small nucleolar RNAs. Cell 1997; 89:799-809. 48. Lange TS, Borovjagin A, Maxwell ES et al. Conserved boxes C and D are essential nucleolar localization elements of U14 and U8 snoRNAs. EMBO J 1998; 17:3176-3187. 49. Samarsky DA, Fournier MJ, Singer RH et al. The snoRNA box C/D motif directs nucleolar targeting and also couples snoRNA synthesis and localization. EMBO J 1998; 17:3747-3757. 50. Shaw PJ, Beven AF, Leader DJ et al. Localization and processing from a polycistronic precursor of novel snoRNAs in maize. J Cell Sci 1998; 111:2121-2128. 51. Lange TS, Borovjagin AV, Gerbi SA. Nucleolar localization elements in U8 snoRNA differ from sequences required for rRNA processing. RNA 1998; 4:789-800. 52. Lange TS, Ezrokhi M, Borovjagin AV et al. Nucleolar localization elements of Xenopus laevis U3 small nucleolar RNA. Mol Biol Cell 1998; 9:2973-2985. 53. Narayanan A, Lukowiak A, Jady BE et al. Nucleolar localization signals of box H/ACA small nucleolar RNAs. EMBO J 1999; 18:5120-5130. 54. Lange TS, Ezrokhi M, Amaldi F et al. Box H and box ACA are nucleolar localization elements of U17 small nucleolar RNA. Mol Biol Cell 1999; 10:3877-3890. 55. Speckmann W, Narayanan A, Terns R et al. Nuclear retention elements of U3 small nucleolar RNA. Mol Cell Biol 1999; 19:8412-8421. 56. Lange TS, Gerbi SA. Transient nucleolar localization of U6 small nuclear RNA in Xenopus laevis oocytes. Mol Biol Cell 2000; 11:2419-2428. 57. Speckmann WA, Terns RM, Terns MP. The box C/D motif directs snoRNA 5'-cap hypermethylation. Nucleic Acids Res 2000; 28:4467-4473. 58. Verheggen C, Mouaikel J, Thiry M et al. Box C/D small nucleolar RNA trafficking involves small nucleolar RNP proteins, nucleolar factors and a novel nuclear domain. EMBO J 2001; 20:5480-5490. 59. Mouaikel J, Verheggen C, Bertrand E et al. Hypermethylation of the cap structure of both yeast snRNAs and snoRNAs requires a conserved methyltransferase that is localized to the nucleolus. Mol Cell 2002; 9:891-901. 60. Verheggen C, Lafontaine DL, Samarsky D et al. Mammalian and yeast U3 snoRNPs are matured in specific and related nuclear compartments. EMBO J 2002; 21:2736-2745. 61. Leung AK, Lamond AI. In vivo analysis of NHPX reveals a novel nucleolar localization pathway involving a transient accumulation in splicing speckles. J Cell Biol 2002; 157:615-629. 62. Narayanan A, Eifert J, Marfatia KA et al. Nuclear RanGTP is not required for targeting small nucleolar RNAs to the nucleolus. J Cell Sci 2003; 116:177-186. 63. Schimmang T, Tollervey D, Kern H et al. A yeast nucleolar protein related to mammalian fibrillarin is associated with small nucleolar RNA and is essential for viability. EMBO J 1989; 8:4015-4024. 64. Henriquez R, Blobel G, Aris JP. Isolation and sequencing of NOP1. A yeast gene encoding a nucleolar protein homologous to a human autoimmune antigen. J Biol Chem 1990; 265:2209-2215. 65. Aris JP, Blobel G. cDNA cloning and sequencing of human fibrillarin, a conserved nucleolar protein recognized by autoimmune antisera. Proc Natl Acad Sci USA 1991; 88:931-935. 66. Tollervey D, Lehtonen H, Carmo-Fonseca M et al. The small nucleolar RNP protein NOP1 (fibrillarin) is required for pre-rRNA processing in yeast. EMBO J 1991; 10:573-583.

The snoRNPs and Related Machines

249

67. Gautier T, Berges T, Tollervey D et al. Nucleolar KKE/D repeat proteins Nop56p and Nop58p interact with Nop1p and are required for ribosome biogenesis. Mol Cell Biol 1997; 17:7088-7098. 68. Wu P, Brockenbrough JS, Metcalfe AC et al. Nop5p is a small nucleolar ribonucleoprotein component required for pre-18 S rRNA processing in yeast. J Biol Chem 1998; 273:16453-16463. 69. Lafontaine DL, Tollervey D. Nop58p is a common component of the box C+D snoRNPs that is required for snoRNA stability. RNA 1999; 5:455-467. 70. Lyman SK, Gerace L, Baserga SJ. Human Nop5/Nop58 is a component common to the box C/D small nucleolar ribonucleoproteins. RNA 1999; 5:1597-1604. 71. Filippini D, Bozzoni I, Caffarelli E. p62, a novel Xenopus laevis component of box C/D snoRNPs. RNA 2000; 6:391-401. 72. Watkins NJ, Segault V, Charpentier B et al. A common core RNP structure shared between the small nucleoar box C/D RNPs and the spliceosomal U4 snRNP. Cell 2000; 103:457-466. 73. Narayanan A, Speckmann W, Terns R et al. Role of the box C/D motif in localization of small nucleolar RNAs to coiled bodies and nucleoli. Mol Biol Cell 1999; 10:2131-2147. 74. Amiri KA. Fibrillarin-like proteins occur in the domain Archaea. J Bacteriol 1994; 176:2124-2127. 75. Newman DR, Kuhn JF, Shanab GM et al. Box C/D snoRNA-associated proteins: two pairs of evolutionarily ancient proteins and possible links to replication and transcription. RNA 2000; 6:861-879. 76. Kuhn JF, Tran EJ, Maxwell ES. Archaeal ribosomal protein L7 is a functional homolog of the eukaryotic 15.5kD/Snu13p snoRNP core protein. Nucleic Acids Res 2002; 30:931-941. 77. Omer AD, Ziesche S, Ebhardt H et al. In vitro reconstitution and activity of a C/D box methylation guide ribonucleoprotein complex. Proc Natl Acad Sci USA 2002; 99:5289-5294. 78. Klein RJ, Misulovin Z, Eddy SR. Noncoding RNA genes identified in AT-rich hyperthermophiles. Proc Natl Acad Sci USA 2002; 99:7542-7547. 79. Kolodrubetz D, Burgum A. Sequence and genetic analysis of NHP2: A moderately abundant high mobility group-like nuclear protein with an essential function in Saccharomyces cerevisiae. Yeast 1991; 7:79-90. 80. Koonin EV, Bork P, Sander C. A novel RNA-binding motif in omnipotent suppressors of translation termination, ribosomal proteins and a ribosome modification enzyme? Nucleic Acids Res 1994; 22:2166-2167. 81. Lafontaine DL, Tollervey D. Birth of the snoRNPs: the evolution of the modification-guide snoRNAs. Trends. BioChem Sci 1998; 23:383-388. 82. Watanabe Y, Gray MW. Evolutionary appearance of genes encoding proteins associated with box H/ACA snoRNAs: cbf5p in Euglena gracilis, an early diverging eukaryote, and candidate Gar1p and Nop10p homologs in archaebacteria. Nucleic Acids Res 2000; 28:2342-2352. 83. Bagni C, Lapeyre B. Gar1p binds to the small nucleolar RNAs snR10 and snR30 in vitro through a nontypical RNA binding element. J Biol Chem 1998; 273:10868-10873. 84. Dragon F, Pogacic V, Filipowicz W. In vitro assembly of human H/ACA small nucleolar RNPs reveals unique features of U17 and telomerase RNAs. Mol Cell Biol 2000; 20:3037-3048. 85. Watkins NJ, Newman DR, Kuhn JF et al. In vitro assembly of the mouse U14 snoRNP core complex and identification of a 65-kDa box C/D-binding protein. RNA 1998; 4:582-593. 86. Jady BE, Kiss T. A small nucleolar guide RNA functions both in 2'-O-ribose methylation and pseudouridylation of the U5 spliceosomal RNA. EMBO J 2001; 20:541-551. 87. Wang C, Query CC, Meier UT. Immunopurified small nucleolar ribonucleoprotein particles pseudouridylate rRNA independently of their association with phosphorylated Nopp140. Mol Cell Biol 2002; 22:8457-8466. 88. Galardi S, Fatica A, Bachi A et al. Purified box C/D snoRNPs are able to reproduce site-specific 2'-O-methylation of target RNA in vitro. Mol Cell Biol 2002; 22:6663-6668. 89. Chamberlain JR, Lee Y, Lane WS et al. Purification and characterization of the nuclear RNase P holoenzyme complex reveals extensive subunit overlap with RNase MRP. Genes Dev 1998; 12:1678-1690. 90. Dragon F, Gallagher JE, Compagnone-Post PA et al. A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 2002; 417:967-970. 91. Dez C, Noaillac-Depeyre J, Caizergues-Ferrer M et al. Naf1p, an essential nucleoplasmic factor specifically required for accumulation of box H/ACA small nucleolar RNPs. Mol Cell Biol 2002; 22:7053-7065. 92. Fatica A, Dlakic M, Tollervey D. Naf1p is a box H/ACA snoRNP assembly factor. RNA. 2002; 8:1502-1514. 93. Yang PK, Rotondo G, Porras T et al. The Shq1p:Naf1p complex is required for box H/ACA small nucleolar ribonucleoprotein particle biogenesis. J Biol Chem 2002; 277:45235-45242.

250

The Nucleolus

94. Lafontaine DL, Tollervey D. Synthesis and assembly of the box C+D small nucleolar RNPs. Mol Cell Biol 2000; 20:2650-2659. 95. Kiss-Laszlo Z, Henry Y, Kiss T. Sequence and structural elements of methylation guide snoRNAs essential for site-specific ribose methylation of pre-rRNA. EMBO J 1998; 17:797-807. 96. Vidovic I, Nottrott S, Hartmuth K et al. Crystal structure of the spliceosomal 15.5kD protein bound to a U4 snRNA fragment. Mol Cell 2000; 6:1331-1342. 97. Nottrott S, Hartmuth K, Fabrizio P et al. Functional interaction of a novel 15.5kD [U4/U6.U5] tri-snRNP protein with the 5' stem-loop of U4 snRNA. EMBO J 1999; 18:6119-6133. 98. Klein DJ, Schmeing TM, Moore PB et al. The kink-turn: a new RNA secondary structure motif. EMBO J 2001; 20:4214-4221. 99. Winkler WC, Grundy FJ, Murphy BA et al. The GA motif: an RNA element common to bacterial antitermination systems, rRNA, and eukaryotic RNAs. RNA. 2001; 7:1165-1172. 100. Baserga SJ, Yang XD, Steitz JA. An intact Box C sequence in the U3 snRNA is required for binding of fibrillarin, the protein common to the major family of nucleolar snRNPs. EMBO J 1991; 10:2645-2651. 101. Watkins NJ, Dickmanns A, Luhrmann R. Conserved stem II of the box C/D motif is essential for nucleolar localization and is required, along with the 15.5K protein, for the hierarchical assembly of the box C/D snoRNP. Mol Cell Biol 2002; 22:8342-8352. 102. Jarmolowski A, Zagorski J, Li HV et al. Identification of essential elements in U14 RNA of Saccharomyces cerevisiae. EMBO J 1990; 9:4503-4509. 103. Huang GM, Jarmolowski A, Struck JC et al. Accumulation of U14 small nuclear RNA in Saccharomyces cerevisiae requires box C, box D, and a 5', 3' terminal stem. Mol Cell Biol 1992; 12:4456-4463. 104. Kufel J, Allmang C, Chanfreau G et al. Precursors to the U3 small nucleolar RNA lack small nucleolar RNP proteins but are stabilized by La binding. Mol Cell Biol 2000; 20:5415-5424. 105. Xia L, Watkins NJ, Maxwell ES. Identification of specific nucleotide sequences and structural elements required for intronic U14 snoRNA processing. RNA 1997; 3:17-26. 106. Hirose T, Steitz JA. Position within the host intron is critical for efficient processing of box C/D snoRNAs in mammalian cells. Proc Natl Acad Sci USA 2001; 98:12914-12919. 107. Terns MP, Grimm C, Lund E et al. A common maturation pathway for small nucleolar RNAs. EMBO J 1995; 14:4860-4871. 108. Bachellerie JP, Nicoloso M, Qu LH et al. Novel intron-encoded small nucleolar RNAs with long sequence complementarities to mature rRNAs involved in ribosome biogenesis. Biochem Cell Biol 1995; 73:835-843. 109. Cavaille J, Nicoloso M, Bachellerie JP. Targeted ribose methylation of RNA in vivo directed by tailored antisense RNA guides. Nature. 1996; 383:732-735. 110. Lowe TM, Eddy SR. A computational screen for methylation guide snoRNAs in yeast. Science 1999; 283:1168-1171. 111. Li HD, Zagorski J, Fournier MJ. Depletion of U14 small nuclear RNA (snR128) disrupts production of 18S rRNA in Saccharomyces cerevisiae. Mol Cell Biol 1990; 10:1145-1152. 112. Dunbar DA, Baserga SJ. The U14 snoRNA is required for 2'-O-methylation of the pre-18S rRNA in Xenopus oocytes. RNA 1998; 4:195-204. 113. Niewmierzycka A, Clarke S. S-Adenosylmethionine-dependent methylation in Saccharomyces cerevisiae. Identification of a novel protein arginine methyltransferase. J Biol Chem 1999; 274:814-824. 114. Wang H, Boisvert D, Kim KK et al. Crystal structure of a fibrillarin homologue from Methanococcus jannaschii, a hyperthermophile, at 1.6 A resolution. EMBO J 2000; 19:317-323. 115. Tollervey D, Lehtonen H, Jansen R et al. Temperature-sensitive mutations demonstrate roles for yeast fibrillarin in pre-rRNA processing, pre-rRNA methylation, and ribosome assembly. Cell 1993; 72:443-457. 116. Ochs RL, Lischwe MA, Spohn WH et al. Fibrillarin: A new protein of the nucleolus identified by autoimmune sera. Biol Cell 1985; 54:123-133. 117. Bortolin ML, Ganot P, Kiss T. Elements essential for accumulation and function of small nucleolar RNAs directing site-specific pseudouridylation of ribosomal RNAs. EMBO J 1999; 18:457-469. 118. Bousquet-Antonelli C, Henry Y, G’Elugne JP et al. A small nucleolar RNP protein is required for pseudouridylation of eukaryotic ribosomal RNAs. EMBO J 1997; 16:4770-4776. 119. Watkins NJ, Gottschalk A, Neubauer G et al. Cbf5p, a potential pseudouridine synthase, and Nhp2p, a putative RNA-binding protein, are present together with Gar1p in all H BOX/ACA-motif snoRNPs and constitute a common bipartite structure. RNA 1998; 4:1549-1568. 120. Henras A, Henry Y, Bousquet-Antonelli C et al. Nhp2p and Nop10p are essential for the function of H/ACA snoRNPs. EMBO J 1998; 17:7078-7090.

The snoRNPs and Related Machines

251

121. Koonin EV. Pseudouridine synthases: Four families of enzymes containing a putative uridine-binding motif also conserved in dUTPases and dCTP deaminases. Nucleic Acids Res 1996; 24:2411-2415. 122. Zebarjadian Y, King T, Fournier MJ et al. Point mutations in yeast CBF5 can abolish in vivo pseudouridylation of rRNA. Mol Cell Biol 1999; 19:7461-7472. 123. Henras A, Dez C, Noaillac-Depeyre J et al. Accumulation of H/ACA snoRNPs depends on the integrity of the conserved central domain of the RNA-binding protein Nhp2p. Nucleic Acids Res 2001; 29:2733-2746. 124. Rozhdestvensky T, Tang T, Tchirkova I et al. Binding of L7Ae protein to the K-turn of archaeal snoRNAs: A shared RNA binding motif for C/D and H/ACA box snoRNAs in Archaea. Nucleic Acids Res 2003; 31:869-877. 125. Lubben B, Fabrizio P, Kastner B et al. Isolation and characterization of the small nucleolar ribonucleoprotein particle snR30 from Saccharomyces cerevisiae. J Biol Chem 1995; 270:11549-11554. 126. Kiss T, Marshallsay C, Filipowicz W. 7-2/MRP RNAs in plant and mammalian cells: Association with higher order structures in the nucleolus. EMBO J 1992; 11:3737-3746. 127. Jacobson MR, Cao LG, Wang YL et al. Dynamic localization of RNase MRP RNA in the nucleolus observed by fluorescent RNA cytochemistry in living cells. J Cell Biol 1995; 131:1649-1658. 128. Lee B, Matera AG, Ward DC et al. Association of RNase mitochondrial RNA processing enzyme with ribonuclease P in higher ordered structures in the nucleolus: a possible coordinate role in ribosome biogenesis. Proc Natl Acad Sci USA 1996; 93:11471-11476. 129. Shuai K, Warner JR. A temperature sensitive mutant of Saccharomyces cerevisiae defective in pre-rRNA processing. Nucleic Acids Res 1991; 19:5059-5064. 130. Lindahl L, Archer RH, Zengel JM. A new rRNA processing mutant of Saccharomyces cerevisiae. Nucleic Acids Res 1992; 20:295-301. 131. Schmitt ME, Clayton DA. Nuclear RNase MRP is required for correct processing of pre-5.8S rRNA in Saccharomyces cerevisiae. Mol Cell Biol 1993; 13:7935-7941. 132. Schmitt ME, Bennett JL, Dairaghi DJ et al. Secondary structure of RNase MRP RNA as predicted by phylogenetic comparison. FASEB J 1993; 7:208-213. 133. Schmitt ME. Molecular modeling of the three-dimensional architecture of the RNA component of yeast RNase MRP. J Mol Biol 1999; 292:827-836. 134. Li X, Frank DN, Pace N et al. Phylogenetic analysis of the structure of RNase MRP RNA in yeasts. RNA. 2002; 8:740-751. 135. Lygerou Z, Allmang C, Tollervey D et al. Accurate processing of a eukaryotic precursor ribosomal RNA by ribonuclease MRP in vitro. Science 1996; 272:268-270. 136. Lygerou Z, Mitchell P, Petfalski E et al. The POP1 gene encodes a protein component common to the RNase MRP and RNase P ribonucleoproteins. Genes Dev 1994; 8:1423-1433. 137. Schmitt ME, Clayton DA. Characterization of a unique protein component of yeast RNase MRP: an RNA-binding protein with a zinc-cluster domain. Genes Dev 1994; 8:2617-2628. 138. Chu S, Zengel JM, Lindahl L. A novel protein shared by RNase MRP and RNase P. RNA 1997; 3:382-391. 139. Dichtl B, Tollervey D. Pop3p is essential for the activity of the RNase MRP and RNase P ribonucleoproteins in vivo. EMBO J 1997; 16:417-429. 140. Stolc V, Altman S. Rpp1, an essential protein subunit of nuclear RNase P required for processing of precursor tRNA and 35S precursor rRNA in Saccharomyces cerevisiae. Genes Dev 1997; 11:2926-2937. 141. Stolc V, Katz A, Altman S. Rpp2, an essential protein subunit of nuclear RNase P, is required for processing of precursor tRNAs and 35S precursor rRNA in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1998; 95:6716-6721. 142. van Eenennaam H, Jarrous N, van Venrooij WJ et al. Architecture and function of the human endonucleases RNase P and RNase MRP. IUBMB Life 2000; 49:265-272. 143. Xiao S, Scott F, Fierke CA et al. Eukaryotic ribonuclease P: a plurality of ribonucleoprotein enzymes. Annu Rev Biochem 2002; 71:165-189. 144. Jarrous N, Wolenski JS, Wesolowski D et al. Localization in the nucleolus and coiled bodies of protein subunits of the ribonucleoprotein ribonuclease P. J Cell Biol 1999; 146:559-572. 145. Bertrand E, Houser-Scott F, Kendall A et al. Nucleolar localization of early tRNA processing. Genes Dev 1998; 12:2463-2468. 146. Kendall A, Hull MW, Bertrand E et al. A CBF5 mutation that disrupts nucleolar localization of early tRNA biosynthesis in yeast also suppresses tRNA gene-mediated transcriptional silencing. Proc Natl Acad Sci USA 2000; 97:13108-13113. 147. Jacobson MR, Cao LG, Taneja K et al. Nuclear domains of the RNA subunit of RNase P. J Cell Sci 1997; 110:829-837. 148. Wolin SL, Matera AG. The trials and travels of tRNA. Genes Dev 1999; 13:1-10.

252

The Nucleolus

149. Cai T, Aulds J, Gill T et al. The Saccharomyces cerevisiae RNase mitochondrial RNA processing is critical for cell cycle progression at the end of mitosis. Genetics 2002; 161:1029-1042. 150. Ridanpaa M, van Eenennaam H, Pelin K et al. Mutations in the RNA component of RNase MRP cause a pleiotropic human disease, cartilage-hair hypoplasia. Cell 2001; 104:195-203. 151. Bonafe L, Schmitt K, Eich G et al. RMRP gene sequence analysis confirms a cartilage-hair hypoplasia variant with only skeletal manifestations and reveals a high density of single-nucleotide polymorphisms. Clin Genet 2002; 61:146-151. 152. Yuan Y, Singh R, Reddy R. Rat nucleolar 7-2 RNA is homologous to mouse mitochondrial RNase mitochondrial RNA-processing RNA. J Biol Chem 1989; 264:14835-14839. 153. Van Eenennaam H, Vogelzangs JH, Lugtenberg D et al. Identity of the RNase MRP- and RNase P-associated Th/To autoantigen. Arthritis Rheum 2002; 46:3266-3272. 154. Venema J, Tollervey D. Ribosome synthesis in Saccharomyces cerevisiae. Annu Rev Genet 1999; 33:261-311. 155. Hughes JM, Ares M, Jr. Depletion of U3 small nucleolar RNA inhibits cleavage in the 5' external transcribed spacer of yeast pre-ribosomal RNA and impairs formation of 18S ribosomal RNA. EMBO J 1991; 10:4231-4239. 156. Enright CA, Maxwell ES, Eliceiri GL et al. 5’ETS rRNA processing facilitated by four small RNAs: U14, E3, U17, and U3. RNA. 1996; 2:1094-1099. 157. Tycowski KT, Shu MD, Steitz JA. Requirement for intron-encoded U22 small nucleolar RNA in 18S ribosomal RNA maturation. Science. 1994; 266:1558-1561. 158. Peculis BA. The sequence of the 5' end of the U8 small nucleolar RNA is critical for 5.8S and 28S rRNA maturation. Mol Cell Biol 1997; 17:3702-3713. 159. Mougey EB, O’Reilly M, Osheim Y et al. The terminal balls characteristic of eukaryotic rRNA transcription units in chromatin spreads are rRNA processing complexes. Genes Dev 1993; 7:1609-1619. 160. Grandi P, Rybin V, Bassler J et al. 90S pre-ribosomes include the 35S pre-rRNA, the U3 snoRNP, and 40S subunit processing factors but predominantly lack 60S synthesis factors. Mol Cell 2002; 10:105-115. 161. Steitz JA, Tycowski KT. Small RNA chaperones for ribosome biogenesis. Science 1995; 270:1626-1627. 162. Ni J The major function of the small nucleolar RNAs is nucleotide modification in ribosomal RNA.: University of Massachusetts at Amherst; 1998. 163. Ni J, Fournier MJ. Sequence-specific methylation of ribonucleic acid. U.S. patent #5,972,705; 1999. 164. Nicoloso M, Qu LH, Michot B et al. Intron-encoded, antisense small nucleolar RNAs: The characterization of nine novel species points to their direct role as guides for the 2'-O-ribose methylation of rRNAs. J Mol Biol 1996; 260:178-195. 165. Ni J, Fournier MJ. Site-specific synthesis of pseudouridine in RNA. U.S. patent #5,989,911; 1999. 166. Ofengand J, Bakin A. Mapping to nucleotide resolution of pseudouridine residues in large subunit ribosomal RNAs from representative eukaryotes, prokaryotes, archaebacteria, mitochondria and chloroplasts. J Mol Biol 1997; 266:246-268. 167. Ofengand J, Fournier MJ. The pseudouridine residues of rRNA: number, location, biosynthesis, and function. In: Grosjean H, Benne R, eds. Modification and Editing of RNA. Washington, DC: ASM Press; 1998:229-253. 168. Maden BE. The numerous modified nucleotides in eukaryotic ribosomal RNA. Prog. Nucleic Acid Res Mol Biol 1990; 39:241-303. 169. Samarsky DA, Fournier MJ. A comprehensive database for the small nucleolar RNAs from Saccharomyces cerevisiae. Nucleic Acids Res 1999; 27:161-164. 170. Tycowski KT, You ZH, Graham PJ et al. Modification of U6 spliceosomal RNA is guided by other small RNAs. Mol Cell 1998; 2:629-638. 171. Ganot P, Jady BE, Bortolin ML et al. Nucleolar factors direct the 2'-O-ribose methylation and pseudouridylation of U6 spliceosomal RNA. Mol Cell Biol 1999; 19:6906-6917. 172. Zhou H, Chen YQ, Du YP et al. The Schizosaccharomyces pombe mgU6-47 gene is required for 2'-O-methylation of U6 snRNA at A41. Nucleic Acids Res 2002; 30:894-902. 173. Davis DR. Biophysical and conformational properties of modified nucleosides in RNA (nuclear magnetic resonance studies). In: Grosjean H, Benne R, eds. Modification and Editing of RNA. Washington, DC: AMS press; 1998:85-102. 174. Decatur WA, Fournier MJ. rRNA modifications and ribosome function. Trends Biochem Sci 2002; 27:344-351. 175. Hansen MA, Kirpekar F, Ritterbusch W et al. Posttranscriptional modifications in the A-loop of 23S rRNAs from selected archaea and eubacteria. RNA. 2002; 8:202-213.

The snoRNPs and Related Machines

253

176. Ofengand J. Ribosomal RNA pseudouridines and pseudouridine synthases. FEBS Lett 2002; 514:17-25. 177. King TH, Liu B, McCully RR et al. Ribosome structure and activity are altered in cells lacking snoRNPs that form pseudouridines in the peptidyl transferase center. Molecular Cell 2003; 11:425-435. 178. Liang WQ, Fournier MJ. U14 base-pairs with 18S rRNA: A novel snoRNA interaction required for rRNA processing. Genes Dev 1995; 9:2433-2443. 179. Michot B, Joseph N, Mazan S et al. Evolutionarily conserved structural features in the ITS2 of mammalian pre-rRNAs and potential interactions with the snoRNA U8 detected by comparative analysis of new mouse sequences. Nucleic Acids Res 1999; 27:2271-2282. 180. Beltrame M, Henry Y, Tollervey D. Mutational analysis of an essential binding site for the U3 snoRNA in the 5' external transcribed spacer of yeast pre-rRNA. Nucleic Acids Res 1994; 22:5139-5147. 181. Beltrame M, Tollervey D. Base pairing between U3 and the pre-ribosomal RNA is required for 18S rRNA synthesis. EMBO J 1995; 14:4350-4356. 182. Hughes JM. Functional base-pairing interaction between highly conserved elements of U3 small nucleolar RNA and the small ribosomal subunit RNA. J Mol Biol 1996; 259:645-654. 183. Sharma K, Tollervey D. Base pairing between U3 small nucleolar RNA and the 5' end of 18S rRNA is required for pre-rRNA processing. Mol Cell Biol 1999; 19:6012-6019. 184. Intine RV, Good L, Nazar RN. Essential structural features in the Schizosaccharomyces pombe pre-rRNA 5' external transcribed spacer. J Mol Biol 1999; 286:695-708. 185. Hartshorne T, Toyofuku W. Two 5'-ETS regions implicated in interactions with U3 snoRNA are required for small subunit rRNA maturation in Trypanosoma brucei. Nucleic Acids Res 1999; 27:3300-3309. 186. Brand RC, Klootwijk J, Van Steenbergen TJ et al. Secondary methylation of yeast ribosomal precursor RNA. Eur J Biochem 1977; 75:311-318. 187. Retel J, van den Bos RC, Planta RJ. Characteristics of the methylation in vivo of ribosomal RNA in yeast. Biochim Biophys Acta 1969; 195:370-380. 188. Udem SA, Warner JR. Ribosomal RNA synthesis in Saccharomyces cerevisiae. J Mol Biol 1972; 65:227-242. 189. Liau MC, Hurlbert RB. The topographical order of 18S and 20S ribosomal ribonucleic acids within the 45S precursor molecule. J Mol Biol 1975; 98:321-332. 190. Maden BE, Salim M. The methylated nucleotide sequences in HELA Cell ribosomal RNA and its precursors. J Mol Biol 1974; 88:133-152. 191. Penman S, Smith I, Holtzman E et al. RNA metabolism in the HeLa cell nucleus and nucleolus. Natl Cancer Inst Monogr 1966; 23:489-512. 192. Muramatsu M, Fujisawa T. Methylation of ribosomal RNA precursor and tRNA in rat liver. Biochim Biophys Acta 1968; 157:476-492. 193. Saveanu C, Bienvenu D, Namane A et al. Nog2p, a putative GTPase associated with pre-60S subunits and required for late 60S maturation steps. EMBO J 2001; 20:6475-6484. 194. Harnpicharnchai P, Jakovljevic J, Horsey E et al. Composition and functional characterization of yeast 66S ribosome assembly intermediates. Mol Cell 2001; 8:505-515. 195. Bassler J, Grandi P, Gadal O et al. Identification of a 60S preribosomal particle that is closely linked to nuclear export. Mol Cell 2001; 8:517-529. 196. Fatica A, Cronshaw AD, Dlakic M et al. Ssf1p prevents premature processing of an early pre-60S ribosomal particle. Mol Cell 2002; 9:341-351. 197. Nissan TA, Bassler J, Petfalski E et al. 60S pre-ribosome formation viewed from assembly in the nucleolus until export to the cytoplasm. EMBO J 2002; 21:5539-5547. 198. Fujiyama S, Yanagida M, Hayano T et al. Isolation and proteomic characterization of human Parvulin-associating preribosomal ribonucleoprotein complexes. J Biol Chem 2002; 277:23773-23780. 199. Fatica A, Tollervey D. Making ribosomes. Curr Opin Cell Biol 2002; 14:313-318. 200. Verschoor A, Warner JR, Srivastava S et al. Three-dimensional structure of the yeast ribosome. Nucleic Acids Res 1998; 26:655-661. 201. Sirum-Connolly K, Mason TL. Functional requirement of a site-specific ribose methylation in ribosomal RNA. Science 1993; 262:1886-1889. 202. Lafontaine DL, Preiss T, Tollervey D. Yeast 18S rRNA dimethylase Dim1p: A quality control mechanism in ribosome synthesis? Mol Cell Biol 1998; 18:2360-2370. 203. Gutgsell NS, Del Campo MD, Raychaudhuri S et al. A second function for pseudouridine synthases: A point mutant of RluD unable to form pseudouridines 1911, 1915, and 1917 in Escherichia coli 23S ribosomal RNA restores normal growth to an RluD-minus strain. RNA 2001; 7:990-998.

254

The Nucleolus

204. Mitchell JR, Cheng J, Collins K. A box H/ACA small nucleolar RNA-like domain at the human telomerase RNA 3' end. Mol Cell Biol 1999; 19:567-576. 205. Rozenski J, Crain PF, McCloskey JA. The RNA modification database: 1999 update. Nucleic Acids Res 1999; 27:196-197. 206. Noon KR, Bruenger E, McCloskey JA. Posttranscriptional modifications in 16S and 23S rRNAs of the archaeal hyperthermophile Sulfolobus solfataricus. J Bacteriol 1998; 180:2883-2888. 207. Bult CJ, White O, Olsen GJ et al. Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii. Science 1996; 273:1058-1073. 208. Hickey AJ, Macario AJ, Conway de Macario E. Identification of genes in the genome of the archaeon Methanosarcina mazeii that code for homologs of nuclear eukaryotic molecules involved in RNA processing. Gene 2000; 253:77-85. 209. Speckmann WA, Li ZH, Lowe TM et al. Archaeal guide RNAs function in rRNA modification in the eukaryotic nucleus. Curr Biol 2002; 12:199-203. 210. Tang TH, Rozhdestvensky TS, d’Orval BC et al. RNomics in Archaea reveals a further link between splicing of archaeal introns and rRNA processing. Nucleic Acids Res 2002; 30:921-930. 211. Kowalak JA, Bruenger E, Crain PF et al. Identities and phylogenetic comparisons of posttranscriptional modifications in 16 S ribosomal RNA from Haloferax volcanii. J Biol Chem 2000; 275:24484-24489. 212. Liang XH, Liu L, Michaeli S. Identification of the first trypanosome H/ACA RNA that guides pseudouridine formation on rRNA. J Biol Chem 2001; 276:40313-40318. 213. Carmo-Fonseca M, Pepperkok R, Carvalho M et al. Transcription-dependent colocalization of the U1, U2, U4/U6, and U5 snRNPs in coiled bodies. J Cell Biol 1992; 117:1-14. 214. Matera AG. Nuclear bodies: Multifaceted subdomains of the interchromatin space. Trends Cell Biol 1999; 9:302-309. 215. Sleeman JE, Lamond AI. Nuclear organization of pre-mRNA splicing factors. Curr Opin Cell Biol 1999; 11:372-377. 216. Gall JG. Cajal bodies: The first 100 years. Annu Rev Cell Dev Biol 2000; 16:273-300. 217. Lewis JD, Tollervey D. Like attracts like: Getting RNA processing together in the nucleus. Science. 2000; 288:1385-1389. 218. Sleeman JE, Lamond AI. Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway. Curr Biol 1999; 9:1065-1074. 219. Jady BE, Darzacq X, Tucker KE et al. Modification of Sm small nuclear RNAs occurs in the nucleoplasmic Cajal body following import from the cytoplasm. EMBO J 2003; 22:1878-1888. 220. Gerbi SA, Lange TS. All small nuclear RNAs (snRNAs) of the [U4/U6.U5] tri-snRNP localize to nucleoli; Identification of the nucleolar localization element of U6 snRNA. Mol Biol Cell 2002; 13:3123-3137. 221. Massenet S, Motorin Y, Lafontaine D et al. Pseudouridine mapping in the Saccharomyces cerevisiae spliceosomal U small nuclear RNAs (snRNAs) reveals that pseudouridine synthase pus1p exhibits a dual substrate specificity for U2 snRNA and tRNA. Mol Cell Biol 1999; 19:2142-2154. 222. Lukowiak AA, Narayanan A, Li ZH et al. The snoRNA domain of vertebrate telomerase RNA functions to localize the RNA within the nucleus. RNA 2001; 7:1833-1844. 223. Mitchell JR, Wood E, Collins K. A telomerase component is defective in the human disease dyskeratosis congenita. Nature 1999; 402:551-555. 224. Pogacic V, Dragon F, Filipowicz W. Human H/ACA small nucleolar RNPs and telomerase share evolutionarily conserved proteins NHP2 and NOP10. Mol Cell Biol 2000; 20:9028-9040. 225. Seto AG, Zaug AJ, Sobel SG, Wolin SL, Cech TR. Saccharomyces cerevisiae telomerase is an Sm small nuclear ribonucleoprotein particle. Nature 1999; 401:177-180. 226. Chen JL, Blasco MA, Greider CW. Secondary structure of vertebrate telomerase RNA. Cell 2000; 100:503-514. 227. Antal M, Boros E, Solymosy F et al. Analysis of the structure of human telomerase RNA in vivo. Nucleic Acids Res 2002; 30:912-920. 228. Wong JM, Kusdra L, Collins K. Subnuclear shuttling of human telomerase induced by transformation and DNA damage. Nat Cell Biol 2002; 4:731-736. 229. Heiss NS, Knight SW, Vulliamy TJ et al. X-linked dyskeratosis congenita is caused by mutations in a highly conserved gene with putative nucleolar functions. Nat Genet 1998; 19:32-38. 230. Vulliamy T, Marrone A, Goldman F et al. The RNA component of telomerase is mutated in autosomal dominant dyskeratosis congenita. Nature 2001; 413:432-435. 231. Luzzatto L, Karadimitris A. Dyskeratosis and ribosomal rebellion. Nat Genet 1998; 19:6-7. 232. Shay JW, Wright WE. Mutant dyskerin ends relationship with telomerase. Science 1999; 286:2284-2285.

The snoRNPs and Related Machines

255

233. Marciniak RA, Johnson FB, Guarente L. Dyskeratosis congenita, telomeres and human ageing. Trends Genet 2000; 16:193-195. 234. Collins K, Mitchell JR. Telomerase in the human organism. Oncogene 2002; 21:564-579. 235. Ruggero D, Grisendi S, Piazza F et al. Dyskeratosis congenita and cancer in mice deficient in ribosomal RNA modification. Science 2003; 299:259-262. 236. Cavaille J, Vitali P, Basyuk E et al. A novel brain-specific box C/D small nucleolar RNA processed from tandemly repeated introns of a noncoding RNA gene in rats. J Biol Chem 2001; 276:26374-26383. 237. Huttenhofer A, Kiefmann M, Meier-Ewert S et al. RNomics: an experimental approach that identifies 201 candidates for novel, small, non-messenger RNAs in mouse. EMBO J 2001; 20:2943-2953. 238. Cavaille J, Seitz H, Paulsen M et al. Identification of tandemly-repeated C/D snoRNA genes at the imprinted human 14q32 domain reminiscent of those at the Prader-Willi/Angelman syndrome region. Hum. Mol Genet 2002; 11:1527-1538. 239. Shimoda M, Morita S, Obata Y et al. Imprinting of a small nucleolar RNA gene on mouse chromosome 12. Genomics 2002; 79:483-486. 240. Neumann B, Kubicka P, Barlow DP. Characteristics of imprinted genes. Nat Genet 1995; 9:12-13. 241. Rougeulle C, Heard E. Antisense RNA in imprinting: Spreading silence through air. Trends Genet 2002; 18:434-437. 242. Sleutels F, Zwart R, Barlow DP. The non-coding Air RNA is required for silencing autosomal imprinted genes. Nature 2002; 415:810-813. 243. Wutz A, Smrzka OW, Schweifer N et al. Imprinted expression of the Igf2r gene depends on an intronic CpG island. Nature 1997; 389:745-749. 244. Horike S, Mitsuya K, Meguro M et al. Targeted disruption of the human LIT1 locus defines a putative imprinting control element playing an essential role in Beckwith-Wiedemann syndrome. Hum Mol Genet 2000; 9:2075-2083. 245. Yi-Brunozzi HY, Easterwood LM, Kamilar GM et al. Synthetic substrate analogs for the RNA-editing adenosine deaminase ADAR-2. Nucleic Acids Res 1999; 27:2912-2917. 246. Teixeira MT, Forstemann K, Gasser SM et al. Intracellular trafficking of yeast telomerase components. EMBO Rep 2002; 3:652-659. 247. Phair RD, Misteli T. High mobility of proteins in the mammalian cell nucleus. Nature 2000; 404:604-609. 248. Warner JR. Nascent ribosomes. Cell 2001; 107:133-136. 249. Weinstein LB, Steitz JA. Guided tours: From precursor snoRNA to functional snoRNP. Curr Opin Cell Biol 1999; 11:378-384. 250. Kressler D, Linder P, de La Cruz J. Protein trans-acting factors involved in ribosome biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:7897-7912. 251. Filipowicz W, Kiss T. Structure and function of nucleolar snRNPs. Mol Biol Rep 1993; 18:149-156. 252. Sollner-Webb B. Novel intron-encoded small nucleolar RNAs. Cell 1993; 75:403-405. 253. Gerbi SA. Small nucleolar RNA. Biochem Cell Biol 1995; 73:845-858. 254. Smith CM, Steitz JA. Sno storm in the nucleolus: new roles for myriad small RNPs [comment]. Cell 1997; 89:669-672. 255. Dahlberg JE, Lund E. The gene and the transcription of the major small nuclear RNAs. In: Birnstiel ML, ed. Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles. New York: Springer-Verlag, 1988:1-37. 256. Tycowski KT, Shu MD, Steitz JA. A mammalian gene with introns instead of exons generating stable RNA products. Nature 1996; 379:464-466. 257. Bortolin ML, Kiss T. Human U19 intron-encoded snoRNA is processed from a long primary transcript that possesses little potential for protein coding. RNA 1998; 4:445-454. 258. Pelczar P, Filipowicz W. The host gene for intronic U17 small nucleolar RNAs in mammals has no protein-coding potential and is a member of the 5'-terminal oligopyrimidine gene family. Mol Cell Biol 1998; 18:4509-4518. 259. Smith CM, Steitz JA. Classification of gas5 as a multi-small-nucleolar-RNA (snoRNA) host gene and a member of the 5'-terminal oligopyrimidine gene family reveals common features of snoRNA host genes. Mol Cell Biol 1998; 18:6897-6909. 260. Tycowski KT, Steitz JA. Non-coding snoRNA host genes in Drosophila: Expression strategies for modification guide snoRNAs. Eur J Cell Biol 2001; 80:119-125. 261. Chang LS, Lin SY, Lieu AS et al. Differential expression of human 5S snoRNA genes. Biochem Biophys Res Commun 2002; 299:196-200. 262. Qu LH, Henras A, Lu YJ et al. Seven novel methylation guide small nucleolar RNAs are processed from a common polycistronic transcript by Rat1p and RNase III in yeast. Mol Cell Biol 1999; 19:1144-1158.

256

The Nucleolus

263. Leader DJ, Clark GP, Watters J et al. Clusters of multiple different small nucleolar RNA genes in plants are expressed as and processed from polycistronic pre-snoRNAs. EMBO J 1997; 16:5742-5751. 264. Qu LH, Meng Q, Zhou H et al. Identification of 10 novel snoRNA gene clusters from Arabidopsis thaliana. Nucleic Acids Res 2001; 29:1623-1630. 265. Barneche F, Gaspin C, Guyot R et al. Identification of 66 box C/D snoRNAs in Arabidopsis thaliana: Extensive gene duplications generated multiple isoforms predicting new ribosomal RNA 2'-O-methylation sites. J Mol Biol 2001; 311:57-73. 266. Brown JW, Clark GP, Leader DJ et al. Multiple snoRNA gene clusters from Arabidopsis. RNA. 2001; 7:1817-1832. 267. Liang D, Zhou H, Zhang P et al. A novel gene organization: Intronic snoRNA gene clusters from Oryza sativa. Nucleic Acids Res 2002; 30:3262-3272. 268. Marker C, Zemann A, Terhorst T et al. Experimental RNomics. Identification of 140 candidates for small non-messenger RNAs in the plant Arabidopsis thaliana. Curr Biol 2002; 12:2002-2013. 269. Brown JW, Echeverria M, Qu LH et al. Plant snoRNA database. Nucleic Acids Res 2003; 31:432-435. 270. Brown JW, Echeverria M, Qu LH. Plant snoRNAs: Functional evolution and new modes of gene expression. Trends Plant Sci 2003; 8:42-49. 271. Kruszka K, Barneche F, Guyot R et al. Plant dicistronic tRNA-snoRNA genes: a new mode of expression of the small nucleolar RNAs processed by RNase Z. EMBO J 2003; 22:621-632. 272. Chanfreau G, Legrain P, Jacquier A. Yeast RNase III as a key processing enzyme in small nucleolar RNAs metabolism. J Mol Biol 1998; 284:975-988. 273. Chanfreau G, Rotondo G, Legrain P et al. Processing of a dicistronic small nucleolar RNA precursor by the RNA endonuclease Rnt1. EMBO J 1998; 17:3726-3737. 274. Petfalski E, Dandekar T, Henry Y et al. Processing of the precursors to small nucleolar RNAs and rRNAs requires common components. Mol Cell Biol 1998; 18:1181-1189. 275. Allmang C, Kufel J, Chanfreau G et al. Functions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J 1999; 18:5399-5410. 276. van Hoof A, Lennertz P, Parker R. Yeast exosome mutants accumulate 3'-extended polyadenylated forms of U4 small nuclear RNA and small nucleolar RNAs. Mol Cell Biol 2000; 20:441-452. 277. Fatica A, Morlando M, Bozzoni I. Yeast snoRNA accumulation relies on a cleavage-dependent/ polyadenylation-independent 3'-processing apparatus. EMBO J 2000; 19:6218-6229. 278. Ooi SL, Samarsky DA, Fournier MJ et al. Intronic snoRNA biosynthesis in Saccharomyces cerevisiae depends on the lariat-debranching enzyme: Intron length effects and activity of a precursor snoRNA. RNA. 1998; 4:1096-1110. 279. Villa T, Ceradini F, Presutti C et al. Processing of the intron-encoded U18 small nucleolar RNA in the yeast Saccharomyces cerevisiae relies on both exo- and endonucleolytic activities. Mol Cell Biol 1998; 18:3376-3383. 280. Reddy R, Ro-Choi TS, Henning D et al. Modified nucleosides of nuclear and nucleolar low molecular weight ribonucleic acid. J Biol Chem 1972; 247:7245-7250. 281. Terns MP, Dahlberg JE. Retention and 5' cap trimethylation of U3 snRNA in the nucleus. Science 1994; 264:959-961. 282. Myslinski E, Segault V, Branlant C. An intron in the genes for U3 small nucleolar RNAs of the yeast Saccharomyces cerevisiae. Science 1990; 247:1213-1216. 283. Chanfreau G, Buckle M, Jacquier A. Recognition of a conserved class of RNA tetraloops by Saccharomyces cerevisiae RNase III. Proc Natl Acad Sci USA 2000; 97:3142-3147. 284. Giorgi C, Fatica A, Nagel R et al. Release of U18 snoRNA from its host intron requires interaction of Nop1p with the Rnt1p endonuclease. EMBO J 2001; 20:6856-6865. 285. Steinmetz E, Conrad N, Brow D et al. RNA-binding protein Nrd1 directs poly(A)-independent 3'-end formation of RNA polymerase II transcripts. Nature 2001; 413:327-331. 286. Cahill N, Friend K, Speckmann W et al. Site-specific cross-linking analyses reveal an asymmetric protein distribution for a box C/D snoRNP. EMBO J 2002; 21:3816-3828. 287. Aittaleb M, Rashid R, Chen Q et al. Structure and function of archaeal box C/D sRNP core proteins. Nat Struct Biol 2003; 10:256-263. 288. King TH, Decatur WA, Bertrand E et al. A well-connected and conserved nucleoplasmic helicase is required for production of box C/D and H/ACA snoRNAs and localization of snoRNP proteins. Mol Cell Biol 2001; 21:7731-7746. 289. Jones K, Gorzynski K, Hales C et al. Direct interaction of the spinal muscular atrophy disease protein SMN with the small nucleolar RNA-associated protein fibrillarin. J Biol Chem 2001; 276:38645-38651. 290. Pellizzoni L, Baccon J, Charroux B et al. The survival of motor neurons (SMN) protein interacts with the snoRNP proteins fibrillarin and GAR1. Curr Biol 2001; 11:1079-1088.

