No title

Current Topics in Membranes and Transport Volume 17 Membrane Lipids of Prokaryotes Advisory Board M . P. Blaustein A...

Author: Author Unknown

6 downloads 709 Views 21MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Current Topics in Membranes and Transport Volume 17 Membrane Lipids of Prokaryotes

Advisory Board

M . P. Blaustein A . Essig

R. K . H . Kinne P. A . Knauf Sir H . L . Kornberg

P. Lauger C. A. Pasternak W . D.Stein W . Stoeckenius K. J . Ullrich

Contributors

John E. Cronmn, Jr. Marina A . Frcludenberg Chris Galanos H o Mtard Goldfine Th( I ina s A . La ng wort hy Volker Lehmann Otto Liideritz Ronald N . McElhanev

Donald L. Melchior Guy Ourisson Shmuel Razin Ernst Th. Rietschel Charles 0. Rock Michel Rohmer Shlomo Rottem Derek H . ShaM,

Current Topics in Membranes and Transport Edited by Felix Bronner

Arnost Kleinzeller

Department of Oral Biology University of Connecticut Health Center Farmington, Connecticut

Department qf Physiology University of Pennsylvania School of Medicine Philadelphia, Pennsylvania

VOLUME 17 Membrane lipids of Prokaryotes

Guest Editors Shrnuel Razin

Shlorno Rottem

Department of Membrane and Ulrrastructure Rearurch The Hebrew, Universit~~-Hadassai Medicul School Jeru.talem Israel

Department of Membrane and Ultrastructure Research The Hebrew University-Hadassah Medicul School Jerubulem, Israel

1982

ACADEMIC PRESS A Subsidiary o f Harcwurt Bract) Jovariovich, Publishers

New York London Paris San Diego San Francisco Sao Paulo Sydney Tokyo Toronto

COPYRIGHT @ 1982, BY A C A D ~ MPRESS, IC INC. ALL RIGHTS RESERVED. N O PART OF THIS PUBLICATION MAY B E REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC O R MKHANICAL, INCLUDING P i i o rocow, HLCORDINC. OR ANY INFORMATION STORAGE AND RErRIEVAL SYSTEM, WITHOUT PERM15SION IN W R l l I N G FROM THE PUBLISHER.

ACADEMIC PRESS, INC. 11 1 Fifth Avenue, N c w York, New York 10003

United Kitrrdom Edition mblished bv ACADEMIC PRESS, INC. (LONDON) LTD. 2 4 / 2 8 Oval Road, London N W 1 I D X

LIBRARY O F CONGRESS CATALOG CARD N U M n E R :

ISBN 0- 12- 15 3 3 17-4 PRINTED IN TIlE UNITED STATES OF AMERICA

82 83 84 85

9 8 7 6 5 4 3 2 1

70 117091

T o the Memory of Yuval Razin

This Page Intentionally Left Blank

Contents List of Contributors, xi Foreword, xiii Preface, xvii Yale Membrane Transport Processes Volumes, xix Contents of Previous Volumes, xxi

Lipids of Prokaryotes-Structure

and Distribution

HOWARD GOLDFINE I. Introduction, 2 11. Structure of the Lipids of Prokaryotes, 2 111. Distribution of Lipids in Prokaryotes, 12 IV. F’rokaryotic Lipids and Phylogeny, 3 1

V. Conclusions, 34 References, 36

Lipids of Bacteria Living in Extreme Environments THOMAS A. LANGWORTHY I . Introduction, 45 Apolar Residues, 49 Ill. Neutral Lipids, 56 IV. Glycolipids, 62 V . Acidic Lipids, 66 VI. Overview, 69 References, 70 11.

Lipopolysaccharides of Gram-Negative Bacteria OTTO LUDERITZ, MARINA A. FREUDENBERG, CHRIS GALANOS, VOLKER LEHMANN, ERNST TH. RIETSCHEL, AND DEREK H. SHAW

I. Introduction, 79 Isolation, Structure, and Biosynthesis of Lipopolysaccharides, 82 I l l . Some Selected Aspects on the Biology of Lipopolysaccharides, 114 IV. Final Remarks, 130 References, 134 11.

vii

viii

CONTENTS

Prokaryotic Polyterpenes: Phylogenetic Precursors of Sterols GUY OURISSON AND MICHEL ROHMER I. Introduction, 154 11. The Sterols of Prokaryotes, 155

H I . The Polyterpenoids of Prokaryotes, 158 IV. The Prokaryotic Polyterpenoids as Phylogenetic Precursors of Sterols, 167 V. Addendum, 177 References, 178

Sterols in Mycoplasma Membranes SHMUEL RAZIN I.

Introduction, 183

11. Cholesterol Uptake, 185 111. Role of Sterols, 191

IV. Conclusions, 200 References, 201

Regulation of Bacterial Membrane Lipid Synthesis CHARLES 0. ROCK AND JOHN E. CRONAN, JR. 1. Introduction, 207 11. Regulation of Membrane Lipid Synthesis, 208 Ill. Conclusions, 226 References, 227

Transbilayer Distribution of Lipids in Microbial Membranes SHLOMO ROTTEM 1. Introduction, 235 11. Assessment of Transbilayer Distribution of Membrane Lipids, 236

111. Transbilayer Distribution of Outer Membrane Lipids, 239 TV. Transbilayer Distribution of Cytoplasmic Membrane Lipids, 244 V. How Lipid Asymmetry Is Maintained, 256 References. 256

Lipid Phase Transitions and Regulation of Membrane Fluidity in Prokaryotes DONALD L. MELCHIOR I. Introduction, 263 11. Lipid Phases, 264 111. Membrane Bilayer Transitions, 267

CONTENTS

IV. Fluidity-Modulating Lipids, 282 V. Patching, 284 VI. Biological Consequences of Membrane State, 292 VII. Biological Control, 299 References. 307

Effects of Membrane Lipids on Transport and Enzymic Activities RONALD N. McELHANEY I. 11. 111. 1V. V. VI.

Introduction, 317 Relevant Properties of Membrane Constituents, 318 Arrhenius Plots of Membrane Transport Systems and Enzymes, 320 Studies of Cells and Membranes, 323 Studies of Isolated Membrane-Bound Enzymes, 362 Conclusions, 369 References. 369

Index. 381

ix


List of Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin. John E. Cronan, Jr., Department of Microbiology. University of Illinois. Urbana, Illinois 61801 (207) Marina A. Freudenberg, Max-Planck-lnstitut fur Immunbiologie, D-78 Freiburg, Federal Republic of Germany (79) Chris Galanos, Max-Planck-Institut fur Immunbiologie, D-78 Freiburg, Federal Republic of Germany (79) Howard Goldfine, Department of Microbiology, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104 ( I ) Thomas A. Langworthy, Department of Microbiology, School of Medicine, University of South Dakota, Vermillion, South Dakota 57069 (45) Volker Lehmann, Max-Planck-lnstitut fur Immunbiologie, D-78 Freiburg, Federal Republic of Germany (79) Otto Luderitz, Max-Planck-Institut fur Immunbiologie. D-78 Freiburg, Federal Republic of Germany (79) Ronald N. McElhaney, Department of Biochemistry, University of Alberta, Edmonton, Alberta, Canada (3 17) Donald L. Melchior, Department of Biochemistry. University of Massachusetts Medical School, Worcester, Massachusetts 01605 (263) Guy Ourisson, Laboratoire de Chimie Organique des Substances Naturelles, Centre de Neurochimie-UniversitC Louis Pasteur, F 67008 Strasbourg, France (153) Shmuel Razin,* Department of Membrane and Ultrastructure Research, The Hebrew University-Hadassah Medical School, Jerusalem, Israel (183) Ernst Th. Rietschel, Max-Planck-lnstitut fur Immunbiologie, D-78 Freiburg, Federal Republic of Germany (79) Charles 0.Rock, Department of Biochemistry, St. Jude Children’s Research Hospital, Memphis, Tennessee 38101 (207)

*Address until September 1 , 1982: Mycoplasma Branch, Bureau of Biologics, Food and Drug Administration, 8800 Rockville Pike, Bethesda, Maryland 20205. xi

xii

LIST OF CONTRIBUTORS

Michel Rohrner, Ecole Nationale Suptrieure de Chimie de Mulhouse, F 68093 Mulhouse, France (153) Shlomo Rottem, Department of Membrane and Ultrastructure Research, The Hebrew University- Hadassah Medical School, Jerusalem, Israel (235) Derek H. Shaw, Northwest Atlantic Fisheries Centre, St. John’s, Newfoundland, Canada

(79)

Foreword Profligacy rather than economy characterizes the design of naturally occurring membrane lipids, or so it would appear from what we know today. If one were to define a structural denominator common to all components of membrane lipid bilayers, would it be more specific than the amphipathic nature or the inherent competence to form closed vesicles? Diversity seems to be the rule. In fact, and this is rarely mentioned, cell membranes containing a single phospholipid species do not seem to exist. It is true that one type usually predominates, e.g., phosphatidylcholine in eukaryotic membranes and phosphatidylethanolamine in bacteria, but they alone seem to be inadequate as matrices for the varied functions membranes are believed to perform. Is it possible that bilayer asymmetry is essential for biological function and that for this reason alone membrane phospholipids are not limited to a single species? Perhaps this intriguing question could be answered if it were possible to create viable bacterial mutants containing phospholipids of a single type. In view of the compositional complexity of natural membrane lipids, it is remarkable that chemically homogeneous liposomes mimic many natural membrane properties, including transport, phase transitions, or effects on membrane-associated enzymes. Clearly there must exist a wide variety of membrane-associated phenomena expressed only in cells which studies with single component model membranes cannot reveal. In the future, attention will have to be paid increasingly to this question, i.e., to the role of the minor membrane phospholipids and their involvement in regiospecific functions. Perhaps bulk phase properties as a function of phospholipid structure have been unduly emphasized. Prokaryotic membrane phospholipids, though by no means simple in composition, are nevertheless much less complex than their eukaryotic counterparts. As for fatty acyl structures, the CI6 and CIS saturated and monounsaturated fatty acids predominate, but with rare exceptions diand polyunsaturated fatty acids are absent. A departure from this general pattern may signify a specialized function. Olefin-derived cyclopropane acids are found in late log and stationary phase Escherichia coli cells. Yet the significance of these branched acids is by no means clear. Escherichia coli mutants lacking the requisite methyl transferase do not seem physiologically impaired. The need for relatively low-melting long-chain acids is met differently by Bacillus species. They produce iso- and anteiso acids xiii

xiv

FOREWORD

even though they have the mechanisms for introducing olefinic bonds into saturated acids. It is also worth noting that some bacteria generate olefink acids by the anaerobic dehydration pathway while others use oxidative desaturation, the universal eukaryotic pathway. A striking departure from the usual fatty acid patterns is found in cells living in extreme environments (temperature, acidity, ionic strength), e.g., extreme halophiles, thermophiles, and also the organisms recently classified as Archaebacteria. Instead of the common fatty acids, they employ as hydrophobic chains phytanyl residues in stable ether linkage. Even more remarkable, in the form of diphytanyl diglycerol tetrdethers they have the proper dimensions and therefore the potential for spanning the membrane bilayer. If they did they would in essence function as lipid monolayers. The exotic structures of the phytanylether lipids appear to represent alternative solutions to membrane rigidity or stability since their presence correlates with the absence of the peptidoglycan cell wall. It has been generally true for phospholipids from all sources that saturated Fdtty acids are esterified at C- I and unsaturated fatty acids at (2-2 of the glycerol moiety. Yet there are exceptions to this rule. Positional inversion in the phospholipids of clostridia has long been known and more recently observed in certain mycoplasmas. Physiological consequences of the nonrandom fatty acyl esterification sites and its inversion have not been recognized and therefore remain unexplained. In bacteria phosphatidylglycerol and phosphatidylethanolamine are the most common phospholipids, whereas phosphatidylcholine is only rarely found, a pattern which distinguishes prokaryotic membranes most strikingly from eukaryotes. Clearly, the bulk and the net charge of the polar head group cannot be trivial but must play a crucial role in the interaction between the cell envelope and the external milieu. If the subject of phospholipid specificity has remained almost entirely unexplored the reason is undoubtedly that it is rarely absolute and difficult to demonstrate. For the futher exploration of this important subject prokaryotes are clearly the cells of choice. Modulation and control of the environment and mutant selection are more readily realized than with eukaryotic cells. Phospholipid biosynthesis is reasonably well understood today, at least the chemistry of the pathways is. The respective enzymology is much less advanced since the component enzymes are membrane-associated and therefore more refractory to purification. For studies of prokaryotic phospholipid biosynthesis. E . coli has for obvious reasons been the organism of choice. However variations from the E . coli pathway are to be expected and have in fact been encountered earlier in clostridia. For mycoplasm a and acholeplasma phospholipid biosynthesis there is at best fragmentary information. Equally or even more uncharted territory is the regulation of phospho-

FOREWORD

xv

lipid biosynthesis in both prokaryotic and eukaryotic cells. It is perhaps not too unreasonable to predict that the control points and the identity of the modifier molecules for the two cell types will be unrelated. Certainly the physical environment and the stimuli to which the respective cells respond have little in common. How little we know in this area is illustrated by the fact that several decades after the discovery of the phenomenon proper, we still do not know how bacterial cells regulate the synthesis of more or less unsaturated phospholipid in response to temperature changes. Regulation may ocur at the stage of unsaturated fatty acid synthesis or glycerophosphate-acyl-CoA transacylation. Conceivably more than one of the component steps is under control and perhaps by the same controlling molecule. For microorganisms a compelling case can be made that phospholipid biosynthesis is coordinated with membrane assembly and macromolecular synthesis. Indeed, substantial evidence exists, at least from in vitro studies. that the magic spot nucleotides (ppGpp) are negative effectors for several of these processes. Sterols are rarely mentioned in conjunction with discussions of prokaryotic lipids, and understandably so. Sterol-producing or -requiring prokaryotes are exceedingly rare, and this fact seemed to support the view that molecules of this type were not invented prior to the appearance of eukaryotic cells along with intracellular membrane-bound organelles. This, as so many generalizations in biology, had to be abandoned even though sterols probably play, whenever they occur, a much more restricted and less specific role in bacteria than they do in higher cells. The formation of the sterol structure in amounts sufficient to affect membrane properties has been observed only in the instance of M r t h y l o c - o c ~ x cups sulatus. But even in this organism the sterol pathway stops short of full development. Equally unique among prokaryotes is the absolute sterol requirement of Mycoplasma species. Studies with these small bacteria have nevertheless provided useful information on sterol structure-function relationships that may be of more general significance even for eukaryotic systems. It has come as somewhat of a surprise that the sterol precursor squalene is quite widely found in prokaryotes including the anaerobic Archaebacteria. Moreover, squalene transformation to pentacyclic triterpenes of the hopane type, traditionally higher plant products, is not uncommon in these organisms. It appears that these early trials of nature to cyclize squalene-without intervention of oxygen-produced molecules that share certain structural and perhaps also functional features with the sterols. Evolutionary "tinkering" with squalene is in fact observable in Methylococurs ccipsulatns, an organism which produces both lanosterol derivatives from squalene epoxide and pentacyclic triterpenes from squalene.

xvi

FOREWORD

During its relatively short history prokaryotic lipid biochemistry has produced a wealth of novel and often unique information. This volume impressively demonstrates the viability and future promise of this field. The discovery of new structures is likely to continue and with less labor than in the past in view of the powerful analytical methods now available. Progress may come more slowly and may be less straightforward in the elucidation of membrane structure-function relationships. Yet this is the area of greatest challenge. Success, whenever it comes, should bring great rewards, including perhaps a better understanding and rationalization of bacterial systematics and phylogeny.

KONRADBLOCH Department of Chemistry Harvard University Cambridge, Massachusetts

Preface The relative simplicity of prokaryotic cells has made them useful in the study of numerous aspects of cellular biology, including membrane structure and function. Moreover, the availability of techniques for genetic manipulation has made possible the controlled alteration of membrane lipids and proteins in ways not yet possible in the case of eukaryotes. A striking example are the studies that provided the first direct demonstration of the bilayer organization of lipids in biological membranes, evidence for which was obtained by changing the fatty acid composition of the plasma membrane of Acholeplasma laidlawii and Escherichia coli. Other studies utilizing prokaryotes have elucidated the physical state and turnover of membrane lipids and their interrelationship with structural and catalytic membrane proteins. Similarly, our understanding of the pivotal role played by cholesterol and congeners in membrane structure and function-a major area of interest for eukaryotic cell membrane research-owes much to the studies of bacterial membranes. Prokaryotic membrane lipid research also has intrinsic interest. Examples are the lipopolysaccharides of gram-negative bacteria, complex molecules that exhibit a wide spectrum of biological properties. The unique lipids found in bacteria that live in extreme environments, such as the Dead Sea, constitute another area of recent research, since they may provide clues to the understanding of how living organisms have adapted to harsh environments. The wide interest in prokaryotic membrane lipids has given rise to many scientific reports and specialized reviews. This volume is the first to have assembled in one source descriptions of the significant advances made in prokaryotic lipid research during the past decade. In addition to providing systematic coverage, we hope the articles in this volume will also give rise to further research. Thus this work will not only serve as a reference source for scholars, teachers, and students, but will stimulate investigators to attempt solving the many problems that remain. The help of expert colleagues was indispensible in collating current knowledge covering such diverse fields as membrane and lipid biochemistry, microbiology, and cell biology. Our special thanks are due to the contributors for their willingness to help make this book a reality.

SHMUEL RAZIN SHLOMO ROTI-EM xvii


Yale Membrane Transport Processes Volumes

Joseph F. Hoffman (ed.). (1978). “Membrane Transport Processes,” Vol. 1. Raven, New York. Daniel C. Tosteson, Yu. A. Ovchinnikov, and Ramon Latorre (eds.). (1978). “Membrane Transport Processes,” Vol. 2. Raven, New York. Charles F. Stevens and Richard W. Tsien (eds.). (1979). “Membrane Transport Processes ,” Vol. 3 : Ion Permeation through Membrane Channels. Raven, New York. Emile L. Boulpaep(ed.). (1980). “Cellular Mechanisms of Renal Tubular Ion Transport”: Volume 13 of Current Topics in Membranes and Transport (F. Bronner and A. Kleinzeller, eds.). Academic Press, New York. William H. Miller (ed.). (1981). “Molecular Mechanisms of Photoreceptor Transduction”: Volume 15 of Current Topics in Membranes and Transport (F. Bronner and A. Kleinzeller, eds.). Academic Press, New York. Clifford L. Slayman (ed.). (1982). “Electrogenic Ion Pumps”: Volume 16 of Current Topics in Membranes and Transport (A. Kleinzeller and F. Bronner, eds.). Academic Press, New York.


Contents of Previous Volumes Volume 1

Volume 3

Some Considerations about the Structure of Cellular Membranes AND MAYNARD M. DEWEY LLOYDBARR The Transport of Sugars across Isolated Bacterial Membranes H. R. KABACK Galactoside Permease of Escherichia coli ADAMKEPES Sulfhydryl Groups in Membrane Structure and Function ASERROTHSTEIN Molecular Architecture of the Mitochondrion DAVIDH. MACLENNAN Author Index-Subject Index

The Na+, K+-ATPase Membrane Transport System: Importance in Cellular Function ARNOLDSCHWARTZ, GEORGEE. LINDENMAYER, AND JULIUSC. ALLEN Biochemical and Clinical Aspects of Sarcoplasmic Reticulum Function ANTHONY MARTONOSI The Role of Periaxonal and Perineuronal Spaces in Modifying Ionic Flow across Neural Membranes W. J. ADELMAN, JR. A N D Y. PALTI Properties of the Isolated Nerve Endings GEORGINA RODRiCUEZ DE LORES ARNAIZ A N D EDUARDO DE ROBERTIS Transport and Discharge of Exportable Proteins in Pancreatic Exocrine Cells: In Vitro Studies J . D. JAMIESON The Movement of Water across Vasopressin-Sensitive Epithelia RICHARD M. HAYS Active Transport of Potassium and Other Alkali Metals by the Isolated Midgut of the Silkworm R. HARVEY AND WILLIAM KARLZERAHN Author Index-Subject Index

Volume 2 The Molecular Basis of Simple Diffusion within Biological Membranes W. R. LIEBA N D W. D. STEIN The Transport of Water in Erythrocytes ROBERTE. FORSTER Ion-Translocation in Energy-Conserving Membrane Systems B. CHANCE A N D M. MONTAL Structure and Biosynthesis of the Membrane Adenosine Triphosphatase of Mitochondria ALEXANDER TZAGOLOFF Mitochondria1 Compartments: A Comparison of Two Models HENRYTEDESCHI Author Index-Subject Index

Volume 4 The Genetic Control of Membrane Transport W. SLAYMAN CAROLYN xxi

xxii Enzymic Hydrolysis of Various Components in Biomembranes and Related Systems MAHENDRA KUMAR JAIN Regulation of Sugar Transport in Eukaryotic Cells HOWAKD E. MORC~AN AND CAROLF. WHITFIELD Secretory Events in Gastric Mucosa RICHARD P. DURBIN Author Index-Subject Index

Volume 5 Cation Transport in Bacteria: K', Na+, and H+ FRANKLIN M. HAROLD AND KARLHEINZ ALTENDORF Pro and Contra Carrier Proteins: Sugar Transport via the Periplasmic GalactoseBinding Protein WINFRIED BOOS Coupling and Energy Transfer in Active Amino Acid Transport EKICH HEINZ The Means of Distinguishing between Hydrogen Secretion and Bicarbonate Reabsorption: Theory and Applications to the Reptilian Bladder and Mammalian Kidney WILLIAM A. BRODSKYA N D THEODORE P. SCHILB Sodium and Chloride Transport across Isolated Rabbit Ileum G. SCHULTZ AND STANLEY PErER F. CURRAN A Macromolecular Approach to Nerve Excitation ICHlJl TASAKl AND EMILIO CARBONE Suhjerr Index

Volume 6 Role of Cholesterol in Biomembranes and Related Systems MAHENDRA KUMAR JAIN Ionic Activities in Cells A. A, LEVA N D W. MCD. ARMSTRONG Active Calcium Transport and Ca*+-Activated ATPase in Human Red Cells H. J . SCHATZMANN The Effect of Insulin on Glucose Transport in Muscle Cells TORBEN CIAUSEN

CONTENTS OF PREVIOUS VOLUMES

Recognition Sites for Material Transport and Information Transfer HAI.VOR N . CHRISTENSEN Subject Index

Volume 7 Ion Transport in Plant Cells E. A. C. MACROBBIE H+ Ion Transport and Energy Transduction in Chloroplasts RICHARDA. DII-LEY AND ROBERT T. GIAQUINTA The Present State of the Carrier Hypothesis P A U IG. LEFEVRE Ion Transport and Short-circuit Technique S. REHM WARREN Sithjecr Index

Volume 8 Chemical and Physical Properties of Myelin Proteins M. A. MOSCARELLO The Distinction between Sequential and Simultaneous Models for Sodium and Potassium Transport P. J. GARRAHAN A N D R. P. GARAY Soluble and Membrane ATPase of Mitochondria, Chloroplasts, and Bacteria: Molecular Structure, Enzymatic Properties, and Functions RIVKAPANETA N D D. RAOSANADI Competition, Saturation, and lnhibitionIonic Interactions Shown by Membrane Ionic Currents in Nerve, Muscle, and Bilayer Systems RoRERr J. FRENCH AND WILLIAM J . A D ~ L M AJ RN. , Properties of the Glucose Transport System in the Renal Brush Border Membrane R. KINNE Suhjert Index

Volume 9 The State of Water and Alkali Cations within the lntracellular Fluids: The Contribution of NMR Spectroscopy ANL) MORDECHAI SHPORER MORTIMER M. CIVAN

xxiii


Electrostatic Potentials at Membrane-Solution Interfaces STUART MCLAUCHIJN A Thermodynamic Treatment of Active Sodium Transport S. ROYCAPLAN A N D A L V I NEssici Anaerobic Electron Transfer and Active Transport in Bacteria WIL N. KONINCS AND JOHANNES BOONSTRA Protein Kinases and Membrane Phosphorylation M. MARLENEH(ISF.YA N r ) MARIANO TAO Mechanism and Physiological Significance of Calcium Transport across Mammalian Mitochondria1 Membranes LRENAMEI.A Thyroidal Regulation of Active Sodium Transport F. ISMAIL-BEIGI Subject Index

Volume 10 Mechanochemical Properties of Membranes E. A. EVANSA N D R. M. HOCHMUTH Receptor-Mediated Protein Transport into Cells. Entry Mechanisms for Toxins, Hormones, Antibodies, Viruses, Lysosomal Hydrolases, Asialoglycoproteins, and Carrier Proteins DAVID M. NEVILLE, JR. A N D TA-MINCHANG The Regulation of Intracellular Calcium ERNESTO CARAFOI.1 A N D MARTIN CROMFTON Calcium Transport and the Properties of a Calcium-Sensitive Potasbium Channel in Red Cell Membranes VIRGIL10 L. L E W A N D HUGOG . FERREIRA Proton-Dependent Solute Transport in Microorganisms A. A. EDDY Subject Index

Volume 11 Cell Surface Glycoproteins: Structure, Biosynthesis, and Biological Functions

The Cell Membrane-A Short Historical Perspective ASERROTHSIEIN The Structure and Biosynthesis of Membrane Glycoproteins J E N N l F E R STURCESS, M A K I OMOSC.AKt.I.1~0,A N D HARRY SCHACHTER Techniques for the Analysis of Membrane Glycoproteins R. L. J U L I A N O Glycoprotein Membrane Enzymes J O H N R. RIORDAN AND GOKDON G. FOKSTNEK Membrane Glycoproteins of Enveloped Viruses RICHAKU C O M P A N S A N I ) MAURICE C. KEMP Erythrocyte Glycoproteins MICHAEL J. A . TA N N ER Biochemical Determinants of Cell Adhesion LLOYDA. CULP Proteolytic Modification of Cell Surface Macromolecules: Mode of Action in Stimulating Cell Growth KENNETH D. NOONAN Glycoprotein Antigens of Murine Lymphocytes MICHELLE LETAKTE Subject Index

w.

Volume 12 Carriers and Membrane Transport Proteins

Isolation of Integral Membrane Proteins and Criteria for Identifying Carrier Proteins MICHAEL J. A. TA N N ER The Carrier Mechanism S. B. HI-ADKY The Light-Driven Proton Pump of Hulobacterium halohium: Mechanism and Function MICHAEL EISENBACH AND S. ROYCAPLAN Erythrocyte Anion Exchange and the Band 3 Protein: Transport Kinetics and Molecular Structure PHILIPA. KNAUF

xxiv The Use of Fusion Methods for the Microinjection of Animal Cells R. G . KULKA A N D A. LOYTER Suhjecf Index

Volume 13 Cellular Mechanisms of Renal Tubular Ion Transport

PART I: ION ACTIVITY AND ELEMENTAL COMPOSITION O F INTRAEPITHELIAL COMPARTMENTS Intracellular pH Regulation WALTER F. BORON Reversal of the pHi-Regulating System in a Snail Neuron R. c . 'rHOMAS How to Make and Use Double-Barreled Ion-Selective Microelectrodes THOMAS ZUETHEN The Direct Measurement of K, CI, Na, and H Ions in Bullfrog Tubule Cells MAMORUFUJIMOTO, KUNIHIKO KOTERA,A N D YUTAKA MATSUMURA lntracellular Potassium Activity Measurements in Single Proximal Tubules of Necturuy Kidney TAKAHIKO KLIHOI'A. BRUCEBIAGI,A N D GERHARD GIEBISCH lntracellular Ion Activity Measurements in Kidney Tubules RAJA N. KHURI lntracellular Chemical Activity of Potassium in Toad Urinary Bladder JOE[. DELONGA N D MORTIMER M. ClVAN Quantitative Determination of Electrolyte Concentrations in Epithelial Tissues by Electron Microprobe Analysis ROGERRICK,ADOLFDORGE, RICHARD BAUER,FRANZ BECK, J U N E MASON, CHRISTIANE ROLOFF, A N D KLAUS THURAU PART 11: PROPERTIES O F INTRAEPITHELIAL MEMBRANE BARRIERS IN THE KIDNEY


Hormonal Modulation of Epithelial Structure JAMESB. WADE Changes in Cell Membrane Surfaces Associated with Alterations of Transepithelial Ion Movement MICHAEL KASHGARIAN The Dimensions of Membrane Barriers in Transepithelial Flow Pathways w. WELLING A N D LARRY DANJ. WELLING Electrical Analysis of Intraepithelial Barriers AND EM[I.EL. BOULPAEP HENRYSACKIN Membrane Selectivity and Ion Activities of Mammalian Tight Epithelia SIMON A . LEWIS,NANCYK. WILLS, A N D DOUGLAS C. EATON Ion Conductances and Electrochemical Potential Differences across Membranes of Gallbladder Epithelium LUISREUSS A Kinetic Model for Ion Fluxes in the Isolated Perfused Tubule BRUCEBIAGI,ERNESTO GONZALEZ, A N D GERHARD GIEBISCH The Effects of Voltage Clamping on Ion Transport Pathways in Tight Epithelia ARTHUR L. F I N N AND PAULA ROGENES Tubular Permeability to Buffer Components as a Determinant of Net H Ion Fluxes G . MALNIC, V . L. COSTASILVA,S. S. CAMPIGI.IA, M. DE MELLOAIRES,A N D G . GIEBISCH Ionic Conductance of the Cell Membranes and Shunts of Necturus Proximal Tubule AND GENJIROKIMURA KENNETH R. SPRING Luminal Sodium Phosphate Cotransport as the Site of Regulation for Tubular Phosphate Reabsorption: Studies with Isolated Membrane Vesicles HEIN] MURER,REINHARD STOLL, CARLAEVERS,ROLFKINNE, AND JEAN-PHILIPPE BONJOUR, HERBERT FLEISCH The Mechanism of Coupling between Glucose Transport and Electrical Potential in the Proximal Tubule: A Study of Potential-

xxv


Dependent Phlorizin Binding to Isolated Renal Microvillus Membranes PETERS. ARONSON Electrogenic and Electroneutral Na Gradient-Dependent Transport Systems in the Renal Brush Border Membrane Vesicle BERTRAM SACKTOR

Volume 14 Carriers and Membrane Transport Proteins

Interface between Two Immiscible Liquids as a Tool for Studying Membrane Enzyme Systems L. I. BOGUSLAVSKY Criteria for the Reconstitution of Ion PART 111: INTRAMEMBRANE Transport Systems CARRIERS AND ENZYMES IN AND ADILE. SHAMOO TRA NSEPITHELI AL TRANSPORT WILLIAM F. TIVOL The Role of Lipids in the Functioning of a Membrane Protein: The Sarcoplasmic ReSodium Cotransport Systems in the Proxiticulum Calcium Pump mal Tubule: Current Developments J . P. BENNETT, K. A. MCCILL,A N D R. K I N N EM. . BARAC,A N D H. M U R E R G. B. WARREN ATPases and Salt Transport in the Kidney The Asymmetry of the Hexose Transfer Tubule System in the Human Red Cell Membrane DE LA MARGARITA PEREZ-GONZALEZ W. F. WIDDAS MANNA,FULGENCIO PROVERRIO, AND Permeation of Nucleosides, Nucleic Acid GUILLERMO WHITEMBURY Bases, and Nucleotides in Animal Cells Further Studies on the Potential Role of an PETERG. W. PLAGEMANN AND Anion-Stimulated Mg-ATPase in Rat ProxROBERT M. WOHLHUETER imal Tubule Proton Transport Transmembrane Transport of Small E. KINNE-SAFFRAN A N D R. K I N N E Peptides Renal Na+- K+-ATPase: Localization and D. M. MATTHEWS AND J. W. PAYNE Quantitation by Means of Its K'-Depenof Epithelial Transport in Characteristics dent Phosphatase Activity Insect Malpighian Tubules REINIER BEEUWKES I11 A N D S. H . P. MADDRELL SEYMOUR ROSEN Subject Index Relationship between Localization of N+K+-ATPase, Cellular Fine Structure, and Volume 15 Reabsorptive and Secretory Electrolyte Transport Molecular Mechanisms of Photoreceptor STEPHEN A. EKNST, Transduction AND CLARAV. RIDDLE, KARLJ . KARNAKY, JR. Relevance of the Distribution of Na+ Pump PART I: T H E ROD PHYSIOLOGICAL RESPONSE Sites to Models of Fluid Transport across Epithelia The Photocurrent and Dark Current of JOHNW. MILLSA N D Retinal Rods DONALD R. DIBONA G. MATTHEWS A N D D. A. BAYLOR Cyclic AMP in Regulation of Renal TransSpread of Excitation and Background Adport: Some Basic Unsolved Questions aptation in the Rod Outer Segment THOMAS P. DOUSA K.-W. YAU,T . D. LAMB,A N D Distribution of Adenylate Cyclase Activity P. A. MCNAUGHTON in the Nephron F. MOREL,D. CHABARDES, Ionic Studies of Vertebrate Rods W. GEOFFREY OWENA N D A N D M. 1MBER.r-TEBOUL Subject Index VINCENT TORRE

xxvi Photoreceptor Coupling: Its Mechanism and Consequences GEOFFREY H. GOLD PART 11: THE CYCLIC NUCLEOTIDE ENZYMATIC CASCADE AND CALCIUM ION First Stage of Amplification in the CyclicNucleotide Cascade of Vision JAMESB. HURLEY, LUBERTSTRYER, A N D BERNARD K.-K. FUNC Rod Guanylate Cyclase Located in Axonemes FLEISCHMAN DARRELL Light Control of Cyclic-Nucleotide Concentration in the Retina THOMAS G. EBREY,PAUL KII.BRIDE, JAMES B. HURLEY. ROGERCALHOON, A N D MUIOYUKI TSUDA Cyclic-GMP Phosphodiesterase and Calmodulin in Early-Onset Inherited Retinal Degenerations G. J . CHADEK, Y. P. Liu, R. T. FLETCHER, G. AGUIRRE. R. SANTOS-ANDERSON, A N D M. T'so Control of Rod Disk Membrane Phosphodiesterase and a Model for Visual Transduction P. A. LIEBMAN ANI) E. N. PUGH,JR. lnteractions of Rod Cell Proteins with the Disk Membrane: Influence of Light, Ionic Strength, and Nucleotides HERMANN KUHN Biochemical Pathways Regulating Transduction in Frog Photoreceptor Membranes M. DERICBOWNDS The Use of Incubated Retinas in Investigating the Effects of Calcium and Other Ions on Cyclic-Nucleotide Levels in Photoreceptors I. COHEN ADOLPH Cyclic AMP: Enrichment in Retinal Cones DEBORA B. FARBER Cyclic-Nucleotide Metabolism in Vertebrate Photoreceptors: A Remarkable Analogy and an Unraveling Enigma G . L. WHEELER, M. W . BITENSKY, A. YAMAZAKI.M. M. RASENICK, AND P. J. STEIN


Guanosine Nucleotide Metabolism in the Bovine Rod Outer Segment: Distribution of Enzymes and a Role of GTP HITOSHISHICHI Calcium Tracer Exchange in the Rods of Excised Retinas ETE Z. Szurs The Regulation of Calcium in the Intact Retinal Rod: A Study of Light-Induced Calcium Release by the Outer Segment H. GOLD A N D GEOFFREY J U A N I . KORENBKOT Modulation of Sodium Conductance in Photoreceptor Membranes by Calcium Ions and cGMP ROwxr T. SOKBI PART 111: CALCIUM, CYCLIC NUCLEOTIDES, AND THE MEMBRANE POTENTIAL Calcium and the Mechanism of Light Adaptation in Rods BRUCEL. BASIIAN AND GORDON L. FAIN Effects of Cyclic Nucleotides and Calcium Ions on Bufo Rods JOEL E. BROWN AND GERALDINE WALOGA The Relation between Ca2+and Cyclic GMP in Rod Photoreceptors STUART A . LIETON A N D JOHNE. DOWLING Limits on the Role of Rhodopsin and cGMP in the Functioning of the Vertebrate Photoreceptor SANFORD E. OSTROY, EDWARD P. MEYERTHOI.EN, PETERJ. STEIN, ROBERTA A . SVOBODA. A N D MEECAN J . WILSON [Ca2+],Modulation of Membrane Sodium Conductance in Rod Outer Segments BURKSOAKLEY I1 A N D L A W R ~ NH. C EPINTO Cyclic-GMP-Induced Depolarization and Increased Response Latency of Rods: Antagonism by Light H . MILLERA N D WII.LIAM GRANTD. NICOL

xxvii


PART IV: AN EDITORIAL OVERVIEW Ca2+and cGMP WII I IAM H.

PART 111. REVERSIBILITY: ATP SYNTHESIS DRIVEN BY ELECTRIC FIELDS

MII.I.ER

Index

Volume 16

Electrogenic Ion Pumps PART I . DEMONSTRATION O F PUMP ELECTROGENICITY IN EUKARYOTIC CELLS Electrophysiology of the Sodium Pump in a Snail Neuron R. C . THOMAS Hyperpolarization of Frog Skeletal Muscle Fibers and of Canine Purkinje Fibers during Enhanced Na+-K+ Exchange: Extracellular K+ Depletion or Increased Pump Current? DAVID C. GADSBY The Electrogenic Pump in the Plasma Membrane of Nircllo ROGERM. SPANSWICK Control of Electrogenesis by ATP. Mg2+. H+, and Light in Perfused Cells of Clzuru MASASHITAZAWA AND TFRUO SHIMMF.N PART 11. THE EVIDENCE IN EPITHELIAL MEMBRANES An Electrogenic Sodium Pump in a Mammalian Tight Epithelium s. A. LEWIS A N D N . K. WILLS A Coupled Electrogenic Na+- K+ Pump for Mediating Transepithelial Sodium Transport in Frog Skin ROBIRT N I E I . S ~ N Transepithelial Potassium Transport in Insect Midgut by an Electrogenic Alkali Metal Ion Pump M I C H AG ~ L. WOLFF.RSBERG~R, WILLIAM R. HARVFY, AND MOIRACIOFFI The ATP-Dependent Component of Gastric Acid Secretion G. SACHS.B. WALLMARK. G . SACCOMANI. E. RABON, H. B. STEWART. D. R. DIBONA, AND T. BFRCLINDH

Effect of Electrochemical Gradients on Active H+ Transport in an Epithelium QAIS Ai.-AwQAri A N D TROY E. DIXON Coupling between H+ Entry and ATP Synthesis in Bacteria PI;.IERC. MALONEY Net ATP Synthesis by H+-ATPase Reconstituted into Liposornes YASUO KAGAWA Phosphorylation in Chloroplasts: ATP Synthesis Driven by A+ and by ApH of Artificial or Light-Generated Origin PETFRGRABFR PART 1V. SOME THEORETICAL QUESTIONS Response of the Proton Motive Force to the Pulse of an Electrogenic Proton Pump ERICHH E I N Z Reaction Kinetic Analysis of CurrentVoltage Relationships for Electrogenic Pumps in Neurosporu and A~~~tuhirluritr DETRIC‘HGRAUMANN, ULF-PETER HANSEN. AND CLIFFORD L. S L A Y M A N Some Physics of Ion Transport HAKOI.D J. M O R O W I ~ ~ PART V. MOLECULAR MECHANISMS OF CHARGE SEPARATION An H+-ATP Synthetase: A Substrate Translocation Concept I . A. Kozi.ov A N D V. P. SKULACHEV Proton Translocation by Cytochrome Oxidase MARI‘ENWIKSTROM Electrogenic Reactions of the Photochemical Reaction Center and the UbiquinoneCytochrome h / c pOxidoreductase P. LESLIEDUTI’ON,PAULMUELLER, DANIEL. P. O’KEEFE, NICELK. PACKHAM, ROGERC . PRINCE, AND DAVID M. TIEDE

xxviii Proton-Membrane Interactions in Chloroplast Bioenergetics R. A. DILLEY,L. J . PROCHASKA, G. M. BAKER.N . E. TANDY, A N D P. A. MILLNER Photochemical Charge Separation and Active Transport in the Purple Membrane BARRY HONlC Mitochondria1 Transhydrogenase: General Principles of Functioning 1. A. KOZLOV Membrane Vesicles, Electrochemical Ion Gradients, and Active Transport H. R. KABACK


PART VI. BIOLOGICAL SIGNIFICANCE OF ELECTROGENIC ION PUMPS The Role of Electrogenic Proton Translocation in Mitochondrial Oxidative Phosphorylation JANNA P. WEHRI.F. Electrogenic Reactions and Proton Purnping in Green Plant Photosynthesis WOIKANGJUNGE The Role of the Electrogenic Sodium Pump in Controlling Excitability in Nerve and Cardiac Fibers MARIOVASSALLE Pumps and Currents: A Biological Perspective FRANKLIN M. HAROI.D Index

.

CURRENT TOPICS IN MEMBRANES A N D TRANSPORT VOLUME 17

Lipids of Prokaryotes-Structure Distribution

and

HOWARD GOLDFINE Department of’ Microbinlog! School of Medicine University of Pennsylvuniu Philadelphia. Pennsyivuniu

I . tntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II . Structure of the Lipids of Prokaryotes . . . . . . . . . . . . . . . . . . . A . The Apolar Chains . . . . . . . . . . . . . . . . . . . . . . . . . B . Polar Lipids with a 1 . 2-Diradyl sn-Glycerol Backbone . . . . . . . . . . C . Other Polar Lipids . . . . . . . . . . . . . . . . . . . . . . . . . D . Nonpolar Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . E . Nonextractable Lipids . . . . . . . . . . . . . . . . . . . . . . . . Ill . Distribution of Lipids in Prokaryotes . . . . . . . . . . . . . . . . . . . . A . Cyanobacteria (Blue-Green Algae) . . . . . . . . . . . . . . . . . . . B . Phototrophic Bacteria . . . . . . . . . . . . . . . . . . . . . . . . C . The Gliding Bacteria and the Sheathed Bacteria . . . . . . . . . . . . . D . Budding and/or Appendaged Bacteria . . . . . . . . . . . . . . . . . E . The Spirochetes . . . . . . . . . . . . . . . . . . . . . . . . . . F. Spiral and Curved Bacteria . . . . . . . . . . . . . . . . . . . . . . G . Gram-Negative Aerobic Rods and Cocci . . . . . . . . . . . . . . . . H . Gram-Negative Facultatively Anaerobic Rods . . . . . . . . . . . . . . I . Gram-Negative Anaerobic Bacteria . . . . . . . . . . . . . . . . . . J . Gram-Negative Cocci and Coccobacilli . . . . . . . . . . . . . . . . . K . Gram-Negative Anaerobic Cocci . . . . . . . . . . . . . . . . . . . L . Gram-Negative Chemolithotrophic Bacteria . . . . . . . . . . . . . . . M . Methane-Producing Bacteria . . . . . . . . . . . . . . . . . . . . . N . Gram-Positive Cocci . . . . . . . . . . . . . . . . . . . . . . . . 0 . Endospore-Forming Bacteria . . . . . . . . . . . . . . . . . . . . . P . Gram-Positive, Non-spore-Forming Rods . . . . . . . . . . . . . . . . Q . Actinomycetes and Related Organisms . . . . . . . . . . . . . . . . . R . Rickettsia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S . Mycoplasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1V . Prokaryotic Lipids and Phylogeny . . . . . . . . . . . . . . . . . . . . . V . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 2 2 5 9 II 12 12 14 14 16

16 17 18 18 20 21 23 23 23 24 25 27 28 28 31 31 31 34 36

1 Copyright 0 1982 by Academic Press. lnc . All rights of reproduction In any form reserved ISBN 0-12-153317-4

2

HOWARD GOLDFINE

1.

INTRODUCTION

As background for the contributions to follow, this article will review the structures of the extractable lipids of prokaryotes and the distribution of these lipids in various groups of organisms. It follows the arrangement of the prokaryotes in the eighth edition of Bergey's Munual ofDeterminative Bucteriology. Since Bergey's Munual is arranged along pragmatic lines, and is not generally concerned with phylogenetic relationships among prokaryotes, an overview of the relationships of lipid compositions to recent synthesis of phylogeny will complete this article. The last two decades have revealed a wealth of information on the membranes of bacteria. The diversity of lipids seen earlier has continued to expand and yet some semblance of order is now becoming clearer. In the last decade the emphasis has shifted from descriptive to biophysical and functional aspects, but many gaps remain in our knowledge of the distribution of lipids in bacteria. The recent description of a new kingdom of prokaryotes, the Archaebacteriae, is based in part on a realization of the uniqueness of their membrane lipids (Fox et al., 1980; Langworthy, this volume). Although bacterial taxonomists are just beginning to use lipids as an aid in classification, it will hopefully be apparent to the reader that these membrane components provide a useful set of characteristics (Shaw, 1974; Lechevalier, 1977). The recent shift of emphasis among microbiologists from research on prokaryotes to eukaryotic unicellular species and to animal and plant cells in culture has lessened the intensity of work in this area. The even more recent rush toward molecular genetics and genetic engineering may perhaps serve to remind us of the importance of our prokaryotic roots.

II. STRUCTURE OF THE LIPIDS OF PROKARYOTES A. The Apolar Chains Since the structural organization of biological membranes depends on the presence of molecules containing from one to several nonpolar chains, linked either directly or indirectly to polar moieties, a description of the apolar moieties provides a useful starting point. Indeed, in the membranes of prokaryotes, as in those of higher organisms, the presence of molecules capable of hydrophobic associations is the essence of these biological structures.

PROKARYOTE LIPIDS: STRUCTURE AND DISTRIBUTION

3

I . FATTYACIDS

The most common apolar structures in prokaryotes are fatty acids, which may be linked to glycerol (most commonly as srz-l,2-di-O-acyl residues), to sugars, to amino acids, and rarely to the amino group of sphingosine. These fatty acids are usually 10 to 20 carbon atoms long with 15 to I8 carbon chains predominating. The major types found are straight chains, which may be even or odd numbered, saturated or monounsaturated; branched chains, predominantly of the isn and arzteiso types; and cyclopropane fatty acids. Examples of these fatty acid types are given in Table I. Although the table lists both oleic and cis-vaccenic acids as examples of 18:l found in prokaryotes, it should be noted that cisvaccenic is more common. This is a consequence of the predominant route o f synthesis of the monounsaturated fatty acids in bacteria, which introduces a double bond during the process of chain elongation, rather than by desaturation, usually between C-9 and C- 10, of preexisting long-chain fatty acids (see Rock and Cronan, this volume). Although polyunsaturated fatty acids occur in bacteria, they are rare; among the exceptions are 5,IO-hexadecadienoic acid in Bacillus /ichen/fnrnris (Fulco, 1974), and several methylene-interrupted polyunsaturated fatty acids in Flexihacter po/ymorphus (Johns and Perry, 1977). The blue-green algae are considered to be transitional between the bacteria and the higher protists such as algae and fungi, because some of them possess the ability to desaturate oleate to give di- and triunsaturated fatty acids (Kenyon et a / . , 1972, Kenyon, 1972). The iso- and witeiso-branched-chain fatty acids are found in many grampositive organisms, for example, Micrococcaceae, Bacillaceae, Corynebacteriurn species, and PropioNihucteriurn species. In these organisms they are often the predominant type of acyl chains. In gram-negative bacteria they are found among scattered groups of organisms (Shaw, 1974; Lechevalier, 1977). Similarly, the cyclopropane fatty acids, which are derived by C-l addition to monounsaturated fatty acyl chains (see Rock and Cronan, this volume), are widely distributed in both gram-negative and gram-positive organisms. The P-hydroxy fatty acids are important constituents of lipopolysaccharides (Luderitz ef al., this volume) and are not usually found in the phosopholipids and glycolipids of the bacterial cell membrane. 2. ALK-1-ENYL CHAINS Alk-1-enyl chains are present in phospholipids of the 1 -0-alk-1 ‘-enyl-2-Oacyl-sn-glycerol-3-P type. These are historically referred to as plasmalogens, since they yield a long-chain fatty aldehyde on acid hydrolysis (see Section II,B, I ) . Although the position on glycerol of the alkenyl moiety has been established in the plasmalogens of animal tissues (Hanahan, 1972), the position in bacteria, with one exception (Hagen and Goldfine, 1967), has not been studied. In general the alk-l-

4

HOWARD GOLDFINE

TABLE I SOMECOMMON FATTYACIDSO F PROKARYOTES Saturated n=

CH,-(CH,)n

COOH Common name

12

I2:O" I4:O

14 16

16:O 18:O

10

Monounsaturated m=

7 9

I

CE1:XCH,)n-CH=CH-(CH2)rn n= 5 5 7

Lauric acid Myristic acid Palmitic acid Stearic acid COOH

cis-9- 16 I cis- 1 1- 18: I cis-9- 18: 1

Palmitoleic acid cis-Vaccenic acid Oleic acid

Branched IS0

CH,- C H - (C H2)T,C OOH

CH,- CH,-

CH- (CH, ), C OOH

Cyciapropane I

n= 5

9

5

m=

Hydroxy a-

17:cycl"." I9:cycl

Lactobacillic acid

OH C H,-

I

(C H, )71-C H-C OOH

OH

8-

CH,-

n = 8

3-OH-

I (CH, ),,-CH-CH,-C

14 :O

OOH

P-Hydroxymyristic acid

Shonhand designations; alternative shorthand designations are given in the form C ,?:,, or C ,?. Shorthand designations; alternative shorthand designations are given in the form C,,:, or 16: 1 A 9 , and so on. Shorthand designations; alternative shorthand designations are 17:cy or I7:cyc. " Although no common name was suggested by the workers who originally described this fatty acid, the name colibacillic acid has been suggested in view of its initial discovery in E . coli (J. Asselineau, T. Kaneshiro, and W. M. O'Leary, personal correspondence). 'I

5


enyl chains have the same structures as the acyl chains of the organisms in which plasmalogens have been analyzed. Thus saturated, monounsaturated, cyclopropane, and iso- and anteiso-branched alk-I -enyl chains have been described (Goldfine and Hagen, 1972; Verkley et u l . , 1975).

3. O-Ai.wL CHAINS Extreme halophiles contain di-0-alkyl phospholipids with polyisopranoid side-chains, which are discussed more fully in the chapter by Langworthy , this volume. Alkyl glycerolipids have also been detected in very small amounts in the anaerobic bacteria that have plasmalogens (Kim et al., 1970; Hagen and Blank, 1970). The alkyl chains have only been examined in two species, and although they bear some qualitative resemblence to the acyl and alk-I-enyl chains, there are quantitative differences (Hagen and Blank, 1970; Kamio ef al., 1969). Tetraethers in which long-chain polyisopranoids are linked at both ends to glycerol or longer chain polyols have recently been found in methanogens and thermoacidophiles. These are described more fully by Langworthy (this volume).

6. Polar Lipids with a 1,P-Diradyl sn-Glycerol Backbone

1 . PHOSPHOLIPIDS The most widely occurring lipids in prokaryotes, as in eukaryotes, are phospholipids of the 1,2-diradyl-sn-glycerol-3-P type. The stereochemistry of these lipids is shown in Fig. 1 . The polar groups linked to phosphate are listed in Table 11. The various classes of diacylphosphatides are by no means equally common. The most widely distributed are phosphatidylglycerol, phosphatidylethanolamine, and cardiolipin. The products of N-methylation of phosphatidylethanolamine-

phosphatidyl-N-methylethanolamine, phosphatidyl-N,N‘-dimethylethanolamine, and phosphatidylcholine-are found mostly in gram-negative species. All of the 0 II

CHZO-C-R, 1 I O

R2C0

-

CH,O-C= O I1 R,CO+

I H CH, O P -X I

0

I H

C-R, I H

I 4

I::

H

CH,OP-X I

0

FIG. I . Structures of a 1 ,Z-diacyl glycerophospholipid (left) and a I-O-alk-l’-enyl-2-acyl glycerophospholipid (i.e., plasmalogen-right). See Table I1 for polar substituents.

6

HOWARD GOLDFINE

Poi

AR

TABLE I1 SUBSTITUENTS O F PROKARYOTIC PHOSPHOLIPIDS" Name of intact phospholipid

-OH + -O-CH,CH,YH, -O-CH,CH,YH(CH:,) --O-CH,CH,N(CH,), --O-CH,CH,N(CH:,):+ o -CH,CH -COO-

Phosphatidic acid Phosphdtidy lethanolamine Phosphatidyl-N-methylethanolamine Phosphatidy I-N.N'-dimethylethanolamine Phosphatidylcholine (lecithin) Phosphdtidykerine

I

NH3

-0 -CH,CHOHCH,OH

Phosphatid ylglycerol

-0-CH,CHOH-CH,

0 -Aminoacyl phosphatidylglycerol

t

i

F?

R-CH-C=O

-0-CH,CHOH-CH,OPO~-O

I R,OCH

Diphosphatidylglycerol (cardiolipin)

I

R,OCH,

I -O-~H-(CHOH)~~HOH

Phosphatidylinositol

CH, €I I I 0-CH-C-CH, I OH

Phosphatidylbutane-2.3-diol

'

Sec Fig. I for complete structures.

methylated ethanolamines are relatively rare, hence their occurrence has taxonomic and probably evolutionary significance. The 0-aminoacyl phosphatidylglycerols (Table 11) are found only in prokaryotes and are more common in gram-positive bacteria. Although they have been reported in a few gram-negative species, there i s still some doubt concerning the identity of these lipids in gram-negative organisms. The most common amino acid in 0-aminoacyl phosphatidylglycerols is lysine, but the occurrence of alanine and ornithine has also been recorded. An unusual lipid containing a glucosaminyl moiety rather than an amino acid has been found in Bacillus inegateriurn and Pseudomonas ovalis Chester (see Shaw, 1975). A butane-2,3-diol analogue of phosphatidylglycerol was found as a major phospholipid of Actinomyces (Streptomyces) olivaceus (see Batrakov and Bergelson, 1978). Other reports of phospholipids with this structure have not appeared.


7

Phosphatidic acid is a biosynthetic intermediate i n the pathways leading to the major phospholipids of bacteria (see Rock and Cronan, this volume), and, as such, is generally found only in trace amounts. Phosphatidylserine is also a biosynthetic intermediate in the pathway leading to phosphatidylethanolamine (see Rock and Cronan, this volume) and is generally not a major lipid of bacterial membranes. However, certain organisms have substantial amounts of phosphatidylserine. For example, van Golde et ul. (1975) have found relatively large amounts of this lipid in anaerobes that ferment lactic acid. Phosphatidylinositol is rarely found in gram-negative bacteria, uncommonly in gram-positive organisms, and most commonly in bacteria related to the actinomycetes, such as Arthrobacter and Corynebacterium, as well as in the actinomycetes themselves (Lechevalier, 1977). In these groups of bacteria, the major forms of inositide are members of a family of phosphatidylinositol mannosides, which may contain from one to five mannose units, the most common being the dimannoside. The first mannose unit is glycosidically linked to the C-2 of the inositol ring, and additional mannose units are added sequentially to the hydroxyl at C-6. There may be more than two acyl residues, but the locations of these are not known (Shaw, 1975). There are many reports of the presence of small amounts of monoacylated glycerophosphatides (lysophosphatides) in bacteria. Since these types of lipids may have a destabilizing effect on bilayers, it has generally been thought that their presence is artifactual and that they probably arise by degradation of the diacyl phospholipids. However, it is known that bacteria contain enzymes capable of forming lysophosphatides, and it is possible that small amounts of these lipids are naturally present and may in fact play important roles in the dynamics of bacterial membranes. Plasmalogens (Fig. 1 ) have been found in prokaryotes only in anaerobic species. These phospholipids are acid labile, readily yielding a long-chain fatty aldehyde on exposure to acid. In alkali, the alk-1-enyl ether is stable, but the acyl chain is cleaved, resulting in the formation of a I-0-alk-1'-enyl glycerol phosphoryl-X lipid. Since the diacyl phospholipids are stable in mild acid and alkali labile, these properties form the basis for one common method for analysis of these classes (Dawson ef al., 1962). The alkyl ether lipids are also alkali stable, and it is necessary to use other criteria, such as I, uptake (Gottfried and Rapport, 1962) or measures of the amount of aldehyde produced, to determine the amount of plasmalogen. The polar head groups on bacterial plasmalogens are somewhat more restricted than those found on the diacyl phosphatides. Plasmalogens with ethanolarnine, N-methylethanolamine, choline, glycerol, and serine have been reported (see Goldfine and Hagen, 1972; van Golde et al., 1975). Since relatively few species have been analyzed, other head groups may be found. In the recently approved nomenclature the ethanolamine-containing plasmalogen is called plas-

8

HOWARD GOLDFINE

menylethanolamine, the choline-containing plasmalogen, plasmenylcholine, and so forth (IUPAC-IUB Commission on Biochemical Nomenclature, 1978). A related lipid having the structure of a glycerol acetal of a plasmalogen has been found in Clostridium hutyricum, I F 0 3852 (Matsumoto et al., 1971) and ATCC 6015 (Khuller and Goldfine, 1974). The former strain has ethanolamine in the polar head group of this lipid, whereas the latter has N-methylethanolamine plus ethanolamine, with the former predominating. In this lipid the R, chain of the plasmalogen has a glycerol substituent: OCH,CHOHCH,OH H,CI

I

O-CH-CH,-RI

As previously noted (Section Il,A,3), 0-alkyl glyceryl ether lipids have also been detected in anaerobic bacteria. They usually represent less than 4% of the total phospholipids.

2. GLYCOSYLDIGLYCERIDES A thorough survey of bacterial glycolipids was published by Shaw (1 975), and some additional material will be found in Lechevalier (1 977). The most common type of glycolipid found in bacteria has the 1,2-diacyl sn-glycerol-3-(sugar)n structure in which n = 2 (Fig. 2). There is considerable variation in the structures of the sugar residues, and certain glycolipids are considered to be characteristic of a given genus. For example, streptococci have glycosyldiglycerides with the Glc(a 1+2)Glc(a I+) substituent; staphylococci and Bui1lu.s sp. have Glc(/3I+6)GIc(/31-+) (Fig. 2B); and lactobacilli and pneumococci have Gal(@I-+2)Glc(cu I+) substituents (Fig. 2A). These glycosyldiglycerides, digalactosyldiglycerides, and dimannosyldiglycerides are the most common bacterial diglycosyldiglycerides (Shaw, 1975). In addition to the diglycosyldigl ycerides , monoglycos yldiglycerides with Glc(a I +) , Glc@ I +), Gal@ 1+), Gal@ I -+I, GIcN@3I-), and glucuronic acid have been characterized. These are usually precursors to the diglycosyldiglycerides and as such do not accumulate, but in some species they do. Tri- and tetraglycosyldiglycerides have also been isolated from bacteria with glucose and galactose as the most frequent terminal sugars (Shaw, 1975). In view of the decreasing lipid solubility of compounds with more than three sugar residues, it is possible that compounds of this type with more than four sugars are present, but not extractable with lipid solvents. In many gram-positive bacteria the glycosyldiglycerides are substituted with a sn-glycerol-1-P moiety in which the glycerol may be acylated with one or two fatty acids (Fig. 2C). The relationship of these glycolipids to lipoteichoic acids is discussed in Section III,N.


9

A

CHOCOR CHZOCOR

HO

on on

CHOCOR I CHzOCOR

0

II CHZOCR

I P

C 0

CHO~R

I

II CH20POCH2

0 11

CHzOCR

G 0 A

HO

on

FIG.2 .

Structures of three representative bacterial glycosyldiglycerides. (A) Galactosyl ( a1 4 2 ) Glucosyl(ol1jdiglyceride). (B) Glucosyl(/31~6)Glucosyl(~ I+diglyceride). (C) Phosphatidylkojibiosyl diglyceride.

C. Other Polar Lipids 1 . ORNITHINE LIPIDS

Several nonphosphate-containing ornithine lipids have been found in bacteria. In one type both the carboxyl group and the amino group of ornithine are linked to fatty acids (Fig. 3A), the carboxyl group through a polyol, and the a-amino group directly in amide linkage. In the second type. which is zwitterionic, the carboxyl group of ornithine is unesterified, but there is a p-hydroxy fatty acid linked to the a-amino group to which other fatty acids are esterified at the hydroxyl group (Fig. 3B). These zwitterionic ornithine lipids are the predominant polar lipids in certain Streptoinyes species, which usually have little or no phosphatidylethanolamine, or in other species under conditions of phosphate limitation (Batrakov and Bergelson, 1978). A third type ofornithine lipid in which the carboxyl group of ornithine is

10

HOWARD GOLDFINE 0 0 II II C-0 (CH2)n-O-CR, I CHNH I

A NH,CH,(CH,),

c=o I

R2

0 II

B

0

c-0

I NH3CH ,(CH,),CHNH

0

I C=O I

0

CH2 II I RZCO-CH I Rl

FIG.3 . Structures of two omithine lipids found in bacteria. See explanation in text.

esterified to an a-hydroxy fatty acid and a 3-hydroxy fatty acid is linked to the a-amino group of ornithine has also been found in Actinomyces (Streptomyces) strain 660-15 (Batrakov and Bergelson, 1978).

2 . GLYCOLIPIDS

In addition to the glycosyldiglycerides and the phosphatidylinositol mannosides, there are a variety of glycolipids that are not readily classified. They are not widely distributed in bacteria and more complete details of their isolation and characterization in addition to primary references can be found in the reviews by Goren (1972), Shaw ( I 974, 1975), and Lechevalier (1977). Some examples are given in Fig. 4. The simplest is a polyacylated glucose (Fig. 4A). The triacylated form shown has been found in members of several groups of prokaryotes. Trehalose 6,6’ dimycolate (“cord factor”) is found in mycobacteria, corynebacteria, and nocardia (Fig. 4B). The R groups in this case are mycolic acids of the general structure OH I

R;-CH-CHCOOH I

R;

In mycobacteria, depending on the species, R’, is a linear alkane, C Z 2or C,,, and R‘?is a complex structure of approximately 60 carbon atoms with hydroxyl, methoxyl, carbonyl, carboxyl, cyclopropane, methyl branches, and carboncarbon double bonds. The intact mycolic acids from mycobacteria have recently been fractionated by high-performance liquid chromatography into homologous series (Quereshi er ul., 1978; Steck et ul., 1978). Trehalose may also be esterified by a series of polyunsaturated fatty acids called phleic acids, which have


A

11

0 CHzOCR I

OH

0

B

OH

c

0 H H II C H 3 C H ( C H ),3C-C-CH20POCH,CH,NH3 I I I I 0 CH3 HO YH 00

c=o (:Hz)fl CH3

FIG. 4.

(A) Triacylglucose. (B) Trehalose 6,6’-dimycolate. “cord-factor” from the C o r w r hactrrium~Mvc‘ohuc.rerium-Noc,ardiagroup. (C) Ethanolamine-containing sphingolipid from bacteroides.

been found in Mycobacrerium phlei. The principal phleic acid is hexatriaconta4,8,12,16,20-pentaenoic acid (Asselineau et a / . , 1972). Also characteristic of mycobacteria are the mycosides, in which 2-0-methylrhamnose is linked through phenol to a complex fatty group with both long-chain polyols and fatty acids (see Goren, 1972, for a review). Goren and his colleagues have also characterized a series of sulfolipids from M. tuherculnsis H37R, in which the 2’ hydroxyl of trehalose is sulfated and acyl groups are located at the 2,3,6, and 6’ positions (Goren, 1972). 3. S P H IN GO LI PI DS

Sphingolipids are rare in prokaryotes; however, they appear to be characteristic of some members of the genus Barteroides. These lipids contain various polar head groups such as phosphorylethanolamine, phosphorylglycerol, and in relatively small amounts, phosphorylglycerophosphate, which are esterified to ceramide (N-acylsphingosine) (Fig. 4C).

D. Nonpolar Lipids The nonpolar lipids of bacteria have often been neglected in general considerations of bacterial membranes. In part this is due to our scanty knowledge, the absence of clear-cut taxonomic correlations, and variation in the reporting of the quantitative analyses of this class of lipids. Most workers have simply stated that a given species has X % nonpolar among the total lipids. These may include

12

HOWARD GOLDFINE

varying amounts of monoacylglycerol , diacylglycerol, hydrocarbons, carotenoids, quinones, free fatty acids, fatty alcohols, waxes, and poly-P-hydroxybutyric acid. Triglycerides are generally not found, or are present in only trace amounts. Sterols, though generally either totally absent or present in very small amounts in most prokaryotes, are important for certain groups of organisms, which are discussed by Durisson and Rohmer and by Razin (this volume). Although certain nonpolar lipids such as the coenzymes Q, vitamins K, and carotenoids are known to play specialized roles in bacterial membranes-for example, in electron transport, oxidative phosphorylation, photosynthesis, and ion and solute transport-the functions, if any, of other neutral lipids remain unclear. Some may simply represent intermediates in pathways related to polar lipid catabolism or metabolism, others may represent a convenient pool of lipid components for eventual utilization; however, the possibility remains that small amounts of neutral lipids may be useful to ensure such membrane properties as stability, fluidity, or the provision of specific lipid structures for interaction with membrane proteins.

E. Nonextractable Lipids In many prokaryotes a portion of the total lipid is not extractable with the usual solvents such as mixtures of chloroform and methanol. However, on acid or alkaline hydrolysis of the nonextractable residue, further amounts of lipid, usually in the form of free fatty acid, are released. Among these bound forms are the lipopolysaccharides of gram-negative bacteria, which are described by Luderitz et al. (this volume), the lipoteichoic acids of gram-positive bacteria (see Section III,N), an acylated mannan found in Micrococcus lysodeikticus (Powell et al., 1975), and the complex waxes D of mycobacteria, corynebacteria, nocardia, and actinomycetes (reviewed by Goren, 1972). In the past decade lipids covalently linked to protein have been described. A major component of the outer membrane of enterobacteria is a lipoprotein of molecular weight 7500, which has a diacylglycerol linked to an N-terminal cysteine through a thioether (Braun and Hantke, 1974). With the recent discovery of protein-linked fatty acids in certain animal viruses (Schmidt et al., 1979), the possibility that this may represent a more general form of membrane organization should be considered.

111.

DISTRIBUTION OF LIPIDS IN PROKARYOTES

Several extensive reviews and compilations of microbial lipid compositions have appeared during the past decade (Goldfine, 1972; O’Leary, 1973; Shaw, 1974; Lechevalier, 1977). The last is the most complete; however, it and the


13

review by Shaw (1974) present qualitative rather than quantitative data. The review by Lechevalier (1977) also describes in some detail effects of pH, medium constituents, growth temperature, and age of cultures on bacterial lipid compositions. It is because of the variability induced by these factors that these authors have eschewed quantitative data. An understanding of the structure and organization of bacterial membranes, however, requires these quantitative relationships, information on the compositions of distinctive membrane fractions, and on the arrangement of lipids in the membranes. The last aspect will be discussed by Rottem in this volume. An unstated premise of this chapter has been that the lipids isolated from bacteria by extraction of whole cells are the lipids present in bacterial cell membranes. In general this conclusion has been supported by many studies on isolated bacterial membranes. Earlier work on gram-positive bacteria showed that membranes obtained from these organisms did contain the same lipids in approximately the same proportions as the whole cells (Vorbeck and Marinetti, 1965; Bishop et ul., 1967); however, specific associations of certain glycolipids with cell walls cannot be discounted (Shaw, 1975). In gram-negative organisms the problem is more complex because their outer membrane, which is part of their cell wall, contains lipids. Many species, especially those capable of photosynthesis, have internal membranes as well. The outer membranes of Escherichia coli and Salmonella typhirnuriurn contain the same phospholipids as the inner membrane, although the ratio of phosphatidylethanolamine to phosphatidylglycerol plus cardiolipin (Osborn et ul., 1972; Diedrich and CotaRobles, 1974; Rottem et ul., 1975) and the ratio of saturated to unsaturated fatty acids (White et al., 1972; Diedrich and Cota-Robles, 1974; Koplow and Goldfine, 1974; Rottem er a/., 1975; Lugtenberg and Peters, 1976) are somewhat higher in the outer than in the inner membrane. Kenyon (1978) has recently reviewed the lipids of photosynthetic bacteria. In several studies the phospholipid composition of the subcellular fractions, including the chromatophores, of these bacteria had similar phospholipid compositions to that of the whole cell (Gorchein, 1964, 1968; Haverkate er d.,1965; Takacs and Holt, 1971), whereas some quantitative differences in the chromatophores and crude membranes of Ectothiorhodospira halophila SL- 1 were observed in an unpublished study (Kenyon, 1978). It would indeed be surprising if the functionally differentiated membrane systems of prokaryotes did not show some differentiation in the compositions of their complex lipids. With these caveats in mind, the lipid compositions of the major groups of prokaryotes will be presented. I shall not duplicate recent reviews that have presented detailed quantitative data on bacterial lipid compositions (see previous discussion). Rather, the broader picture will hopefully emerge through the use of selected examples. However, it is important to realize that even within a bacterial genus, significant differences in lipid composition have been found.

14

HOWARD GOLDFINE

A. Cyanobacteria (Blue-Green Algae) The blue-green algae studied resemble green algae and the photosynthetic apparatus of higher plants in their complex lipid composition. The major lipids are phosphatidylglycerol, monogalactosyldiglyceride, digalactosyldiglyceride, and sulfoquinovosyldiglyceride [SQDG; 1,2-diacyl-sn-glycero-3-(6-sulfo-~u-oquinovopyranoside)]. They lack phosphatidylethanolamine and phosphatidylcholine (Nichols r t u l . , 1965). It is interesting to note that in fatty acid composition some of the blue-green algae resemble bacteria in the absence of polyunsaturated fatty acids, whereas others-especially , but not exclusively, the filamentous types-have polyunsaturated fatty acids (Kenyon and Stanier, 1970).

B. Phototrophic Bacteria The eighth edition of Bergey's Munuul of Determinative Bacteriology (Buchanan and Gibbons, 1974) divides the phototrophic bacteria (Rhodospirillales) into three families. Two of the best-studied members of the Rhodospirillaceae family, the purple nonsulfur bacteria, which are representative in terms of their lipid compositions, are Rhodospirillum ruhrurn and Rhodopseudomows sphueroides. Several of the more recent analyses of the former agree that the major polar lipids of light- and dark-grown cells are phosphatidylethanolamine, phosphatidylglycerol, cardiolipin, and a nonphosphate-containing ornithine lipid (Fig. 3B) (Hirayama, 1968; Depinto, 1967). Earlier reports of the presence of phosphatidylcholine were not confirmed (Brooks and Benson, 1972). Rp. sphueroides has phosphatidylethanolamine,phosphatidylglycerol, phosphatidylcholine, and an omithine lipid (Fig. 3B) as its major lipids. Small amounts of SQDG and cardiolipin were also reported. Fig. 5A presents the phospholipid composition of R p . sphueroides diagrammatically. The lipid composition of Rp. capsulatu is similar, and lecithin increases at the expense of the other major lipids in dark-grown cells (Steiner et al., 1970). There have been fewer studies of the purple sulfur bacteria (Chromatiaceae). The major lipids of Chromatium viiiosum are phosphatidylethanolamine and phosphatidylglycerol, with smaller amounts of cardiolipin and some glucose and mannose-containing lipids (see Kenyon, 1978, for references). Although the lipids of Thiocapsu roseopersirina and Ectothiorhodospira halophila, an extreme halophile, have been studied, no complete analyses are available. The latter organism does not have the d i - 0 alkylglycerol ether lipids characteristic of Halobacterium (see Langworthy, this volume). Its major lipid is phosphatidylglycerol, with smaller amounts of phosphatidylethanolamine and unknown phospholipids. Although total analyses of the green sulfur bacteria (Chlorobiaceae) are also lacking, it is of interest that several species have been found to have the monogalactolipid and SQDG charac-

15


w

0

C

rn Gal OG

Rhodopseudomonos sphaeroides a

Hyphornfcrobiurn vulgore N0-52Ib

Treponerno pollidum Kazon 5'

(light-grown)

@ CL Azofobacter

Pseudomonos

ieptospfro potocd

ogr1is (log)'

oerugfnosa'

G

I

H

@ CL

Aqroboclerium tumefociens flog)

'

Escherrchjo COll

Serratio rnorcescens'

(log)'

J

Proteus vulgaris

(Ioq)'

qonorrhoeoe

*

Megasphaera elsdenii

FIG. 5. Polar lipid compositions of gram-negative bacteria. Nonionic phospholipids are not shaded. Anionic phospholipids are stippled. Phosphatidylserine is shown with diagonal hatching. Unmarked areas of the diagrams represent unknown lipid components, mixtures of small amounts of known lipids, or a combination of the two. PE, Phosphatidylethanolamine; LPE, lysophosphatidylethanolamine; PlaE, plasmalogen form of PE; PME, phosphatidyl-N-methylethanolamine; PDME, phosphatidyl-N,N-dimethylethanolamine;PC, phosphatidylcholine; PG, phosphatidylglycerol; CL, cardiolipin; PS, phosphatidylserine; PlaS, plasmalogen form of PS; DG, diglyceride. References: "Gorchein. 1968; Russell and Harwood, 1979; "Goldfine and Hagen, 1968; Johnson or al., 1970b; "Johnson et ul., 1970a; '~Hancock and Meadow, 1969; 'Randle cf a / . , 1969; "Senff ef a / . , 1976: Beebe and Wlodkowski, 1976; "van Golde ef a / . , 1975. (

teristic of higher plants, in addition to phosphatidylglycerol. No phosphatidylethanolamine was found in pure cultures (Kenyon, 1978). The major fatty acids of photosynthetic bacteria are 16:0, 16:1, and 18:l. Substantial amounts of 14:O are found in the Chrornatiaceae (Kenyon, 1978).

16

HOWARD GOLDFINE

C. The Gliding Bacteria and the Sheathed Bacteria So little is known about the lipids of the Myxobacteriales that it would be premature to generalize about the membrane composition of this interesting group of organisms, which is generally capable of forming fruiting bodies. The phospholipids of a marine organism, tentatively identified as a Sporocytophaga sp., were shown to be typical of gram-negative bacteria, consisting principally of phosphatidylethanolamine (76%) and phosphatidylglycerol (20%) Oliver and Colwell, 1973). Oral isolates of Capnocyrophagu and the related Sporocytophuga were recently found to have considerable amounts of neutral lipids in addition to phospholipids. Phosphatidylethanolamine, lysophosphatidylethanolamine, and two unidentified phospholipids were the major polar lipid components of three species of Cupnocyrophaga. An ornithine lipid was also detected. In addition to phosphatidylethanolamine and the lyso analogue, phosphatidylserine was found in Sporocytophaga (Holt et al., 1979). An unusual class of lipids, based on the sulfonolipid capnine (2-amino-3-hydroxy15-methylhexadecane-I -sulfonic acid), has been found as a major constituent of the cell envelope of Capnocytophuga (Godchaux and Leadbetter, 1980). These lipids are similar to 1 -deoxyceramide-1 -sulfonic acid previously found as a minor component of the diatom Nirzschin alba (Anderson et a / . , 1978). Some species have the free form of capnine, but most have N-acylcapnines in which the acyl moieties are rich in 2- and 3-hydroxy groups, and have methyl branches (Godchaux and Leadbetter, 1981). Capnines have been found in a variety of other gliding bacteria including Cytophngu johnsonae, Vitreoscilh Jrercoruriu, Flexibacter, and Sporocytophaga myxococcoiries (W. Godchaux, personal communication). The structures of the capnines and N-acyl capnines are analogous to those of sphingosine and ceramide (see Fig. 4C). The principal phospholipids of the membranes of M\;xococcus xanthus were found to be phosphatidylethanolamine (76%), and phosphatidylglycerol (9%) (Omdorff and Dworkin, 1980). The lipid composition of the sheathed bacteria has received less attention.

D. Budding and/or Appendaged Bacteria Among this large group, data are available for Hyphornicrobiurri and Cuulobucter; the former divide by budding at the tips of their hyphae, whereas the latter produce adherent stalks, but divide by binary fission. Hyphomicrohiutn has an unusual mixture of phospholipids characterized by a high proportion of phosphatidyl-N,N’-dimethylethanolamine, which represented 36% of the phospholipids in strain NQ-521 and a similarly large proportion of the lipids in three other strains (Hagen et a / . , 1966). The other phospholipids in strain NQ-521 were phosphatidylcholine, phosphatidylethanolamine, and

17


polyglycerol phosphatides (Fig. 5B). The phospholipids are very rich in cis- 1 1 18:1 (cis-vaccenic acid) (Auran and Schmidt, 1972). Caulobacrer crrscrntiis has a phospholipid composition that is highly unusual for a gram-negative organism. The major phospholipid is phosphatidylglycerol (>80%). Cardiolipin, acylphosphatidylglycerol, and lysophosphatidylglycerol have been identified as minor phospholipids (Contreras et al., 1978; Jones and Smith, 1979; DeSiervo and Homola, 1980). A major proportion (-50%) of the polar lipids are glucosyldiglyceride and other glycolipids (DeSiervo and Homola, 1980).

E. The Spirochetes These flexible, helically coiled bacteria are divided into several genera on the basis of morphological and biochemical characteristics. Determinations of lipid composition were hampered by the difficulties encountered in the laboratory cultivation of these organisms. The major groups have now been studied and several generalizations have emerged. Spirochaeta are characterized by the presence of phosphatidylglycerol and cardiolipin, the absence of the N-containing phosphatides, and the presence of glycolipids of the monoglycosyldiglyceride type, which usually are close to 50% of the polar lipids. Thus their lipids resemble those of many gram-positive organisms. Treponeina have phosphatidylcholine and (usually) phosphatidylethanolamine, with minor amounts of polyglycerolphosphatides. They have the same type of glycolipids as the Spirochaeta, in similarly large amounts (Fig. SC) (Livermore and Johnson, 1974). Some of the treponemes have alk-1 -enyl acyl (plasmalogen) phospho- and glycolipids (Meyer and Meyer, 1971; Matthews et ul., 1979). Leptospira have neither glycosyldiglycerides nor phosphatidylcholine. The principal phospholipid is phosphatidylethanolamine, which may be accompanied by polyglycerolphosphatides (Fig. 5D) (Johnson et al., 1970a). Borrrlia hermsi has been shown to synthesize phosphatidylcholine, phosphatidylglycerol, monogalactosyldiglyceride, and cholesterylglucosides (Livennore et al., 1978). TABLE I11 CHARACTERI~TIC. POLARLiPrns OF T H E SPIROCHETES Lipid Monoglycosyldiglyceride Cholesterylglucosides Phosphatidylchol ine

Phocphatidylethanolamine ‘I

Spirochneru

Treponumu

+

+

~

~

Leptospiru

-

+ +

Except T . prillidum (Nichols virulent strain) (Matthews ef a l . , 1979)

Borreliii

+ + + +

18

HOWARD GOLDFINE

The lipid composition of B . hermsi is therefore similar to that of the treponemes, except that the latter do not have cholesterylglucosides. Table I11 presents a comparison of the lipid compositions of the spirochetes. These marked differences suggest that these groups diverged early in their evolution, which is consistent with other data (Fox et ul., 1980).

F. Spiral and Curved Bacteria This group has had little attention. A marine species, Spirillum linum, has a typical gram-negative phospholipid complement consisting of phosphatidylethanolamine (75%) and phosphatidylglycerol (23%) (Oliver and Colwell, 1973). The lipids of Campylobacter (Vihrio),fetus were reported to contain the same two major phospholipids plus small amounts of phosphatidylserine, phosphatidylinositol, and digalactosyldiglyceride (Tornabene and Ogg, 197 1). These cells had considerable amounts of neutral lipid, especially when in the coccoid form.

G. Gram-Negative Aerobic Rods and Cocci 1 . PSEUDOMONADACEAE Among the family Pseudomonadaceae, the genus Pseudomnnas has been studied most intensively. Most species have a phospholipid composition characteristic of gram-negative organisms, consisting of phosphatidylethanolamine, which is the most abundant lipid, phosphatidylglycerol, and cardiolipin (see Shaw, 1974, for references). The phospholipid composition of P . aeruginosu is shown in Fig. 5E. Some species, including P. ueruginosa, have been reported to contain small amounts of phosphatidylcholine (see Goldfine, 1972, for references), and a zwitterionic ornithine lipid (Fig. 3B) was found in several species sensitive to ethylenediaminetetraacetic acid when the cells were grown on nutrient agar (Wilkinson, 1970), but not when grown in nutrient broth (Wilkinson et ul., 1973). Dramatic increases in the ratio of the ornithine lipid to the phospholipids were shown in phosphate-limited cultures of P . jluorescens NCMB 129 (Minnikin and Abdolrahimzadeh, 1974a). Four species, P . diminuta, P . multophilia, P. vesicularis, and P . rubescens, have glycosyldiglycerides containing both glucose and glucuronic acid. Although still designated Pseudnmonas in the eighth edition of Bergey’s Manuul (Buchanan and Gibbons, 1974), the first three were placed in a distinct group based on their growth-factor requirements, and the last is no longer included in this genus, again strengthening the taxonomic value of lipid-compositional studies (Shaw, 1974; Lechevalier, 1977; Wilkinson and Galbraith, 1979). Gluconnbacter, another genus in the Pseudomonadaceae, has been reported to contain phosphatidylcholine and an


19

uncharacterized ornithine lipid in addition to the usual lipids of the genus Pseudomonas (Tahara et a / . , 1976; Heefner and Claus, 1978).

2. AZOTOBACTERACEAE Like the Pseudomonadaceae, the Azotobacteraceae are aerobic gram-negative rods, but they are capable of fixing nitrogen. The phospholipids of two species of Azotobucter, A . agilis and A . vinelundii, have been studied. The former has small amounts of phosphatidyl-N-methylethanolamineand phosphatidylcholine, in addition to major amounts of phosphatidylethanolamine and phosphatidylglycerol (Fig. 5F). The latter does not appear to possess the N-methylation system needed for the synthesis of phosphatidylcholine (Randle et ul., 1969; Jurtshuk and Schlech, 1969). When induced to encyst by addition of @-hydroxybutyrate to the medium, A . vinelundii accumulates 5-n-alkylrcsorcinols and their galactose derivatives. The alkyl chains are C,, and C 2 , (Reusch and Sadoff, 1979). These unusual lipids appear to become major membrane components during encystment because the phospholipids decrease to less than 10% of the total (Reusch and Sadoff, 1981). Little work appears to have been done on the membrane lipids of other genera in this family.

3 . RHIZOBIACEAE Rhizobia are capable of infecting the roots of leguminous plants to produce nodules, and agrobacteria infect diverse species of plants and produce gall hypertrophies. All rhizobia examined have phosphatidylcholine in addition to the usual lipids found in gram-negative bacteria. I n two strains examined by Gerson and Patel ( 1 9 7 3 , phosphatidylcholine represented 5 1 % and 21 % of total phospholipid in the free-living form, and 33% and 29% in the bacteroid isolated from root nodules. Phosphatidylinositol was also found in these strains, but not in R. juponicum (Bunn and Elkan, 1971) and R . leguminosururn (Faizova et al., I97 I ). All species of Agrobuc/eriutn likewise have phosphatidylcholine (Goldfine and Ellis, 1964), and A . fumefuciens (Fig. 5G) (Kaneshiro and Marr, 1962; Randle et ul., 1969) is typical. In this organism phosphatidylcholine increased to 28% of the phospholipid in stationary cells at the expense of its precursors (Randle e / ul., 1969).