The snoRNPs and Related Machines

257

291. Terns MP, Terns RM. Macromolecular complexes: SMN—the master assembler. Curr Biol 2001; 11:R862-864. 292. Paushkin S, Gubitz A, Massenet S et al. The SMN complex, an assemblyosome of ribonucleoproteins. Curr Opin Cell Biol 2002; 14:305-312. 293. Fischer U, Liu Q, Dreyfuss G. The SMN-SIP1 complex has an essential role in spliceosomal snRNP biogenesis. Cell 1997; 90:1023-1029. 294. Liu Q, Fischer U, Wang F et al. The spinal muscular atrophy disease gene product, SMN, and its associated protein SIP1 are in a complex with spliceosomal snRNP proteins. Cell 1997; 90:1013-1021. 295. Carvalho T, Almeida F, Calapez A et al. The spinal muscular atrophy disease gene product, SMN: A link between snRNP biogenesis and the Cajal (coiled) body. J Cell Biol 1999; 147:715-728. 296. Tremblay A, Lamontagne B, Catala M et al. A physical interaction between Gar1p and Rnt1p is required for the nuclear import of H/ACA small nucleolar RNA-associated proteins. Mol Cell Biol 2002; 22:4792-4802. 297. Bohmann K, Ferreira J, Lamond A. Mutational analysis of p80 coilin indicates a functional interaction between coiled bodies and the nucleolus. J Cell Biol 1995; 131:817-831. 298. Ochs R, Stein TJ, Tan E. Coiled bodies in the nucleolus of breast cancer cells. J Cell Sci 1994; 107:385-399. 299. Buonomo SB, Michienzi A, De Angelis FG et al. The Rev protein is able to transport to the cytoplasm small nucleolar RNAs containing a Rev binding element. RNA 1999; 5:993-1002. 300. Michienzi A, Li S, Zaia JA et al. A nucleolar TAR decoy inhibitor of HIV-1 replication. Proc Natl Acad Sci USA 2002; 99:14047-14052. 301. Samarsky DA, Ferbeyre G, Bertrand E et al. A small nucleolar RNA:ribozyme hybrid cleaves a nucleolar RNA target in vivo with near-perfect efficiency. Proc Natl Acad Sci USA 1999; 96:6609-6614. 302. Michienzi A, Cagnon L, Bahner I et al. Ribozyme-mediated inhibition of HIV 1 suggests nucleolar trafficking of HIV-1 RNA. Proc Natl Acad Sci USA 2000; 97:8955-8960. 303. Liu B, Ni J, Fournier MJ. Probing RNA in vivo with methylation guide small nucleolar RNAs. Methods 2001; 23:276-286.

258

The Nucleolus

CHAPTER 14

Ribosomal Subunit Assembly Jesús de la Cruz, Dieter Kressler and Patrick Linder

Summary

T

he biogenesis of eukaryotic ribosomes is a very dynamic process, which requires a precise spatial and temporal coordination of processing, modification and assembly events that take place within preribosomal particles travelling from the nucleolus to the cytoplasm. Most of our current knowledge on ribosome biogenesis comes from studies with the yeast Saccharomyces cerevisiae. The combined use of powerful proteomic and yeast genetic tools has led to the identification of many of the different nonribosomal factors involved in the process. These factors, termed trans-acting factors, include the enzymes and their cofactors that process and modify the precursors of the ribosomal RNAs, the adapters for nuclear export of preribosomal subunits and a large set of factors of unclear function, generally referred to as ribosomal subunit assembly factors. The ribosomal subunit assembly factors can be grouped into two categories. First, energy-consuming enzymes, such as RNA helicases, AAA ATPases or GTPases, that promote the modulation of RNA-RNA, RNA-protein and protein-protein interactions during maturation of ribosomal subunits. Second, nonenergy consuming proteins that are likely required to stabilize appropriate RNA-protein structures, prevent unproductive interactions and/or help r-proteins and other trans-acting factors to assemble. The recent advances in the biochemical purification of distinct preribosomal particles have not only provided a refined model of the ribosomal subunit assembly pathway but also a handle to tackle the current challenges in the field, which consist in the identification of the precise location of trans-acting factors within preribosomal particles and the understanding of their specific substrates and timing of action.

Introduction Ribosomes are ubiquitous ribonucleoprotein particles (RNPs) that are responsible for the fundamental process of protein synthesis.1 Eukaryotic 80S ribosomes are composed of a small 40S and a large 60S ribosomal subunit (r-subunit). In the yeast Saccharomyces cerevisiae, the 40S r-subunit contains one rRNA (18S) and 32 unique r-proteins while the 60S r-subunit is composed of three rRNA species (25S, 5.8S and 5S) and 46 r-proteins.2 Recently, the structure of the 80S ribosome from S. cerevisiae has been solved at about 15 Å resolution by cryo-electron microscopy reconstruction,3 and it reveals that the cores of the eukaryotic r-subunits are similar to those of the bacterial and archaebacterial r-subunits.4,5 This reinforces the notion that the basic mechanism of protein synthesis is highly conserved throughout all kingdoms. Despite the structural and functional conservation, eukaryotes seem to have evolved a distinct mechanism for the biogenesis of ribosomes. In the early 1970s, some laboratories were able to promote bacterial ribosomal self-assembly in vitro using purified rRNA and r-protein components.6,7 This has also been accomplished with archaebacteria.8,9 More recently, experimental evidence suggests the existence of a few nonribosomal factors (so-called trans-acting The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Ribosomal Subunit Assembly

259

factors) that may assist bacterial r-subunit assembly in vivo.10,11 In clear contrast, no reconstitution experiment has been described for eukaryotic ribosomes except for Dictyostelium discoideum.12 Still, in vitro reconstitution of r-subunits from D. discoideum does not occur autonomously but instead requires addition of small amounts of nuclear extract.12 In S. cerevisiae, about 100 small nucleolar RNAs (snoRNAs) and more than 150 protein trans-acting factors have been so far described to function in ribosome biogenesis, and most likely not all of them have been identified yet.13-15 These trans-acting factors participate in a series of complicated nuclear, from the nucleolus across the nucleoplasm to the nuclear pore complexes (NPCs), and cytoplasmic events. These events cover all aspects of ribosome biogenesis, and they include the processes that occur after rDNA transcription (see Chapters 8 and 9) like precursor rRNA (pre-rRNA) processing (see Chapter 11 and 12), pre-rRNA modification (see Chapter 13), r-subunit assembly (this Chapter) and r-subunit export (see Chapter 14). This chapter is aimed at reviewing the current understanding of r-subunit assembly by providing insights into the potential functions of the so far characterized trans-acting assembly factors. We will also discuss the recent advances accomplished by systematic and specific proteomic analyses that have allowed the characterization of defined preribosomal particles. We will focus on the S. cerevisiae r-subunit assembly pathway since most of our current knowledge on ribosome biogenesis comes from studies with this yeast, however, as will also be discussed, the basic outline of ribosome synthesis is highly conserved throughout the eukaryotic kingdom.

Methods to Approach Ribosomal Subunit Assembly The assembly of r-subunits has so far been mainly examined in S. cerevisiae by applying different experimental tools. These include diverse biochemical assays, useful molecular and reverse genetic methods that are more or less exclusive for this microorganism, visual screens and powerful proteomic analyses.

Biochemical Studies The analysis of r-subunit assembly in yeast was initiated exclusively by biochemical methods. These methods had been previously applied to the study of the synthesis of ribosomes in HeLa cells.16 They were based on the isolation of crude extracts enriched in high molecular weight RNPs from yeast protoplasts labelled with [3H]-uridine for different times.17 The protoplasts were incubated at low temperature (15º C), conditions that were known to slow down the process of ribosome formation.18 After labelling, the sedimentation pattern of the RNPs was analyzed by centrifugation through linear sucrose gradients. These studies identified three major preribosomal particles that sediment at about 90S, 66S and 43S.18 The predominantly labelled rRNA species found in these RNPs were the 35S, 27S and 20S pre-rRNA, respectively.18 Kinetic data of the labelled material indicated a precursor-product relationship between the 90S particle and the two other particles.18 The 66S and the 43S particles were the precursors of the mature 60S and 40S r-subunits, respectively.19,20 These experiments also demonstrated that the 90S and 66S preribosomal particles were exclusively present in the nucleus, whereas the 43S particle was mainly present in the cytoplasmic fraction.19 Interestingly, measurement of buoyant densities of the different particles in CsCl gradients indicated that they contained more protein than the mature r-subunits, suggesting the involvement on nonribosomal proteins in the r-subunit assembly reactions.18 Finally, the course of the assembly of the different r-proteins into preribosomal particles was studied mainly by monitoring the kinetics of incorporation of individual r-proteins into the different described nuclear and cytoplasmic r-particles. To do so, yeast protoplasts were labelled for different times (15 to 120 min) with [3H]-amino acids and the incorporation of label into r-proteins within preribosomal particles or mature r-subunits was then normalised to the incorporation of [14C]-labelled amino acids of uniformly labelled protoplasts (120 min).21 These experiments clearly indicated that most r-proteins assemble with the nuclear preribosomes at an early stage during ribosome maturation.21 Only, a

260

The Nucleolus

limited number of r-proteins assemble at later steps during ribosome maturation or are even added in the cytoplasm.21 Although, it seems that r-subunit assembly is irreversible, there are reports of few r-proteins that can exchange on mature 60S r-subunits in vivo.22,23 Due to the lack of in vitro reconstitution assays for eukaryotic ribosomes, the above mentioned studies could only be complemented by approaches that examined the abilities of the different r-proteins to either bind to pre-rRNA molecules in vitro24 or dissociate from purified ribosomal particles by increasing concentration of ions.25,26 Although the following years of biochemical studies did not allow establishment of the precise order of events during yeast ribosome maturation, they allowed us to draw a low-resolution picture of the r-subunit assembly pathway (see Fig. 1 and its legends).

The Power of Yeast Genetics Early biochemical work indicated that the process of ribosome synthesis in yeast was similar to that in higher eukaryotes.17 This finding validated the use of yeast as a model system to which, powerful genetic tools could easily be applied. These included the screens of collections of cold-sensitive (cs) and thermo-sensitive (ts) mutants for altered ratios of free 40S to 60S r-subunits or defective pre-rRNA processing (for examples see refs. 27-29), isolation of high-copy suppressors (for an example see ref. 30), functional analyses of genes encoding putative enzymes or proteins containing specific motifs (for examples see refs. 31, 32) or synthetic lethal screens (for examples see refs. 33, 34). The application of different genetic methods has permitted the identification of many of the different trans-acting factors that are so far known to be involved in ribosome biogenesis. The roles of these factors have been inferred on the basis of their mutant phenotype with respect to pre-rRNA metabolism. Some of them are predicted to encode the enzymatic components of the processing machinery, endo- and exonucleases, or the proteins that directly facilitate the processing reactions (see Chapter 11 and 12). In addition, the analyses of trans-acting factor mutants have allowed a better characterization of the pre-rRNA processing pathway, its different highly ordered steps and the exact position of the cleavage sites within the pre-rRNA intermediates.14,15 Other trans-acting factors are part of the about 100 snoRNPs, which mostly guide early methylation and pseudouridylation modifications to specific sites in the rRNA (see Chapter 13), although four of them (U3, U14, snR10 and snR30) are also involved in early pre-rRNA processing reactions.14,35,36 However, most of the identified trans-acting factors have no precise function in pre-rRNA processing or modification, and have been generally called r-subunit assembly factors. Mutation in or depletion of any of these factors has an impact on the pre-rRNA processing pathway since the processes of pre-rRNA processing and r-subunit assembly are intimately linked.14 Similarly, mutation in or depletion of r-proteins lead to a pre-rRNA processing defect.37-39 Two classes of terminal phenotypes have been observed for r-subunit assembly mutants. (i) In many mutants, pre-rRNAs are apparently made but are highly unstable. This phenotype suggests that preribosomal particles are rapidly disassembled and their components rapidly degraded in the absence of the correct function of any of the r-subunit assembly factors belonging to this phenotypic class. (ii) In many other mutants, processing of certain pre-rRNAs appears to be blocked, leading to a slight accumulation of one or more pre-rRNA intermediates or aberrant species. Normally, this accumulation represents only a minor fraction of the flux of RNA through the processing pathway, and then, as above, non or improperly processed pre-rRNA seems to be degraded. This suggests that preribosomal particles are also rapidly degraded in the absence of the correct function of any of the factors belonging to this second phenotypic class. However, the kinetics of blocking pre-rRNA processing may be significantly faster than that of pre-rRNA degradation, thus allowing to some extent for preribosomal particles or aberrant dead-end products to slightly accumulate. Only very few mutant strains accumulate pre-rRNAs to a level that represents a significant fraction of the abundance of the mature rRNA. These mutants are usually defective in bona fide pre-rRNA nucleases (i.e., exosome or Nob1p).40,41

Ribosomal Subunit Assembly

261

Figure 1. The processosome model for r-subunit assembly in S. cerevisiae. In the nucleolus, the 18S, 5.8S and 25S rRNAs are transcribed by RNA Pol I as a large precursor. Mature sequences are separated by two internal transcribed sequences (ITS1 and ITS2) and flanked by two external transcribed sequences (5' ETS and 3’ETS). The 5S rRNA is transcribed by RNA Pol III as a precursor which has an extension of a few nucleotides at its 3' end. The 5S rDNA gene is located between two nontranscribed spacers (NTS1 and NTS2). About two-hundred repeats of this array are found on the long arm of chromosome XII. Clear arrows indicate the origin and direction of transcription. Following transcription, the 35S pre-rRNA associates with many early assembling small and large subunit r-proteins and nonribosomal factors to form a 90S preribosomal particle. According to this model, the 5S rRNA is already present in the 90S particle. Processing at the early sites A0, A1 and A2 releases the 66S and 43S preribosomal particles containing the 27S and the 20S pre-rRNAs, respectively. Late assembly of r-proteins and nonribosomal factors occurs within these particles. The 43S preribosomal particle is rapidly exported through the nuclear pore to the cytoplasm where the final maturation steps in the synthesis of the 18S rRNA take place. The 66S preribosomal particle remains in the nucleus until maturation of both 5.8S and 25S rRNAs is completed. Then, the particle is exported to the cytoplasm where assembly of mature 60S r-subunits is completed by the incorporation of some r-proteins. Nucleolus, grey square. Preribosomal particles and mature r-subunits, light balloons. Nuclear envelope, rods.

262

The Nucleolus

In addition, genetic analyses of the many trans-acting factors involved in ribosome synthesis have helped researchers conclude that a considerable autonomy exists for the synthesis of 40S and 60S r-subunits. Most factors required for synthesis of 40S r-subunits are not required for synthesis of 60S r-subunits. This occurs because the maturation steps to 25S and 5.8S rRNAs can be initiated independently of the early processing steps that liberate the 20S precursor to the mature 18S rRNA.14,42 Moreover, mutants in factors required for the synthesis of 60S r-subunits still produce substantial amounts of 40S r-subunits. However, in almost all these mutants, there is a delay in processing at the early cleavage sites and therefore, in production of mature 18S rRNA. 14,43 These observations were reconciled in the so-called “processosome” model.13,14 In this model, factors for assembly of both 40S and 60S r-subunits bind to the early 35S pre-rRNA to form the identified 90S preribosomal complex. This was thought to serve as a quality control mechanism to ensure the rapid and efficient maturation of r-subunits only when most ribosomal and nonribosomal components were present in the 90S preribosomal particle (see Fig. 1 and its legend). However, although it is clear that pre-rRNA processing and r-assembly are highly coordinated, this hypothesis seems not to be fully correct, as will be discussed below.

Visual Screens Another approach to identify trans-acting factors involved in r-subunit assembly has consisted in the application of visual screens. In principle, these screens were thought to identify factors directly involved in r-subunit export. Screens for 60S r-subunit export mutants relied on the nuclear or cytoplasmic localization of different 60S r-proteins fused to the green fluorescent protein (GFP).44-46 Screens for 40S r-subunit export mutants were based on cytoplasmic processing of the 3' end of the 18S rRNA.47,48 The results of these screens indicate that, in common with many other nuclear transport processes, the export of r-subunits requires the GTPase Ran, the nuclear export receptor Xpo1p/Crm1p and components of the NPC.45,48 In addition, these screens suggest that the processes of r-subunit assembly and r-subunit export are linked since many of the identified mutants are most likely trans-acting factors involved in r-assembly reactions rather than export factors (for examples see refs. 49, 50). Interestingly, there are also examples of mutant strains, which are not clearly affected in r-subunit assembly but are strongly impaired in 60S r-subunit export.51 Uncoupling both processes would ensure a rapid progress in the understanding of r-subunit export (see Chapter 15 for further details).

Proteomic Analyses All of the above mentioned methods have led to a great advance in the understanding of ribosome assembly, which is however not comparable to the breakthrough that the application of the modern biochemical approaches has allowed. These approaches consisted of the use of epitope-tagged trans-acting factors for affinity purification of specific preribosomal particles. From the different methods, the tandem affinity purification (TAP), due to its stringency and ease of handling, has served as the basis for allowing this major advance.52 The RNA composition of the purified preribosomal particles is assayed by northern blot and primer extension analyses and the RNA is quantified by phosphorimaging techniques.53 The protein composition of the preribosomal particles is determined by protein separation (i.e., 2D-PAGE or SDS-PAGE) and subsequent mass-spectrometric analysis.54 In addition to the purification of preribosomal complexes using selected trans-acting factors as bait (for examples see refs. 51, 55), large-scale approaches have led to the isolation of more than two hundred multiprotein complexes, amongst them, preribosomal particles or stable subcomplexes thereof.56,57 Using these novel biochemical analyses, it has been possible to identify many new protein trans-acting factors that had eluded the different sophisticated genetic screens and reverse genetic methods.15,58 In addition, the use of these methods has increased substantially, due to the tremendous information on the composition of distinct early, late and exported preribosomal particles, the resolution of our picture of the r-subunit assembly pathway. The current model for the r-subunit assembly pathway is shown in Figures 2-4 and it will be discussed in a further section of this chapter.

Ribosomal Subunit Assembly

263

Compilation of Trans-Acting Factors Involved in r-Subunit Assembly As mentioned above, the combined power of yeast genetics and proteomics has provided us with many proteins and snoRNAs involved in ribosome synthesis. The identification of new factors has advanced very fast in the past three years; since the last comprehensive reviews in 199913,14 over 50 additional protein trans-acting factors have been described.15 However, most likely not all of them have been identified yet. For a list of trans-acting factors involved in ribosome synthesis and links to other databases see http://www.expasy.org/linder/proteins.html. What are the characteristics that define a trans-acting factor as an assembly factor rather than as a factor involved in pre-rRNA processing, pre-rRNA modification or r-subunit export? As discussed above, this question cannot be easily answered. The assembly of preribosomal particles is linked to pre-rRNA processing and coupled to some extent to export of preribosomal subunits. Therefore, the terminal phenotypes of mutants in trans-acting factors involved in r-subunit assembly are mostly consistent with indirect effects on pre-rRNA processing. Regarding r-subunit export, many defects observed in mutant strains might also be the result of abortive assembly of preribosomal particles. From these observations, it seems reasonable to define what are not the characteristics of r-subunit assembly factors. For instance, trans-acting factors with experimental or predicted nuclease activity as well as their cofactors are not bona fide assembly factors.40,59-61 In addition, the snoRNPs involved in 2'-O-ribose methylation and pseudouridine formation as well as the putative methyltransferases, which are likely involved in base methylation, are neither true assembly factors.13 However, this seems not to be the case for the U3 snoRNP, which has been reported to prevent premature formation of the central pseudoknot in the 18S rRNA that may otherwise block further assembly of the 90S preribosomal complex.14,62 We do not know if other snoRNPs, most of which may associate cotranscriptionally with the 35S pre-rRNA, also have a function in preventing unwanted folding arrangements (see Chapter 13 for further details). Moreover, several putative methyltransferases might also, as discussed below, have a quality control function during r-subunit assembly. Why are r-subunit assembly factors required? A quick view to the structure of the ribosome at the atomic resolution reveals a very compact machine. Most of the mass of the ribosome is rRNA, which forms complex, almost inaccessible tertiary structures at the interior of the r-subunits.3 An important role of r-proteins is to stabilize the tertiary structure of the rRNA.63 Many r-proteins fold into globular bodies with linear extensions that generally lack any secondary structure.3,4 The globular domains are found largely on the exterior of the r-subunits, while the extensions penetrate into the interior, resulting in an intertwining of RNA and protein in the centre of the subunits.4 It is hard to imagine that such a compact object could assemble from its components efficiently in anything other than a highly ordered manner. In this ordered process, certain pre-rRNA folding must be prevented until late in the r-assembly pathway since it would otherwise impair access for the pre-rRNA processing, modification and/or assembly factors. Therefore, key regions of the pre-rRNA might be maintained in a relatively loose structure until processing of the mature rRNA is almost completed. What might be the function of the r-subunit assembly factors? In the scenario presented above, two categories of factors are envisaged to assist ribosome assembly. First, energy-consuming enzymes and their cofactors would be required to rearrange intramolecular pre-rRNA, pre-rRNA-protein, or protein-protein structures. Second, nonenergy consuming RNA- or protein-binding factors would be required to either stabilize all these different structures or prevent the premature and inappropriate formation of an unproductive rRNA, rRNA-protein or protein-protein interaction. In addition, a subset of factors may be involved in aspects that are more related to the recycling of trans-acting factors, or the trafficking of preribosomal particles from the nucleolus to the nucleoplasm and/or the cytoplasm. In the following sections, we attempt to rationalize the function of the many different r-assembly factors. However, it needs to be mentioned that, unfortunately, in most cases the postulated functions are rather speculative and not inferred from solid experimental evidence.

264

The Nucleolus

RNA Helicases The putative RNA helicases of the DEAD-box and related protein families (about 50 members in yeast) represent the largest class of protein trans-acting factors involved in ribosome biogenesis. These ubiquitous protein families are defined by several evolutionarily conserved motifs, and their members are involved in most, if not all, aspects of RNA metabolism.64,65 Enzymatically, many of these proteins have an RNA-dependent ATPase activity in vitro, while a small number have been shown to possess the ability to unwind RNA duplexes in an ATP-dependent manner.66 Their unwinding activity together with their strong structural homology to DNA helicases has prompted researchers to generally refer to these enzymes as RNA helicases.67 However, a processive RNA unwinding activity has solely been documented for the vaccinia virus NPH-II helicase, which is essential for the replication and transcription of the double-stranded RNA (dsRNA) genome of the virus.68 Considering the RNA structures, mostly short inter- and intramolecular duplexes, which might be encountered by putative RNA helicases during ribosome biogenesis, it is clear that a “true helicase” activity is not even needed. Indeed, except for the RNA genomes of certain viruses, a continuous dsRNA is very rare in biological systems.69 Moreover, as recently reported, it is very likely that the putative RNA helicases involved in ribosome biogenesis will not only be required for the ATP-dependent dissociation of RNA-RNA structures, but also for the active modulation of specific RNA-protein or even protein-protein interactions.69,70 In any case, putative RNA helicases must be considered as energy-dependent modulators of RNA within a natural RNP context. To date, 18 putative RNA helicases have been directly implicated in ribosome synthesis in S. cerevisiae.13,65,71-73 So far, neither precise substrates nor clear interacting partners have been elucidated for these proteins. Moreover, an in vitro ATP-dependent RNA-unwinding activity has not been reported for any of the helicases involved in ribosome biogenesis. However, although their exact roles are not known, the terminal phenotypes observed upon depletion or inactivation of the different putative RNA helicases indicate their possible function during ribosome biogenesis. (i) Putative RNA helicases may assist discrete steps in the pre-rRNA processing pathway by rendering the pre-rRNA substrates accessible to specific endo- or exonucleases. Thus, putative RNA helicases may act as cofactors for a given nuclease. It has been suggested that this may be the function of the putative RNA helicases Dob1p and Dbp3p.59,60 Since these two putative RNA helicases might rather act as pre-rRNA processing than as assembly factors, they will not be further discussed in this chapter (see Chapter 12 for details). (ii) With over one hundred snoRNAs in yeast and even more in mammalian cells, there are many potential helicase substrates. The theoretical RNA-unwinding activity of these enzymes could establish, and/or dissociate snoRNA-pre-rRNA base pairings, which are, in most cases, mutually exclusive with the final folding of the rRNA in the mature ribosome.62,74 No particular helicase has been described to contribute to the pre-rRNA modification reactions catalyzed by snoRNPs, but interestingly, few putative RNA helicases have been found to be physically or genetically associated with snoRNAs. Dhr1p is the sole putative RNA helicases for which a physical association with a snoRNP has been reported.72 Protein A-tagged Dhr1p specifically and stably coprecipitates U3 snoRNA, the efficiency of precipitation being similar to that obtained when Nop56p, which is a core C/D-box protein,75 is used as bait. It has been speculated that the U3 base-pairing with the 35S pre-rRNA targets Dhr1p to the 90S preribosomal particle, where it functions in the formation of the central pseudoknot in 18S rRNA.72 Unfortunately, no evidence has been reported to support this function. On the other hand, genetic observations suggest a functional interaction between Rok1p and snR1076 and between Dbp4p and U14 snoRNA.77 As above, further experiments are required to define the functional link between these two putative RNA helicases and snoRNAs. (iii) Finally, it has been speculated that many of the putative RNA helicases are involved in rearrangements of RNA-RNA, RNA-protein and protein-protein interactions that would then facilitate the recruitment, remodelling or dissociation of trans-acting factors and r-proteins within preribosomal particles.13,66 A set of different rearrangements in a well-defined order may also ensure efficiency and accu-

Ribosomal Subunit Assembly

265

racy during ribosome synthesis. In addition to Dbp4p, Dhr1p and Rok1p, whose possible functions have been discussed above, there are still another five putative RNA helicases required for production of 40S r-subunits, namely Dbp8p, Dhr2p, Fal1p, Has1p and Rrp3p.72,73,78-80 In the absence of the correct function of any of these five proteins, the early cleavages at sites A1 and A2 are inhibited and the cleavage at site A0 is more or less delayed. In addition, none of these putative RNA helicases are needed for the accumulation of any of the four snoRNAs required for pre-rRNA processing. In conclusion, it is likely that these putative RNA helicases are involved in specific rearrangement reactions within 90S preribosomal particles, however, the uniformity of terminal phenotypes does not allow us to make any statements about their precise functions. In clear contrast, in addition to Dob1p and Dbp3p, whose possible functions have been discussed above, the putative RNA helicases required for production of 60S r-subunits can be grouped into two categories with respect to their putative functions as factors promoting structural rearrangement reactions within preribosomal particles. The first category includes Dbp6p, Dbp7p, Dbp9p and maybe Mak5p.31,81-83 Inactivation of any of these four proteins leads to loss of 27S pre-rRNA species. It has been speculated that in the absence of their correct function, certain structural rearrangements do not take place or they are delayed, leading to disassembly and degradation of early preribosomal particles to mature 60S r-subunits.13 The second category comprises Dbp10p and Spb4p and maybe Dbp2p and Drs1p.13,71,84,85 Inactivation of Dbp10p and Spb4p leads to a clear accumulation of 27SB precursors. This phenotype has also been interpreted as a direct consequence of the lack or the delay of specific structural rearrangements within late preribosomal particles to mature 60S r-subunits. Unfortunately, up to now no biochemical assays are available to analyze the precise environment of putative RNA helicases or to define the structural rearrangements mediated by putative RNA helicases. Therefore, it has so far not been possible to demonstrate that the putative RNA helicases involved in r-subunit assembly indeed act as energy-dependent modulators of structures within preribosomal particles. Nevertheless, in favour of such a model speaks the fact that mutation of the ATPase motifs abolishes the in vivo function of several of the above putative RNA helicases.78,86,87 Yet another unsolved question is why there are so many putative RNA helicases required for r-subunit assembly. There is almost a complete lack of functional redundancy between the different putative RNA helicases since genetic experiments clearly indicate that in general putative RNA helicases cannot substitute each other even when overexpressed. So far, it has only been reported that Dbp9p is a high dosage suppressor of the slow-growth phenotype and the 60S r-subunit deficiency of certain dbp6 alleles.31 However, overproduction of Dbp9p can not bypass the lethality of the dbp6 null strain.31 The realistic challenge for putative RNA helicases is the identification, by genetic and/or biochemical means, of their precise substrates.

AAA ATPases AAA ATPases fall into a large family (about 50 members in yeast) of ubiquitous ATPases, which are easily recognized by their strong sequence conservation in the so-called AAA domain.88 These proteins, some of them self-associate to form hexameric structures, are involved in many cellular processes as their name indicates (AAA ATPases, ATPases associated with various cellular activities). These include proteolysis, vesicle and organelle transport, disassembly of stable protein-protein complexes, microtubule organization, DNA replication and intracellular motility.89,90 The AAA ATPases associated to the proteolytic machines seem to work by unfolding target proteins so that they are accessible to the protease activity.91 Most interestingly, many AAA ATPases use the energy derived from ATP hydrolysis to dissociate stable protein-protein interactions. This has been well reported for the role of the AAA ATPase katanin in the disassembly of microtubules.92 Finally, there are also examples of AAA ATPases that exploit the energy of ATP hydrolysis to generate unidirectional motion along a track, as exemplified by dynein, which is a microtubule-based motor.88

266

The Nucleolus

How is the energy release of ATP hydrolysis coupled to all these different activities? The model to explain unidirectional movement involves a cycle of association and dissociation of an AAA ATPase complexed with its cargo to multiple substrates along a track. Disassembly of protein-protein interactions is explained by the unfolding of a structural element of the target complex at or near the binding interface between its components. In this case, the hydrolysis of ATP may cause a conformational change of the shape of the AAA ATPase structure that can be transmitted to the bound substrate, allowing by this means the unfolding of the substrate. A similar scenario might be suspected for AAA ATPases associated with proteolytic machines. It is reasonable to assume that enzymes with the ability to disassemble protein complexes will be required for r-subunit assembly or export. In addition, AAA ATPases may also mediate the movement of preribosomal particles from the nucleolus through the nucleoplasm to the NPCs.93 To date, two putative AAA ATPases have been described to participate in ribosome biogenesis: Rix7p50 and Rea1p.94 Both proteins are essential for viability and they associate with preribosomal precursor particles to 60S r-subunits.50,94 However, only Rix7p has been phenotypically analyzed so far. The rix7-1 ts-mutant was isolated in a screen for trans-acting factors involved in 60S r-subunit export.50 This mutant was also, due to the rapid degradation of 27SB pre-rRNAs, defective in the production of 60S r-subunits. Considering its putative enzymatic activity, Rix7p may be involved in a structural rearrangement required for the correct assembly of 60S r-subunits.50 This function resembles that of putative RNA helicases from the Dbp6p group (see above), although in the absence of Dbp6p, the 27SA pre-rRNAs are not formed or are unstable.81 Due to the fact that putative AAA ATPases can function as disassemblers of protein complexes, it cannot be discarded that Rix7p may also facilitate the dissociation of a subset of trans-acting factors from late preribosomal precursors to mature 60S r-subunits that otherwise might impede the use of processing site C2. Alternatively, Rix7p may assist transport of pre60S ribosomal particles to or through the NPC. All these different possibilities are not mutually exclusive, and as for RNA helicases, the function of putative AAA ATPases will remain obscure until their substrates have been identified.

GTPases GTPases are members of a broad family of ubiquitous proteins that regulate a collection of different cellular processes, which are as diverse as protein traffic and cytoskeleton organization (i.e., Ran and Rho,95,96) translation (i.e., IF2, EF-Tu and EF-G97,98) or signal transduction (i.e., heterotrimeric G proteins.99) GTPases are identified by the presence of the so-called G domain, which consists of five conserved polypeptide loops that form contact sites with GTP and Mg2+.100,101 GTPases cycle between two conformational states, a GTP-bound and a GDP-bound form.102 Normally, release of bound GDP and subsequent binding of GTP is catalyzed by guanine nucleotide exchange factors.96 A common model to explain the function of GTPases has been suggested. In this model, the energy of GTP hydrolysis is used to induce a conformational switch in GTPases that changes their binding affinities for their effector molecules, as exemplified for G-protein coupled receptor systems.102 Alternatively, the conformational switch can change interactions between the different domains of the GTPase itself, as exemplified by EF-G.97 In any case, GTPases can be considered as modulators of protein-protein and even RNA-protein interactions.100 It is reasonable to assume that such modulators will be required for r-subunit assembly or export. In this sense, GTPases can serve as energy sources for assembling r-proteins onto the pre-rRNA or properly positioning other trans-acting factors in the preribosomal particles. To date, some putative GTPases have been described to function during the biogenesis of both mature 40S and 60S r-subunits. Bms1p is the sole member of this family required for the production of mature 40S r-subunits.103,104 Bms1p is an essential nucleolar protein that binds GTP but has a very modest GTPase activity in vitro (mentioned in ref. 103). Interestingly, another yeast protein, Tsr1p, shows significant homology to Bms1p but lacks the G domain.103 There is additional evidence to indicate that Tsr1p and Bms1p are not functional equivalents

Ribosomal Subunit Assembly

267

(further details in ref. 103). So far, six members of the broad family of GTPases have been described to function in 60S r-subunit biogenesis. The putative GTPases Nog1p, Nug1p and Nog2p/Nug2p are all essential nucleolar proteins associated with late preribosomal precursors to 60S r-subunits.51,105,106 Unfortunately, no phenotypic analysis has been reported for conditional nog1 mutants. Different data imply Nug1p and Nog2p/Nug2p in export reactions of pre60S ribosomal complexes. Indeed, conditional nug1 and nug2 mutants have reduced amounts of mature 60S r-subunits due to a nucleoplasmic accumulation of pre60S ribosomal precursors.51,106 However, these mutants are only slightly impaired in pre-rRNA processing. Since Nug1p and Nug2p are good candidates for trans-acting factors involved in export of pre60S particles, their function will not be further discussed in this chapter (see Chapter 15). An interesting link between the yeast Ran GTPase homologue Gsp1p107 and the exosome (see Chapter 13 for further details about the exosome), which is a complex of 3'→5' exonucleases required for the formation of the mature 3' end of 5.8S rRNA,108 has been reported. Gsp1p and the exosomal subunit Dis3p/Rrp44p interact both physically and genetically,107 and strikingly, gsp1 mutants, but so far no other mutants affecting nucleocytoplasmic transport, display pre-rRNA processing defects that are similar to those observed upon inactivation of the exosome or its nuclear cofactor Dob1p.107 Therefore, it is unlikely that the pre-rRNA processing phenotypes of gsp1 mutants could be the indirect result of defects in nucleocytoplasmic transport of pre60S particles or recycling of trans-acting factors. Unfortunately, how Gsp1p may regulate the activity of the exosome is largely unknown. Finally, two cytoplasmic putative GTPases, Efl1p/Ria1p and Kre35p, participate in 60S r-subunit formation.94,109,110 The role of Kre35p, which is physically associated with a cytoplasmic precursor of mature 60S r-subunits, has so far not been further characterized.94 In contrast, the function of Efl1p is well documented. Efl1p is a bona fide GTPase whose activity is stimulated by 60S r-subunits.110 Its closest yeast homologue, Eft1/2p, is the EF-G-like ribosome translocator factor that operates during elongation of translation.98 Deletion of EFL1 is not lethal but confers a severe growth phenotype and causes an underaccumulation of 60S r-subunits.109,110 Interestingly, these defects are suppressed by point mutations in TIF6 or mild overexpression of wild-type Tif6p. Tif6p is a nucleolar trans-acting factor involved in 60S r-subunit synthesis109,110 and it has been described to impede the unspecific binding of both mature r-subunits in vitro.111 It has nicely been demonstrated that Efl1p promotes the dissociation of Tif6p from 60S r-subunits in vitro. In vivo, nucleolar Tif6p is redistributed to the cytoplasm in mutant strains lacking Efl1p.110 Altogether, these data are consistent with a model where Tif6p accompanies pre60S ribosomal particles from the nucleolus to the cytoplasm. There, the GTPase activity of Efl1p triggers a structural rearrangement within pre60S ribosomal particles, which directly or indirectly facilitates the release of Tif6p and thus, promotes the first step during its recycling to the nucleolus. The overall homology of Efl1p to Eft1/2p suggests that their binding sites on the ribosome are conserved. If this were the case, the association of Efl1p, Tif6p or both proteins with cytoplasmic pre60S r-subunits could also provide a quality control to ensure that these pre60S r-subunits are not engaged in translation until their maturation has been completed.

Methyltransferases Some trans-acting factors involved in ribosome biogenesis may act as methyltransferases as suggested by the occurrence of distinct motifs (motifs I, post-I, II and III) within their primary sequence that are the hallmarks of S-adenosylmethionine-dependent methyltransferases.112 This family includes Dim1p113,114 and Rrp8p,33 both involved in 40S r-subunit synthesis, and Ncl1p,115 Nop2p116 and Spb1p,117,118 which are required for 60S r-subunit synthesis. Since Nop1p is likely the bona fide site-specific 2'-O-methyltransferase of the ribose moieties,119 it is probable that all these putative methyltransferases are site-specific base methylases. However, no specific methylation defects have been observed or tested in their mutant strains, except for Dim1p, which dimethylates the 3' end of 18S rRNA within the 20S pre-rRNA at positions

268

The Nucleolus

m26 A1779 m26 A1780.113 In addition, it seems that 2'-O-ribose methylation in 25S rRNA at position UmGm2922 remains low upon Nop2p depletion,116 however, considering the identification of the snoRNA snR52, which is complementary to UmGm2922, it is unlikely that Nop2p methylates this site. The role of putative methyltransferases during r-subunit assembly remains unclear, although inactivation of these proteins, excluding Ncl1p, leads to distinct pre-rRNA processing defects that are similar to those observed in cells depleted of other trans-acting factors with a more obvious role in assembly reactions. It has been assumed that these pre-rRNA processing defects appear as secondary effects, perhaps as a consequence of the improper association of other trans-acting factors within preribosomal particles. For example, conditional dim1 mutations lead to inhibition of early pre-rRNA processing steps in addition to late dimethylation defects.114 Surprisingly, the pre-rRNA processing defects are suppressed when the rRNA unit is transcribed by RNA Pol II from the PGK promoter.114,120 This suggests that Dim1p is not directly involved in pre-rRNA processing. Instead, Dim1p might associate cotranscriptionally with the nascent RNA Pol I transcribed pre-rRNA and simultaneously with some trans-acting factors involved in the early pre-rRNA processing reactions. In the absence of Dim1p, these other trans-acting factors are not efficiently recruited, thus leading to early pre-rRNA processing defects. Interestingly, RNA Pol I has been shown to be associated with some trans-acting factors like Nop1p, Cbf5p, Nhp2p or Rrp5p;121 moreover, it might be a common strategy to interconnect subsequent processes as also exemplified by the coupling of RNA Pol II transcription to premRNA processing events and mRNA export.122-125 However, no evidence for a similar model exists so far to explain the pre-rRNA processing defects observed in mutants with inactivated Rrp8p, Nop2p or Spb1p.

Trans-Acting Factors Related to r-Proteins Some trans-acting factors contain domains that display significant similarity to RNA-binding motifs of different r-proteins. For instance, Rrp5p, which is the only protein well-reported to be required for both 18S and 5.8S/25S rRNA synthesis, contains 12 N-terminal repeats of the S1 RNA-binding domain.126 In addition, there are nucleolar trans-acting factors with extensive segments of identity or similarity with r-proteins throughout their entire primary sequences. Amongst them are Rlp7p, which is homologous to Rpl7Ap an Rpl7Bp,49,127 Rlp24p, which is a homologue of Rpl24Ap and Rpl24Bp,106 Mrt4p, which is homologous to Rpp0p55 and Imp3p, which is related to Rps9Ap and Rps9Bp.128 Two other proteins, Nhp2p and Snu13p, which are core components of H/ACA-box and C/D-box snoRNAs, respectively, show certain homology to Rpl8p.129,130 Considering the high degree of homology found in all these examples, it has been assumed that these trans-acting factors and their r-protein counterparts successively bind the same rRNA structure but, while the trans-acting factors bind to the rRNA site within a preribosomal particle, the r-proteins bind the same site within the mature r-subunit. In agreement with this model, the functional environment of Imp3p and Rlp7p corresponds to that of Rps9p and Rpl7p, respectively. Rps9p associates early with a 90S preribosomal complex131 and Imp3p is a component of the U3 snoRNP that functions in early pre-rRNA processing reactions.128 Ribosomal protein Rpl7p is a component of an early 66S preribosomal particle,94 and intriguingly, Rlp7p is required for an early step during 60S r-subunit biogenesis.49,127 The functional analysis of Rlp24p and Mrt4p has not yet been performed but these two proteins are likely associated with early pre60S particles.55,94 However, both Rpl24p and Rpp0p likely assemble onto pre60S particles very late in the nucleus or in the cytoplasm.21,106 It will be interesting to test whether Rlp24p and Mrt4p dissociate from pre60S ribosomal particles just prior to export or in the cytoplasm.

Assemblers of r-Proteins The function of some trans-acting factors is genetically or physically linked to some r-proteins. These factors are Krr1p,30 Tom1p132 and Rrp7p,133 which are all involved in the production of 40S r-subunits and Rrb1p,134,135 Sqt1p136 and Rsa1p,34 which are required for 60S r-subunit

Ribosomal Subunit Assembly

269

biogenesis. All of them, except Rsa1p, are essential factors. The rsa1 null mutant is viable but displays a temperature dependent slow-growth phenotype.34 Conditional krr1, tom1 and rrp7 mutants are blocked in pre-rRNA processing at sites A0, A1 and A2 and therefore, they are impaired in 18S rRNA synthesis.30,132,133 Both growth and pre-rRNA defects of conditional krr1 mutants are partially suppressed by high dosage of Rps14Ap. However, overexpression of Rps14p does not suppress the lethality of the krr1 null mutant.30 Similarly, both the growth and pre-rRNA processing defects of two tom1 ts-mutants are partially suppressed by overexpression of Rps0Ap and Rps0Bp.132. On the other hand, the lethality and pre-rRNA defects of the rrp7 null mutant are partially suppressed by overexpression of the nearly identical Rps27Ap and Rps27Bp. The simplest explanation for this suppression by overexpression is that Krr1p, Tom1p and Rrp7p are required for efficient association of Rps14p, Rps0p or Rps27p with preribosomal particles, respectively. Disfunction of any of these trans-acting factors might entail a block in processing at the early cleavage sites as an indirect consequence of an improper assembly of Rps14p, Rps0p or Rps27p within 90S preribosomal particles. Thus, elevated levels of these r-proteins may improve the efficiency of their assembly when their putative assemblers are not fully functional or absent. Unfortunately, although pertinent, the physical association between Tom1p and Rrp7p and their respective high copy r-protein suppressors have not been investigated. However, recent data show that both Krr1p and its high dosage suppressor Kri1p bind to Rps14p.137 Genetic and physical interactions have also been described between the r-protein Rpl10p and the cytoplasmic protein Sqt1p.136 Overexpression of Sqt1p suppresses the slow-growth phenotype of dominant negative rpl10 truncation mutants.136 Depletion of Sqt1p results in decreased levels of Rpl10p on free 60S r-subunits. Finally, Sqt1p interacts strongly with Rpl10p in a yeast two-hybrid assay.136 A likely explanation for all these data is that Sqt1p is involved in the assembly of Rpl10p into almost mature cytoplasmic 60S r-subunits. However, it has recently been reported that Rpl10p assembles onto nascent 60S r-subunits in the nucleoplasm,45 furthermore, this association is a prerequisite for nuclear export of pre60S r-subunits.45,138 The trans-acting factor Rsa1p, which has also been described to assist the assembly of Rpl10p, is a nucleoplasmic protein.34 It will be interesting to test if there is a physical interaction between Rsa1p and Sqt1p, and thus, if Sqt1p is a nucleocytoplasmic shuttling protein. In agreement with this hypothesis, Sqt1p has been found associated with early 66S preribosomal particles.94 The trans-acting factor whose function has most recently been linked to an r-protein is the essential Rrb1p protein.134,135 Rrb1p physically interacts with the free form of the r-protein Rpl3p.134,135 This interaction is specific, since no other r-proteins interact with Rrb1p, and the complex between Rpl3p and Rrb1p is stoichiometric.135 Rrb1p shuttles between the cytoplasm and the nucleus, although at steady-state it is predominantly nucleolar. Overproduction of Rrb1p leads to a nuclear accumulation of both Rpl3p and Rrb1p. However, the two proteins become predominantly cytoplasmic in the absence of active translation.134 Finally, genetic analyses indicate that depletion of Rrb1p results in decreased levels of 60S r-subunits as a consequence of delayed 27SB pre-rRNA processing.134 Decreased levels of 60S r-subunits have been observed in two rpl3 mutants,86 although their pre-rRNA defects have not yet been studied. Altogether, these data are consistent with a model where Rrb1p is the assembler of Rpl3p onto a precursor to mature 60S r-subunits. Since Rrb1p associates very weakly with preribosomal particles,135 the assembly of Rpl3p into a defined preribosomal particle, most likely the early E1 particle (see below),94 must be accompanied by the dissociation of Rrb1p. Interestingly, Rrb1p contains a WD-repeat, a conserved protein motif that mediates protein-protein interaction139 and that it is also present in Rrp7p and Sqt1p. It will be interesting to learn whether the WD-repeat motif is a common feature of other, as yet unknown, putative assemblers of r-proteins.