4. METHYLOMONADACEAE A N D O T H E RMETHANE-UTILIZING BACTERIA Many of these organisms are characterized by an obligate requirement for one-carbon organic compounds as a source of carbon. Me/hvlomotzns methanolicu has a phospholipid composition similar to that of E . coli (Goldberg and Jensen, 1977) (see Fig. 5H),but Methylococcus cupsulutus has

20

HOWARD GOLDFINE

8% phosphatidylcholine and Methylosinus trichosporium is rich in N-methylethanolamine and N,N’-dimethylethanolamine, as well as choline phosphatides (Makula, 1978). However, Weaver ef NI. (1975) reported mainly phosphatidylglycerol and phosphatidylethanolamine in the latter organism. Unlike M . capsulatus and M . rnethanolica, which have intracytoplasmic membranes consisting of vesicular disks organized into bundles (type I), M . trichosporium has type I1 membranes, which are characteristically arranged either at the periphery of the cells or paired and extending throughout the cells. These two groups of methane-utilizing organisms also differ in the pathways utilized for 1-carbon assimilation (Makula, 1978). Two other strains of methane-utilizing bacteria, LaPaz and OBT, also had high levels of N,N’-dimethylethanolamine and choline- or N-monomethylethanolamine phosphatides (Makula, 1978). A facultative methylotrophic organism that has type I1 intracytoplasmic membranes when grown on methane, but none when grown on glucose or methanol, Methylobucterium orgunophiliurn, has phosphatidyl-N,N’-dimethylethanolamine, phosphatidylcholine, and phosphatidylethanolamine, with a somewhat higher proportion of the methylated bases in methanol- and glucose-grown cells than in methane-grown cells (Patt and Hanson, 1978).

5. HALOBACTERIACEAE These organisms, which require above 2 M sodium chloride for growth, have unusual polyisopranoid ether lipids that are described by Langworthy (this volume). 6. GRAM-NEGATIVE AEROBIC RODS A N D Cocci AFFILIAIION

OF

UNCERTAIN

Most of the organisms of this group, which includes Alculigenes, Acetobarter, Brucellu, and Bordetella, have typical gram-negative lipid compositions. Whereas Alculigenes (Lechevalier, 1977) and Bordetella pertussis do not have phosphatidylcholine, Brucella ubortus and Brucella melitensis are both rich in this lipid and have in addition the N-methylated intermediates between phosphatidylethanolamine and phosphatidylcholine (Thiele and Schwinn, 1973). Brucella and Bordetella also have ornithine lipids of the type illustrated in Fig. 3A.

H. Gram-Negative Facultatively Anaerobic Rods This group of organisms includes some of the most familiar prokaryotes, such as E . coli, Salmonella, Shigella, Klebsiella, Serratia, and Proteus, which with


21

several other genera are grouped together as the Enterobacteriaceae, and another family, Vibrionaceae, which includes Vihrio and Aeromonas with several other genera. Typical phospholipid patterns are illustrated in Fig. 5H and 51. Salrnonella is very similar to E . coli in having mainly phosphatidylethanolamine and lesser amounts of phosphatidylglycerol and cardiolipin (Ames, 1968). The ratios of phosphatidylglycerol and cardiolipin are often variable depending on growth phase, osmolarity, and carbon source, but such changes may be more directly related to growth rate (Merlie and Pizer, 1973). Some enteric organisms, such as Proreus (Goldfine and Ellis, 1964) and Yersinia species (Tornabene, 1973), have small amounts of the N-methylated ethanolamine phosphatides, but no lecithin (Fig. 5J). The phospholipids of a number of Vibrio species mainly of marine origin, but including V . cholerue, were analyzed by Oliver and Colwell (1973). With the exception of V . marinus, all vibrios had typical gram-negative phospholipid compositions: phosphatidylethanolamine (60-80%), phosphatidylglycerol ( 1530%), and small amounts of lysophosphatidylethanolamine and cardiolipin. V . marinus had more phosphatidylglycerol (54%) than phosphatidylethanolamine (23%). The phospholipid composition of Chrornobacterium violaceum (Randle et al., 1969) and Huemophilus influenzae Rd (Sutrina and Scocca, 1976) is very similar to that of E . coli log-phase cells (Fig. 5H), except that no cardiolipin was detected in the latter organism.

I. Gram-Negative Anaerobic Bacteria 1 . BACTEROIDACEAE

Many species of Bacteroides have been shown to contain sphingolipids, a unique finding among prokaryotes (White et a/., 1969; Fritsche and Thelen, 1973). Miyagawa et al. (1978) examined 15 species of Bacteroides and found evidence for sphingolipids in 10, but the other 5-B. succinogenes, B . furcosus, B. hypernegus, B. umylophilus, and B. multiacidus-did not have sphingolipids. On this basis it was proposed that the inclusion of these species among the bacteroides is in some doubt. The major sphingolipids in the bacteroides have been characterized as ceramide phosphorylethanolamine (Fig. 4C) and ceramide phosphorylglycerol. The major phosphoglycerides are phosphatidylethanolamine and smaller amounts of the polyglycerolphosphatides, which represent from one-third to one-half of the polar lipids (White et al., 1969; Stoffel et al., 1975). Even within a species, B . melaninogenicus, considerable variation was seen in the relative proportions of the sphingolipids and diacylphospholipids (Rizza el al., 1970). Among the organisms containing sphingolipids, there are branched chains in both the acyl groups and the long-chain bases.

22

HOWARD GOLDFINE

2 . Desirlfbvibrio

AND

Butyrivihrio

Two species of Desulfovibrio, D . clesulfuricuns Norway and D . vulgaris, have typical gram-negative phospholipid patterns consisting of 61 to 72% phosphatidylethanolamine, 20 to 21 5% phosphatidylglycerol, and smaller amounts of cardiolipin. D . desu~uricunshas, in addition, 1 I % phosphatidylserine (Makula and Finnerty, 1974). D . gigas differs considerably in having a phosphatidylethanolamine to phosphatidylglycerol ratio of 30:70, and in having an ornithine lipid of the zwitterionic type (Fig. 3B). The ornithine lipid represented 78% of the total lipid (Makula and Finerty, 1975). Butyrivibrio, obligately anaerobic bacteria of the rumen, appear to have. a unique group of lipids, which have been recently characterized by a group of workers at Babraham, Cambridge, England. Three unusual features are worth emphasizing. The polar lipids of these organisms have various short-chain fatty acids. For example, there are n-butyryl esters of phosphatidylglycerol, and lipids in which a galactose residue of a galactolipid is esterified with butyrate. In all of the species examined, alk-l -enyl acyl substituents are found on the diglyceride moieties (Clarke et ul., 1976). The most unusual feature of one such lipid is the cross-linking of two “diglyceride ” moieties with a long-chain dicarboxylic acid that has a vicinal dimethyl substitution at the center of the chain. These fatty acids have been named diabolic acids (Fig. 6) (Klein et ul., 1979). Thus the intact lipid can be thought of as a dimer of a plasmalogenic glycosyldiglyceride and a plasmalogenic phosphatidylglycerol in which the glycerol group is esterified with butyrate. The two plasmalogens are cross-linked by the diabolic acid. It is not known whether the lipid spans the membrane of these cells or is bent into a hairpin structure (Hazlewood et ul., 1980). Selenomonas ruminantium, another anaerobic rumen bacterium, contains ethanolamine phosphatides as the major polar lipid class in its cytoplasmic mem-

H,C [CH2],3CH=CH-O-hH,

n n

FIG. 6 . Structure of a diabolic acid-containing phospholipid isolated from Butyrivibrio S2 grown in the presence of palmitic acid. The R group esterified to the galactose is a butyroyl residue. The butyroyl group on the glycerol residue may be replaced by a palmitoyl group (Clarke er a / . , 1980). Two molecules of sn- 1 -alkenylglycero-3-pbospho-m1 ‘-glycerol butyroyl ester may also be linked through a diabolic acid.


23

brane. These exist in the diacyl, alk-I-enyl acyl, and alkyl acyl forms in the ratio 1:0.5:0.2 (Kamio and Takahashi, 1980). J. Gram-Negative Cocci and Coccobacilli Recent work on the membrane lipids of Neisseriu gonorrhoeae (Fig. 5K) and Brunhomellu caturrhalis has shown that gram-negative cocci, which like the gram-negative rods have inner and outer membranes, have phospholipids typical of gram-negative bacteria consisting largely of phosphatidylethanolamine, phosphatidylglycerol, and cardiolipin. Some lysophosphatidylethanolamine has been found, but this was largely attributed to the action of an endogenous phospholipase A (Senff et al., 1976; Beebe and Wlodkowski, 1976; see Lechevalier, 1977, for earlier references). Phosphatidylcholine was found in some strains of N . gonorrhoeue (Sud and Feingold, 1975) and in B . cuturrhalis (Beebe and Wlodkowski, 1976). Other gram-negative, cocci such as Acinetobucter sp. HO 1 -N (previously Micrococcus ceriJiccin.7) and Acirietobacter sp. MJT/F5/ 199A, similarly have typical gram-negative phospholipid patterns, but in strain HOI-N there is 12 to 3 1% phosphatidylcholine in addition to phosphatidylethanolamine and cardiolipin (Makula and Finerty, 1970; Thorne et ul.. 1973).

K. Gram-Negative Anaerobic Cocci The major membrane lipids of Veillonellu purvulu and Megusphaeru elsdenii have, like those of many anaerobes, a large proportion of alk-I-enyl acyl phospholipids. In addition, these organisms are unusual in that a major component of the phospholipids are serine phosphatides (Fig. 5L). This was found to be characteristic of anaerobes that ferment lactate (van Golde et al., 1975).

L. Gram-Negative Chemolithotrophic Bacteria This group of organisms is divided into those oxidizing ammonia or nitrite, those metabolizing sulfur, and those depositing iron or manganese oxide. Two of the nitrifying bacteria, which contain intracytoplasmic membrane systems, were found to have mainly phosphatidylethanolamine (67-78%) and polyglycerolphosphatides (17-1 8%) among their phospholipids. One of these, Nitrosococcus (Nitrosocystis) oceanus, has 3% phosphatidylcholine, whereas the other, Nitrosomonus europaeu, had no phosphatidylcholine (Hagen et al., 1966). Two species of Nitrobucter with internal membranes also have phosphatidylcholine (Auran and Schmidt, 1972).

24

HOWARD GOLDFINE

TABLE IV LIPIDSOF Thiobucillus

Group %GC I

56-57

I1 62-66:

111 SO-52

Fatty acids

Species

PE

PME

PC

C-14 and C-16 predominate C- I5 and C- 17 predominate

T . necipulitrrnu.7 7'..firmridtitis T . fhiopurus T . nowllus T . inti,rmedius" T . rhiooxicluns (log)

44' 20 65 25 58 20

23' 42

-

Most are C-14 or shorter

~~

Polyglycerol phosphatides 33 36 35 33 29 39

1.5

-

-

7

3s

14

-

36'

-

~

~~~

~

~~

%GC unknown. Values for phospholipids (% of lipid P) are the middle of the ranges given by Barridge and Shively (1968), Shively and Renson (1967). and Short r i nl. (1969). Increases in stationary phase (Shively and Benson, 1967; Agate and Vishniac. 1973). "

I

Among the organisms metabolizing sulfur, Thiobacillus and Sm/jo/~olohushave been studied intensively with respect to their membrane lipids. The latter group lives at high temperature and low pH. Its unusual lipids are described by Langworthy (this volume). In the eighth edition of Bergey's Manuul (Buchanan and Gibbons, 1974), the thiobacilli have been classified according to their lipid fatty acids and the percentage guanine-cytosine (CC) of their DNA. As can be seen in Table IV, the distribution of phospholipids does not follow any clear pattern. PhosphatidylN-methylethanolamine is found in members of each group. Phosphatidylcholinc is present in only one member of group 11, 7'.novellus, and one member of group 1 , T . (Ferrobucillus)ferrooxiduns. T . thiooxicluns has a zwitterionic ornithine lipid (Fig. 3B), but it is not known if this lipid is present in other thiobacilli.

M. Methane-Producing Bacteria The lipids of the methanogens, which are similar to those of the thermoacidophiles and related to those of the halobacteria, are discussed by Langworthy (this volume). They are characterized by branched chains (phytanyl) in ether linkage present in phospho- and glycolipids. It has recently been proposed that these organisms be grouped in a separate kingdom, the Archaebacteriae (Woese et d . , 1978).

25


N. Gram-Positive Cocci Differences in the cell wall composition and structure of the gram-positive and gram-negative groups of organisms are also reflected in major differences in the lipid compositions of their membranes. As noted above, most groups of gramnegative bacteria have phosphatidylethanolamine alone or together with the N-methylated ethanolamine phosphatides as the major nonionic group of polar lipids, which usually represent more than half of the total amphipathic lipid in these cells (Fig. 5). In gram-positive organisms the situation is reversed. Among the aerobic, facultatively anaerobic, and anaerobic cocci, for example, the major polar lipids are anionic phospholipids including phosphatidylglycerol, which is usually the most abundant, and cardiolipin. These may be accompanied by an 0-aminoacyl phosphatidylglycerol or phosphatidylinositol (Whiteside et ul., 1971; see Goldfine, 1972, and Lechevalier, 1977, for additional references). An early analysis of the lipids of Micrococcus lysodeiktici*s (lucteus) (Macfarlane, 1961a,b) is given in Fig. 7A. The glycolipid was identified as dimannosyldiglyceride (Lennarz and Talamo, 1966). The analyses of Whiteside e f al. (1971), based on 32P labeling, agree qualitatively, but the ratio of cardiolipin to phosphatidylglycerol is much higher (39:45). These authors found a similarly high ratio of cardiolipin to phosphatidylglycerol in M . refrugenus, other Micrococcus sp., and Surcinufluvu. It should also be noted that many of these organisms have

C

A

IGicDG

Muacoccus

lysodeikt,cus'

Slreptococcus pneurnonme t-192Ap

Slophylococcus aureus /logic

t?OCill"S

cereus

F

Baollus subti1,r'

80C~llUS

megateriurn'

L octabacillus pbntarumg

FIG. 7. Polar lipid compositions of gram-positive bacteria. PI, phosphatidylinositol. For other abbreviations, see legend to Fig. 5 . Lysyl-PG and ornithine-PG, which are cationic phospholipids, are shown with light stippling. Rcferences: ' I Macfarlane, 1961a.b; Brundish ef a/., 1967; Joyce r f d., 1970; Shaw, 1975; "Houtsmuller and van Deenen, 1963; "Bishop rt al., 1967; 'Bertsch af u l . . 1969; '' Exterkate t'r o l . , 107 I ; L . p l t r n t u r u t n contains an unknown proportion of galactosylglucosyldiglyceride (Shaw, 1975).

26

HOWARD GOLDFINE

a substantial amount of hydrocarbons ranging from 17 to 22% of total lipid (Tornabene er al., 1970). It is apparent from work with other gram-positive cocci that the lipid composition may vary considerably with growth stage and pH (Lechevalier, 1977). In Staphylococcus aureus, phosphatidylglycerol predominates during log phase (Fig. 7C), but the relative and absolute amount of lysylphosphatidylglycerol increases in stationary phase and/or at low external pH, whereas the absolute amount of phosphatidylglycerol decreases. Thus the ratio of the two lipids can reverse (Houtsmuller and van Deenen, 1965; Gould and Lennarz, 1970). In Planococcus, a group of motile, gram-positive cocci, some phosphatidylethanolamine is present along with cardiolipin and phosphatidylglycerol (Komura et ul., 1975a). In the streptococci the major phospholipids are also cardiolipin and phosphatidylglycerol, which may be accompanied by aminoacyl phosphatidylglycerol and glycolipids, predominantly diglucosyldiglyceride (Fig. 7B) (Goldfine, 1972; Shaw, 1975). The relative amounts of these lipids may vary with growth rates (Carson et al., 1979) and with growth phase (Chiu and Hung, 1979). Carson et al. (1979) noted that the predominant neutral lipid, diacylglycerol, and cardiolipin accumulated relative to cellular mass as the rate of growth decreased. At the shortest doubling time (30 minutes) the anionic phospholipids predominated. In addition to phospholipids and glycosyldiglycerides, many gram-positive cocci contain another class of lipids, which have been designated phosphoglycolipids (Pieringer and Ganfield, 1975). These are sn-glycerol- 1 -phosphate derivatives of diglycosyldiglycerides in which the glycerol-1 -phosphate may be acylated with fatty acids as in phosphatidylkojibiosyl diglyceride (Fig. 2C). These lipids may serve as a hydrophobic anchor for membrane teichoic acids in which there is a glycerol-P polymer linked to the disaccharide. The polymer usually has 20 to 40 glycerol-P units (Fischer et ul., 1980), which may be substituted with a variable number of disaccharide and alanine moieties (Pieringer and Ganfield, 1975). Many gram-positive cocci have an internal vesicular or tubular localized membrane system, the mesosome. It appears that the lipids of the mesosome and the plasma membrane are qualitatively similar; however, certain lipid fractions may be concentrated in the mesosome and the mesosome may have a higher content of lipid relative to dry weight than the plasma membrane (Thomas and Ellar, 1973; Beining et nl., 1975). With a few exceptions, most of the gram-negative organisms discussed earlier have mixtures of long-chain saturated and monounsaturated fatty acids. In many organisms, the monounsaturated fatty acids undergo conversion to cyclopropane fatty acids after they have been incorporated into membrane lipids (Law, 1971). In many gram-positive cocci, especially the Micrococcaceae and the anaerobic Peptococcaceae (Whiteside et ul., 1971; Shaw, 1974; Lambert and Armfield,


27

1979), unsaturated fatty acids are replaced by branched-chain fatty acids of the iso- and anteiso type.

0. Endospore-Forming Bacteria The aerobic and facultative genus Bncillits lies in phospholipid compositional terms between the aerobic gram-positive cocci and the gram-negative bacteria. In many species there is a considerable amount of phosphatidylethanolamine,which usually comprises from 20 to 45% of the total phospholipid (Fig. 7D-F) compared to the 55 to 80% phosphatidylethanolamine plus the N-methylated ethanolamine phosphatides found in many gram-negative bacteria. Conversely, the phosphatidyl-glycerol family occupies a more prominent place in the bacilli, sometimes including aminoacyl phosphatidylglycerol or a glucosaminyl phosphatidylglycerol, as in B. meguteriurn (Fig. 7F). As in many gram-positive cocci, the acyl chains of the lipids of Bacillus are usually branched (60 to 90% of the total fatty acids) (Kaneda, 1977), plus straight-chain saturated, and in a number of species these are accompanied by monounsaturated fatty acids. As in the gram-positive cocci, there may be considerable variation in lipid composition within a species, depending on environmental factors and stage of growth. Fatty acid desaturation is under temperature control in some Bucillus species (Fulco, 1974). Changes in the branched-chain amino acid composition of the medium affects the ratio of iso- to anteiso-acyl chains in the lipids (Kaneda, 1977). The effects of phosphate and magnesium limitation on the composition of the polar lipids of Bucillus can also be quite dramatic. In cheniostat cultures of B . sirhtilis (Marburg), phosphate limitation led to an increase in diglucosyldiglyceride at the expense of the phospholipids, especially phosphatidylethanolamine (Minnikin et ul., 1972). Less striking quantitative changes were noted in studies of B . sirbtilis var. niger, in which phosphate limitation led to an increased ratio of cardiolipin to phosphatidylglycerol at pH 7.0. Mg2’ -limited cultures had relatively more phosphatidylglycerol than phosphate-limited cultures, and phosphatidylethanolamine and lysyl-phosphatidylglycerol were not seen at pH 8.0 (Minnikin and Abdolrahimzadeh, 1974b). Although there has been considerable work on the acyl chains of clostridia (Goldfine, 1964; Moss and Lewis, 1967; Chan et a / . , 1971), and on the plasmalogen content of the lipids (Baumann et ul., 1965; Kamio et ul., 1969), there has not been sufficient work on the intact lipids to derive a general picture. However, it is already clear that there will be differences in the phospholipid cornposition of members of this genus when such a picture emerges. Baumann et al. ( 1965) found both diacyl and plasmalogen forms of phosphatidylethanolamine phosphatidyl-N-methylethanolamine, and phosphatidylglycerol in C.

28

HOWARD GOLDFINE

hutyricum, and a second ether lipid type in this organism was later characterized as a glycerol acetal of the ethanolamine and N-methylethanolamine plasmalogens (see Section II,B,l) (Matsumoto er al., 1971; Khuller and Goldfine, 1974). Macfarlane (1962) found a more gram-positive-like lipid pattern of C. welchii (perfringens), which had principally aminoacyl phosphatidylglycerol , phosphatidylglycerol , and cardiolipin. Interestingly, the presence of phosphatidylserine synthetase and decarboxylase in addition to phosphatidylglycero-P-synthetase was recently reported to this organism (Carman and Wieczorek, 1980).The relationship of these enzymes to the lipid composition of this organism remains to be clarified.

P. Gram-Positive, Non-spore-Forming Rods The lactobacilli are especially rich in phosphatidylglycerol and cardiolipin. As determined by :p2P,labeling, the range for 10 species was phosphatidylglycerol, 55-83% of lipid P; and cardiolipin, 3-15%. Eight of ten species also had lysylphosphatidylglycerol, representing 3 to 32% of lipid P. L . plantarum (Fig. 7G), therefore, presents a typical pattern (Exterkate et al., 1971). In addition to phospholipids, lactobacilli also have galactosylglucosyldiglycerides (Shaw, 1975). It should also be noted that lactobacilli lipids contain the saturated, monounsaturated, and cyclopropane fatty acids more typical of gram-negative than gram-positive bacteria. The related B$dobacterium, which is now classified with the actinomycetes (Buchanan and Gibbons, 1974), differs from the lactobacilli in having much more cardiolipin, less phosphatidylglycerol, a galactosyldiglyceride with an sn-glycerol-1-P substituted sugar, and alanylphosphatidylglycerol (Exterkate et al., 1971; Veerkamp and van Shaik, 1974).

Q. Actinomycetes and Related Organisms 1 . CORYNEBACTERIA The corynebacteria are a large and apparently heterogeneous group of organisms whose taxonomy is still in a state of flux. Recent work on the fatty acids, phospholipids, and mycolic acids of these organisms has provided information of taxonomic importance. One group of organisms, of which C . diphtheriae is an example, has 52-60% GC in its DNA, meso-diaminopimelic acid in its peptidoglycans, corynomycolic acids, and saturated and monounsaturated straight-chain fatty acids in their extractable lipids. There is a high cardiolipin to phosphatidylglycerol ratio, and phosphatidylinositol dimannosides (PIM) among the extractable polar lipids. They may in addition have phosphatidylinositol and phosphatidylethanolamine (Komura et al., 1975b; Lechevalier, 1977). A second


29

group has 69-70% GC in its DNA, diaminobutyric acid in its peptidoglycan, no mycolic acids, and mainly unteiso-15:0, unteiso-l7:0, and iso-16:O fatty acids. The polar lipids are largely cardiolipin, phosphatidylglycerol, and one to three uncharacterized, probably mannose-containing glycolipids. The latter are not of the PTM type (Komura et a l . , 197Sb; Collins and Jones, 1980). This group contains several plant pathogens such as C. tritici, C. imnicurn, C . sepedonicum, and C . rnichigutierrse. A third group, which has 66-71% GC, and ornithine in its cell wall, has an even higher proportion of a n t e i x - and iso-branched fatty acids, and the same phospholipids and several uncharacterized glycolipids as in the second group (Collins et a / . , 1980). These authors have suggested that on the basis of their cell wall and lipid compositions, C . fklccumfaciens, C. poinsettiae, C . bctue, and C . oortii should probably be included in the genus Curtobucterium. 2.

A R T HROBAC rER A N D PROPlONlBACTERl UM

These organisms also have large amounts of branched-chain fatty acids in addition to straight-chain saturated fatty acids. Arthrobacter phospholipids, like those of the other actinomycete-related groups, consist of a high proportion of cardiolipin, less phosphatidylglycerol, PIM, and in some organisms, phosphatidylinositol (Komura et a / . , 197Sb; Lechevalier, 1977). Three species were shown to have both galactosyldiglyceride and mannosyldiglyceride by Shaw and Stead ( 1 971). In A . crystallopoietes the glycolipids represented approximately 25% of the total lipid. The lipids of propionibacteria have not been extensively studied. P . shermunii was found to contain phosphatidylglycerol, phosphatidylinositol, and a diacylinositol mannoside. The presence of PIM in these organisms has not been established (Shaw, 1975). An isolated report of the presence of substantial amounts of aldehydogenic lipids, presumably plasmalogens, in anaerobic propionibacteria (Kamio e t a / . , 1969) has not been confirmed, as yet. 3 . ACTINOMYCEI ALES u . Mycobucrerium. The cell wall of mycobacteria has been an object of intense interest for many years because of its hydrophobic nature and its presumed role in the pathogenicity of these organisms. Underlying the wall is a typical 8-nm-thick cytoplasmic membrane (Ratledge, 1976). Since the cell wall is complex in structure and contains a number of unusual lipids of high molecular weight, it has not been easy to define the components of the membrane, the object of our interest, and distinguish them from the wall lipids. It is known that certain of the complex lipids such as the mycolic acids (see Section I,C,2), are covalently linked to the peptidoglycan through an arabinogalactan in the wax D structure (Goren, 1972;

30

HOWARD GOLDFINE

Ratledge, 1976). The mycosides, which are either phenolic glycolipids (mycosides A and B) or peptidoglycolipids (mycosides C ) , are also thought to be located peripherally (Goren, 1972). The major phospholipid, cardiolipin, is thought to be localized principally in the cell membrane, whereas the PIM are mostly associated with the cell wall (Akamatsu er ul., 1966). In this study phosphatidylethanolamine was found to be only slightly enriched in the cytoplasmic membrane fractions relative to the cell wall fraction. It is possible that, as in the outer membranes of gram-negative cells, certain phospholipids are associated with glycolipids in a wall fraction of the mycobacteria. Unlike the mycolic acids, with their very long-chain fatty acids, the phospholipids of mycobacteria are rich in acyl chains of ordinary length with C,, and C,, saturated and monounsaturated chains predominating (Goren, 1972). In addition, tuberculostearic acid (10-methylstearic acid) is widely distributed among actinomycetes and related groups of organisms (Lechevalier, 1977). Mycobacteria are relatively rich in triglycerides, and longer-chain (CZoto C Z 6 fatty ) acids are found on the 3-position of glycerol in these lipids (Ratledge, 1976). h. Nocurdiu. Like corynebacteria and mycobacteria, nocardia have complex nocardomycolic acids, which have chains of intermediate length between those of the longer eumycolates of mycobacteria and the shorter corynomycolates (Lechevalier, 1977). Their phospholipids are also similar, with cardiolipin and PIM predominating, along with somewhat less phosphatidylethanolamine (Khuller, 1977; Trana e t a / . . 1980; Lechevalier, 1977). In addition, small amounts of :v2Pi-labeledphosphatidylglycerol were found in all species of nocardia examined by Komura et ul. (1975b). One species, N . coefiucu, contains phosphatidylcholine as a major phospholipid (Yano e t u f . , 1969; Khuller and Brennan, 1972). Acylated trehaloses (cord factor) have also been found in nocardia (Lechevalier, 1977), strengthening the relationship of this group of organisms to the mycobacteria and corynebacteria. c. Strepiompces. The lipids of Streptomyces were recently reviewed (Batrakov and Bergelson, 1978). As in the related actinomycetes, cardiolipin, phosphatidylethanolamine, and PIM are the major phospholipids (Batrakov and Bergelson, 1978). In some species cardiolipin predominates; however, in S. griseus, phosphatidylethanolamine is 30-40% of total phospholipid, depending on the age of the culture (Talwar and Khuller, 1977). In this study PIM increased from 14.5 to 24% of total phospholipids as the cultures aged. The streptomyces may also contain several unusual polar lipids, such as the finding in one species of a butane-2,3-diol analogue of phosphatidylglycerol, in which the 4-carbon analogue replaces the unacylated glycerol (Table 11). An ornithine lipid of the zwitterionic type (Fig. 3B) has been found in two Streptomyces species, in one of which it appears to replace phosphatidylethanolamine (Batrakov and Bergelson, 1978). These authors have shown that in Actinomyces (Streptomyces) olivaceus the ornithine lipid can replace phosphatidylethanolamine when the cells are


31

grown in phosphate-deficient media. Some streptomyces have lysine lipids that resemble the ornithine lipids with substituents at both the carboxyl and a-amino groupings. The ratio of ornithine to lysine lipids varies in S. sioyuensis (Kimura and Otsuka, 1969).

R. Rickettsia These obligate, intracellular parasites have a cell envelope that resembles that of gram-negative species in its morphology. In addition, they have muramic acid and diaminopimelic acid, which are characteristic of prokaryotic cell walls (Buchanan and Gibbons, 1974). Rickertsiu prowazeki grown in chicken embryo yolk sacs has a typical gram-negative phospholipid composition consisting of phosphatidylethanolamine (60-70%), phosphatidylglycerol (20%), and phosphatidylcholine ( 15%). Small amounts of phosphatidylserine and cardiolipin were also detected (Winkler and Miller, 1978). Since these authors believe that the phosphatidylcholine is host derived based on its :"P-specific activity in labeling experiments, some caution must be exercised in the evaluation of the ratios of the other phospholipids.

S. Mycoplasma The polar lipids of the cell wall-less Mycoplasrna resemble those of grampositive bacteria in their high concentration of phosphatidylglycerol, cardiolipin, and glycosyldiglycerides. Phosphoglycolipids (Section 111,N; Fig. 2 ) have also been identified (Razin, 1978; Smith, 1979).

IV.

PROKARYOTIC LIPIDS AND PHYLOGENY

The outlines of prokaryotic membrane lipid compositions have emerged with increasing clarity during the past decade. The organization of Section 111 was based on the current arrangement of bacteria in Bergey's Manual (Buchanan and Gibbons, 1974), which is divided on pragmatic grounds into 19 groups and does not indicate the relatedness of the various groups of prokaryotes. It is, therefore, of some interest to examine these bacterial lipid compositions in the light of current work on bacterial phylogeny. The recent summary of the work of Fox, Woese, Wolfe and their colleagues, which was previously scattered in a number of papers, presents the opportunity to make such a comparison (Fox et ul., 1980). This phylogeny has been constructed on the basis of an extensive examination of 16 S ribosomal RNA sequences. It departs in several important ways from traditional taxonomies. For example, cell shape is shown not to be a

32

HOWARD GOLDFINE

workable criterion for relatedness; most spherical bacteria are seen to fall into groupings defined by nonspherical organisms. Mycoplasma, which had been assigned a distinct division in earlier phylogenies, are considered by Fox et (11. ( I 980) to be wall-less offshoots of the clostridial branch (see below). There are other recent phylogenies based on such molecular characteristics of prokaryotes as various protein sequences, cell wall analyses, 5 S rRNA, and DNA-RNA hybridization. Fox e1 ul. (1980) state that their scheme is in reasonable agreement with most of them. As can be seen in Fig. 8, the 16 S rRNA data indicate that most present-day gram-negative organisms have descended from a common ancestral group of purple photosynthetic bacteria. From this trunk there are three major branches. One includes Purucoccus and Rhizobi~ctn along with the purple nonsulfur Rhodopseudotrionus species. An examination of the lipids of these organisms (Section III,B) reveals that in addition to the usual phosphatidylethanolamine, phosphatidylglycerol, and cardiolipin of most gram-negative organisms, many organisms in this subline also have phosphatidylcholine, which is sometimes accompanied by an ornithine lipid. In another major branch derived from the purple bacteria are grouped the enteric organisms, the pseudomonads, and the vibrios along with the purple sulfur bacteria. Of the two sublines of this branch, the enteric line has organisms that generally have the simplest gram-negative polar lipid composition consisting of phosphatidylethanolamine plus polyglycerolphosphatides, whereas the other major subline, which contains the pseudomonads and acinetobacter, includes a few organisms that have a small proportion of phosphatidylcholine or zwitterionic ornithine lipids in addition to the common gram-negative phospholipids. Descr@vibrio, a separate branch, contains species with the simplest gramnegative lipid pattern and one species, D. gigus, that has a large amount of a zwitterionic ornithine lipid. Thus we see that all the gram-negative bacteria presumably derived from purple bacterial ancestors have a common pattern of phosphatidylethanolamine plus polyglycerolphosphatides, to which have accreted in some branches and sub-branches the enzymatic capacities to N-methylate phosphatidylethanolamine and to form the ornithine lipids. According to Fox et (11. (1980) the cyanobacteria (blue-green algae) and the green sulfur bacteria form separate evolutionary groups, and their chloroplastlike lipid compositions reflect this. The major extractable polar lipids are galactosyldigl yceride, pol ygl ycerolphosphatides , and SQDG . Phosphatidylethanolamine and phosphatidylcholine have not been found in these organisms. The spirochaetes are also considered to be in a separate group. As noted above (Section III,E), the lipids of the Spirochaeta are similar to those of gram-positive bacteria. They contain polyglycerol phosphatides, glycosyldiglycerides, and no ethanolamine phosphatides. The Treponernu, which do not appear in this scheme, have both ethanolamine and choline phosphatides along with glycosyl-

33


Thcrmophilcr

1

ancestral state

FIG. R . The phylogeny of prokaryotes according to Fox et ul. (1980). Copyright 1980 by the American Association for the Advancement of Science.

diglycerides, and may, therefore, be more closely related to grani-negative organisms. Lcptospiru, as noted by Fox et u / . (1980), do not cluster with Spirochuetu, on the basis of their 16 S rRNA. They also differ in their lipid composition, which is much like that of gram-negative organisms (Fig. 5D, Table 111). The other major group of eubacteria in the scheme of Fox rt ul. (1980) includes all gram-positive eubacteria except for a sniall group of cocci. From this gram-positive trunk two major branches emerge. One includes the “actinomyces” group, which gives rise to one line that includes the arthrobacter group, a second that includes the Cor?.tiehacteriurn -M~cohuc.trriurii-Noc~urdiu (CMN) group, and a third, the streptomyces group. We have seen (Section JIJ,Q) how similar the extractable lipids of these three major groups of organisms are. They generally have a high proportion of cardiolipin and PIM. They may also have phosphatidylinositol in small amounts. Phosphatidylethanolamine appears in the mycobacteria, nocardia, and streptomyces. The CMN group is also distinguished by the presence of their unique mycolic acids.

34

HOWARD GOLDFINE

Another major branch of the gram-positive trunk is the ancestral “clostridial” line, which gives rise to one branch including both the genus Bacillus and the lactobacilli. The “intermediate” character of the lipids of the genus Bacillus was noted in Section 111,O. In composition they lie between the major groups of gram-negative and gram-positive organisms. In most of these organisms, there is a high proportion of both phosphatidylethanolamine and the polyglycerolphosphatides, characteristically there are diglucosyldiglycerides, and in some species either aminoacyl- or glucosaminyl-phosphatidylglycerol.It should be noted that Fox et al. ( 1980) also consider Staphylococcus epidermidis and Streptococcus luctis as close relatives of Bacillus. Streptococci generally have mainly polyglycerolphosphatides and glycosyldiglycerides, and in this regard resemble the lactobacilli. They also do not have the branched-chain fatty acids characteristic of Bacillus and of staphylococci. Both the bacilli and clostridia are relatively “deep” genera in that the association coefficients of the 16 S rRNA sequences indicate considerable evolutionary distances (Fox er al., 1980). The relatively incomplete information we currently have on the lipids of the clostridia similarly indicate considerable divergencies. Fox et al. (1980) consider the mycoplasma to be a subgroup of clostridia. Their gram-positive lipid composition, consisting largely of polyglycerolphosphatides plus glycolipids (Razin, 1978), is consistent with this grouping. The presence of large amounts of sterols in many mycoplasma is discussed by Razin (this volume). The “archaebacteria, ” originally separated from the main lines of bacterial descent by Woese, Fox, Wolfe, and their colleagues (Fox et al., 1980) on the basis of their cell walls, which contain no muramic acid; their tRNAs, which differ in the thymine, pseudouridine, cytidine loop; their distinctive RNA polymerases; and in the presence of several unusual coenzymes, also have unusual phospholipids. These are described by Langworthy (this volume). Thus, for all major prokaryotic lines of descent, membrane lipid compositions agree well with the phylogenetic tree proposed by Fox et al. (1980). The placement of the bacteroids with their unusual sphingolipids and the anaerobic gramnegative organisms with their alk-1-enyl acyl lipids, in this evolutionary scheme, will be of considerable future interest.

V. CONCLUSIONS The diversity of prokaryotic lipids stands as further testimony to the great age of this group of organisms. Similar conclusions have come forth from studies of other cellular macromolecules (see Fox et a l . , 1980, for references), the geological record, and from a consideration of biosynthetic pathways (Goldfine and Bloch, 1963). At the present time, it is more difficult to relate prokaryotic lipid composition


35

and membrane function. The presence of acyl and alk-I-enyl chains containing double bonds, methyl branches, or small rings is critical for the maintenance of membrane fluidity (Melchior, this volume), which in turn is important for the function of membrane enzymes and transport systems. Variation in polar head groups may also affect membrane fluidity and the association of certain lipid classes with these enzymes and transport systems, and in some cases the specificity of these systems for a particular lipid class may be of equal significance (McElhaney , this volume). In this chapter the location of membrane lipids with respect to intracellular membranes has been discussed briefly. It is now becoming evident that lipids are often distributed asymmetrically in the two leaflets of a given membrane. This topology is described in detail by Rottem (this volume). As organisms evolved, the need for elaboration of the cytoplasmic membranes into specialized intracytoplasmic membranes containing increased surface area for photoreception or greater electron-transport capacity may have required the evolution of lipids that permit or facilitate the required membrane infolding and pinching off. We have previously discussed the much greater frequency of occurrence of phosphatidylcholine in organisms with complex intracytoplasmic membranes (Hagen et al., 1966). A group of organisms that provides an exception to this correlation is the Rhizobiaceae (Section III,G,3). These organisms interact with their host plant cells in order to form tumors or root nodules. It is tempting to speculate that the evolution of the phosphatidylethanolamine methylation pathway in this group has served to promote these bacteria-host interactions. Goren (1977) has reviewed the evidence that the sulfatides of mycobacteria (Section II,C,2) aid in the intracellular growth of pathogenic strains by interfering with phagolysosome formation. As noted in Section 111, some organisms have evolved dual lipid-biosynthetic capacities, which allow them to substitute glycosyldiglycerides or ornithine lipids for phosphatidylethanolamine in the absence of phosphate. It is clear that within limits of size and charge, polar lipids may replace one another in biological membranes. Exogenous fatty acyl chains and alk-1-enyl chains may also be substituted for the naturally occurring chains in bacterial auxotrophs (chapters by Melchior and McElhaney, this volume). Much more work on membrane mutants will be needed before a more complete understanding of the multifaceted roles of prokaryotic lipids can be attained. Work on prokaryotic organisms has not only provided an abundance of new insights into their membrane structure and functions, it is also leading the way to a more complete understanding of eukaryotic cell membranes. ACKNOWLEDGMENTS

I should like to expresb my appreciation to Dr. G. P. Hazlewood for the use of material prior to publication, and to Roseann Femia for able assistance in the preparation of this manuscript.

36

HOWARD GOLDFINE

REFERENCES Agate, A. D., and Vishniac, W. (1973). Changes in phospholipid composition of 7hiubaeillus neapolitunus during growth. Arch. Mikrohiol. 89, 247-255. Akamatsu, Y . , Ono, Y . , and Nojima, S. (1966). Phospholipid patterns in subcellular fractions of Mycobucteriirrn phlei. J . Biochern. (Tokyo) 59, 176-1 82. Ames, G. F. (1968). Lipids of Subnoriella ~phimirriurti and Escherichia coli: Structure and metabolism. J . Bacteriul. 95, 833-843. Anderson, R., Kates, M., and Volcani, B. E. (1978). Identification of the sulfolipids in the nonphotosynthetic diatom Nitzschin alba. Biochim. Biophys. Actu 528, 89-106. Asselineau, C. P., Montrozier, H. L., Prome, J.-C., Savagnac, A. M., and Welby, M. (1972). Etude d’un glycolipide polyinsature synthetise par Mycobacterium phlei. Eur. J . Biochem. 28, 102109. Auran, T. B., and Schmidt, E. L. (1972). Similarities between Hyphomicrobium and Nitrohacter with respect to fatty acids. J . Bucteriol. 109, 450-451. Barridge, J. K., and Shively, J . M. (1968). Phospholipids of the thiobacilli. J . Bucteriul. 95, 2182-2185. Batrakov, S . G . , and Bergelson, L. D. (1978). Lipids of the Streptornycetes Structural investigation and biological interrelation. Chem. Phys. Lipids 21, 1-29. Baumann, N . A , , Hagen, P.-O., and Goldfine, H. (1965). Phospholipids of Clostridium huyricurn. Studies on plasmalogcn composition and biosynthesis. J . B i d . Chern. 240, 1559-1567. Beebe, J . L., and Wlodkowski, T . J . (1976). Lipids of Branhumella cntarrhalis and Neisseriu goriorrhoeae. J . Bucteriol. 127, 168-178. Beining, P. R., Huff, E., Prescott, B., and Theodore, T. S . (1975). Characterization of the lipids of mesosoinal vesicles and plasma membranes from Staphylococcus aurerrs. J . Bacteriol. 121, 137-143. Bertsch, L. L., Bonsen, P. P. M . , and Kornberg, A. (1969). Biochemical studies of bacterial sporulation and germination XIV. Phospholipids in Bacillus rnegateriitrn. J . Bacteriol. 98, 75-81. Bishop, D. G . , Rutberg, L., and Samuelsson, B. (1967). The chemical composition of the cytoplasmic membrane of Burillus subtilis. Eur. J . Biocehm. 2, 448-453. Braun, V . , and Hantke, K. (1974). Biochemistry of bacterial cell envelopes Annu. Rev. Biuchern. 43, 89-121. Brooks, J . L., and Benson, A . A. (1972). Studies on the structure of an ornithine-containing lipid from Rhodospirillum rubrum. Arch. Biochern. Biophys. 152, 347-355. Brundish, D. E., Shaw, N., and Baddiley, J. (1967). The phospholipids of Pneumococcus I-I92R, A.T.C.C. 12213. Some structural rearrangements occurring under mild conditions. Biochem. J . 104, 205-211. Buchanan, R. E., and Gibbons, N. E., eds. (1974). “Bergey’s Manual of Determinative Bacteriology,” 8th ed. Williams & Wilkins, Baltimore, Maryland. Bunn, C. R . , and Elkan. G. H. (1971). The phospholipid composition of Rhizobium ,japonicum. Can. J . Microbiol. 17, 291-295. Carman, G. M., and Wieczorek, D. S . (1980). Phosphatidylglycerophosphate synthase and phosphatidylserine synthase activities in Clostridiurn perfringens. J . Bacteriol. 142, 262-267. Carson, D., Pieringer, R. A., and Daneo-Moore, L. (1979). Effect of growth rate on lipid and lipoteichoic acid composition in Streptococcus ,faeciurn. Biochim. Biophys. Acta 575, 225233. Chan, M., Himes, R. H., and Akagi, J . M. (1971). Fatty acid composition of thermophilic, mesophilic, and psychrophilic clostridia. J . Bacteriol. 106, 876-881. Chiu, T.-H., and Hung. S.-A. (1979). Effect of age on the membrane lipid conrposition of Streptococcus sariguis. Biochirn. Biophys. Acta 558, 267-272.


37

Clarke, N. G., Hazlewood, G . P., and Dawson, R. M. C. (1976). Novel lipids of Butyrivibrio spp. Chem. Phys. Lipids 17, 222-232. Clarke, N . G., Hazlewood, G. P . , and Dawson, R. M. C. (1980). Structure of diabolic acidcontaining phospholipids isolated from Butyrivibrio Sp. Biochem. J . 191, 561-569. Collins, M. D., and Jones, D. (1980). Lipids in the classification and identification of coryneform bacteria containing peptidoglycan based on 2,4-diaminobutyric acid. J . Appl. Bacteriol. 48, 459--470. Collins, M . D., Goodfellow, M., and Minnikin, D. E. (1980). Fatty acid, isoprenoid quinone, and polar lipid composition in the classification of Curtohacreriurn and related taxa. J . Gen. Microhiol. 118, 29-37. Contreras, I . , Shapiro, L., and Henry, S . ( 1 978). Membrane phospholipid composition of Caulohacter crcsceiztus. J . Bucteriol. 135, 1130-1 136. Dawson, R. M. C., Hemington, N., and Davenport, J . B. (1962). Improvements in the method of determining individual phospholipids in a complex mixture by successive chemical hydrolyses. Biochem. J . 84, 497-501. Depinto, J . A. (1967). Omithine-containing lipid in Rhodospirillum rubrum. Biochim. Biophys. Actu 144, 113-117. DeSiervo, A. J . , and Homola, A. D. (1980). Analysis of Cuulubacter crescenrus lipids. J . Bucteriol. 143, 1215-1222. Diedrich, D. L., and Cota-Robles, E. H. (1974). Heterogeneity in lipid composition of the outer membrane and cytoplasmic membrane of Pseutlomonas BAL-31. J . Barterid. 119, 10061018. Exterkate, F. A , . Otten, B. J . , Wassenberg, H. W . , and Veerkamp, J. H. (1971). Comparison ofthe phospholipid composition of Bifidohurrerium and Lactobacillus strains. J . Bucteriol. 106, 824-829. Faizova, G. K., Borodulina, Y . S . , and Samsonova, S . P. (1971). Lipids in nodular bacteria (Rhizohium leguminosurum). Microbiology ( E n g l . Trunsl.) 40, 41 1-413. Fischer, W . , Koch, H. U . , Rosel, P. Fiedler, F., and Schmuck, L. (1980). Structural requirementsof lipoteichoic acid carrier for recognition by the poly (ribitol phosphate) polymerase from Stuphy/ococcus uureus H. A study of various lipoteichoic acids, derivatives, and related compounds. J . B i d . Chem. 255, 4550-4556. Fox, G. E . , Stackebrandt, E . . Hespell. R. B., Gibson, J.. Maniloff, J . , Dyer, T. A., Wolfe. R . S . , Balch, W. E., Tanner, R . S . , Magrum, L. J . , Zablem, L. B., Blakemore, R., Gupta, R., Bonen, L., Lewis, B. J . , Stahl, D. A . , Luehrsen. K . R., Chen, K. N., and Woese, C. R . (1980). The phylogeny of prokaryotes. Science 209,457-463. Fritsche, D . , and Thelen, A. (1973). Die Abgrenzung der Genera Bacteroides and Sphaerophorus auf komplexen Lipoide. Zentrulhl. Bakteriol., Hyg. Parusitenkd. Infectionkr. Abt. I : Orig., Reihr A 223, 356-365. Fulco, A. J. (1974). Metabolic alterations of fatty acids. Annu. Rev. Biochem. 43, 215-241. Gerson, T . , and Patcl, J . J . (1975). Neutral lipids and phospholipids of free-living and bacteroid forms of two strains of Rhiiohirrm infective on Lorus peduneularus. A p p l . Microbid. 30, 193- 198. Godchaux, W . , 111, and Leadbetter, E. R . (1980). Capnocytophaga spp. contain sulfano-lipids that are novel in procaryotes. J . Barterid. 144, 592-602. Godchaux, W . , I l l , and Leadbetter, E. R. (1981). Sulfonolipids of gliding bacteria: Structure of N-acylcapnine. Fed. Proc.. Fed. A m . So(,. E x p . B i d . 40, 1845. Goldberg, I . , and Jensen, A . P. (1977). Phospholipid and fatty acid composition of methanolutilizing bacteria. J . Bacteriol. 130, 535-537. Goldfine, H. ( 1964). Composition of the aldehydes of Clostridium butyricum plasmalogens. Cyclopropane aldehydes. J . B i d . Chem. 239, 2130-2134. I

38

HOWARD GOLDFINE

Goldfine. H. (1972). Comparative aspects of bacterial lipids. A t / \ , . Micro/?. Physiol. 8, 1-58. Goldfine, H , and Bloch, K . (1963).Oxygen and biosynthetic reactions. I n “Control Mechanisnls in Respiration and Fermentation” (B. Wright. e d . ) , pp. 81-103. Ronald Press, New York. Goldfine, H.. and Ellis, M. E. (1964). N-Methyl groups in bacterial lipids. J . Btrcteriol. 87, 8-15. Goldfine, H., and Hagen, P:0. (1968). N-Methyl groups in bacterial lipids. 111. Phospholipids of hyphomicrobia. J . Btrcreriol. 95, 367-375. Goldfine, H , and Hagen. P - 0 . ( 1972). Bacterial plasmalogens. I n “Ether Lipids: Chemistry and Biology” (F. I.. Snyder. e d . ) . pp. 329-350. Academic Press, New York. Gorchein, A . ( 1964). Ornithine in R h o ~ / o ~ ~ s r i r t / ~ ~ myizeroitks. o,,lrs Bioclrirn. Hiophys. Acfa 84, 356-358. Gorchein, A . (1968). The separation and identification of the lipids of R h ~ J d ~ ~ ~ . ~ ~ l t d ~ J ~ ? l ~ J , r c l . c bpheroides. Proc. K . Soc. London, Ser.. B 170, 279-297. Goren, M. B. (1972). Mycobacterial lipids: Selected topics. B u r r e r i d . Re,,. 36, 33-64. Goren, M . B. (1977). Phagocyte lysosomes: Interaction with infectious agents, phagosomes, and experimental perturbations in function. Annu. Rei.. Microhid. 31, 507-533. Gottfried, E. L., and Rapport, M. M. (1962). The biochemistry of plasmalogens. I . Isolation and characterization of phosphatidal choline, a pure native plasmalogen. J . B i o / . Chern. 237, 329-333. Could. R . M. and Lennar7, W. J . (1970). Metabolism OF phosphatidylglycerol and lycyl phosphatidylglycerol in Sffipl7yioeuc.c.trsuureus. 1. Bucterinl. 104, 1 135- I 144. Hagen. P.-0.. and Blank, M. L. (1Y70). Minor lipid components of Clostridium huQv+rrm. J . Am. Oil Clirrn. S o r . . 47, 83A. Hagen, P.-O., and Goldfine, H (1967). Phospholipids of CIosrridiirrn huryricurn. Further studies on the origin of the aldehyde chains of plasmalogens. J . B i d . C‘herrr. 242, 5700-5708. Hagen, P.-O Goldfine, H., and Williams, P. J . L. (1966). Phospholipids of bacteria with extenaive intracytoplasmic membranes. Science 151, 1543-1544. Hanahan, D. J . (1972). Ether-linked lipids: Chemistry and methods of measurement. It7 “Ether Lipids: Chemistry and Biology” (F. L. Snyder, ed.), pp. 25-50. Academic Press, New York. Hancock. 1. C . , and Meadow, P. M. (1969). The extractable lipids of Psertdumunus urrrrginoscr. Riochirn. Biopl7y.s. Actti 187, 366-378 Haverkate, F., Teulings, F. A . G. and van Deenen, L. L. M . (1965). Studieson the phospholipidsof photosynthetic microorganisms. / I . Proc. K . Ned. A k u d W ~ I LSer. B : PI7y.5. Sci. 68, 154-

.

159.

Hazlewood, G. P., Clarke, N . G . . and Dawson, R . M. C. (1980). Complex lipids of a lipolytic and general-fatty-acid-requiring Burj’rii’ihrio sp. isolated from the ovine rumen. Biochem. J . 191, 555 -560. Heefner. D. L . , and Claus, G. W. (1978). Lipid and fatty acid composition of Gluconoborrrr o.ry?.dum before and after intracytoplasmic membrane formation. J . Bucleriol. 134, 38-47. Hirayama, 0. (1968). Lipids and lipoprotein complex in Photosynthetic tissues. Part IV. Lipids and pigments of photosynthetic bacteria. Agric. Bio/. Chen7. 32, 34-41. Holt, S . C., I>oundowlakis. J., and Takacs, B . J. (1979). Phospholipid composition of gliding bacteria: Oral isolates of C ‘ ~ r / ~ ~ r ~ ~ ~ : \ ’ t ~compared ~ / J I t u g Nwith . S ~ ~ ~ r ~ ~ c : v t o p Infecr. h u g t r . lrnrnun, 26, 305-3 10. Houtsmuller, U . M . T.. and van Deenen, L. L. M. (1963). Metabolism of phospholipids in B~cil/u.\ cereus. Biochem. J . 88, 43P. Houtsmuller, U. M. T.. and van Deenen. L. L. M. (1965). On the amino acid esters of phospharidyl glycerol from bacteria. Bioc.hirn. B i o p h y . ~ .A d u 106, 564-576. IUPAC-IUB Commission on Biochemical Nomenclature. (1978). J . Lipid R r s . 19, 1 14-126. Johns, R . B . , and Perry, G . J . (1977). Lipids of the marine bacterium. Flrxihucter pdymorphu.\. Arch. Microhiol. 114, 267-271


39

Johnson, R. C., Livermore, B. P., Walhy, J . K . , and Jenkin, H. M. (1970a). Lipids of parasitic and saprophytic leptospires. Inficr. /nmirn. 2 , 286-291. Johnson, R. C., Livermore, B . P., Jenkin, H . M . , and Eggebraten, L. (1970h). Lipidsof Treponemcr p d l i d u m Kazan 5. Infecr. Immun. 2, 606-609. Jones, D. E., and Smith, J. D. (1979). Phospholipids ofthe differentiating Cuiilohucre Cun. J . Bioiliem. 57, 424-428. Joyce, G . H., Hamnmond. R . K . , and White, D . C. (1970). Changes in membrane lipid composition in exponentially growing S/erph?./eJc~oc.c.cf.\ urrrerrs during the shift from 37 to 25 C. J . Bucrerid. 104, 323-330. Jurtshuk, P., and Schlech. B . A. (1969). Phospholipids of Azorobucrer i i n e l u d i i . J . Bercrrriol. 97, 1507- 1508. Kamio, Y . , and Takahashi, H. (1980). Isolation and characterization of outer and inner membranes of Sc,lenovwriu.s nrminctnrirr,,i; Lipid composition. J . Bucrcriol. 141, 888-898. Kamio, Y . , Kanegasaki, S . , and Takahashi. H . (1969). Occurrence of plasmalogens in anaerobic bacteria. J . Geti. A p p l . Microhiol. 15, 439-451. Kaneda, T. (1977). Fatty aci,ds of the genus Bocillm; An example of branched-chain preference. Buireriol. Rev. 41, 39 1-4 18. Kaneshiro. T . , and Marr, A . G. ( 1962). Phospholipids of A:orohacter ugilis. Agrobucirrirrm rum&ciens, and E.sc,heric,hicr coli. J . Li[~ie/Re.\. 3, 184- IX9. Kenyon, C. N . (1972). Fatty acid composition ofuniccllular strains of blue-green algae. J . Bucteriol. 109, 827-834. Kenyon, C . N . (1978). Complex lipids and fatty acids in photosynthetic bacteria. In "The Photosynthetic Bacteria" (R. K . Clayton and W. R. Sistronm. eds.), pp. 281-313. Plenum, New York. Kenyon, C. N . , and Stanier, R . Y . (1970). Possible evolutionary significance of polyunsaturated fatty acids in blue-green algae. Norrrre (London) 227, 1164-1 166. Kenyon, C. N . , Rippka, R . , and Stanier, R . Y. (1972). Fatty acid compobition and phyaiological properties of some filamentous blue-green algae. Arch. Mikrohini. 83, 216-236. Khuller, C. K . (1977). Phospholipid composition of Norccrrcliu species. Iritlicrn J . Meel. Kes. 65, 657-660. Khuller, G. K . , and Brennan. P. J . (1972). The polar lipids of some species of Norurcfiu. J . Gen. Microhid. 73, 409-41 2 . Khuller, G. K . , and Goldfine, H . (1974). Phospholipids of Clo.\rrrcliron huhricirni V . Effects of growth temperature on fatty acid, alk-I-enyl ether group, and phospholipid cornposition. J . Lipid R e s . 15, S00-507. Kim, K . C.. Kamio, Y . , and Takahashi, H . (1970). Glyceryl ether phospholipid in anaerobic bacteria. J . Gen. A/?/?/.M i c ~ ~ h i o16, l . 321 -325. Kimura, A , , and Otsuka, H . ( 1969). The changes o f lysine-and ornithine-lipids i n S/repfo/n.wes .sio>~ueri.sis. Agric. B i d Chcni. 33, 781 -7X4. Klein. R . A , , Hazlewood, G . P.. Kemp. P.. and Dawson. R . M . C. (1979). A new series of long-chain dicarboxylic acids with vlcinal dimethyl branching found as major components of thc lipids of BitrJriL.ihrio sp. Hioc-hcn. J . 183, 691-700. Komura, I . , Yamada, K . , and Konmagata. K . (197%). Taxonomic significance of phospholipid composition in aerobic gram-po\itive cocci. J . Gen. A/'[?/. Mie,rohin/. 21, 97- 107. Komura, I.. Yamada, K . . Shin-lchiro. 0.. and Koniagata, K . (1975b). Taxonomic significance of pho,pholipid\ in coryneforin and nocardiofornm bacteria. J . Gen. A / ? / > / .Mic~rohiol.21, 25 I 261. Koplow, J . , and Goldfine. H . (1974). Alterations i n the outer membrane of the cell envelope of heptose-deficient mutants o f E . s d t e r i ~ ~ / icid~ir. J . Bcrc,rerio/. 117, 527-543.

40

HOWARD GOLDFINE