Miscellaneous A quick look at the complete list of the currently known trans-acting factors involved in ribosome synthesis clearly shows that most of them have not been covered in the previous

270

The Nucleolus

sections. Some of these trans-acting factors are predicted to have enzymatic activities. For example, Rcl1p is homologous to 3' phosphate-cyclases, 140 Tom1p is homologous to ubiquitin-protein ligases132 and Rrp10p/Rio1p, Rio2p and Hrr25p seem to be protein kinases;141,142 however, their proposed enzymatic activities do not allow us to conclude what their precise function during ribosome biogenesis is. Thus, the biochemical function of most trans-acting factors involved in ribosome synthesis remains unclear. Some of them contain RNA-binding motifs (i.e., S1-, Brix-, GAR-, RRM- motifs, etc.) and other factors bear motifs that may participate in protein-protein interactions (i.e., TPR, coil-coiled, WD-repeats, etc.); however, their functions cannot be readily inferred from these recognizable domains. One example is the family of four nucleolar proteins that contains the so-called Noc-domain. Noc1p/Mak21p, Noc2p and Noc3p are each required for 60S r-subunit biogenesis143,144 and Noc4p has been shown to associate with 90S preribosomal particles.131 Analysis of pre-rRNA in noc4 ts-mutants indicates that Noc4p is required for the early cleavages at sites A0 to A2 and, thus, for 40S r-subunit biogenesis.145 Noc4p forms a heterodimer with Nop14p, which has been described to be also involved in 40S r-subunit biogenesis.146 Moreover, the Rps2p-GFP reporter for the 40S r-subunits is strongly accumulated in the nucleus of noc4 ts-mutants at the nonpermissive temperature.145 Therefore, it appears that Noc4p plays a role in an early step during biogenesis of 40S r-subunits that is necessary to mediate their subsequent export to the cytoplasm. Analysis of pre-rRNAs in conditional noc1, noc2 and noc3 mutants indicates that the Noc1p, Noc2p and Noc3p are not primarily required for pre-rRNA processing.143 These three proteins interact physically in two stable RNA-independent heterodimeric complexes, one containing Noc1p and Noc2p and the other Noc2p and Noc3p. Genetic experiments confirm that these complexes are functionally relevant. Moreover, the two complexes associate with different preribosomal particles in different subnuclear regions: while the Noc1p/Noc2p complex associates with early 66S preribosomal particles in the nucleolus, the Noc2p/Noc3p complex associates with nucleoplasmic pre60S particles. Interestingly, 66S preribosomal particles accumulate in the nucleolus in the noc1 mutant or in the nucleoplasm in the noc3 mutant. In addition, overexpression of the Noc-domain leads to a dominant negative effect on cell growth and results in nucleolar accumulation of preribosomal particles without any detectable pre-rRNA processing defect. Altogether, these results have permitted the establishment of a model where the Noc-proteins facilitate the vectorial movement of both pre40S and pre60S r-subunits from the nucleolus to the nucleoplasm. The dissociation of Noc1p from Noc2p and its replacement by Noc3p may trigger this movement. Whether the mechanism of this movement is passive (retention system) or active (i.e., by the help of an energy-consuming factor) is still unknown. Another trans-acting factor that allows movement of preribosomal particles is the export adapter Nmd3p45,138 (see Chapter 15). The discovery of the Noc and export adapter proteins suggests that the spatial distribution of preribosomal particles within the nucleus is regulated by trans-acting factors. In addition, there is convincing data to suggest that the timing of pre-rRNA processing can also be regulated by trans-acting factors. Ssf1p and Ssf2p are two nearly identical and functionally redundant nucleolar trans-acting factors required for 60S r-subunit synthesis.147 Both proteins associate very early with preribosomal particles to mature 60S r-subunits and they dissociate prior to or at C2 cleavage. Interestingly, in the absence of these two proteins (deletion of SSF2, depletion of Ssf1p), the 27SA2 pre-rRNA is prematurely cleaved at site C2, thus, inhibiting the synthesis of 27SB and 7S pre-rRNAs but allowing the formation of the 25.5S pre-rRNA. However, it seems that the 25.5S pre-rRNA is rapidly degraded in the absence of the correct 7S pre-rRNA, and thus, synthesis of both mature 5.8S and 25S rRNA is affected upon depletion of Ssf1p.147 Ssf1p and Ssf2p belong to a family of six evolutionarily conserved proteins that all harbour the so-called Brix-domain, which contains the σ70-like RNA binding motif.32,148 Interestingly, all these proteins are involved in either 60S r-subunit (Ssf1p, Ssf2p, Rpf1p, Rpf2p and Brix1p) or 40S r-subunit (Imp4p) biogenesis.32,128,147 However, there is no evidence suggesting that any Brix-proteins other than Ssf1p/Ssf2p coordinate timing of pre-rRNA processing reactions.

Ribosomal Subunit Assembly

271

Another interesting factor is Tif6p, which is an essential nuclear protein that is homologous to the mammalian eukaryotic initiation factor eIF6. However, Tif6p seems not to be required for translation;111 instead, it is involved in 60S r-subunit biogenesis.111,149 It has been reported that Tif6p binds in vitro to mature 60S r-subunits and prevents their association with mature 40S r-subunits.111 In vivo, Tif6p is associated with nuclear and cytoplasmic pre60S ribosomal particles94 and Efl1p facilitates its dissociation from cytoplasmic pre60S ribosomal particles.110 Incorporating all experimental evidence, it can be assumed that association of Tif6p, Efl1p or both proteins prevent almost complete cytoplasmic pre60S ribosomal particles from engaging into polysomes and thus from interfering with protein synthesis.

From the Processosome to the Tollervey Model As mentioned above, the combination of selective purification methods with standard techniques for RNA identification and mass spectrometry for protein identification has allowed the characterization of many different preribosomal particles. The purified particles are extremely complex and they contain many nonribosomal proteins, with about a quarter of them previously not having been characterized as trans-acting factors. Due to this complexity, the use of standardized conditions of cell growth, cell lysis and protein purification has been necessary to exclude composition discrepancies of similar particles simply because of differences in handling. In this sense, the effort of the Hurt laboratory in the purification and characterization of many different particles deserves to be specially mentioned (ref. 94 and references therein). All the purified particles are likely to reflect different snapshots of the continuous process of ribosome synthesis. Thus, the proteomic approach has allowed us to propose a new outline of the ribosome synthesis pathway at a much higher resolution compared to the previously available picture based only on sucrose gradient analyses. In addition, as will be described below, the current outline of ribosome synthesis is clearly different from the previously established one. Comparison of the compositions of the distinct preribosomal particles also allows for the conclusion that many trans-acting factors undergo cycles of association, dissociation and recycling from the preribosomal particles. Two further conclusions can also be made. First, the so far identified preribosomal particles lack many trans-acting factors that were predicted to be associated with them, especially enzymes such as many endo- and exonucleases involved in pre-rRNA processing reactions and several putative RNA helicases. This likely reflects the transient nature of the association of these enzymes with preribosomal particles, their association in substoichiometric amounts or their loss during purification. Second and paradoxically, the r-proteins present in these particles have not yet been carefully analyzed. This is in part due to the physicochemical properties of r-proteins, which are small; a limited number of peptides can be analyzed by mass spectrometry. The r-proteins are highly basic, which impedes good resolution during separation by 2D- and SDS-PAGE. In addition, r-proteins are common contaminants in purified complexes; therefore, severe criteria are required to assign a specific r-protein to a given complex. In this section, we describe the current knowledge about the pathway of r-subunit assembly.

The 90S Preribosomal Particles

Two recent reports have approached the direct study of the 90S preribosomal particle.131,150 In one report, Pwp2p and its associated proteins were used as baits to purify complexes that were enriched in small subunit r-proteins over large subunit r-proteins.131 The Pwp2p-containing particle sediments faster than 80S ribosomes, contains the 35S pre-rRNA, more than 30 nonribosomal proteins (around 10 r-proteins from the small subunit and about 5 r-proteins from the large subunit).131 In another report, a large U3 RNP has been purified by applying a purification strategy that was based on the simultaneous epitope-tagging of Mpp10p, a U3-specific protein, and Nop58p, which is a protein common to all C/D-box snoRNAs.150 This complex contains the U3 snoRNA and 28 nonribosomal proteins, but lacks 35S pre-rRNA and r-proteins. It has been suggested that the U3 RNP corresponds to a particular precursor in 90S preribosomal assembly (see Fig. 2 and its legend).

272

The Nucleolus

Figure 2. Current model for the formation and maturation of the 90S preribosomal particle. During transcription of the 35S pre-rRNA, a large U3 RNP assembles at the 5' end of the nascent pre-rRNA. Concomitantly, many r-proteins (small squares), many additional trans-acting factors involved in assembly of 40S r-subunits and few trans-acting factors required for the assembly of 60S r-subunits (small circles) associate to form a 90S preribosomal particle. It is likely that most covalent nucleotide modifications carried out by the C/D-box and H/ACA-box snoRNAs occur on the nascent pre-rRNA. Note that the 90S preribosomal particle lacks most trans-acting factors required for 60S r-subunit synthesis. The early cleavages at sites A0, A1 and A2 lead, by dissociation of most 40S r-assembly factors, to the formation of a 43S preribosomal particle, which is further matured as indicated in (Fig. 3). Subsequent to cleavage at site A2, an early 66S preribosomal particle is formed by association of a large number of 60S r-assembly factors; the further maturation of this particle is depicted in (Fig. 4). This figure indicates the protein composition of the large U3 RNP and it highlights those trans-acting factors and r-proteins that are stably associated with several of the purified complexes related to 90S preribosomal particles. Trans-acting factors with putative enzymatic activity are underlined and those factors and r-proteins found in 66S particles or with a function in synthesis of 60S r-subunits are shown in light grey and italic. Note that a functional implication in ribosome biogenesis has not been reported for all the indicated factors. See text for details. For simplicity, we do not use the recommended nomenclature for yeast proteins (e.g., Bms1 versus Bms1p). Preribosomal particles, light balloons.

Ribosomal Subunit Assembly

273

The RNA components of the Pwp2p-containing particle have been analyzed and, as expected, it contains the 35S pre-rRNA. Unfortunately, although pertinent, neither the levels of the 33S nor the 32S pre-rRNA were examined.131 Protein complexes that are more enriched in either one of these two pre-rRNAs than in 35S pre-rRNA are expected to represent intermediates in the maturation of the 90S particle to the earliest 43S preribosomal complex. Small RNAs were also analyzed: a significant amount of U3 snoRNA is present in the Pwp2p-particle, however, the U14 snoRNA is only substantially enriched in the complex that has been purified using an epitope-tagged version of Nop58p.131 The presence of either snR30 or snR10, which have both a role in the processing of 35S pre-rRNA at sites A0 to A2, was not addressed. Interestingly, the 90S preribosomal particle lacks the 5S rRNA.131 This suggests that, contrary to earlier reports,20 its assembly does not occur at the level of 90S preribosomes.20 In addition, no precipitation of precursors to the 25S rRNA was obtained. Moreover, no or very little precipitation of the 20S precursor to 18S rRNA is detected except for the preribosomal particle purified using a TAP-tagged Enp1p.131 This reflects that, except for the Enp1p-containing complex, the analyzed proteins do not precipitate components from other preribosomal particles than 90S preribosomes. As expected, many known trans-acting factors required for the early cleavages on the 35S pre-rRNA were found in the Pwp2p-containing particle.131,150 These comprise the four proteins common to C/D-box snoRNPs (Nop1p, Nop56p, Nop58p and Snu13p) and seven out of the nine proteins that are specific for the U3 snoRNP (Sof1p, Mpp10p, Imp3p, Imp4p, Dhr1p, Rrp9p and Bms1p). Rcl1p was only found in sub-stoichiometric amounts and Lcp5p was not found.131 The stoichiometry of the four proteins common to C/D-box snoRNPs was identical to that of the U3-specific proteins. In addition, none of the proteins common to H/ACA-box snoRNPs were identified in stoichiometric amounts. These two findings confirm the notion that most of the snoRNA-mediated modifications of rRNAs occurs cotranscriptionally and prior to the completion of the 35S pre-rRNA36,62 (see Fig. 2). About 20 novel proteins were also described, some of them common to the U3 RNP. Several of these novel proteins harbour WD-repeats, which have been described to participate in protein-protein interactions139 and two of these novel proteins, Utp7p and Utp14p, contain an adenylate- and an ATP/GTP-binding site, respectively.150 Interestingly, almost no factors involved in the assembly of 60S r-subunits were identified in the representative 90S preribosomal particles. This finding is in agreement with the absence of precursors to mature 25S rRNA within these particles. Therefore, it must be concluded that the 60S r-subunit synthesis machinery does not significantly assemble onto the 35S pre-rRNA. This is completely in disagreement with the predictions of the processosome model and provides a new picture of r-subunit assembly (compare top of the Fig. 1 with Fig. 2). Consistently, only few large subunit r-proteins are found within the different precursor particles to 40S r-subunits 131 suggesting that these do also only largely associate with pre60S ribosomal particles. As proposed by Fatica and Tollervey,15 the simplest explanation for this phenomenon is that the preribosomal assembly pathway reflects the 5'→3' directionality of pre-rRNA transcription.15 Indeed, there is enough evidence to conclude that the U3 RNP forms the so-called “terminal balls”, which are found at the tip of each elongating 35S pre-rRNA transcript, that are observed by electron microscopy of chromatin spreads150 (see top of Fig. 2). In addition, these results suggest that while the cis-acting elements required for the assembly of most r-proteins and trans-acting factors involved in 40S r-subunit biogenesis are available before 35S pre-rRNA transcription has been completed, the cis-acting elements required for the assembly of r-proteins and trans-acting factors involved in 60S r-subunit biogenesis might be hidden until pre-rRNA processing at site A2 has occurred.

The 43S Preribosomal Particles There is so far only one report that describes the isolation and characterization of 43S preribosomal particles.142 The analysis of the RNA composition of the different 90S particles

274

The Nucleolus

revealed that these particles contain no or little 20S pre-rRNA.131 Of the different TAP-tagged proteins used for purifying 90S particles, only the nuclear Enp1p coprecipitates significant amounts of 20S pre-rRNA.131 Moreover, primer extension showed that 20S pre-rRNA associated with Enp1p had undergone adenine dimethylation by Dim1p, which is reported to occur in the cytoplasm.131,151 Thus, it has been speculated that Enp1p is one of the factors that binds to nucleolar 90S but remains associated with cytoplasmic 43S preribosomal particles.131 Indeed, reexamination of the sedimentation of Enp1p-TAP on sucrose gradients revealed two distinct peaks, one at about 90S and a second one at about 40S. Analyses of the 40S peak indicated that only a few proteins were enriched, namely Enp1p, Dim1p, Hrr25p, Nob1p, Rrp12p, Rio2p, Tsr1p and Yor145cp.142 Before this report, only very few trans-acting factors that are specifically required for the processing of 20S to 18S had been described. In addition to Dim1p, which modifies the 20S pre-rRNA in the cytoplasm, these factors are expected to be associated with 43S preribosomal subunits. These include the small subunit r-proteins Rps0p, Rps14p and Ubi3p,137,152,153 the nuclear factors Tsr1p, Rrp12p and Nob1p41,103 and the cytoplasmic proteins Hcr1p and Rrp10p/ Rio1p.141,154 Interestingly, Rio2p is a homologue of Rrp10p/Rio1p.142 Nob1p is a protein with putative nuclease and RNA-binding domains. These features make it a candidate to be the long-sought 18S rRNA 3' endonuclease.41 Similarly to Enp1p, TAP-tagged Nob1p precipitates substantial amounts of adenine dimethylated 20S pre-rRNA species but small amounts of 35S pre-rRNA and U3 snoRNA.41,142 Rrp12p is unusual in that it seems not only to be involved in 20S pre-rRNA processing to 18S rRNA but also in efficient 3' processing of 7S pre-rRNA to mature 5.8S rRNA.41 In the specific work by the Hurt laboratory, complexes containing Dim1p, Hrr25p, Nob1p, Rio2p, Rrp12p or Tsr1p were purified and analysed.142 All these complexes contain similar protein and RNA compositions (see Fig. 3) and may represent components of the 43S preribosomal particles in the nucleus and/or cytoplasm. Rrp10p or Hcr1p could not be found, thus, suggesting that they are only transiently associated. All these complexes contain 20S pre-rRNA, but dimethylated 20S pre-rRNA is only found abundantly in Enp1p-, Hrr25pand Rio2p-containing complexes. Substantial amounts of 35S pre-rRNA are found associated with Dim1p-, Hrr25p- and Enp1p-containing particles. Finally, most of these proteins coprecipitate mature 18S rRNA, which likely reflects association with almost mature 40S r-subunits. These experiments were complemented by analyzing the subcellular localization of some of the above mentioned proteins. Enp1p was found predominantly in the nucleolus, Dim1p, Hrr25p, Rrp12p and Tsr1p were mostly nuclear although some cytoplasmic signal was detected. Finally, Nob1p, Rio2p and Yor145cp are predominantly cytoplasmic, however, while Rio2p and Yor145cp are likely incorporated into nuclear pre40S ribosomal particles, Nob1p seems to interact only with cytoplasmic pre40S ribosomal particles.142 Altogether, these analyses are consistent with a model where Dim1p, Enp1p, Tsr1p and maybe Rrp12p associate with 90S preribosomal particles and remain associated with nuclear 43S preribosomal particles after the early cleavages within the 35S pre-rRNA. Rio2p and maybe Hrr25p and Yor145cp associate only with nuclear 43S preribosomal particles, while Nob1p seems to interact only with late cytoplasmic pre40S ribosomal particles. Two other conclusions can be made. (i) Most of the trans-acting factors associated with the 90S preribosomal particle must dissociate following cleavages at sites A0, A1 and A2. Interestingly, many of these factors seem to still remain associated with the excised 5' ETS region that it is released by cleavage at site A0.41 (ii) Pre40S particles seem to lack AAA ATPases and other energy-consuming enzymes found associated with pre60S ribosomal particles. This finding, together with the fact that no previous reports describe a requirement for such enzymes in processsing 20S pre-rRNA to mature 18S rRNA, suggests that only few structural rearrangements are needed to finish the maturation of 40S r-subunits from early 43S preribosomal particles.

Ribosomal Subunit Assembly

275

Figure 3. Current model for the formation and maturation of the 43S preribosomal particle. The 43S preribosomal particle is formed after cleavage of the 35S pre-rRNA at sites A0, A1 and A2 and dissociation of most 40S r-assembly factors. Only few trans-acting factors (in light grey) from the 90S preribosomes are still associated with the 20S pre-rRNA-containing 43S preribosomal particle. Export of this particle requires Xpo1p, but so far no factor associated with this particle has been described to contain a nuclear export signal. Trans-acting factors with putative enzymatic activity are underlined. Nob1p might be the endonuclease cleaving the 20S pre-rRNA at site D. Note that the order of association and dissociation of factors is based on the protein composition of different purified particles. For many of these factors a functional implication in ribosome biogenesis has not yet been confirmed. Nucleolus, grey square. Preribosomal particles and mature r-subunits, light balloons. Nuclear envelope, rods.

276

The Nucleolus

The 66S Preribosomal Particles Undoubtedly, the maturation pathway of precursor particles to mature 60S r-subunits is the best defined so far. Many studies have reported several distinct complexes, amongst them the pioneer studies that validated the large-scale proteomic complex purification approaches.51,55 Importantly, several of the selected nonribosomal proteins chosen for purification are located in different cellular compartments ranging from the nucleolus, through the nucleoplasm to the cytoplasm. This has made it possible to obtain a series of snapshots of distinct precursor particles to mature 60S r-subunits. The current model for 60S r-subunit assembly distinguishes at least five intermediates, which are designated early 1 (E1), early 2 (E2), medium (M), late (L) and cytoplasmic (see Fig. 4). The so-called E1 intermediate is defined by its purification with a TAP-tagged version of Ssf1p and more recently by its purification with Nsa3p-TAP.94,147 This complex sediments in a broad peak from 60S to 90S. The most abundant RNA species in these particles is the 27SA2 pre-rRNA followed by 27SB and 7S pre-rRNAs. Almost no 35S pre-rRNA is detected. Interestingly, mature 5S rRNA coprecipitates with this and further particles. Since Rpl5p, which is the sole 5S rRNA associated r-protein,155 is also found in Nsa3p-containing particles, it can be concluded that the assembly of the Rpl5p-5S RNP occurs in early precursors to 60S r-subunits. This is consistent with genetic studies that show that different 5S rRNA mutants only affect the efficiency of 25S rRNA maturation.156 Therefore, the proteomic data are in disagreement with the prediction of the processosome model which assigns the assembly of Rpl5p-5S RNP into a 90S preribosomal particle (compare Fig. 1 and Fig. 4). Regarding its protein composition, the E1 particle comprises about 50 nonribosomal proteins and 28 r-proteins from the 60S subunit (see Fig. 4). Amongst the different nonribosomal proteins, there are some previously characterized factors involved in early steps during the synthesis of 60S r-subunits, including the exonuclease Xrn1p61 and the nucleolar proteins Nop4p, Rrp1p, Brx1p and Erb1p.32,157-159 In addition, the E1 particle includes a large number of factors proposed to participate in late steps during 60S r-subunit synthesis, such as the putative RNA helicases Dbp10p and Spb4p,84,160 the putative methyltransferases Nop2p and Spb1p,116,117 the nucleolar proteins Nip7p and Tif6p,110,161 and even the cytoplasmic protein Sqt1p.136 This suggests that the timepoint of association of the different trans-acting factors with the precursor particles to 60S r-subunits does not really reflect the timepoint of function. Similarly, at least Dim1p, Enp1p and Rrp12p seem to already associate with 90S preribosomal particles when they are required for late steps during processing or modification of 20S pre-rRNA (see above). Recruitment of late r-assembly factors in early particles may operate as a quality control mechanism that ensures that all processing factors needed for subsequent steps are associated correctly in order to rapidly and effectively mature both r-subunits. Interestingly, amongst the different nonribosomal proteins associated with the E1 particle, there are only few trans-acting factors involved in the early pre-rRNA processing reactions at sites A0, A1 and A2. These include Nop56p and Nop58p, which are components of the C/ D-box snoRNPs,75,162 Rrp8p, which is a putative methyltransferase,33 Rrp9p, which is a U3-specific protein163 and Nop14p.146 It is worthwhile to mention that a nop56 ts-mutant inhibits early steps in pre-rRNA processing but also leads to a strong delay of processing of 27SA and 27SB pre-rRNAs.164 This result suggests that Nop56p might also have a function within preribosomal precursors to 60S r-subunits. However, no role during 60S r-subunit synthesis has been assigned to Rrp8p, Rrp9p and Nop14p. This implies that there exists some association between the 90S and the 66S E1 preribosomal particles. Rrp5p should also be specially mentioned since it is the only trans-acting factor known to be required for pre-rRNA processing at both the early A0 to A2 sites and at site A3, which is the first step that is specific to the pathway leding to mature 5.8S and 25S rRNAs.126 Another interesting protein found associated with the E1 particle is the putative RNA helicase Has1p. Mutation in or depletion of Has1p leads to a deficit in 40S r-subunits due to a clear inhibition of the early cleavages within the 35S pre-rRNA.73 This is consistent with the presence of Has1p in some purified particles

Ribosomal Subunit Assembly

277

Figure 4. Current model for the formation and maturation of 66S preribosomal particles. A series of distinct particles are predicted to be intermediates during the synthesis of 60S r-subunits. These are termed, according to their position in the pathway, early 1 (E1), early 2 (E2) and middle (M) 66S preribosomal particles and late (L) and cytoplasmic 60S preribosomal particles. The E1 particle is the result of the association of more than 30 nonribosomal proteins and most large subunit r-proteins with the 27SA pre-rRNAs. Evidence suggests that the Rpl5p-5S RNP assembles with E1 particles. Presence of the 27SB pre-rRNA defines the E2 particle. Processing of 27SB pre-rRNA at site C2, which generates the 25.5S and 7S precursors, leads to the formation of the M particle. Exonucleolytic trimming of the 7S pre-rRNA to the 3' end of the mature 5.8S rRNA and of the 25.5S pre-rRNA to the 5' end of the mature 25S rRNA precede the formation of the 60S preribosomal L particle. Nuclear export of the L particle is promoted by the interaction of the nuclear export adapter Nmd3p, which contains a nuclear export signal, with the export receptor Xpo1p. The cytoplasmic 60S preribosomal particle undergoes final assembly reactions that lead to the formation of mature translation-competent 60S r-subunits. The order of the association and dissociation of the different trans-acting factors is based on the protein composition of the distinct purified complexes. Some factors (i.e., Arx1p) are associated with all preribosomal particles, many with different particles (i.e., Spb4p) and others with only a single particle (i.e., Nop4p). Trans-acting factors with putative enzymatic activity are underlined. Note that many pre-rRNA processing factors, such as endo- and exonucleases are not present in these particles. For many of the trans-acting factors listed in this figure, a functional implication in ribosome biogenesis has not yet been confirmed. The r-proteins most likely present in the different preribosomal particles are also shown. Nucleolus, grey square. Preribosomal particles and mature r-subunits, light balloons. Nuclear envelope, rods.

278

The Nucleolus

related to 90S preribosomes.131 In addition, we have also detected a mild accumulation of 27SA3 and 27SB pre-rRNAs after depletion of Has1p.73 Therefore, these results suggest that Has1p may have a role in both 40S and 60S r-subunit formation. While its interaction with different 66S particles is quite stable, its interaction with the 90S preribosome may be rather transient. The so-called E2 complex is defined by the Nop7p-containing particle.55 Nop7p runs from 60S to 80S in sucrose gradients. The RNA composition of this particle reveals a clear enrichment of 27SB, 25.5S and 7S pre-rRNAs. In clear contrast to the E1 particle, there is no significant association with the 27SA2 pre-rRNA. This reflects that pre-rRNA processing at site A3 occurs prior to the formation of the E2 particle. Regarding its protein composition, only few differences are found with respect to the E1 particle. This suggests that only few trans-acting factors dissociate from the E1 particle following cleavage at site A3 and that only few others join the E2 particle (see Fig. 4). Rrp5p is an example of a factor that may dissociate since it is only required for cleavage at site A3 on the pathway to mature 5.8S and 25S rRNAs. The M particle is represented by the purification of complexes associated with TAP-tagged versions of the nuclear Nug1p, Rix1p and Sda1p.94 These particles sediment in a more or less confined peak at 60S in sucrose gradients. Their RNA composition clearly shows that they are deficient in 27SA pre-rRNAs. While the Nug1p-containing particle is clearly enriched in 27SB pre-rRNA, mildly enriched in 25.5S and 7S pre-rRNAs, and lacks mature 5.8S and 25S rRNAs, the Rix1p-containing particle does not have substantial amounts of 27SB pre-rRNA, but contains abundant amounts of 25.5S and 7S pre-rRNA and some mature 5.8S and 25S rRNAs. Sda1p did not detectably coprecipitate any pre-rRNA, leading to the conclusion that it is predominantly associated with the mature 5.8S and 25S rRNAs. Altogether, this suggests that the M particles are the snapshots for processing at site C2, 3' trimming of 7S pre-rRNA to 5.8S rRNA and 5' trimming of 25.5S pre-rRNA to 25S rRNA. Regarding protein composition, there is a clear reduction of complexity from the Nug1p- to the Sda1p-containing particle. This reduction of complexity suggests an additional step of dissociation of trans-acting factors concomitant to late pre-rRNA processing reactions (see Fig. 4). Amongst the dissociating factors are Ssf1p and Ssf2p, which prevent the premature pre-rRNA cleavage at site C2.147 Also, during the progression from the Nug1p- to the Sda1p-containing particle, the Noc1p/Noc2p complex is replaced by the Noc2p/Noc3p complex, which finally dissociates from the Sda1p-containing intermediate.94 This is in agreement with the fact that Nug1p is a nuclear protein but Sda1p is predominantly nucleoplasmic.51,165 Few factors associate with the M particles, including Rix1p, which has been described to be required for nuclear export of pre60S particles.51 Pre60S L particles are defined by the purification of complexes associated with TAP-tagged versions of Arx1p and Nog2p.94 These complexes run at a confined peak at 60S in sucrose gradients. Analyses of RNA composition reveal, for the first time, clear accumulation of mature 5.8S and 25S rRNAs, thus suggesting that maturation of both rRNAs is largely completed at the nuclear stage of the Arx1p-containing particle. Only 10-15 nonribosomal proteins were found to coprecipitate with Arx1p. It is worthwhile to mention that Mtr2p and Nmd3p, which are required for nuclear export of pre60S particles,45,51,138 are amongst these proteins. The cytoplasmic pre60S precursor corresponds to the Kre35p-containing complex. It contains exclusively the mature 5.8S and 25S rRNAs and most of the large subunit r-proteins. Rpp0p, which forms together with Rpp1p and Rpp2p the ribosomal stalk,166 is only detected in cytoplasmic 60S preribosomal particles. This suggests that Rpp0p, like Rpp1p and Rpp2p, mainly assemble late in the 60S r-subunit pathway, however, Rpp0p is not an exchangeable r-protein.166 Some r-proteins from the small subunit coprecipitate with Kre35p, thus, suggesting some association with 40S r-subunits in the cytoplasm.94 Few nonribosomal proteins remain associated with this particle, including Nmd3p, Sqt1p and Arx1p (see Fig. 4). Cytoplasmic Efl1p must have transiently associated with the Kre35p-containing particle since it lacks Tif6p.94 The final dissociation of these trans-acting factors and the assembly of the ribosomal stalk will convert the cytoplasmic particles into mature translation-competent 60S r-subunits.

Ribosomal Subunit Assembly

279

Ribosomal Subunit Assembly in Higher Eukaryotes Ribosome synthesis is not as well characterized in other organisms as it is in yeast, mainly due to the lack of comparable tools for functional analysis, including powerful genetic manipulations. Recently, proteomic analyses of human nucleoli have been reported.167-170 This has provided us with the first catalogue of human trans-acting factors involved in ribosome synthesis. Rapid progress in the understanding of human ribosome synthesis is expected within the next years to stem from the purification of distinct mammalian preribosomal particles (for an example see ref. 171) and the application of small interference RNA technology to suppress the expression of particular mammalian genes (see ref. 172 and references therein). Experimental evidence suggests that the basic outline of ribosome synthesis is conserved throughout eukaryotes and even in archaebacteria. (i) The primary structure of the pre-rRNA and the processing reactions are similar among a wide range of eukaryotes.173 In addition, the assembly reactions follow identical kinetics in yeast and HeLa cells.16,17 (ii) Many of the trans-acting factors are well conserved in evolution from yeast to man, corroborating the notion that ribosome assembly is related in these organisms. Strikingly, many of the eukaryotic trans-acting factors have also homologues in the genomes of different archaebacteria but not of eubacteria (for an example see ref. 174). (iii) Indeed, there are examples of similarity between the effects on pre-rRNA processing caused by inactivation of homologous (orthologous) trans-acting factors in yeast and higher eukaryotic cells.159,175,176 These similarities suggest that these homologues perform similar or identical functions in the different eukaryotes and that they are therefore most likely orthologues. (iv) Finally, there are even reports describing the functional complementation of yeast trans-acting factor null mutants by, for example, the respective orthologous human, mouse or Drosophila genes (for examples see refs. 140, 149, 177).

Conclusions and Perspectives From the above description, it is clear that, ribosome assembly is a very dynamic process. Many different structural rearrangements are expected to occur within the distinct preribosomal particles. The high complexity of the process itself together with its compartmentalization in eukaryotes offer an explanation for the involvement of the large number of essential nonredundant trans-acting factors. In the last few years many new trans-acting factors have been identified, first by using classical and reverse yeast genetics and more recently by the purification of preribosomal particles. The proteomic approach has also allowed to refine our understanding of the r-subunit assembly pathway. Importantly, we have learned that ribosome synthesis appears to be a biphasic process in which maturation of 40S and 60S r-subunits is separated. However, we have only scratched the surface of ribosome biogenesis. A clear molecular role for most of the identified trans-acting factors is still missing. Indeed, we do hardly know the precise substrate for any of these factors. This knowledge is essentially required in order to understand the different structural rearrangements that take place within the preribosomal particles during the assembly reactions. The thorough application of currently available techniques as well as the development of novel assays should allow to deal with purified assembly intermediates or fractions of them in order to know the precise location of trans-acting factors, their substrate specificity and timing of action.

Acknowledgments We thank our colleagues who have contributed unpublished results to this book chapter and we apologize to all those whose work could not be cited owing to space limitations. J. de la Cruz acknowledges the Spanish Ministry of Science and Technology (BMC2001-2660) and the Andalusian Regional Government (CVI 271) for support. Work in the Linder laboratory is supported by grants from the Swiss National Foundation.

280

The Nucleolus

References 1. Lafontaine DL, Tollervey D. The function and synthesis of ribosomes. Nature Rev Mol Cell Biol 2001; 2:514-520. 2. Planta RJ, Mager WH. The list of cytoplasmic ribosomal proteins of Saccharomyces cerevisiae. Yeast 1998; 14:471-477. 3. Spahn CM, Beckmann R, Eswar N et al. Structure of the 80S ribosome from Saccharomyces cerevisiae-tRNA-ribosome and subunit-subunit interactions. Cell 2001; 107:373-386. 4. Ban N, Nissen P, Hansen J et al. The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution. Science 2000; 289:905-920. 5. Wimberly BT, Brodersen DE, Clemons WM et al. Structure of the 30S ribosomal subunit. Nature 2000; 407:327-339. 6. Mizushima S, Nomura M. Assembly mapping of 30S ribosomal proteins in E. coli. Nature 1970; 226:1214-1218. 7. Rohl R, Nierhaus KH. Assembly map of the large subunit (50S) of Escherichia coli ribosomes. Proc Natl Acad Sci USA 1982; 79:729-733. 8. Sánchez ME, Londei P, Amils R. Total reconstitution of active small ribosomal subunits of the extreme halophilic archaeon Haloferax mediterranei. Biochim Biophys Acta 1996; 1292:140-144. 9. Sánchez ME, Urena D, Amils R et al. In vitro reassembly of active large ribosomal subunits of the halophilic archaebacterium Haloferax mediterranei. Biochemistry 1990; 29:9256-9261. 10. Maki JA, Schnobrich DJ, Culver GM. The DnaK chaperone system facilitates 30S ribosomal subunit assembly. Mol Cell 2002; 10:129-138. 11. Woolford J. Chaperoning ribosome assembly. Mol Cell 2002; 10:8-10. 12. Mangiarotti G, Chiaberge S. Reconstitution of functional eukaryotic ribosomes from Dictyostelium discoideum ribosomal proteins and RNA. J Biol Chem 1997; 272:19682-19687. 13. Kressler D, Linder P, de la Cruz J. Protein trans-acting factors involved in ribosome biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19:7897-7912. 14. Venema J, Tollervey D. Ribosome synthesis in Saccharomyces cerevisiae. Annu Rev Genet 1999; 33:261-311. 15. Fatica A, Tollervey D. Making ribosomes. Curr Opin Cell Biol 2002; 14:313-318. 16. Warner JR, Soeiro R. Nascent ribosomes from HeLa cells. Proc Natl Acad Sci USA 1967; 58:1984-1990. 17. Warner JR. The assembly of ribosomes in yeast. J Biol Chem 1971; 246:447-454. 18. Trapman J, Retèl J, Planta RJ. Ribosomal precursor particles from yeast. Exp Cell Res 1975; 90:95-104. 19. Trapman J, Planta RJ. Maturation of ribosomes in yeast. I. Kinetic analysis by labelling of high molecular weight rRNA species. Biochim Biophys Acta 1976; 442:265-274. 20. Trapman J, Planta RJ, Raué HA. Maturation of ribosomes in yeast. II. Position of the low molecular weight rRNA species in the maturation process. Biochim Biophys Acta 1976; 442:275-284. 21. Kruiswijk T, Planta RJ, Krop JM. The course of the assembly of ribosomal subunits in yeast. Biochim Biophys Acta 1978; 517:378-389. 22. Zinker S, Warner JR. The ribosomal proteins of Saccharomyces cerevisiae Phosphorylated and exchangeable proteins. J Biol Chem 1976; 251:1799-1807. 23. Dick FA, Eisinger DP, Trumpower BL. Exchangeability of Qsr1p, a large ribosomal subunit protein required for subunit joining, suggests a novel translational regulatory mechanism. FEBS Lett 1997; 419:1-3. 24. Yeh L-CC, Lee JC. Yeast ribosomal proteins L4, L17, L20 and L25 exhibit different binding characteristics for the yeast 35S precursor rRNA. Biochim Biophys Acta 1998; 1443:139-148. 25. El-Baradi TTAL, Raué HA, de Regt VCHF et al. Stepwise dissociation of yeast 60S ribosomal subunits by LiCl and identification of L25 as a primary 26S rRNA binding protein. Eur J Biochem 1984; 144:393-400. 26. Lee JC, Anderson R, Yeh YC et al. Extraction of proteins from Saccharomyces cerevisiae ribosomes under nondenaturing conditions. Arch Biochem Biophys 1985; 237:292-299. 27. Mitchell P, Petfalski E, Tollervey D. The 3' end of yeast 5.8S rRNA is generated by an exonuclease processing mechanism. Genes Dev 1996; 10:502-513. 28. Warner JR, Udem SA. Temperature sensitive mutations affecting ribosome synthesis in Saccharomyces cerevisiae. J Mol Biol 1972; 65:243-257. 29. Ripmaster TL, Vaughn GP, Woolford Jr JL. DRS1 to DRS7, novel genes required for ribosome assembly and function in Saccharomyces cerevisiae. Mol Cell Biol 1993; 13:7901-7912. 30. Sasaki T, To-e A, Kikuchi Y. Yeast Krr1p physically and functionally interacts with a novel essential Kri1p, and both proteins are required for 40S ribosome biogenesis in the nucleolus. Mol Cell Biol 2000; 20:7971-7979.

Ribosomal Subunit Assembly

281

31. Daugeron M-C, Kressler D, Linder P. Dbp9p, a putative ATP-dependent RNA helicase involved in 60S-ribosomal-subunit biogenesis, functionally interacts with Dbp6p. RNA 2001; 7:1317-1334. 32. Wehner KA, Baserga SJ. The sigma(70)-like motif: A eukaryotic RNA binding domain unique to a superfamily of proteins required for ribosome biogenesis. Mol Cell 2002; 9:329-339. 33. Bousquet-Antonelli C, Vanrobays E, Gélugne J-P et al. Rrp8p is a yeast nucleolar protein functionally linked to Gar1p and involved in pre-rRNA cleavage at site A2. RNA 2000; 6:826-843. 34. Kressler D, Doère M, Rojo M et al. Synthetic lethality with conditional dbp6 alleles identifies the previously uncharacterized RSA1 gene, encoding a protein involved in a late nucleoplasmic step of 60S-ribosomal-subunit assembly. Mol Cell Biol 1999; 19:8633-8645. 35. Filipowicz W, Pogacic V. Biogenesis of small nucleolar ribonucleoproteins. Curr Opin Cell Biol 2002; 14:319-327. 36. Kiss T. Small nucleolar RNA-guided post-transcriptional modification of cellular RNAs. EMBO J 2001; 20:3617-3622. 37. Moritz M, Paulovich AG, Tsay Y-F et al. Depletion of yeast ribosomal proteins L16 or rp59 disrupts ribosome assembly. J Cell Biol 1990; 111:2261-2274. 38. Moritz M, Pulaski BA, Woolford Jr JL et al. Assembly of 60S ribosomal subunits is perturbed in temperature sensitive yeast mutants defective in ribosomal protein L16. Mol Cell Biol 1991; 11:5681-5692. 39. Vilardell J, Warner JR. Ribosomal protein L32 of Saccharomyces cerevisiae influences both the splicing of its own transcript and the processing of rRNA. Mol Cell Biol 1997; 17:1959-1965. 40. Mitchell P, Petfalski E, Shevchenko A et al. The exosome: A conserved eukaryotic RNA processing complex containing multiple 3'→5' exoribonucleases. Cell 1997; 91:457-466. 41. Fatica A, Oeffinger M, Dlakic M et al. Nob1p is required for cleavage of the 3' end of 18S rRNA. Mol Cell Biol 2003; 23:1798-1807. 42. Venema J, Tollervey D. Processing of preribosomal RNA in Saccharomyces cerevisiae. Yeast 1995; 11:1629-1650. 43. Tollervey D. Trans-acting factors in ribosome synthesis. Exp Cell Res 1996; 229:226-232. 44. Hurt E, Hannus S, Schmelzl B et al. A novel in vitro assay reveals inhibition of ribosomal nuclear export in Ran-cycle and nucleoporin mutants. J Cell Biol 1999; 144:389-401. 45. Gadal O, Strauß D, Kessl J et al. Nuclear export of 60S ribosomal subunit depends on Xpo1p and requires a nuclear export sequence-containing factor, Nmd3p, that associates with the large subunit protein Rpl10p. Mol Cell Biol 2001; 21:3405-3415. 46. Stage-Zimmermann T, Schmidt U, Silver PA. Factors affecting nuclear export of the 60S ribosomal subunit in vivo. Mol Biol Cell 2000; 11:3777-3789. 47. Moy TI, Silver P. Nuclear export of the small ribosomal subunit requires the Ran-GTP cycle and certain nucleoporins. Genes Dev 1999; 13:2118-2133. 48. Moy TI, Silver PA. Requirements for the nuclear export of the small ribosomal subunit. J Cell Sci 2002; 115:2985-2995. 49. Gadal O, Strauss D, Petfalski E et al. Rlp7p is associated with 60S preribosomes, restricted to the granular component of the nucleolus, and required for pre-rRNA processing. J Cell Biol 2002; 157:941-951. 50. Gadal O, Strauss D, Braspenning J et al. A nuclear AAA-type ATPase (Rix7p) is required for biogenesis and nuclear export of 60S ribosomal subunits. EMBO J 2001; 20:3695-3704. 51. Baßler J, Grandi P, Gadal O et al. Identification of a 60S preribosomal particle that is closely linked to nuclear export. Mol Cell 2001; 8:517-529. 52. Rigaut G, Shevchenko A, Rutz B et al. A generic protein purification method for protein complex characterization and proteome exploration. Nature Biotechnol 1999; 17:1030-1032. 53. Venema J, Planta RJ, Raué HA. In vivo mutational analysis of ribosomal RNA in Saccharomyces cerevisiae. In: Martin R, ed. Protein synthesis: Methods and protocols. Totowa: Humana Press, 1998:257-270. 54. Kumar A, Snyder M. Protein complexes take the bait. Nature 2002; 415:123-124. 55. Harnpicharnchai P, Jakovljevic J, Horsey E et al. Composition and functional characterization of yeast 66S ribosome assembly intermediates. Mol Cell 2001; 8:505-515. 56. Gavin AC, Bosche M, Krause R et al. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 2002; 415:141-147. 57. Ho Y, Gruhler A, Heilbut A et al. Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 2002; 415:180-183. 58. Warner JR. Nascent ribosomes. Cell 2001; 107:133-136. 59. de la Cruz J, Kressler D, Tollervey D et al. Dob1p (Mtr4p) is a putative ATP-dependent RNA helicase required for the 3' end formation of 5.8S rRNA in Saccharomyces cerevisiae. EMBO J 1998; 17:1128-1140.

282

The Nucleolus

60. Weaver PL, Sun C, Chang T-H. Dbp3p, a putative RNA helicase in Saccharomyces cerevisiae, is required for efficient pre-rRNA processing predominantly at site A 3. Mol Cell Biol 1997; 17:1354-1365. 61. Henry Y, Wood H, Morrissey JP et al. The 5' end of yeast 5.8S rRNA is generated by exonucleases from an upstream cleavage site. EMBO J 1994; 13:2452-2463. 62. Tollervey D, Kiss T. Function and synthesis of small nucleolar RNAs. Curr Opin Cell Biol 1997; 9:337-342. 63. Ramakrishnan V, White SW. Ribosomal protein structures: Insights into the architecture, machinery and evolution of the ribosome. Trends Biochem Sci 1998; 23:208-212. 64. Linder P. DEAD-box proteins. Curr Biol 2001:R887. 65. de la Cruz J, Kressler D, Linder P. Unwinding RNA in Saccharomyces cerevisiae: DEAD-box proteins and related families. Trends Biochem Sci 1999; 24:192-198. 66. Tanner NK, Linder P. DExD/H box RNA helicases: from generic motors to specific dissociation functions. Mol Cell 2001; 8:251-262. 67. Linder P, Daugeron M-C. Are DEAD-box proteins becoming respectable helicases? Nature Struct Biol 2000; 7:97-99. 68. Jankowsky E, Gross CH, Shuman S et al. The DExH protein NPH-II is a processive and directional motor for unwinding RNA. Nature 2000; 403:447-451. 69. Jankowsky E, Gross CH, Shuman S et al. Active disruption of an RNA-protein interaction by a DExH/D RNA helicase. Science 2000; 291:121-124. 70. Linder P, Tanner KN, Banroques J. From RNA helicases to RNPases. Trends Biochem Sci 2001; 26:339-341. 71. Bond AT, Mangus DA, He F et al. Absence of Dbp2p alters both nonsense-mediated mRNA decay and rRNA processing. Mol Cell Biol 2001; 21:7366-7379. 72. Colley A, Beggs JD, Tollervey D et al. Dhr1p, a putative DEAH-box RNA helicase, is associated with the box C+D snoRNP U3. Mol Cell Biol 2000; 20:7238-7246. 73. Emery B, de la Cruz J, Linder P. Unpublished data. 2002. 74. Decatur WA, Fournier MJ. rRNA modifications and ribosome function. Trends Biochem Sci 2002; 27:344-351. 75. Lafontaine D, Tollervey D. Synthesis and assembly of the box C+D small nucleolar RNPs. Mol Cell Biol 2000; 20:2650-2659. 76. Venema J, Bousquet-Antonelli C, Gelugne J-P et al. Rok1p is a putative RNA helicase required for rRNA processing. Mol Cell Biol 1997; 17:3398-3407. 77. Liang W-Q, Clark JA, Fournier MJ. The rRNA-processing function of the yeast U14 small nucleolar RNA can be rescued by a conserved RNA helicase-like protein. Mol Cell Biol 1997; 17:4124-4132. 78. Daugeron M-C, Linder P. Characterization an mutational analysis of yeast Dbp8p, a putative RNA helicase involved in ribosome biogenesis. Nucleic Acids Res 2001; 29:1144-1155. 79. Kressler D, de la Cruz J, Rojo M et al. Fal1p is an essential DEAD-box protein involved in 40S-ribosomal-subunit biogenesis in Saccharomyces cerevisiae. Mol Cell Biol 1997; 17:7283-7294. 80. O’Day CL, Chavanikamannil F, Abelson J. 18S rRNA processing requires the RNA helicase-like protein Rrp3. Nucleic Acids Res 1996; 24:3201-3207. 81. Kressler D, de la Cruz J, Rojo M et al. Dbp6p is an essential putative ATP-dependent RNA helicase required for 60S-ribosomal-subunit assembly in Saccharomyces cerevisiae. Mol Cell Biol 1998; 18:1855-1865. 82. Daugeron M-C, Linder P. Dbp7p, a putative ATP-dependent RNA helicase from Saccharomyces cerevisiae, is required for 60S ribosomal subunit assembly. RNA 1998; 4:566-581. 83. Zagulski M, Kressler D, Bécam A-M et al. Mak5p, which is required for the maintenance of the M1 dsRNA virus, is encoded by the yeast ORF YBR142w and involved in the biogenesis of the 60S subunit of the ribosome. 2003. Submitted. 84. Burger F, Daugeron M-C, Linder P. Dbp10p, a putative RNA helicase from Saccharomyces cerevisiae required for ribosome biogenesis. Nucleic Acids Res 2000; 28:2315-2323. 85. Adams CC, Jakovljevic J, Roman J et al. Saccharomyces cerevisiae nucleolar protein Nop7p is necessary for biogenesis of 60S ribosomal subunits. RNA 2002; 8:150-165. 86. Kressler D, Linder P. Unpublished data. 2002. 87. Oh J-Y, Kim J. ATP hydrolysis activity of the DEAD box protein Rok1p is required for in vivo ROK1 function. Nucleic Acids Res 1999; 27:2753-2759. 88. Neuwald A, Aravind L, Spouge JL et al. AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res 1999; 9:27-43. 89. Vale RD. AAA proteins: Lords of the ring. J Cell Biol 2000; 150:F13-F19. 90. Patel S, Latterich M. The AAA team: Related ATPases with diverse functions. Trends Cell Biol 1998; 8:65-71.