Lambert, M. A. S . , and Arnifield, A . Y . (1979). Differentiation ot Peptoc~occusand Peptosfreptococc'ur by gas-liquid chromatography of ccllular fatty acids and metabolic products. J . C h i . Microbiol. 10, 464-476. Law, J . H. (1971). Biosynrhesis of cyclopropane rings. A r r . Chem. Res. 4, 199-203. Lechevalier. M . P. (1977). Lipids in bacterial taxonomy-A taxonomist's view. CRC Crit. Rev. Mic,rohiul. 5 , 109-210. Lennarz, W . J . . and Talamo. B. (1966). The chemical characterization and enzymatic synthesis of mannolipids in MicrococcuJ lysodeikricus. J . Biol. Chern. 241, 2707-2719. Livermore. B . P., and Johnson, R. C. (1974). Lipids ofthe Spirochuetalest Comparison of the lipids of several members of the genera Spirochaeta. Treponema, and Leptospira. J . Bucleriol. 120, 1268-1273. Livermore. B. P . Bey, K. F.. and Johnson, R. C. (1978). Lipid metabolism of Borrelia hennsi. /nject. /rwnun. 20, 215-220. Lugtenberg. E. J . J . . and Peters, R. (1976). Distribution of lipids in cytoplasmic and outer rnembranes of Escherichitr coli K12. Biochirn. Biophys. Acta 441, 38-47. Macfarlane, M. G . (1961a). Composition of lipid from protoplast membranes and whole cells of Micrucoi,cus l~sodeikticrrs.Biochem. J . 79, 4P-5P. Macfarlane, M. C. (1961b). Isolation of a phosphatidylglycerol and a glycolipid from Micrococcus lvsodeiktiws cells. Biuchem. J . 80, 45P. Macfarlane. M . G . (1962) Characterization of lipoamino-acids as 0-amino-acid esters of phosphatidylglycerol. Narurr (London) 196, 136- 138. Makula, R. A. ( 1978). Phospholipid composition of methane-utilizing bacteria J . Barterin/. 134, 77 1-777 Makula, R. A , , and Finnerty, W . R. (1970). Microbial assimilation of hydrocarbons. Identification of phospholipids. J . Burteriol. 103, 348-355. Makula, R. A., and Finnerty, W . R. (1974). Phospholipid composition of Dr.rulfovihrio species. J , Barrerid. 120, 1279-1283. Makula, R. A., and Finnerty, W . R. (1975). Isolation and characterization of an ornithine-containing lipid from Desulfovibrio gigas. J . Bacteriol. 123, 523-529. Matsurnoto, M., Tamiya, K., and Koizumi, K . (1971). Studies on neutral lipids and a new type of aldehydogenic ethanolamine phospholipid in Clostridium buryricum. J . Biochem. (Tokyo) 69, 61 7-620. Matthews, H. M., Yang, T.-K.. and Jenkin, H. M. (1979). Unique lipid composition of Treponerna p d l i d i m (Nichols virulent strain). Iqfect. lmmun. 24, 713-719. Mcrlie, J . P., and Pizer, L. I . (1973). Regulation of phospholipid synthesis in Escherichia coli by guanosine tetraphosphate. J . Bacterial. 116, 355-366. Meyer, H . , and Meyer, F. (1971). Lipid metabolism in the parasite and free-living spirochetes Treponemti pallidurn (Reiter) and Treponema zuelzerae. Biochim. Biophys. Acta 231,93- 106. Minnikin, D. E., and Abdolrahimzadeh, H. (1974a). The replacement of phosphatidylethanolamine and acidic phospholipids by an ornithine-amide lipid and a minor phosphorus-free lipid in Pseurlomonas~uorescensNCMB 129. FEES Lett. 43, 257-260. Minnikin, D. E., and Abdolrahimzadeh, H. (1974b). Effect of pH on the proponions of polar lipids, i n chemostat cultures of Bacillus subtilis. J . Bucteriol. 120, 999- 1003. Minnikin, D. E , Abdolrahimzadeh, H . , and Baddiley, J. (1972). Variation of polar lipid composition of Baci1lu.s srrhfilis (Marburg) with different growth conditions. FEES Lett. 27, 16-18. Miyagawa, E., Azunia, R., and Suto, T. (1978). Distribution of sphingolipids in Bacteroides species. J . Gen. Appl. Microbiol. 24, 341-348. Moss, C. W . , and Lewis, V . J . (1967). Characterization of clostridia by gas chromatography. I . Differentiation of species by cellular fatty acids. Appl. Microbiol. 15, 390-397.


41

Nichols, B . W., Harris, R. V . , and James, A. T. (1965). The lipid metabolism of blue-green algae. Biochem. Biophys. Res. Comtiruti. 20, 256-262. O’Leary, W. M. (1973). Lipoidal contents of specific microorganisms. I n “Handbook of Microbiology” (A. I. Laskin and H. A. Lechevalier, eds.), Val. 11, pp. 275-327. CRC Press, Cleveland, Ohio. Oliver, J . D., and Colwcll, R. K. (1973). Extractable lipids of‘ gram-negative marine bacteria: Phospholipid composition. J . Buctrriol. 114, 897-908. Omdorff, P. E., and Dworkin. M . (1980). Separation and properties of the cytoplasmic and outer membranes of vegatative cells of M~xococcrrsxcrirfhus. J . Buctrriol. 141, 914-927. Osborn, M. J . , Gander, J . E . , Parisi, E . , and Carson, J . (1972). Mechanism of assembly of the outer membrane of Salmonellu tvphimurium. J . B i d . Chrm. 247, 3962-3972. Patt, T . E . , and Hanson, R. S. (1978). lntracytoplasmic membrane, phospholipid and sterol content of Methylobac.ierium or,qatii~phi~rrrn cells grown tinder different conditions. J. Bac,frriol. 134, 636-644. Pieringer, R . A., and Ganfield, M.-C. W. (1975). Phosphatidylkojibiosyl diglyceride: Metabolisin and function as an anchor in bacterial cell membranes. Lipids 10, 421-426. Powell, D. A., Duckworth, M., and Baddiley, J . (1975). A membrane-associated lipomannan in micrococci. Biochem. J . 151, 387-397. Qureshi, N., Takayama, K . , Jordi, H. C . , and Schnoes, H. K . (1978). Characterization of the purifieti components of a new homologous series of a-mycolic acids from Mycvbacterium trrhercrrlosis H37Ra. J . Bin/. Chem. 253, 541 1-5417. Randle, C. L . , Albro, P. W . , and Dittmer, J. C. (1969). The phospholipid composition of gramnegative bacteria and the changes in composition during growth. Biochim. Biophys. Acta 187, 214-220. Ratledge, C. (1976). The physiology of the mycobacteria. Adv. Microb. Physiol. 13, 115-234. Razin, S. (1978). The mycoplasmas. Microbiol. Rev. 42, 414-470. Reusch. R. N . , and Sadoff, H. L. (1979). 5-n-Alkylresorcinols from encysting Azotobacter vinelundii: Isolation and characterization. J . Bactrriol. 144, 448-453 Reusch, R. N . , and Sadoff. H. L. (1981). Unique lipids in membranes of Aiofnhacter virielundii cysts. Absfr., Annu. Meer. Am. Soc. Micruhiol. p. 166. Rizza, V . , Tucker, A. N., and White, D. C. (1970). Lipids of Bacteruides meluninogenicus. J . Bucferiol. 101, 84-91. Rottem, S . , Hasin, M . , and Razin, S. (1975). The outer membrane of Proteus mirubilis. 11. The extractable lipid fraction and electron paramagnetic resonance analysis of the outer and cytoplasmic membranes. Biochim. Biophy. Acra 375, 395-405. Russell, N. J . , and Harwood, J . L. (1979). Changes in the acyl lipid composition of photosynthetic bacteria grown under photosynthetic and non-photosynthetic conditions. Biochem. J . 181, 339-345. Schmidt, M . F. G . , Brancha, M . , and Schlesinger, M. J . (1979). Evidence for covalent attachment of fatty acids to Sindbis Virus glycoproteins. Proc. Natl. Acad. Sci. U.S.A. 76, 1687-1691, Senff, L. M., Wegener, W . S., Brooks, G. F., Finnerty, W. R . , and Makula, R . A. (1976). Phospholipid composition and phospholipase A activity o f Neisseriu gonorrhoeae. J . Bacteriol. 127, 874-880. Shaw, N. (1974). Lipid composition as a guide to the classification of bacteria. Adv. Appl. Microbiol. 17, 63-108. Shaw, N. (1975). Bacterial glycolipids and gtycophospholipids. Adv. Microb. Physiot. 12, 141- 167. Shaw, N . , and Stead, D. (1971). Lipid composition of some species of Arthrobacter. J . Bacreriol. 107, 130-133. Shively, J . M., and Benson, A. A. (1967). Phospholipids of Thiobacillus thiooxidans. J . Bacferiol. 94, 1679-1683.

42

HOWARD GOLDFINE

Short, S . A.. White. D. C . , and Aleem, M. I . H. (1969). Phospholipid inetabolisni in Ferrohuci//u.\ /erromiduti.\. J . Buctcriol. 99, 142- I 50. Smith, P. F. (1979). The composition of membrane lipids and lipopolysaccharides. I n "The Mycoplasmas" (M. F. Barile and S . Razin. eds.), Vol. I , pp. 23 1-259. Academic Press. New York. Steck, P. A . , Schwartz. €3 A., Rosendahl, M. S . , and Gray. G. R . (1978). Mycolic acids. A reinvestigation. J . Biol. Chem. 253, 5625-5629. Steiner, S . , Sojka, C. A,, Conti. S . F., Geat, H., and Lester, R . L. (1970). Modification of membrane composition in growing photosynthetic bacteria. Biochirn. Biophys. Actu 203, 571 -574 Stoffel, W., Dittmar, K . , and Wilmes, R. (197% Sphingolipid metabolism in Bocteroideacear. HO{J{Jr-Sry/u,-'s z. Phxsiol. Chem. 356, 71 5-725. Sud. I. J . , and Feingold, D. S . (1975). Phospholipids and fatty acids of NtJissrritr gonorrhorue. J . Bucteriol. 124, 713-717. Sutrina, S . L.. and Scocca, J . J . (1976). Phospholipids of Harmophilus influrnzar Rd. during exponential growth and following the development of competence lor genetic transformation. J . & t i . Mictohrol. 92, 410-412. Tahara. Y . , Yamada, Y ., and Kondo, K . (1976). Phospholipid composition of G/uc.oriohncrur crrirrriA. A , y r w B i d . Chr/rt. 40, 2355-2360. 'l'akacs, B. J . . and Holt, S . C. (1971). Thioctrpscr /7orit/crnu; a cytological physical, and chemical characterization. I I . Physical and chemical characteristics of isolated and reconstituted chrornatophores. Biochim. Biophys. Actu 233, 278-295. Talwar, P., and Khuller. G. K . (1977). Effect of age on the major phospholipids of Streptotnwrs griseu.i. fndiun J . Bioc.hcw. Biophys. 14, 85-86. Thiele, 0. W., and Schwinn, G. (1973). The free lipids of Brucellu wic/ireii.sis and Bor-&rr//u /irr/rcssis. E u r . J . Bioc,hewt. 34, 333-344. Thomas. T . D., and Ellar, D. J. (1973). Properties of plasma and mesosomal membranes isolated from Microcacws /ysodrikticrts: Rates of synthesis and characterization of lipids Biochirn. Biophyr. Acru 316, 180-195. Thome. K . J . I . , Thornley. M. J . , and Glauert, A. M. (1973). Chemical analysis of the outer membrane and other layers of the cell envelope of Acirrerobuctrr sp. J . Bncteriol. 116, 41 0-41 7. Tornabene, T. C. (1973). Lipid composition of selected strains of Yersiriicr pestis and Yersinitr p,irudor~rhei-ert/oSiS.Biochirn. B i o p h w . Acttr 306, I 73- 185. Tornabene, T. G.. and Ogg. J . E. (1971). Chromatographic studies of the lipid components of Vihrio ,fetus. Biochir~t.Biophys. Actu 239, 133-141. Tornahene, T. G., Morrison, S. J . , and Kloos, W. E. (1970). Aliphatic hydrocarbon contents of various members of the family Micrococraceue. Lipids 5, 929-937. Trana, A. K . , Khuller, G. K . , and Subrahmanyam, D. (1980). Metabolism of phospholipids in Nocardia pcJ/yc~hmino~ene.r. J . Gen. Microhid. 116, 89-92. van Golde, L. M. G . , Akkermans-Kryawijk, J . , Franklin-Klein, W . , Lankhorst, A , , and Prins, R. A . (1975). Accumulation of phosphatidylserine in strictly anaerobic lactate fermenting bacteria. FEES Lett 53, 57-60. Veerkamp, J . H.. and vim Sheik, F. W . (1974). Biochemical changes in Brfidohuc,rcriurn hifidus var. penn.r$i~trnicits after cell wall inhibition. VII. Structure of the galactosyldiglycerides. Biochirr?. B i o p h n . Actu 348, 370-387. Verkley, A. J., Ververgaert, P. H. J. T., Prins, R. A., and van Golde, L. M. G. (1975). Lipid-phase transitions of the strictly anaerobic bacteria Veillonclla parvula and Anaeroiihrio ripolytica. J . B u c t ~ r i ~124, l . 1522-1528. Vorheck. M. L., and Marinetti, G. V . (1965). Intracellular distribution and characterization of the lipids of Srreptocuccit.s fiec~ulis (ATCC 9790). Biochemistry 4, 296-305.


43

Weaver. T . L., Patrick. M. A , . and Dugan, P. R. (1975). Whole-cell and membrane lipids o f t h e methylotrophic bacterium Mrt/?\/o.siurc.\ tri[./ro.s/’~~riri,,r. J . Hrrcter-fol. 124, 602-605 White. D. A , , Lennari, W . J . , and Schnaitman, C . A . (1971). Distribution of lipids in the wall and cytoplasmic mernbrane subfractions of the cell envelope of E.sc/rrric,/rftr< d . J . Bocrcriol. 109, 686-690. White, D. C., Tucker, A. N . , and Sweeley. C. C . (1969). Characterization of the iso-branched sphinganines IroiTi the ceramide phospholipids of Buc~teroit/~.s f ~ f ~ , / ~ f J f ; r r ~ ~ ~ ~ , Jh’ioc,/i;r?i. ii(,i~.\. Bio/)/r,ys. Actti 187, 527 532 Whiteside, 7 L., deSiervo, A . J . , and Salton, M . R . (1971). Use of antibody to membrane adenosine triphosphalase in the study of bacterial relatiomhipa. J . Bucterrd. 105, 957-967. Wilkinson. S . G . (1970).Cell walk of P.\c’rrc/oJ,?o,rtr.s species sensitive to ethylenediaminetetraacelic acid. J . Bercteriol. 104, 1035-1044. Wilkinson, S . G . , and Galbraith, L. ( l Y 7 9 ) . Polar lipid of P.\rtrc/orrrori~rsi,r.sic.rt/uri.s.Presence of a ~i. Actu 575, 244-254. heptosyldiacylglycerol. B i o ~ h i ~Riophys. Wilkinson. S . G . , Galbraith, L . , and Lightfool, G . A . (1973). Cell walls. lipids, and lipopolysaccharides of P.~rrrc/o~rrr~~ro.s species. E i ~ r ../. H r o c h P r , r . 33, 158- 174 Winkler, H. H., and Miller, E. T. (197X). Phospholipid composition of Rtrkmrici prowcreki grown i n chicken embryo yolk sacs. J . Buc.rr,-io/. 136, 175-178. Woese. C. R . . Magruin. 1.. J . . and Fox, G . E. (1978). Archaebacteria. J . Mn/. E i d . 11. 245-252. Yano, I . , Furukawa, Y ., and Kuaunose. M. (1969). Phospholipids of Noc.urdiu w e / i c i ~ ~Ju. . Beic.teriol. 98, 124- 130.


CURRENT TOPICS IN MEMBRANES AND TRANSPORT. VOLUME 17

Lipids of Bacteria Living in Extreme Environments THOMAS A . LANGWORTHY Department of Microbiologv School qf Medicine Uriiversit~of Sourh Dakota Vermrllion. South Dakoro

I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Organisms and Environments . . . . . . . . . . . . . . . . . . . . . I1 . Apolar Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . FattyAcids . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Isopranyl Glycerol Ethers . . . . . . . . . . . . . . . . . . . . . . 111. Neutral Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Isoprenoid Derivatives . . . . . . . . . . . . . . . . . . . . . . . B . Other Neutral Lipid Components . . . . . . . . . . . . . . . . . . . IV . Glycolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Glycosyldiacylglycerols . . . . . . . . . . . . . . . . . . . . . . . B . Tetrahydroxyhacteriohopane Glycosides . . . . . . . . . . . . . . . . C . Isopranyl Glycerol Ether Glycosides . . . . . . . . . . . . . . . . . . D . Other Polar Lipids . . . . . . . . . . . . . . . . . . . . . . . . . V . Acidic Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Phospholipids . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Phosphoglycolipids . . . . . . . . . . . . . . . . . . . . . . . . . C . Sulfolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI . O v e r v i e w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

1

45 46 49 49 51

56 56 62 62 62 63 64 65 66 66 67 68 69 70

INTRODUCTION

The isolation of new and metabolically diverse species of bacteria during the past decade has renewed a fundamental interest in the ecology. biogeochemistry. physiology. and evolution of bacteria from extreme environments (Heinrich. 1976; Brock. 1978; Brierley. 1978; Kushner. 1978a; Shilo. 1979). Considerable 45

Copyrlghf @ 1982 by Academic Press. Inc All right5 of reproduction in any form rmerved ISBN 0-12-153317-4

46

THOMAS A. LANGWORTHY

interest has developed in the membrane structure of these bacteria, since their membranes must carry out normal physiological processes, yet must intercede against such hostile environmental parameters as extreme temperatures, pH, salinity, radiation, pressure, and dryness. To be sure, the lipids of a vast number of bacteria from extreme environments have not been investigated. Yet it is becoming clear that although the membrane lipids of many of the bacteria thus far examined are quite ordinary, the lipid structures of some of these organisms-notably the isopranyl ether lipids of thermoacidophilic, halophilic, and methanogenic bacteria-are not only unusual, but are changing our perceptions of the supramolecular lipid bilayer model as the universal membrane lipid matrix (Langworthy, 1Y77a,b, 1979a,b). These lipids are additionally promoting our understanding of biochemical, biogeochemical, and cellular evolution (Rohmer et ul., 1979; Tornabene et ul., 1979; Holzer et ul., 1979; Fox et ul., 1980). In view of the variable information between different groups of organisms, this article is necessarily incomplete and can only provide in broad outline the nature of the membrane lipids of bacteria from extreme environments. It is the intent of this article to present a descriptive summary of the more unusual lipid structures from obligately thermophilic, psychrophilic, acidophilic, thermoacidophilic, halophilic, and methanogenic bacteria.

The Organisms and Environments Microbial populations in naturally occurring extreme environments are quite limited, yet a fairly large number and physiological variety of bacteria have been isolated. Although a considerable number of species are able to survive or tolerate exposure to extreme environmental parameters, this article is restricted to the lipids of those bacteria that have an obligatory requirement for their extreme or unusual condition. A brief description follows of the major genera of bacteria in which some aspects of the lipids have been investigated. Thermophilic bacteria have been isolated from such natural habitats as volcanic regions, geothermal soils, and hot springs where temperatures may reach 90"C, as well as from mining waste dumps and soil (Brock, 1978; Tansey and Brock, 1978; Castenholz, 1979; Zeikus, 1979). Moderate thermophiles that have been examined include spore-forming, gram-positive, aerobic Bucillus species, principally B . steurotheniiophilus, which grow optimally at 50-65°C and between 37 and 70°C (Allen, 1953) and several anaerobic Clostridium species, with optimum growth including C. turturivorurn and C. thcrmosarchurol~tici~~n, at 55°C (Chan et ul., 1971). Extreme thermophiles include the Bucillus species B . culdolyticus, B . culdovelox, and B. culdotenux, which grow between 70 and 85°C (Heinen and Heinen, 1972). Members of the genus Thermus, which are

LIPIDS OF BACTERIA IN EXTREME ENVIRONMENTS

47

gram-negative, non-spore-forming rods, include T . uquuricus (Brock and Freeze, 1969) and T . jlaviis (Saiki et ul., 1972), which grow between 40 and 80°C with an optimum at 70°C, whereas T . thermophilus (Oshima and Imahori, 1974) grows between 47 and 85°C and has an optimum at 75°C. Some species of Thermovnicrobiurn (Jackson et ul., 1973; Phillips and Perry, 1976) have also been investigated. These are gram-negative, pleomorphic rods growing between 45 and 85°C and optimally at 60-70°C. Methanobacterium thermouutotrophirum is an example of an obligately anaerobic, thermophilic methanogen that grows between 40 and 75°C with an optimum at 70°C (Zeikus and Wolfe, 1972). Several recent reviews, in addition to those just mentioned, concerning the biochemistry of thermophily and including aspects of thennophile lipids have appeared (Zuber, 1976: Friedman, 1978; Amelunxen and Murdock, 1978; Ljungdahl, 1979). Psychrophilic bacteria populate polar regions and arctic waters where temperatures approach 0°C (Morita, 1975; Baross and Morita, 1978; lnniss and Ingraham, 1978). The lipids of some psychrotrophic bacteria, which will grow below 20°C but have growth optima at higher temperatures, have been examined (Cullen er ul., 1971; Gill and Suisted, 1978). Studies on truly psychrophilic bacteria, which grow only between 0 and 20"C, have been limited primarily to several psychrophilic marine pseudomonads (Brown and Minnikin, 1973), and Vibrio sp. (Bhakoo and Herbert, 1979), Serrutiu sp. (Kates and Hagen, 1964), anaerobic clostridia (Sinclar and Stokes, I964), and Micrococcus cqophitus (Russell, 1971 ). Acidophilic bacteria, represented by the chemolithotrophic Thiobaciilus, are found associated with acidic regions such as acid mine drainage and acidic mining waste dumps (Lundgren et ul., 1964, 1974; Langworthy, 1978a; Tuovinen and Kelly, 1978). The lipids of two species that have been investigated in some detail include T . thiooxicluns and T .jierrooxickrns, which grow optimally at pH 2 and between pH 1 and 4. They are obligately autotrophic, non-sporeforming, gram-negative rods that obtain energy from the oxidation of iron and sulfur with the concomitant production of sulfuric acid. Several thermophilic Thiohuc.illcis species have been reported (Brierley, 1978; Brock, 1978), as well as alkalophilic Bacillus species (Langworthy, 1978a), which grow optimally at pH 10, but the lipids of these organisms have not been detailed. Thermoacidophilic bacteria have been isolated from acid hot springs, solfotara soils, and self-heating coal refuse piles where temperatures and acidity range from 55 to 85°C and pH 1-3. These organisms must thus contend with high temperature and low pH simultaneously. Thermoacidophiles are composed of three morphologically and physiologically distinct types. Bucillus ucidocdduriiis is a spore-forming rod that grows within the range of 40-70°C and pH 2-6 and optimally at 60-65°C and pH 3 (Darland and Brock, 1971). Sulfolobiis ucio'oculrlurius, a facultative autotroph capable of growth on sulfur and iron,

48


possesses an atypical cell wall and grows within the limits of 55-85°C and pH 2-5 and optimally at 75°C and pH 3 depending on strains (Brock et ul., 1972; de Rosa et al., 1975a). Perhaps the most unusual organism is Thermuplusmu acidophilurn, a wall-less mycoplasma, whose membrane is directly exposed to its hot acid environment (Belly et ul., 1973; Langworthy, 1979a). It grows within the limits of40-62°C and pH 1-4 and optimally at 59°C and pH 2. Additionally, hydrogen ions are specifically required for maintaining cellular integrity, as Thrrrnopla.smu is lysed by neutrality (Smith et d., 1973). Both 7hermoplusmu and Su/fo/obus are characterized by the possession of isopranyl ether lipids. The lipids of thermoacidophilic bacteria have been the subject of several recent reviews (Langworthy 1978b, 1979b, 1980a, 1981). Halophilic bacteria inhabit solar salt flats, brine, and hypersaline lakes where salt concentrations approach saturation (Dundas, 1977; Bayley and Morton, 1978; Kushner, 1978b; Lanyi, 1979). The extreme halophiles comprise the rodshaped Hulobacteriuni species H. cutirubrum, H . Izulobium, H . salinarium, and H . rnurisrnortui, as well as the coccal forms Surcinu litoralis and S. morrhuue. These organisms grow optimally in 20-25% NaCl and between salt concentrations of 15 and 3 0 8 . Like Sulfolobus, the halophiles possess an atypical cell wall structure. Like Thermoplasmu, which requires protons, halophiles require sodium ions for structural integrity, being lysed by low salt concentrations. In addition, the discovery (Oesterhelt and Stoeckenius, 1973) that H . halobium contains bacteriorhodopsin, a photosensitive purple pigment that converts light to chemical energy, has generated considerable interest in the bioenergetics of halophilic bacteria (Caplan and Ginzburg, 1978). The halophilic bacteria possess isopranyl ether lipids, which have been the most fully established of any of the bacteria from extreme environments, principally by Kates and associates. The chemistry of these lipids has been extensively reviewed by Kates (1972, 1978), Kates and Kushwaha (1976), and Kates and Kushwaha (1978). Methanogenic bacteria are strictly anaerobic organisms whose metabolism is based on the formation of methane from carbon dioxide and hydrogen, formate, acetate, or ethanol. They are found in sewage, bogs, and sediments, and they comprise a variety of Methunobacteriurn, Methanosarcina, Methunospiritlum, and Merhanococcus species (Zeikus, 1977; Balch et al., 1979). They, too, possess an atypical cell wall structure and isopranyl ether lipids (Tornabene and Langworthy, 1979). Although they live in more of an unusual than an extreme environment, the methanogenic bacteria are included in this review because of the recent realization of the close phyletic relationship between Thermoplasma, Sulfolobus, halophiles, and methanogens. Based on 16 S rRNA sequence analyses, the presence of ether lipids, and the absence of typical cell walls, Woese and associates have proposed that this group of bacteria be given the name Archaebacteria. According to their view this group represents a line of evolutionary

49


descent different from either prokaryotic or eukaryotic cells (Woese and Fox, 1977; Woese et al., 1978; Fox et al., 1980).

II. APOLAR RESIDUES

A. Fatty Acids Influence of temperature on the physical state of membrane lipids has focused considerable attention on the fatty acyl moieties of the acylglycerol residues in thermophilic and psychrophilic bacteria. In general, the fatty acid profiles extend the trends that are reflected in mesophilic bacteria, such as Escherichia coli, which have been subjected to short-term shifts in temperature (Marr and Ingraham, 1962). Psychrophiles are characterized by high proportions of unsaturated and shorter saturated fatty acids, whereas thermophiles possess a high content of longer saturated and predominantly isobranched acids. This trend is in general concert with the anhydrous melting points of the fatty acids, although the actual melting points of the ester-linked acids in the membrane are certainly influenced by cooperative interactions with other molecules. Examination of extremely thermophilic Thermus species reveal that isoC,, fatty acid is most abundant (50-61%), followed by isoc,,, and together the isoC,, and isoC,, pair accounts for about 7 5 4 5 % of the fatty acid content (Heinen et al., 1970; Ray et a l . , 1971a; Oshima and Miyagawa, 1974; Oshima, 1978). Oshima and Miyagawa (1974) found that the ratio of isoC,, to isoC,, acids increased in T . thermophilus grown at increasing temperatures from 49 to 82"C, indicative of chain elongation at higher temperatures. Ray el al. (1971a), however, observed a slight decrease in isoC and concomitant increase in the nC,, and isoC,, fatty acid content with increasing temperatures of T . aquaticus grown between 50 and 75°C. In the extreme thermophiles B . caldotenax, B . caldovelox, and B . caldolyticus, branched fatty acids represent about 80% of the total, consisting mainly of isoc,,, isoC,,, and isoC,, acids. A pronounced shift from isoC,, to isoC,,, and also isoC,, to nC,,, was demonstrated on increasing growth temperatures from 45 to 80°C (Heinen and Heinen, 1972; Weerkamp and Heinen, 1972; Hasegawa er a / . , 1980). The fatty acids of several hydrocarbonutilizing thermophiles, apparently closely related to species of Thermomicrobium, were examined by Merkel and Perry (1977). The fatty acid distribution varied depending on growth substrate, but consisted mostly of CIS-,c16-7 and C,,-branched and nC,, fatty acids. The branched acids were reported to have the anteiso rather than the iso configuration. The most dramatic change resulted after growth on n-heptadecane, which caused a large shift from the branched fatty acids to nC,, and nC,,, which together represented 50-70% of the total fatty

,,

50


acids. Among moderate thermophiles, R. stearotherinophilirs contains isoC,, and isoC,,acids, but the total of this pair (34-64%) is substantially less than in extreme thermophiles, and considerable amounts of tic,,, is0 and unteisoc,,, and unteisoC,, are present as well (Cho and Salton, 1966; Daron, 1970;Yao et ul., 1970;Shen et al., 1970;00 and Lee, 1971;Oshima and Miyagawa, 1974). Similar trends are apparent in the moderately thermophilic Closfridiuin examined by Chan etul. (1971).The main fatty acids were nC,,, iiC,,, and predominantly isoc,,. Small quantities (8-10%)of a new unsaturated, C,,,cyclopropane fatty acid, identified as 12,13-methylene-9-tetradecenoicacid, were found. This is the first reported occurrence of an unsaturated cyclopropane fatty acid in bacteria. Psychrophiles, in contrast to thermophiles, are distinguished by large quantities of monoenoic acids, mainly nC,,:, and HC,,:,. The psychrophilic marine pseudomonads examined by Brown and Minnikin (1973)were grown at 10°C and contained a simple fatty acid profile consisting of nC,,. tic,,:,,and small amounts of nC,,,,. When grown between 10 and 20"C,the fatty acid profiles remained unchanged, suggesting a lack of a mechanism whereby fluidity may be controlled in these organisms. Micrococcus cryphilus, grown at either 20°C or O"C,contained 95% nC,,,, and nC,,:,, but Russell (1971)observed a 4-fold increase in the nC,,:, to nC,,:, ratio when growth temperature was changed from 20°C to O"C, indicative of chain shortening. Kates and Hagen (1964)reported that a Serrutiu-like psychrophile, grown at 5°C or 10"C, had large amounts of tic,, and nC,,,, but not j7Cl,:l or cyclopropane fatty acids as in its psychrotrophic counterpart, S. niarcescens. Bhakoo and Herbert (1979)investigated the fatty acids of four different marine vibrios grown between 0 and 15°C.The isolates contained no fatty acids longer than 17 carbons. Two of the isolates increased the proportions of nC,,,, , nC,,:, , and nC,,,, on lowering growth temperatures. One responded by chain-length shortening by increasing the amount of nC,,,, , but one isolate contained 60% nC,,,,, which did not change at all in response to temperature. The psychrophilic Clostridium examined by Chan et a / . (1971) contained nC and nC (40%) and a large quantity (45%) of unsaturated cyclopropane fatty acids, mainly 12,13-methylene-9-tetradecanoicacid. Acidophilic Thiobacillus species are characterized by a high content (nearly 50%) of C,,-cyclopropane fatty acids (Levin, 1971, 1972). A C,, P-hydroxy fatty acid, 3-hydroxyhexadecanoate (Knoche and Shively, 1972) and a CIS,cyclopropane hydroxy acid, cis- 1 I ,12-methylene-2-hydroxyoctadecanoicacid (Knoche and Shively, 1969) are also found in covalent linkage to the ornithinecontaining lipid of these organisms (see Section IV,D). The thermoacidophile, B . acidoculdarius, contains both C,,-branched fatty acids, like thermophiles, and a prevalence of cyclized fatty acids like acidophiles, However, the major fatty acids (50-90%) are composed of the alicyclic, w-cyclohexyl, C,, and C19acids, 1 1-cyclohexylundecanoate and 13cyclohexyltridecanoate, which are biosynthesized from glucose via the shikimate


51

pathway (de Rosa et ai.,1972; Oshima and Ariga, 1975; Oshima et al., 1978). Although de Rosa et crl. (1974) discuss a complex relationship between the fatty acid composition, temperature, pH, and metabolism, Oshima and Ariga ( I 975) could find no significant alterations in the fatty acid composition in cells grown at different temperatures and pH values. The function of cyclohexyl fatty acids becomes even less clear, since a mutant of B . suhtilis, which is able to synthesize cyclohexyl fatty acids on addition of appropriate precursors, is not provided with any special advantage for growth at high temperature or low pH. In fact, these fatty acids caused a decrease in the transition temperature of the lipids (Blume Pt ui., 1978). It has been suggested (Blume ot ui., 1978) that the bulkiercyclohexyl rings in the interior of the membrane may provide optimal packing and orientation with the numerous triterpene derivatives present in B . ac.itlocalr1r~riu.s(see Section III,A,3,4 and Ourisson and Rohmer, this volume). Cyclohexyl fatty acids are not restricted to B . acidocaldurius but also occur in hydrocarbonutilizing Mycohucterium and Nocardii species grown on n-alkyl-substituted cycloparaffins (Beam and Perry, 1974).

6. lsopranyl Glycerol Ethers Unlike any other organisms, diacylglycerol residues are absent in the halophilic, methanogenic, and thermoacidophilic archaebacteria, Thertnoplusrna and Su(fo1obu.s. The apolar residues consist either of two C,, or two C,,, fully saturated and isopranoid-branched hydrocarbons in ether linkages to glycerol as either di-0-phytanyl glycerol (1) or tetra-0-di(biphytany1) diglycerol derivatives (2). Halophiles possess the diether (Kates, 1972, 1978), whereas Ther~noplastna (Langworthy, 1977a) and Sulfolohirs (de Rosa et N I . , 1977b) contain tetraethers, and methanogens possess both diether and tetraethers (Tornabene and Langworthy, 1979).

Di-0-phytanyl glycerol was first recognized to constitute the sole apolar residue in halophilic bacteria through the extensive studies of Kates and associates (reviewed by Kates, 1972, 1978). The 0-alkyl groups were found to consist of

52


the C,,-hydrocarbon, phytane, and the glycerol to have the sn-2,3 configuration opposite that of naturally occurring diacylglycerols. The di-0-phytanyl glycerol (1) was thus shown to be 2,3-di-0-(3R, 7R, 1 l R , 15-tetramethylhexadecyl)-snglycerol. Initial studies on Thermoplasrna and Sulfolobus (Langworthy et al., 1972, 1974) revealed the sole presence of ether lipids. These contained glycerol but instead had C ,,,-hydrocarbon chains. Their structural assembly as diglycerol tetraethers (2) has been only recently established and confirmed (Langworthy , 1977a; de Rosa ef al., 1977b, 1980e; Yang and Haug, 1979). The diglycerol tetraether structure consists of two sn-2,3-glycerol molecules bridged through ether linkages by two identical pairs of C,,-terminal diols with the resultant primary hydroxyl groups of the glycerols in the trans configuration. The C,,hydrocarbon chains have been shown by de Rosa et al. (1977a,b, 1980e) to have the w,w-biphytanyl skeleton made up of two C B,rphytanyl units joined “head to head” at the 16,16‘-geminal ends. The diglycerol tetraethers (MW 1300) are therefore the structural equivalent of two molecules of di-0-phytanyl glycerol (MW 650) that have been condensed by covalent linkage through the 16,16’terminal ends of their 0-phytanyl side-chains. The C,,-biphytanyl chains also differ in the additional feature that they may contain up to four cyclopentane rings (de Rosa et ul., 1977a,b, 1980e). The series of diols constituting the ether linkages to glycerol may be the acyclic biphytane, C4,HR202(3), as already noted; the monocyclic-C,,H,,O, (4); bicyclic-C,,H,,O, ( 5 ) ; tricyclicC,,H,,O, (6);or tetracyclic-C,,H,,O, (7) biphytane derivatives. There are H0.

OH

HO

OH


53

HO

% % H O

(7)

therefore five different molecular species of tetraethers that are theoretically ,;, MW 1300- 1 194, depending on cyclization in the identipossible, C Jl 172-ll,R0 cal pair of biphytane chains. However, only (3), (4), and ( 5 ) are the main biphytnne constituents of diglycerol tetraethers in Thermoplasma and Sulfolobus. The more cyclized biphytanes (6) and (7)are primarily associated with the second specialized class of tetraether peculiar to Suljolobus. This more polar tetraether, representing 50-75% of the total ethers in Sulfolobus-depending on heterotrophic or autotrophic growth-contains a pol yo1 substituted for one of the glycerol molecules in the tetraether assembly (Langworthy et a l . , 1974; Langworthy, 1977b, 1978b; de Rosa et al., 1980a). De Rosa et al. (1980a) have shown the polyol to be a C,-branched nonitol, called by them calditol, giving rise to a Cg2H l,ixO12, MW 1472-1464, calditol glycerol tetraether (8). The extent to which pentacyclic rings do occur appears dependent on the strains involved, as

well as the growth temperature (Langworthy et ol., 1972, 1974; de Rosa et a l . , 1976a, 1980d; Yang and Haug, 1979; Furuya et al., 1980). In Thermoplasma, acyclic (3) and monocyclic (4) biphytanes are predominant, whereas in Sulfolobus, which grows at much higher temperatures, the monocyclic (4), bicyclic (3,tricyclic (6), and tetracyclic (7)biphytanes are most pronounced. De Rosa et al. (1980d) and Furuya et a l . (1980) have also shown that Sulfolobus increases the amount of cyclization in the biphytanyl chains with increasing temperatures from 50 to 85°C; suggesting a role in membrane stabilization at high temperatures. Thus the apolar residues of Thermoplasma are composed of diglycerol tetraethers, whereas Sulfolobus contains approximately equal amounts of more highly cyclized diglycerol tetraethers and calditol glycerol tetraethers. However, both organisms do possess small quantities of di-0-

54


phytanyl glycerol associated with the polar lipids (Langworthy , I979b, and unpublished). A11 methanogenic bacteria so far investigated contain di-0-phytanyl glycerol, but depending on genera, contain diglycerol tetraethers as well. Of nine different species representing four different genera examined (Tornabene ct ul., 1978; Makula and Singer, 1978; Tornabene and Langworthy, 1979). the coccal forms, Methanococcus and Methatiosarcinu, contain only di-0-phytanyl glycerol, whereas the rod- and spiral-shaped methanogens, MethntinbaL.tL.riiirn and Methnrios~~irilht~i, possess di-0-phytanyl glycerol (38-72%) and diglycerol tetraether (28-62%). Cyclization is absent and the tetraether contains only acyclic biphytanyl chains (3). Di(biphytany1) diglycerol tetraether (2) is the only molecular species of tetraether so far detected in methanogens. Thus the occurrence of both ethers in methanogens, which grow under normal physiological conditions (albeit anaerobically), indicates that the ethers of halophiles and thermoacidophiles cannot be viewed as an adaptation for survival in hypersaline or hot acid environments. These lipids are however, well suited for such purpose (Kates, 1972, 1978; Langworthy, 1977a, 1979a). Rather, these lipids reflect a phyletic relationship and a more profound evolutionary development in which these organisms share a common evolutionary episode distinctly different from other cells (Fox et ul., 1980). The distribution of diethers and tetraethers among halophilic, methanogenic, and thermoacidophilic archaebacteria is summarized in the table. The discovery of tetraether lipids within the archaebacteria is of considerable interest in terms of molecular organization and membrane biogenesis. The di0-phytanyl glycerol residues of halophiles allow for the formation of a typical membrane lipid bilayer through interaction of separate and opposite phytanyl residues, the only constraint being that the chain length is invariably fixed at 20 carbons. Tetraethers, however, accounting for the majority of the membrane hydrocarbon of Thermnplusinu and Sulfolobus, approximate 45-75 in length depending on cyclization in the hydrocarbon chains and span the membranes, which average about 70 8, in width (Langworthy, 1977a, 1978b, 1979a,b, 1980a). Therefore, these two organisms, along with regions within the membranes of those methanogens containing tetraether, can be considered to possess a covalently cross-linked, or sealed, membrane bilayer created by virtue of the extension of the C,,-hydrocarbon chains across the membrane in covalent linkage to glycerol residues on the inner and outer membrane faces. These membranes cannot, therefore, be considered to comprise a lipid bilayer in the strict sense of the word, but are structural equivalents of an amphiphilic monolayer that has been condensed at the center joining both halves of the bilayer together. Correlating well with a monolayer membrane, Thermoplasrna and Sulfolobus fail to freeze-fracture tangentially to yield inner and outer membrane faces, but instead characteristically cross-fracture perpendicularly through the membrane as ex-

55

LIPIDS OF BACTERIA IN EXTREME ENVIRONMENTS D I P H Y T A NGI.YCEROI. YL ETHER A N D DICI.YCEROL TETRAETHER DISTRIEIJTION I N HAI.OPHII.IC. METHANOGENK, A N D THI:RMOACIDOPHILIC BACTERIA" Organism

Diethcr (%)

I00 I00

Hulohucterium cutiruhrum Hulnhuc~trriumhtilohiitm Holohwterium sulinrrrium Hulohuctrriurn mcirismortui Surc,rnu literulis Surt,inu morrhuur

100 1 00

Mrthunosurcinci burkeri Mrthanor~oc~c~u.~ strain PS Mcjthrinococcus vunnirllr Mr~thunoha~.terium ruminutiuni M- I Methonohuc~tc,riumruminritiunr PS Methanobacterium thrrmoti utntrophic.rtm Methunobwtrrium strain M.v.H. Me~thuno.spirrllirtnhungutiG Methunospirillum strain AZ

I00 99.9 71.8 44.7 44.5 43.5 40.5

Thrrmoplasmu uc~idophilum Su!/olobus cic~idc~caldurtus "

I00 1 00

I 00

Tetraether (47.) 0 0 0 0

0 0 0

37.5

0 0.I 28.2 55.3 55.5 56.5 59.5 62.4

10.0 5 .O

90.0 95.0

Data from Kates (197X); Tornabene and Langworthy (1979); Langworthy (1979b. IY80a, 19x1).

pected of a monolayer assembly (Langworthy, 1979a). In light of a monolaycr membrane, the role of cyclization in the tetraethers of Therrnoplustnu and SU/,folohus might also be explained. Since the hydrocarbon chains comprising the tetraethers are fixed at 40 carbon atoms linked to glycerol on each end of the chain, cyclization would reduce rotational freedom within the chains and thereby the interior of the membrane. In addition, cyclization provides an effective means of controlling chain length and simultaneously tetraether and membrane width. Thus cyclization may provide a mechanism of controlling internal mernbrane viscosity and condensing the membrane in response to temperature and permeability in a fashion similar to cholesterol (see Razin, this volume). Cyclization is absent in the tetraethers of methanogens, but perhaps membrane fluidity and permeability may be controlled by altering proportions of diether and tetraethers, although the effect of temperature and ions has not been reported. The question of membrane asymmetry is also an interesting one since the tetraethers are themselves symmetrical molecules. Substitution of either one or both of the primary hydroxyl groups on opposite ends of the tetraether molecule will determine asymmetry. First indications (Kuswaha et ui., 1981a,b) are that both types of substitution exist, wherein glycolipids contain carbohydrate linked to only one side, whereas the acidic phosphoglycolipids contain carbohydrate attached to one

56


side and the phosphate radical to the opposite end of the tetraether (see Sections IV,C and V,B). Furthermore, diether and tetraether biosynthesis clearly involves the isopentenyl pyrophosphate pathway, at least to the C,,-geranylgeraniol pyrophosphate intermediate (Kates, 1978; Langworthy, 1979a; de Rosa et ul., I980b; see Fig. 15 in the article by Ourisson and Rohmer, this volume). From this point the biosynthetic steps are unknown. However, the biphytane chains of tetraethers are condensed “head to head” through the geminal ends of two C,,, residues rather than tail to tail through the terminal phosphates of two C2,geranylgeraniol pyrophosphates as in carotenoid synthesis. Thus, combined with the fact that tetraethers are the structural equivalents of two covalently linked diether molecules, tetraether biosynthesis could occur via “head to head” condensation between the two diether molecules, or in fact two polar lipid derivatives of diethers to yield unsubstituted tetraethers or tetraether complex polar lipids (Langworthy, 1979a, 1980a; de Rosa et al., 1980e; Kushwaha e f al., 1981a,b). The rapid turnover of the small quantity of di-0-phytanyl glycerol in Thennoplusmu indirectly adds support to this hypothesis (Langworthy, 1980b). Tetraether biosynthesis is clearly unusual and its elucidation should provide a new route of hydrocarbon biosynthesis.

111.

NEUTRAL LIPIDS

A. lsoprenoid Derivatives The neutral lipids of halophilic, thermoacidophilic, and methanogenic archaebacteria have been the most fully elucidated among the bacteria from extreme environments. They represent approximately 10-30% of the total lipids and are composed almost exclusively of isoprenoid derivatives. The neutral lipids of other bacteria have been largely ignored, but several constituents have been identified in the thermoacidophile B . acidocaldarius, the acidophile T . ferrooxidans, and the thermophile T . aquaticus. Neutral lipids of these organisms range respectively from about 16% in B . acidocaldarius to 60% in the extreme thermophile T . aquaticus. The neutral lipids that have been identified can be broadly grouped into seven major classes based on chain length as follows: C,,-isoprenoids (geranylgeraniol, phytanes, phytanyl ethers, retinal), C,,-isoprenoids (pentaisoprenalogues), C,,isoprenoids (squalene, hopanes), C,,-isoprenoids (tetrahydroxybacteriohopane), C,,,-isoprenoids (carotenes), and C,,-isoprenoids (bacteriorubrins, polyprenols). I . C,,,-~SOPRENOIDS

a. Geranylgeruniol. Geranylgeraniol(9), containing one cis double bond, constitutes the main C2, component of the halophilic bacteria (Kushwaha et ul.,

57


1975a; Kushwaha and Kates, I978b). All-trans-geranylgeraniolhas been reported as a minor component in Sulfolobus (de Rosa et al., 1980b). Since geranylgeraniol may be an intermediate in diether and tetraether biosynthesis, it is likely to occur in Thermoplasma and methanogenic archaebacteria as well.

b. Phytanes. Although halophilic bacteria contain trace amounts of phytanol (Kushwaha and Kates, 1978b), phytane and its unsaturated homologues dominate the C,,-isoprenoids of Thermoplasma, Sulfolobus, and methanogenic bacteria. Methanogens contain mainly phytane and phytaene, Sulfolobus phytadiene and phytatriene, whereas phytatetraene predominates in Thermoplusma (Tornabene e t a / . , 1979; Holzer ef al., 1979). c. Di-0-Phytunyl Glycerol. Primarily associated with the polar lipids, the diether is present in the free form amounting to about 8% of the neutral lipid fraction of halophilic bacteria (Kushwaha e f al., 1975a; Kushwaha and Kates, 1978b). The small quantities of di-0-phytanyl glycerol in Thermoplasma and Sulfolobus are associated with the polar lipids and have not been detected free in the neutral lipid fraction (Langworthy, 1979b, 1980a, also unpublished). De Rosa et al. (l976b) reported the occurrence of an unusual tri-0-phytanyl glycerol ether containing partially or fully saturated phytanyl chains as a minor component of Sulfolobus neutral lipids. d. Retinal. All-trans-retinal (10) as a minor component has only been reported in the neutral lipids of pigmented halophiles (Kushwaha and Kates, 1973; Kushwaha et al., 1974; Kates, 1978). Its presence is dependent on growth conditions being produced anaerobically in the presence of light. It is associated principally with the purple membrane in the retinal-protein complex, bacteriorhodopsin. Of considerable interest is whether retinal or an analogue may occur in other arc hae bac teria .

2. C

2

,

-

I

~

~

~

~

~

~

~

~

~

~

Acyclic pentaisoprenes with a continuous range of hydropentaisoprene derivatives are relatively major neutral lipid species in Thermoplasma, Sulfolobus, and various strains of methanogenic archaebacteria (Tomabene et al., 1979; Holzer

58


ul., 1979). In Thet-rnoplasma the C,,H,, pentaene is predominant, whereas fully saturated C,,H,, is the major pentaisoprene in Sulfolobus. The full range of C 2;,H3 2 - - 1 2 pentaisoprenes is found among different species of methanogens.

CI

3 . C:~O1SO PKE N 0 11)s

u . Squalenes. The presence of squalenes and hydrosqualene derivatives as the major acyclic isoprenoid neutral lipids is a feature that distinguishes archaebacteria. Squalenes, representing about 36% of the neutral lipids of halophiles grown aerobically, have been identified as C,,H,,,, all-trans-squalene (11); C ,JI ;,2r all-trans-dihydrosqualene (12); C :,,$I ;,4, all-trans-tetrahydro squalene (13); and C:,,,H 4 X , dehydrosqualene (14) (Tornabene et ul., 1969; Kramer et al., 1972; Kushwaha et al., 1972). The relative proportions vary among halophiles (Kushwaha et al., 1974), and the ratio of squalene to dihydro- and tetrahydrosqualene decreased proportionately when cells are grown anaerobically in the light (Kushwaha ef al., 1975b) or under microaerobic conditions (Tornabene, 1978).

The squalenes of methanogens represent between 64 and 95% of the total neutral lipids (Tornebene et al., 1978, 1979). These are composed of a continuous range of C & 32-,(i,, hydrosqualenes from dihydrosqualene up to and including


59

decahydrosqualene, but squalene. dihydro-, and tetrahydrosqualene are the predominant species. Squalene itself is predominant in Thc.i-r,io~ltr.c./~ltr, but octaand decahydrosqualene are the major species in Sulfi~lobiis(Tornabene ~t a [ . , 1979). Dehydrosqualene is absent in methanogens, Tlwrmoplu.smc~, and Sirl,folohu.s. Although characteristic, squalenes are not restricted to archaebacteria. De Rosa e/ u l . ( 1973) identified squalene in the therinoacidophile R . acicloculclurius, which is the likely precursor to the pentacyclic hopanoid triterpenes that characterize this organism (Ourisson rt ul.. 1979). Squalenes have now been 1978; detected in a few aerobic and anaerobic eubacteria as well (Amdur e/ d., Mercer P / al.. 1979). h. Hopunes. Pentacyclic triterpenes of the hopane family are now known to have a widespread occurrence in a variety of microorganisms (Ourisson c/ ul., 1979; Rohmer rt u l . , 1979). The function of this class of lipids, which may be the structural equivalent and phylogenetic precursors of sterols, is considered in detail by Ourisson and Kohmer (this volume). Among the bacteria from extreme environments considered herein, only one, B . uciclocultlarius, possesses hopanes in the neutral lipids. De Rosa ~t ul. (1973) demonstrated that free hopanes account for about 0.3% of the cell dry weight of this bacterium and consist of about 86% hop-22(29)-ene (15), trace amounts of hop-17(21)-ene (16), and 4% hopane (17).

60

4. C


3

5

-

I

~

~

~

~

~

~

~

~

~

~

In addition to squalene and hopanes, B . acidocaldarius possesses a third triterpene derivative, tetrahydroxybacteriohopane (18). This polar compound contains the hopane nucleus but is substituted at C-29 with n-l,2,3,4-tetrahydroxypentane (Langworthy and Mayberry, 1976). It equals nearly 1.5% of the cell dry weight but only a small amount exists in the free form. It serves primarily as a major new type of aglycone in the glycolipids of the organism (Langworthy et al., 1976; see Section IV,B). O H OH

5. C&OPRENOIDS With the exception of the pigmented halophiles, carotenoids of bacteria from extreme environments have not been well defined. Halophiles possess low concentrations of lycopersene, cis- and trans-phytoene, cis- and trans-phytofluene, lycopene, neo-a-carotene, and p- and neo-p-carotene (Kushwaha et al., 1972; Kushwaha and Kates, 1973). The low concentrations of carotenoids suggest that they might serve as biosynthetic precursors to retinal or bacteriorubrins (Kates, 1978). The neutral lipids of the extreme thermophile T . uquaticus were shown by Ray et al. (1971b) to be composed of about 8% phytoene, 7% A-carotene, and 75% very polar carotenoids, which were not identified. Although the distribution remained the same. the total carotenoid content increased I .8-fold on increasing the growth temperature from 50 to 75"C, suggesting a role in membrane stabilization at high temperatures. Yellow and orange carotenoids were also noted in the high neutral lipid content of the acidophile T . ferrooxidans by Short et al. (1969), but were not investigated. 6. LIPOQUINONES Menaquinones-7, -8, and -9, containing C3sr C4,,, and C4,-prenyl chains, respectively (19), have been demonstrated in several aerobic bacteria from extreme

61


habitats. Thermoplasma contains menaquinone-7 (Langworthy et a / . , 1972); halophiles, menaquinone-8 (Tornabene et al., 1969; Kushwaha rt al., 1974, 1975a); and B . acidocaldurius contains menaquinone-9 (de Rosa ef a / . , 1973). Ray et ul. (1971b) found that T . uquuticus contained all three menaquinones, 7-9. Menaquinones were found to be absent in T . ferrooxidans (Short et al., 1969), but it contained coenzyme 4-8 as the sole lipoquinone. De Rosa et al. (1975b) found that Surolobus contains an unusual triterpenoid 4,7thianaphthenequinone, which is the first occasion for this compound to be detected in a natural source. 7. C ,,,-ISOPRENOIDS Polyprenols containing more than 45 carbons have only been described in halophilic bacteria and in B . acidoculdarius. Pigmented halophiles contain, in decreasing proportions, the tetrahydroxy , C,,,-noncyclic carotenoid, bacteriorubrin (20); the C,,-triol, mono-anhydrobacteriorubrin (21); and the C,,-diol, bisanhydrobacteriorubrin (22); these are associated primarily with the red membrane fraction (Kushwaha et al., 1974, 1975a,b). The prenol fraction from B . acidoculdarius was shown by de Rosa et al. (1973) to contain a series of polyprenols between 9 and 12 isoprene units. The C,, and C,, species were predominant, consisting of a-cis-, all-rrans-, and a-tert-prenyl derivatives. OH

OH

OH

OH

62


B. Other Neutral Lipid Components I n addition to the major isoprenoids just discussed, Holzer et a / . ( 1 979) have described a series of methyl-branched isoprenes and isopranes (C,,-C30) and n-alkanes (C,9-C32)that are present in small quantities in Thermoplasmu, Sulfolobus, and methanogenic bacteria. Kushwaha and Kates (1978a) showed low levels of mevalonic acid in a number of halophiles but this is not surprising because the lipids of these bacteria contain exclusively isoprenoid chains. The neutral lipids of halophiles, however, do contain significant amounts of a nonisoprenoid compound, indole (Kushwaha et ul., 1977). The physiological significance of indole and its intracellular presence is unknown.

IV. GLYCOLIPIDS Considering the fairly large variety of bacteria from extreme habitats, glycolipids have only been investigated in a few thermophilic and thermoacidophilic bacteria, with most attention centered on the archaebacteria. However, it is becoming evident that within the thermophilic eubacteria and thermophilic archaebacteria, carbohydrate-derived lipids constitute the major lipid class.

A. Glycosyldiacylglycerols Although Short et al. (1969) detected no glycolipids at all in the mesophilic acidophile T . Jerrooxiduns, the thermophilic acidophile B . acidocalclarius has a glycolipid content of about 64% (Langworthy ef al., 1976). The major glycolipids are glucosyl-glucosamidyl-diacylglycerol derivatives (23),and comprise about 70% of the total glycolipid fraction. They consist of approximately 25% Glcp(P I +4)GlcNacyl(P I +1 )diacylglycerol, 41 9% of Glcp(p 1-+4)GlcNacyl(p I + I)monoacylglycerol, and trace amounts of Glcp(@1 -+4)GlcNacyl(P 1- 1)glycerol. The configuration of the glycerol residue is uncertain but the ester- and amide-linked fatty acids have a similar distribution. HOCHv

HOCH:!

OH

NH I

c =o I

R

(23)

HC- 0 - C - R I 0 HpC-0-C-R


63

Extreme thermophiles of the genus Thermus, including T . thermophilus, T . Juvus, and T . uquuticus, contain an unusual tetraglycosyldiacylglycerol that constitutes 50-70% of the total lipids in the organisms (Oshima and Yamakawa, 1974; Oshima and Ariga, 1976; Oshima, 1978; Pask-Hughes et a/., 1977). Oshima and Yamakawa (1974) found the compound to be a galactosyl- galactosyl-glucosamidyl-glucosyl-diacylglycerol with a terminal galactofuranose residue (24). The structure has now been established as Gal,f(P1-2)Galp ((-~1+6)GlcN(15- methylhexadecanoyl) (pl-2)Glcp ( a l ~ l ) d i a c y I g l y c e r o l al, though the glycerol configuration is uncertain (Oshima and Ariga, 1976). In this lipid, isoC,, is the sole fatty acid in amide linkage to glucosamine. Additionally, the glycolipid content of a number of Thhrrmus species increases almost 2-fold at temperatures increasing from 50 to 80°C (Ray et ul., 1971b; Oshima, 1978). This correlates with the observations of Pask-Hughes c2t al. (1977) and Sharom et ul. (1976) that increasing glycolipid concentrations or glycolipids with an increasing number of sugar residues renders phospholipid bilayers increasingly rigid. Thus the high glycolipid content, as well as the presence of tetraglycosyldiacylglycerol, and, in fact, glycosylglycerols in general, may play a major role in stabilizing membranes to thermal or other environmental stresses.

B. TetrahydroxybacteriohopaneGlycosides

B . acidocald~irhscontains, in addition to glucosyl glucosamidyldiacylglycerols, an N-acylglucosaminyl-tetrahydroxybacteriohopane,which constitutes about 25% of the glycolipid fraction of cells grown at pH 3 and 60°C (Langworthy et al., 1976; Langworthy and Mayberry, 1976). N-acylglucosamine, which is joined by a P-glycosidic linkage to the primary -OH of the tetrahydroxybacteriohopane aglycone, has been established as 1-( O-~-N-acylglucosaminy1)-2,3.4tetrahydroxypentane-29-hopane (25). Poralla and associates ( 1980; Kannenberg et a/.. 1980), demonstrated that the glycolipid as well as the free aglycone produced a condensing effect similar to cholesterol in synthetic monolayer membranes, suggested that the lipid may function in diminishing diffusion of H ’ ions through the membrane.

64


NH I

c=o

C. lsopranyl Glycerol Ether Glycosides 1 . DIET'HER DERIVATIVES

Di-0-phytanylglycerol ether glycolipids have been identified in halophilic and methanogenic archaebacteria. Most extreme halophiles encountered contain the tri gl ycosyldiether Galp ( p1+6)Manp ( a1-2)Glcp ( a1+1 )2,3, - di- 0 - phytanylsn-glycerol (Kates and Deroo, 1973; Kates, 1978). The glycolipid per se occurs in lesser amounts, but the acidic sulfate derivative is one of the major lipids in these organisms (see Section V,C). However, the lipids of one halophile from the Dead Sea, H . mirismortui, contain about 1 1 % of a novel triglycosyldiether, Glcp (pi-6)Manp (a14 2 ) G l c p ( a l b1)2,3-di- 0-phytanyl- sn-glycerol (Evans et ul., 1980). In the first report on the nature of the complex lipid structures in methanogenic bacteria that contain both diethers and tetraethers, Kushwaha et ul. (1981a,b) have shown that M . hungutei possesses two new galactofuranosyl-containing diglycosyldiethers: Calf@ 1+6)Galf(p I +1)2,3-di-O-phytanyl-sn-glycerol and Glcp ( a1 +2)Gal,f(a 1- 1)2,3-di-O-phytanyl-sn-glycerol. These accounted for 2 and 17% of the total lipids, respectively. 2. TETRAETHER DERIVATIVES Tetraether glycolipids, identified thus far in archaebacteria, include those of M . hungatei, Sulfolobus, and Thermoplasma. The methanogen M . hungatei contains two diglycosyltetraether glycolipids, which equal less than 1% of the total lipids (Kushwaha et al., 1981a,b). The same disaccharides as in the diether analogues are glycosidically linked to one -OH of the diglycerol tetraether, with the other -OH radical remaining free, existing as Ga!f'(P 1 +6)Galf(p 1- 1)O-[diglyceryltetraetherl-OH and Glcp ( a1+2)Galf(pl- 1)-0-[diglyceryltetraetherl-OH. The two glycolipids representing 68% of Sulfolobus lipids are based on the two types of diglycerol and calditol glycerol tetraether species (Langworthy et al., 1974; de Rosa et at., 1980~). A diglycosyl diglycerol tetraether and glycosyl

65


calditol glycerol tetraether are present in about equal proportions in heterotrophically grown cells, but the calditol glycerol tetraether glycolipid is the major derivative in cells grown autotrophically (Langworthy, 1977b). The two glycolipids have been partially characterized as Glcp(P+)Galp(P-+)-O-[diglyceryltetraetherl-OH and Glcp(P-+)-0-[caldityIglyceryltetraether]lOH. The glucosyl-galactosyl disaccharide is linked to one side of the diglycerol tetraether, whereas glucose is linked to one of the -OH groups of the calditol radical in the calditol glycerol tetraether. The glycolipid fraction of Thennc~plusmcirepresents 25% of the lipids but none of the six diglycerol tetraether glycolipids have been identified (Langworthy et ai., 1972). However, an unusual linear lipoglycan (MW 5300) containing 24 mannose, one glucose, and terminating in a diglycerol tetraether has been isolated from Thermoplusrnu (Mayberry-Carson et ul., 1974). The molecule, which can be considered to be a glycolipid with an extended 25-sugar chain, accounts for 3% of the cell dry weight. Its structure (26) has been fully established by Smith ( 1980) to be [ Manp ( a1+2)Manp ( a1-4)Manp ( a1-+3)],-Glcp (al-+l)-O-[diglyceryltetraetherl-OH in which the sugar chain is attached to one side of the diglycerol tetraether molecule. The lipoglycan has physical properties similar to gram-negative lipopolysaccharides (Maybemy-Carson et ul., 1975) and is located on the cell surface (Mayberry-Carson et al., 1978). Its finding in Thermoplasma has led to the isolation of similar diacylglycerol lipoglycans in the Acholeplasrnu species (Smith et ul., 1976). CH,OH

CH,OH

SH,OH

FH,OH

1

CH,OH

D. Other Polar Lipids An ornithine-containing lipid was found by Shively and Knoche ( 1 969) among the lipids of the acidophile T . thiooxiduns. Its structure (27), in which 3-hydroxyhexadecanoic acid is amide-linked to the amino group of omithine and cis- 1 I , 12-methylene-2-hydroxyoctadecanoicacid is ester-linked to the 3-OH group, has been established (Knoche and Shively, 1969, 1972; Hilker et ul., 1978). The biological significance of this lipid is unknown. O=C-CH-(CH, ),-CH-CH-(CH,);-CH, COOH

0 OH H, N-(cH,),-cH-NH-c-cH,-~H-(cH,),IcH, II

I

\

I

CH,

66


V.

ACIDIC LIPIDS

Acidic lipids of eubacteria from extreme environments are composed largely of ordinary phospholipids including either diphosphatidylglycerol (DPG), phosphatidylglycerol (PG), phosphatidylethanolamine (PE), phosphatidylserine (PS), phosphatidylinositol (Pi), phosphatidic acid (PA), or phosphatidyl-N-monomethylethanolamine (PME) as major constituents. Within the halophilic and methanogenic archaebacteria, the diether-derived phospholipids occur as analogues of the respective diacylglycerol phosphatides. However, in the thermoacidophilic and methanogenic archaebacteria possessing tetraethers, the tetraether phospholipids established thus far exist almost exclusively as phosphoglycolipids in which the carbohydrate and phosphoryl radicals are attached asymmetrically to opposite ends of the tetraether molecule. Additionally, acidic sulfolipids occur in several eubacterial and archaebacterial species.

A. Phospholipids 1 . PHOSPHATIDES Among the thermophilic Bacillus species, B . steurothermophilus (Card ef d., 1969; Card, 1973) and B . caldotenux (Hasegawa et d.,1980), phospholipids composed mainly of DPG, PG, and PE constitute 60-90% of the total lipids. Hasegawa et u / . (1980) noted a substantial increase in lower melting PG and a decrease in the higher melting PE content when the growth temperature of B. cu/do/yticus was lowered from 65 to 45°C. The extreme thermophile T . apaticus contains DPG, PG, PI, and PA, representing only 20% of the phospholipid fraction (Ray et a / . , 1971b). It contained, in addition, a major unidentified phospholipid having a minimum molecular weight of 1800, which possessed phosphate, three fatty acids, one glycerol, and a long-chain unsaturated amine. It was also noted that the phospholipid content increased 2-fold in cells grown from 50 to 75°C. The four psychrophilic Vibrio species examined by Bhakoo and Herbert (1979) all contained DPG, PG, and PE, but two of the isolates possessed significant quantities of PS. Changes in the phospholipid distribution of some isolates were noted at 20"C, the upper temperature limit for growth, suggesting a thermal-sensitive impairment of phospholipid synthesis. It was also shown that total phospholipid levels increased markedly on decreasing the growth temperature to 0"C, but Cullen et (I/. (1971) found no change in the phospholipid composition of a psychrotrophic Pseudotnonasfluorescens species on decreasing temperatures. The mesophilic acidophiles, T . thiooxidans and T . ferrooxidans, have been shown by Shively and Benson (1967) and Short et a/. (1969) to have phospholipids made up mostly of DPG, PG, PE, and DME. The presence of PS


67

reported by Korczynski et ( I / . (1967) could not be confirmed. The phospholipids of T . ,ferrooxiiluti.s were found to have a slow rate of tnetabolism and no differences in proportions or turnover were found during growth at either pH I .5 or 3.5 (Short (’t u l . , 1969). Of the total lipids from the thermoacidophile B . Nc,iik,ctildurii,.s, 20% are acidic lipids, of which DPG, PG, and traces of PA and PE account for 57% (Langworthy et ( I / . , 1976). The remaining 43% consists of a sulfonolipid (see Section V,C,2). 2. DIETHER ANAI.OGI~ES The phospholipids of extremely halophilic bacteria are composed of the di0-phytanyl glycerol ether analogues of phosphatidyl glycerolphosphate (PCP), PG, and phosphatidylglyceryl- 1 ‘-sulfate (PGS), which constitute about 65 and 4% of the total acidic lipids, respectively (Kates, 1972, 1978; Kates and Kushwaha, 1976). The PGS is exclusively associated with the purple membrane fraction (Kushwaha r t ul.. 1975b). The Dead Sea halophile H . rriarisrirortiri contains the same phospholipids but differs from the other halophiles examined by having a significantly higher amount ( 1 7%) of PGS, perhaps compensating 1980; see for the deficit of any glycolipid sulfate in this organism (Evans et d., Section V , C , I ) . In M . hungutei, a small amount of PG (5%) was found, and this is the only di-0-phytanyl glycerol ether analogue to be reported thus far in methanogens (Kushwaha rt ul., I9Xla,b).

B. Phosphoglycolipids Phosphoglycolipids are the major constituents of archaebacteria that contain tetraether residues. In M . hungutc’i the phosphoglycolipids occur as glyceryl phosphoryl derivatives of the two diglycerol tetraether glycolipids in which sn3-glyceryl phosphate is linked to the free -OH group of the tetraether moiety (Kushwaha et u l . , I98 1a.b). The two phosphoglycolipids have been established as Galfw 1 -6)GalfG 1- 1)-0-[diglyceryltetraetherl-0-PO,-glycerol and Glcp(cu l b 2 ) G a l f w l-+ 1)-0-[diglyceryltetraether]-0-P03-glycerol.They represent 14 and 50% of the total lipids, respectively. These compounds are clearly the structural analogues of the di-0-phytanyl glycerol glycolipids that have been covalently condensed “head to head” with the di-0-phytanyl phosphatidylglycerol present in the organism. This lends strong support to the view that tetraether lipids are derived biogenically via condensation through the geminal ends of the C,,,-phytanyl chains of either free diethers or the complex lipids themselves. The three phospholipids of Su(fo/obirs are inositol phosphate derivatives, representing about 21% of the total lipids (Langworthy et al., 1974; de Rosa et ul.,

68

THOMAS A. IANGWORTHY

1 9 8 0 ~ ) .Present in close proportion are the tetraether analogue of phosphatidylinositol, inositol-OP0,-[diglyceryltetraetherl-OH,and the inositolphosphoryl derivatives of the two partially characterized glycolipids; Glcp(P+)Galp(P-)- 0-[diglyceryltetraetherl-OH and Glcp(P-)-O- [calditylglyceryltetraetherl-OH. The location of inositol phosphate residues on the latter two glycolipids has not been established. Phosphoglycolipids constitute about 57% of the total lipids from Thrrmoplusmu, all containing glycerol phosphate residues (Langworthy et ul., 1972). Although only partly characterized at a time prior to the recognition of tetraethers (Langworthy, 1977a), Thermoplusmu can be described as containing a glycerylphosphoryl monoglycosyl diglycerol tetraether that accounts for 80% of the lipid phosphorus and nearly half of the entire lipids of the organism. Four minor components include amine-containing diglycerol tetraether phosphoglycolipids. Thus, among archaebacteria containing tetraethers, essentially the total complex lipids (glycolipids plus acidic lipids) contain carbohydrate residues.

C. Sulfolipids 1 . SULFATIDES

Sulfate-containing glycolipids occur in Sulfolobus and most halophilic archaebacteria. About 6% of the acidic lipids of Su(fo1obus are composed of the sulfate derivative of its partially characterized tetraether glycolipid, Glcp(P--+)-0-[calditylglyceryltetraetherl-OH,but the location of the sulfate residue has not been determined (Langworthy rt ul., 1974; Langworthy, 1977b; de Rosa et ul., 1 9 8 0 ~ ) Most . halophiles examined, with the exception of H . mcirismortui, contain the triglycosyl diether sulfate (28) identified as -O:, SO3-Galpp 1+6)-Manp(a 1-+2)Glcp(a 1- 1)2,3-di-O-phytanyI-sn-glycerol(Kates and Deroo, 1973; Kates, 1978). It represents about 25% of the lipids, and thus combined with the phospholipids makes essentially all of the polar lipids of extreme halophiles acidic. Its function is unknown, but it is associated with the purple membrane (Kushwaha rt ul., 1975b), and it has been speculated that it might serve as a proton donor for the functioning of the purple membrane as a light-driven proton pump (Kates and Kushwaha, 1978). CH*OH

qojq

OH

o&

CH,-O-C,,H,,

I

I

CH-O-C2,JH41

69


2.

SULFONOLIPrDS

Sulfonolipids, containing the C-SO, bond, rarely occur in nonphotosynthetic bacteria (Haines, 197 1). However, the thermoacidophile B . ucidocaldarius contains a sulfonolipid (29) that appears identical to the plant sulfonoquinovosyldiacylglycerol, -03S-6-quinovosyl(a 1+1) 1,2-di-0-acyl-sn-glycerol (Langworthy et al., 1976). It represents nearly half (43%) of the acidic lipids and 8% of the total lipids of the organism. Its function is unknown, but since sulfonolipids are the most acidic lipids known, being ionized at all pH values, it was speculated (Langworthy et al., 1976) that it might serve in H+-ion exclusion or perhaps as a cation exchanger in the acidic environment. 0

I1

CH,SOJ

CH2-O-C--R I CH -0- C--R

I

0-CHz

II

0

OH

(29)

VI.

OVERVIEW

Although studies are necessarily incomplete, it is clear that bacteria from extreme environments contain many ordinary as well as quite unusual lipid structures. Perhaps the most significant class of lipids found to occur are the isopranyl ether lipids. Although of interest in their own right in terms of their individual biochemistry and how they might function in membrane stabilization in hostile environments, they also extend significant insight into our perceptions of taxonomy, evolution, biogeochemistry, and the molecular organization of membranes. The isopranyl ether lipids serve not only as a chemical marker for identifying archaebacteria, but also strongly support the taxonomic relationship of these organisms, which appear to have evolved through a line of descent different from either eukaryotic or prokaryotic cells. Moreover, the same range of archaebacterial isoprenoids, including phytane and “head to head” linked biphytanes, have recently been found in sediments (Anderson et al., 1977), kerogen (Michaelis and Albrecht, 1979), shale (Chappe et al., 1979), and petroleum (Moldowan and Seifert, 1979). Thus the fact that Thermoplasma, Sulfolobus, halophiles, and methanogens live in environments presumably dominating earlier periods of the Earth’s geological evolution suggests that the isoprenoids found in sediments and petroleum could have been synthesized directly by archaebacteria. Perhaps some of these organisms are actively involved in petrogenesis. Simi-

70


larly, the geochemical occurrence of hopanoids can now be ascribed to a bacterial origin (Rohmer rt (41.. 1979). In terms of molecular organization, the elevated content of carbohydratecontaining lipids in many of the bacteria from extreme environments indicates that the often neglected glycolipid class may have a significant function in controlling membrane stability. In addition, the molecular organization of membrane proteins, energy transduction, and transport of the tetraether-derived monolayer membranes will be of considerable interest, and will also provide a useful model for assessing our ideas of normal lipid bilayer systems. The currently recognized and yet to be discovered bacteria from extreme or unusual environments should continue to provide a rich source of material for study. New lipids will surely be uncovered to test the ingenuity and patience of the investigator. ACKNOWLEDGMENTS The author thanks R. Uecker for structural illustrations and J . Ratzlaff for editorial assistance. Portions of the author’s work described herein were supported by a grant from the National Science Foundation (PCM-7809351). REFERENCES Allen, M. B. (1953). The thermophilic aerobic sporeforming bacteria. Bacteriol. R e v . 17, 125-173. Amdur, R . H., Szabo. E. I . , and Socransky, S . S . (1978). Presence of squalene in gram-positive bacteria. J . Bacreriol. 135, 161-163. Amelunxen, R . E., and Murdock. A. L. (1978). Mechanisms of thermophily. CRC Crir. Rev. Microhiol. 6 . 343-393. Anderson, R . , Kales, M. , Baedeckcr, M . J .. Kaplan, I. R . , and Ackman, K.G . ( 1977). The stereoisomeric compositionof phytiinyl chains in lipids of Dead Sea sediments. Gf,o~.him.Cosmoc.him. A&l 41, 1381-1390. Balch, W. E., Fox, G. E., Magrum, L. J . , Woese. C. R., and Wolfe, R. S . (1979). Methanogens: Reevaluation of a unique biological group. Microhid. Re,,. 43, 260-296. Baross, J . A , , and Morita, R. Y . (1978). Life at low temperatures: Ecological aspects. I n “Microbial Life in Extreme Environments” (D. J . Kushner, ed.), pp. 9-71, Academic Press, New York. Bayley, S. T . , and Morton, R. A. (1978). Recent developments in the molecular biology of extremely halophilic bacteria. CRC Crit. Rev. Mirrohiol. 6 , 151-205. Beam, H . W., and Perry, J . J . (1974). Microbial degradation and assimilation of ti-alkyl-substituted cycloparaffins. J . Butterid. 118, 394-399. Belly, R. T., Bohlool. B. B., and Brock, T . D. (1973). The genus 7 h u r m o p l u s m . Ariri. N . Y . Acud. Sri. 225, 94-107. Bhakoo, M., and Herbert, R. A. (1979). The effects of temperature on the fatty acid and phospholipid composition of four obligately psychrophilic Vihrio spp. A r r h . Microbiol. 121, 121-127. Blume, A , , Dreher, R . . and Poralla, K . (1978). The intluence of branched-chain and o-alicyclic fatty acida on the transition temperature of Bucillrr.~suhrilis lipids. Biochim. Biophvs. Acru 512, 489-494. Brierley, C. L. (1978). Bacterial leaching. CRC Crir. Re),. Microbiol. 6 , 207-262. Brock, T. D. (1978). “Thermophilic Microorganisms and Life at High Temperatures. ’’ SpringerVerlag, Berlin and New York.


71

Brock, T. D., and Freeze. H. (1969). Theritirrs uqiruricirs gen. n . and sp. n . , a nonsporulating extreme thermophile. J . Bucteriol. 98, 289-297. Brock, T. D., Brock, K . M . , Belly, R. T . , and Weiss, R. L. (1972). Su/folohu.s: A new genus of sulfur-oxidizing bacteria living at low pH and high temperature. Arch. Mierohiol. 84, 54-68. Brown, C . M., and Minnikin, D. E. (1973). The effect of growth temperature on the fatty acid composition of some psychrophilic marine pseudomonads. J . G m . Microhiol. 75, IX. Caplan, S. R., and Ginzburg, M., eds. (1978). “Energetics and Structure of Halophilic Microorganisms. ElsevieriNorth-Holland Publ., Amsterdam. Card, G. L. (1973). Metabolism of phosphatidylglycerol, phoaphatidylethanolamine, and cardiolipin of Btrcillrrs stecirothermophilus. J . Bucteriol. 114, I 1 25- I 137. Card, G. L., Georgi, C. E . , and Militzer. W . E. (1969). Phospholipids from Bacillirs .rtearotherinophilus. J . Bacteriol. 97, I 86- 192. Castenholz, R. W. (1979). Evolution and ecology of thermophilic microorganisms. In “Strategies of Microbial Life in Extreme Environments” (M. Shilo, ed.), pp. 373-392. Verlag Chemie, Weinheim. Chan, M . , Himes, R . H.. and Akagi, J . M . (1971). Fatty acid composition of thermophilic, mesophilic and psychrophilic cloatridia. J . Bacteriol. 106, 876-881. Chappe, B. (Chap Sim). Michaelis, W., Albrecht, and Ourisson, G. (1979). Fossil evidence for a novel series of archaebacterial lipids. Naturwissenschuften 66, 522-523. Cho, K. Y.,and Salton, M . R. J . (1966). Fatty acid composition of bacterial membrane and wall lipids. Bioehiin. Biophys. Acta 116, 73-79. Cullen, J . , Phillips, M. C.. and Shipley, G . G. (1971). The effectsoftemperature on the composition and physical properties of the lipids of Psriidotnonasfluorescens.Biochein. J . 125, 733-742. Darland. G . , and Brock. T. D. (1971 ). Bucillu.~trcirloc.alrlcrriIrs sp. nov., an acidophilic thermophilic spore-forming bacterium. J . Gen. Microhiol. 67, 9- 15. Daron, H. H. (1970). Fatty acid composition of lipid extracts of a thermophilic Bacillus species. J . Bacteriol. 101, 145-151. de Rosa, M., Gambacorta, A , , Minale, L., and Bu’Lock, J. D. (1972). The formation of w-cyclohexyl-fatty acids from shikimate in an acidophilic thermophilic bacillus. Biocheni. J . 128, 751-754. de Rosa, M., Gambacorta, A , , Minale, L.. and Bu’Lock J . D. (1973). Isoprenoids of Bacillrts ucidoculdurirrs. Phytocheniistry 12, 1 1 17- 1 123. de Rosa, M., Gambacorta, A , , and Bu’Lock, J . D. (1974). EffectsofpH and temperature on the fatty acid composition of Bacillus ricidoc~uldarirts.J . Bacteriol. 117, 212-214. de Rosa, M., Gambacorta, A., and Bu’Lock, J . D. (1975a).Extremely thermophilic acidophilic bacteria convergent with Sulfolohirs acidocaldurirts. J . Gen. Microbiol. 86, 156- 164. de Rosa, M., Gambacorta. A , , and Minale, L. (197%). A terpenoid 4.7-thianaphthenequinone from an extremely thermophilic and acidophilic micro-organism. J . Chem. Soc., Chem. Coiiirnun. pp. 392-393. de Rosa, M., Gambacorta. A , , and Bu’Lock, J . D. (1976a).The Caldariella group of extreme thermoacidophile bacteria: Direct comparison of lipids in Sulfi,lohus, Thermoplasmu, and the MT strains. Phyrochemistry 15, 143-145. de Rosa, M . , de Rosa, S . , Gambacorta, A , , and Bu’Lock, J. D. (1976b). Isoprenoid triether lipids from Caldariella. Phytocheiiiistry 15, 1996-1997. de Rosa, M., de Rosa, S . , and Gambacorta, A . (1977a). ‘:’C-NMR assignments and biosynthetic data for the ether lipids of Caldariellu. Phyrochcmistg 16, 1909-1912. de Rosa, M., de Rosa, S . , Gambacorta, A . , Minale, L., and Bu’Lock, J. D. (1977b). Chemical structure of the ether lipids of thermophilic acidophilic bacteria of the Calrluriella group. Phytochemistry 16, 1961-1965. de Rosa, M., de Rosa, S . , Gambacorta, A . , and Bu’Lock, J . D. (1980a). Structure of calditol, a new ”

72


branched-chain nonitol. and of the derived tetraether lipids in thermoacidophile archaebacteria of the Culduriellu group. Phytochernistp 19, 249-254. de Rosa, M., Gambacorta, A,, and Nicolaus, B. (1980b). Regularity of isoprenoid biosynthesis in the ether lipids of archaebacteria. Phytochernistry 19, 791 -793. de Rosa, M., Esposito, E., Gambacona, A,, Nicolaus, B., and Bu’Lock, J. D. (1980~).Complex lipids of Culdurieflu ucidophilu, a thermoacidophile archaebacterium. Phytochernistv 19, 821 -825. de Rosa, M., Eposito, E., Gambacorta, A,, Nicolaus, B., and Bu’Lock, J . D. (1980d). Effects of temperature on ether lipid composition of Culduriello ucidophilu. Phytochemistry 19, 827831. de Rosa, M., Gambacorta, A.. Nicolaus, B., Sodano, S . , and Bu’Lock, J . D. (1980e). Structural regularities in tetraether lipids of Culluriello and their biosynthetic and phyletic implications. P hytochernistry 19, 833-836. Dundas. 1. E. D. (1977). Physiology of the Hulubucteriaceue. Adv. Microh. Physiol. 15, 85-120. Evans, R. W . , Kushwaha, S . C., and Kates, M. (1980). The lipids of Hulobacteriurn rnurisrnorrui. an extremely halophilic bacterium in the Dead Sea. Eiochim. BiophJs. Arra 619, 533-544. Fox, G. E., Stackenbrandt, E., Hespell, R. B., Gibson, J., Maniloff, J., Dyer, T. A,, Wolfe, R . S., Balch, W. E., Tanner. R. S . . Magrum, L. J., Zablen, L. B., Blakemore, R., Gupta, R., Bonen, L., Lewis, B. J . , Stahl, D. A., Luehrsen. K . R., Chen. K . N., and Woese, C. R. (1980). The phylogeny of prokaryotes. Science 209, 457-463. Friedman, S . M., ed. (1978). “Biochemistry of Thermophily.” Academic Press, New York. Furuya, T . , Nagumo, T.. Itoh, T . . and Kaneko, H. (1980). The effect of growth temperature on the lipids in an extremely thermoacidophilic bacterium, TA-I. Agric. B i d . Chern. 44,517-521. Gill, C. 0 . .and Suisted. J . R. (1978). The effects of temperature and growth rate on the proportion of unsaturated fatty acids in bacterial lipids. J . Gen. Microhiol. 104, 31-36. Haines, T. H. (1971). The chemistry of the sulfolipids. Prog. Chern. Furs Other Lipids 2,297-345. Hasegawa. Y.. Kawada, N., and Nosoh, Y,(1980). Change in chemical composition of membrane of Eaciltus ctrltlotmux after shifting the growth temperature. Arch. Microhiol. 126, 103- 108. Heinen, U. J . , and Heinen, W. (1972). Characteristics and properties of a caldo-active bacterium producing extracellular enzymes and two related strains. Arch. Mikrubiul. 82, 1-23. Heinen, W.. Klein, H. P., and Volkmann, C. M. (1970). Fatty acid composition of Tlterrnrrs tiquuticu.~at different growth temperatures. Arch. Mikrohiol. 72, 199-202. of Microbial Adaptation. ” Heinrich. M. R . , ed. (1976). “Extreme Environments-Mechanisms Academic Press, New York. Hilker. D. R . , Gross. M. L.. and Knocke, H. W . (1978). The Interpretation ofthc mass spectrum of an ornithine-containing lipid from Thiohacillus thiooxidnns. Hiurned. Muss Spectrum. 5 , 64-7 I . Holzer, G . , Orb, J., and Tornabene, T. G. (1979). Gas chromatographic-mass spectrometric analysis of neutral lipids from methanogenic and thermoacidophilic bacteria. J . Chrornarogr. 186, 795-809. Inniss. W . E.. and Ingraham. J . L. (1978). Microbial life at low temperatures: Mechanisms and molecular aspects. In “Microbial Life in Extreme Environments” (D. J. Kushner, ed.), pp. 73-104. Academic Press, New York. Jackson, T . J . , Ramalcy. R . F., and Meinschein, W. G. (1973). Thcrrnornicrubiurn, a new genus of extremely thermophilic bacteria. Inr. J . Syst. Eac~trriol.2.1.28-36. Kannenbcrg. E . . Poralla, K.. and Blume, A. (1980). A hopanoid from thermo-acidophilic Eucillus ucidoculdurius condenses membranes. Nuturwissenchuften 67,458-459. Katcs, M. (1972). Ether-linked lipids in extremely halophilic bacteria. I n “Ether Lipids: Chemistry and Biology” (F. Snyder, ed.). pp. 351-398. Academic Press, New York.


73

Kates, M. (1978). The phytanyl ether-linked polar lipids and isoprenoid neutral lipids of extremely halophilic bacteria. f r o g . Chtm. Futs Othor Lipids 15, 301-342. Kates, M . , and Deroo, P. W. (1973). Structure determination of the glycolipid sulfate from the extreme halophile Hu/ohoctrrium c.utiruhrum. J . Lipid Res. 14,438-445. Kates, M., and Hagen, P . - 0 . (1964). Influence of temperature on fatty acid composition of psychrophilic and mesophilic Srrrutiu species. Cun. J . Biochrm. 42, 481 -488. Kates. M., and Kushwaha, S. C. (1976). The diphytanyl glycerol ether analogues of phospholipids and glycolipids in membranes of Hulohrrcterium cutiruhrum. In “Lipids” (R. Paoletti, G . Procellati, and G. Jacini, eds.), Vol. I , pp. 267-275. Raven, New York. Kates, M., and Kushwaha. S . C. (1978). Biochemistry of the lipids of extremely halophilic bacteria, I n “Energetics and Structure of Halophilic Microorganisms” ( S . R. Caplan and M. Ginzburg, eds.), pp. 461-480. ElaevieriNorth-Holland Publ., Amsterdam. Knoche, H. W . , and Shively. J . M. (1969). The identification of cis-l I , 12-methylene-2hydroxyoctadecanoic acid from Thiolmcillus rhiooxid(rns. J . B i d . Chem. 244,4773-4778. Knoche, H. W.. and Shively. J . M. (1972). The structure of an ornithine-containing lipid from Thiob(1cillu.s thi[)[).~i(/~ln.s. J . Bioi. Cht~m.247, 170- 178. Korczynski, M. S . , Agate. A . D., and Lundgren, D. G . (1967). Phospholipids from thechemoautotroph Ferrohtrc.i//rrsfi.rroo.ri(/uns. Bioc.hc,m. BiophyJ, Ros. Commun. 29, 457-462. Kramer, J. K. C.. Kushwaka, S . C.. and Kates. M. (1972). Structure determination of squalene. dihydrosqualene and tctrahydrosqualene in Ha/ohucteriurn cutiruhrum. B i ( ~ ~ h i rBiophys. n. Actrr 270, 103-1 10. Kushner. D. J . , ed. (1978a). “Microbial Life in Extreme Environments.” Academic Press, New York. Kushner, D. J . (1978b). Life in high salt and solute concentrations: Halophilic bacteria. In “Microbial Life in Extreme Environments” (D.J . Kushner, ed.), pp. 3 17-368. Academic Press, New York. Kushwaha, S. C . , and Kates, M. (1973). Isolation and identification of “bacteriorhodopsin” and minor C ,,,carotenoids in HalohrrctcvYum curiruhrum. Riochim. Biopliy.~.Ac.tn 316, 235-243. Kushwaha, S . C., and Kates, M. (1978a). Mevalonic acid concentrations in halophilic bacteria. Phyrochrmisrry 17, 1793. Kushwaha, S. C., and Kates, M. (lg78h). 2.3-Di-0-phytanyl-sn-glyceroland prenols from extremely halophilic bacteria. Phvtocliernistry 17, 2029-2030. Kushwaha, S . C . , Pugh. E. L.. Kramer, J . K . G . . and Kates, M. (1972). Isolation and identification of dehydrosqualene and C ,,,-carotenoid pigments in Hulohuctrrium cutirubrum. Biochirn. Biophys. A m 260, 492-506. Kushwaha, S . C . , Gochnauer, M . B . . Kushner, D. J . , and Kates, M. (1974). Pigments and isoprenoid compounds in extremely and moderately halophilic bacteria. Can. J . Mic.robio/. 20, 24 I-245. Kushwaha, S. C., Kramer, J . K . C . . and Kates, M. (1975a). Isolation and characterization of C,,,carotenoid pigments and othcr polar isoprenoids from Hulohuc,tc,rium cutiruhrum. Biorhim. Biophy.~.Ai.ru 398, 303-3 14. Kushwaha, S. C., Kates, M., and Martin, W. G. (197%). Characterization and composition of the purple and red membrane from ~ ( i ~ ~ ) ~ r r ( .cutit-uhrum. ~ e r i ~ ~ , nCun. J . Biochrm. 53, 284-292. Kushwaha, S. C., Kates, M.. and Kramer, J . K . G. (1977). Occurrence of indole in cells of extremely halophilic bacteria. C‘un. J . Mic.rohiol. 23, 826-828. Kushwaha, S. C.. Kates. M.. Sprott, G . D., and Smith, I . C. P. ( I 98 la). Novel complex polar lipids froin the methanogen Mc,rhrrno.c/Jtril/umhungatei. S(,ienc,r 21 1 , I 163-1 164. Kushwaha, S. C., Kates. M . , Sprott. G . D.. and Smith, I . C. P. (I981b). Novel polar lipids from the methanogen Mc~thufiospirillu,,~ hungutei GPI. Bioc,him. Bioph,~.~. Acta 664, 156- 173.

74

THOMAS A. LANGWOATHY

Langworthy, T. A. ( 1977a). Long-chain diglycerol tetraethers from Thermoplasmu ucidophilum. Bioclrim. Biophys. Acru 487, 37-50. Langworthy, T. A. (l977b). Comparative lipid composition of heterotrophically and autotrophically grown Sulfolobus acidocaldurius. J . Bacteriol. 130, 1326- 1332. Langworthy, T. A . (1978a). Microbial life in extreme pH values. In “Microbial Life in Extreme Environments” (D. J. Kushner, ed.), pp. 279-315. Academic Press, New York. Langworthy, T. A. ( I 978b). Membranes and lipids of extremely thermoacidophilic microorganisms. In “Biochemistry of Thermophily” (S. M. Friedman, ed.), pp. 11-30. Academic Press, New York . Langworthy, T . A. (1979a). Special features of thennoplasmas. In “The Mycoplasmas” (M. F. Barile and S. Razin, eds.), Vol. I , pp. 495-513. Academic Press, New York. Langworthy, T. A . (l979b). Membrane structure of thermoacidophilic bacteria. In “Strategies of Microbial Life in Extreme Environments” (M. Shilo, ed.), pp. 417-432. Verlag Chemie, Weinheim. Langworthy, T . A. ( 1 980a). Archaebacterial membrane assembly. i n “Dissipative Structures and Spatistemporal Organization Studies in Biomedical Research” (G. P. Scott and J. M. McMillin, eds.), pp. 82-102. Iowa State Univ. Press, Ames. Langworthy, T . A. (1980b). Turnover of di-0-phytanyl glycerol in Thermoplasma. Absrr. Conf. In?. Org. Mycoplasmol. 3rd, 1980, p. 151. Langworthy, T. A. (1981). Diglyceryl tetraether lipids. In “Ether Lipids: Biomedical Aspects” (H. K . Mangold and F. Paltauf. eds.). Academic Press, New York (in press). Langworthy, T. A.. and Mayberry, W. R . (1976). A I ,2,3,4-tetrahydroxy pentane-substituted pentacylcic triterpene from Bacillus acidocddarius. Biuchirn. Biophys. Actu 431, 570-577. Langworthy, T. A., Smith, P. F., and Mayberry, W. R. (1972). Lipids of Thermoplusmcc ucidophilum. J . Racreriol. 112, 1 193-1200. Langwonhy, T . A., Mayberry, W. R . , and Smith, P. F. (1974). Long chain diether and polyol dialkyl glycerol triether lipids of Sulfolohus acidocaldurius. J . Bacteriol. 119, 106- I 16. Langworthy, T. A., Mayberry, W. R . , and Smith, P. F. (1976). A sulfonolipid and novel glucosamidyl glycolipids from the extreme tbermoacidophile Bucillus acidocaldarius. Biochim. Biophys. Actu 431, 550-569. Lanyi. J . K . (1979). Physiochemical aspects of salt-dependence in halobacteria. In “Strategies of Microbial Life in Extreme Environments” (M. Shilo, ed.), pp. 93-107. Verlag Chemie, Weinheim. Levin, R. A. (1971). Fatty acids o f Thir~liacillusthiooxidans. J . Bucreriol. 108, 992-995. Levin, R. A. (1972). Effect of cultural conditions on the fatty acid composition of Thiobacillus novellus. J . Bacreriol. 112, 903-909. Ljungdahl, L. G . (1979). Physiology of thermophilic bacteria. Adv. Microb. Physiol. 19, 149-243. Lundgren, D. G., Andersen, K. J., Remsen, C. C., and Mahoney, R. P. (1964). Culture, structure and physiology of the chcmoautotroph Ferrohacillus ferrooxidans. Drv. I d . Microbial. 6 , 250-259. Lundgren, D. G . , Vestal, J . R . , and Tabita, F. R. (1974). The iron oxidizing bacteria. In “Microbial Iron Metabolism” (J. B. Neilands, ed.), pp. 457-473. Academic Press, New York. Makula. R. A., and Singer, M. E. (1978). Ether-containing lipids of methanogenic bacteria. Bioclzrm. Biophys. Res. Commun. 82,716-722. Marr, A. G., and Ingraham, J. L. (1962). Effect of temperature on the composition of fatty acids in Esc-herichia c d i . J . Brwreriol. 84, 1260- 1267. Mayberry-Carson, K. J . , Langworthy, T. A , , Mayberry, W. R., and Smith, P. F. (1974). A new class of lipopolysaccharide from Thrrmoplasma ucidophilum. Biochim. Biophy,r. Acta 360, 2 17-229.


75

Mayberry-Carson, K . J . . Roth, 1. L . , and Smith, P. F. (1975). Ultrastructure of lipopolysaccharide isolated from Thrrmopkrsmti trcidophihm. J . Bncteriol. 121, 700-703. Mayberry-Carson, K . J . , Jewell. M. J . , and Smith, P. F. (1978). Ultrastructural location of Thermop/osmcr trcidophilum surface carbohydrate by using concanavalin A. J . Bac~teriol. 133, 15 10-15 13. Mercer, E. I . , Modi, N., Clarke, D. J . . and Morris. J. C. (1979). The occurrence and location of squalene in Clostridium pusrcrtn‘unum. J . Gon. Miuohiol. 1 1 1, 437-440. Merkel, G . J . , and Perry, J. J . (1977). Effect of growth substrate on thermal death of thermophilic bacteria. Appl. Environ. Microhiol. 34, 626-629. Michaelis, W . , and Albrecht. P. (1979). Molecular fossils of Archaebacteria in kerogen. Naturwissenshufrcn 66, 420-422. Moldowan, J . M., and Seifert, W. K . (1979). Head to head linked hydrocarbons in petroleum. S c i r n w 204, 169- 17 1 Morita. R. Y. (1975). Psychrophilic bacteria. Bacterial. R t v , 39, 144-167. Oesterhelt, D . , and Stoeckenius, W . (1973). Functions of new photoreceptor memhranc. Proc. N u t / . Aeud. Sci, U . S . A . 70, 2853-2857. 00,K. C . , and Lee. K . L. ( 1971). The lipid content of RociMus .Ffearothrrrnophifusat 37“and at 55”. J . G m . Microhiol. 69. 287-289. Oshima, M . (1978). Structure and function of membrane lipids in thermophilic bacteria. I n “Biochemistry of Thermophily” ( S . M. Friedman. ed.). pp. 1-10. Academic Press, New Y ork . Oshima, M., and Ariga, T . (1975). w-Cyclohexyl fatty acids in acidophilic thermophllic bacteria. J . B i d . Chem. 250, 6963-6968. Oshima, M . , and Ariga, T. (1976). Analysis of the anomeric configuration of a galactofuranose containing glycolipid from an extreme thermophile. FEBS Lrtt. 64, 440-442. Oshima, M., and Miyagawa, A. (1974). Comparative studies on the fatty acid composition of moderately and extremely thermophilic bacteria. Lipids 9 , 476-480. Oshima, M., and Yamakawa, T . (1974). Chemical structure of a novel glycolipid from an extreme thermophile, Flnv~htrc~terium thc~rmophilun7.Biochrmistry 13, I 140- I 145. Oshima, M . , Sakaki, Y . . and Oshima. T . (1978). w-Cyclohexyl fatty acids in acido-thermophilic bacterial membranes and phage capsids. In “Biochemistry of Thermophily” ( S . M. Friedman. ed.), pp. 31-44. Academic Press, New York. Oshima, T.. and Imahori, K. (1974). Description of Thmnus /h<wmophi/rr.s(Yoshida and Oshima) Comb. nov.. a non-sporulating thermophilic bacterium from a Japanese thermal spa. Ini. J. Syst. Bac,tc,riol. 24, 102- I 12. Ourisson, G . , Albrecht. P , and Rohmer. M. (1979). The hopanoids. Pure Appl. Chcm. 51, 709729. Pask-Hughes. R. A . , Mozaffdry, H . , and Shaw, N. (1977). Glycolipids in procaryotic cells. Bioc,hrm. Soc.. T r m r . 5 , 1675-1677. Phillips, W. E., Jr.. and Perry, J . J. (1976). Thmnomicwhium fi).sfrri sp. nov. a hydrocarbonutilizing obligate thermophile. IN/. J . Sjw. Boc.;eriol. 26, 220-225. Pordk+, K . , Kannenberg, E . , and Blume, A. (1980). A glycolipid containing hopanc isolated from the acidophilic, thermophilic Bac~illu.rnc,idoccildrrriu.s, has a cholesterol-like function in mcmbranes. FEBS Lcrt. 113, 107-1 10. Ray, P. H., White, D. C.. and Brock. T. D. (1971a). Effect of temperature on the fatty acid composition of Therrnu.\ ciquuticu. J . Bnctrriol. 106. 25-30. Ray, P. H . , White, D. C., and Brock, T . D . (1971b). Effect of growth temperature on the lipid composition of Therrnus tiqictrtic~rrs.J . Btrc./c,riol. 108, 227-235. Rohmer, M., Bovier. P . . and Ourisson. C. (1979). Molecular evolution of biomembranes: Structural

76


cquivalents and phylogenetic precursors of sterols. Prnc. N u t / . Accrd. P i . U.S.A. 76, 847851. Russell. N. J (1971). Alteration in fatty acid chain length in MicrocYJc.cus c,ryophilu.s grown at different temperatures. Biochim. B i o p h y . A&i 231, 254-256. Saiki. T . . Kimura. R., and Arima, K. (1972). Isolation and characterization of extremely thermophilic bacteria from hot springs. A g r i c . B i d . Chem. 36, 2357-2366. Sharom. F., Barratt, D. G . , Thede, A. E., and Grant, C. W. M. (1976). Glycolipids in model membranes; spin label and freeze-etch studies. Bioc-him. Biophys. Acra 455, 485-492. Shen. P. Y.. Coles, E . , Foote, J . L., and Stenesh, J . (1970). Fatty acid distribution in mesophilic and thermophilic strains of the genus Burillus. J. Rucreriol. 103, 479-48 1. Shilo, M., ed. ( I 979). “Strategies of Microbial Life in Extreme Environments.” Verlag Chemie, Weinheim. Shively, J . M.. and Benson, A. A . (1967). Phospholipids of Thiohacillus rhiooxiduns. J. Bac’furicd. Y4, 1679- 1683. Shively, J . M., and Knoche. H. W. (1969). Isolation of an ornithine-containing lipid from Thiohuc~il/us rhiotzriilrrn.s. J. Bocteriol. 98, 829-830. Short, S . A., White, D. C., and Aleem, M. I. H. (1969). Phospholipid metabolism in Ferrohuri/lirs ferrooxiduns. J. Brrc,terio/. 99, 142-1 50. Sinclair. N. A . , and Stokes, J. L. (1964). Isolation of obligately anaerobic psychrophilic bacteria. J. Bucrrriol. 87, 562-565. Smith. P. F. (1980). Sequence and glycosidic bond arrangement of sugars in lipvpulysaccharide from Tliermoplasmu acidophilum. Biochim. Biuphys. Actti 619, 367-373. Smith, P. F., Langworthy, T. A , , Maybeny, W. R., and Houghland, A. E. (1973). Characterization of the membranes of Thermopkusmci uckhphilitm. J. Barrerid. 116, 1019-1028. Smith, P. F., Langworthy, T . A , , and Mayberry, W. R. (1976). Distribution and composition of lipopolysaccharides from mycoplasmas. J. Bacferiol. 125, 9 16-922. Tansey, M. R . , and Brock, T. D. (1978). Microbial life at high temperatures: Ecological aspects. In “Microbial Life in Extreme Environments”(D. J . Kushner, ed.). pp. 160-216. Academic Press, New York. Tornabene. T . G . (1978). Non-aerated cultivation of Halohac~reriumcutirrthrum and its effect on cellular squalenes. J . M o l . Evol. 11, 253-257. Tornabene, T . G . , and Langworthy, T . A . (1979). Diphytanyl and dibiphytanyl glycerol ether lipids of methanogenic archaehacteria. Science 203, 5 1-53. Tornahene, T . G.. Katcs, M . , Gelpi, E., and Oro, J . (1969). Occurrence of squalene, di-, and tetrahydrosqualenes and vitamin MK, in an extremely halophilic bacteriuni, Halobacterium cmriru/wurn. J. Lipid Res. 10, 294-303. Tomabene, T . G . , Wolfe, R. S . , Balch, W. E., Holzer, G., Fox, G. E., and Orb, J . (1978). Phytanyl-glycerol ethers and squalene in the archaebacterium Methrmohucteriurn thtrmotrutorrophicum. J . M u / . Evol. 1 1 , 259-256. Tornabene, T . G . , Langworthy, T. A.. Holzer, G., and Orb, J . (1979). Squalenes, phytanes and other isopranoids as major neutral lipids of methanogenic and thermoacidophilic “archaebacteria.” J . M i d . Evol. 13, 73-83. Tuovinen, O . , and Kelly. D. P. (1978). Metabolic transitions in cultures of acidophilic thiobacilli. In “Metallurgical Applications of Bacterial Leaching and Related Microbiological Phenomena” (L. E. Murr, A . E. Torma, and J . A . Brierley, eds.), pp. 61-81. Academic Press, New York. Weerkamp, A,. and Heinen, W . (1972). Effect of temperature on the fatty acid composition of the extreme thermophiles. Bucillus c.rrl&J/-vticus and 3uci//irs culdormur. J . Burteriiil. 109, 443446. Woese, C. R., and Fox, G. E. (1977). Phylogenetic structure of the prokaryotic domain: The primary kingdoms. Proc. Nut/. Acnd. Sci. U.S.A. 74, 5088-5090.


77

Woese. C. R . , Magrum, L. J . , and Fox, G. E. (1978). Archaebacteria. J . M o l . Evol. 11, 245-252. Yang, L. L., and Haug, A . (1979). Structure of membrane lipids and physico-biochemical properties of the plasma membrane from Thermoplusmu ui,idophi/um. adapted to growth at 37°C. Biochim. Biophys. Acta 513, 308-320. Yao, M., Walker, H. H., and Lillard, D. A . (1970). Fatty acids from vegetative cells and spores of Bac,i/lus stL.arothermt,philus. J . Bacteriol. 102, 877-878. Zeikus, J . G. (1977). The biology of methanogenic bacteria. Bucteriul. Rev. 41, 514-541. Zeikus, J . G.( 1979). Thermophilic bacteria: Ecology, physiology and technology. Enzyme M i i w h . Ti,chno/. I , 243-251. Zeikus, J . G., and Wolfe, R . S. (1972). Mrth~inobai,teriumthermoautotrophicum sp. n., an anaerobic. autotrophic extreme thermophile. J . Bacteriol. 109, 707-7 13. Zuber, H., ed. (1976). “Enzymes and Proteins from Thermophilic Microorganisms.” Birkhaeuser, Basel.