Ribosomal Subunit Assembly

283

91. Reid BG, Fenton WA, Horwich AL et al. ClpA mediates directional translocation of substrate proteins into the ClpP protease. Proc Natl Acad Sci USA 2001; 98:3768-3772. 92. Hartman JJ, Vale RD. Microtubule disassembly by ATP-dependent oligomerization of the AAA enzyme katanin. Science 1999; 286:782-785. 93. Brown JD. Ribosome biogenesis: Stripping for AAAction? Curr Biol 2001; 11:R710-R712. 94. Nissan TA, Bassler J, Petfalski E et al. 60S preribosome formation viewed from assembly in the nucleolus until export to the cytoplasm. EMBO J 2002; 21:5539-5547. 95. Kuersten S, Ohno M, Mattaj IW. Nucleocytoplasmic transport: Ran, beta and beyond. Trends Cell Biol 2001; 11:497-503. 96. Gulli MP, Peter M. Temporal and spatial regulation of Rho-type guanine-nucleotide exchange factors: the yeast perspective. Genes Dev 2001; 15:365-379. 97. Green R. Ribosomal translocation: EF-G turns the crank. Curr Biol 2000; 10:R369-R373. 98. Rodnina MV, Savelsbergh A, Wintermeyer W. Dynamics of translation on the ribosome: molecular mechanics of translocation. FEMS Microbiol Rev 1999; 23:317-333. 99. Smith FD, Scott JD. Signaling complexes: junctions on the intracellular information super highway. Curr Biol 2002; 12:R32-R40. 100. Leipe DD, Wolf YI, Koonin EV et al. Classification and evolution of P-loop GTPases and related ATPases. J Mol Biol 2002; 317:41-72. 101. Bourne HR, Sanders DA, McCormick F. The GTPase superfamily: conserved structure and molecular mechanism. Nature 1991; 349:117-127. 102. Sprang SR. G proteins, effectors and GAPs: structure and mechanism. Curr Opin Struct Biol 1997; 7:849-856. 103. Gelperin D, Horton L, Beckman J et al. Bms1p, a novel GTP-binding protein, and the related Tsr1p are required for distinct steps of 40S ribosome biogenesis in yeast. RNA 2001; 7:1268-1283. 104. Wegierski T, Billy E, Nasr F et al. Bms1p, a G-domain-containing protein, associates with Rcl1p and is required for 18S rRNA biogenesis in yeast. RNA 2001; 7:1254-1267. 105. Park J-H, Jensen BC, Kifer CT et al. A novel nucleolar G-protein conserved in eukaryotes. J Cell Sci 2000; 114:173-185. 106. Saveanu C, Bienvenu D, Namane A et al. Nog2p, a putative GTPase associated with pre60S subunits and required for late 60S maturation steps. EMBO J 2001; 20:6475-6484. 107. Suzuki N, Noguchi E, Nakashima N et al. The Saccharomyces cerevisiae small GTPase, Gsp1p/Ran, is involved in 3' processing of 7S-to-5.8S rRNA and in degradation of the excised 5'-A0 fragment of 35S pre-rRNA, both of which are carried out by the exosome. Genetics 2001; 158:613-625. 108. Mitchell P, Tollervey D. Musing on the structural organization of the exosome complex. Nature Struct Biol 2000; 7:843-846. 109. Becam AM, Nasr F, Racki WJ et al. Ria1p (Ynl163c), a protein similar to elongation factors 2, is involved in the biogenesis of the 60S subunit of the ribosome in Saccharomyces cerevisiae. Mol Genet Genomics 2001; 266:454-462. 110. Senger B, Lafontaine DL, Graindorge JS et al. The nucle(ol)ar Tif6p and Efl1p are required for a late cytoplasmic step of ribosome synthesis. Mol Cell 2001; 8:1363-1373. 111. Basu U, Si K, Warner JR et al. The Saccharomyces cerevisiae TIF6 gene encoding translation initiation factor 6 is required for 60S ribosomal subunit biogenesis. Mol Cell Biol 2001; 75:1453-1462. 112. Niewmierzycka A, Clarke S. S-Adenosylmethionine-dependent methylation in Saccharomyces cerevisiae. J Biol Chem 1999; 274:814-824. 113. Lafontaine D, Delcour J, Glasser A-L et al. The DIM1 gene responsible for the conserved m6,2Am6,2A dimethylation in the 3'-terminal loop of 18S rRNA is essential in yeast. J Mol Biol 1994; 241:492-497. 114. Lafontaine D, Vandenhaute J, Tollervey D. The 18S rRNA dimethylase Dim1p is required for preribosomal RNA processing in yeast. Genes Dev 1995; 9:2470-2481. 115. Wu P, Brockenbrough JS, Paddy MR et al. NCL1, a novel gene for a nonessential nuclear protein in Saccharomyces cerevisiae. Gene 1998; 220:109-117. 116. Hong B, Brockenbrough JS, Wu P et al. Nop2p is required for pre-rRNA processing and 60S ribosome subunit synthesis in yeast. Mol Cell Biol 1997; 17:378-388. 117. Kressler D, Rojo M, Linder P et al. Spb1p is a putative methyltransferase required for 60S ribosomal subunit biogenesis in Saccharomyces cerevisiae. Nucleic Acids Res 1999; 27:4598-4608. 118. Pintard L, Kressler D, Lapeyre B. Spb1p is a yeast nucleolar protein associatd with Nop1p and Nop58p that is able to bind S-adenosyl-L-methionine in vitro. Mol Cell Biol 2000; 20:1370-1381. 119. Galardi S, Fatica A, Bachi A et al. Purified box C/D snoRNPs are able to reproduce site-specific 2'-O- methylation of target RNA in vitro. Mol Cell Biol 2002; 22:6663-6668. 120. Lafontaine DLJ, Preiss T, Tollervey D. Yeast 18S rRNA dimethylase Dim1p: A quality control mechanism in ribosome synthesis? Mol Cell Biol 1998; 18:2360-2370.

284

The Nucleolus

121. Fath S, Milkereit P, Podtelejnikov AV et al. Association of yeast RNA polymerase I with a nucleolar substructure active in rRNA synthesis and processing. J Cell Biol 2000; 149:575-590. 122. Cullen BR. Nuclear RNA export pathways. Mol Cell Biol 2000; 20:4181-4187. 123. Strasser K, Masuda S, Mason P et al. TREX is a conserved complex coupling transcription with messenger RNA export. Nature 2002; 417:304-308. 124. Proudfoot NJ, Furger A, Dye MJ. Integrating mRNA processing with transcription. Cell 2002; 108:501-512. 125. Linder P, Stutz F. mRNA export: Travelling with DEAD box proteins. Curr Biol 2001; 11:R961-963. 126. Venema J, Tollervey D. RRP5 is required for formation of both 18S and 5.8S rRNA in yeast. EMBO J 1996; 15:5701-5714. 127. Dunbar DA, Dragon F, Lee SJ et al. A nucleolar protein related to ribosomal protein L7 is required for an early step in large ribosomal subunit biogenesis. Proc Natl Acad Sci USA 2000; 97:13027-13032. 128. Lee SJ, Baserga SJ. Imp3p and Imp4p; two specific components of the U3 small nucleolar ribonucleoprotein that are essential for 18S rRNA processing. Mol Cell Biol 1999; 19:5441-5452. 129. Watkins NJ, Gottschalk A, Neubauer G et al. Cbf5p, a potential pseudouridine synthase, and Nhp2p, a putative RNA-binding protein, are present together with Gar1p in all H BOX/ACA-motif snoRNPs and constitute a common bipartite structure. RNA 1998; 4:1549-1568. 130. Watkins NJ, Segault V, Charpentier B et al. A common core RNP structure shared between the small nucleoar box C/D RNPs and the spliceosomal U4 snRNP. Cell 2000; 103:457-466. 131. Grandi P, Rybin V, Bassler J et al. 90S preribosomes include the 35S pre-rRNA, the U3 snoRNP, and 40S subunit processing factors but predominantly lack 60S synthesis factors. Mol Cell 2002; 10:105-115. 132. Tabb AL, Utsugi T, Wooten-Kee CR et al. Genes encoding ribosomal proteins Rps0A/B of Saccharomyces cerevisiae interact with TOM1 mutants defective in ribosome synthesis. Genetics 2001; 157:1107-1116. 133. Baudin-Baillieu A, Tollervey D, Cullin C et al. Functional analysis of Rrp7p, an essential yeast protein involved in pre-rRNA processing and ribosome assembly. Mol Cell Biol 1997; 17:5023-5032. 134. Iouk TL, Aitchison JD, Maguire S et al. Rrb1p, a yeast nuclear WD-repeat protein involved in the regulation of ribosome synthesis. Mol Cell Biol 2001; 21:1260-1271. 135. Schaper S, Fromont-Racine M, Linder P et al. A yeast homolog of chromatin assembly factor 1 is involved in early ribosome assembly. Curr Biol 2001; 11:1885-1890. 136. Eisinger DP, Dick FA, Denke E et al. SQT1, which encodes an essential WD domain protein of Saccharomyces cerevisiae, suppresses dominant-negative mutations of the ribosomal protein gene QSR1. Mol Cell Biol 1997; 17:5146-5155. 137. Woolford J. Personal communication. 2002. 138. Ho JH-N, Kallstrom G, Johnson AW. Nmd3p is a Crm1p-dependent adapter protein for nuclear export of the large ribosomal subunit. J Cell Biol 2000; 151:1057-1066. 139. Smith TF, Gaitatzes C, Saxena K et al. The WD repeat: a common architecture for diverse functions. Trends Biochem Sci 1999; 24:181-185. 140. Billy E, Wegierski T, Nasr F et al. Rcl1p, the yeast protein similar to the RNA 3'-phosphate cyclase, associates with U3 snoRNP and is required for 18S rRNA biogenesis. EMBO J 2000; 19:2115-2126. 141. Vanrobays E, Gleizes PE, Bousquet-Antonelli C et al. Processing of 20S pre-rRNA to 18S ribosomal RNA in yeast requires Rrp10p, an essential nonribosomal cytoplasmic protein. EMBO J 2001; 20:4204-4213. 142. Schäfer T, Strauss D, Petfalski E et al. The path from nucleolar 90S to cytoplasmic 40S preribosomes. EMBO J 2003; 22:1370-1380. 143. Milkereit P, Gadal O, Podtelejnikov A et al. Maturation and intranuclear transport of preribosomes requires Noc proteins. Cell 2001; 105:499-509. 144. Edskes HE, Ohtake Y, Wickner RB. Mak21p of Saccharomyces cerevisiae, a homolog of human CAATT-binding protein, is essential for 60S ribosomal subunit biogenesis. J Biol Chem 1998; 273:28912-28920. 145. Milkereit P, Strauss D, Bassler J et al. A Noc complex specifically involved in the formation and nuclear export of ribosomal 40S subunits. J Biol Chem 2003; 278:4072-4081. 146. Liu PC, Thiele DJ. Novel stress-responsive genes EMG1 and NOP14 encode conserved, interacting proteins required for 40S ribosome biogenesis. Mol Biol Cell 2001; 12:3644-3657. 147. Fatica A, Cronshaw AD, Dlakic M et al. Ssf1p prevents premature processing of an early pre60S ribosomal particle. Mol Cell 2002; 9:341-351. 148. Eisenhaber F, Wechselberger C, Kreil G. The Brix domain protein family - a key to the ribosomal biogenesis pathway? Trends Biochem Sci. 2001; 26:345-347.

Ribosomal Subunit Assembly

285

149. Sanvito F, Piatti S, Villa A et al. The β4 integrin interactor p27BBP/eIF6 is an essential nuclear matrix protein involved in 60S ribosomal subunit assembly. J Cell Biol 1999; 144:823-837. 150. Dragon F, Gallagher JE, Compagnone-Post PA et al. A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 2002; 417:967-970. 151. Brand RC, Klootwijk J, van Steenbergen TJ et al. Secondary methylation of yeast ribosomal precursor RNA. Eur J Biochem 1977; 75:311-318. 152. Finley D, Bartel B, Varshavsky A. The tails of ubiquitin precursors are ribosomal proteins whose fusion to ubiquitin facilitates ribosome biogenesis. Nature 1989; 338:394-401. 153. Ford CL, Randal-Whitis L, Ellis SR. Yeast proteins related to the p40/laminin receptor precursor are required for 20S ribosomal RNA processing and the maturation of 40S ribosomal subunits. Cancer Res 1999; 59:704-710. 154. Valasek L, Hasek J, Nielsen KH et al. Dual function of eIF3j/Hcr1p in processing 20S pre-rRNA and translation initiation. J Biol Chem 2001; 276:43351-43360. 155. Deshmukh M, Tsay Y-F, Paulovich AG et al. Yeast ribosomal protein L1 is required for the stability of newly synthesized 5S rRNA and the assembly of 60S ribosomal subunits. Mol Cell Biol 1993; 13:2835-2845. 156. Dechampesme A-M, Koroleva O, Leger-Silvestre I et al. Assembly of 5S ribosomal RNA is required at a specific step of the pre-rRNA processing pathway. J Cell Biol 1999; 145:1369-1380. 157. Sun C, Woolford Jr JL. The yeast NOP4 gene product is an essential nucleolar protein required for pre-rRNA processing and accumulation of 60S ribosomal subunits. EMBO J 1994; 13:3127-3135. 158. Fabian GR, Hopper AK. RRP1, a Sacharomyces cerevisiae gene affecting rRNA processing and production of mature ribosomal subunits. J Bacteriol 1987; 169:1571-1578. 159. Pestov DG, Stockelman MG, Strezoska Z et al. ERB1, the yeast homolog of mammalian Bop1, is an essential gene required for maturation of the 25S and 5.8S ribosomal RNAs. Nucleic Acids Res 2001; 29:3621-3630. 160. de la Cruz J, Kressler D, Rojo M et al. Spb4p, an essential putative RNA helicase, is required for a late step in the assembly of 60S ribosomal subunits in Saccharomyces cerevisiae. RNA 1998; 4:1268-1281. 161. Zanchin NIT, Roberts P, DeSilva A et al. Saccharomyces cerevisiae Nip7p is required for efficient 60S ribosome subunit biogenesis. Mol Cell Biol 1997; 17:5001-5015. 162. Lafontaine DLJ, Tollervey D. Nop58p is a common component of the box C+D snoRNPs that is required for snoRNA stability. RNA 1999; 5:455-467. 163. Venema J, Vos HR, Faber AW et al. Yeast Rrp9p is an evolutionarily conserved U3 snoRNP protein essential for early pre-rRNA processing cleavages and requires box C for its association. RNA 2000; 6:1660-1671. 164. Gautier T, Bergès T, Tollervey D et al. Nucleolar KKE/D repeat proteins Nop56p and Nop58p interact with Nop1p and are required for ribosome biogenesis. Mol Cell Biol 1997; 17:7088-7098. 165. Buscemi G, Saracino F, Masnada D et al. The Saccharomyces cerevisiae SDA1 gene is required for actin cytoskeleton organization and cell cycle progression. J Cell Sci 2000; 113:1199-1211. 166. Zurdo J, Parada P, van den Berg A et al. Assembly of Saccharomyces cerevisiae ribosomal stalk: Binding of P1 proteins is required for the interaction of P2 proteins. Biochemistry 2000; 39:8929-8934. 167. Andersen JS, Lyon CE, Fox AH et al. Directed proteomic analysis of the human nucleolus. Curr Biol 2002; 12:1-11. 168. Ospina JK, Matera AG. Proteomics: The nucleolus weighs in. Curr Biol 2002; 12:R29-R31. 169. Pederson T. Proteomics of the nucleolus: More proteins, more functions? Trends Biochem Sci 2002; 27:111-112. 170. Scherl A, Couté Y, Déon C et al. Functional proteomic analysis of human nucleolus. Mol Biol Cell 2002; 13(11):4100-9. 171. Fujiyama S, Yanagida M, Hayano T et al. Isolation and proteomic characterization of human Parvulin-associating preribosomal ribonucleoprotein complexes. J Biol Chem 2002; 277:23773-23780. 172. Tuschl T. Expanding small RNA interference. Nature Biotechnol 2002; 20:446-448. 173. Eichler DC, Craig N. Processing of eukaryotic ribosomal RNA. Prog Nucleic Acid Res Mol Biol 1994; 49:197-239. 174. Terns MP, Terns RM. Small nucleolar RNAs: Versatile trans-acting molecules of ancient evolutionary origin. Gene Expression 2002; 10:17-39. 175. Allmang C, Petfalski E, Podtelejnikov A et al. The yeast exosome and human PM-Scl are related complexes of 3'→5' exonucleases. Genes Dev 1999; 13:2148-2158. 176. Strezoska Z, Pestov DG, Lau LF. Bop1 is a mouse WD40 repeat nucleolar protein involved in 28S and 5.8S rRNA processing and 60S ribosome biogenesis. Mol Cell Biol 2000; 20:5516-5528. 177. Jansen RP, Hurt EC, Kern H et al. Evolutionary conservation of the human nucleolar protein fibrillarin and its functional expression in yeast. J Cell Biol 1991; 113:715-729.

286

The Nucleolus

CHAPTER 15

Nuclear Export of Ribosomal Subunits Arlen W. Johnson

Introduction

R

ibosomal subunit assembly is a complex process (see Chapters 11-14) that depends on transcription of rRNA in the nucleus and translation of r-proteins in the cytoplasm to provide the constituent molecules for the ribosome. In most cell types the pool of free ribosomal proteins available for incorporation into subunits is small, requiring ongoing translation to support ribosome biogenesis. Consequently, ribosomal subunit export has been difficult to study experimentally and has lagged behind our understanding of other RNA export pathways. However, the last several years have seen rapid progress due to the development of in vivo assays and proteomic analysis of protein complexes in the budding yeast Saccharomyces cerevisiae. This Chapter will describe our current understanding of subunit export, but many questions remain to be answered. Considering the present pace of advance in ribosome export, the next several years should fill in many of the gaps in our current understanding. The export pathway of ribosomal subunits can be broken down into different steps, beginning with the release of the nascent subunits from the nucleolus. Release presumably requires the correct assembly of the subunits, the subject of a previous chapter in this book. Consequently, rRNA processing and ribosome assembly will be discussed here only as they are thought to pertain to export. Release from the nucleolus is also likely to require the disassembly of complexes of processing factors, some of which may function in part as nucleolar retention factors to prevent the premature release of subunits during assembly. The nascent subunits must be directed through the nucleoplasm to the nuclear pore complexes (NPCs) for export to the cytoplasm. Export out of the nucleus requires the assembly of an export complex, composed minimally of the subunit, a nuclear export signal (NES) provided by a ribosomal protein or in trans by an adapter protein that binds to the subunit, an export receptor, and the GTP-bound form of the small GTPase Ran (Ran•GTP). Although translocation through the NPC does not require energy, GTP hydrolysis by Ran ensures that transport is unidirectional. The subunits that are released from the nucleolus are not fully mature, and additional rRNA processing and protein assembly events that occur in the nucleoplasm and cytoplasm may be coupled with export. Once in the cytoplasm, trans-acting export factors are released from the subunits and recycled to the nucleus. As discussed below, many of these events are likely to be regulated by nucleotide hydrolysis. Assembly and export of the large (60S) subunit is largely independent of the small (40S) subunit. Because we currently know much more about export of the 60S subunit compared to the 40S subunit, this chapter will be biased toward 60S subunit export. Nevertheless, many of the general ideas are likely to pertain to the small subunit as well.

The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Nuclear Export of Ribosomal Subunits

287

Figure 1. The nuclear export pathway of the large subunit. Assembly of ribosomal subunits in the nucleolus requires the import of newly synthesized ribosomal proteins from the cytoplasm and their assembly onto the nascent rRNA transcripts. Release from the nucleolus is associated with the release of many of the trans-acting nucleolar factors required for subunit biogenesis and the addition of shuttling factors including Tif6p and possibly Nmd3p. Further changes in the subunit composition occur in the nucleoplasm. Export out of the nucleus requires the assembly of general export factors, including CRM1 and Ran•GTP. General export factors are released immediately upon entry into the cytoplasm whereas 60S-specific shuttling factors may remain associated until later steps. Two cytoplasmic GTPases, Efl1p and Lsg1p in yeast, interact with the subunit to mediate recycling of export factors.

Export of the Large Subunit Nuclear Export Assays Initial experiments examining ribosome export were carried out by microinjection of radiolabeled ribosomes into nuclei of Xenopus oocytes. These experiments showed that ribosomal subunit export is energy-dependent, unidirectional and saturable, indicating that it is receptor-mediated.1-5 Ribosome export has also been monitored using resealed rat liver nuclei.6 More recently, in vivo assays for export of the large subunit have been developed in yeast using fusions of Green Fluorescent Protein (GFP) to the ribosomal proteins Rpl25p7 or Rpl11p.8

Release from the Nucleolus Nuclear export of ribosomal subunits begins with release of the nascent subunit from the nucleolus. Ribosomal subunit assembly within the nucleolus requires a large number of trans-acting processing factors (reviewed in Chapter 14). Mass-spectrometric analysis of complexes associated with pre60S particles from yeast has identified over 100 associated proteins.9-13 Attempts have been made to order the association and release of these proteins with nucleolar pre60S particles.10 However, this analysis is complicated by the possible exchange or loss of loosely associated proteins during preparation and the likelihood that complex composition will be biased by the choice of protein used for affinity purification. Subsequent to release from the nucleolus and during export, the subunit is much less complex14 (see Fig. 1). Thus, release

288

The Nucleolus

from the nucleolus is accompanied by significant remodeling of the complex involving the removal of many nucleolar factors and the addition of proteins required for export. Release is also likely to be coordinated with the correct assembly of the particle and completion of specific rRNA processing steps that are monitored by quality control checks. This is consistent with the finding that various mutants defective in late steps of 60S biogenesis do not appear to release stable subunits from the nucleolus (see below). Ultrastructural analysis also suggests that release follows the completion of certain processing steps, suggesting a requirement for or at least coordination between correct processing and entering the export pathway. Using in situ hybridization to examine the nuclear distribution of 27S prerRNA, Gleizes et al found that internal transcribed spacer 2 (ITS2), an indicator of 27S prerRNA, was 90% nucleolar and 10% nucleoplasmic.15 In contrast, 25% of the hybridization of a 25S rRNA probe was nucleoplasmic, suggesting that the final processing of 27S to 25S occurs predominantly before release from the nucleolus. Nevertheless, mutations in some proteins required for subunit assembly, including Gar1p and Nop1p, cause release of improperly processed subunits into the nucleoplasm,13,15 although these subunits are not competent for export and remain in the nucleus where they must be degraded. The nucleolar release of defective subunits observed in some biogenesis mutants may not be indicative of a lack of fidelity in the biogenesis pathway, but rather it may be the result of the enormous flux of defective subunits overwhelming the degradation pathway.

Nucleotidases in pre60S Transport As mentioned above, over 100 trans-acting factors as well as a comparable number of snoRNAs are required for the processing and assembly of the large subunit. However, the majority of these factors appear to be released before or as the subunit leaves the nucleolus. In addition, processing of 27S and 7S prerRNAs and protein exchanges may be associated with release from the nucleolus. Thus, one might expect that significant conformational changes of the subunit and large-scale disassembly of processing complexes are required for release of the subunit. These events likely require extensive energy-driven remodeling of the complex. Not surprisingly, several nucleotide-dependent proteins are also required for multiple steps in export of the large subunit, beginning with its release from the nucleolus (see Table 1). Two AAA-type ATPases, Rix7p and Mdn1p/Rea1p, have been implicated in 60S export.9,16 Proteins of this class of ATPases form hexameric rings whose conformation changes upon nucleotide binding (see 17 for review). These molecules can be thought of as “molecular crowbars”, capable of disassembling stable protein complexes. rix7 temperature-sensitive mutants accumulate Rpl25-eGFP in the nucleolus at nonpermissive temperature, suggesting a role for Rix7p in subunit release.16 Interestingly, Rix7p-GFP was observed to redistribute from the nucleolus in stationary phase cells, which are inactive in assembling ribosomes, transiently to the nuclear periphery upon return to growth, followed by a steady state distribution throughout the nucleus in actively growing cells. These results suggest that Rix7p functions on the pre60S particle as it passes from the nucleolus through the nucleoplasm, although Rix7p does not appear to form a stable complex with pre60S subunits.16 The accumulation of Rpl25-eGFP in rix7 mutants may reflect a defect in removing and recycling nucleolar factors that leave the nucleolus bound to the pre60S particle. A second AAA-ATPase, Mdn1p/Rea1p, is associated with the pre60S particle in the nucleoplasm.9,14 This protein is remarkable for its size of 559 kDa, making it the largest protein in yeast. Mdn1p/Rea1p is an essential protein but its possible function in ribosome biogenesis has not been described. In addition to ATPases at least three GTPases (Nog1p, Nog2p/Nug2p and Nug1p) are found associated with pre60S particles in the nucleolus and/or nucleoplasm. Each of these proteins is essential and required for 60S biogenesis. Mutations in NOG1 result in nucleolar accumulation of Rpl25-eGFP (Kallstrom, Hedges and Johnson, in press) whereas Rpl25-eGFP accumulates throughout the nucleus in nug1 and nug2 mutants.9,13 Sequence comparison of these GTPases indicates that Nug1p and Nog2p/Nug2p are related to each other13 and are

Nuclear Export of Ribosomal Subunits

289

Table 1. Yeast proteins involved in 60S subunit export Protein

Orf

Function

Protein Associated Localization rRNAs

Localization of Rpl25-eGFP in Mutant

Nog1p Nug2p Nug1p Rix1p Noc1p/Mak21p Noc2p/Rix3p Noc3p Rlp7p/Rix9p Ria1p/Efl1p

YPL093w YNR053c YER006w YHR197w YDR060w YOR206w YLR002c

no,nu no,nu no,nu nu no no,nu nu no nuc or cyt1

no nu nu nu no no no,nu no nu

Tif6p

YPR016C

Lsg1p/Kre35p3 Nmd3p Rix7p Mdn1p/Rea1p

YGL099w YHR170w YLL034c YLR106c

GTPase unknown unknown unknown unknown unknown unknown unknown GTPase, recycling Tif6p 60S binding, anti association GTPase export adapter AAA-ATPase AAA-ATPase

YNL163c

27S, 7S 35S, 27S, 7S 35S, 27S, 7S 27S, 7S

nu, cyt2 cyt cyt,nu 2 no,nu nu

nu 25S, 5.8S 25S, 5.8S

no no,nu4 no

1. This protein was reported to be nuclear91 or cytoplasmic.88 2. Shuttling protein. 3. Kallstrom and Johnson, submitted. 4. The localization of Rpl25-eGFP is nucleolar or nuclear depending on the allele of NMD3.

members of an ancient subclass of the Ras GTPase superfamily in which the G-protein motifs are circularly permuted.18 The related essential bacterial protein YjeQ has recently been shown to have GTPase activity in vitro. Although its cellular function is unknown, YjeQ may be involved in translation.19 We do not yet know the function of these NTPases, but they may mediate protein exchanges on the pre60S particle to provide checks in the assembly pathway. Genetic screens in yeast to identify interacting proteins and the identification of the binding sites of these proteins on the pre60 complex should provide insight to their function. Although the relatedness of Nug1p and Nog2p/Nug2p suggests the possibility that these proteins exchange with one another on the same site on the nascent subunit, affinity purification experiments indicate that both proteins are present simultaneously on the subunit.9,13

Protein Exchanges in the Large Subunit Among the large number of trans-acting nucleolar factors required for subunit assembly, the nucleolar 66S complex contains several transiently-associating proteins that are related to integral subunit proteins of the mature cytoplasmic subunit. These include Cic1p/Nsa3p (Yhr052wp),12 related to Rpl1p; Rlp7p, related to Rpl7p;20,21 and Rlp24p, related to Rpl24p.13,15 Each of these 66S-associated proteins is essential, whereas Rpl1p and Rpl7p but not Rpl24p are essential.22 Because of the similarities of the primary sequences of these pairs of proteins, it has been suggested that they are exchanged for each other during subunit maturation.13,21 Both Rlp7p and Rlp24p are present in relatively late pre60S particles with Rlp7p being released prior to Rlp24p.13 Interestingly, an Rlp7 mutant (rix9) was identified in a screen for 60S export mutants.21 Rlp7p is found in the granular component of the nucleolus, along with a fraction of Nog2p (see above);21 although Nog2p-bound pre60S particles appear to contain

290

The Nucleolus

Rpl7p and not Rlp7p.13 Rlp7 mutants accumulate Rpl25-eGFP in the nucleolus leading to the suggestion that the exchange of Rlp7p for Rpl7p may be one of the events that signals release of the nascent subunit from the nucleolus.21 Rlp24p appears to be released later than Rlp7p, as it is found on Nog2p-associated particles, and Rpl24p (also referred to as L30 and encoded by two genes in yeast) was identified as one of the last proteins to assemble into the subunit.23 However, Rlp24p may be released before loading the export adapter Nmd3p (see below) as mass-spec analysis of Nmd3p-associated particles identified Rpl24p but not Rlp24p (Kallstrom and Johnson, unpublished). The requirement of Rlp24p for intranuclear subunit transport or nuclear export has not been reported. In comparison to Rpl24p, Rlp24p contains an extended C-terminus, predicted to form a coiled-coil, a protein interaction motif. This raises the possibility that Rlp24p, through its C-terminus, helps to anchor the subunit in the nucleolus. Release of Rlp24p, would then be required for subunit release. Since Rpl24p is not essential, it may simply serve as a patch to cover the Rlp24p binding site.

The NOC Proteins Noc1p, Noc2p and Noc3p are essential proteins that were originally found in a large nucleolar complex containing RNA polymerase I.24 A noc2 mutant (rix3-1) was also found in a screen for ribosome export mutants.25 These Noc proteins are highly conserved throughout eukaryotes. Noc1p and Noc3p contain a common 45 aa motif, a noc domain, shared with another yeast protein Noc4p, a protein found in a pre40S particle.26 Noc2p forms complexes with Noc1p in the nucleolus and with Noc3p in the nucleoplasm. Mutations in NOC1 and NOC2 or overexpression of an oligomerized noc domain lead to accumulation of Rpl25-eGFP in the nucleolus,25 suggesting that these proteins are required for release of the subunit from the nucleolus. However, the apparent sedimentation of Noc1p with 35S rRNA and its presence in a RNA polymerase I-containing complex suggests an earlier function in subunit biogenesis. On the other hand, noc3 mutants accumulate Rpl25-eGFP in the nucleoplasm as well as the nucleolus, consistent with the nucleoplasmic localization of Noc3p. Whether the Noc3p/ Noc2p dimer binds to the 66S particle or to a nucleoplasmic pre60S particle remains to be determined. Considering that Noc2p interacts with both Noc1p and Noc3p in separate complexes, the exchange of Noc3p for Noc1p on the Noc2p-bound pre60S may be necessary for subunit release from the nucleolus.25

Export to the Cytoplasm Nuclear Export Mechanisms Nuclear envelopes of eukaryotic cells spatially separate nuclear rRNA processing and ribosome assembly from translation in the cytoplasm. The regulation of transport of macromolecules and RNPs such as ribosomes through NPCs provides cells with a mechanism to ensure that only mature, fully-processed cargo is exported. Although 40S and 60S subunits are assembled on rRNAs derived from a common primary transcript, they are exported independently from the nucleus as separate subunits. This conclusion is supported by numerous genetic and biochemical findings. For example, 40S assembly precedes and is distinct from 60S assembly.26,27 40S subunits are assembled and exported faster than 60S subunits,28,29 consistent with the greater abundance of 60S over 40S subunits in the nucleoplasm.15 In both yeast and mammalian cells, mutants that affect the biogenesis of one of the subunits often have little effect on the production of the other subunit.30,31 During nuclear export, cargo molecules are recruited to the NPCs by specialized export receptor proteins. Export receptors of the importin-β family bind to nuclear export signals (NESs) on the cargo molecules in the presence of Ran•GTP, the nuclear GTP-bound form of the small GTPase Ran. After translocation through the NPC to the cytoplasm, the intrinsic GTPase activity of Ran is activated by a cytoplasmic GTPase-activating protein (RanGAP) and Ran•GTP binding proteins (RanBP1 and RanBP2). The hydrolysis of GTP on Ran results in

Nuclear Export of Ribosomal Subunits

291

release of cargo from the export receptor. The unidirectional nature of transport by importin-β receptors is controlled by the steep gradient of Ran-GTP across the nuclear envelope (for reviews see refs. 32-37). Nuclear export signals are generally short stretches of amino acids that interact with export receptors thereby promoting transport of cargo molecules. These signals may be encoded in the primary sequence of a cargo protein, as is the leucine-rich NES of the cAMP-dependent protein kinase inhibitor (PKI).38 However, an NES can also be provided by an adapter protein that bridges the interaction between cargo and an export receptor.32,37 For example, HIV-1 Rev is an adapter for viral RNA export,3 that binds to incompletely spliced viral transcripts and recruits the export receptor CRM1. Because CRM1, a member of the importin β-like family of transport receptors, does not bind RNA directly, it recognizes a leucine-rich NES on Rev.39,40

NMD3 Is the Adapter for the Large Subunit Nmd3p was initially identified in a 2-hybrid interaction screen with the nonsense mediated decay factor Upf1p.41 However, Nmd3p does not appear to play a direct role in mRNA degradation.42 Nmd3p was also identified in a screen for extragenic suppressors of a conditional rpl10 mutant.43 Nmd3p is a predominantly cytoplasmic protein that binds to free 60S subunits and is required for the production of stable 60S subunits. Conditional nmd3 mutants process rRNA with wild-type kinetics but the resulting 25S rRNA is extremely unstable (t1/2=4 min).42 NMD3-like proteins are found throughout Archaea and Eukarya, but the eukaryotic proteins possess a C-terminal extension. Deletion analysis of Nmd3p showed that this domain contains a basic nuclear localization sequence and a leucine-rich NES that are responsible for the nucleo-cytoplasmic shuttling of NMD3 protein. Deletion of the NES of Nmd3p results in a dominant negative phenotype,44 due to accumulation of mutant Nmd3p in the nucleus and retention of 60S subunits.45,46 Importantly, when the mutant protein was supplied with the NES of PKI, both the export of Nmd3p and the cytoplasmic accumulation of 60S subunits were restored. Thus, export of Nmd3p is required for export of 60S subunits. Similar results have been obtained using comparable mutants of human NMD3 expressed in HeLa cells or in Xenopus oocytes.47 Therefore, the same basic mechanism for export of 60S subunits is conserved between humans, frogs and yeast.

Loading of the Adapter NMD3 Several reports have characterized the protein composition of affinity-purified nuclear pre60S particles.9,12-14,48 Pre60S-containing particles affinity purified with Nug1p, Nog2p/Nug2p, Rix1p and Sda1p9,13,14 appear to be nucleoplasmic but lack Nmd3p, suggesting that Nmd3p loads late in the nuceloplasm. However, the localization of these complexes is inferred from the localization of the tagged proteins used for affinity purification. Subunits associated with a protein that is found in both the nucleolus and nucleoplasm could be derived from either compartment. Some of the proteins found in the Nug1p-bound pre60S complex were also found, in addition to ribosomal proteins, in preparations of NPCs from yeast.49 The lack of Nmd3p in these preparations may suggest that these proteins represent nucleolar contamination rather than nascent subunits as they are translocated through the NPC. On the other hand, several lines of evidence suggest that Nmd3p loads onto the nascent subunit in the nucleolus, or perhaps as it emerges from the nucleolus. First, Rpl25-eGFP accumulates in the nucleolus in conditional nmd3 mutants whereas nmd3 mutants that specifically disrupt interaction with Crm1p lead to nucleoplasmic accumulation of the pre60S particle (Hedges and Johnson, unpublished). Pre60S particles coimmunoprecipitated with Nmd3∆100 (deleted of the C-terminal 100 aa, including the NES) contain mature 25S rRNA but also a number of nucleolar proteins (Kallstrom and Johnson, unpublished). In human cells, hNMD3 is found throughout the nucleus and accumulates in the nucleolus upon treatment with LMB.47 Interestingly, treatment with actinomycin D, which blocks synthesis of rRNA, also leads to

292

The Nucleolus

Figure 2. Assembly of the 60S export complex. The nuclear export signal (NES) for the large subunit is provided by the adapter protein Nmd3p. Nmd3p may be recruited to the subunit through its interaction with Rpl10p. Because Rpl10p is closely associated with the A site of the large subunit and is one of the last proteins to be loaded onto the subunit, an Nmd3p binding site composed in part of Rpl10p could provide a means for structural proofreading of the subunit. The interaction of CRM1 with the NES of Nmd3p is stabilized by Ran•GTP and possibly additional accessory factors. In the cytoplasm, disassembly of the export complex is triggered by activation of the GTPase activity of Ran.

accumulation of NMD3 in nucleoli, suggesting that exit of hNMD3 from nucleoli depends on the flux of ribosomal particles from the nucleolus.47 Deletion analysis of Nmd3p suggests that its interaction with 60S subunits is multivalent (Hedges and Johnson, unpublished). However, the r-protein Rpl10p appears to be one component of the binding site, as conditional rpl10 mutants are suppressed by overexpression of or dominant mutants of NMD3.43,50 Furthermore, Nmd3p copurified with a truncated Rpl10p when coexpressed in E. coli.46 Rpl10p is an exchangeable protein that binds to 60S subunits late in their assembly.23,51 Rpl10p (L10e in archaeal ribosomes) is positioned close to the peptidyl-transferase site on the large subunit52,53 and is required for subunit joining54 (see Fig. 2). Thus, binding of Nmd3p may monitor the structure of the 60S subunit surrounding the active site, ensuring that only correctly assembled subunits have access to the export pathway. Interestingly, rpl10 conditional mutants accumulate nascent subunits in the nucleoplasm.46 This is in contrast to nmd3 conditional mutants that accumulate subunits in the nucleolus. Analyses of the protein composition of pre60S particles12-14 are consistent with Rpl10p loading prior to Nmd3p binding. One explanation of this paradox is that Rpl10p is not absolutely required to recruit Nmd3p but is required for subsequent recruitment of Crm1p and Ran-GTP (see Fig. 2). Thus, in the absence of Rpl10p, Nmd3p could bind to and stabilize the nascent subunit but without Rpl10p, the NES of Nmd3p is not accessible for assembling an export complex.

Nuclear Export of Ribosomal Subunits

293

Quality Control and Subunit Export Due to their complex structure and central function in translating the genetic code, ribosomes are likely to undergo a quality control check before they are released into the pool of translating subunits in the cytoplasm. Nuclear translation has been proposed as a mechanism for mRNA proofreading (reviewed in ref. 55). If it occurred, such translation in the nucleus could also be used as a functional check for newly made ribosomes. However, definitive experiments demonstrating nuclear translation have not been presented (for a discussion see refs. 56-58) and several facts indicate that nuclear translation is unlikely to occur in yeast. For example: the association of Tif6p with the nuclear pre60S subunit would prevent its joining with the small subunit; and the final processing of 20S to 18S rRNA, a step necessary for a functional 40S subunit, occurs in the cytoplasm in yeast.29 Furthermore, nuclear translation, if it occurs, is not required for a functional check of ribosomes in vertebrate cells because nuclear export of the large subunit in Xenopus oocytes occurs in the presence of various inhibitors of protein synthesis.59 A more plausible mechanism for quality control of nascent subunits is “structural proofreading”, in which binding of essential export factors (such as NMD3) depends on the assembly of a binding site whose structure is intimately related to the function of the subunit (reviewed in ref. 59). Such a mechanism of quality control would coordinate export with assembly of structurally correct subunits.

CRM1 Is the Export Receptor for NMD3 and the 60s Subunit The NES of Nmd3p was mapped by deletion analysis to the extreme C-terminus of Nmd3p.45 This region contains a highly conserved sequence (INIDELLDELD) which is similar to leucine-rich NESs recognized by Crm1. A 21 aa peptide containing this sequence was sufficient for export of a reporter molecule (Hedges and Johnson, unpublished). The similarity to leucine-rich NESs suggested that Crm1 was the receptor for Nmd3p and the 60S subunit. In most eukaryotic cells, the interaction of CRM1 with its ligand is inhibited by the drug leptomycin B (LMB), which reacts with a highly conserved cysteine residue in CRM1.60 Although Crm1p of S.cerevisiae is resistant to LMB, alteration of threonine 539 to cysteine renders it LMB-sensitive.61 Using this mutant, treatment with LMB led to nuclear accumulation of Nmd3p as well as the 60S reporter Rpl25-GFP.45,46 In contrast, the conditional mutant xpo1-1, did not efficiently block export of 60S subunits7,8 or Nmd3p45 at restrictive temperature. This may be due to the incomplete inactivation of xpo1-1 function at restrictive temperature or to a stress response to elevated temperature that leads to transient inhibition of ribosome biogenesis in yeast.62 Recent work has shown that CRM1 is the export receptor for 60S subunits in vertebrate cells as well. Treatment of Xenopus oocytes with LMB or microinjection of an NES peptide conjugate inhibits subunit export, monitored by the disappearance of labeled rRNA in the cytoplasm and its retention in the nucleus.47

Nucleoplasmic- and Export-Associated Assembly and Processing The residence time of the pre60S subunit in the nucleoplasm and the time required for its incorporation into translating polysomes are greater than that for the 40S subunit. This has led to the speculation that the 60S subunit undergoes additional modifications and/or structural rearrangements that are rate limiting. Such events could in principal occur in the nucleus and may affect export. As discussed below, Rpl10p is required for nuclear export of the large subunit and is thought to recruit the export adapter Nmd3p.46 The nucleoplasmic protein Rsa1p genetically interacts with Rpl10p and deletion of RSA1 leads to decreased levels of Rpl10p on free 60S subunits.63 Thus, Rsa1p may facilitate the loading of Rpl10p in the nucleus. Rpl10p is an exchangeable protein64 making it technically difficult to observe its accumulation in the nucleus on nascent subunits. Nevertheless, Rpl10p does accumulate in rrp44-1 mutants,46 containing a defect in the exosome. Since the exosome degrades aberrant rRNA species65 as well as participates in

294

The Nucleolus

trimming the 3'-end of 5.8S rRNA (see Chapter 12 by Raué, or Chapter 14 by de la Cruz), it is possible that a defect in degradation or processing allows accumulation of a particle that contains a pre-rRNA species as well as Rpl10p. One additional processing step that may be linked to export is the final 3'-trimming of 5.8S rRNA. In Xenopus oocytes, inhibition of 60S export leads to nuclear accumulation of 6S prerRNA.47 This final trimming to 5.8S is a cytoplasmic event, as the 6S pre-rRNA of subunits recovered from the nuclei of these oocytes is processed upon microinjection into the cytoplasm (Lund and Dahlberg, personal communication). In yeast, it is not known whether the final processing of 5.8S occurs in the nucleus or cytoplasm,66 however, subunits containing 3'-extended forms of 5.8S are incorporated into polysomes in rrp6 mutants,67 indicating that if this step is linked to export, it is not obligatory for export.

Assembly of the Export Complex The binding of Crm1p to leucine-rich NES-containing proteins is stabilized by the binding of Ran-GTP, thereby promoting the assembly of export complexes in the nucleus where there is a high concentration of Ran-GTP. In some cases the stable interaction of Crm1p with cargo requires an additional cofactor such as RanBP3 (Yrb2p in yeast).68,69 Accordingly, mutants in Ran (Gsp1p in yeast) and in proteins required for Ran cycling, such as Rna1p (Ran-GAP) or Prp20p (Ran-GEF), inhibit export of the large subunit.7,8 Attempts to reconstitute an in vitro export complex containing Crm1p, Nmd3p and Ran•GTP, with or without cytoplasmic 60S subunits, have not yet been successful (Kallstrom and Johnson, unpublished). The most likely explanation for this is that an additional factor(s) is required to stabilize the complex. Although the most likely candidate would be RanBP3, based on its known function in stabilizing Crm1-ligand interactions, yrb2 mutants do not show a significant defect in 60S export in yeast.70 It is also possible that CRM1 access to the NES of Nmd3p is regulated by an additional nuclear protein that monitors binding of Nmd3p to the 60S subunit. Such an accessory factor would not be present in the reconstituted system.

Crossing the Nuclear Envelope Translocation of ribosomal subunits through the NPC is the least understood step of the export pathway. Indeed, the mechanism of translocation of any molecule is not well understood, although several models have been presented (for recent reviews see 71-74). Small molecules can pass through the NPC by simple diffusion, whereas larger molecules (>20-40kDa) are transported by facilitated diffusion, requiring interaction with transport receptors. All cargo must pass through the central channel of the NPC which has a measured diameter of approximately 44 nm in Xenopus oocyte nuclei.75,76 The upper limit for diameter of cargo that could be transported was initially determined to be 26 nm.77 Remarkably, recent experiments suggest that molecules with diameters of up to 39 nm can pass through NPCs in Xenopus.78 This would be sufficient to accommodate a ribosomal subunit with a diameter, of 25-30 nm79 without significant conformational changes, consistent with the notion that the nuclear particles do not undergo significant structural changes associated with export. Translocation itself can be divided into three steps, docking at the NPC, translocation, and release of cargo into the cytoplasm. The docking of cargo at the NPC is mediated by interactions between the receptor molecules and phenylalanine-glycine (FG) repeats of the nucleoporins, the proteins of the NPC. Mutations in various nucleoporins cause defects in 60S export7,8,15 indicating that a subset of nucleoporins is specific for 60S export. Some of the proteins within the NPC form subcomplexes with distinct functions in separate transport pathways. For example, Nup82p-Nup159p-Nsp1p form a subcomplex that is necessary for 60S export.15 On the other hand, particular mutants of Nup159p are defective for mRNA export but not for pre60S export, indicating separation of these pathways. Nuclear export of mRNA requires the export factor TAP (Mex67p in yeast) and its cofactor NXT (Mtr2p in yeast) (see 80 for review).