CURRENT TOPICS IN MEMBRANES AND TRANSPORT, VOLUME 17

Lipo polysacc harides of G ram-Neg ative Bacteria OTTO LUDERITZ, MARINA A . FREUDENBERG, CHRIS GALANOS, VOLKER LEHMANN, ERNST TH. RIETSCHEL, A N D DEREK H . SHAW* Mux-PIunc,k-lnstitut f u r Irnmunhiologic, Freihurg, Federul Repuhlic of' Gertnrrn~ und

*Northwesf Atlantic Fisheries Centre John's, Newfoundlund, Cunudu

St.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolation, Structure, and Biosynthesis of Lipopolysaccharides . . . . . . . A. Isolation and Purification of Lipopolysaccharides . . . . . . . . . . B. Structure and Biosynthesis of the 0 Chains . . . . . . . . . . . . C . Structure and Biosynthesis of the Core . . . . . . . . . . . . . D. Structure and Biosynthesis of Lipid A . . . . . . . . . . . . . . 111. Some Selected Aspects on the Biology of Lipopolysaccharides . . . . . . A. Endotoxic and Immunogenic Properties of Lipid A . . . . . . . . . B. Physicochemical and Structural Prerequisites for Biological Activities of Lipopolysaccharide . . . . . . . . . . . . . . . . . . . . . . . C . The Fate of Lipopolysaccharides (Lipid A) in Experimental Animals . . IV. Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . I.

11.

1.

. . .

. . . . . .

. . . . . . . . . . .

79 82 82 83 90 91 1 I4 I14 1 I8

123 130 134

INTRODUCTION

Lipopolysaccharides (LPS) form a large, unique class of macromolecules representing a characteristic attribute of gram-negative bacteria. Associated with proteins, they are located in the outer leaflet of the outer membrane of the bacterial cell (Nikaido and Nakae, 1979). In this exposed position on the cell surface, lipopolysaccharides are involved in the interaction of the cell with the environment. Thus contact of the bacterium with the immune system leads to the stimulation of specific antibodies directed predominantly against determinant 79

Copyright 0 1982 hy Academic Press, lnc All rights of reproduction in any form reserved ISBN 0-12-153317-4

80

OTTO LUDERITZ ET AL.

structures of the lipopolysaccharide. Hence, lipopolysaccharides represent the main surface antigens of gram-negative bacteria; these antigens have for many years been known as the 0 antigens (Morgan and Partridge, 1941; Tal and Goebel, 1950; Luderitz ei al., 1966b). A synonymously used term for lipopolysaccharide is endotoxin. The injection of lipopolysaccharide-containingbacteria or of purified lipopolysaccharide into experimental animals causes a wide spectrum of so-called endotoxic reactions. In contrast to the specific, delayed immune response, these effects are, in general, nonspecific and acute, and would include such phenomena as fever, changes in the white cell counts, shock, and death after high doses of lipopolysaccharide (Kadis et al., 1971; Kass and Wolff, 1973; Schlessinger, 1977). Lipopolysaccharides have attracted scientific interest in the past mainly because of their 0-antigenic and endotoxic properties. Lipopolysaccharides often act as specific receptor sites for bacteriophages, and this system has been intensively studied as a model for phage-bacterium interaction (Lindberg, 1977; Braun and Hantke, 1981). The important physiological role of lipopolysaccharide itself has recently been recognized with the demonstration that the outer membrane of the cell acts as a barrier for passive penetration of various compounds (Nikaido, 1979; Nikaido and Nakae, 1979). Low-molecular-weight nutrients and excretion products (< 600 daltons) may pass through the barrier but products of higher molecular weight (such as antibiotics and other poisons) are prevented from crossing the membrane. Lipopolysaccharides are essential for maintaining the integrity of the cell wall (Henning, 1975). Bacteria that, due to a mutational defect, are not able to synthesize lipopolysaccharide, are no longer able to reproduce, and cease growing. Another quality of lipopolysaccharides has been elucidated. It appears that lipopolysaccharides are of importance in the normal physiological bacterium-host relationship. Lipopolysaccharides have the ability to activate and suppress lymphocyte functions (Melchers, 1980; Koenig and Hoffmann, 1979; McGhee et d.,1980). They stimulate polyclonal B lymphocytes to differentiation, proliferation, and secretion of immunoglobulin. This may have special significance, when the omnipresence of gram-negative bacteria in the gut and in the environment is considered. Humans and animals must deal with lipopolysaccharides throughout their lives. The fact that in germ-free animals the immune system is poorly developed indicates that the host requires this continuous exposure for the development of vital physiological systems, such as the immune apparatus (see Schwab, 1977). In the interrelationship of plants and gram-negative bacteria the specific role of rhizobia lipopolysaccharides is presently under active discussion (Wolpert and Albersheim, 1976). There is evidence that the binding of the bacterial cells to the roots is mediated by specific interactions of the lipopolysaccharides with lectins of the plants (see Section Il,B,4). In summary, the lipopolysaccharides of gram-

LIPOPOLYSACCHARIDES OF GRAM-NEGATIVE BACTERIA

81

negative bacteria exhibit a plurality of effects and functions. They represent the 0 antigens, they are the endotoxins expressing their harmful activities during infection, they have indispensable functions for the bacterium and, possiblywhen symbiosis with animals and some plants is considered-also for the host. Just how this class of macromolecules is capable of inducing such a wide variety of effects has been the subject of considerable investigations and speculation. Now, since the structural principles of many lipopolysaccharides have been evaluated, more sophisticated attempts can be undertaken, both to identify biologically active moieties in the molecule, and to study the mode of action of lipopolysaccharide, when it acts as an antigen, an adjuvant, a toxin, and a factor indispensable for !he life of bacteria, and possibly also for the life of other organisms. Of the various lipopolysaccharides studied to date, those of Sulrnonellu are probably the most thoroughly investigated. The results obtained with Salmonetlu lipopolysaccharides may serve as the base to which results on lipopolysaccharides from other bacterial families may be compared. Figure 1 shows the schematic representation of the structure of Salmonella lipopolysaccharides. It contains a lipid region, the lipid A, and a long, covalently linked heteropolysaccharide that, according to composition, structure, and mode of biosynthesis, can be subdivided into the core and the 0-specific chain. These three regions are not only distinct in their chemical structure, but also in their biological and functional properties. In the course of this chapter, we will illustrate that the viability of the cells is dependent on a minimal core-lipid A structure, the 0-antigenic specificity is determined by structures of the 0 chain, and the endotoxic principle of the molecule is expressed by lipid A . The first part of this chapter (Section 11) describes general aspects of the

1

0 Monosaccharide,

---

P h o s p h a t e , ru

Ethanolamine

Long C h a i n l H y d r o x y l F a t t y Acid

FIG. I. Schematic structure of Sulmonelllr lipopolysaccharides. The number of nonhydroxylated and hydroxylated fatty acids given in this scheme is arbitrary.

82


0-specific chains and the core, and principles of their biosynthesis; the structure of Salmonella lipid A will then be discussed, together with the present knowledge of its biosynthesis; and, finally, lipid A's of other gram-negative bacteria, which have recently been investigated will be described. The second part of the chapter (Section 111) will deal with some selected aspects of the biological properties of lipopolysaccharides.

11.

ISOLATION, STRUCTURE, AND BIOSYNTHESIS OF LIPOPOLYSACCHARIDES

A. Isolation and Purification of Llpopolysaccharides A number of extraction procedures for the isolation of lipopolysaccharides or lipopolysaccharide-protein complexes have been described (for a summary, see Galanos er al., 197713; Wilkinson, 1977). The method of choice is usually the phenoUwater procedure (Westphal et al., 1952), which yields a water-soluble extract that is subsequently purified by high-speed centrifugation. These lipopolysaccharide preparations in general contain small amounts of contaminants, such as protein (about 1%). It has occasionally been observed that lipopolysaccharides, which are more lipophilic in nature, partition mainly into the phenol phase during phenol/water extraction. This is also true for lipopolysaccharides from R(rough-)-form bacteria. For the extraction of R mutants, the phenol/chloroform/petroleum ether procedure has been proved to be highly efficient and specific (Galanos et al., 1969; Galanos and Liideritz, 1982). Water-soluble preparations of high purity are thus obtained. Because of the presence of carboxyl, phosphoryl, and ethanolamine residues in the molecule, lipopolysaccharides are amphoteric. The overall charge is negative. In the original lipopolysaccharide preparation, negatively charged groups are neutralized by Na+, K + , Mg2+, and Ca2+, the mixture of cations being variable depending on the culture medium used. In addition, polyamines synthesized by the bacterium are present. It has been found that the nature of cations present in lipopolysaccharides greatly influences their state of aggregation and, as a consequence, their biological activities (see Section III,B, 1). For special purposes, therefore, i t IS useful to convert lipopolysaccharides into a uniform salt form. This is achieved by electrodialysis of the lipopolysaccharide, whereby cations are removed, and the acidic form of the lipopolysaccharide is obtained. Neutralization with base then leads to a defined salt form of the lipopolysaccharide (Galanos and Luderitz, 1975). Recently, the purified lipopolysaccharide of Salmonella abortus equi has been prepared through a number of specific purification steps, including electrodialysis and conversion to the sodium salt form (Galanos et al., 1 9 7 9 ~ ) .This highly purified S . abortus eyui lipopoly-


83

saccharide is available as a standard prepartion in the form of ampules containing 1.0 and 0.1 pg/ml of LPS under the name of Novo-Pyrexal (Henna1 Chemie, K . Hermann. Reinbeck B. Hamburg). It will also soon be available in solid form. It should be kept in mind that in nature lipopolysaccharides d o not occur in pure form. Lysis of bacteria leads to either water-soluble or insoluble complexes containing lipopolysaccharide, proteins, phospholipids, and other components of the bacterial cell. Such complex mixtures are representative of natural endotoxin. Since all the components may exhibit their own endotoxin or endotoxin-like effects, and, moreover, may act synergistically or inhibitory, it is of prime importance to use clearly defined preparations when the mechanisms of endotoxin effects or relationships between structure and biology are to be studied. Lipopolysaccharide must be free of contaminants, proteins, murein, and nucleic acids or, alternatively. defined mixtures of the components should be used. This is a prerequisite for obtaining reproducible results in biological investigations.

B. Structure and Biosynthesis of the 0 Chains 1 . STRUCTURE OF 0 CHAINS

As indicated in Fig. I , the 0-specific chains of lipopolysaccharides are made up of repeating units of identical oligosaccharides (Robbins and Uchida, 1962). These units usually contain different constituents, thus the 0 chain represents a heteropolysaccharide. In some cases, however, the repeating units may contain an oligomer of a single sugar type though i n a distinct linkage sequence, hence repeating units can also be recognized. In these cases the 0 chain represents a homopolysaccharide, for instance, a mannan in Escherichia coli 09 (Prehm et al., 1976b), and a galactan in Klebsiellu 08 (Curvall et ul., 1973). It should be noted that 0 chains have also been identified containing a homopolymer with only one type of linkage such as the PI ,2-linked poly 6-deoxy-~-altropyanose isolated from Yersiniu rnterocolifica (Hoffman et u / . , 1980a). The 0 chains contain the irnmunodeterminant structures against which the anti-0 antibodies formed during infection or on immunization are directed (Luderitz e f ul., 1966b, 1971). Each bacterial serotype synthesizes a unique lipopolysaccharide, characterized by a specific composition and structure of the 0 chain, and by an individual 0 antigenicity. There consequently exist in nature as many distinct lipopolysaccharides as there exist bacterial serotypes. This number is certainly very high. But as far as we know today, irrespective of their individual detailed composition and structure, all lipopolysaccharides are built up according to the structural principle illustrated in Fig. I , with lipid A , core, and 0 chain. Figure 2 shows, as an example, the structure of the lipopolysaccharide of S. fyphirnuriurn. As indicated, lipid A, core, and 0 chain are interlinked by

84

I- -


- -

-

-

--

- - - - I Core

Poiysocchrrde (kgron

i7)

1- - - - - - -

- -

- -I

/

8 I‘

D-GlcN- ( F A )

D-GlcNp- ( F A )

k

-

Lipid A

(Region

z)1- -1

FIG. 2. Structure of the lipopolysaccharide of Sa/tnnnellu fvphimurium (0-antigen factors 4,5.12), as derived from studies by the groups of A . M. Staub, M . J . Oshorn, H. Nikaido, P. W . Robbins, B. Lindberg, and the authors of this chapter. Dotted linkages, incomplete substitutions;_p, pyranose form; L-cr-D-Hep, L-glycero-rr-u-munnoheptose-equivalent to p-L-o-Hep or p-Hep; KDO, 2-keto-3-deoxyoctonate: P, phosphate; FA, fatty acids.

covalent linkages. The 0 chain contains repeating units of a pentasaccharide with mannosyl, rhamnosyl, and galactosyl residues in the main chain, and acetylabequosyl and glucosyl residues as branches. The oligosaccharide (I), which is formed and polymerized during biosynthesis, represents the “biological” repeating unit. Due to the acid lability of the rhamnosyl (and abequosyl) linkage, mild acid hydrolysis of this lipopolysaccharide yields the “chemical” repeating unit (2) with a different sequence (Luderitz et ul., 1966b). Chemical analysis of lipopolysaccharides in general will reveal chemical repeating units, and only by chance are they identical with the biological ones. A be

1

Man

Glc

1

Gal

Glc +

1

Rha (1)

.--f

Gal

(Abe) +

1

Man (2 )

+

Rha

85


The S. typhimuriurn 0-antigen factors 4.5, and 12 of the Kauffmann-White Salmonella classification scheme are determined primarily by the immunodominant sugars D-abequose , 2- 0-acetyl-D-abequose and D-glUCOSe, respectively. The complete antibody-combining sites (0 factors) comprise the immunodominant sugars together with the adjacent 3 to 4 sugar units in their distinct conformations (Luderitz ef al., I97 1 ; Nghiem and Staub, 1975). In recent years, an increasing number of lipopolysaccharides from various bacterial families have been studied with respect to their 0 chains, and a great diversity of constituents, linkage types, and structural peculiarities has become apparent. During this time, many new sugar classes and sugar derivatives have been discovered, together with sugar alcohols, acidic compounds, and even amino acids (Wilkinson, 1977; Jann and Westphal, 1975; Jann and Jann, 1977; Weckesser el al., 1979). In the past, more than a decade was required for the complete analysis of one 0 chain (e.g., S. rjphirnuriurn). Thanks to the advances made in carbohydrate chemistry and particularly in structural analysis, these studies are now accomplished much more efficiently. B. Lindberg in Stockholm and N. Kochetkov in Moscow and their schools have elaborated new methods of polysaccharide analyses, especially “methylation analysis’’ and nuclear magnetic resonance spectrometry (Lindberg e f a l . , 1975; Lindberg, 1979; Kochetkov and Chizhov, 1966; Jennings and Smith, 1980). There are few laboratories in the world performing structural studies on lipopolysaccharides that have not cooperated with one of these laboratories, and an increasing number of lipopolysaccharide structures will be available in the future. Table I lists about 70 gram-negative serotypes, which have had the detailed structure of their 0 chains determined mainly during the past 10 years (reviewed by Wilkinson, 1977; Orskov erul.. 1977; Jann and Jann, 1977, 1978). Similarly, a great number of bacterial capsular antigens have been investigated (Jann and Jann, 1977). ~

2 . HETEROGENEITY O F 0 CHAINS The true repeating units (Abe-Man-Rha-Gal of Fig. 2 ) in 0 chains are strictly constant. In contrast, modifying substitiients (e.g., acetyl groups), and sidechain sugars (e.g., glucose) added in later steps of biosynthesis (see Section 11,B,3), are often not present in equimolar ratios (Wright and Kanegasaki, 1971). Furthermore, their presence or absence is often subject to variation (e.g., antigenic conversion by phages). This incomplete substitution of 0 chains is one of the causes for microheterogeneity of lipopolysaccharides. Another important type of‘ heterogeneity of 0 chains has recently been revealed. Hurlbert, Makela, and Leive and their co-workers (Hurlbert and Hurlbert, 1977; Palva and Makela, 1980; Goldman and Leive, 1980; see also Jann ef al., 1975) have independently shown that radiolabeled lipopolysaccharide separates on SDS-polyacrylamide electrophoresis into a number of de-

o n 0 LUDERITZ ET AL.

86

TABLE I BACTERIAL SEROTYPES WHOSE0 - C H A I N STRUCTURE HASBEENEVALUATED 0 Chains Derived from

Salmonella S . purutvphi A S . typhimurium

S. ryphimuriitrn Col 11 S . bredenev

S. rnreritidis S . strusbourg S . zuerich S . muenstrr S . anntum

S. newingroti S . illinois S . simftenherg

S.friedrnau S . minnc~sotu S . goclrsherg S . mil ~ a u k e e

Escherichia E . coli 08 09 020 032 058 069 075 086 0100 0111 0124 0141

References

Hellerqvist et a / . (1971~) Hellerqvist i’t ul. (1968, 1969~); Liideritz et crl. (1966b) Hoffmann et ul. ( I980b) Bagdian et d.(1966); Hellerqvist et d. ( I969b) Hellerqvist r t d . (1970h. 1971a. 1 9 7 2 ~ ) Hellerqvist e t a / . (1970b, 1971a. 1972~) Tinelli and Staub (1960); Bagdian et ul. ( 1967) Hellerqvist et ( I / . (1969a, 1971d) Hellerqvist et al. (1970a. 1971d) Nghiem and Staub ( I 975) Hellerqvist et i d . ( I97 I e) Robbins and Uchida (1962); Hellerqvist OI d.(1971b) Robbins and Uchida (1962); Hellerqvist et c d . ( 197 1 b) Robbins and Uchida (1962) Staub and Girard (1965); Hellerqvist et u / . (19710 Simmons et N / . (1965a) Liideritz rt a / . ( I 966a) Simmons et ul. (1 96%) Liideritz et id. (1965)

Reske and Jann ( I 972) Prehm et a / . ( 1 976b) Vasiliev and Zakharova (1976) Jann et ul. (1971) Dmitriev et a / . ( 1 9 7 7 ~ ) Erbing et al. (1977a) Erbing et ul. (1978) Springer (1971) Jann et a / . (1970) Edstrom and Heath (1967) Dmitriev et al. (1976a) Jann i’t ul. (1966)

Shigella Sh. dysenteriae Type 1

Dmitriev et al. (1976b) (continued)

87

LIPOPOLYSACCHARIDESOF GRAM-NEGATIVE BACTERIA

TABLE I (Continued) 0 Chains Derived from

Type 4 Type 5 Type 6 Type 8 Type 9 Type 10 Sh.jlexneri All known serotypes and aubserotypes Variant X Variant Y

References Dmitriev rt a / . (1977a) Dmitriev rt ul. (1977d); Kochetkov et n l . (1977) Dmitriev et ul. ( 1 977e) Dmitriev et ul. (l977b) Dmitriev et ul. (1975h) Dmitriev et ul. (1978a) Dmitriev er (11. (1978h) Dmitriev et ul. (19770 Lindberg rt ul. ( 1973) Kenne et ul. (1977a,b, 1978a)

Sh. hoydii

Type 4 Sh. boydii Type 6 Sh. sonnei Phase I Phase I1 Sh. newcastle

Lvov et nl. (1980) Dmitriev rt ul. ( I 9753) Kenne et ul. (1980) Kontrohr and Kocsis (1978) Dmitriev et crl. (1979)

Citrobacter Cirrobacter sp. 396

Jann et ul. (1978)

Klebsiella All known 0 groups 1 - I2

Erhing et ul. (1977h)

Serratia S.murcescens 08 Bizio

Tarcsay et ul. (1973) Wang and Alaupovic (1973)

Proteus P . mirabilis (D52)

Gmeiner ( 1977)

Yersinia Y . pseudotuberculosis All 10 known sero groups and subgroups (I-1V) Type 111 (pathogenic strain) Y . enterocoliticu Ye 128

Samuelsson et 01. ( 1974) Hellerqvist et u l . (1972a,h) Gorshkova et a / . (1980) Hoffman rt ul. ( 1980a)

Vibrio V . cholerue (Inabe)

Redmond (1979); Kenne et ul. (1978b)

Pseudomonas Ps. ueruginosu group 7 Pr. maltophilia P s . cepuciu

Dmitriev et a / . (1980) Neal and Wilkinson (1979) Knirel et a / . (1980)

88

OTTO LUDERITZET AL.

fined (double) bands. From their investigations, these authors conclude that a lipopolysaccharide preparation consists of a family of molecules differing in the length of the 0 chain, that is, in the number of repeating units. This method reveals the presence of free core-lipid A (i.e., lipopolysaccharide devoid of repeating units) and a spectrum of species with up to 40 repeat units. In the lipopolysaccharide of a S. t ~ ~ ~ strain ~ ~(Palva i and ( ~Makela, ~ ~ 1980), ~ ~ for? instance, there exists an accumulation of lipopolysaccharide species with 20 to 35 repeat units (30%), but also a large proportion of lipopolysaccharide with unsubstituted core (60%); double bands are assumed to be due to differing phosphate substitutions in the core or in lipid A . The distribution of lipopolysaccharide species is dependent on the strain and on growth conditions. This method has disclosed new aspects of the microheterogeneity of 0 chains and certainly will be used for revealing new structure-function relationships, e.g., the role of 0-chain length in virulence. 3. PRINCIPLES O F 'THE BIOSYNTHESIS OF 0 CHAINS

0-chain biosynthesis is catalyzed by a multienzyme system that utilizes a membrane-bound polyisoprenoid compound as a carrier for the glycosyl residues. In Sufrnoneliu groups A , D, E (Robbins and Wright, 1971; Wright and Kanegasaki, 1971) and C,, C, (Shibaev e t u l . , 1979), the synthesis, according to the now classical pathway, begins with the reversible transfer of galactose I -phosphate from UDPgalactose to the phosphorylated carrier lipid. Sequential transfer of additional sugars to the lipid-linked intermediate leads to the formation of the 0-chain repeating unit linked through a pyrophosphate bridge to the carrier molecule. The repeating units are then polymerized and eventually modified by substituents. Polymerization occurs by constantly transferring the growing chain to a newly activated repeating unit (growth at the reducing end). This results in a long-chain polymeric intermediate that is still linked at the reducing end to the carrier lipid. In a final step the 0 chains are transferred from the lipid carrier to the independently synthesized core lipid A to form the completed molecule. Recently, a second pattern of 0-chain synthesis has been shown to occur in Sulrnonella groups C, and L (Makela and Stocker, 19811, and in E . coli 08 and 09 (Schmidt et ul., 1976; Kanegasaki and Jann, 1979; Jann et u l . , 1979). In this case the synthesis appears to proceed by sequential addition of the sugars to the nonreducing end of the growing chain, which again is linked to a carrier lipid. I t has further been demonstrated that in E . coli 08 and 09, the polymannose 0-chain synthesis starts in fact with a glucose unit at the reducing end, and in SuOno/iellu group C, glyceraldehyde (Gmeiner, 1975) or Mannose (Heasley , 1981) has been identified at the reducing end of the 0 chain (Gmeiner, 1975), which is assumed to play an analogous starter role to the aforementioned


89

glucose unit. In these systems, in contrast to the “classical” system, a gene of the rfe locus is required for 0-chain synthesis (as well as for the synthesis of TI antigen and enterobacterial common antigen (ECA) (Schmidt et al., 1976; see Section II,B,4), whereas the gene of the rfc locus, which codes for 0-chain polymerase, is nor needed. The genetics of lipopolysaccharide synthesis have been reviewed recently (Makela and Stocker, 1981). Acquisition of an understanding of the mechanisms of lipopolysaccharide biosynthesis and the underlying genetics have made it feasible to synthesize congenic strains differing only in one precisely known lipopolysaccharide feature. Structure-function relationships of parent and altered strains and their 1ipopoIysaccharides can now be studied, for example the effect of repeat-unit structure on virulence or its role in protection.

4. CHANGES IN

AND

REPLACEMENT OF 0 C H A I N S

a . Conversions Due to Extrachromosomal Elements. Lipopolysaccharide modification can occur through genes not linked to the chromosome, but to phages or plasmids. In their classical work, Robbins and Uchida (1962) and Robbins et al. (1965) elucidated the mechanisms underlying antigenic conversion by phage. When bacteria of Salmonella group E, are lysogenized by the converting phage the bacteria acquire the serological specificity of group E,; that is, the lipopolysaccharide 0 chain is converted from a-linked to P-linked repeat units. Genes of the phage code for a protein to act as inhibitor of the old a-polymerase, and also for a new P-polymerase. Many conversion systems have been studied in Salmonellu by A. M. Staub and her group, but lysogenic conversion has also been observed in other Entcrobacteriaceae (reviewed by Luderitz rt al. , 1966b, 1971; Nghiem and Staub, 1975; Makela and Stocker, 1981). Plasmid-linked conversions have only recently been recognized. A most dramatic plasmid effect has been described by Hoffman et al. (1980b). S . typhimurium (as well as Salmonella group D and E species) infected with the plasmid Col Ib drd 2 (a mutant Col Ib derepressed in colicin production) was shown to synthesize a lipopolysaccharide different in its 0 chain from the wild type (see Fig. 2), containing a new repeat unit (3) Rha-Glc-Man-tGlcNAc

t Gal

(3)

Plasmid-controlled expression of Migella sorinei phase I antigen has recently been recognized (Kopeck0 rt ul.. 1980). Rhizobia have the ability to infect leguminous plants, and in some cases, lipopolysaccharide seems to play a decisive role in the infection. Nodulating bacteria of Rhizobium trifolii have recently been shown to contain two different “Nod” plasmids. Nod-negative mutants

90

OTTO LUOERITZ ET AL.

have lost the capability of inducing nodulation and their lipopolysaccharide exhibits a changed composition (Zurkowski and Lorkievicz, 1979; R . Russa and E. Rietschel, unpublished data). b. SIT Variation. Salmonellu T forms were identified by Kauffmann (1956); they had been isolated from clinical cases and were then recognized as mutants of the infecting Salmonella bacteria of groups B and D. T1 and T2 forms were identified. They are characterized by a lipopolysaccharide carrying, instead of 0 chain, polymers of ribose and galactose (partly in the furanose form) in the case of T1 (Berst et a / . , 1969), and substituted glucosamine in the case of T2 (Bruneteau et al., 1974). T I - and T2-determining genes (rft and rfn, respectively) can be transferred experimentally to other strains, which then express T specificity (Sarvas, 1967; Valtonen et al., 1976). c. Lipopolysaccharide-Bound Enterobacterial Common Antigen (ECA). The haptenic ECA is a glycolipid present in all members of Enterobacteriaceae and responsible for the common serological cross-reactions among them. Its structure has been determined as a polymer of the disaccharide N-acetylglucosamine-Nacetylmannosaminuronic acid, containing small amounts of fatty acids. In its immunogenic form, which rarely occurs naturally, ECA is linked to the core of lipopolysaccharide, replacing 0 chains. Only those core types not containing terminal N-acetylglucosamine were found to function as acceptors for ECA (i.e., the core types R 1 , R4, and to some extent K-12 of Fig. 3). Genetic manipulation allows the construction of strains lacking ECA, or expressing haptenic or immunogenic ECA (for reviews, see Makelii and Mayer, 1976; Mayer and Schmidt, 1979). d . Lipopolysaccharide-Bound Capsular Polysaccharide. The possibility that capsular polysaccharides may replace 0 chains in E . coli has been suggested and discussed (Galanos ef al., 1977a; qrskov e f al., 1977).

C. Structure and Biosynthesis of the Core 1 . STRUCTURE O F CORETYPES

Diversity in lipopolysaccharide structures is usually associated with the 0 chains; core structures are more uniform. The core of S. typhimuriurn (as shown in Fig. 2) is, for instance, common to all Sulmonella species and also occurs in other enterobacterial lipopolysaccharides. The Salmonella core contains a lipid A-distal hexose oligosaccharide consisting of D-glucosamine, D-glucose, and D-galactose, and an inner lipid A-proximal region consisting of an oligosaccharide of the core-specific sugars, L-glycero-D-rnanno-heptose (L-D-Hep) and 2-keto-3-deoxy-~-manno-octonate (KDO or dOclA), each forrning a branched trisaccharide (Liideritz et al., 1966a; Osborn, 1966; Hellerqvist and Lindberg, 1971). As indicated in Fig. 2, the exact linkages in the KDO


91

trisaccharide region have not been definitely established (Droge et ul., 1970; chemistry and biology of KDO have been reviewed by Unger, 1981). KDO I links the polysaccharide to lipid A in a relatively acid-labile (ketosidic) linkage (for designation of the core sugar units see legend of Fig. 4). The 0 chains are attached to the subterminal glucose I1 unit of the core. A characteristic feature of the core is its substitution by such polar groups as phosphate and ethanolamine. Furthermore, the carboxyl groups of KDO contribute to the acidic character of lipopolysaccharides. This negative change is believed to be physiologically important. Lipopolysaccharides may be interlinked to other components of the cell wall through divalent cations and polyamines in ionic linkages, thus providing integrity and stability to the cell wall. These cations and amines are also usually found in lipopolysaccharide preparations associated with the core and lipid A . As with the 0 chains and lipid A, the core region of lipopolysaccharides exhibits a high degree of heterogeneity. Substituents of the main chain of the inner core (Hep 111, P, EtN, and possibly N-acetylglucosamine and galactose IT) are not necessarily present in molar amounts. Furthermore, alkali-labile substituents on the core such as glycine have occasionally been detected (Gentner and Berg, 1971; see also Hellerqvist and Lindberg, 1971). Recent evidence indicates the presence of an acid-labile substituent linked to the branched galactose (Fumahara and Nikaido, 1980). The nature of these substituents is unknown. In recent years five other core types have been identified in Enterobacteriaceae, the core types coli R1LR4 and K-12. As seen in Fig. 3, all differ in the hexose region of the core but are similar in the Hep-KDO region (Janssen e r a / . , 1981). The overall structural similarities are obvious. As indicated in Table 11, these cores have been identified in serotypes of E . c d i , Shigellu, Arizoriu, and others. Table 111 lists strains whose core structures have been studied; some of them are R mutants with incomplete biosynthesis of the core. Compositional analyses of the core from various bacterial species have revealed the presence of such unusual constituents as u-glycero-D-nrannoheptose. uronic acids (Kotelko ei al., 1974), or amino acids (Drewry et ul., 1975). and, in the case of Vibrio choler~ie strains, fructose (Redrnond r f ul., 1973; Jann et ul., 1973; Raziuddin, 1980) or seduheptulose ( K . W. Broady, unpublished data) may be present. These probably replace KDO in linking the core to lipid A . Lipopolysaccharides containing a core devoid of heptose and/or KDO have also been identified (summarized by Galanos et ul., 1977b; Wilkinson, 1977). The analysis of core structures in S-form lipopolysaccharides is hampered by the fact that the core represents only a relatively small part of the whole molecule. Recognition of the presence of nonsubstituted core stubs in S-form lipopolysaccharides (see Section If,B,2) offers the possibility of studying core oligosaccharide devoid of 0 chains in cases where Ra mutants (see next paragraph) are not available. Treatment of S-form

92

OTTO LUDERITZ ET AL

Salmonella

Ra

Gal

GlcNAc

@-@-EtN

HIP

KDO-?**@-EtN

R1

R2 R3 E.coli

R4

014

Gal

Gal

t2la

141p

HIP

,J?la_-

Gal $b Glc -!$~Glc*Hep

K-12

G!cNAc Gic ' ' 2 ' G k

-p]

t - E t N $-EtN

E coli K12 161

t$p'$%KW-(1(00)

Gal

Hep & H i p

pl-3 %GIC+H~~&H~~-(KDO)~% 1.64

Rha . @-EtN .

i ;

L--6--J FIG 3 Structures of enterobacterial core types identifed thub far Ra Osbom (1966). Ludentz et ul. (1966a); Hellerqvist and Lindberg (1971). R1: Feige and Stirm (1976); Feige (1977); Jansson ef al. (1981); Katzenellenbogen and Romanowska (1980). R2: Hammerling etal. (1971). R3: Johnston el al. (1967); Jansson et a / . (1979). R4: Feige ef u l . (1977). E . coli K-12: Mayer et u / . (1976); Prehm e/ al. (1976a). For designation of sugars, see Fig. 2 . P-Hep, 1-a-D-Hep (according to IUPAC-IUB rules).

93


COKF

coli R I

Ra Salmonella Arizona

TABLE I1 TYPI:S~ D l ~ . N T l F l f IN ~ l ) ENT~~KOHACTERIACEAE" coli R4

coli R 3

coli R2

E . coli 01 1 I Citr(JbU&'r 9b. 10

E . coli 08 E . C Y J / ~09 E . c d i 08 E . c w f i 0100 E. cdi C Sh. sonnri phase II Sh. hoydii type 3 S h . jlcxneri type 6

coli K-12

E . coli 014 E . u d i K-12 Sh. dysenteriac type I

Sh. ,/lr.rneri type 4 b

" The classification is based on serological relationships. reactivity with Con A . comparison of phage patterns. and chemical analyses. For references see Galanos P I ul. (1977b).

Strains

Structure''

References

( P , E t N ) KDO Hep I I Glc - Glc - Hep - Hep KDO KDO

E . culi B E . coli K-12

-

Ter-Mutants CR 3 4 E . coli 0 I 1 I

X I

Glc- Hep--Hep-KDO X = H, Hep, or Glc - GlcN Gal - Glc

I

E . coli ATCC 12408

(Hep - Hep),

Prehm ef al. (1975) Ohkawa (1980) Blache ef a / . (1981)

-

Fuller rt d . ( I 973)

KDO

-

KDO

-

A0

Morton and Stewart ( 1972)

Hep- P

P rofrus rnirahilis

Glc - H!?p- KDO

Radziejewska-Lebrecht rt ul. (1980)

GlC - G!c

Pseudomonus arruginosci

Glc

-

Rha - Glc - - - GhlN - - -Hep - Hep - KDO - KDO I (P,EtN)

ma11 2

Drewry rt ctl ( 1 975)

Gtc NAc,

Borderella pertussis

Xunthomonas wwn.si.y

,Hep G~CUA" Man Man Glc - Man - KDO ~

I

P I

GalUA - Amide "

Details of linkages are omitted

Chaby et d.(1977)

W. A. Volk (unpublished data)

94

Om0 LUDERITZ ET AL.

lipopolysaccharide with mild acid will release a mixture of polysaccharides, which subsequently is separated on Sephadex G-50 into 0 chain-substituted and nonsubstituted core (Miiller-Seitz e t ul., 1968). The core is thus obtained in a pure form, though it is probable that degradation takes place under the acidic conditions needed to separate lipid A from polysaccharide i n lipopolysaccharides that do not contain KDO (Shaw, 1982). 2. LIPOPOLYSACCHAKIDES O F R MUTANTS

Evaluation of the core structure of Salmonella lipopolysaccharides was greatly facilitated when it was found that so-called R mutants are defective in lipopolysaccharide biosynthesis. Depending on the defect, these mutants synthesize complete or incomplete core structures, still linked to lipid A , but all devoid of the 0 chain. Figure 4 shows a series of R-form lipopolysaccharides derived from Salmonella mutants defective in different steps of the biosynthetic pathway. Similar series of R mutants have also been isolated from E . coli, Shigrllu, Proteus, and others. Ra mutants are defective in 0-chain synthesis, and produce lipopolysaccharides with complete core. Rb through Re mutants are core defective and synthesize lipopolysaccharides with incomplete core. Re lipopolysaccharides (Re glycolipids) are the most defective and contain only KDO and lipid A . The different R mutants and their respective R lipopolysaccharides can be differentiated and identified by means of lectins (Ahamed et ul., 1980), and antibodies (Ruschmann and Niebuhr, 1972; Nixdorff and Schlecht, 1972), as the terminal sugar sequences, which are different, are recognized by these tools. Recently developed immunoadsorbants make the isolation of pure monospecific R antibodies possible (see Section III,A,2). As well as being recognized by chemical analysis of their lipopolysaccharides, the mutants can also be identified by their phage pattern (Wilkinson et ul., 1972). Since the linkage of KDO to lipid A is not affected by mutation, and since this linkage is acid labile, mild acid

FIG. 4. The structures of 0-chain-defective (chemotype Ra) and core-defective (chemotypes Rb-Re) lipopolysaccharides of Salmonella R mutants, as derived from chemical, biosynthetic, and genetic investigations of the groups of M. J . Osbom, H. Nikaido, B . A . D. Stocker, P. H. Makela, and the present authors. For designation of sugars. see Fig. 2. Hep. 1.-glycero-o-mannoheptosc; DHep. o-glyccro-o-mannoheptose. The Rb, stmcmre seems to occur in some Rb, lipopolysaccharides (Hellerqvist and Lindberg, 1971). R, represents an unusual chemotype and is not an intermediate of lipopolysaccharide biosynthesis. The Re, glycolipid has been detecied in a mutant of E . coli B . For simplicity, these products of R mutants are often called lipopolysaccharides, although they are rather lipooligosaccharides. being termed sometimes R glycolipids according to Liideritz rt u / . ( 1 969). Designation of the sugar units: Glc N

Gal I1 H e p 111

I

I

KDO 111 I

Glc 11- Gal I- Glc I- H e p 11- H e p I- KDO 11- KDO I- Lipid A


95

96

OTTO LuDERlTZ ET AL.

hydrolysis will liberate lipid A from all these lipopolysaccharides. The Re lipopolysaccharide is of course the richest source of free lipid A . In recent investigations, Gmeiner and Schlecht (1979, 1980) have analyzed the cell walls of S . typhimurium S - and R-form mutants comparatively for peptidoglycan, lipopolysaccharide, phospholipids, proteins, and covalently linked lipoprotein. The results showed that with increasing defect the strains incorporated increasing amounts of R lipopolysaccharide into their cell walls. In Re mutants, the lipopolysaccharide content of the cell wall was about four times higher than that of the wild type. Also, the content of phospholipid increased, whereas that of protein decreased. Based on their findings, the authors discuss the occurrence of different molecular organizations of the outer membrane in the various strains, this being reflected in their different properties in biological systems (e.g., sensitivity to detergent and antibiotics; for a controversial viewpoint regarding the amounts of lipopolysaccharide, see Nikaido and Nakae, 1979). Occurrence of R forms is rare in nature. Although they can be cultivated under laboratory conditions, they appear to be less resistant to in vivo situations. They are generally nonvirulent, are phagocytized without opsonization by macrophages (Roantree, 1971) and amebas (Gerisch et al., 1967), and are sensitive to toxic agents. They have been found, however, in urinary tract infections (Ryan e t a / . , 1973; Westenfelder et a l . , 1977). 3. BIOSYNTHESIS O F 'I H E

CORE

It is evident that the lipopolysaccharide structures of Fig. 4 represent intermediates in the pathway of core-lipid A synthesis, which would proceed from Re to Ra by sequential addition of the constituents. That this mechanism is valid in r + \ v is concluded from experiments where many of the transfer reactions have already been accomplished in ivirro (Wright and Kanegasaki, 1971; Osborn and Rothfield, 1971). In some cases, the respective sugar transferases have been purified and, recently, also cloned (Creeger et al.. 1979). The activated forms of the sugars are known. In similar manner to 0-chain biosynthesis, a lipid carrier is involved. In this case, the carrier is lipid A , which remains linked to the core. Thc genetics of core biosynthesis have also been evaluated (for review, see Make18 and Stocker, 1981). 0-chain synthesis proceeds independently of the synthesis of the core and therefore core-defective mutants also make 0 chains. Incomplete core does not, however, act as receptor for 0 chains, and thus the synthesized 0 chains are not transferred and remain bound to the polyisoprenoid carrier. The 0 chains can then be isolated by phenol/water extraction, because the pyrophosphate linkage to the lipid carrier is cleaved by this treatment (Beckmann et d.,1964; Kent and Osborn, 1968). An R mutant has been identified whose defect results in the synthesis of an


97

R-form lipopolysaccharide not representing a natural precursor of the parent lipopolysaccharide (Rx of Fig. 4). This R-form lipopolysaccharide contains D-glycero-D-manno-heptose (D-D-Hep) linked to KDO-lipid A (Lehmann et a / ., 1973). According to a proposed sequence of reactions, sedoheptulose-7phosphate is an intermediate of heptose synthesis in the bacterial cell. It is isomerized in the bacterial cell to ~-~-Hep-7-phosphate and subsequently converted to the activated nucleotide-diphospho derivative, NDP-D-D-Hep, which is epimerized in position 6 to form the activated form of L-D-Hep. L-D-Hep is then transferred to KDO. The diphosphonucleotide has been identified as ADP (Kontrohr and Kocsis, 1981). The above mutant is presumably defective in the synthesis of the enzyme, NDP-D-D-heptose 6-epimerase. The reaction sequence stops at the level of NDP-D-D-Hep. Hep I transferase may recognize the D-I?IUII/ZO configuration of this “unnatural” heptose. D-D-Hep is, under these conditions, transferred to KDO. D-D-Hep, however, does not act as acceptor for glucose I transfer. The original mutant was leaky, and also synthesized in addition the correct lipopolysaccharide. Recently, a nonleaky E . coli K-12 mutant of this type has been isolated (Coleman and Leive, 1979). It is interesting to note that strains of the photosynthetic bacterium Rhodopsrudornonus gelutinosu synthesizes naturally a lipopolysaccharide containing D-D-Hep, KDO, and lipid A as the only constituents (Drews e f a / . , 1978; Weckesser et ul., 1979). Furthermore, polysaccharides resembling 0 chains have been isolated from these organisms by phenol/water extraction. As the authors discuss, these lipopolysaccharides .d polysaccharides could represent ancient forms of lipopolysaccharide and capsules, respectively, or, alternatively, these strains are results of mutations (i.e., core-defective R forms) that could survive. Another example of the occurrence of R-form bacteria in nature are Yersinia pesfis organisms, whose lipopolysaccharide contains glucose, L-glycero- and D-glycero-D-manrw-heptose, KDO, and lipid A (Hartley rt ul., 1974).

D. Structure and Biosynthesis of Lipid A Historically, structural investigations of Sulmonella lipopolysaccharides started with studies on the 0 chains, which then led to identification and investigation of the core, and only recently have detailed structural studies on lipid A been performed, even though the presence of a covalently linked lipid component in lipopolysaccharides had long been recognized (Boivin et ul., 1933; Morgan and Partridge, 1941; Tal and Goebel, 1950; Ikawa et ul., 1953; Westphal and Liideritz, 1954). This delay was caused mainly by technical difficulties existing at that time in the field of lipid analysis and in particular by difficulties regarding this lipid species. The resumption of structural work on lipid A in the last decade was greatly stimulated by the recognition of lipid A as the endotoxic principle in the lipopolysaccharide molecule.

98


Due to the acid lability of the KDO linkage, lipid A can be liberated from lipopolysaccharides containing KDO by mild acid treatment. Free lipid A thus obtained is a water-insoluble precipitate, soluble in organic solvents such as chloroform. Similar conditions of cleaving lipid A are applicable to lipopolysaccharides in which ketoses (fructose and sedoheptulose in V . cholerae) mediate the linkage to lipid A (Redmond et al., 1973; K . W. Broady, personal cornmunication). Great difficulties, however, are encountered with lipopolysaccharides containing a relatively acid-stable linkage to lipid A, and in these cases either degraded free lipid A or lipid A still linked by polysaccharide fragments is obtained. 1. STRUCTURE O F Salmonella LIPIDA

Figure 5 shows the structure of Salmonella lipid A with the actual positions of the ester-bound acyl residues left open. Lipid A represents an unusual phospholipid in that it contains a central disaccharide of two u-glucosamine residues linked /ill .6. Both glucosamine residues are substituted by a phosphate group: one is bound in ester linkage to C-4 of glucosamine I1 (the nonreducing glucosamine); the other one is bound to C-1 of glucosamine I (the reducing glucosamine), thus occupying the glycosidic hydroxyl group and rendering the

- ...-.

.

.

FIG.5. Structure of lipid A from Salmondla. Dotted linkages indicate incomplete substitutions


99

molecule nonreducing. In the lipopolysaccharide, the terminal KDO unit of the core is attached to C-3 of glucosamine I1 (Gmeiner et al., 1969, 1971; Rietschel et a l . , 1977). The P-GlcN-GlcN-P region of lipid A is termed the lipid A backbone. It represents a phosphorylated polyol and therefore resembles the hydrophilic region found in normal phospholipids (e.g., glycerophosphate). In common with other phospholipids, the backbone of lipid A is also substituted by acyl residues. Salmonella lipid A contains the following 7 moles of long-chain fatty acids: 4 moles of D-3-hydroxytetradecanoic acid (3-OH-14:0, P-hydroxymyristic acid) and about 1 mole each of dodecanoic (12:0, lauric acid), tetradecanoic (14:0, myristic acid), and hexadecanoic acid ( I 6 9 , palmitic acid). Recently additional small amounts of L-2-hydrpxytetradecanoic acid (2-OH-14:0, a-hydroxymyristic acid) and 3-hydroxydodecanoic acid (3-OH- 12:0, P-hydroxylauric acid) were identified. Careful analysis has shown that in many preparations tested the sum of 14:O and 2-OH-14:0 equals 1 mole equivalent (Rietschel et ul., 1972; Bryn and Rietschel, 1978). T w o moles of 3-OH-14:0 are linked to the amino groups of the disaccharide. The remaining 5 moles of fatty acids are ester linked. The exact positions of the ester-bound fatty acids in lipid A can be identified only partly by the methods presently available. There is strong evidence that only the two ester-bound P-hydroxymyristoyl residues are linked directly to the lipid A backbone, where 3 hydroxyl groups are available; one hydroxyl group, therefore, would remain free. It has also been shown that 14:O plus 2-OH-14:O either substitute one of the two ester-bound 3-OH- l4:O residues, or, alternatively, are distributed on both (Rietschel et u l . , 1972). Finally, there is preliminary evidence (Rietschel ct a l . , 1981; Wollenweber d a / . , 1981) that 12:O and 16:O are bound to the 3-hydroxyl groups of the two amide-linked 3-OH-14:O fatty acids in a still-unknown distribution (not shown in Fig. 5 ) . If these findings prove to be correct, all of the 3-hydroxylated fatty acids would substitute the backbone glucosamines directly, two of them in amide, and two in ester linkage as would have been predicted from the lipid A precursor structure (see Section ll,D,6, Fig. 10). The nochydroxylated fatty acids (and 2-OH-14:O) would then be linked to the hydroxy groups of the hydroxy acyl residues. Similarly to 0 chains and core, lipid A also exhibits structural heterogeneity (Nowotny, 1971; Chang and Nowotny, 1975; Banerji and Alving, 1979), one reason being a partial substitution of the phosphate groups by amino compounds, the so-called polar head groups (Muhlradt et ul., 1977). In approximately 50% of the lipid A molecules of Sultnonella, the phosphate group linked to glucosamine I is substituted by a phosphorylethanolamine residue with free amino group. In about 30 to 60% of the lipid A molecules, the phosphate bound to glucosamine I1 is substituted by 4-amino-4-deoxy-L-arabinose (4-AraN), the linkage being through C - l . The amino group and the hydroxyl groups of 4-AraN appear to be