Nuclear Export of Ribosomal Subunits

295

Formation of the heterodimeric TAP/NXT complex induces binding to nucleoporins to facilitate export. Because of its role in mRNA export it would seem unlikely that NXT/Mtr2p plays a direct role in ribosomal subunit export. However, in yeast mtr2-33, and mex67-5 mutants show allele-specific synthetic lethality with nmd3 mutants.9,45 Furthermore, Mtr2p was found associated with late nucleoplasmic pre60S complexes that also contained Nmd3p.14 Thus, Mtr2p may facilitate interaction of the 60S export complex with the NPC. Alternatively, these results could be due to indirect interactions between the ribosome and mRNA export pathways. Although these studies are suggestive of specific physical interactions important for ribosomal subunit export, additional experiments are needed to understand how the ribosomal subunit export complex docks at the NPC. Once docked, cargo translocates through the central pore in an energy-independent facilitated diffusion process. The pore appears to be a loose dynamic sieve of hydrophobic FG repeats that acts as a permeability barrier. Nevertheless, the flux of molecules that can pass through an NPC is remarkable, measured at approximately 100 MDa/s and >1000 translocation events/s per NPC.81 In yeast, which synthesize approximately 2000 ribosomes per minute62 and has approximately 150 NPCs/nucleus,82 one calculates an export rate of one subunit per 2.25 sec. Three models have been proposed to explain translocation through the pore (reviewed in ref. 74). In the affinity gradient model,83 translocation is driven by receptors binding to successively higher affinity binding sites on nucleoporins as they pass through the NPC. In the Brownian affinity gate model49 the interaction of a receptor with nucleoporins is thought to increase the time cargo spends at the entrance to the channel, driving its diffusion through the channel. Finally, the selective phase model84 posits that transport receptors, through weak hydrophobic interactions with the FG repeats of the nucleoporins, partition into the hydrophobic phase of the channel, ferrying cargo with them. The hydrophobic permeability barrier of the pore is reestablished as the cargo/receptor complex passes though. Regardless of the model, considering the hydrophobic nature of the nucleoporins comprising the channel, it is difficult to imagine how RNPs as large and hydrophilic as the ribosome can be accommodated in this environment. Using an assay monitoring ribosome export from resealed rat liver nuclei, Schlatter et al found that HSP90 was required for efficient export85 and suggested that it facilitates remodeling of the subunit during translocation, perhaps helping it to partition into the hydrophobic environment. Although HSP90 function was also found to be required for efficient export of 60S subunits microinjected into Xenopus oocyte nuclei,85 a role for HSP90 (Hsp82p in yeast) has not been reported.

Packaging the Complex for Passage through NPC The ribosomal subunits may be the bulkiest cargo that passes through the NPC. At a diameter of 25-30 nm, they approach the diameter of the channel. Consequently, it is important that subunits do not interact with each other or with other large complexes that may cause steric hindrance during translocation through the pore. Indeed, the separate export of the large and small subunits suggests the presence of factors that prevent their premature interaction. One such factor is Tif6p. Initially identified by its ability to prevent subunit interaction, Tif6p was thought to regulate subunit joining during translation initiation.86 More recently, Tif6p was shown to be required for subunit biogenesis in the nucleolus and for transport of the large subunit.87,88 Considering its anti-association function, it may be required to maintain a free subunit during transport. Other large complexes that are assembled and exported from the nucleus may pose a similar problem if their interaction with ribosomal subunits is not blocked. For example, signal recognition particle (SRP), is assembled in the nucleus.89,90 Because SRP can bind ribosomes weakly in the absence of nascent peptide, it may be necessary to prevent such interaction in the nucleus for efficient ribosomal subunit export.

296

The Nucleolus

Recycling Export Factors At least three shuttling proteins that bind to the nascent 60S subunit are exported along with the large subunit. These are the adapter Nmd3p,45,46 Tif6p (see above)88 and Arx1p, a nonessential protein related to methionyl-aminopeptidases 14(Hung and Johnson, unpublished). These proteins must be released from the subunit in the cytoplasm for their recycling to the nucleus. However, unlike the export receptor CRM1, which is released by RanGAP-dependent hydrolysis of Ran•GTP immediately upon entry into the cytoplasm,32-36 these proteins remain associated with the free subunit after release of Crm1. Because these three proteins are not present in polysomes, they are released prior to or upon subunit joining. Tif6p is distributed between the nucleus and cytoplasm,87,88 but accumulates in the cytoplasm in cells mutant for the GTPase Efl1p/Ria1p.88 Dominant mutants of Tif6p suppress the growth defect of efl1 mutants88,91 and restore the nuclear pool of Tif6p.88 Thus, Efl1p appears to mediate release of Tif6p from the subunit in the cytoplasm. A failure in this step leads to depletion of the nuclear pool of Tif6p and a block in 60S subunit biogenesis. In fact, Rpl25-eGFP accumulates in the nucleoplasm in efl1 mutants, indicating a block in a late step of subunit export.88 Efl1p is related to eukaryotic translation elongation factors, suggesting that Efl1p binding to the GTPase center on the large subunit triggers the release of Tif6p. Because Efl1p does not bind Tif6p directly, Efl1p binding to the large subunit may induce a conformational change that reduces the affinity of Tif6p for its binding site on the subunit.88 In the cytoplasm Nmd3p remains associated with the free exported subunits before their incorporation into 80S ribosomes. Nmd3p also binds to free 60S subunits recycling during rounds of translation.92 The steady state distribution of Nmd3p is largely cytoplasmic, consistent with its persistence on cytoplasmic free 60S subunits, which exceed the number of pre60S subunits in the nucleus. The mechanism of release of Nmd3p is not known but it could be dependent on the cytoplasmic GTPase Lsg1p/Kre35p (Kallstrom, Hedges and Johnson, in press).14 Although Lsg1p is related to the nuclear GTPases Nug1p and Nug2p/Nog2p,13 Lsg1p appears to be restricted to the cytoplasm. Interestingly, lsg1 mutants accumulate the 60S reporter Rpl25-eGFP in the nucleolus, indicating a failure to recycle an export factor. Because Tif6p, Nug1p, Nog1p and Nog2p/Nug2p all appear to remain nuclear in lsg1 mutants, it is possible that a defect in Nmd3p recycling accounts for the export defect of lsg1 mutants (Kallstrom, Hedges and Johnson, in press). Could the recycling of export factors be coupled to translation initiation? Proteins on the nascent 60S subunit in the cytoplasm, including Nmd3p, Tif6p and Lsg1p, are largely absent from translating 80S ribosomes. Moreover, Nmd3p also binds to mature, cytoplasmic 60S subunits as they cycle between the termination and initiation steps of protein synthesis.42 Since these proteins are not present in 80S ribosomes, it is possible that some (or all) of them are released concurrently with subunit joining.

Export of 40S Subunits As discussed above, export of the large subunit requires Crm1 and RanGTP. Although the small subunit is exported independently of the large subunit, export of the small subunit also depends on the export receptor Crm1 and RanGTP. This has been shown in yeast utilizing an in situ RNA localization assay to monitor ITS1. The final processing of 20S to 18S rRNA occurs in the cytoplasm in yeast and the resulting fragment, the 5'-portion of ITS1, is degraded by Xrn1p.93 Thus, 5'-ITS1 accumulates in the cytoplasm in xrn1 mutants, whereas it accumulates in the nucleoplasm in mutants defective in 40S export.70,94 Using this assay, Moy and Silver showed that a conditional xpo1-1 mutant or treatment of LMB-sensitive yeast with LMB resulted in the nuclear accumulation of 5'-ITS1.70,94 In addition, overexpression of a dominant Ran mutant or mutations in the Ran GTPase activating protein, Prp20p/Mtr1p, or in the RNA-GTP exchange factor, Rna1p, caused accumulation of 5'-ITS1.70,94 Similar results in yeast were obtained for prp20/mtr1 and rna1 mutants in an assay monitoring the nuclear

Nuclear Export of Ribosomal Subunits

297

accumulation of pre18S.15 The export pathway for the 40S subunit appears to be conserved throughout eukaryotes. In Xenopus oocytes, microinjection of either a PKI NES conjugate or dominant negative Ran, or treatment with LMB leads to nuclear accumulation of 18S rRNA.47 In contrast to the large subunit, efficient export of the small subunit also requires Yrb2p (yeast RanBP3).70 Yrb2p presumably enhances the interaction of Crm1 with the 40S export complex, even though Yrb2p is not essential in yeast. Although a screen for yeast mutants defective in 5'-ITS1 export identified a number of nucleoporins in addition to Ran effectors (for a comprehensive discussion of these mutants see 70, 94), factors specific to 40S export were not identified. Thus, we do not yet know what protein provides the NES for the 40S subunit. However, the inability to reconstitute Crm1 binding to the small subunit in vitro70 could indicate that the NES is provided in trans by an adapter protein and not by an integral component of the 40S subunit. More recently, a 40S export assay has been described that employs a GFP fusion to the yeast small subunit protein Rps2p.95 However, the accumulation of the reporter molecule in 40S biogenesis mutants95 may limit the utility of this assay in identifying new mutants. Proteomic analysis of protein complexes provides an alternative to genetic screens. A pre40S particle has been affinity purified from yeast and analyzed by mass-spectrometry.26 This complex contains a large number of nonribosomal proteins that partially overlap with those described for the small subunit processome,27 and may represent the 90S precursor particle.96 This complex contains the GTPase Bms1p, an essential protein required for 40S biosynthesis97,98 and Noc4, related to Noc1p and Noc3p, required for intranuclear transport of the pre60S subunit (see above). Mutations in two other proteins in this complex, Kre33p and Krr1p, cause transient nucleolar accumulation of the small subunit reporter Rps2p-eGFP, suggesting that Kre33p and Krr1p may be required for movement of the pre40S complex out of the nucleolus.26 However, it is also possible that transient accumulation is due to a defect in 40S assembly because ribosome processing mutants have been shown to affect the 40S export assay.94,95 Furthermore, these two proteins appear to be released from the pre40S complex in the nucleus as they do not coprecipitate 20S rRNA. Among the pre40S associated proteins characterized, only the predominantly nuclear protein Enp1p was found associated with pre40S complexes containing dimethylated 20S prerRNA.26 Since this methylation is a cytoplasmic event, Enp1p appears to be a shuttling protein that exits the nucleus bound to the pre40S particle. Thus Enp1p may be a bona fide export factor. The characterization of additional proteins associated with Enp1p-bound pre40S particles will likely yield the export adapter for the 40S subunit. Assuming that Enp1p is a shuttling protein and that export of the 40S subunit is mediated by an adapter protein, there must be mechanisms for recycling these factors to the nucleus as there are for release of export factors from the 60S subunit. It is possible that the association of initiation factors during translation initiation provides a final quality control check of the subunit. Along these lines, Hcr1p from yeast was found as a high copy suppressor of a mutation in RPG1,99 encoding the largest subunit of translation initiation factor eIF3, which is required for multiple activities including dissociation of 80S and promoting the formation of the 43S preinitiation complex. Hcr1p also physically interacts with eIF3.99 Intriguingly, deletion of HCR1 results in decreased levels of 40S subunits.100 Thus, Hcr1p may couple 40S biogenesis and translation initiation, perhaps acting during a final quality control step to monitor binding of eIF3 to the 40S subunit. A defect in this step may result in destruction of the subunit. Likewise, the release of export factors could also be coupled to recruitment of initiation factors during such a functional check of the subunit.

Acknowledgements I would like to thank members of my lab as well as J. Dahlberg and E. Lund for stimulating discussions.

298

The Nucleolus

References 1. Giese G, Wunderlich F. In vitro ribosomal ribonucleoprotein transport. Temperature-induced “graded unlocking” of nuclei. J Biol Chem 1983; 258(1):131-135. 2. Bataille N, Helser T, Fried HM. Cytoplasmic transport of ribosomal subunits microinjected into the Xenopus laevis oocyte nucleus: a generalized, facilitated process. J Cell Biol 1990; 111(4):1571-1582. 3. Fischer U, Huber J, Boelens WC et al. The HIV-1 Rev activation domain is a nuclear export signal that accesses an export pathway used by specific cellular RNAs. Cell 1995; 82(3):475-483. 4. Khanna-Gupta A, Ware VC. Nucleocytoplasmic transport of ribosomes in a eukaryotic system: is there a facilitated transport process? Proc Natl Acad Sci USA 1989; 86(6):1791-1795. 5. Pokrywka NJ, Goldfarb DS. Nuclear export pathways of tRNA and 40 S ribosomes include both common and specific intermediates. J Biol Chem 1995; 270(8):3619-3624. 6. Hassel I, Cezanne V, Trevino C et al. Export of ribosomal subunits from resealed rat liver nuclear envelopes. Eur J Biochem 1996; 241(1):32-37. 7. Hurt E, Hannus S, Schmelzl B et al. A novel in vivo assay reveals inhibition of ribosomal nuclear export in Ran-cycle and nucleoporin mutants. J Cell Biol 1999; 144(3):389-401. 8. Stage-Zimmermann T, Schmidt U, Silver PA. Factors affecting nuclear export of the 60S ribosomal subunit in vivo. Mol Biol Cell 2000; 11(11):3777-3789. 9. Baßler J, Grandi P, Gadal O et al. Identification of a 60S preribosomal particle that is closely linked to nuclear export. Mol Cell 2001; 8(3):517-529. 10. Fatica A, Cronshaw AD, Dlakic M et al. Ssf1p prevents premature processing of an early pre-60S ribosomal particle. Mol Cell 2002; 9(2):341-351. 11. Gavin AC, Bösche M, Krause R et al. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 2002; 415(6868):141-147. 12. Harnpicharnchai P, Jakovljevic J, Horsey E et al. Composition and functional characterization of yeast 66S ribosome assembly intermediates. Mol Cell 2001; 8(3):505-515. 13. Saveanu C, Bienvenu D, Namane A et al. Nog2p, a putative GTPase associated with pre-60S subunits and required for late 60S maturation steps. EMBO J 2001; 20(22):6475-6484. 14. Nissan TA, Baßler J, Petfalski E et al. 60S pre-ribosome formation viewed from assembly in the nucleolus until export to the cytoplasm. EMBO J 2002; 21(20):5539-5547. 15. Gleizes PE, Noaillac-Depeyre J, Léger-Silvestre I et al. Ultrastructural localization of rRNA shows defective nuclear export of preribosomes in mutants of the Nup82p complex. J Cell Biol 2001; 155(6):923-936. 16. Gadal O, Strauss D, Braspenning J et al. A nuclear AAA-type ATPase (Rix7p) is required for biogenesis and nuclear export of 60S ribosomal subunits. EMBO J 2001; 20(14):3695-3704. 17. Vale RD. AAA proteins. Lords of the ring. J Cell Biol 2000; 150(1):F13-19. 18. Leipe DD, Wolf YI, Koonin EV et al. Classification and evolution of P-loop GTPases and related ATPases. J Mol Biol 2002; 317(1):41-72. 19. Daigle DM, Rossi L, Berghuis AM et al. YjeQ, an essential, conserved, uncharacterized protein from Escherichia coli, is an unusual GTPase with circularly permuted G-motifs and marked burst kinetics. Biochemistry 2002; 41(37):11109-11117. 20. Dunbar DA, Dragon F, Lee SJ et al. A nucleolar protein related to ribosomal protein L7 is required for an early step in large ribosomal subunit biogenesis. Proc Natl Acad Sci USA 2000; 97(24):13027-13032. 21. Gadal O, Strauss D, Petfalski E et al. Rlp7p is associated with 60S preribosomes, restricted to the granular component of the nucleolus, and required for pre-rRNA processing. J Cell Biol 2002; 157(6):941-951. 22. Baronas-Lowell DM, Warner JR. Ribosomal protein L30 is dispensable in the yeast Saccharomyces cerevisiae. Mol Cell Biol 1990; 10(10):5235-5243. 23. Kruiswijk T, Planta RJ, Krop JM. The course of assembly of ribosomal subunits in yeast. Biochem. Biophys. Acta. 1978; 517:378-389. 24. Fath S, Milkereit P, Podtelejnikov AV et al. Association of yeast RNA polymerase I with a nucleolar substructure active in rRNA synthesis and processing. J Cell Biol 2000; 149(3):575-590. 25. Milkereit P, Gadal O, Podtelejnikov A et al. Maturation and intranuclear transport of pre-ribosomes requires noc proteins. Cell 2001; 105(4):499-509. 26. Grandi P, Rybin V, Bassler J et al. 90S pre-ribosomes include the 35S pre-rRNA, the U3 snoRNP, and 40S subunit processing factors but predominantly lack 60S synthesis factors. Mol Cell 2002; 10(1):105-115. 27. Dragon F, Gallagher JEG, Compagnone-Post PA et al. A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 2002; 417(6892):967-970.

Nuclear Export of Ribosomal Subunits

299

28. Trapman J, Planta RJ. Maturation of ribosomes in yeast. I Kinetic analysis by labelling of high molecular weight rRNA species. Biochim Biophys Acta 1976; 442(3):265-274. 29. Udem SA, Warner JR. The cytoplasmic maturation of a ribosomal precursor ribonucleic acid in yeast. J Biol Chem 1973; 248(4):1412-1416. 30. Venema J, Tollervey D. Ribosome synthesis in Saccharomyces cerevisiae. Annu Rev Genet 1999; 33(261):261-311. 31. Strezoska Z, Pestov DG, Lau LF. Bop1 is a mouse WD40 repeat nucleolar protein involved in 28S and 5.8S rRNA processing and 60S ribosome biogenesis. Mol Cell Biol 2000; 20(15):5516-5528. 32. Mattaj IW, Englmeier L. Nucleocytoplasmic transport: the soluble phase. Annu Rev Biochem 1998; 67(265):265-306. 33. Nakielny S, Dreyfuss G. Transport of proteins and RNAs in and out of the nucleus. Cell 1999; 99(7):677-690. 34. Pemberton LF, Blobel G, Rosenblum JS. Transport routes through the nuclear pore complex. Curr Opin Cell Biol 1998; 10(3):392-399. 35. Weis K. Nucleocytoplasmic transport: cargo trafficking across the border. Curr Opin Cell Biol 2002; 14(3):328-335. 36. Görlich D, Kutay U. Transport between the cell nucleus and the cytoplasm. Annu Rev Cell Dev Biol 1999; 15(607):607-660. 37. Lund E, Dahlberg JE. Direct and indirect roles of Ran GTP in nuclear export of RNAs in higher eukaryotes. In: Rush M, D’Eustachio P, eds. The Small GTPase Ran. Norwell: Kluwer Academic Publishers, 2001:59-83. 38. Wen W, Meinkoth JL, Tsien RY et al. Identification of a signal for rapid export of proteins from the nucleus. Cell 1995; 82:463-473. 39. Fornerod M, Ohno M, Yoshida M et al. CRM1 is an export receptor for leucine-rich nuclear export signals. Cell 1997; 90(6):1051-1060. 40. Stade K, Ford CS, Guthrie C et al. Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 1997; 90(6):1041-1050. 41. He F, Jacobson A. Identification of a novel component of the nonsense-mediated mRNA decay pathway by use of an interacting protein screen. Genes & Dev 1995; 9:437-454. 42. Ho J, Johnson AW. NMD3 encodes an essential cytoplasmic protein required for stable 60S ribosomal subunits in Saccharomyces cerevisiae. Mol Cell Biol 1999; 19(3):2389-2399. 43. Karl T, Önder K, Kodzius R et al. GRC5 and NMD3 function in translational control of gene expression and interact genetically. Curr Genet 1999; 34(6):419-429. 44. Belk JP, He F, Jacobson A. Overexpression of truncated Nmd3p inhibits protein synthesis in yeast. RNA 1999; 5:1055-1070. 45. Ho JHN, Kallstrom G, Johnson AW. Nmd3p is a Crm1p-dependent adapter protein for nuclear export of the large ribosomal subunit. J Cell Biol 2000; 151(5):1057-1066. 46. Gadal O, Strauss D, Kessl J et al. Nuclear export of 60S ribosomal subunits depends on Xpo1p and requires a nuclear export sequence-containing factor, Nmd3p that associates with the large subunit protein Rpl10p. Mol Cell Biol 2001; 21(10):3405-3415. 47. Trotta C, Lund E, Kahan L et al. Coordinated Nuclear Export of NMD3 and 60S Ribosomal Subunits. in press. 48. Fatica A, Tollervey D. Making ribosomes. Curr Opin Cell Biol 2002; 14(3):313-318. 49. Rout MP, Aitchison JD, Suprapto A et al. The yeast nuclear pore complex: Composition, architecture, and transport mechanism. J Cell Biol 2000; 148(4):635-651. 50. Zuk D, Belk JP, Jacobson A. Temperature-sensitive mutations in the Saccharomyces cerevisiae MRT4, GRC5, SLA2 and THS1 genes result in defects in mRNA turnover. Genetics 1999; 153(1):35-47. 51. Zinker S, Warner JR. The ribosomal proteins of Saccharomyces cerevisiae. Phosphorylated and exchangeable proteins. J Biol Chem 1976; 251(6):1799-807. 52. Ban N, Nissen P, Hansen J et al. The complete atomic structure of the large ribosomal subunit at 2.4 A resolution. Science 2000; 289(5481):905-920. 53. Spahn CMT, Beckmann R, Eswar N et al. Structure of the 80S ribosome from Saccharomyces cerevisiae - tRNA-ribosome and subunit-subunit interactions. Cell 2001; 107(3):373-386. 54. Eisinger DP, Dick FA, Trumpower BL. Qsr1p, a 60S ribosomal subunit protein, is required for joining of 40S and 60S subunits. Mol Cell Biol 1997; 17(9):5136-5145. 55. Wilkinson MF, Shyu AB. RNA surveillance by nuclear scanning? Nat Cell Biol 2002; 4(6):E144-147. 56. Bohnsack MT, Regener K, Schwappach B et al. Exp5 exports eEF1A via tRNA from nuclei and synergizes with other transport pathways to confine translation to the cytoplasm. EMBO J 2002; 21(22):6205-6215.

300

The Nucleolus

57. Nathanson L, Xia T, Deutscher MP. Nuclear protein synthesis: A re-evaluation. RNA 2003; 9(1):9-13. 58. Dahlberg JE, Lund E, Goodwin EB. Nuclear translation: What’s the evidence? RNA 2003; 9(1):1-8. 59. Johnson AW, Lund E, Dahlberg J. Nuclear Export of Ribosomal Subunits. TiBS 2002; 11:580-585. 60. Kudo N, Matsumori N, Taoka H et al. Leptomycin B inactivates CRM1/exportin 1 by covalent modification at a cysteine residue in the central conserved region. Proc Natl Acad Sci USA 1999; 96(16):9112-9117. 61. Neville M, Rosbash M. The NES-Crm1p export pathway is not a major mRNA export route in Saccharomyces cerevisiae. EMBO J 1999; 18(13):3746-3756. 62. Warner JR. The economics of ribosome biosynthesis in yeast. Trends Biochem Sci 1999; 24(11):437-440. 63. Kressler D, Doère M, Rojo M et al. Synthetic lethality with conditional dbp6 alleles identifies Rsa1p, a nucleoplasmic protein involved in the assembly of 60S ribosomal subunits. Mol Cell Biol 1999; 19(12):8633-8645. 64. Dick FA, Eisinger DP, Trumpower BL. Exchangeability of Qsr1p, a large ribosomal subunit protein required for subunit joining, suggests a novel translational regulatory mechanism. Febs Lett 1997; 419(1):1-3. 65. Allmang C, Mitchell P, Petfalski E et al. Degradation of ribosomal RNA precursors by the exosome. Nucleic Acids Res 2000; 28(8):1684-1691. 66. Faber AW, Van Dijk M, Raue HA et al. Ngl2p is a Ccr4p-like RNA nuclease essential for the final step in 3'-end processing of 5.8S rRNA in Saccharomyces cerevisiae. RNA 2002; 8(9):1095-1101. 67. Briggs MW, Burkard KT, Butler JS. Rrp6p, the yeast homologue of the human PM-Scl 100-kDa autoantigen, is essential for efficient 5.8 S rRNA 3' end formation. J Biol Chem 1998; 273(21):13255-13263. 68. Englmeier L, Fornerod M, Bischoff FR et al. RanBP3 influences interactions between CRM1 and its nuclear protein export substrates. EMBO Rep 2001; 2(10):926-932. 69. Lindsay ME, Holaska JM, Welch K et al. Ran-binding protein 3 is a cofactor for Crm1-mediated nuclear protein export. J Cell Biol 2001; 153(7):1391-1402. 70. Moy TI, Silver PA. Requirements for the nuclear export of the small ribosomal subunit. J Cell Sci 2002; 115(Pt 14):2985-2995. 71. Conti E, Izaurralde E. Nucleocytoplasmic transport enters the atomic age. Curr Opin Cell Biol 2001; 13(3):310-319. 72. Kuersten S, Ohno M, Mattaj IW. Nucleocytoplasmic transport: Ran, beta and beyond. Trends Cell Biol 2001; 11(12):497-503. 73. Macara IG. Transport into and out of the nucleus. Microbiol Mol Biol Rev 2001; 65(4):570-594. 74. Paschal BM. Translocation through the nuclear pore complex. Trends Biochem Sci 2002; 27(12):593-596. 75. Akey CW, Radermacher M. Architecture of the Xenopus nuclear pore complex revealed by threedimensional cryo-electron microscopy. J Cell Biol 1993; 122(1):1-19. 76. Hinshaw JE, Carragher BO, Milligan RA. Architecture and design of the nuclear pore complex. Cell 1992; 69(7):1133-1141. 77. Dworetzky SI, Lanford RE, Feldherr CM. The effects of variations in the number and sequence of targeting signals on nuclear uptake. J Cell Biol 1988; 107(4):1279-1287. 78. Pante N, Kann M. Nuclear pore complex is able to transport macromolecules with diameters of about 39 nm. Mol Biol Cell 2002; 13(2):425-434. 79. Wool IG. The structure and function of eukaryotic ribosomes. Annu Rev Biochem 1979; 48:719-754. 80. Cullen BR. Nuclear RNA export. J Cell Sci 2003; 116(Pt 4):587-597. 81. Ribbeck K, Görlich D. Kinetic analysis of translocation through nuclear pore complexes. EMBO J 2001; 20(6):1320-1330. 82. Winey M, Yarar D, Giddings Jr TH, et al. Nuclear pore complex number and distribution throughout the Saccharomyces cerevisiae cell cycle by three-dimensional reconstruction from electron micrographs of nuclear envelopes. Mol Biol Cell 1997; 8(11):2119-2132. 83. Ben-Efraim I, Gerace L. Gradient of increasing affinity of importin beta for nucleoporins along the pathway of nuclear import. J Cell Biol 2001; 152(2):411-417. 84. Ribbeck K, Gorlich D. The permeability barrier of nuclear pore complexes appears to operate via hydrophobic exclusion. EMBO J 2002; 21(11):2664-2671. 85. Schlatter H, Langer T, Rosmus S et al. A novel function for the 90 kDa heat-shock protein (Hsp90): facilitating nuclear export of 60 S ribosomal subunits. Biochem J 2002; 362(Pt 3):675-684.

Nuclear Export of Ribosomal Subunits

301

86. Russell DW, Spremulli LL. Purification and characterization of a ribosome dissociation factor (eukaryotic initiation factor 6) from wheat germ. J Biol Chem 1979; 254(18):8796-8800. 87. Basu U, Si K, Warner JR et al. The Saccharomyces cerevisiae TIF6 gene encoding translation initiation factor 6 is required for 60S ribosomal subunit biogenesis. Mol Cell Biol 2001; 21(5):1453-1462. 88. Senger B, Lafontaine DL, Graindorge JS et al. The nucle(ol)ar Tif6p and Efl1p are required for a late cytoplasmic step of ribosome synthesis. Mol Cell 2001; 8(6):1363-1373. 89. Ciufo LF, Brown JD. Nuclear export of yeast signal recognition particle lacking Srp54p by the Xpo1p/Crm1p NES-dependant pathway. Curr Biol 2000; 10(20):1256-1264. 90. Grosshans H, Deinert K, Hurt E et al. Biogenesis of the signal recognition particle (SRP) involves import of SRP proteins into the nucleolus, assembly with the SRP-RNA, and Xpo1p-mediated export. J Cell Biol 2001; 153(4):745-761. 91. Bécam AM, Nasr F, Racki WJ et al. Ria1p (Ynl163c), a protein similar to elongation factors 2, is involved in the biogenesis of the 60S subunit of the ribosome in Saccharomyces cerevisiae. Mol Gen Genet 2001; 266(3):454-462. 92. Ho JH-N, Kallstrom G, Johnson AW. Nascent 60S subunits enter the free pool bound by Nmd3p. RNA 2000; 6(11):1625-1634. 93. Stevens A, Hsu CL, Isham KR et al. Fragments of the internal transcribed spacer 1 of pre-rRNA accumulate in Saccharomyces cerevisiae lacking 5'→3' exoribonuclease 1. J Bacteriol 1991; 173(21):7024-7028. 94. Moy TI, Silver PA. Nuclear export of the small ribosomal subunit requires the Ran-GTPase cycle and certain nucleoporins. Genes Dev 1999; 13(16):2118-2133. 95. Milkereit P, Strauss D, Bassler J et al. A noc complex specifically involved in the formation and nuclear export of ribosomal 40 s subunits. J Biol Chem 2003; 278(6):4072-4081. 96. Trapman J, Retel J, Planta RJ. Ribosomal precursor particles from yeast. Exp Cell Res 1975; 90(1):95-104. 97. Gelperin D, Horton L, Beckman J et al. Bms1p, a novel GTP-binding protein, and the related Tsr1p are required for distinct steps of 40S ribosome biogenesis in yeast. RNA 2001; 7(9):1268-1283. 98. Wegierski T, Billy E, Nasr F et al. Bms1p, a G-domain-containing protein, associates with Rcl1p and is required for 18S rRNA biogenesis in yeast. RNA 2001; 7(9):1254-1267. 99. Valasek L, Hasek J, Trachsel H et al. The Saccharomyces cerevisiae HCR1 gene encoding a homologue of the p35 subunit of human translation initiation factor 3 (eIF3) is a high copy suppressor of a temperature-sensitive mutation in the Rpg1p subunit of yeast eIF3. J Biol Chem 1999; 274(39):27567-27572. 100. Valasek L, Hasek J, Nielsen KH et al. Dual function of eIF3j/Hcr1p in processing 20 S pre-rRNA and translation initiation. J Biol Chem 2001; 276(46):43351-43360.

302

The Nucleolus

CHAPTER 16

Proteomics of the Nucleolus Yun Wah Lam, Archa H. Fox, Anthony K.L. Leung, Jens S. Andersen, Matthias Mann and Angus I. Lamond

Abstract

T

he nucleolus is the most prominent organelle in the mammalian nucleus. It is assembled around rDNA genes and is the site of rDNA transcription and ribosome subunit assembly. However, the presence in nucleoli of proteins with no obvious relationship with ribosome synthesis suggests that we still do not completely understand the full range of nucleolar functions. The high density and structural stability of the nucleolus make it an ideal organelle to isolate. Nucleoli can be isolated from HeLa cells that are biochemically, morphologically and functionally intact and mass spectrometry analysis shows that over 350 proteins can be identified reproducibly. These proteins, which likely represent most of the human nucleolar proteome, show considerable functional diversity. The many novel factors and separate classes of proteins identified support the view that the nucleolus may perform additional functions beyond its known role in ribosome subunit biogenesis. Future studies will expand our knowledge of the nucleolar proteomes in other model organisms and will provide a more detailed quantitative picture of the levels of each protein and how this changes under a range of metabolic conditions.

Introduction Advances in mass spectrometry (MS) techniques and the completion of the human genome project enable the large-scale identification of proteins (see ref. 1 for a recent review). This powerful approach, known as proteomics, is rapidly becoming widely used to unravel the protein composition in complex mixtures ranging from purified biochemical complexes such as the human spliceosome2-5 to the complete protein complement of an entire organism such as the malaria parasite P. falciparum.6,7 The isolation and characterisation of organelles by subcellular fractionation is a well-established technique in cell biology. Many organelles have been isolated and analysed in the past century (see ref. 8 for reviews and protocols). These studies have provided invaluable information on the functions and properties of each organelle. With new developments in the proteomic approach, it is now possible to determine the major protein composition of various cytoplasmic organelles such as mitochondria,9 Golgi apparatus10,11 and chloroplast thylakoid membrane.12 The isolation of subnuclear structures, in contrast with these cytoplasmic organelles, is more difficult because nuclear compartments are not surrounded by membrane. Despite this limitation, isolation of several nuclear compartments, such as the nuclear envelope,13 nuclear pore complexes,14 interchromatin granule clusters15 and Cajal bodies16 have been reported. The most well studied nuclear organelle, however, is the nucleolus, whose high density and structural stability allow effective purification using a straightforward procedure (see below). The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

Proteomics of the Nucleolus

303

The human nucleolus is a prominent nuclear substructure assembled around tandemly repeated ribosomal genes (rDNA genes) on chromosomes 13, 14, 15, 21 and 22, and is the site of rDNA transcription and ribosomal synthesis (see Chapter 1). Visualised under EM, the nucleolus is morphologically separated into three distinct substructures: fibrillar centres (FC), which are surrounded by dense fibrillar components (DFCs), and granular components (GC) (see Chapter 2 of this book). The ability to isolate nucleoli in large scale provides an excellent starting material for identifying, purifying and studying proteins in this nuclear compartment. For example, early attempts have been made to fractionate isolated nucleoli, resulting in separation of protein complexes with differential transcriptional activities.17 In another study, specific exoribonuclease activity was purified from nucleoli isolated from Ehrlich ascites tumour cells.18 More recently, proteins from isolated nucleoli with affinity to silver nitrate were studied, leading to identification of a novel protein.19 While the role of the nucleolus as a site of ribosome subunit synthesis is well established (see Chapter 1), recently an increasing number of proteins with no obvious relationship with ribosomes have also been detected in nucleoli (see other chapters). This new information emphasizes the need to characterize the nucleolar proteome in detail if we are to understand fully its biological role. Recently, two reports on the proteomic study of the human nucleolus were published,20,21 shedding new light on this nuclear organelle and demonstrating the impact of the MS techniques on cell biology.

Isolation of Nucleoli A prerequisite for the proteomic study of a cellular component is the ability to isolate it intact, in high purity and in large quantities. The relatively high density and structural stability of the nucleolus, as compared with other cellular structures, facilitates its efficient isolation. Since the initial successful attempts to purify nucleoli from human tumour cells and rodent liver cells in the early 60s (e.g., refs. 22, 23), numerous studies have been reported on the characterisation of isolated nucleoli. Nucleoli have been purified from a large variety of mammalian tissues, including liver, thyroid24 and brain25 and from cells of nonmammalian species such as Xenopus26 and Tetrahymena (e.g., ref. 27). The nucleolar isolation procedure is robust, and therefore the general strategy has been essentially unchanged over almost 40 years. Isolated cell nuclei are subjected to sonication, whose power is adjusted so that the nucleoli remain intact while the rest of the nuclei are fragmented. Then the nucleoli are isolated by centrifugation through a density gradient based on their high density compared with other nuclear components. Modifications of the procedure are needed for nucleoli from different cell types and organisms. For example, the procedure for isolating nucleoli from adherent HeLa cells was not suitable for suspension cultured HeLa S3 cells (our unpublished results). A critical factor is the salt concentration, especially, magnesium ion concentration, used in the buffer during sonication, because the structural integrity of nucleoli decreases if the salt concentration is too low.28 However, if the Mg concentration is too high, nuclei cannot be efficiently disrupted by sonication16 and hence the purity of the isolated nucleoli is compromised. In our studies, nuclei from adherent HeLa cells were sonicated in 0.35M sucrose containing 0.5mM MgCl2, which lies between the large range of Mg concentrations reported in other recent work involving nucleolar isolation.29,21 It is essential to assess the purity and intactness of the isolated nucleoli before MS analysis. We have assessed the quality of isolated HeLa nucleoli using several criteria. First, the fraction isolated contained round or ovoid particles of uniform size; more than 95% of which can be labelled by the RNA dye Pyronin Y and by anti-nucleolar antibodies. These particles are morphologically similar to the nucleoli in intact HeLa cells, as judged by both light and electron microscopy (Fig. 1). The ultrastructure of the isolated nucleoli shows that the internal nucleolar substructures (FC, DFC and GC) remain intact. The purity of the isolated nucleoli was further confirmed by western blotting, which shows that proteins known to be present outside

304

The Nucleolus

Figure 1. Purified nucleoli are morphologically similar to nucleoli in intact cells. Both transmission electron microscopy (A) and Fluorescence light microscopy (B) were used to image nucleoli within intact HeLa cells (left) and nucleoli purified from HeLa cells (right). The central panels are enlargements of the indicated nucleoli in the intact cells. The isolated nucleoli are morphologically intact, retaining a clearly defined granular compartment (GC), dense fibrillar centre (DFC) and fibrillar centre (FC). Scale bars are 1 µm.

the nucleolus (e.g., NUP62 as shown in Fig. 2) were virtually undetectable in the isolated nucleoli, while known nucleolar proteins (e.g., nucleolin and fibrillarin) were highly enriched (Fig. 2). This was confirmed by an initial MS analysis, conducted to estimate the purity of the isolated nucleoli. Of the 80 proteins found in the screen, many were known nucleolar proteins,

Proteomics of the Nucleolus

305 Figure 2. Purified nucleoli are enriched in known nucleolar antigens and do not contain nonnucleolar proteins. Proteins from preparations of sonicated whole nuclei, nucleoplasm and nucleoli were separated by gel electrophoresis, transferred to nitrocellulose and immunolabelled with anti-NUP62 (1:5000, Babco) anti-nucleolin (1:1000, G. Dreyfuss) and anti-fibrillarin (1:10000, F. Fuller-Pace) antibodies. The nucleolar fraction is enriched in nucleolin and fibrillarin, but not in the nonnucleolar protein NUP62.

while obvious protein contaminants were absent. The detection of fibrillarin and nucleolin in the nonnucleolar fraction by Western blotting is likely due to the physiological presence of these proteins in a diffuse pool in the nucleoplasm.30 It is, however, also possible that there may be some “leakage” of nucleolar proteins during the isolation procedure. One indication of the intactness of the isolated HeLa nucleoli is that these nucleoli can incorporate BrUTP efficiently in vitro (Fig. 3A). The incorporated BrUTP is located in distinct foci inside the nucleoli, similar to the published nascent RNA pattern in nucleoli (e.g., ref. 31). The in vitro incorporation of BrUTP in isolated nucleoli was inhibited by actinomycin D, but not by α-amanitin, which selectively inhibits RNA polymerase II (Fig. 3B). This confirms that nucleolar BrUTP inocrporation was due to the action of RNA polymerase I, and that the isolated nucleoli, in agreement with previous reports, were transcriptionally active29 (see also ref. 32 for an early review). The properties of isolated nucleoli may vary greatly according to cell types, as Vandalaer et al28 reported that nucleoli isolated from ELT cells using the method we adopted showed damage, especially in the fibrillar centres, and were inefficient in transcription, while conversely we found that the isolation method they employed gave an extremely low yield of nucleoli and therefore was unsuitable for use with HeLa cells. Altogether, the combination of morphological, biochemical and functional studies demonstrate it is possible to isolate nucleoli that are both structurally and functionally intact.

Separation of Nucleolar Proteins for MS Analysis The complex mixture of nucleolar proteins can be separated using various methods before analysis by MS. Traditionally (e.g., ref. 2), proteins are separated by 2D-PAGE, and each protein spot on the gel is cut out and digested with trypsin before being subjected to MS analysis. Although 2D-PAGE provides excellent resolution for proteins with certain molecular weight (MW) and isoelectric point (pI) ranges, it selects against proteins with more extreme sizes and pI values. In order to select those proteins, 1D-PAGE has been used with different buffer systems that give maximum resolution for proteins in different MW ranges.20 The resolution of protein separation in 1D PAGE gels, though inferior to that in a 2D gel, was sufficient, because it was demonstrated that electrospray ionization MS (ESI-MS) combined with tandem MS (see below) was able to identify multiple proteins, even in complex mixtures.20 The additional advantage of electrophoretic separations of nucleolar proteins is that changes in protein levels in nucleoli isolated from cells cultured in different conditions can be compared. For example, Figure 4 shows the 1D and 2D profiles of nucleolar proteins from HeLa cells

306

The Nucleolus

Figure 3. Nucleoli purified from HeLa cells are transcriptionally active. A. (i) fluorescence micrograph of purified nucleoli following incubation with Br-UTP and subsequent immuno-labelling of the incorporated Br-UTP. Arrows indicate transcription foci. (ii) shows the same region also counterstained with Pyronin Y to highlight the location of nucleoli. B. Purified nucleoli were incubated in the presence of Actinomycin D (1µg/ml, 1 hour) and α-amanitin (100µg/ml, 1 hour), followed by incubation with Br-UTP and subsequent immuno-labelling. Scale bars are 5 µm.

cultured in the absence and presence of the RNA polymerase I inhibitor Actinomycin D. The proteins exhibiting differential expression levels were readily revealed and identified (see below). An alternative to protein separation before analysis by mass spectrometry is an approach which involves in-solution digestion of intact nucleoli, followed by analysis of the resulting peptide mixture by liquid-chromatography (LC) coupled directly to mass spectrometry. Peptides eluting from the LC column are ionized via electrospray ionization and transferred with high efficiency into the mass spectrometer for analysis by single (MS) and tandem (MS/MS) mass spectrometry. The combination of efficient peptide separation and sensitive sequencing of individual peptides makes this LC MS/MS approach an extremely powerful technique for the characterization of complex mixtures of peptides. Recent developments in instrumentation and software for automated data acquisition combined with miniaturized columns have dramatically increased the sensitivity and the number of peptides that can be sequenced during a single experiment. It is now possible to analyse mixtures of thousands of peptides by nanoLC MS/MS, making it a high-throughput method for the characterization of nucleoli, as compared with the approaches discussed above.

Proteomics of the Nucleolus

307

Figure 4. The nucleolar proteome changes when cellular transcription is inhibited. Nucleoli were isolated as described from both control cells and HeLa cells that had been incubated with medium containing Actinomycin D (1µg/ml for 4 h), the proteins solubilised in 1X LDS sample buffer (Invitrogen), then subjected to electrophoresis on 1D gels (4-12% Bis-Tris). Alternatively, for 2D SDS/PAGE, nucleolar proteins were separated first by isoelectric focussing (pH4-7), using the IPGphor system (Amersham), and then by 1D electrophoresis (4-12% Bis-Tris). Gels were stained with Colloidal Blue (Invitrogen) according to the manufacturers instructions, or were immuno-blotted with anti-coilin antibody to examine the different levels of p80/coilin, a protein known to enrich in nucleoli under these conditions. Those protein bands that were enriched in the nucleoli purified from Actinomycin D-treated cells were excised from the 1D gel and in-gel digested with trypsin. The resulting peptides were analysed by MALDI-MS and nanoES tandem MS, subjected to database searching and the identified protein names and accession numbers are indicated at left.

MS Analysis and Protein Identification In our first attempt to study the nucleolar proteome, we analysed the proteins separated by both 1D and 2D gels by MALDI-TOF MS and ESI-MS. A layered analytical strategy was designed to allow online interrogation of complex peptide mixtures and to identify proteins directly in genome sequences.20 Unambiguous sequence identification was established by directed sequencing of a minimum of two peptides for each protein, followed by matching expected and measured peptide fragment ions. 271 proteins were identified using this method, of which about 30% are novel proteins. Data on the identified proteins are available online in a searchable database (see below). The development of the nanoLCMS/MS technique allows sensitive and rapid analysis of crude mixtures of peptides separated by liquid chromatography coupled to tandem mass spectrometry. We have used this method to reanalyse HeLa nucleolar proteins. Only a small percentage of proteins were identified only by the in-gel digestion method, while a larger number of proteins were identified exclusively by nanoLCMS/MS, indicating LCMS/MS is a powerful additional technique. A similar conclusion was reached when nanoLCMS/MS was used

308

The Nucleolus

Figure 5. Distribution of conserved motifs and putative functional categories of the identified proteins in the nucleolar proteome. Nucleolar proteins, identified using either 1D/2D gel data (bottom) or LCMS/MS data (top), were divided into groups as indicated above. The two different datasets produced very similar pie-charts.

recently to reanalyse human spliceosomal proteins.4 Interestingly, however, despite such an increase in the number of identified nucleolar proteins, when the proteins were categorized in terms of their functions (see below), the distribution of functional categories was not altered significantly (Fig. 5). This suggests that while the nucleolar proteome we identified may not be complete, it nonetheless fairly reflects the distribution of protein categories in the nucleolus. Even when more nucleolar proteins are discovered in the future, their functional distribution is unlikely to change dramatically unless there are major advances in the functional annotation of the human genome. The functional categories indicated by proteomic studies are therefore likely to be a realistic reflection of HeLa nucleolar functions. The most striking feature of the functional distribution of the nucleolar proteome is the high proportion of novel factors, a surprising fact for an organelle intensively investigated for at least a century. Of the known proteins, the most common functional motifs found in these proteins are nucleic acid and nucleotide binding domains. The DEAD-box helicase motifs characteristic of the superfamily of RNA dependent ATPases were present in about 5% of the nucleolar proteome, suggesting that the control of RNA base pairing interactions may be an important feature of nucleolar function. The major function known for the nucleolus is the

Proteomics of the Nucleolus

309

transcription and processing of rRNAs and their subsequent assembly into ribosomal subunits. Consistent with this, the nucleolar proteome includes many ribosomal proteins, processing factors, and components required for transcription of the rRNA gene clusters, as well as human homologues of genes known to be involved in these processes in other organisms. This analysis has also identified novel proteins that may be involved in these processes, such as NNP8, a protein related to the yeast pre-rRNA processing factor Ski6p, and the putative exoribonucleases NNP15 and NNP11. However, not all the proteins appear to have functions associated with the known roles of the nucleolus. For example, we have also identified several protein translation factors, including eIF4A, eIF5A, eIF6, and ETF1 (peptide chain release factor subunit 1), which bind to active ribosomes in the cytoplasm but were not known to preassemble in the nucleolus with the individual ribosomal subunits. The functional distribution of nucleolar proteins is further confirmed by a second, independent proteomic study of isolated HeLa nucleoli, using a combination of in-gel digestion and LCMS/MS approaches.21 In this study, 213 proteins were identified in the HeLa nucleolus, which shared an almost identical functional distribution as in the study of Andersen et al.20 However, Scherl et al reported an apparently low degree of overlap between the two published studies: only 133 proteins were reported by Scherl et al to be identical to the proteins identified using 1D and 2D gels by Andersen et al, suggesting that the proteomic approach was not highly reproducible.33 However, reanalysis of the data shows that the overlap was actually much higher when we compare the two lists based on sequence homology. We conclude that the number of proteins common to the two studies was actually 163. If we include the expanded proteome we have identified by LCMS/MS, the degree of agreement is even more pronounced, with only17 proteins unique to Scherl et al. Such a high degree of reproducibility is remarkable considering that Scherl et al used 10 fold lower concentration of Mg ions than used by Andersen et al in their nucleolar isolation protocol, which is known to affect the integrity of nucleoli.28 In summary, the nucleolar proteomes established by two independent groups shows remarkably high agreement and provides a clear overview of the main protein components of nucleoli.

Validation of MS Data Many of the proteins identified by MS are known nucleolar proteins or homologues of nucleolar proteins in other species. There is still a large number of proteins that are either previously unidentified or have not been shown to be localised in the nucleolus. To confirm whether these proteins are indeed localised in nucleoli, and not contaminants, we systematically tag these proteins with fluorescent proteins and examine their subcellular localisation in cells following transient transfection. With the large number of proteins identified in the nucleolus, it was proved impractical to clone the full-length cDNAs and tag them with fluorescent proteins for every gene identified. Therefore, we selected a subset of the uncharacterised proteins and examined their subcellular expression pattern in order to estimate the specificity of proteins identified in our samples. So far, 18 proteins identified by MS in isolated HeLa nucleoli have been tagged. Figure 6 shows examples of the expression patterns of some of the tagged proteins. Only three tagged proteins were not detected in the nucleolus at all. This demonstrates that at least 80% of the identified proteins were at steady state associated with nucleoli in interphase cells and comfirmed that the proteomic approach is a reliable method for discovering novel nucleolar proteins. We note that the small number of tagged proteins that were not localised in the nucleolus were not necessarily contaminants, because a protein may only be accumulated in the nucleolus during a particular phase of the cell cycle or under specific metabolic conditions. For example, in vivo microscopy has shown that many proteins rapidly cycle between the nucleolus and nucleoplasm (e.g., ref. 30). The isolated nucleoli may therefore contain factors that are predominantly localised in the nucleoplasm but transiently cycle through the nucleolus. MS is sufficiently sensitive to detect these low abundance nucleolar proteins. For example, PSP1, a novel protein identified in the nucleolar proteome, was present in a previously unknown nuclear

310

The Nucleolus

Figure 6. Tagged nucleolar candidate proteins, identified by MS, accumulate within nucleoli in HeLa cells, albeit with varying patterns. Each panel shows a confocal fluorescence micrograph of HeLa cells expressing YFP fusion proteins, with the corresponding Nomarski image shown on the left. Constructs expressed were as follows (names and accession numbers in brackets): (A) YFP-PES (Pescadillo, Hs.13501), (B) YFP-HP1γ (Heterochromatin Protein 1 gamma isoform, Hs.278554). (C) YFP-NHPX (Hs.182255) and (D) YFP-PWP1 (nuclear phosphoprotein similar to S. cerevisiae PWP1, Hs.172589). Small arrows indicate nucleoli, large arrowheads indicate the localization of the fusion proteins and open arrowheads highlight small nuclear bodies also labeled by the fusion proteins (Cajal bodies in the case of YFP-NHPX). Scale bars, 5 µm.

domain termed “paraspeckles” and was apparently not localized to nucleoli, as judged by fluorescence microscopy.34 However, drug treatment and FLIP experiments confirmed that this protein cycled dynamically between the nucleoplasm and nucleoli, in a transcription-dependent manner (Fig. 7).34 To fully confirm the presence of a protein in the nucleolus, therefore, it is necessary to take this into consideration and perform FLIP experiments on the transfected cells.

Presentation and Publication of Data The large amount of data acquired from proteomic studies requires a systematic way to analyze and integrate it with the information already deposited in publicly available databases (e.g., Swissprot, EBI and Entrez). To facilitate the analyses, several databases were downloaded to our UNIX server (bioinformatics.msiwtb.dundee.ac.uk) and were interrogated with software such as standalone BLAST.35,36 Briefly, the peptide sequences obtained from the MS data were assigned to a particular protein sequence either by direct BLAST searches or, more commonly, using the MASCOT database. The UNIGENE entries derived from the obtained protein sequences provide a starting point leading to various other relevant databases, such as LocusLink and OMIM, to provide the cognate genomic and literature data, while the information deposited in the Swissprot database provides information about physical properties of each protein, such as pI and MW. The resulting information was integrated by customized PERL scripts specially written for the project and is presented in a searchable database written in Macromedia FLASH action script. The database is available online at http:// www.dundee.ac.uk/lifesciences/lamonddatabase/ (Fig. 8).

Perspectives The nucleolus can be isolated effectively from mammalian cells using a simple and straightforward procedure. This makes the nucleolus a model nuclear organelle for proteomic studies. The continuing advances in MS techniques toward high sensitivity and automation enable

Proteomics of the Nucleolus

311

Figure 7. Inhibition of transcription (by treatment with the drug Actinomycin D) results in the relocalisation of Paraspeckle Protein 1 (PSP1) to the nucleolus. Fluorescence micrograph of sections through HeLa cells expressing YFP-PSP1 before(left panel) and after (right panel) incubation with media containing Actinomycin D (1µg/ml) for 4 h. YFP-PSP1 fluorescence is shown in the bottom panel and cellular RNA (including nucleoli) achieved by staining with Pyronin Y is shown in the upper panel. Arrows indicate YFP-PSP1 paraspeckles and relocalised YFP-PSP1 at the nucleolar periphery. Scale bars 5 µm.

identification of most of the proteins present in isolated nucleoli. The basic map of HeLa nucleolar proteins is therefore now largely charted. However, the identified nucleolar proteome does not include every human protein that may either associate, or interact, with nucleoli in vivo, and some previously reported nucleolar proteins, especially factors involved in regulation of RNA polymerase I transcription, were not detected so far by proteomic methods. These proteins may have such low abundance that they have escaped detection using the methods employed so far. To expand the nucleolar proteome to include these proteins, isolated nucleoli will in future be separated by more than one method before MS analysis so that the sensitivity of MS is used to the fullest. For example, nucleolar proteins can be separated first by 1D PAGE, and proteins in gel slices then trypsin-digested before being subjected to LCMS/MS. Such an additional separation step has proved to be efficient in detecting low abundance peptides in complex biological mixtures.6 Affinity purification using probes specific for known nucleolar proteins can also be used to enrich for other interacting proteins. Some proteins may be associated relatively weakly with the nucleolus and therefore may be lost during the isolation procedure. It will be interesting to analyse HeLa nucleoli isolated using various levels of stringency, such as different concentrations of salts and detergent, in order to identify loosely bound proteins.

312

The Nucleolus

Figure 8. The Nucleolar Protein Database (http://www.dundee.ac.uk/lifesciences/lamonddatabase/). This web-based database displays interactive entries for each nucleolar protein identified in our studies. Information for each protein includes a summary of the known features, genomic location, unigene entry and proteome. Nucleolar proteins can be searched by name, category (eg. shown here are the RNA-modifying enzymes and related proteins) or molecular weight. The database is continually updated to include newly identified nucleolar proteins.

Future analyses of HeLa cell nucleoli will also include some cell type-specific nucleolar proteins, through the analysis of nucleoli purified from a variety of sources, including primary cells and cell lines derived from different tissues. Work has also been initiated to study the proteomes of nucleoli from other species, such as A. thaliana, in order to investigate the functional and biochemical conservation of the nucleolar proteome. Some proteins may interact with nucleoli only under specific metabolic conditions and therefore have not been detected in current studies. For example, it will be important to isolate and analyse nucleoli from cells at specific cell cycle stages, after DNA damage and during senescence. One challenge of these experiments is the need to detect not only the identity of proteins but also to quantitate the changes in their abundance under different conditions. We have previously successfully identified 11 proteins, including PSP1, that were found by 1D PAGE to accumulate in the nucleolus after Actinomycin D treatment20 (Fig. 4). Recently, a method called ‘stable isotope labelling with amino acids in cell culture’ or ‘SILAC’, has been described.37 This method can be adopted to quantitate the changes in the nucleolar proteome in different experimental conditions. In conclusion, although more work remains to be done, we believe that the human nucleolar proteome detailed so far represents a significant advance toward defining a comprehensive inventory of nucleolar proteins. These data should be of value for future studies on the range of biological roles performed by the nucleolus as well as the mechanisms involved in its assembly and function. Future studies will expand our knowledge of the nucleolar proteome in other

Proteomics of the Nucleolus

313

model organisms and will provide a more detailed quantitative picture of the levels of each protein and how this changes under a range of metabolic conditions.

References 1. Aebersold R, Mann M. Mass spectrometry-based proteomics. Nature 2003; 422: 198-207. In Press. 2. Neubauer G, King A, Rappsilber J et al. Mass spectrometry and EST-database searching allows characterization of the multi-protein spliceosome complex. Nat Genet 1998; 20:46-50. 3. Reichert VL, Le Hir H, Jurica MS et al. 5' exon interactions within the human spliceosome establish a framework for exon junction complex structure and assembly. Genes Dev 2002; 16:2778-2791. 4. Rappsilber J, Ryder U, Lamond AI et al. Large-scale proteomic analysis of the human spliceosome. Genome Res 2002; 12:1231-45. 5. Zhou Z, Licklider L, Gygi SP et al. Comprehensive proteomic analysis of the human spliceosome. Nature 2002; 419:182-185. 6. Lasonder E, Ishihama Y, Andersen JS et al. Analysis of the Plasmodium falciparum proteome by high-accuracy mass spectrometry. Nature 2002; 419:537-42. 7. Florens L, Washburn M, Raine JD et al. A proteomic view of the Plasmodium falciparum life cycle. Nature 2002; 419:520-6. 8. Chapters 34-55. In: Spector DL, Goldman R, Leinwand LA, eds. Cells: A Laboratory Manual. New York: Cold Spring Harbor Laboratory Press, 1997. 9. Pflieger D, Le Caer J, Lemaire C et al. Systematic identification of mitochondrial proteins by LC-MS/MS. Anal Chem 2002; 74:2400-6. 10. Taylor RS, Wu CC, Hays LG et al. Proteomics of rat liver Golgi complex: minor proteins are identified through sequential fractionation. Electrophoresis 2000; 21:3441-59. 11. Bell AW, Ward MA, Blackstock WP et al. Proteomics characterization of abundant Golgi membrane proteins. J Biol Chem 2001; 276:5152-65. 12. Gomez SM, Nishio JN, Faull KF et al. The chloroplast grana proteome defined by intact mass measurements from liquid chromatography mass spectrometry. Mol Cell Proteomics 2002; 1:46-59. 13. Dreger M, Bengtsson L, Schoneberg T et al. Nuclear envelope proteomics: novel integral membrane proteins of the inner nuclear membrane. Proc Natl Acad Sci USA 2001; 98:11943-8. 14. Cronshaw JM, Krutchinsky AN, Zhang W et al. Proteomic analysis of the mammalian nuclear pore complex. J Cell Biol 2002; 158:915-27. 15. Mintz PJ, Patterson SD, Neuwald AF et al. Purification and biochemical characterization of interchromatin granule clusters. EMBO J 1999; 18:4308-20. 16. Lam YW, Lyon CE, Lamond AI. Large-scale isolation of Cajal bodies from HeLa cells. Mol Biol Cell 2002; 13:2461-73. 17. Rosenberg N. Isolation of an MSP1-derived transcriptionally active nucleolar particle. Exp Cell Res 1986; 165:41-52. 18. Lasater LS, Eicher DC. Isolation and properties of a single-strand 5'-3' exoribonuclease from Ehrlich ascites tumor cell nucleoli. Biochemistry 1984; 23:4367-73. 19. Vandelaer M, Thiry M, Goessens G. AgNOR proteins from morphologically intact isolated nucleoli. Life Sci 1999; 64:2039-47. 20. Andersen JS, Lyon CE, Fox AH et al. Directed proteomic analysis of the human nucleolus. Curr Biol 2002; 12:1-11. 21. Scherl A, Coute Y, Deon C et al. Functional proteomic analysis of human nucleolus. Mol Biol Cell 2002; 13:4100-9. 22. Muramatsu M, Smetana K, Busch H. Quantitative aspects of isolation of nucleoli of the Walker carcinosarcoma and liver of the rat. Cancer Res 1963; 25:693-697. 23. Maggio R. Some properties of isolated nucleoli from guinea-pig liver. Biochim Biophys Acta 1966; 119:641-4. 24. Voets R, Lagrou A, Hilderson H et al. RNA synthesis in isolated bovine thyroid nuclei and nucleoli. alpha-Amanitin effect, a hint to the existence of a specific regulatory system. Hoppe Seylers Z Physiol Chem 1979; 360:1271-83. 25. Banks SP, Johnson TC. Developmental alterations in RNA synthesis in isolated mouse brain nucleoli. Biochim Biophys Acta 1973; 294:450-60. 26. Saiga H, Higashinakagawa T. Properties of in vitro transcription by isolated Xenopus oocyte nucleoli. Nucleic Acids Res 1979; 6:1929-40. 27. Matsuura T, Higashinakagawa T. In vitro transcription in isolated nucleoli of Tetrahymena pyriformis. Dev Genet 1992; 13:143-50.

314

The Nucleolus

28. Vandelaer M, Thiry M, Goessens G. Isolation of nucleoli from ELT cells: a quick new method that preserves morphological integrity and high transcriptional activity. Exp Cell Res 1996; 228:25-31. 29. Cheutin T, O’Donohue MF, Beorchia A et al. Three-dimensional organization of active rRNA genes within the nucleolus. J Cell Sci 2002; 115:3297-307. 30. Chen D, Huang S. Nucleolar components involved in ribosome biogenesis cycle between the nucleolus and nucleoplasm in interphase cells. J Cell Biol 2001; 153:169-76. 31. Masson C, Bouniol C, Fomproix N et al. Conditions favoring RNA polymerase I transcription in permeabilized cells. Exp Cell Res 1996; 226:114-25. 32. Onishi T, Muramatsu M. Techniques of in vitro RNA synthesis with isolated nucleoli. Methods Cell Biol 1978; 19:301-15. 33. Roix J, Misteli T. Genomes, proteomes, and dynamic networks in the cell nucleus. Histochem Cell Biol 2002; 118:105-16. 34. Fox AH, Lam YW, Leung AK et al. Paraspeckles: a novel nuclear domain. Curr Biol 2002; 12:13-25. 35. Altschul SF, Gish W, Miller W et al. Basic local alignment search tool. J Mol Biol 1990; 215:403-10. 36. Altschul SF, Madden TL, Schaffer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997; 25:3389-402. 37. Ong SE, Blagoev B, Kratchmarova I et al. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics 2002; 1:376-86.

CHAPTER 17

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus Thilo Sascha Lange

Abstract

V

arious prominent small RNA species traffic through nucleoli during their maturation process en route to their final cellular destination. These include tRNAs, RNAse P RNA, MRP RNA, SRP RNA, telomerase RNA, viral RNAs and also spliceosomal snRNAs. The nucleolus is the “White House” of the cell and part of a network that includes many types of nuclear bodies such as Cajal Bodies, speckles, snurposomes, PML bodies, gems or the recently described paraspeckles. The nucleolus communicates within the complex nuclear network by transiently sequestering RNAs, maturing RNPs and a variety of other cellular factors. Nucleolar functions other than in ribosome biogenesis remain elusive. With respect to spliceosomal snRNAs (U1, U2, U4, U5, U6) the nucleolus is favoured as the place where 2'-O-methylation and pseudouridylation of nucleotides of U6 and at least of some nucleotides of other snRNAs such as U2 are carried out. These modification reactions are assisted by trans-acting guide snoRNAs. A majority of the internal modifications of U1, U4 and U5, however, have been suggested to be assisted by scaRNAs (small RNAs related to snoRNAs) within Cajal bodies, rather than in nucleoli. In addition to these internal modifications which have been postulated to facilitate dynamic interactions during splicing, another maturation step has been linked to the nucleolar body in yeast: It appears to be the site where posttranscriptional 5’-cap hypermethylation of snRNAs and snoRNAs is carried out. Yet another function ascribed to the nucleolus with regard to RNA maturation is the assembly of RNAs and proteins into RNPs, such as for the signal recognition particle (SRP). It is possible that some steps in the assembly of spliceosomal factors also may occur in the nucleolus. Details to the mechanism of nucleolar traffic have been published or postulated: It is mediated by proteins that transport the snRNA to and/or anchor it within the nucleolus by binding to RNA specific intrinsic sequence elements. It has been suggested that the NHPX/ 15.5 kD protein by binding to U4 snRNA, the LSm protein complex by binding to U6 snRNA and the Sm protein complex by binding to U2 snRNA play such a role. Localization of snRNA components of the spliceosomal [U4/U6.U5] tri-snRNP takes place independently of cytoplasmic steps of snRNP maturation and of association with one another during formation of the spliceosome or during recycling after splicing. Further studies on the mechanism and role of intranuclear sorting of spliceosomal snRNAs are underway; they focus on the relationship between nucleoli and Cajal bodies, which, depending on the state of differentiation of a cell, might have an overlap in function with or have assimilated functions from nucleoli.

The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

316

The Nucleolus

Introduction A fascinating feature of the nucleolus is the multitude of small RNA species, such as tRNAs, RNAse P RNA, SRP RNA, telomerase RNA, viral RNAs and others which transit through this plurifunctional nuclear organelle. A given RNA species after synthesis in the nucleoplasm and during maturation, can engage in complex interaction with other cellular compartments or nuclear bodies, in addition to nucleoli, such as Cajal Bodies, speckles, PML bodies, gems (for review see refs. 2-5) or the recently described paraspeckles.6 Therefore, it is challenging to understand the meaning, mechanism, and kinetics of intracellular and intranuclear RNA movement. Currently, identification of the composition of the nucleolus by proteomics and RNomics13-15 go hand in hand with in vivo and in vitro studies to reveal the driving force of what has been called the nucleolar trafficking, localization, shuttling, association or sequestering of RNAs. In this chapter recent observations of a transient interaction of a number of spliceosomal small nuclear RNAs (snRNAs) with the nucleolus will be introduced by discussing first the hypothesized role of the nucleolus for the maturation of snRNAs and, second, the mechanism of nucleolar localization of snRNAs. The variety of cell types (e.g., the absence, presence, or differentiation of Cajal Bodies with functional interaction to the nucleolus) and variety of methods utilized in nucleolar studies do not allow the construction of an ultimate model but provide us with preliminary concepts about the relationship between spliceosomal factors and nucleoli.

The Relationship between Spliceosomal RNAs and the Nucleolus Nucleolar Maturation of RNA Polymerase III Transcribed Spliceosomal U6 snRNA The first evidence that the maturation of spliceosomal snRNAs can include a nucleolar phase was directly linked to the function of snoRNAs carried out in the nucleolus. In the last decade snoRNAs were described to be essential trans-acting guide RNAs during ribosomal RNA (rRNA) processing and modification. SnoRNAs of the box C/D family act as guide RNAs for 2'-O-methylation of rRNA while snoRNAs of the box H/ACA family guide rRNA pseudouridylation; certain members of both snoRNA families are required for cleavages within pre-rRNA (see chapters by Gerbi et al; Bertrand and Fournier; Raué). More recently Ganot et al18 and Tycowski et al19 suggested that, similarly, internal posttranscriptional modifications of splicosomal U6 snRNA are carried out in the nucleolus: Biochemical fractionation of mammalian tissue culture cells showed that all factors essential for 2'-O-methylation (of a total of 8 nucleotides) and pseudouridylation (3 nucleotides) of U6 snRNA are functionally active in nucleoli.18 Moreover in Xenopus oocytes various box C/D snoRNAs that serve as guide RNAs in the 2’-O-methylation of U6 nucleotides have already been identified and localize to nucleoli themselves.19 U6 snRNA is the first example of a non-rRNA molecule whose modification is guided by snoRNAs. The evidence that U6 snRNA traffics through nucleoli for modification is highly suggestive; nevertheless other box C/D snoRNAs are found and seem to mature in Cajal bodies prior to their nucleolar affiliation25-27 and U6 snRNA has been previously visualized in Cajal bodies by in situ hybridization4,31,32 where assembly of snRNPs of the cellular transcription-, splicing-, and processing- machinery is believed to occur.4,35 On the other hand the hypothesis that certain steps of U6 maturation take place in nucleoli18,19 is strengthened by direct snRNA localization assays carried out in Xenopus oocytes. After injection of in vitro transcripts of fluorescein-labeled snRNAs into oocyte nuclei, it was confirmed by fluorescence microscopy of nucleolar preparations from these nuclei that U6 snRNA can specifically and transiently shuttle through nucleoli.36,37 Similar to U6 snRNA, all spliceosomal RNAs contain post-transcriptionally synthesized 2'-O-ribose-methylated nucleotides and pseudouridines, which are confined to the function-

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

317

ally important regions of snRNAs known to interact with pre-mRNAs, other snRNAs or spliceosomal proteins.44-46 The presence of pseudouridines with hydrogen-bonding capacity and hydrophobic 2'-O-methyl groups might facilitate dynamic molecular interactions required for spliceosome function, which so far has been shown to be the case for U2 snRNA47. In studies on the intranuclear localization of snRNAs other than U6, the situation is more complicated since they include a cytoplasmic phase of maturation, unlike U6. Thus, a brief summary (see textbox) of the life-cycle of the RNAs which eventually assemble into the spliceosome will be given:

Milestones in the Life-Cycle of snRNAs Spliceosomal RNAs (except U6) after nuclear transcription by RNA polymerase II are exported to the cytoplasm by binding to an adaptor protein called PHAX1. In the cytoplasm the RNAs assemble with the heteroheptameric Sm-protein complex assisted by the activity of the SMN (survial of motor neurons) protein complex. Cytoplasmic interaction of the SMN complex with snRNPs is transient. The SMN complex, however, is also found in the nucleus in Cajal bodies and in bodies called gems (gemini of coiled bodies) containing other factors such as Gemin2-4, all of which directly interact with several Sm proteins.7-12 After cytoplasmic assembly with Sm proteins, snRNA precursors undergo 3’-end trimming and their 5'-methylguanosine cap is hypermethylated.5,12,16,17 This trimethylguanosine 5’-cap contributes to the nuclear re-import of the snRNAs as part of a bi-partite signal together with Sm proteins and is recognized by Snuportin-1 in a rather complex mechanism involving import factor Importin beta5.20-22 Non-spliceosomal U7 snRNA, which functions in histone pre-mRNA processing, follows the same pathway. After re-import to the nucleoplasm, the snRNPs accumulate in interchromatin granule clusters and various nuclear bodies2,23,24 (including Cajal bodies and nucleoli; for details see next section). In contrast, the lifecycle of U6 does not include a cytoplasmic phase: Upon transcription the first protein to associate with U6 snRNA is La which generally binds to the 3'-termini of nascent RNA polymerase III transcripts and a number of viral RNAs.28-30 Subsequently, the 3'-end of U6 undergoes maturation to a terminal poly U stretch and conversion of the terminal OH-group to a 2',3'-cyclic phosphate.33,34 Various proteins including Prp24/SART3 form the U6 snRNP and La is replaced by an heteroheptameric Sm-like ("LSm") protein complex which facilitates the formation of the [U4/U6] di-snRNP and the [U4/U6.U5] tri-snRNP.38-43 LSm proteins are still present in the tri-snRNP while Prp24/SART3 is not; both are implicated in the recycling of snRNPs during subsequent spliceosomal cycles.38,39,48-51 A fully assembled [U4/U6.U5] tri-snRNP eventually contains at least 23 proteins: In addition to the seven Sm proteins and seven LSm proteins, functional counterparts of at least 9 proteins are present in the tri-snRNP in yeast and human cells.55-58 Eventually the spliceosome is formed by the ordered interaction of the U1 and U2 snRNPs, the [U4/U6.U5] tri-snRNP particle and of non-snRNP splicing factors.12,55,56,59,60

Do All Spliceosomal RNAs Traffic through the Nucleolus? The cellular location responsible for the site-specific synthesis of modified nucleotides in RNA polymerase II transcribed snRNAs, still remains an enigma. Depending on the cell system and method used, RNA II polymerase transcribed spliceosomal U1, U2, U4, and U5 snRNAs have been observed in various nuclear bodies. These include Cajal bodies4,31,52-54 as well as nucleoli4,17,36,37,52,53 and snRNPs localize to both compartments.52,61,62 Generally, it should be said that in situ hybridization or labeled transcripts derived from plasmids after transformation of these snRNAs often showed nucleolar signals just above background. The only snRNA, however, which has not been reported in nucleoli at all is U7, a non-spliceosomal snRNA which functions in histone pre-mRNA 3’ processing.63,64 In Xenopus oocytes U4 snRNA and U5 snRNA can localize to both, nucleoli as well as Cajal bodies, with its preference depending on cytoplasmic passage as shown in Figure 1. This was

318

The Nucleolus

Figure 1. Localization of U4 and U5 snRNA to nucleoli and Cajal bodies of Xenopus laevis oocytes. Fluorescein-labelled in vitro-transcribed wild type U5 snRNA (WT U5), wild type U4 snRNA (WT U4), a U4 mutant deficient for nucleolar localization (U4 NoLE) or a synthetic RNA serving as negative control (control RNA) were injected into Xenopus laevis oocytes (see Methods in refs.37,66 for details). Either 4 hours after cytoplasmic injection or 2 hours after injection directly into the nucleus, nuclear spreads were prepared67 and analyzed by phase contrast (PC) or fluorescence microscopy (FL-green). Nucleoli can be distinguished from other non-chromosomal nuclear bodies since only the nucleoli contain DNA which is visualized by staining (DAPI positive-blue). Cajal bodies, which do not contain DNA (DAPI negative) and often are associated with B-snurposomes4 are indicated by arrows. After nuclear injection, wild type U4 and U5 snRNA preferentially localized to nucleoli which are stained strongly; Cajal bodies are stained weakly (compare arrow for DAPI and FL). After cytoplasmic injection the wild type snRNAs exhibit a stronger preference to localize to Cajal bodies but also localize to nucleoli. The NoLE mutant or the control RNA do not stain either body after cytoplasmic injection (see above) or nuclear injection.66 Lampbrush chromosomes (see U4 NoLE mutant panel; PC, DAPI) and B-snurposomes (which are DAPI negative) are not labelled by any of the injected RNAs. Bar is 10 mm.

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

319

demonstrated by fluorescence microscopy of nucleolar preparations after injection of fluorescein-labeled snRNA in vitro transcripts into either Xenopus oocyte nuclei or the cytoplasm. The transcripts carried a 5’-monomethyl G cap like their in vivo counterparts and exhibit normal nucleo/cytoplasmic traffic.37,65,66 Interestingly, after nuclear injection wild type U4 and U5 snRNA stain nucleoli strongly and Cajal bodies weakly but after cytoplasmic injection and nuclear import both snRNAs exhibit a stronger preference to localize to Cajal Bodies but still localize to nucleoli (Fig. 1). In contrast, a U4 mutant deficient for nucleolar localization or a synthetic RNA serving as negative control (control RNA) in both scenarios do not stain either nuclear body, confirming the specificity of the observed signals. The internal controls given in experiments like this, especially the lack of signals for mutant small RNAs versus their wild type counterpart,37,53,66 suggest that nucleolar localization is specific and support previous observations of some snRNA association with nucleoli by in situ hybridisation or in other systems.4,17,36,37,52,53 Similarly, expression of various tagged Sm proteins in HeLa cells indirectly suggest that snRNPs (after cytoplasmic passage) specifically associate with Cajal bodies as well as nucleoli.52,54 On a cautionary note, it should be said that the detection and function of a given guide snoRNA or snRNA in Cajal bodies and/or nucleoli might vary depending on the differentiation state of a given cell type: Depending on the cell system, Cajal bodies can be tightly associated with or even localized within nucleoli; they can vary in number or presence and while some transformed cells and also immortal cell lines exhibit Cajal bodies in high numbers, various mammalian somatic cells do not have Cajal bodies (see refs. 2, 35, 68). Interestingly, there is deductive evidence that the presence of maturing snRNPs might directly influence the numbers of Cajal bodies in some cell systems: Overexpression of the snRNP protein SmB in primary mammalian cells lacking Cajal bodies has been shown to induce Cajal body formation.54

A Role of the Nucleolus in Post-Transcriptional Modification of All Spliceosomal RNAs? Recently, it has been suggested that post-transcriptional 2'-O-methylation and pseudouridylation of RNA polymerase II-specific spliceosomal snRNAs, in contrast to the modification of U6, takes place in the Cajal body rather than in the nucleolus.46 In that study, the authors (Darzacq et al.) characterized novel putative guide RNAs with both box C/D and/ or H/ACA motifs of snoRNA that are predicted to direct both 2'-O-methylation and/or pseudouridylation of the RNA pol II-specific U1, U2, U4 or U5 snRNAs. Unexpectedly, the new guide RNAs as well as the previously characterized U85 box C/D-H/ACA RNA which assists in modification of U5 snRNA seem to accumulate specifically in Cajal bodies and not in nucleoli in a human transformed cell line (Hela cells).46,69 Therefore, in contrast to snoRNAs, they comprise a novel class of small nuclear RNAs, called the small Cajal body-specific RNAs (scaRNAs). So far, twelve putative guide scaRNAs have been linked with the synthesis of 12 2'-O-methylated nucleotides and two pseudouridines in the U1, U2, U4 and U5 snRNAs which led Darzacq et al46 to suggest that the synthesis of most, and perhaps all, 2'-O-methylated nucleotides and pseudouridines in pol II-specific spliceosomal snRNAs is directed by scaRNAs. Apparently, these RNAs, despite the presence of two independent nucleolar localization signals of snoRNAs (the C/D and H/ACA motif25,27,70-75), do not localize to the nucleolus. Since box C/D snoRNAs are believed to transit through Cajal bodies before they accumulate in the nucleolus,26,27,76 it has been hypothesized that the C/D motif targets scaRNAs to the Cajal body, where a putative retention factor inhibits the export of these RNAs to nucleoli.46 However, identification of a box H/ACA scaRNA which can carry out one specific pseudouridylation in U2 snRNA demonstrates that even some RNAs lacking C and D boxes can also accumulate in the Cajal body46 in addition to their postulated destination—the nucleolus. Direct evidence from modification studies of chimeric RNAs and U2 in vitro transcripts in Xenopus oocyte nuclei assigns the function of modification of at least some sites of U2 snRNA to the nucleolus.53 Thus, both organelles, Cajal bodies as well as nucleoli, appear to be the sites

320

The Nucleolus

where particular 2'-O-methylations and pseudouridylations of U2 snRNA are carried out. Potentially the modification of some nucleotides of the other snRNAs for which a guide scaRNA has not yet been identified could still be carried out in the Cajal Body and/or the nucleolus by guide snoRNAs. Other observations support the hypothesis that both the Cajal body as well as the nucleolus, fulfill functions in the maturation of snRNPs. Expression of various tagged Sm proteins in HeLa cells suggested that RNA polymerase II derived snRNPs (after cytoplasmic passage) transiently localize to both organelles before they eventually accumulate in nuclear speckles for further function in splicing52,.54 This unidirectional pathway through Cajal bodies and nucleoli leads to full maturation and assembly of these snRNPs. Accordingly, fully matured snRNPs after cell division immediately localize to speckles but not to Cajal bodies or to nucleoli upon entering the nucleus.52

Other Nucleolar Functions during Maturation of snRNPs? In addition to the modifications discussed above a further function is carried out by nucleolar bodies in yeast. Recent data revealed that Tgs1p, an evolutionarily conserved protein essential for post-transcriptional hypermethylation of the 5’-caps of both snRNAs and snoRNAs is localized to this substructure of the yeast nucleolus.17 Moreover, deletion of the TGS1 gene leads to a splicing defect that correlates with nucleolar retention of U1 snRNA and Mouaikel et al17 suggest a trafficking pathway in which yeast snRNAs and snoRNAs cycle through the nucleolus to posttranscriptionally undergo 5’-cap hypermethylation. In Hela cells the human homolog of the methyltransferase, hTgs1, and trimethylated U3 snoRNA are concentrated in Cajal bodies rather than nucleoli.26 In any case, while the nucleolar body in yeast may be the site of 5’-cap hypermethylation of snRNAs, this step of maturation of pol II transcribed snRNAs in higher eucaryotes occurs in the cytoplasm as one prerequisite for nuclear reimport (see textbox, previous section). Yet another role ascribed to the nucleolus with regard to RNA maturation in general is the association with proteins to RNPs, such as in SRP assembly.77,78 It can not be excluded that, similarly, some steps in the assembly of other small nuclear RNPs, including spliceosomal factors are carried out in the nucleolus in which hundreds of proteins and various small RNAs can be located.13-15 Cajal bodies, which are equally complex in their composition, have already been suggested to participate in the intranuclear sorting and assembly of various small nuclear RNPs of the cellular transcription-, splicing-, and processing-machinery.4,35

Mechanisms of Nucleolar Trafficking of Spliceosomal RNAs A Cis-Acting Nucleolar Localization Element (NoLE) Essential for Targeting of U6 snRNA to the Nucleolus The hypothesis that certain small nuclear RNPs specifically localize to nucleoli during their maturation and/or assembly is supported by the identification of specific cis-acting sequences within a particular RNA required for nucleolar targeting. Earlier approaches to understand the mechanism by which small RNAs are targeted to nucleoli were based on the identification of such sequences or motifs essential for nucleolar traffic, the so called Nucleolar Localization Elements (NoLEs—see Table 1 and references therein). These NoLEs are believed to be recognized by proteins that either transport the RNA or RNP, respectively, from the nucleoplasm to the nucleolus and/or anchor it within the nucleolus. Currently, a variety of studies are underway to determine which proteins of a given RNP bind to these NoLEs and, thus, are implicated in nucleolar shuttling. The studies to define elements essential for nucleolar targeting of a given small RNA species, as summarized in Table 1 (also see chapter by Bertrand and Fournier), were based on approaches such as the detection of a given mutated or chimeric RNA by microscopy after expression from a transfected plasmid or after injection of labeled fluorescent in vitro transcripts into tissue culture cells or Xenopus laevis oocytes.

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

321

Table 1. Cis-acting nucleolar localization elements (NoLEs) of small RNA species

snoRNAs

RNA Species Associated with the Nucleolus [Function]

Nucleolar Localization Element (NoLE)

box C/D snoRNAs [2'-O-methylation and/or rRNA processing]

box C/D motif 25,27,70,71,73

box H/ACA snoRNAs [pseudouridylation and/or rRNA processing]

box H, box ACA 72,74,75

MRP RNA [5.8S rRNA processing; rRNA processing] RNAse P RNA [5’ processing of pre-tRNA] Diverse RNAs

SRP RNA Telomerase RNA tRNAs

snRNAs

To-antigen binding site 79

To-antigen binding site 80,81

Alu domain, Helix 8 of S-domain 82 box H, box ACA 83,84 localizes to nucleoli but NoLEs N/D 85

U6 snRNA

3’-end, LSm-binding site36,37

U2 snRNA

Sm-binding site53

U5 snRNA

other than Sm-binding site66

U4 snRNA [mRNA splicing]

NHPX/15.5 kD binding site37,66

In the case of U6 snRNA, the only region required for nucleolar localization after injection of U6 in vitro transcripts into Xenopus oocyte nuclei is the 3'-end. This 3’-NoLE of U6 is both essential and sufficient for nucleolar localization.37 It was shown that U6 snRNA after mutation of the 3’-NoLE loses its ability to localize to both nucleoli as well as to Cajal bodies37, suggesting that this sequence may be both a Cajal body localization element (CaBLE) and a NoLE. This element acts differently than a second class of NoLEs, namely the C/D motif of snoRNAs which is a potential regulator for transit during snoRNP maturation from Cajal bodies (necessary for export) to nucleoli (element essential for import/localization).25-27,76 In U1 snRNA a third type of a cis-acting localization sequence has been described where sequences at the 5' end of U1 are essential for export of U1 snRNA from Cajal bodies, but in this case do not contribute to nucleolar localization.4 As a fourth variety, the Sm site in U7 snRNA is essential for RNA localization to Cajal bodies but U7, which functions in histone pre-mRNA 3’ processing and not in splicing, is not found in nucleoli at all.63 Although, several proteins make direct contact with U6 snRNA during conversion from free U6 to the tri-snRNP, only the La protein and the LSm protein complex bind to the 3’-end (the NoLE) of U6 snRNA. La is the first protein to associate with U6 snRNA upon transcription, it has been observed in nucleoli, and binds to precursors of various RNA polymerase III

322

The Nucleolus

and some RNA polymerase II transcripts to guarantee a normal pathway of maturation (see textbox, previous section). Intriguing as the idea may be, it appears unlikely that La mediates nucleolar localization of certain other RNA polymerase III transcripts86 (e.g., SRP RNA, RNase P RNA). It was ruled out that this is the case for U6 snRNA, since conversion of the 3’-OH group of U6 snRNA to a 3'-phosphate, which prevents association with the La protein, did not affect U6 localization to nucleoli (or Cajal bodies) in Xenopus oocytes.37 In the in vivo situation, during the maturation of U6 snRNA, La is replaced with the Sm-like protein complex, LSm2-8 (see textbox, previous section), which is a candidate to initiate nucleolar shuttling of U6 snRNA.37 It might be interesting to note that this LSm complex in Xenopus also binds to a box C/D snoRNA (U8) but not to U3 or U14 nor to box C/D NoLE's common to these RNAs;87 thus the LSm protein is not essential for nucleolar targeting of U8 but rather pursues custom-tailored interactions with various unrelated small RNAs for different functions.

Signals Essential for Nucleolar Targeting of RNA Polymerase II Transcribed Spliceosomal RNAs The RNA polymerase II transcribed snRNAs for which the mechanism of nucleolar localization has been studied in detail (and so far solely in Xenopus oocytes) are U2, U4, and U5 snRNA. Mutational analysis suggested that nucleolar localization of U4, U5 as well as U6 snRNA is initiated for the individual snRNP and independent of the formation of a di- or tri-snRNP (which occurs during splicing and subsequent snRNP recycling for further spliceosomal cycles; see textbox, previous section): Mutant U4 and U6 molecules that lack sites essential for base-pairing to one another88,89 still localize to nucleoli and, moreover, depletion of endogenous U6 snRNA does not disturb nucleolar association of U4 or U5 snRNA wild type transcripts.37,66 An additional conclusion of these studies is that NoLE sequences of snRNAs are fundamentally distinct from internal sequences essential for [U4/U6] snRNP formation, while all of the above are important for further function in splicing. Similarly, functionally important regions of box C/D snoRNAs involved in pre-rRNA processing are distinct from the box C/D NoLEs essential for nucleolar targeting of these RNAs.25,27,70,71,73-75 Since formation of a [U4/U6.U5] snRNP is not a qualifying event for nucleolar localization, a candidate factor to assist shuttling of either one of these snRNAs through the nucleolus should recognize it individually before engagement with other snRNAs. In the case of U4 the only intrinsic sequence relevant for nucleolar localization coincides with the binding site for the NHPX/15.5 kD protein and just 5 nucleotides essential for binding to the protein act as the essential NoLE for U466 (also see U4 NoLE mutant Fig. 1). Indeed, the NHPX/15.5 kD protein has been shown to interact in vivo with a form of U4 snRNA that is not stably associated with U6 snRNA90 and U4 snRNA by itself is sufficient for specific and stable interaction with this protein.91 Moreover, mutant U4 with an intact binding site for the NHPX/15.5 kD protein but incapable of association with U6 localizes to the nucleolus while mutants of the NHPX/15.5 kD binding site do not localize to nucleoli anymore.37 Therefore, it can be suggested that the NHPX/15.5 kD protein mediates nucleolar localization of U4 before engagement in a di- or tri- snRNP. This hypothesis is strengthened by the proposal that the NHPX/ 15.5 kD protein might nucleate the assembly of proteins of the [U4/U6] snRNP.12,92 Interestingly, the NHPX/15.5 (Snu13p in yeast) not only binds to the NoLE of spliceosomal U4 snRNA but also to the C/D core structure of snoRNAs,76,93 equally essential for nucleolar localization (see chapter by Bertrand and Fournier). Thus, it can be hypothesized that NHPX/ 15.5 mediates nucleolar localization of several classes of small RNA. Intriguingly, labeled NHPX/ 15.5 kD protein after in vivo expression in various cell lines was shown to associate with nucleoli after being routed transiently to speckles. It followed this unidirectional pathway not only in primary cells that lack Cajal bodies, but also in other cell lines where labeled NHPX/15.5 kD first localized to speckles, then to Cajal Bodies and to nucleoli which both remain stained.90 Curiously, this unidirectional movement is reciprocal to the route of nuclear maturation of

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

323

snRNPs that transiently first localize to both organelles before they eventually accumulate in nuclear speckles for further function in splicing.52,54 Accordingly, a function of NHPX/15.5 kD in intranuclear shuttling of various small RNAs might be rather complex and, thus, has not yet been fully elucidated. Recent data obtained from Xenopus oocytes suggest that Sm proteins, commonly associating with RNA polymerase II transcribed snRNAs during their cytoplasmic maturation (see textbox, previous section), are not candidate factors to play a role for U4 or U5 nucleolar association. It was confirmed that the mutation of the Sm binding site prevented Sm protein binding to either snRNA, and it abolished nuclear import after injection of fluorescently labeled U4 or U5 transcripts into the cytoplasm; nevertheless these mutant molecules were still able to localize to nucleoli when injected directly into the nucleoplasm.66 In contrast, for U2 snRNA, in the same Xenopus oocyte system, it recently has been be suggested that nucleolar localization (but not association with Cajal bodies) is dependent on the Sm binding site.53 An explanation for the different observations for U2 snRNA as compared to U4 or U5 snRNA is not at hand. It is known, however, that structural features unique to U1 and U5 snRNAs individually influence the otherwise conserved Sm binding site.94 Moreover, the Sm site seems not to enable another pol II transcribed snRNA, U7, to localize to nucleoli but instead is essential for localization of U7 snRNA exclusively to Cajal bodies.63 In the case of U7, two proteins in the heteroheptameric Sm complex are different than in other snRNPs which might bear an explanation for different function of the U7 Sm site as well as different sub-nuclear localization.64 Since the Sm site is not essential for nucleolar localization of U4 and U5 and kinetic analysis suggested that nucleolar localization of these snRNAs in Xenopus oocytes can occur before export from and re-import to the nucleus,37,95 a cytoplasmic phase might not be required prior to nucleolar localization. This is in line with the observation that snRNAs do not need a hypermethylated 5'-methylguanosine cap to localize to nucleoli,66 but they need the cap for efficient re-import into the nucleus.16,66,95 The various models for the mechanism of nucleolar traffic of spliceosomal RNAs are summarized in Figure 2. Nucleolar localization is mediated by protein(s) that either transport the snRNA to and/or anchor it within the nucleolus by binding to RNA specific intrinsic NoLEs. It has been suggested that the NHPX/15.5 kD protein by binding to U4 snRNA, the LSm protein complex by binding to U6 snRNA and the Sm protein complex by binding to U2 snRNA play such a role (for further details see legend of Fig. 2).

Concluding Remarks Based on the available data it is reasonable to hypothesize that some internal modifications of spliceosomal RNAs are carried out by guide snoRNPs in nucleoli. Cajal bodies, depending on the cell system, might have an overlap in function or have assimilated functions from nucleoli. Accordingly, the interaction between Cajal bodies and nucleoli, among other sub-nuclear structures, deserves to be analyzed in more detail. The nucleolus has been favored as the place where 2'-O-methylation and pseudouridylation of nucleotides of U6 snRNA and at least some modifications of U2 snRNA occur. Apparently, for U1, U4 and U5 snRNA these posttranscriptional modifications take place in Cajal bodies; potentionally, modification of some sites to which a guide scaRNA has not been identified could still be carried out in nucleoli. The internal modifications of spliceosomal RNAs have been postulated to facilitate formation of dynamic molecular interactions required in splicing. Thus, they are likely carried out before assembly of complex snRNPs, such as the [U4/U6.U5] tri-snRNP, which is part of the spliceosome. This hypothesis is supported by the notion that U2, U4, U5, and U6 can localize to nucleoli as individual RNPs, which also holds true for Cajal bodies. While the nucleolar body in yeast appears to be the site of posttranscriptional 5’-cap hypermethylation of snRNAs and snoRNAs, no other nucleolar functions during snRNA maturation have been described in higher eucaryotes. Further studies are necessary to establish a durable model to the role, kinetics, and mechanism of nucleolar traffic of spliceosomal snRNAs.

324

The Nucleolus

Figure 2. Models for nucleolar localization of spliceosomal RNAs of the [U4/U6.U5 tri-snRNP]. U6 snRNA, upon transcription by RNA polymerase III in the nucleoplasm, is stabilized by the La protein, which binds to the 3’-NoLE of U6 (left side of model). La, subsequently is replaced by the LSm protein complex. This facilitates localization of the U6 snRNP to the nucleolus. RNA polymerase II transcribed spliceosomal snRNAs such as U4 and U5 can localize to nucleoli at various points during maturation. In Xenopus they associate with nucleoli (and to a lesser extent with Cajal bodies) before passage to the cytoplasm (see path/arrow I in figure). Nucleolar localization of U4 and U5 can still occur after the cytoplasmic step (see dotted path/ arrow II), but then the affinity for Cajal bodies is more pronounced (see Fig. 1). U4 and U5 localization to nucleoli does not require 5’-cap tri-methylation or association with Sm proteins occurring after export to the cytoplasm. For U4 snRNA binding of the NHPX/ 15.5 kD protein (15.5 kD in figure) which associates with U4 snRNA early during maturation (and also with the NoLE of box C/D snoRNAs) appears to be essential for nucleolar traffic. Nucleolar localization of all snRNA components of the spliceosomal [U4/U6.U5] tri-snRNP particle takes place independent of an association with one another but might still occur after nuclear re-import (see dotted path/arrow III) and subsequently during or after formation of a complex multi-snRNP. Eventually, splicing ensues by the ordered interaction of the U1 and U2 snRNPs and the [U4/U6·U5] tri-snRNP particle which recycle during subsequent spliceosomal cycles (see textbox, previous section). For cells with Cajal bodies, nucleolar localization of snRNAs might take place before and/or after transit through the Cajal body where assembly and intranuclear sorting of various small nuclear RNPs is believed to take place.

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

325

Acknowledgment This work is dedicated to Hedwig I. Lange-Berenrot. The research in this article was supported by the National Science Foundation (MCB Grant 0091166 to T.S.L.).

References 1. Ohno M, Segref A, Bachi A et al. PHAX, a mediator of U snRNA nuclear export whose activity is regulated by phosphorylation. Cell 2000; 101(2):187-198. 2. Matera AG. Nuclear bodies: multifaceted subdomains of the interchromatin space. Trends Cell Biol 1999; 9(8):302-309. 3. Wu Z, Murphy C, Wu CH et al. Snurposomes and coiled bodies. Cold Spring Harb Symp Quant Biol 1993; 58:747-754. 4. Gall JG, Bellini M, Wu Z et al. Assembly of the nuclear transcription and processing machinery: Cajal bodies (coiled bodies) and transcriptosomes. Mol Biol Cell 1999; 10(12):4385-4402. 5. Narayanan U, Ospina JK, Frey MR et al. SMN, the spinal muscular atrophy protein, forms a pre-import snRNP complex with snurportin1 and importin beta. Hum Mol Genet 2002; 11(15):1785-1795. 6. Fox AH, Lam YW, Leung AK et al. Paraspeckles: A novel nuclear domain. Curr Biol 2002; 12(1):13-25. 7. Matera AG. Of coiled bodies, gems, and salmon. J Cell Biochem 1998; 70(2):181-192. 8. Young PJ, Le TT, thi Man N et al. The relationship between SMN, the spinal muscular atrophy protein, and nuclear coiled bodies in differentiated tissues and cultured cells. Exp Cell Res 2000; 256(2):365-374. 9. Liu Q, Dreyfuss G. A novel nuclear structure containing the survival of motor neurons protein. EMBO J 1996; 15(14):3555-3565. 10. Fischer U, Liu Q, Dreyfuss G. The SMN-SIP1 complex has an essential role in spliceosomal snRNP biogenesis. Cell 1997; 90(6):1023-1029. 11. Liu Q, Fischer U, Wang F et al. The spinal muscular atrophy disease gene product, SMN, and its associated protein SIP1 are in a complex with spliceosomal snRNP proteins. Cell 1997; 90(6):1013-1021. 12. Will CL, Lührmann R. Spliceosomal UsnRNP biogenesis, structure and function. Curr Opin Cell Biol 2001; 13(3):290-301. 13. Andersen JS, Lyon CE, Fox AH et al. Directed proteomic analysis of the human nucleolus. Curr Biol 2002; 12(1):1-11. 14. Hüttenhofer A, Kiefmann M, Meier-Ewert S et al. RNomics: an experimental approach that identifies 201 candidates for novel, small, non-messenger RNAs in mouse. EMBO J 2001; 20(11):2943-2953. 15. Dundr M, Misteli T. Nucleolomics: an inventory of the nucleolus. Mol Cell 2002; 9(1):5-7. 16. Mattaj IW. Cap trimethylation of U snRNA is cytoplasmic and dependent on U snRNP protein binding. Cell 1986; 46(6):905-911. 17. Mouaikel J, Verheggen C, Bertrand E et al. Hypermethylation of the cap structure of both yeast snRNAs and snoRNAs requires a conserved methyltransferase that is localized to the nucleolus. Mol Cell 2002; 9(4). 18. Ganot P, Jady BE, Bortolin ML et al. Nucleolar factors direct the 2'-O-ribose methylation and pseudouridylation of U6 spliceosomal RNA. Mol Cell Biol 1999; 19(10):6906-6917. 19. Tycowski KT, You ZH, Graham PJ et al. Modification of U6 spliceosomal RNA is guided by other small RNAs. Mol Cell 1998; 2(5):629-638. 20. Fischer U, Sumpter V, Sekine M et al. Nucleo-cytoplasmic transport of U snRNPs: definition of a nuclear location signal in the Sm core domain that binds a transport receptor independently of the m3G cap. EMBO J 1993; 12(2):573-583. 21. Huber J, Cronshagen U, Kadokura M et al. Snurportin1, an m3G-cap-specific nuclear import receptor with a novel domain structure. EMBO J 1998; 17(14):4114-4126. 22. Hamm J, Darzynkiewicz E, Tahara SM et al. The trimethylguanosine cap structure of U1 snRNA is a component of a bipartite nuclear targeting signal. Cell 1990; 62(3):569-577. 23. Lewis JD, Tollervey D. Like attracts like: getting RNA processing together in the nucleus. Science 2000; 288(5470):1385-1389. 24. Sleeman JE, Lamond AI. Nuclear organization of pre-mRNA splicing factors. Curr Opin Cell Biol 1999; 11(3):372-377. 25. Samarsky DA, Fournier MJ, Singer RH et al. The snoRNA box C/D motif directs nucleolar targeting and also couples snoRNA synthesis and localization. EMBO J 1998; 17(13):3747-3757.

326

The Nucleolus

26. Verheggen C, Lafontaine DL, Samarsky D et al. Mammalian and yeast U3 snoRNPs are matured in specific and related nuclear compartments. EMBO J 2002; 21(11):2736-2745. 27. Narayanan A, Speckmann W, Terns R et al. Role of the box C/D motif in localization of small nucleolar RNAs to coiled bodies and nucleoli. Mol Biol Cell 1999; 10(7):2131-2147. 28. Rinke J, Steitz JA. Association of the lupus antigen La with a subset of U6 snRNA molecules. Nucleic Acids Res 1985; 13(7):2617-2629. 29. Kunkel GR, Maser RL, Calvet JP et al. U6 small nuclear RNA is transcribed by RNA polymerase III. Proc Natl Acad Sci USA 1986; 83(22):8575-8579. 30. Chang YN, Kenan DJ, Keene JD et al. Direct interactions between autoantigen La and human immunodeficiency virus leader RNA. J Virol 1994; 68(11):7008-7020. 31. Carmo-Fonseca M, Pepperkok R, Carvalho MT et al. Transcription-dependent colocalization of the U1, U2, U4/U6, and U5 snRNPs in coiled bodies. J Cell Biol 1992; 117(1):1-14. 32. Matera AG, Ward DC. Nucleoplasmic organization of small nuclear ribonucleoproteins in cultured human cells. J Cell Biol 1993; 121(4):715-727. 33. Lund E, Dahlberg JE. Cyclic 2',3'-phosphates and nontemplated nucleotides at the 3' end of spliceosomal U6 small nuclear RNA’s. Science 1992; 255(5042):327-330. 34. Terns MP, Lund E, Dahlberg JE. 3'-end-dependent formation of U6 small nuclear ribonucleoprotein particles in Xenopus laevis oocyte nuclei. Mol Cell Biol 1992; 12(7):3032-3040. 35. Gall JG. Cajal bodies: the first 100 years. Annu Rev Cell Dev Biol 2000; 16:273-300. 36. Lange TS, Gerbi SA. Transient nucleolar localization Of U6 small nuclear RNA in Xenopus laevis oocytes. Mol Biol Cell 2000; 11(7):2419-2428. 37. Gerbi SA, Lange TS. All small nuclear RNAs (snRNAs) of the [U4/U6.U5] tri-snRNP localize to nucleoli; Identification of the nucleolar localization element of U6 snRNA. Mol Biol Cell 2002; 13(9):3123-3137. 38. Achsel T, Brahms H, Kastner B et al. A doughnut-shaped heteromer of human Sm-like proteins binds to the 3'-end of U6 snRNA, thereby facilitating U4/U6 duplex formation in vitro. EMBO J 1999; 18(20):5789-5802. 39. Mayes AE, Verdone L, Legrain P et al. Characterization of Sm-like proteins in yeast and their association with U6 snRNA. EMBO J 1999; 18(15):4321-4331. 40. Pannone BK, Xue D, Wolin SL. A role for the yeast La protein in U6 snRNP assembly: evidence that the La protein is a molecular chaperone for RNA polymerase III transcripts. EMBO J 1998; 17(24):7442-7453. 41. Salgado-Garrido J, Bragado-Nilsson E, Kandels-Lewis S et al. Sm and Sm-like proteins assemble in two related complexes of deep evolutionary origin. EMBO J 1999; 18(12):3451-3462. 42. Seraphin B. Sm and Sm-like proteins belong to a large family: identification of proteins of the U6 as well as the U1, U2, U4 and U5 snRNPs. EMBO J 1995; 14(9):2089-2098. 43. Vidal VP, Verdone L, Mayes AE et al. Characterization of U6 snRNA-protein interactions. Rna 1999; 5(11):1470-1481. 44. Massenet S, Mougin A, Branlant C. Posttranscriptional modifications in the U small nuclear RNAs. In: Grosjean HaB, R., ed. Modification and Editing of RNA. Washington, DC: ASM Press, 1998:201-227. 45. Reddy R, Busch H. Small nuclear RNAs: RNA sequences, structure and modifications. In: Birnsteil ML, ed. Structure and Function of major and minor small nuclear ribonucleoprotein particles. Berlin: Springer-Verlag; 1988:1-37. 46. Darzacq X, Jady BE, Verheggen C et al. Cajal body-specific small nuclear RNAs: a novel class of 2'-O-methylation and pseudouridylation guide RNAs. EMBO J 2002; 21(11):2746-2756. 47. Yu YT, Shu MD, Steitz JA. Modifications of U2 snRNA are required for snRNP assembly and pre-mRNA splicing. EMBO J 1998; 17(19):5783-5795. 48. Bell M, Schreiner S, Damianov A et al. p110, a novel human U6 snRNP protein and U4/U6 snRNP recycling factor. EMBO J 2002; 21(11):2724-2735. 49. Schneider C, Will CL, Makarova OV et al. Human U4/U6.U5 and U4atac/U6atac.U5 tri-snRNPs exhibit similar protein compositions. Mol Cell Biol 2002; 22(10):3219-3229. 50. Raghunathan PL, Guthrie C. A spliceosomal recycling factor that reanneals U4 and U6 small nuclear ribonucleoprotein particles. Science 1998; 279(5352):857-860. 51. Staley JP, Guthrie C. Mechanical devices of the spliceosome: motors, clocks, springs, and things. Cell 1998; 92(3):315-326. 52. Sleeman JE, Lamond AI. Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway. Curr Biol 1999; 9(19):1065-1074. 53. Yu YT, Shu MD, Narayanan A et al. Internal modification of U2 small nuclear (sn)RNA occurs in nucleoli of Xenopus oocytes. J Cell Biol 2001; 152(6):1279-1288.

Trafficking of Spliceosomal Small Nuclear RNAs through the Nucleolus

327

54. Sleeman JE, Ajuh P, Lamond AI. snRNP protein expression enhances the formation of Cajal bodies containing p80-coilin and SMN. J Cell Sci 2001; 114(Pt 24):4407-4419. 55. Behrens SE, Lührmann R. Immunoaffinity purification of a [U4/U6.U5] tri-snRNP from human cells. Genes Dev 1991; 5(8):1439-1452. 56. Will CL, Lührmann R. Protein functions in pre-mRNA splicing. Curr Opin Cell Biol 1997; 9(3):320-328. 57. Gottschalk A, Neubauer G, Banroques J et al. Identification by mass spectrometry and functional analysis of novel proteins of the yeast [U4/U6.U5] tri-snRNP. EMBO J 1999; 18(16):4535-4548. 58. Stevens SW, Abelson J. Purification of the yeast U4/U6.U5 small nuclear ribonucleoprotein particle and identification of its proteins. Proc Natl Acad Sci USA 1999; 96(13):7226-7231. 59. Konarska MM, Sharp PA. Association of U2, U4, U5, and U6 small nuclear ribonucleoproteins in a spliceosome-type complex in absence of precursor RNA. Proc Natl Acad Sci USA 1988; 85(15):5459-5462. 60. Wassarman DA, Steitz JA. Interactions of small nuclear RNA’s with precursor messenger RNA during in vitro splicing. Science 1992; 257(5078):1918-1925. 61. Lyon CE, Bohmann K, Sleeman J et al. Inhibition of protein dephosphorylation results in the accumulation of splicing snRNPs and coiled bodies within the nucleolus. Exp Cell Res 1997; 230(1):84-93. 62. Sleeman J, Lyon CE, Platani M et al. Dynamic interactions between splicing snRNPs, coiled bodies and nucleoli revealed using snRNP protein fusions to the green fluorescent protein. Exp Cell Res 1998; 243(2):290-304. 63. Wu CH, Murphy C, Gall JG. The Sm binding site targets U7 snRNA to coiled bodies (spheres) of amphibian oocytes. Rna 1996; 2(8):811-823. 64. Pillai RS, Will CL, Lührmann R et al. Purified U7 snRNPs lack the Sm proteins D1 and D2 but contain Lsm10, a new 14 kDa Sm D1-like protein. EMBO J 2001; 20(19):5470-5479. 65. Fischer U, Darzynkiewicz E, Tahara SM et al. Diversity in the signals required for nuclear accumulation of U snRNPs and variety in the pathways of nuclear transport. J Cell Biol 1991; 113(4):705-714. 66. Gerbi SA, Borovjagin AV, Odreman FE et al. U4 snRNA nucleolar localization requires the NHPX/ 15.5-kD protein binding site but not Sm protein or U6 snRNA association. J Cell Biol 2003; 162(5):821-832. 67. Gall JG, Murphy C, Callan HG et al. Lampbrush chromosomes. In: Kay BaPH, ed. Methods Cell Biol. Vol 36. New York: Academic Press, 1991:149-166. 68. Spector DL, Lark G, Huang S. Differences in snRNP localization between transformed and nontransformed cells. Mol Biol Cell 1992; 3(5):555-569. 69. Jady BE, Kiss T. A small nucleolar guide RNA functions both in 2'-O-ribose methylation and pseudouridylation of the U5 spliceosomal RNA. EMBO J 2001; 20(3):541-551. 70. Lange TS, Borovjagin A, Maxwell ES et al. Conserved boxes C and D are essential nucleolar localization elements of U14 and U8 snoRNAs. EMBO J 1998; 17(11):3176-3187. 71. Lange TS, Borovjagin AV, Gerbi SA. Nucleolar localization elements in U8 snoRNA differ from sequences required for rRNA processing. Rna 1998; 4(7):789-800. 72. Lange TS, Ezrokhi M, Amaldi F et al. Box H and box ACA are nucleolar localization elements of U17 small nucleolar RNA. Mol Biol Cell 1999; 10(11):3877-3890. 73. Lange TS, Ezrokhi M, Borovjagin AV et al. Nucleolar localization elements of Xenopus laevis U3 small nucleolar RNA. Mol Biol Cell 1998; 9(10):2973-2985. 74. Narayanan A, Lukowiak A, Jady BE et al. Nucleolar localization signals of box H/ACA small nucleolar RNAs. EMBO J 1999; 18(18):5120-5130. 75. Ruhl DD, Pusateri ME, Eliceiri GL. Multiple conserved segments of E1 small nucleolar RNA are involved in the formation of a ribonucleoprotein particle in frog oocytes. Biochem J 2000; 348 Pt 3:517-524. 76. Verheggen C, Mouaikel J, Thiry M et al. Box C/D small nucleolar RNA trafficking involves small nucleolar RNP proteins, nucleolar factors and a novel nuclear domain. EMBO J 2001; 20(19):5480-5490. 77. Grosshans H, Deinert K, Hurt E et al. Biogenesis of the signal recognition particle (SRP) involves import of SRP proteins into the nucleolus, assembly with the SRP-RNA, and Xpo1p-mediated export. J Cell Biol 2001; 153(4):745-762. 78. Politz JC, Yarovoi S, Kilroy SM et al. Signal recognition particle components in the nucleolus. Proc Natl Acad Sci USA 2000; 97(1):55-60. 79. Jacobson MR, Cao LG, Wang YL et al. Dynamic localization of RNase MRP RNA in the nucleolus observed by fluorescent RNA cytochemistry in living cells. J Cell Biol 1995; 131(6 Pt 2):1649-1658.

328

The Nucleolus

80. Jacobson MR, Cao LG, Taneja K et al. Nuclear domains of the RNA subunit of RNase P. J Cell Sci 1997; 110 ( Pt 7):829-837. 81. Jarrous N, Wolenski JS, Wesolowski D et al. Localization in the nucleolus and coiled bodies of protein subunits of the ribonucleoprotein ribonuclease P. J Cell Biol 1999; 146(3):559-572. 82. Jacobson MR, Pederson T. Localization of signal recognition particle RNA in the nucleolus of mammalian cells. Proc Natl Acad Sci USA 1998; 95(14):7981-7986. 83. Mitchell JR, Cheng J, Collins K. A box H/ACA small nucleolar RNA-like domain at the human telomerase RNA 3' end. Mol Cell Biol 1999; 19(1):567-576. 84. Lukowiak AA, Narayanan A, Li ZH et al. The snoRNA domain of vertebrate telomerase RNA functions to localize the RNA within the nucleus. Rna 2001; 7(12):1833-1844. 85. Bertrand E, Houser-Scott F, Kendall A et al. Nucleolar localization of early tRNA processing. Genes Dev 1998; 12(16):2463-2468. 86. Maraia RJ. La protein and the trafficking of nascent RNA polymerase iii transcripts. J Cell Biol 2001; 153(4):F13-18. 87. Tomasevic N, Peculis BA. Xenopus LSm proteins bind U8 snoRNA via an internal evolutionarily conserved octamer sequence. Mol Cell Biol 2002; 22(12):4101-4112. 88. Vankan P, McGuigan C, Mattaj IW. Roles of U4 and U6 snRNAs in the assembly of splicing complexes. EMBO J 1992; 11(1):335-343. 89. Vankan P, McGuigan C, Mattaj IW. Domains of U4 and U6 snRNAs required for snRNP assembly and splicing complementation in Xenopus oocytes. EMBO J 1990; 9(10):3397-3404. 90. Leung AK, Lamond AI. In vivo analysis of NHPX reveals a novel nucleolar localization pathway involving a transient accumulation in splicing speckles. J Cell Biol 2002; 157(4):615-629. 91. Nottrott S, Hartmuth K, Fabrizio P et al. Functional interaction of a novel 15.5kD [U4/U6.U5] tri-snRNP protein with the 5' stem-loop of U4 snRNA. EMBO J 1999; 18(21):6119-6133. 92. Vidovic I, Nottrott S, Hartmuth K et al. Crystal structure of the spliceosomal 15.5kD protein bound to a U4 snRNA fragment. Mol Cell 2000; 6(6):1331-1342. 93. Watkins NJ, Segault V, Charpentier B et al. A common core RNP structure shared between the small nucleoar box C/D RNPs and the spliceosomal U4 snRNP. Cell 2000; 103(3):457-466. 94. Jarmolowski A, Mattaj IW. The determinants for Sm protein binding to Xenopus U1 and U5 snRNAs are complex and non-identical. EMBO J 1993; 12(1):223-232. 95. Fischer U, Darzynkiewicz E, Tahara SM et al. Diversity in the signals required for nuclear accumulation of U snRNPs and variety in the pathways of nuclear transport. J Cell Biol 1991; 113:705-714.

CHAPTER 18

Nontraditional Roles of the Nucleolus Mark O.J. Olson

Summary

I

n addition to its well-established role in ribosomal subunit assembly, unexpected new functions of the nucleolus have been discovered in recent years. These include interaction with viral components, regulation of tumor suppressor and oncogene activities, cell cycle regulation in yeast, signal recognition particle assembly, modification of small RNAs, control of aging and modulating telomerase function. Some of these nontraditional functions are recognized as genuine activities of the nucleolus, whereas the significance of others is not well understood. Additional progress in defining the precise role of the nucleolus in the nontraditional activities is expected in the next decade.

Introduction After the initial description of the nucleolus, more than a century passed before it was established that the nucleolus is the primary location for ribosome assembly in eukaryotes. Now, several decades after the development of that major biological concept, a series of puzzling observations have begun to accumulate. A handful of different proteins and RNAs with no obvious relationship to ribosome biogenesis have been discovered in nucleoli from a variety of organisms. These findings have raised the question: does the nucleolus do something more than assemble ribosomes? As this possibility gained greater acceptance, there has been a general realization that the nucleolus is indeed a plurifunctional subnuclear body,1 although the extent of the added functionality remains a subject of debate. Nonetheless, the discovery of the new nucleolar capabilities has not only enlightened us about the roles of multiple subnuclear compartments in carrying out nuclear processes, but it has also been a major factor in reinvigorating scientific interest in the nucleolus. The nontraditional roles of the nucleolus fall into several categories: 1. 2. 3. 4.

processing, modification or assembly of cellular components, participation in regulatory mechanisms, sequestration of molecules for protective purposes and molecules that are clearly present in the nucleolus, but have poorly defined roles.

Many examples of the final category are not covered in this chapter because of the uncertainty about the significance of their nucleolar location. For a discussion of role of the nucleolus in modifying spliceosomal RNAs the reader is referred to the chapter by Lange. Figure 1 summarizes several of the novel molecules found in the nucleolus and the possible reasons for their presence.

Viral Components in the Nucleolus Viruses are dependent on host cell machinery for their survival and propagation. Consequently, viral components are found in most cellular subcompartments, including the nucleolus. In fact, Hiscox2 goes so far as to suggest that for many viruses, the nucleolus could be an The Nucleolus, edited by Mark O.J. Olson. ©2004 Eurekah.com and Kluwer Academic / Plenum Publishers.

330

The Nucleolus

Figure 1. Nontraditional roles of the nucleolus. Several examples are illustrated in the figure. A variety of cellular processes occur in the nucleolus including tRNA maturation and partial assembly of signal recognition particles (SRPs) and telomerase complexes. The yeast cell cycle is regulated through nucleolar sequestration of the cdc14p protein phosphatase in the RENT complex, which also contains the proteins Sir2p and Net1p. The activity of tumor-suppressor protein p53 is also regulated by sequestration in the nucleolus of components that control its degradation (MDM2 and E2F1), which bind to the ARF protein. The HIV-1 Rev protein interacts with protein B23 in the nucleolus and also recruits the nuclear export factor hCRM1 and nucleoporins Nup98 and Nup214 to the nucleolus. Proteins Sir3p and Sir4p relocate from telomeres to nucleoli during aging in yeast, presumably in response to accumulation of extrachromosomal circles of rDNA. Sir2p also participates in chromatin silencing and suppression of extrachromosomal circle formation. Adapted from: Olson MO, Hingorani K, Szebeni A. Int Rev Cytol 2002; 219:199-266.

important gateway to infection. One of the first viral proteins that attracted attention to the possibility that the nucleolus plays a role in its function was the HIV-1 Rev protein (reviewed by Kjems and Askjaer3). When HIV proteins were first characterized, researchers were surprised to find that Rev localizes predominantly4,5 and Tat partially6,7 to nucleoli. The nucleolar localization of Rev could be due, at least in part, to its specific interaction with nucleolar protein B23,8 which is proposed to serve as a molecular chaperone for Rev, possibly inhibiting its tendency to aggregate and improving its mobility in the cell.9 However, since ongoing transcription by RNA polymerase I is required for its nucleolar localization, Rev probably also interacts with preribosomal RNA (pre-rRNA).10 Although a mutant form of the Rev protein that does not spend time in the nucleolus is capable of regulating HIV-1 mRNA splicing,11 the nucleolar location could be important for optimal Rev activity; i.e., for storage or maintaining a threshold level of Rev, preventing its nonproductive shuttling or protecting it from degradation.10 Rev is also capable of recruiting nucleoporins Nup98 and Nup214 as well as the nuclear export factor hCRM1 to the nucleolus.12 This leaves open the possibility that a

Nontraditional Roles of the Nucleolus

331

complex consisting of Rev, hCRM1 and the two nucleoporins assembles in the nucleolus and that Nup98 and Nup214 might facilitate movement of the complex to the nuclear pore. While it is reasonably well established that Rev and Tat spend time in the nucleolus, there is also some evidence that the HIV-1 mRNA passes through the nucleolus during its life cycle. A hammerhead ribozyme that specifically cleaves the HIV-1 mRNA was engineered into U16 snoRNA, which resulted in the ribozyme accumulating in nucleoli.13 In HIV-1 infected cells expressing this ribozyme, there was significant suppression of HIV-1 replication. Although the above studies provide indirect evidence for at least a transient nucleolar location of HIV-1 mRNA, it has not been possible to directly detect this message in the nucleolus. Studies by Cmarko et al14 using in situ hybridization showed HIV RNAs accumulating in clusters along with Rev in the nucleoplasm and at the nuclear pores, but they were unable to find a signal for it in nucleoli. Thus, the involvement of the nucleolus in the life cycle of HIV-1 message needs further investigation. More recently, Michienzi et al15 constructed a vector that fused the TAR element to U16 snoRNA (the Tat protein binds the TAR element of the HIV-1 mRNA). The U16TAR chimeric RNA exclusively localized to the nucleolus and served as a decoy to trap Tat in that compartment. More importantly, when this construct was expressed in human T lymphoblastoid CEM cells, it effectively inhibited HIV-1 replication. These results could be interpreted to suggest that nucleolar trafficking of Tat is critical for HIV-1 replication. It is also interesting that the affinity of TAR for Tat is enhanced by its interaction with cyclin T1.16 Using FRET analyses Marcello et al17 showed that Tat and cyclin T1 are in close proximity to each other in the nucleolus, lending further support to the idea that nucleolar location is important for Tat function. Thus, both Tat and the HIV-1 mRNA seem to travel through the nucleolus, but the function of this traffic is not well understood. There are several other examples of viral proteins located in the nucleolus during parts of their life cycles. The coronavirus nucleoprotein (N), which is involved in several aspects of viral replication, is found in both cytoplasmic and nucleolar compartments. Two possible roles for the nucleolar localization of the N protein have been proposed.18 First, while in the nucleolus, its interaction with ribosomal proteins could cause dissociation of the coronavirus core and release of the genomic RNA. Second, the N protein could subvert host cell translation by disrupting formation of new ribosomes or interfering with the cell cycle. The N protein might then enhance the translation of the viral RNA by binding to its 5’ end, which could recruit existing ribosomes. When the adenovirus protein V accumulates in the nucleolus during infection it causes two prominent nucleolar proteins, nucleolin and protein B23 to relocate to the cytoplasm.19 Although adenovirus infection inhibits both synthesis and processing of pre-rRNA, protein V by itself does not have this inhibitory effect. Adenovirus disruption of the cellular balance could occur by interfere with the synthesis and processing of rRNA, resulting in disruption of more general cellular processes.19 Alternatively, the virus could enhance its own replication by relocating nucleolin, which, in turn, could interfere with replication through its ability to repress transcription.20 Two hepatitis delta virus antigens (HDAgs) localize to nucleoli of human hepatocyte-derived cells.21 These form a complex, which contains both proteins B23 and nucleolin. Expression of the HDAgs in the human cell lines not only up-regulates the level of B23 mRNA, but interaction of the antigens with protein B23 also enhances the hepatitis delta virus replication. Whether the modulation of viral replication is due to its ability to bind nucleic acids22 or its molecular chaperone activity9 or both remain to be determined. Less is known about the function of a number of other viral proteins that localize to the nucleolus. These include the UL3 protein of herpes simplex virus type 223 and human papillomavirus type 16 E7 protein.24 It is interesting that the latter protein interacts with the retinoblastoma-1 protein, which is capable of regulating rDNA transcription (see chapter by Cavanaugh et al).

332

The Nucleolus

Regulation of Tumor Suppressor and Oncogene Activities The tumor-suppressor protein p53 is generally acknowledged to be the primary guardian of the metazoan genome.25 It does this by inducing cells to arrest or die in response to various forms of stress, including DNA damage. But how is the activity of p53 regulated? Recent work suggests that at least partial regulation is achieved by nucleolar sequestration mechanisms.26 The levels of p53 are controlled by a balance between synthesis and degradation; the protein is degraded in the cytoplasm after interaction with the oncoprotein MDM2. MDM2 has intrinsic E3 ligase activity that conjugates ubiquitin to p53, which marks the protein for degradation and promotes its export to the cytoplasm where it is degraded in proteasomes. A nucleolar protein called p19ARF (p14ARF in humans; hereafter designated as ARF), forms a complex with MDM2, which attenuates p53 degradation by inhibiting its ubiquitination. When MDM2 and ARF are coexpressed, MDM2 increasingly becomes localized to nucleoli. The highly basic regions of ARF are important for nucleolar localization.27 ARF is predominantly localized to the granular component of the nucleolus and its presence there is dependent on continuing transcriptional activity.28 These observations led to a proposed mechanism of controlling p53 levels through the tethering of the MDM2 component of its degradative system in the nucleolus. However, more recent studies using human cells indicate that certain versions of ARF can form a complex with MDM2 in extranucleolar regions.29 This work suggests that nucleolar localization is not absolutely essential for ARF activity toward MDM2, at least in the human system. Although the actions of ARF could be carried out strictly in the nucleoplasm, the authors suggest that the nucleolar location might be a means of stabilizing or storing ARF, or even providing a backup system for controlling p53 levels. Reinforcing this view is a study by David-Pfeuty and Nouvian-Dooghe30 showing that the levels of ARF and its location are highly variable and dependent on cell type and physiological conditions. ARF partitions between the nucleolus and nucleoplasm in subconfluent, epithelial-like HeLa cells, but it becomes nucleoplasmic in confluent cells. However, in HeLa cell lines that express much higher levels of ARF, the protein is almost exclusively nucleolar during interphase. Furthermore, ARF disappears nearly completely from the cell during metaphase, but rapidly accumulates in nucleoli again during early G1 phase. These studies support the idea that the action of ARF is in the nucleoplasm rather than the nucleolus. Work by Lin and Lowe31 showing that MDM2 is not found in the nucleolus following cell cycle arrest provides additional evidence that ARF interacts with MDM2 in the nucleoplasm. The accumulated evidence suggests that although nucleolar sequestration of MDM2 plays a role in p53 regulation, it is probably not the sole mechanism. Another condition that causes MDM2 to translocate to the nucleolus is inhibition of proteasome activity.32 It is curious that proteasome inhibition also causes the accumulation of p53 as well as MDM2 in the nucleolus. 33 The latter translocation is dependent on phosphoinositol-3-kinase (PI-3kinase). Although the interpretation of these results is unclear, it is possible that in the absence of a clear route to the proteasome this complex of proteins follows a default pathway to the nucleolus. The above studies describe mechanisms by which p53 levels are controlled by regulating its degradation. However, some cells lack p53 and in these, tumor suppression is achieved in part by targeting transcription factors for degradation. ARF is capable of mediating tumor suppression by targeting some members of the E2F (E2F1, -2, and –3) family of transcription factors that promote cell cycle progression.34 Proteolysis of E2F1 is dependent on the presence of a functional proteasome as is the case with p53. Co-expression of ARF and E2F1 also results in the relocalization of E2F1 to the nucleolus, but how this event triggers degradation has not been determined. Taken together, the evidence suggests that ARF acts in more than one way to promote tumor suppression in response to oncogenic signals. How ARF can protect p53 from degradation in one instance and target other transcription factors for degradation in another remains unresolved.

Nontraditional Roles of the Nucleolus

333

Further studies on the regulation of p53 have uncovered more interacting partners. One of these is a protein called nucleostemin.35 Nucleostemin is present in embryonic and adult stem cells, primitive bone marrow cells and cancer cells, but not in differentiated cells of most adult tissues. Prior to terminal cell division, this protein is abruptly down regulated during differentiation. Nucleostemin contains two GTP-binding motifs and a N-terminal basic domain that is required for nucleolar localization. The authors propose that nucleostemin is present in the nucleolus when it is in the non-GTP bound form; dissociation from the nucleolus occurs upon GTP binding. Nucleostemin was shown to interact with p53 by both GST pull-down and coimmunoprecipitation experiments. The binding of p53 by nucleostemin is proposed to promote cell proliferation; in contrast, its down-regulation might facilitate exit from the cell cycle by releasing p53 to act as a tumor suppressor. As with many nontraditional nucleolar proteins, neither the significance of its nucleolar localization nor what it is recognizing in the nucleolus is currently understood. The abundant nucleolar protein, nucleolin, also interacts with p53 in response to the stress of ionizing radiation or the radiomimetic agent, camptothecin.36 After γ-irradiation, nucleolin translocates from the nucleolus to the nucleoplasm; however, this translocation does not occur in p53-null cells. This suggests that the nucleolin relocation is p53-dependent, which is further supported by the detection of a nucleolin-p53 complex after the irradiation treatment. Stress also mobilizes nucleolin to interact with replication protein A (RPA) to suppress chromosomal replication.37 In this case, the stress response seems to be mediated by p53 whereby replication is suppressed by translocating nucleolin out of its normal nucleolar site. It is curious that topoisomerase I undergoes a nucleolar to nucleoplasmic translocation in response to DNA damage that is also p53-dependent.38 The fact that nucleolin and topoisomerase I interact with each other39 raises the possibility that a larger complex is mobilized during certain kinds of stress responses. An unanswered question is how p53 triggers the translocation of these proteins from the nucleolus to the nucleoplasm. The Fas/JNK pathway of apoptosis is mediated by a protein called Daxx, which acts as a transcriptional repressor of several genes.40 These include the MMP1 and Bcl-2 genes that are regulated by the transcriptional activator ETS1. The activity of the latter protein is inhibited by binding to Daxx. The Daxx protein itself distributes to various parts of the nucleus, including the nucleolus. Lin and Shih41 found that Daxx function is modulated by a protein called MSP58 (58 kDa microspherule protein). Overexpression of MSP58 results in the recruitment of Daxx to the nucleolus; the sequestration of Daxx in the nucleolus via MSP58 results in the derepression of Daxx regulated genes. The extensive studies with p53 and the developing story with Daxx suggest that nucleolar involvement in signaling pathways is more common than previously believed.

Cell Cycle Regulation in Yeast Nucleolar sequestration is part of the system that controls cell division in both S. cerevisiae and S. pombe. The exit from mitosis in these organisms is controlled by cyclin-dependent kinase (CDK) inactivation. In S. cerevisiae a key player in this process is the protein phosphatase Cdc14p, which both promotes the degradation of a cyclin subunit and the accumulation of a protein kinase inhibitor. Cdc14p is regulated by sequestration in the nucleolus for most of the cell cycle, but it is released from that site during anaphase.42 The protein responsible for its nucleolar anchoring has been identified as Cfi1p (also called Net1p), a protein related to Reg1p, a regulatory subunit for protein phosphatase 1. Cdc14p is part of a larger nucleolar complex, which not only contains Cfi1p/Net1p, but also Sir2p and possibly Nan1p.43 This is termed the RENT complex, an acronym for “regulator of nucleolar silencing and telophase exit.” The release of Cdc14p is controlled by a signaling cascade that is designated the mitotic exit network (MEN).44 The MEN consists of nine or more proteins including GTPases, protein kinases and nucleotide exchange factors that promote release of Cdc14p from the nucleolus, probably through phosphorylation events. Several cellular events appear to regulate the activity

334

The Nucleolus

of MEN: the position of the nucleus relative to the bud, the status of sister chromatid separation, microtubule integrity and DNA damage. Thus, MEN is not only a signaling cascade, but it is also a system of sensors that ensures that mitotic exit is initiated only after sister chromatid separation has begun and that there is segregation of the genetic material between the mother and daughter cells. In the fission yeast, S. pombe, mitotic events are also controlled by nucleolar sequestration, but by a somewhat different mechanism than found in S. cerevisiae. In this case, there is a homologue of Cdc14p called clp1/flp1, which regulates the phosphorylation of cdc2, the sole CDK in S. pombe. In contrast to Cdc14p in S. cerevisiae, whose regulatory effects are seen in the later stages of mitosis, clp1/flp1 is released from the nucleolus early in mitosis. Although it is not known what causes the release of clp1/flp1 from the nucleolus, a signaling cascade called SIN (septation initiation network) is essential for sustaining its exclusion from the nucleolus. Although similar strategies are used to control exit from mitosis in both species of yeast, we don’t know for certain whether the nucleolar location of these cell cycle regulators is absolutely essential for their functions.

Signal Recognition Particle Assembly One of the better-characterized nontraditional nucleolar processes is the maturation of the signal recognition particle (SRP). This complex of a small RNA and several proteins targets nascent secretory proteins to the endoplasmic reticulum. The possible involvement of the nucleolus in SRP assembly was first suggested in studies by Jacobson and Pederson,45 who found that when SRP RNA was injected into the nuclei of mammalian cells, it initially and rapidly localized in nucleoli, after which there was a gradual decline in nucleolar signal and an increase in its presence in the cytoplasm . Later, three out of four of the SRP proteins of the S domain (SRP19, SRP68 and SRP72) but not SRP54 were found in the nucleolus.46 A plausible reason for the nucleolar location is that SRP assembly might utilize some of the ribosome assembly machinery. However, Politz et al47 found that the SRP RNA is not present to any significant extent in the fibrillar centers or the dense fibrillar components of the nucleolus. Although some of the SRP RNA was found in the granular component, most of it was contained in regions of the nucleolus that are not involved with ribosome biogenesis. Pederson and Politz48 originally suggested that SRP assembly might be coordinated with ribosome assembly as a mutual quality control mechanism for both classes of particles. The more recent data makes this possibility less likely. On the other hand, it is entirely possible that SRP assembly utilizes some of the multiple molecular chaperones and helicases found in the nucleolus49,50 in this process. The yeast SRP, is a ribonucleoprotein complex consisting of an RNA molecule (scR1) and six proteins. Grosshans et al51 found four of the six proteins in the nucleolus (Srp14p, Srp21p, Srp68p, and Srp72p); these are the SRP core proteins. The Srp54p protein is exclusively cytoplasmic, but Sec65p is both nucleolar and nucleoplasmic. The SRP core proteins are imported into the nucleolus using the ribosomal protein import receptors Pse1p and Kap123p/Yrb4p and it is proposed that the core proteins act as RNA chaperones to aid in the correct folding and stabilization of scR1. The SRP RNA is only transiently located in the nucleolus as an intermediate product in combination with the core proteins to give the pre-SRP particle. The latter particle is then exported to the cytoplasm via the nuclear export receptor, Xpo1p and a subset of nucleoporins. The final assembly of the yeast SRP is performed in the cytoplasm, as in mammalian species.

Nucleolar Processing of RNA pol III Synthesized Transcripts The nucleolus is involved in the some stages of the life cycles of certain tRNAs. The processing of tRNA occurs in a series of steps, which are regulated spatially as well as temporally. The reactions include the removal of the 5’ leader from pre-tRNAs, 3’ end cleavage, splicing and multiple nucleotide modifications.52,53 The La protein binds to the 3’ end of the newly synthesized transcript at an early stage, and this complex serves as the substrate for the ribonucleoprotein enzyme RNase P, which removes the 5’leader sequence. Using probes against pre-tRNA

Nontraditional Roles of the Nucleolus

335

introns, Bertrand et al54 demonstrated the nucleolar presence of the precursors for tRNALeu3 and tRNATrp as well as the dimeric transcript of tRNAArg/tRNAAsp. However, it is unlikely that all of the 5’ processing of tRNAs occurs in the nucleolus and it is not clear whether any 3’ tRNA processing takes place in that location.52 Not only is pre-tRNA processing taking place in the nucleolus, assembly of the processing enzyme, RNase P, also seems to be occurring there. The RNA component of RNAse P has been localized primarily in the nucleolus, although the protein subunits of the enzyme are also found in both nucleoli and Cajal bodies.55 In S. cerevisiae, the RNA subunit of RNase P, RPR1, is initially synthesized as a precursor (pre-RPR1), from which flanking sequences on both ends are removed during or after subunit assembly. Srisawat et al56 found that pre-RPR1 localizes to the nucleolus. This suggests that the processing of the precursor RNA and part of the assembly process take place in the nucleolus. A role of the nucleolus in tRNA biosynthesis is supported indirectly by studies on tRNA gene mediated silencing of RNA polymerase II (pol II) transcription. In certain cases, tRNA-class polymerases (pol III) negatively regulate neighboring RNA pol II promoters in S. cereviseae.57 Curiously, the transcription mediated silencing (tgm) is independent of the orientation of the tRNA gene and shows no requirement for binding to either the upstream pol II factors or the pol II holoenzyme. This phenomenon was shown in one case to be linked with the Cbf5 protein. The latter protein is a probable pseudouridine synthetase, which is associated with snoRNAs and is involved in ribosomal RNA maturation (see Chapters 12 and 13). A mutant that affects the expression of this protein has been shown to suppress tgm silencing. Since the biogenesis of at least some tRNAs begins in the nucleolus, this finding raised the question of a possible link between tgm silencing and the nucleolus. The nucleoli of the CBF5 mutant are slightly fragmented and the pre-tRNAs are localized in the nucleoplasm instead of nucleoli. This loss could be attributed to an inadequate supply of snoRNPs playing transport or structural roles. The fragmentation of the nucleoli could be due to incomplete rRNA pseudouridylation, resulting in a disruption of nucleolar organization. The authors also raise the possibility that some tRNA transcription takes place in the nucleolus. Hence, the sequestered localization of tRNA genes might antagonize pol II mediated transcription of the nearby genes. Another posttranscriptional modifications of small RNAs, isopentenylation, may also be performed in the nucleolus. One form of the enzyme (Mod5p-II) that catalyses the formation of isopentenyl adenosine is present in the nucleoli of yeast.58 However, its nucleolar localization is not essential for the production of the modified tRNA; therefore, the functional significance of the nucleolar presence of this enzyme is not clear.

Tissue-Specific Expression of Small Nucleolar RNAs Some snoRNAs are uniquely present in nucleoli of certain tissues, but they probably have nothing to do with ribosome biogenesis. Three C/D-box snoRNAs and one H/ACA-box snoRNA that are exclusively expressed in mouse or human brains have been identified by Cavaille et al.59 Unlike all other snoRNAs isolated to date, these show no potential for base pairing with pre-rRNA, even though they reside in nucleoli. One of the C/D box snoRNAs is partially complementary to the serotonin 2C receptor mRNA, suggesting that it plays a role in the processing of this mRNA. It is also curious that the H/ACA box snoRNA is encoded in an intron for the brain-specific serotonin 2C receptor gene. The genes of the brain-specific snoRNAs are subjected to imprinting; i.e., they are expressed from only one chromosome, either the maternal or paternal allele.60 These novel snoRNAs have been implicated in human neurological disorders, including the Prader-Willi and Angelman syndromes. In the Prader-Willi syndrome, three of the novel C/D type snoRNAs are not expressed, which is apparently due to deletion of the whole locus in which the imprinted genes are contained.61 Although their functional targets have not yet been identified, these snoRNAs must be very important to the normal development of the individual. Since the tissue-specific snoRNAs are not complementary to rRNA, they might serve as guides for modification of certain messenger RNAs.

336

The Nucleolus

Yeast as a Model System for Aging A number of studies suggest a linkage between aging and events occurring in the nucleolus, especially in yeast. Most of the work has focused on silencing factors called Sir (silent information regulatory) proteins, which operate in silent mating loci, in subtelomeric regions and in the repeated genes of the rDNA locus. The observation that stimulated interest in using yeast as a model system for studying aging was a mutation in the SIR4 gene that lengthened lifespan in S. cerevisiae.62 The silencing proteins Sir3p and Sir4p relocate from telomeres to nucleoli in the strain carrying this mutation. However, another gene, UTH4, also extends lifespan and it is necessary for the relocation of the Sir proteins to the nucleolus. More importantly, the Sir complex also relocates from telomeres to nucleoli as wild-type cells age. Mutations in another gene, SGS1, cause premature aging in yeast and also induce a redistribution of Sir3p from the telomeres to the nucleolus. Mutations in SGS1, which codes for a helicase, also cause nucleolar fragmentation; this occurs because of an accumulation of extrachromosomal rDNA circles (ERCs). The movement of proteins from telomeres to the nucleolus might be a mechanism for protection against damage caused by formation of ERCs. Defossez et al63 found that mutations in the gene for a nucleolar protein called Fob1p slows the production of the ERCs and also extends the lifespan of yeast. The wild type Fob1p causes unidirectional blocks in rDNA replication forks; these blocks seem to induce chromosomal breaks, which consequently trigger premature aging. Another Sir protein, Sir2p, is a regulator of the silencing of rDNA repeats as well as a sub-telomeric repressor molecule.64 It also forms a complex with Sir3p and Sir4p in the HM loci.65 Sir2p is part of the RENT complex and its location in the nucleolus is dependent on Net1p.43 Accumulated evidence suggests a connection between Sir2p activity in silencing and yeast longevity. For example, mutations in SIR2 that increase the rate of ERC formation reduce the life span.66 In contrast, expression of an additional copy of the SIR2 gene extends the yeast life span. The enhanced level of Sir2p is believed to result in increased rDNA stability and/or silencing, which suppresses the formation of ERCs, thereby slowing down the aging process. A curious phenomenon is that reduced caloric intake extends the life span of yeast and that this effect is dependent on Sir2p.67 We now have a reasonable understanding of how Sir2p combats ageing in yeast, but how caloric intake might regulate this process is puzzling. The story begins with the discovery that Sir2p is a histone deacetylase that is involved in rDNA chromatin silencing (reviewed by Shore68). This is consistent with the idea that histones in the deacetylated state are associated with silenced chromatin. The novelty of the histone deacetylase activity of Sir2p is that it is a NAD-dependent enzyme.69 Initially it was thought that NAD is simply a regulator of the enzyme, but it was later shown to have a direct role in the catalytic mechanism; i.e., the acetyl groups removed from the lysine residues of the histones are transferred to NAD+ to form the product, 1-O-acetyl-ADP-ribose. How does this explain the effect of caloric restriction on the longevity of yeast? It has been proposed that reduced caloric intake either increases the availability of NAD70 or increases the flux of NAD through a salvage pathway.71 Another possibility is that the enzyme is activated by the removal of nicotinamide, which is an inhibitor of Sir2p.72 Koubova and Guarente73 suggest that when glucose levels are lowered there is a shift from fermentation to respiration, which converts more NADH to NAD. In any event, increasing NAD concentration and/or removal of an inhibitor would enhance the activity of the enzyme, thereby increasing the level of gene silencing through histone deacetylation.

Aging in Mammals Do the activities of the yeast Sir complex serve as a good model of aging in vertebrates? Mammalian equivalents to the Sir2 proteins have been isolated (Sir2α and hSir2 in mouse and human, respectively) by Luo et al74 and Vaziri et al75; these also localize to nucleoli. In fact, there are at least seven Sir2 homologs in humans (reviewed by Sinclair71). Koubova and Guarente73 speculate that Sir2 proteins might act in two areas: sensing caloric intake to regulate hormone levels and to

Nontraditional Roles of the Nucleolus

337

slow age-dependent degeneration of organs. The activity of at least one member of the human Sir2 family, SIRT1, seems to be related to the latter process. SIRT1 associates with and deacetylates p53 with a consequent negative regulation of its ability to activate transcription.76 This, in turn, prevents cellular senescence and apoptosis induced by DNA damage or other forms of stress. However, the nucleolus does not seem to have a significant role in the relationship between the Sir2 proteins and aging in mammals. Although small amounts of ERCs can be found in mammalian cells, their abundance does not appear to correlate with aging.71 Thus, the mechanism by which Sir2 relates to aging may be very different in mammals than in yeast. The RecQ family of DNA helicases is associated with cancer and premature aging.77 Several of these helicases, including the SGS1 protein in yeast (see above) and the proteins involved in Bloom and Werner syndromes in humans are located in nucleoli, at least part of the time. These proteins participate in homologous recombinational repair78 and in telomere maintenance.79

Bloom Syndrome Bloom Syndrome is an autosomal recessive condition resulting in genomic instability and leading to increased frequencies of cancer. BLM is the RecQ DNA helicase that is altered in Bloom Syndrome and is normally found in the nuclear domain 10 in the nucleoplasm. However, it shifts to the nucleolus during S phase where it colocalizes with the Werner syndrome protein, WRN.80 Consistent with its proposed role in maintaining the stability of repeated sequence elements, BLM is also present in a subset of telomeres. The DNA helicase activity of BLM is necessary to ensure that replication from telomeres can proceed smoothly subsequent to DNA damage, thereby lowering the occurrence of cancers.81 This activity is also required to reduce recombinatorial events within rDNA repeats, which may lead to loss of chromosomal integrity. Although BLM partially colocalizes with WRN, the two proteins seem to exist in functionally distinct complexes.82 Both of these proteins exhibit the capacity to resolve triple helices, which are distributed throughout the genome and the loss of this activity could lead to genomic instability.

Werner Syndrome

The WRN helicase, associated with premature aging in Werner syndrome (WS),83 has been localized to transcriptionally active nucleoli, suggesting a link between rDNA transcription and the functioning of the nucleoli. The WRN protein accelerates the transcription of ribosomal RNA; consequently, WS fibroblasts (with mutated WRN) have a decreased level of rRNA transcription.84 This effect on rDNA transcription appears to be due to a direct interaction with one of the subunits of RNA polymerase I, RPA40, which co-immunoprecipitates with WRN. It is proposed that the effect of WRN is to unwind not only double-stranded rDNA, but also rRNA-rDNA heteroduplexes. WRN also interacts with DNA polymerase δ and recruits it to the nucleolus, suggesting a role of WRN in DNA replication.85 Finally, there is accelerated methylation of ribosomal RNA genes during cellular senescence of Werner syndrome fibroblasts although this is not directly linked to alterations in the WRN protein.86 The WRN protein localizes to nucleoli in HeLa cells, but translocates to nucleoplasmic foci when cells are treated with agents that cause S-phase arrest such as hydroxyurea.87 These foci coincide with those formed by the replication protein A and they probably represent stalled replication forks. This is consistent with a role for WRN in processing potentially recombinogenic replication forks. WRN also is found in telomeres where it colocalizes with the telomere binding protein TRF2.79 This suggests that the WRN helicase activity participates in telomere maintenance, possibly in concert with other telomere repair enzymes. These studies raise the question of why WRN is found in nucleoli at all. It is interesting that the Werner’s syndrome protein (WRN) is localized to nucleoli in human cells, but the equivalent protein in mouse is not.88 This is apparently due to sequence differences between the human and mouse protein in the C-terminal region that directs it to the nucleolus.89 Furthermore, in human cells the majority of WRN is not localized to the nucleolus.90 More definitive studies are needed to clarify the significance of the nucleolus in the function of this protein.

338

The Nucleolus

Telomerase in the Nucleolus Although there is clear evidence for redistribution of telomeric components to nucleoli as yeast age (see above), the findings in S. cerevisiae may not apply to aging in vertebrates. However, evidence for a normal presence of telomeric components in nucleoli of mammalian cells has been accumulating over the past several years. Mitchell et al91 observed that a small proportion of telomerase RNA localizes to nucleoli of HeLa cells. In addition, Narayanan et al92 showed that fluorescently labeled telomerase RNA also localizes to nucleoli when injected into Xenopus oocytes. These studies led to the suggestion that the nucleolar association of the enzyme is related to the fact that telomerase RNA contains a domain that resembles the H/ACA box of small nucleolar RNAs. It was also hypothesized that the telomerase RNP is partly assembled in the nucleolus. This contention is supported by studies showing that two yeast telomerase components, Est2p (a telomerase reverse transcriptase subunit) and Est1p, accumulate in the nucleolus when overexpressed.93 Interestingly, the complex between Est2p and the TLC1 telomerase RNA is localized to the nucleoplasm; the formation of this complex could trigger its release from the nucleolus. Additional clues about the role of the nucleolus in telomerase function have come from studies on the protein dyskerin, which is mutated in the human disease dyskeratosis congenita (DKC). Telomere shortening is a characteristic of all forms of DKC. Dyskerin associates with snoRNAs of the H/ACA family as well as with telomerase RNA in the nucleolus.94 This protein is a pseudouridine synthase, formerly known as Nap57 in rat95 and homologous to the yeast protein Cbf5p.96 At least three additional proteins, GAR1P, NHP2P and NOP10P are also common to both telomerase and the H/ACA snoRNPs.97 The fact that DKC cells have less telomerase RNA and lower levels of telomerase activity than normal cells supports the idea that dyskerin is important for the assembly and/or processing of telomerase, possibly in the nucleolus. Several possible roles of dyskerin in telomerase biogenesis are discussed by Collins and Mitchell.98 Although it is generally agreed that mutated dyskerin causes premature telomere shortening, there are conflicting reports on the effects on ribosome biogenesis. One study, which utilized DKC cells, indicated that mutations in dyskerin do not seem to affect pre-rRNA transcription, maturation or the proliferative capacity of the cells, although they show increasing levels of apoptosis as related to the number of cell divisions.99 However, Ruggero et al100 found that in cells from hypomorphic Dkc1 mutant mice there was a markedly reduced level of pseudouridylation of ribosomal RNA and an accumulation of preribosomal RNA intermediates. Thus, it is reasonably certain that dyskerin is involved in both ribosome biogenesis and in telomerase assembly. A more remarkable observation is that the nucleolus seems to regulate telomerase activity by controlling its distribution in subnuclear compartments. Wong et al101 showed that a GFP fusion of telomerase is present in nucleoli during G1 phase, but it partially translocates to the nucleoplasm in late S/G2 phase, apparently coinciding with the time of telomere synthesis. However, in tumor and transformed cell lines, there is essentially no telomerase present in nucleoli, which could be a major factor in the immortalization of these cells. The authors suggest that telomerase is normally sequestered in nucleoli to prevent its potential interaction with substrates other than telomeres. In contrast, telomerase reassociates with the nucleolus in response to treatment of the cells with ionizing radiation. This is most likely a protective measure to prevent telomerase from attempting to repair the double-stranded DNA breaks when telomeres are not replicating. What regulates the cycle of nucleolar localization and release remains an important subject for further study.

Concluding Remarks The nontraditional components in the nucleolus may be classified not only by function, but also by our level of comprehension about nature’s rationale for placing them there. In the case of small nuclear- and tRNAs, guide RNAs and enzymes for methylation, pseudouridylation

Nontraditional Roles of the Nucleolus

339

and processing are already abundantly present in the nucleolus; the cell utilizes available machinery to put the finishing touches on these RNAs . However, why some small RNAs spend time in the nucleolus and others don’t remains an open question. Because of its structure the nucleolus would seem to be a reasonable place to sequester key regulatory molecules such as is seen with the regulation of cell division in yeast. In at least one instance, the telomerase RNP, the nontraditional component shares a set of proteins with a nucleolar sub-assembly for modifying preribosomal RNA. An added benefit of the kinship of telomerase with other nucleolar components is its propensity to be sequestered in the nucleolus, which acts as a protective mechanism to avert inappropriate activities of the enzyme. The reasons for the apparent partial assembly of the signal recognition particle in the nucleolus are elusive; however, it is possible that this process takes place in the nucleolus because of presence of numerous chaperone and helicase activities. For some of the nonconventional roles there remains a great deal of uncertainty about whether the nucleolus is really necessary to carry out these functions; e.g., it is unclear whether the nucleolar location essential for regulation of p53 activity or what role the nucleolus really plays in the life cycles of viruses. What the newly discovered roles show is that the nucleolus is not simply an isolated structure, but that it interacts and communicates with the rest of the cell at many different levels. As our understanding of the traditional role of the nucleolus in ribosome biogenesis approaches maturity, the current and developing research tools should foster the rapid growth of the nontraditional nucleolar field out of its early adolescence during the years ahead.

Acknowledgements The author acknowledges the helpful discussions with and critical reading of the manuscript by current and former members of his laboratory: A. Szebeni, N. Huang, S. Negi, M. Wallace, K. Hingorani and M. Dundr.

References 1. 2. 3. 4.

Pederson T. The plurifunctional nucleolus. Nucleic Acids Res 1998; 26(17):3871-3876. Hiscox JA. The nucleolus—a gateway to viral infection? Archives of Virology 2002; 147(6):1077-1089. Kjems J, Askjaer P. Rev protein and its cellular partners. Adv Pharmacol 2000; 48:251-298. Kubota S, Siomi H, Satoh T et al. Functional similarity of HIV-I rev and HTLV-I rex proteins: identification of a new nucleolar-targeting signal in rev protein. Biochem Biophys Res Commun 1989; 162(3):963-970. 5. Malim MH, Bohnlein S, Hauber J et al. Functional dissection of the HIV-1 Rev trans-activator— derivation of a trans-dominant repressor of Rev function. Cell 1989; 58(1):205-214. 6. Ruben S, Perkins A, Purcell R et al. Structural and functional characterization of human immunodeficiency virus tat protein. J Virol 1989; 63(1):1-8. 7. Siomi H, Shida H, Maki M et al. Effects of a highly basic region of human immunodeficiency virus Tat protein on nucleolar localization. J Virol 1990; 64(4):1803-1807. 8. Fankhauser C, Izaurralde E, Adachi Y et al. Specific complex of human immunodeficiency virus type 1 rev and nucleolar B23 proteins: dissociation by the Rev response element. Mol Cell Biol 1991; 11(5):2567-2575. 9. Szebeni A, Olson MO. Nucleolar protein B23 has molecular chaperone activities. Protein Sci 1999; 8(4):905-912. 10. Dundr M, Leno GH, Hammarskjold ML et al. The roles of nucleolar structure and function in the subcellular location of the HIV-1 Rev protein. J Cell Sci 1995; 108(Pt 8):2811-2823. 11. McDonald D, Hope TJ, Parslow TG. Posttranscriptional regulation by the human immunodeficiency virus type 1 Rev and human T-cell leukemia virus type I Rex proteins through a heterologous RNA binding site. J Virol 1992; 66(12):7232-7238. 12. Zolotukhin AS, Felber BK. Nucleoporins Nup98 and Nup214 participate in nuclear export of human immunodeficiency virus type 1 Rev. J Virol 1999; 73(1):120-127. 13. Michienzi A, Cagnon L, Bahner I et al. Ribozyme-mediated inhibition of HIV 1 suggests nucleolar trafficking of HIV-1 RNA. Proc Natl Acad Sci USA 2000; 97(16):8955-8960. 14. Cmarko D, Boe SO, Scassellati C et al. Rev inhibition strongly affects intracellular distribution of human immunodeficiency virus type 1 RNAs. J Virol 2002; 76(20):10473-10484. 15. Michienzi A, Li S, Zaia JA et al. A nucleolar TAR decoy inhibitor of HIV-1 replication. Proc Natl Acad Sci USA 2002; 99(22):14047-14052.

340

The Nucleolus

16. Wei P, Garber ME, Fang SM et al. A novel CDK9-associated C-type cyclin interacts directly with HIV-1 Tat and mediates its high-affinity, loop-specific binding to TAR RNA. Cell 1998; 92(4):451-462. 17. Marcello A, Cinelli RA, Ferrari A et al. Visualization of in vivo direct interaction between HIV-1 TAT and human cyclin T1 in specific subcellular compartments by fluorescence resonance energy transfer. J Biol Chem 2001; 276(42):39220-39225. 18. Hiscox JA, Wurm T, Wilson L et al. The coronavirus infectious bronchitis virus nucleoprotein localizes to the nucleolus. J Virol 2001; 75(1):506-512. 19. Matthews DA. Adenovirus protein V induces redistribution of nucleolin and B23 from nucleolus to cytoplasm. J Virol 2001; 75(2):1031-1038. 20. Yang TH, Tsai WH, Lee YM et al. Purification and characterization of nucleolin and its identification as a transcription repressor. Mol Cell Biol 1994; 14(9):6068-6074. 21. Huang WH, Yung BY, Syu WJ et al. The nucleolar phosphoprotein B23 interacts with hepatitis delta antigens and modulates the hepatitis delta virus RNA replication. J Biol Chem 2001; 276(27):25166-25175. 22. Dumbar TS, Gentry GA, Olson MO. Interaction of nucleolar phosphoprotein B23 with nucleic acids. Biochemistry 1989; 28(24):9495-9501. 23. Yamada H, Jiang YM, Zhu HY et al. Nucleolar localization of the UL3 protein of herpes simplex virus type 2. J Gen Virol 1999; 80(8):2157-2164. 24. Zatsepina O, Braspenning J, Robberson D et al. The human papillomavirus type 16 E7 protein is associated with the nucleolus in mammalian and yeast cells. Oncogene 1997; 14(10):1137-1145. 25. Sharpless NE, DePinho RA. p53: good cop/bad cop. Cell 2002; 110(1):9-12. 26. Zhang YP, Xiong Y. Control of p53 ubiquitination and nuclear export by MDM2 and ARF. Cell Growth & Differentiation 2001; 12(4):175-186. 27. Rizos H, Darmanian AP, Mann GJ et al. Two arginine rich domains in the p14ARF tumour suppressor mediate nucleolar localization. Oncogene 2000; 19(26):2978-2985. 28. Lindström MS, Klangby U, Inoue R et al. Immunolocalization of human p14ARF to the granular component of the interphase nucleolus. Exp Cell Res 2000; 256(2):400-410. 29. Llanos S, Clark PA, Rowe J et al. Stabilization of p53 by p14ARF without relocation of MDM2 to the nucleolus. Nat Cell Biol 2001; 3(5):445-452. 30. David-Pfeuty T, Nouvian-Dooghe Y. Human p14(Arf): an exquisite sensor of morphological changes and of short-lived perturbations in cell cycle and in nucleolar function. Oncogene 2002; 21(44):6779-6790. 31. Lin AW, Lowe SW. Oncogenic ras activates the ARF-p53 pathway to suppress epithelial cell transformation. Proc Natl Acad Sci USA 2001; 98(9):5025-5030. 32. Xirodimas D, Saville MK, Edling C et al. Different effects of p14ARF on the levels of ubiquitinated p53 and Mdm2 in vivo. Oncogene 2001; 20(36):4972-4983. 33. Latonen L, Kurki S, Pitkanen K et al. p53 and MDM2 are regulated by PI-3-kinases on multiple levels under stress induced by UV radiation and proteasome dysfunction. Cellular Signalling 2002; 15(1):95-102. 34. Martelli F, Hamilton T, Silver DP et al. p19ARF targets certain E2F species for degradation. Proc Natl Acad Sci USA 2001; 98(8):4455-4460. 35. Tsai RYL, McKay RDG. A nucleolar mechanism controlling cell proliferation in stem cells and cancer cells. Genes & Development 2002; 16(23):2991-3003. 36. Daniely Y, Dimitrova DD, Borowiec JA. Stress-dependent nucleolin mobilization mediated by p53-nucleolin complex formation. Mol Cell Biol 2002; 22(16):6014-6022. 37. Daniely Y, Borowiec JA. Formation of a complex between nucleolin and replication protein a after cell stress prevents initiation of DNA replication. J Cell Biol 2000; 149(4):799-809. 38. Mao Y, Mehl IR, Muller MT. Subnuclear distribution of topoisomerase I is linked to ongoing transcription and p53 status. Proc Natl Acad Sci USA 2002; 99(3):1235-1240. 39. Bharti AK, Olson MOJ, Kufe DW et al. Identification of a nucleolin binding site in human topoisomerase I. J Biol Chem 1996; 271(4):1993-1997. 40. Chang HY, Yang X, Baltimore D. Dissecting Fas signaling with an altered-specificity death-domain mutant: requirement of FADD binding for apoptosis but not Jun N- terminal kinase activation. Proc Natl Acad Sci USA 1999; 96(4):1252-1256. 41. Lin DY, Shih HM. Essential role of the 58-kDa microspherule protein in the modulation of Daxx-dependent transcriptional repression as revealed by nucleolar sequestration. J Biol Chem 2002; 277(28):25446-25456. 42. Visintin R, Hwang ES, Amon A. Cfi1 prevents premature exit from mitosis by anchoring Cdc14 phosphatase in the nucleolus. Nature 1999; 398(6730):818-823.

Nontraditional Roles of the Nucleolus

341

43. Shou WY, Seol JH, Shevchenko A et al. Exit from mitosis is triggered by Tem1-dependent release of the protein phosphatase Cdc14 from nucleolar RENT complex. Cell 1999; 97(2):233-244. 44. Bardin AJ, Amon A. Men and sin: what’s the difference? Nat Rev Mol Cell Biol 2001; 2(11):815-826. 45. Jacobson MR, Pederson T. Localization of signal recognition particle RNA in the nucleolus of mammalian cells. Proc Natl Acad Sci USA 1998; 95(14):7981-7986. 46. Politz JC, Yarovoi S, Kilroy SM et al. Signal recognition particle components in the nucleolus. Proc Natl Acad Sci USA 2000; 97(1):55-60. 47. Politz JC, Lewandowski LB, Pederson T. Signal recognition particle RNA localization within the nucleolus differs from the classical sites of ribosome synthesis. J Cell Biol 2002; 159(3):411-418. 48. Pederson T, Politz JC. The nucleolus and the four ribonucleoproteins of translation. J Cell Biol 2000; 148(6):1091-1095. 49. Andersen JS, Lyon CE, Fox AH et al. Directed proteomic analysis of the human nucleolus. Curr Biol 2002; 12(1):1-11. 50. Scherl A, Coute Y, Deon C et al. Functional proteomic analysis of human nucleolus. Mol Biol Cell 2002; 13(11):4100-4109. 51. Grosshans H, Deinert K, Hurt E et al. Biogenesis of the signal recognition particle (SRP) involves import of SRP proteins into the nucleolus, assembly with the SRP-RNA, and Xpo1p-mediated export. J Cell Biol 2001; 153(4):745-761. 52. Wolin SL, Matera AG. The trials and travels of tRNA. Genes Dev 1999; 13(1):1-10. 53. Carmo-Fonseca M, Mendes-Soares L, Campos I. To be or not to be in the nucleolus. Nat Cell Biol 2000; 2(6):E107-E112. 54. Bertrand E, Houser-Scott F, Kendall A et al. Nucleolar localization of early tRNA processing. Genes Dev 1998; 12(16):2463-2468. 55. Jarrous N, Wolenski JS, Wesolowski D et al. Localization in the nucleolus and coiled bodies of protein subunits of the ribonucleoprotein ribonuclease P. J Cell Biol 1999; 146(3):559-571. 56. Srisawat C, Houser-Scott F, Bertrand E et al. An active precursor in assembly of yeast nuclear ribonuclease P. RNA 2002; 8(10):1348-1360. 57. Kendall A, Hull MW, Bertrand E et al. A CBF5 mutation that disrupts nucleolar localization of early tRNA biosynthesis in yeast also suppresses tRNA gene-mediated transcriptional silencing. Proc Natl Acad Sci USA 2000; 97(24):13108-13113. 58. Tolerico LH, Benko AL, Aris JP et al. Saccharomyces cerevisiae Mod5p-II contains sequences antagonistic for nuclear and cytosolic locations. Genetics 1999; 151(1):57-75. 59. Cavaillé J, Buiting K, Kiefmann M et al. Identification of brain-specific and imprinted small nucleolar RNA genes exhibiting an unusual genomic organization. Proc Natl Acad Sci USA 2000; 97(26 PT 1):14311-14316. 60. Bachellerie JP, Cavaille J, Huttenhofer A. The expanding snoRNA world. Biochimie 2002; 84(8):775-790. 61. Nicholls RD, Knepper JL. Genome organization, function, and imprinting in Prader-Willi and Angelman syndromes. Annu Rev Genomics Hum Genet 2001; 2:153-175. 62. Guarente L. Link between aging and the nucleolus. Genes Dev 1997; 11(19):2449-2455. 63. Defossez PA, Prusty R, Kaeberlein M et al. Elimination of replication block protein Fob1 extends the life span of yeast mother cells. Mol Cell 1999; 3(4):447-455. 64. Cockell MM, Gasser SM. The nucleolus: Nucleolar space for RENT. Curr Biol 1999; 9(15):R575-R576. 65. Moazed D, Kistler A, Axelrod A et al. Silent information regulator protein complexes in Saccharomyces cerevisiae: a SIR2/SIR4 complex and evidence for a regulatory domain in SIR4 that inhibits its interaction with SIR3. Proc Natl Acad Sci USA 1997; 94(6):2186-2191. 66. Kaeberlein M, McVey M, Guarente L. The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two different mechanisms. Genes Dev 1999; 13(19):2570-2580. 67. Lin SJ, Defossez PA, Guarente L. Requirement of NAD and SIR2 for life-span extension by calorie restriction in Saccharomyces cerevisiae. Science 2000; 289(5487):2126-2128. 68. Shore D. The Sir2 protein family: A novel deacetylase for gene silencing and more. Proc Natl Acad Sci USA 2000; 97(26):14030-14032. 69. Tanner KG, Landry J, Sternglanz R et al. Silent information regulator 2 family of NAD- dependent histone/protein deacetylases generates a unique product, 1-O-acetyl-ADP-ribose. Proc Natl Acad Sci USA 2000; 97(26):14178-14182. 70. Defossez PA, Lin SJ, McNabb DS. Sound silencing: the Sir2 protein and cellular senescence. Bioessays 2001; 23(4):327-332. 71. Sinclair DA. Paradigms and pitfalls of yeast longevity research. Mech Ageing Dev 2002; 123(8):857-867.

342

The Nucleolus

72. Bitterman KJ, Anderson RM, Cohen HY et al. Inhibition of silencing and accelerated aging by nicotinamide, a putative negative regulator of yeast sir2 and human SIRT1. J Biol Chem 2002; 277(47):45099-45107. 73. Koubova J, Guarente L. How does calorie restriction work? Genes Dev 2003; 17(3):313-321. 74. Luo J, Nikolaev AY, Imai S et al. Negative Control of p53 by Sir2alpha Promotes Cell Survival under Stress. Cell 2001; 107(2):137-148. 75. Vaziri H, Dessain SK, Eaton EN et al. hSIR2(SIRT1) Functions as an NAD-Dependent p53 Deacetylase. Cell 2001; 107(2):149-159. 76. Smith JS. Human Sir2 and the “silencing’ of p53 activity. Trends in Cell Biology 2002; 12(9):404-406. 77. Mohaghegh P, Hickson ID. DNA helicase deficiencies associated with cancer predisposition and premature ageing disorders. Hum Mol Genet 2001; 10(7):741-746. 78. Thompson LH, Schild D. Recombinational DNA repair and human disease. Mutat Res 2002; 509(1-2):49-78. 79. Opresko PL, von Kobbe C, Laine JP et al. Telomere-binding protein TRF2 binds to and stimulates the Werner and Bloom syndrome helicases. J Biol Chem 2002; 277(43):41110-41119. 80. Yankiwski V, Marciniak RA, Guarente L et al. Nuclear structure in normal and Bloom syndrome cells. Proc Natl Acad Sci USA 2000; 97(10):5214-5219. 81. Neff NF, Ellis NA, Ye TZ et al. The DNA helicase activity of BLM is necessary for the correction of the genomic instability of bloom syndrome cells. Mol Biol Cell 1999; 10(3):665-676. 82. Brosh RM Jr, Majumdar A, Desai S et al. Unwinding of a DNA triple helix by the Werner and Bloom syndrome helicases. J Biol Chem 2001; 276(5):3024-3030. 83. Yu CE, Oshima J, Wijsman EM et al. Mutations in the consensus helicase domains of the Werner syndrome gene. Werner’s Syndrome Collaborative Group. Am J Hum Genet 1997; 60(2):330-341. 84. Shiratori M, Suzuki T, Itoh C et al. WRN helicase accelerates the transcription of ribosomal RNA as a component of an RNA polymerase I-associated complex. Oncogene 2002; 21(16):2447-2454. 85. Szekely AM, Chen YH, Zhang CY et al. Werner protein recruits DNA polymerase δ to the nucleolus. Proc Natl Acad Sci USA 2000; 97(21):11365-11370. 86. Machwe A, Orren DK, Bohr VA. Accelerated methylation of ribosomal RNA genes during the cellular senescence of Werner syndrome fibroblasts. FASEB J 2000; 14(12):1715-1724. 87. Constantinou A, Tarsounas M, Karow JK et al. Werner’s syndrome protein (WRN) migrates Holliday junctions and co- localizes with RPA upon replication arrest. EMBO Rep 2000; 1(1):80-84. 88. Marciniak RA, Lombard DB, Johnson FB et al. Nucleolar localization of the Werner syndrome protein in human cells. Proc Natl Acad Sci USA 1998; 95(12):6887-6892. 89. Suzuki T, Shiratori M, Furuichi Y et al. Diverged nuclear localization of Werner helicase in human and mouse cells. Oncogene 2001; 20(20):2551-2558. 90. Shiratori M, Sakamoto S, Suzuki N et al. Detection by epitope-defined monoclonal antibodies of Werner DNA helicases in the nucleoplasm and their upregulation by cell transformation and immortalization. J Cell Biol 1999; 144(1):1-9. 91. Mitchell JR, Cheng J, Collins K. A box H/ACA small nucleolar RNA-like domain at the human telomerase RNA 3' end. Mol Cell Biol 1999; 19(1):567-576. 92. Narayanan A, Lukowiak A, Jády BE et al. Nucleolar localization signals of Box H/ACA small nucleolar RNAs. EMBO J 1999; 18(18):5120-5130. 93. Teixeira MT, Forstemann K, Gasser SM et al. Intracellular trafficking of yeast telomerase components. EMBO Reports 2002; 3(7):652-659. 94. Mitchell JR, Wood E, Collins K. A telomerase component is defective in the human disease dyskeratosis congenita. Nature 1999; 402(6761):551-555. 95. Wang C, Query CC, Meier UT. Immunopurified small nucleolar ribonucleoprotein particles pseudouridylate rRNA independently of their association with phosphorylated Nopp140. Mol Cell Biol 2002; 22(24):8457-8466. 96. Kiss T. Small nucleolar RNA-guided post- transcriptional modification of cellular RNAs. EMBO J 2001; 20(14):3617-3622. 97. Mason PJ. Stem cells, telomerase and dyskeratosis congenita. Bioessays 2003; 25(2):126-133. 98. Collins K, Mitchell JR. Telomerase in the human organism. Oncogene 2002; 21(4):564-579. 99. Montanaro L, Chilla A, Trere D et al. Increased mortality rate and not impaired ribosomal biogenesis is responsible for proliferative defect in dyskeratosis congenita cell lines. J Invest Dermatol 2002; 118(1):193-198. 100. Ruggero D, Grisendi S, Piazza F et al. Dyskeratosis congenita and cancer in mice deficient in ribosomal RNA modification. Science 2003; 299(5604):259-262. 101. Wong JM, Kusdra L, Collins K. Subnuclear shuttling of human telomerase induced by transformation and DNA damage. Nat Cell Biol 2002; 4(9):731-736.

Index A

D

Acanthamoeba castellanii 88, 92, 102, 104, 111, 112, 123, 124 Acidosis 18 Actinomycin D (AMD) 17, 25, 33, 34, 44, 79, 291, 306, 307, 310 Adapter protein 270, 286, 291, 293, 297 Aging 8, 67, 87, 237, 329, 331, 336-338 Archaea 223-225, 227, 230, 234, 235, 239, 241, 242, 291 ARF 331, 332 ATPase 212, 213, 242, 258, 264, 265, 266, 274, 288, 308

Dense fibrillar component (DF, DFC) 2, 11-18, 21, 23-26, 30, 31, 43, 51, 90, 107, 154, 155, 157-162, 164, 166, 167, 199, 245, 303, 305, 334 Dexamethasone (DEX) 98, 112, 113 DNA 11, 12, 14-16, 18, 23, 24, 26, 32, 34, 37, 44, 48, 49, 58, 63, 64, 66, 67, 77, 82, 88-91, 94, 96, 98, 99, 101-109, 111, 112, 116-128, 130, 133-135, 137, 142, 144, 160, 170-172, 180, 188, 216, 230, 237, 242, 264, 265, 312, 319, 332-334, 337, 338 DNA methyltransferase 80, 96 Double-stranded RNA (dsRNA) 264 Drosophila melanogaster 128

B Bcl-2 333 Bloom syndrome 337

C Cajal body 7, 36, 174, 178, 223-227, 229, 234, 236, 237, 239, 240, 242-245, 311, 315-317, 319-323, 325, 335 C/D 36, 171, 174, 188, 189, 200, 201, 203, 209, 212, 223-245, 264, 268, 271, 273, 276, 316, 319-322, 325, 335 Chaperone 2, 6, 170, 171, 179, 182, 184, 186, 188, 205, 212, 213, 224, 227, 230, 232, 235, 239, 242, 243, 330, 331, 334, 339 Chromatin 6, 10-13, 15, 18, 22, 24, 26, 31, 43, 46, 48-50, 52, 54, 66, 67, 73-87, 90, 91, 96, 108, 112, 121, 122, 127, 131, 137-139, 142, 144-146, 151-153, 155, 167, 209, 238, 273, 331, 336 condensed 10-13, 15, 18, 24, 26, 46, 48, 50, 52, 167 Core promoter element (CPE) 93-95, 101-104, 106, 118, 125 Coronavirus nucleoprotein (N) 331 CPBF 107, 108, 118 Cyclin-dependent protein kinase (CDK) 5, 41, 54, 109, 333, 334

E 3'-ETS 58, 64, 170, 186, 200, 204, 207, 260 5'-ETS 47, 58, 59, 170, 176, 202, 204, 205 E. coli 2, 60, 61, 63, 85, 132, 138, 141, 189, 232, 234, 292 Endonuclease 73, 74, 78, 207, 211, 240, 274 Exonuclease 74-78, 81, 185, 186, 200, 204, 211, 215, 240, 260, 264, 267, 271, 276 Exosome 186, 204, 211, 212, 214, 215, 240, 260, 267, 293 Export adapter protein 270 Export receptor 6, 214, 262, 276, 286, 290, 291, 293, 296, 334 External transcribed spacer (ETS) 58, 59, 90, 92, 108, 164, 170, 173, 174, 176-182, 184, 186, 188, 200, 202, 204, 205, 207, 260, 274

F Factor C* 98, 99, 112, 122 Fibrillar center (FC) 2, 11-14, 16-18, 21, 23-25, 30-32, 37, 43, 90, 107, 154-162, 164, 166, 199, 303, 305, 334 Fibrillarin 33, 36, 37, 43, 44, 47, 50-52, 90, 171, 201, 227, 234, 238, 240-243, 245, 304, 305 Fluorescence recovery after photobleaching (FRAP) 5, 32-34, 52

344

G Granular component (GC) 2, 11-14, 16-18, 21, 23-25, 31, 43, 61, 63, 68, 90, 92, 106, 154, 158, 160-162, 164, 167, 171, 199, 289, 303, 305, 332, 334 GTPase 212-214, 258, 262, 266, 267, 286, 288-290, 293, 296, 297, 333 Guide snoRNA see Small nucleolar RNA

H H/ACA 171, 200, 201, 203, 209, 211, 212, 223-227, 229-231, 234-244, 268, 273, 316, 319, 320, 335, 338 Helicase 2, 6, 83, 87, 109, 182, 184, 211, 242, 258, 264, 265, 266, 271, 276, 308, 334, 336, 337, 339 Hepatitis delta virus antigen (HDAgs) 331 Herpes simplex virus 331 Histone deacetylase (HDAC) 79-81, 86, 96, 115, 127, 138, 139, 336 HIV 245, 291, 330, 331 HOT1 74, 75, 83, 85, 87 HTA1 144 HTB1 144

I Imp4 family 215 Imprinted snoRNA see Small nucleolar RNA Interferon 12 Intergenic spacer (IGS) 58-60, 63-68, 74, 75, 81, 82, 90-92, 95, 96, 106, 120, 121 Internal transcribed spacer 58, 59, 68, 90, 92, 154, 164, 170, 173, 200, 230, 288 ITS1 58, 60, 68, 170, 184-186, 188, 189, 202, 204-206, 212, 260, 296, 297 ITS2 58, 68, 170, 186, 188, 204, 206, 207, 212, 213, 260, 288 Intranucleolar chromatin 11

K Ku/E1 BF 107, 108, 118

M MAP kinase (MAPK) 107, 116 MDM2 331, 332

The Nucleolus Methylation 2, 6, 50, 58, 67, 80, 86, 144, 171, 174, 200-203, 225, 227, 229, 231, 232, 235, 237, 238, 241, 260, 263, 267, 268, 297, 315, 316, 319, 320, 323, 325, 337, 338 Methyltransferase 80, 96, 144, 212, 263, 267, 268, 276, 320 Mitochondrial RNA-processing (MRP) 164, 185, 188, 189, 204, 205, 211, 212, 224-228, 230, 231, 235, 315, 320 Mitosis 5, 7, 17, 29, 33, 37, 41-44, 46-54, 64, 79, 90, 116, 119, 120, 126, 127, 141-143, 333, 334 Mitotic exit network (MEN) 333, 334 MRP snoRNA see Small nucleolar RNA

N NHPX protein 36 NIPP1 34, 36 Nuclear envelope 22, 23, 25, 31, 43, 44, 50, 90, 141, 142, 146, 274, 276, 290, 291, 294, 302 Nuclear export 6, 35, 213-215, 258, 262, 269, 274, 276, 278, 286, 287, 290, 291, 293, 294, 330, 331, 334 Nuclear export receptor 262, 334 Nuclear export signal (NES) 35, 214, 274, 276, 286, 290-294, 297 Nuclear import 242, 319, 323 Nuclear inhibitor of PP1 34 Nuclear localization sequence (NLS) 35, 291 Nuclear pore complex (NPC) 3, 6, 210, 212-214, 259, 262, 266, 286, 290, 291, 294, 295, 302 Nuclease 66, 74, 77, 78, 83, 86, 90, 120, 121, 184-186, 188, 202, 205, 211, 230, 240, 260, 263, 264, 274 Nucleolar assembly 41-43, 50, 52-54 Nucleolar disassembly 42 Nucleolar localization element (NoLE) 319-323, 325 Nucleolar localization signal (NoLS) 35, 36, 106, 243, 319 Nucleolar protein database 313 Nucleolar ultrastructure 5, 21, 22, 67 Nucleolin 33, 43, 66, 170, 176, 178-180, 188, 304, 305, 331, 333 Nucleolus derived foci (NDF) 46, 50, 53, 54 Nucleolus organizer region (NOR) 1, 4, 5, 10-17, 24, 25, 29, 37, 44, 46, 48, 50, 51, 53, 54, 59, 67, 89, 90, 107, 119, 171 Nucleostemin 333

345

Index

O Oncogene 7, 125, 127, 329, 332

P p16 108 Paraspeckle protein 1 (PSP1) 34, 309, 310, 312 Perinatal asphyxia 18 Pol I 6, 21, 24, 25, 31-33, 36, 37, 41, 42, 44, 46, 49, 50, 52, 54, 89, 93, 100, 126, 128, 130-147, 260, 268 see also RNA polymerase I Pol II 24-26, 54, 55, 130-137, 142-147, 268, 319, 320, 323, 335 see also RNA polymerase II Pol III 24, 128, 131, 132, 135, 136, 140, 260, 334, 335 see also RNA polymerase II Poly-ADP-ribose polymerase (PARP) 16 Polymerase sssociated factors PAF49 100 PAF51 100 PAF53 33, 98, 100, 112, 114, 122 Pore complex 6, 210, 212, 214 Potorous tridactylis 46 Prenucleolar bodies (PNBs) 5, 33, 41, 42, 44, 50-54 Preribosomal particles 66S 259, 260, 268-270, 273, 276 90S 260, 262, 264, 265, 269-271, 273, 274, 276 Pre-ribosomal RNA (Pre-rRNA) 1-3, 5, 6, 21, 33, 36, 37, 41, 43, 46, 50, 52-54, 66, 89, 95, 96, 98, 109, 154-156, 159, 161, 162, 164, 166, 170-174, 176-202, 204, 205, 207, 209-216, 223-227, 230, 232, 233, 235, 239, 259, 260, 262-271, 273, 274, 276, 278, 279, 288, 294, 297, 309, 316, 322, 330, 331, 335, 338, 339 Pre-rRNA processing 6, 33, 164, 166, 199, 201, 202, 204, 207, 211, 214, 216, 260, 262-265, 267-271, 273, 274, 276, 278, 279, 309, 322 Processome 180, 209-212, 215, 297 Promoter 16, 31-33, 36, 58, 59, 63, 64, 66, 67, 73-78, 80-86, 90-96, 99-112, 115, 117, 118, 120-126, 128, 130, 131, 135-137, 139, 143, 145, 246, 268, 335

Protein kinase 5, 107, 109, 116, 125, 127, 140, 141, 209, 270, 291, 333 Protein phoshatase 1 (PP1) 34, 36 Proteomic 4, 7, 34, 36, 37, 188, 258, 259, 262, 263, 271, 276, 279, 286, 297, 302, 303, 308-311, 316 Pseudouridylation 2, 6, 50, 171, 200, 225, 229, 235, 241, 260, 315, 316, 319, 320, 323, 335, 338 Psoralen 73-79, 81, 82, 84, 86, 87, 90, 138, 145, 179, 180

R Rapamycin 140, 141 Replication fork barrier (RFB) 73-75, 77, 80-86, 128, 130 Ribonucleoprotein (RNP) 4, 16, 26, 46, 49, 89, 90, 119, 171, 178, 180, 200, 207, 214, 224, 225, 227, 231-235, 237, 239-242, 258, 259, 264, 271, 273, 276, 290, 295, 315, 320, 323, 325, 334, 338, 339 Ribosomal assembly 271, 273 Ribosomal DNA (rDNA) 1-7, 12, 15, 17, 21, 24, 25, 29, 32, 33, 35-37, 41-44, 46, 48-50, 52, 54, 58-69, 73-96, 98-112, 114-128, 130, 131, 134-147, 154-157, 160, 162, 172, 184, 186, 200, 201, 209, 259, 260, 302, 303, 331, 336, 337 chromatin 6, 49, 67, 73, 78, 79, 82, 85, 87, 91, 131, 137, 139, 142, 144-146, 209, 336 core promoter 102, 108 spacer 64 transcription 6, 7, 17, 24, 25, 29, 33, 36, 41-44, 46, 49, 50, 52, 58, 77, 79, 80, 88-90, 92-96, 98, 100-104, 107-112, 114-119, 121-127, 130, 131, 136-141, 143, 160, 172, 259, 302, 303, 331, 337 Ribosomal protein (r-protein) 2, 3, 6, 21, 26, 31, 33, 35, 41, 54, 86, 88, 91, 92, 115, 120, 140-142, 154, 170, 171, 173, 174, 179, 188, 199, 207, 209, 210, 212-215, 227, 233, 235, 241, 258-260, 262-264, 266, 268, 269, 271, 273, 274, 276, 278, 286, 287, 291, 292, 297, 309, 331, 334 Ribosomal protein-like 214

346 Ribosomal RNA (rRNA) 2-6, 21, 24-27, 2933, 35-37, 41-44, 46, 47, 49, 50, 52-54, 58-64, 66-68, 73-83, 84-92, 95, 96, 99, 101, 102, 104, 108, 109, 111, 115, 116, 118-128, 130, 131, 134, 135, 137-146, 154-157, 159-162, 164, 166, 167, 170174, 176-189, 199-202, 204-207, 209216, 223-240, 245, 246, 258-260, 262271, 273, 274, 276, 278, 279, 286, 288, 290, 291, 293, 294, 296, 297, 309, 316, 320, 322, 330, 331, 335, 337, 338 Ribosomal subunit 1, 2, 3, 6, 31, 33, 41, 42, 54, 90, 92, 111, 171, 173, 199, 213, 214, 232, 233, 235, 258, 259, 263, 274, 279, 286, 287, 294, 295, 309, 329 assembly 259, 260, 262, 263, 265, 266, 268, 271, 273, 276, 279 export 259, 262, 263, 266 Ribosome 1-7, 10-18, 21, 22, 25, 26, 29, 33, 41-43, 54, 59, 61, 63, 69, 79, 80, 86, 88, 90-92, 110, 111, 116, 119, 120, 122, 128, 131, 138, 140, 141, 146, 154, 164, 166, 171, 173, 174, 180, 181, 183, 188, 199-202, 207, 209-216, 232, 233, 237-239, 245, 246, 258-260, 262-267, 269-271, 273, 274, 276, 279, 286-288, 290, 292, 293, 295-297, 302, 303, 315, 329, 331, 334, 335, 338, 339 assembly 1, 2, 4, 7, 12, 33, 128, 141, 146, 173, 188, 207, 216, 232, 233, 262, 263, 279, 286, 290 assembly complex (RAC) 188, 207 export 12, 214 subunit 6, 41, 302, 303 RNA 1-7, 10, 16, 21, 23-27, 29-33, 35-37, 41, 42, 44, 46, 48-50, 52-54, 60, 62, 63, 65, 73, 77-79, 83-92, 95-105, 107-128, 130-135, 138, 141, 142, 154-160, 162, 164, 166, 167, 170, 171, 173, 178, 184, 185-188, 199, 200, 202, 209-212, 215, 223-227, 229-232, 234-246, 258-260, 262-266, 268, 270, 271, 273, 274, 276, 278, 279, 286, 290, 291, 296, 303, 305, 306, 308, 310, 311, 313, 315-317, 319-323, 325, 329-331, 334, 335, 337-339 modification 7

The Nucleolus polymerase I (pol I) 2, 5, 6, 10, 21, 25, 31-33, 36, 37, 42, 44, 49, 50, 52, 73, 85-92, 95-105, 107-128, 131, 154-160, 162, 164, 166, 171, 199, 200, 202, 260, 268, 290, 305, 306, 311, 330, 337 polymerase II (pol II) 6, 23, 32, 86, 88, 91, 92, 95-98, 102, 107-109, 126, 127, 130, 162, 164, 236, 242, 268, 305, 317, 319, 320, 322, 323, 325, 335 polymerase III (pol III) 24, 77, 91, 96, 123, 128, 200, 260, 316, 317, 321, 322, 325, 334 splicing 16, 26, 320, 330 RPA43 33, 132-134 Rrn3 33, 36, 99, 112, 113, 115, 122 rRNA methylation 235

S Saccharomyces carlsbergensis 22 Saccharomyces cerevisiae 22-26, 73-75, 85-87, 92, 101, 121-123, 126, 128, 131, 132, 135, 137, 138, 144, 146, 199-202, 207, 209, 210, 237, 242, 244, 258-260, 264, 286, 311, 333-336, 338 Schizosaccharomyces pombe 22-26, 137, 188, 207, 231, 333, 334 Septation initiation network (SIN) 334 Serine/threonine protein phoshatase 1 33 SET1 144 Signal recognition particle (SRP) 7, 26, 295, 315, 316, 320, 322, 329, 331, 334, 339 Silencing 54, 55, 58, 67, 80, 81, 86, 87, 96, 114, 121, 131, 136, 142-146, 331, 333, 335, 336 Silent information regulatory (Sir) 336 Sir protein 336 SIR2 81, 85, 143, 144, 336 SL1 46, 49, 88, 90, 92, 94, 96, 99, 100, 102-104, 106, 107, 110, 114, 115, 117, 118, 122, 123, 127, 137, 139, 142 Small Cajal bodies (scaRNAs) 223-226, 236, 237, 315, 319, 320, 323 Small nuclear RNA (snRNA) 26, 180, 223-227, 231, 235-238, 240, 242-244, 315-317, 319-323, 325

347

Index Small nucleolar RNA (snoRNA) 2-4, 6, 36, 43, 50, 170-172, 174-189, 200-203, 205, 209, 211, 214, 223-227, 229-232, 234-246, 259, 263-265, 268, 271, 273, 274, 288, 315, 316, 319-323, 325, 331, 335, 338 guide 171, 184, 188, 224, 225, 227, 229, 231, 232, 237, 239, 245, 246, 315, 319, 320 MRP 185, 188, 189, 226, 227, 230 imprinted 237 Spacer 46, 58, 59, 61, 63, 64, 66-68, 73-76, 81, 82, 90-93, 95, 96, 106, 108, 111, 120, 121, 154, 164, 170, 171, 173, 174, 182, 188, 200-202, 204-207, 213, 235, 239, 260, 288 Spliceosomal 308, 315, 316, 317, 319, 320, 322, 323, 325, 329 Spliceosomal RNA (sRNA) 223, 224, 229, 230, 234, 235, 239, 241, 264, 316, 317, 319, 320, 322, 323, 325, 329

T TAF1 108, 109, 126 Tandem affinity purification (TAP) 207, 209-211, 262, 273, 274, 276, 278, 294, 295 Target of rapamycin (TOR) 140, 141 Telomerase 7, 27, 234, 237, 238, 315, 316, 320, 329, 331, 338, 339 Telomerase RNA 234, 237, 238, 315, 316, 320, 338 Terminator 58, 63, 64, 66, 90, 92, 96, 106, 121, 131 TIF-IA 33, 98, 99, 112, 113, 122, 127, 137, 139 TIF-IC 98, 100, 110, 112, 122 Tollervey model 271 Topoisomerase 34, 90, 108, 118, 120, 121, 125-127, 134, 135, 333

Transcription factors 6, 33, 41, 46, 48, 76, 80, 82, 85, 90, 91, 94, 96, 98-102, 105, 107, 108, 110, 111, 113, 118-120, 122-128, 131, 134-137, 139, 141-143, 145, 146, 332 TFIC 98, 99, 112, 113, 115, 122 TFIIB 108, 123 TFIID 92, 101, 103, 108, 109, 114, 123 TFIIE 108 TFIIF 108 TFIIH 88, 108, 109, 126, 136 Tumor suppressor 7, 127, 329, 332, 333

U Upstream 16, 31, 58, 59, 63, 64, 66, 67, 76-78, 81, 83, 93-96, 100-102, 119, 120, 122-125, 127, 128, 130, 131, 135-137, 145, 170, 176, 177, 179, 180, 182, 185, 186, 201, 204-206, 227, 230, 238, 335 Upstream binding factor (UBF) 16, 31, 36, 46, 48, 67, 88, 90, 92, 94, 96, 98, 100-107, 109-111, 114-127, 137, 164 Upstream control element (UCE) 93, 124 Upstream promoter element (UPE) 77, 93-95, 103, 106, 108

V Vascular smooth muscle cell (VSMC) 116

W Werner’s syndrome 35, 67, 337

X Xenopus laevis 37, 86, 95, 104, 120, 121, 175, 177, 319, 320

MOLECULAR BIOLOGY INTELLIGENCE UNIT

Landes Bioscience, a bioscience publisher, is making a transition to the internet as Eurekah.com.

Biotechnology Intelligence Unit Medical Intelligence Unit Molecular Biology Intelligence Unit Neuroscience Intelligence Unit Tissue Engineering Intelligence Unit

OLSON

INTELLIGENCE UNITS

MBIU

The chapters in this book, as well as the chapters of all of the five Intelligence Unit series, are available at our website.

The Nucleolus

9 780306 478734