100


nonsubstituted. This rare sugar had been previously identified as a component of Sultnonella lipopolysaccharides (Volk et al., 1970; Volk and Luderitz, 1981), but it was shown only recently that it is a constituent of the lipid A component of the molecule (Miihlradt et al., 1977; Rietschel et al., 1977; Hase and Rietschel, 1977). The formula in Fig. 5 contains some details that are still under investigation. Thus the pyranose form of glucosamine I is not proved, neither is its anomeric configuration known. The same applies for 4-AraN. Earlier determinations of fatty acids in lipid A had indicated the presence of three equivalents of 3-OH-14:O. New methods of analysis, however, have led to the identification of an additional mole (for a detailed discussion, see Rietschel et a l . , 1981). It was previously anticipated that lipid A units were interlinked by pyrophosphate bridges (Luderitz et al., 1973; Rosner et ul., 1976); this has recently been shown by "P-NMR spectrometry to be incorrect (Muhlradt et al., 1977; Rosner et al., 1979a,b). This may also be concluded from the fact that the phosphate groups of lipid A are in many cases quantitatively substituted by polar head groups. Most lipid A analyses have been performed with a lipopolysaccharide from a Salmonella tninnesota Re mutant, but sufficient data have been accumulated to conclude that all serotypes of this genus contain an identical lipid A . Apart from the uncertainties already mentioned, the lipid A formula seems to be correct, and in several laboratories lipid A's of other Enterobacteriaceae have been studied with analogous results (see Section Il,D,4). In most cases, however, the applied analytical methods were the same as those originally used for Salmorrellu lipid A. Recently, however, the structure of E . cofi K-12 lipid A has been investigated with quite different methods of analysis, including chemical and enzymatic degradations as well as :j'P-NMR spectroscopy. The results of this work of Khorana and his co-workers (Rosner et al., 1979a,b,c) agree excellently with those on Salmonella lipid A-apart from details specific to E . coli. This is satisfying. It is suggested that when new lipid A's are studied, the Salmonella lipid A should always be analyzed in parallel, in order that new findings with other lipid A's are recognized as specific for the strain or lead to a modification of the original model of lipid A structure. I N T H E STRUCTURE O F LIPIDA 2. C O N D I T I O NVARIATION AL

Under normal physiological conditions of growth, the composition and structure of lipid A seems to be relatively constant, but effects of extreme conditions in bacterial growth, for instance, have not been systematically studied. Changes hitherto observed concerned the lipid A head groups. When comparing lipid A isolated from different bacterial batches of the same strain, Muhlradt et al. (1977) could detect variations in the degree of substitution by head groups. These authors conclude that gram-negative bacteria-as it is known for grampositives-are able to modify their surface charge by addition or omission of


101

substituents in order to adapt to the ionic environment. An analogous mechanism of adaptation may apply to polar core substituents. Characteristic changes in the pattern of ester-bound fatty acids have been observed when the bacteria were grown at low temperature. In Proteus rnirabilis, grown at 15"C, the hexadecanoic (16:O) acid content of lipid A was markedly decreased, whereas the unsaturated analogue, hexadecenoic acid (16:l ) , appeared (Rottem et al., 1978). A change in the composition of ester-linked fatty acids was also observed in E . coli K-12. When grown at 1 2 T , the normal dodecanoic acid was decreased and replaced by the unsaturated palrnitoleic acid (van Alphen et ul., 1979). Identical changes have recently been observed with lipopolysaccharides from Salmonefla S and R forms grown at 12°C (H. Wollenweber, S. Schlecht, and E. T. Rietschel, unpublished data). Unsaturated fatty acids are rarely found in lipid A's. The appearance of unsaturated fatty acids in lipid A probably influences its fluidity, as was demonstrated in the following system. Lipopolysaccharide of E . rofi K-12 represents a cofactor for protein d of E . coli K-12, which functions as the receptor for phage K3 (in the case of E . coli K-12 and phage T4 the lipopolysaccharide acts as the receptor, but it requires protein Ib to exhibit activity; Henning and Jann, 1979; Datta et al., 1977). When complexed with lipopolysaccharide this protein d inactivates the phage. It was shown that complexes with lipopolysaccharide from bacteria grown at 37°C inactivated the phage only above 20"C, whereas complexes with lipopolysaccharide from 12°C growth were active at above 10°C (van Alphen et al., 1979). The lack of activity below these temperatures is attributed to the influence of the respective lipopolysaccharide on fluidity. Analogous results were obtained with corresponding whole cells. This system demonstrates in a most elegant way the influence of lipid environment on the function of an outer membrane protein. 3. CONFORMATION OF Sulmonella LIPIDA Empirical force-field calculations are now being applied to obtain information on the conformation of lipid A disaccharide. The approach includes calculations using glucopyranoses, p-maltose, /3-cellobiose, and P-gentiobiose as models (Melberg, 1979; Melberg and Rasmussen, 1979; Rasmussen, 1980). Using the methods of Ramachandran, computer calculations have been performed on the (KDO),-lipid A molecule (Formanek, 1978; Formanek and Weidner, 1980). The permitted torsion angles of the ketosidic linkages in the trisaccharide, and of the linkage to lipid A, as well as the angles of the phosphate residue bound to glucosamine 11, have been determined and proved to be highly restrictive. Only one terminal KDO residue would exhibit free rotation. Furthermore, X-ray diffraction studies indicated a very compact molecule with a lattice periodicity of about 4.1 A and a hexagonal order of the fatty acids, much like other phospholipids (Wawra et ul., 1979; Ueki et al., 1979; Rottem and Leive, 1977). On the basis of these data, an atomic model has been built (Fig. 6). It repre-

102



103

sents the Sulmonellu Re-form lipopolysaccharide containing KDO and lipid A (see Figs. 4 and 5 ) . The lipid A backbone is substituted by 4-aminoarabinose, phosphorylethanolaniine, and 4 moles of P-hydroxyniyristic acid, esterified by 1 mole each of lauric, myristic, and palmitic acid (see Section II,D,l). The fatty acids are oriented in parallel and exhibit together with the lipid A backbone a length of about 22 A; the KDO trisaccharide is about 5 A in length, in accordance with X-ray diffraction data (Formanek and Weidner, 1980; Veki pf ul., 1979). Recently two classes of binding sites for divalent cations have been identified in lipopolysaccharides (Schindler and Osborn, 1979). The first class, of low binding affinity, was attributed to phosphate groups of the KDO-lipid A region; the second class, of higher affinity, was related to the KDO trisaccharide representing a specific site of interaction with divalent cations. In accordance with the experimental findings, the orientation of the carboxyl groups of the KDO residues in the model allows the binding of Ca2+ and Mg2+. The model clearly shows the amphipathic nature of the molecule. In addition, with an accumulation of polar groups it also exhibits amphoteric properties. Its integration in the bacterial outer membrane can be visualized, as can its reactivity with artificial and natural membranes. 4. STRUCTURAI. FEATURES O F LIPIDA’s T H A N Salmonella

FROM

BACTERIA OTHER

Although detailed structures have been completely evaluated for only a few lipid A’s, many results are available regarding constituents and partial structures (reviewed by Galanos er ul., 1977b; Rietschel rt u l . , 1981), and they allow first conclusions to be drawn on the structural constancy or variability of lipid A’s from different bacterial families. In order to study lipid A structures in a comparative manner, standard procedures have been worked out that allow screening of different lipid A’s (Rietschel, 1982). Amide- and ester-bound fatty esters as well as ester-bound 3 - 0 acylhydroxy fatty acids, are sequentially released under specified differentiated hydrolysis or transesterification conditions and subsequently identified by gas-liquid chromatography (GLC) and mass spectrometry (MS). The optically isomeric form of hydroxy fatty acids is determined by GLC after their conversion into diastereomeric derivatives. For studying the lipid A backbone and head-group substituents, classical degradation pathways have been worked out, starting from the corresponding lipopolysaccharide, the final product being the central glucosamine disaccharide FIG. 6 . Atomic model of Srrlmor~rlla lipid A . This model was built according to the structure given in Fig. 5 and to the results on the conformation of lipid A (Forrnanek, 1978; Formanek and Weidner. 1980). All nonhydroxylated fatty acids have been attached in ester linkages to the Thydroxy groups of the P-hydroxymyristoyl residues (see text).

104


in the reduced form (Hase, 1982; Volk and Luderitz, 1982). The intermediates of degradation are purified by electrophoresis and analyzed for their respective constituents. In the following, results of these investigations are summarized. They have been reviewed in more detail by Galanos et ul. (1977a) and by Rietschel et al. (1981). The lipid A backbone (P-GlcN-GlcN-P, see Fig. 5 ) is a common structure. Besides being identified in all Enterobacteriaceae investigated thus far (six genera), it has also been found in two plant pathogens, in two anaerobes, and in two photosynthetic bacteria (Rietschel et al., 1977). Some exceptions will be discussed later (Section 11,DS). In most cases the backbone is substituted by at least one head group. There is no variation regarding the substituent of the phosphate on glucosamine 11: either 4-aminoarabinose (Salmonellu, P . mirahilis, Y . enterocoliticu, Chrotnobucterium violaceum, Rhodospirillum tenue), or no head group ( E . c d i , V . cholerae) is present. In contrast, variation does occur regarding the substituent of the phosphate on glucosamine I: phosphate ( E . coli), phosphorylethanolamine (Su/motzella. V . cholerue), D-glUCOSamine (Chr. violureum), and D-arabinofuranose (Rhsp. tenue) have been found, as well as no substituent ( P . rnirubilis, Y . enterocolitica). Partial as well as quantitative substitutions occur, this leading again to microheterogeneity in the lipid A molecule (and to lipid A species devoid of head groups as in the case of E . coli). The head groups are not acylated and the amino groups where they occur are free. It was found, without exception, that the amide-bound fatty acids are saturated, 3-hydroxylated, and of the D form (for a new specific chemical synthesis of D-3-hydroxy fatty acids, see Tai el ul., 1980). In some rare cases a mixture of two, three, or four types of ~ - 3 - 0 Hfatty acids have been found in amide linkage. The following N-acyl residues have been identified: 3-OH- 1 O:O, 3-OH- 12:0, 3-OH-14:0, 3-OH-16:0, 3-OH-18:0, 3-OH-I I-Me-12:0, 3-OH-15-Me-16:0, 3-OH-17:O. A unique pattern of amide-bound fatty acids is present in the lipid A components of Vihrio anguillarum (D. H . Shaw, unpublished data) and Rhodopseudomorias sphueroides ATCC I7023 (W. Strittmatter, unpublished data), which both contain 3-oxotetradecanoic acid (besides 3-OH-14:0). Since 3-hydroxy fatty acids are rarely found in other lipids of gram-negative bacteria and in biological fluids, they can serve as a specific marker for lipid A and endotoxin. A great variation regarding ester-linked fatty acids is seen in different lipid A's (for references, see Wilkinson, 1977; Galanos et ul., 1977a; Drews e.'a/., 1978; Weckesser et d., 1979). Nonhydroxylated, 2-, and 3-hydroxylated nonbranched, iso-, and unteiso fatty acids of chain lengths varying from 10 to 20 carbon atoms have been identified. 3-Hydroxy fatty acids are of the 11 form, and 2-hydroxy fatty acids are of the I. form. It appears as though nonsaturated fatty acids may also occur (Broady et al., 1981; Raziuddin, 1980a,b). Ester-bound 3-acyloxyacyl residues as present in Salmonellu lipid A (see Fig. 5 ) , are frequently but not


105

generally found; when they do occur, however, the substituted acyl ester is always a 3-hydroxy fatty acid, whereas the substituting fatty acid may be non-, 2-, or 3-hydroxylated. That hydroxy fatty acids are exclusively linked directly to the lipid A backbone (as in Salmonellu lipid A) is therefore not a general phenomenon. Whether nonhydroxylated fatty acids are linked exclusively to hydroxylated fatty acids (as proposed for Salmonella lipid A) is presently being investigated. Since the overall fatty acid spectrum of lipid A’s is, within certain limits, constant for members of a genus or a family, this may prove to be an additional tool in taxonomic classification of the bacteria (Nikaido, 1970; Jantzen et ul., 1975; 1976; Rietschel and Liideritz, 1980; Wilkinson and Caudwell, 1980).

5. PRESENTLY KNOWN LIPIDA STRUCTURES Aside from Salmonella and closely related enterobacterial lipid A’s, detailed structures are presently available for the lipid A’s of Chr. violuceum, Rhsp. tenue, Rhodopseudomonus viridis, and Rhps. palustris (Liideritz et al., 1978). Lipid A of Chr. violuceum (Fig. 7) contains a backbone identical to that of Sulmonellu (P-GN-PI ,6-GN-P), and this is substituted by ~-3-OH-12:0in amide linkage, and ~-3-OH-10:0,~-2-OH-12:0,12:0, and 16:O in ester linkage. The head groups, 4-amino-~-arabinoseand D-glucosamine, are present in molar ratios and are neither 0- nor N-substituted (Hase and Rietschel, 1977). This lipid A, therefore, represents an unusual acylated tetrasaccharide with four amino sugars linked together either glycosidically or by phosphate bridges to give a nonreducing molecule. The glucosamine I I-P I-glucosamine is reminiscent of a trehalose type of structure. Lipid A of Rhsp. tenue (Fig. 8) also contains a backbone identical to that of Salmonella, which is acylated by amide-linked ~-3-OH-10:0 and ester-linked ~-3-OH-10:0,14:0, and 16:O. The phosphate groups are completely substituted by 4-amino-~-arabinose and D-arabinose, respectively, the latter being present in the furanose form. A third substituent, D-glucosamine, is linked to glucosamine I at position 4 (Tharanathan et ul., 1978; Weckesser et

L-L-AraN 1 @--_D-GlcNp

-

FIG. 7.

a

D-GlcN -1@ l Q - G k N

Structure of lipid A from Chrornohacterium violuceum

O n 0 LUDERITZ ET AL.

106

FIG. 8. Structure

of lipid A from Riiodospirillitrrr t m w .

al., 1979). The structure of this lipid A resembles a pentasaccharide with glycosidic linkages present in an inner-branched glucosamine trisaccharide, and two phosphate bridges linking the remaining two monosaccharides in such a way that the molecule is nonreducing. Lipid A of Rhps. viridis (Fig. 9) has a structure very different from other lipid A's. I t is devoid of phosphate and glucosamine, and instead contains a 2,3-diamino-2,3-dideoxy-~-glucose@-aminoglucosamine derivative, glucosamine), which is substituted by ~-3-OH-14:0and by acetyl groups. A lipid A structurally indistinguishable from that of Rhps. viridis is present in Rhps. pulustris (and probably in Rhps. sulfiviridis). As will be shown (Section IIl,B,3), these types of unusual lipid A's are neither endotoxically active, nor do they serologically cross-react with Sulrnorzellu lipid A (Galanos et u l . , 1 9 7 7 ~ ; Weckesser et uf.,1979). 2,3-Diaminoglucose has also been found in the lipid A of Pseudominus dirninuta and Ps. vesicirluris (Wilkinson and Taylor, 1978). Lipid A components devoid of phosphate have been identified in Chrornatiurri

107


vinosum, Thiocupsu roseopersicinu, and Rhodomicrohiurn vunniclii (Weckesser et ul., 1979). 6. F I N A L STEPSI N

THE

BIOSYNTHESIS ot, Sulmonellci LIPIDA

As lipid A is the endotoxic principle of lipopolysaccharides, its study represents a key area of research. However, the only source of free lipid A is lipopolysaccharide, from which it has to be cleaved by acid hydrolysis with possible undetermined minor degradation. In order to circumvent this possibility, the search for mutants defective in the synthesis of the KDO was of obvious prime interest, as it was expected that such mutants would synthesize only the lipid A portion of the molecule. Conventional phage-selection techniques, which had been employed successfully for the isolation of mutants defective in more distal regions of the polymer, failed to yield such mutants, suggesting that KDO-lipid A might be essential for the maintenance of the structural or functional integrity of the cell. Accordingly two approaches have been devised to isolate conditional Sulmotiellu mutants with defects in this part of the molecule. One approach was based on the dependence of the desired mutant on an exogenous source of KDO, both for synthesis of complete lipopolysaccharide and for growth (Rick and Osborn, 1972; Osborn r t d.,1974). The mutant from S. fyphimurium obtained in this manner contained an altered KDO-8-P synthetase that catalyzes the reaction i)-arabinose-5-P

+

phoaphoenolpyruvate + KDO-X-P

+ P,

This enzyme had a K , value for arabinose-5-P 35 times higher than the wild-type enzyme, and because of this the normal pool of arabinose-5-P was not sufficient for KDO synthesis when the reaction occurred at 37°C. Exogenous D-arabinose-5-P was required to maintain an internal concentration of this substrate sufficient to support both KDO synthesis and growth. It was subsequently found that the mutation was also temperature sensitive and at 42°C the mutant phenotype was expressed even in the presence of arabinose-5-P. In a second approach, KDO-lipid A mutants were selected for temperature sensitivity, both in synthesis of lipopolysaccharide and in growth (Osborn et ul., 1974; Lehmann et u l . , 1977; Osborn, 1979). Most of these mutants proved to be defective in KDO metabolisni. At the permissive temperature they synthesized KDO and lipopolysaccharide and grew normally, but when such cultures were shifted to the nonpermissive temperature, KDO synthesis ceased. As a consequence growth continued in these mutants for one to two generations at a normal rate but then ceased. Analysis of KDO-deficient mutants after incubation at the nonpermissive temperature led to the identification of new products synthesized by the cells during the absence of KDO (Rick rt ul., 1977; Lehmann, 1977). These substances were subsequently demonstrated to be incomplete lipid A .

108

a

C

d


G-3-OH-U:O D-3-OH-1L:O

I

D-3W-lL:0 D3OH-lL:O 16:O 16:O

1D-3-OH-1LD D-3-OH-M:O

e

1

D-3QH-14 0

FIG. 10. Incomplete lipid A molecules isolated from Salmonella and E . coli mutants. Products (a). (b). and (d) accumulate in KDO-defective Sulmonella mutants, products (a) and (c) in phosphatidylglycerol-defective E . roli K-12 mutants. The underacylated lipopolysaccharide (e) is formed from product a by KDO-defective Salmonellu mutants following a shift from nonpermissive to permissive conditions in the presence of cerulenin, which inhibits de novo synthesis of fatty acids.


109

Structural analysis showed that the lipid A precursor I molecule (Fig. 10a) contained the diphosphorylated PI ,6-diglucosamine backbone, substituted by 3-OH-14:O residues, two in amide linkage and two (Lehmann, 1977) in ester linkage (only one was found by Rick et a / . , 1977), and additionally by small amounts of ester-linked 16:O. As was expected, the precursor I , due to the defective KDO synthesis, does not contain KDO, but it was also shown to lack the saturated fatty acids 12:0, 14:0, and most of 16:O present in complete lipid A . More recently it was found that similar incomplete lipid A molecules were accumulated in a temperature-sensitive E . coli K-12 mutant unable to synthesize phosphatidylglycerol at the nonpermissive temperature (Nishijima, 1980; Nishijima and Raetz, 1979; Nishijima et al., 1981). Formation of phosphatidylglycerol in this mutant was prevented by two mutations. One mutation in the pgsA gene reduced the activity of the phosphatidylglycerol synthetase to less than 5% of the wild-type level but still permitted normal growth of the mutant. A second mutation in the pgsB gene rendered the pgsA strain temperature sensitive both in synthesis of residual phosphatidylglycerol and in growth. The incomplete lipid A isolated from this E . coli mutant could be separated into two fractions. Fraction I contained a phosphorylated glucosamine backbone substituted by 4 moles of 3-OH-14-0 (Fig. lOa), and fraction II contained in addition 2 moles of esterified 16:O (Fig. 10c). It should be noted that palmitic acid is a minor component of E . coli K-12 derived lipid A (van Alphen er af., 1979; Boman and Monnez, 1975). Genetic studies have revealed that fraction I accumulation is primarily associated with the pgsB mutation, whereas fraction I1 builds up when both mutations, pgsA and pgsB, are expressed. These results suggest that the incorporation of 16:O into fraction 1, and phospholipid biosynthesis are somehow coordinated. Furthermore, since these lipid A fractions are similar to those isolated from KDO-deficient mutants and lack KDO residues, it is possible that incorporation of KDO requires an appropriate phospholipid composition in the membrane. The incomplete lipid A produced by Salmonella mutants conditionally defective in the synthesis of KDO was shown to represent an intermediate in the synthesis of the KDO-lipid A region of lipopolysaccharides, and it acts as an efficient acceptor of KDO residues from cytidine monophosphate (CMP)-KDO in virro (Munson et a l . , 1978). Using particulate cell envelope fractions or a partially purified detergent-soluble fraction as source of the enzyme, a single reaction product was obtained containing two KDO residues. Since lipopolysac-

During conversion of (a) to (e) there is a transient accumulation of (d). Complete lipid A contains in addition to 3-OH-14:0 the nonhydroxylated fatty acids 12:0, 14:0, and 16:O in the case of Salmonella, and 12:0, 14:0, and traces of 16:O in the case of E . coli K-12. References: (a) Rick et al., 1977; Lehmann, 1977; Nishijima, 1980. (b) Jxhmann e / a / . , 1978. (c) Nishijima, 1980. (d) Walenga and Osborn, 1980a. (e) Walenga and Osborn, 1980b.

110


charide has been shown to contain three KDO residues in a branched trisaccharide structure (Fig. 2), the enzymatic product may correspond to lipid A precursor with either the main-chain KDO disaccharide, the branch unit, or a mixture of both. It is not known why the in vztro system failed to add the third KDO residue, but it is possible that the third KDO unit is transferred only after incorporation of the nonhydroxylated fatty acids. It has been known for many years from the work of Heath et ul. (1966) that KDO is transferred from CMPKDO to a chemically 0-deacylated lipid A preparation from E . coli, a product resembling the lipid A precursor in structure. I n vivo, the incomplete lipid A molecules are rapidly converted to lipopolysaccharide when the cultures of the KDO-defective mutants are shifted from nonpermissive to permissive conditions. Other intermediate products that accumulate transiently during this conversion have been isolated and identified. One of these products, lipid A precursor 11 (Fig. lob), resembled precursor I , but carried, in similar manner to complete lipid A , the two polar head groups 4-aminoarabinose and phosphotylethanolamine (Lehmann et u l . , 1978). Substantial amounts of precursor I1 have also been found when the mutants were incubated at intermediate growth temperature (Lehmann and Rupprecht, 1977). Another intermediate product was a derivative of lipid A precursor 1 containing two residues of KDO (Fig. 10d). It was indistinguishable in composition and chromatogaphic properties from the product obtained by enzymatic addition of KDO to isolated lipid A precursor (Walenga and Osborn, 1980a). This intermediate was produced transiently in vivo after shift of the culture to permissive conditions. It could be completed to lipopolysaccharide when the culture was returned to the nonpermissive temperature, where its continued formation is interrupted (Walenga and Osborn, 1980a). From pulse-chase experiments the early steps of lipopolysaccharide biosynthesis can be visualized (Fig. 11). Precursor I consisting of the phosphorylated diglucosamine backbone substituted by four P-hydroxymyristic acids is the acceptor for polar groups of the lipid A region, including 4-aminoarabinose, phosphorylethanolamine, and the KDO residues (Osborn, 1979). Since the glycolipid of Re mutants comprises the KDO-lipid A portion of lipopolysaccharides and contains the full complement of 0-acyl chains (Liideritz er ul., 1973), it was anticipated that the transfer of lauric, myristic, and palmitic acid would be the next step after the addition of KDO. Furthermore, it was assumed that full acylation would be obligatory prior to subsequent extension of the core polysaccharide chain. Recent results, however, have shown (Walenga and Osborn, 1980b) that the latter assumption is incorrect by studying the effect of cerulenin on the conversion of the acyl-deficient lipid A precursor to lipopolysaccharide. Concentrations of cerulenin that caused greater than 95% inhibition in the de novo synthesis of fatty acids and lipopolysaccharide had no


111

FIG. 1 1 . Possible pathways (I. II. 111) of bio5yntheais of Solrrrorrrlki lipopolysaccharide (LPS). Products a , b, and d have been isolated from Salrtiotic~llamutants (Fig. 10). KD0,-lipid A has been isolated from an E . ( o l i B mutant (Prehm ('t ( I / . , 1975). It is possible that product d could be isolated because the parent strain of the KDO-defective mutant had an additional mutation affecting the transfer of the third KDO residue (Walenga and Osborn, 19ROa). Whether product d and KD0,-lipid A (pathway I) or the theoretical product KDO,,-a (pathway 11) are the natural inlerniediates cannot be decided. Pathways I and I1 lead to lipopolysaccharide molecules devoid of the polar head groups in lipid A; pathway 111 proceeds via product b and leads to lipopolysaccharide carrying these groups.

effect on the rate or extent of the conversion of the preformed lipid A precursor to lipopolysaccharide (devoid of nonhydroxylated fatty acids, Fig. I Oe). These results indicate that incorporation of the nonhydroxylated 0-acyl residues of lipid A is not necessary for the extension of the core and the 0 chains, and that under certain conditions lipopolysaccharide lacking these fatty acids can be synthesized (Fig. IOe). These fatty acids may, however, have an essential function for the survival of the cells. If this is true, the Re lipopolysaccharide represents the minimal structure required to maintain bacterial viability. The pathway responsible for the biosynthesis of the v-3-hydroxy fatty acids of lipid A remains an unsolved problem. Humphreys et al. (1972) have shown that after incubation of particulate fractions of Pseudomonas aeruginosa with radioactive decanoic or dodecanoic acyl-CoA, labeled 3-OH-1O:O o r 3-OH- I2:O, respectively, were found in the lipopolysaccharide. Whether hydroxylation of the fatty acids occurs prior to or after transfer, remains open to discussion. If the hydroxy fatty acids are formed by the usual @-oxidation pathway, the initial 3-hydroxy acids should have the L configuration, but since the lipid A-linked 3-hydroxy fatty acids are of the D form, the involvement of an epimerase such as that already identified (Overath rt a / . , 1967) is indicated. Kawahara ef al. (1979) have studied the biosynthesis of bacterial 2-hydroxy fatty acids and have demonstrated that in Pseudomonas ovulis, cultivated in the presence of IxO,, dodecanoic acid incorporates lXOto form the 2-OH-12:O present in the lipid A of the Iipopolysaccharide. 3-OH-12:0, which is also a constituent of this lipid A, was found to be unlabeled. Although oxygen incorporation into lipid A-bound fatty acids has not been proved, this finding could explain the identical position of 14:O and 2-OH-14:O on ester-bound 3-OH-14:O in Salinonella lipid A , and may also explain the phenomenon wherein different strains

112

O n 0 LoDERITZ ET AL.

containing different ratios of 14:O and 2-OH-14:0, always have the sum of both fatty acids equal to approximately 1 equivalent (Bryn and Rietschel, 1978). A N D INTEGRATION 01.LIPOPOLYSACCHARIDES 7. TRANSLOCATION I N 1 0 T H E O U T E R MEMBRANE

Major emphasis has recently been placed on lipopolysaccharide translocation and insertion into the outer membrane. The development of techniques for separation of this outer membrane from the underlying cytoplasmic membrane has offered a new approach to investigations of these processes (Osborn et ul., 1972a). Lipopolysaccharides pulse-labeled in vivn appeared initially in the cytoplasmic membrane, but were rapidly transferred to the outer membrane during a subsequent chase. This was true for both core sugars and 0 chains (Osborn e t a / . , 1972b). The enzymes of 0-chain biosynthesis were shown to be entirely in the cytoplasmic membrane, and the enzymes that synthesize the core region of lipopolysaccharides were probably in the same location, though they tended to redistribute during fractionation (Osborn et a / ., 1972a). From these experiments Osborn et u l . concluded that lipopolysaccharides are synthesized on the inner membrane and subsequently translocated to the other membrane. Furthermore, in contrast to translocation of phospholipids, which appears to be readily reversible (Jones and Osborn, 1977a,b), the overall process of lipopolysaccharide translocation is unidirectional-that is, lipopolysaccharide molecules that are integrated into the outer membrane are no longer accessible to the inner-membrane enzymes (Osborn et al., 1972a,b). It is not known whether this apparent unidirectionality is imposed by the mechanism of inter- or transmembrane translocation per se, or by subsequent lipopolysaccharide-protein or lipopolysaccharide-phospholipid interaction within the outer leaflet of the outer membrane. Lipopolysaccharide-protein interactions in the outer membrane have been clearly demonstrated (Yu and Mizushima, 1977; Yamada and Mizushima, 1980; Schindler and Rosenbusch, 1978; Schweizer et ul., 1978; Henning and Jann, 1979). The current understanding of translocation of lipopolysaccharides postulates a transfer from the inner to the outer membrane at sites of contact between the two, the so-called zones of adhesion initially described by Bayer ( I 975) and Muhlradt et a / . (1973, 1974). However, the detailed molecular mechanisms of translocation and integration are still only poorly understood (for models see Osborn, 1979). As the carbohydrate portion of the lipopolysaccharide must for biosynthetic reasons face inward i n t o the cytoplasm initially, but outward after translocation, the translocation process may require transmembrane movement (flip-flop) as well as intermembrane transfer. Mutants in lipopolysaccharide synthesis have facilitated studies directed to the question of the possible role of the lipopolysaccharide structure in biosynthetic


113

processes. Clearly, neither completion of the core nor of the 0 chains is a necessary prerequisite for translocation and integration in the outer membrane, since such lipopolysaccharide translocation occurs in a variety of mutants including Re mutants. The KDO-deficient lipid A precursors accumulated in the mutants are, however, only incorporated very slowly into the outer membrane (Osborn et ul., 1980), providing strong evidence that the structural features of the internal region of the lipopolysaccharide molecules are indeed important (Osborn, 1979). Slow incorporation of lipid A intermediates into the outer membrane does not give a definitive answer to the question of whether the primary defect in assembly lies at the level of translocation or arises from failure of subsequent interactions required for stable integration. The latter possibility appears, however, unlikely on the basis of experiments on liposome-mediated incorporation of exogenous lipid A precursor into the outer membrane of intact cells. No significant transfer to the inner membrane and no detectable conversion to lipopolysaccharide occurred under conditions where extensive intermembrane translocation of liposome-derived phospholipids took place (Jones and Osbo’rn, 1977b). In order to define the smallest translocation unit, lipopolysaccharide intermediates were tested for their translocation properties. Lipid A precursor substituted by two KDO residues accumulates in the inner membrane and is not translocated (Walenga and Osborn, 1980a). In contrast, the KDO-lipid A structure of Re mutants does not accumulate and is translocated at near-normal rates into the outer membrane (Osborn et ul., 1980). Re lipopolysaccharides contain the nonhydroxylated fatty acids lauric, myristic, and palmitic acid and a third KDO residue, which are missing in the structure of the KDO lipid A precursor, though lauric, myristic, and palmitic acid have been demonstrated to be unnecessary for normal rates of translocation (Walenga and Osborn, 1980b). Lipopolysaccharides formed in the presence of cerulenin lack these fatty acids and are nevertheless translocated into the outer membrane. The head groups of lipid A, such as 4-aminoarabinose or phosphorylethanolamine, which may be present in the intermediates, are found in variable and nonstoichiometric amounts in lipopolysaccharides (Lehmann and Rupprecht, 1977; Muhlradt et al., 1977), and are therefore also unlikely to be essential. Therefore, a hypothetical molecule containing KDO trisaccharide linked to lipid A precursor appears to be the smallest unit capable of being translocated at normal rate into the outer membrane. The KDO residues have been shown to provide a high-affinity binding site for divalent cations (Schindler and Osborn, 1979), but the exact relationship of this site to translocation or integration remains unknown. Lipopolysaccharides inserted into the outer membrane are known to diffuse laterally within this membrane, from the insertion loci, to cover the entire surface. This diffusion appears to be dependent on the structure of lipopolysaccharide. Thus the translational diffusion coefficient of lipopolysaccharides of the

114


chemotype Rc is similar to that of phospholipids ( D = 10 !' cm2 SS') as measured 1980). In by fluorescence photobleaching recovery techniques (Schindler et d., contrast, S-form lipopolysaccharides are more restricted (Miihlradt ef ul., 1974; Rottem and Leive, 1977; Leive, 1977), and appear to be organized at least partially in domains.

111.

SOME SELECTED ASPECTS ON THE BIOLOGY OF LIPOPOLYSACCHAIDES

There is a vast and still expanding literature on the biological effects of lipopolysaccharides, and a comprehensive review on this topic is outside of the scope of this article. The reader is referred to recent reviews (Galanos et a / . , 1977a; Berry, 1977; Rietschel ef d . , 1980b, 1981; Morrison and Ulevitch, 1978; Jirillo and Fumarola, 1979; Bradley, 1979; Agarwal, 1980). and monographs (edited by Kadis et ul., 1971; Kass and Wolff, 1973; Schlessinger, 1977, 1980). Only selected aspects of the topic will be discussed here.

A. Endotoxic and Immunogenic Properties of Lipid A PRINCIP1.E 1. LIPID A , T H E ENDOTOXIC

Ob

LIPOPOLYSACCHARIDES

Lipid A is the endotoxic principle of lipopolysaccharides. This was initially postulated in the early 1950s as a conclusion from the fact that lipid A has been found as the only component common to all lipopolysaccharides from the Enterobacteriaceae (Westphal and Liideritz, 1954). A further strong indication was suggested by the observation that R-mutant lipopolysaccharides, lacking most of the polysaccharide chain (Fig. 4), and even the Re lipopolysaccharide, containing only KDO and lipid A , were all equally active as endotoxins (Liideritz et ul., 1966a). However, only recently could direct and definitive proof be offered that lipid A, devoid of 0 chain and core, represents the endotoxic moiety of lipopolysaccharides (Galanos et al., 1971a, 1977a,b). Free lipid A , obtained by mild acid hydrolysis of lipopolysaccharide and rendered water soluble by electrodialysis and subsequent neutralization, was shown to exhibit all the endotoxic reactions displayed by the complete parent lipopolysaccharide with the exceptions to be noted. This has been confirmed in various laboratories for many biological test systems, and is now generally accepted. It has been found, however, that in contrast to lipopolysaccharide, free lipid A does not induce necrosis and regression of tumors in mice (Wasilauskas and Cameron, 197 1 ; Tanamoto et al., 1979; Westphal et al., 1979, 1981). Since the elucidation of the structure of lipid A , several teams have started a

115


TABLE 1V SYNTHETIC LIPIDA MODEISUBSTANCES WITH ENDOTOXIN ACTIVITIES Compound

N-Palmitoyl-o-glucosamine Myristic acid-bovine serum albumin complex

N-Palmitoyl- and N-myristuyl-pglucosamine O-Palmitoyl-dextranphosphate Maltose tetrapalmitate 3-hydroxymyristoyl hydroxamate N-Myristoyl- and N-o.1 -3hydroxymyristoyl-o-glucosamine 6-phosphate

Activity B-Cell mitogen Lethality for mice. induction of reciprocal endotoxin tolerance Adjuvant effect, protection against radiation Antitumor activity Antitumor activity Serological crossreactivity with lipid A Serological crossreactivity with lipid A

Reference Rosenstreich er ul. ( 1 974) Bradley ( I 976)

Behling et d.(1976)

Suzuki et ul. (1977) V. Nigam (personal communication) Lugowski and Romanowska (1974) Gorbach et ul. (1979)

chemical synthesis of this component. Some groups have taken the approach of synthesizing model compounds resembling lipid A partial structures or compounds being endowed with physicochemical properties characteristic of lipid A . These products exhibit endotoxin-like activities and they are listed in Table I V . 2 . IMMUNOGENICITY O F L I P I DA ANTI-LIPID A ANTISERA

AND

PROPER^ I E S O F

In contrast to lipopolysaccharide-bound lipid A , free lipid A complexed or linked to a suitable carrier is immunogenic. I n the original approach, bacterial cells carrying lipid A on the surface were used for immunization (Galanos et u l . , 1971b, 1977a,b). For the preparation of these immunogens, the gram-negative bacterial cell was treated with acetic acid in order to remove the polysaccharide portion of the lipopolysaccharide, thus unmasking lipid A and exposing it on the surface. These cells were then coated with an excess of external free lipid A . Immunization of experimental animals with such immunogens leads to the formation of specific high-titer antisera, and this procedure has been widely adopted. It must be emphasized, though, that these antisera also contain antibodies evoked by other antigens exposed after the acid treatment (e.g., lipoprotein). Therefore, other forms of immunogenic lipid A have been developed. Lipid A complexed with bovine serum albumin (Gorbach et ul., 1979) or coated on erythrocytes (Mattsby-Baltzer and Kaijser, 1979) has been used for obtaining

116


anti-lipid A antisera. Liposomes with actively incorporated free lipid A proved to represent a valuable immunogen (Schuster et ul., 1979; Banerji and Alving, 1979, 1981; C. Galanos, unpublished data); these antisera, however, also contained antibodies against phospholipids of the liposomes. Such antibodies were not induced by lipid A-free liposomes, but reacted with them (Schuster et ul., 1979). The possibility of conjugating antigenically active lipid A fragments to protein carriers will be discussed later. Anti-lipid A antibodies are detected and measured by the passive-hemolysis test, using sheep erythrocytes coated with lipid A or alkali-treated lipid A (Galanos et al., 197 I b). An enzyme-linked immunosorbent assay (ELISA) has also been published (Jay, 1978; Mattsby-Baltzer and Kaijser, 1979; Fink and Kozak, 1980; Fink and Galanos, 1981). Lipopolysaccharide does not evoke anti-lipid A antibodies, nor do these react with lipopolysaccharide. Due to the close structural relationships between lipid A’s of Enterobacteriaceae and other gram-negative bacteria, complete serological cross reactions are seen among free lipid A’s of Enterobacteriaceae and other families (Galanos et a l . , 1977b; Rietschel et al., 1981; for exceptions see Section III,B,3). The immunodominant structure of lipid A comprises the linkage region of glucosamine and amide-bound hydroxy fatty acid (Liideritz et al., 1973). This was concluded from the observations that free lipid A and 0-deacylated free lipid A exhibit equal serological activity, and furthermore that a fragment of lipid A (obtained from Re lipopolysaccharide by mild hydrazine treatment and subsequent hydrolysis), containing the glucosamine-disaccharide devoid of all but one amide-bound hydroxy fatty acid (linked probably to glucosamine I, see Fig. 12), is still antigenically highly active (Liideritz et al., 1973; Galanos et al., 1977a). This agrees with the observation that Salmonella free lipid A and precursor lipid A show complete cross-reaction (see later), and that some synthetic compounds exhibit lipid A specificity (Table IV). Thus Lugowski and Romanowska (1974) have shown that 3-hydroxymyristoyl hydroxamate acts as an inhibitor of a lipid A/anti-lipid A system. In analogy, Gorbach et u l . (1979) tested the following compounds for their inhibitory activity: (1) N-acetylglucosamine; (2) N-acetylglucosamine 6-phosphate; (3) N-myristoyl-, and (4) N-o,L-P-hydroxymyristoylglucosamine 6-phosphate. Whereas compounds ( I ) and (2) proved to be inactive, an almost equal inhibitory activity could be demonstrated with compounds (3) and (4). Unfortunately, the two optical antipodes present in (4)-that is, the natural (D) and the unnatural (L) hydroxy acid-have not been tested separately; therefore, it is, not possible to define the contribution of the stereochemistry of the hydroxyl group of the amide-bound fatty acid to antigenic specificity. Further, these results do not permit a conclusion regarding the role of the anomeric configuration of glucosamine in specificity. In all these cases, the synthetic lipid A analogues

117


exhibited an activity approximately 100 to 1000 times less than expected for free lipid A. The aforementioned lipid A product of mild hydrazine treatment contains a free amino group on one of the two glucosamine units. This amino group has been used as a functional group to couple, with the aid of glutaraldehyde, this hapten to suitable carriers (Fig. 12). With edestin o r other proteins, very efficient, nontoxic immunogens are obtained that on immunization lead to highly specific anti-lipid A antibodies. Such vaccines may be valuable for clinical use. Binding of the fragment to a solid carrier has afforded an immunoadsorbant, which made possible the isolation of anti-lipid A antibodies (see also Lugowski and Romanowska, 1974). The efficiency of this procedure has been demonstrated and pure lipid A antibodies can now be produced for use in biological investigations (C. Galanos and D. Nerkar, unpublished data). A similar strategy has been applied to S- and R-form lipopolysaccharides for obtaining nontoxic 0 and R haptens. In this case, complete hydrazinolysis was performed and water-soluble preparations were obtained containing the respective 0 and/or R region linked to the lipid A backbone with free amino groups on the glucosamine residues available for coupling. Artificial immunogens and immunoadsorbants have been prepared from, for instance, hydrazine-treated lipopolysaccharides of different R chemotypes (Nixdorff et ul., 1975; C. Galanos, unpublished data). In this way pure monospecific R antibodies have been obtained after adsorption on different columns. Many animal species respond readily to primary or booster injections of acidtreated and lipid A-coated bacterial cells with the production of anti-lipid A antibodies (rabbits, goats, dogs, rats, vervets, baboons) (Galanos ef ul., 1971b). These animal species and humans often contain natural antibodies with lipid A

LPS

Hydrazine 1 h, 60'

acet. acid

A PLGlcN-pl.6-GlcN-!-P

I

0.5 h, 100'

Q

Glutaraldehyde Carrier

0

CH

C.0

p 2 $H2

p 2 H$-OH

FH2

HC=N-Corricr

[$"Z)lO CH3

FIG.12. Preparation of lipid A-specific immunogen (camer. protein) and immunoadsorbant (carrier, AH-Sepharose).

118


specificity in addition to a large variety of anti-0 specificities. In mice, however, naturally occurring antibodies with lipid A specificity have never been detected, although their sera also contain various anti-0 specificities. Moreover, mice fail to show a primary anti-lipid A antibody response, even after booster injections. Under specified conditions, however, mice will respond, and high anti-lipid A titers have been obtained using two injections of lipid A-coated bacteria with a time interval between the injections of longer than 3 weeks (Freudenberg, 1975; Galanos et al., 1971b). The presence of anti-lipid A antibodies in normal animals and humans and their increase in patients with gram-negative infections indicate that under certain natural conditions, lipid A becomes sufficiently exposed to express its immunogenicity . Biological activities of anti-lipid A antisera and their possible clinical role and application have been discussed recently (Rietschel and Galanos, 1977; Galanos et a l . , 1977a; Blake ct d.,1980; Westenfelder et al., 1977).

B. Physicochemical and Structural Prerequisites for Biological Activities of Lipopolysaccharide 1 . PHYSICAL S T A T E A N D BIOLOGICAL ACTIVITY OF LIPO PO L Y sACC H A R I DE s

Isolated lipopolysaccharide in solution forms aggregates (micelles) due to nonpolar interactions of the lipid A component. Recently, an additional cause of aggregation in which divalent cations and polyamines form intermolecular ionic linkages with the acidic groups of the molecule (phosphate, carboxyl groups of KDO) has been identified. The conversion of lipopolysaccharides into uniform salt forms by electrodialysis (see Section II,A) offered the possibility of investigating the influence of different cations on the degree of aggregation and, furthermore, of studying the effect of aggregation on the activity of lipopolysaccharide in various in vitro and in vivo biological tests (Galanos, 1975; Galanos et a / . , 1979a). The results of these investigations are summarized in Fig. 13. S. ubortus eyui lipopolysaccharide was converted into different salt forms and various parameters were rested. The state of aggregation (represented by sedimentation coefficient) and the solubility of the preparations were characteristic of the respective cation. Further, individual biological activities changed with increasing aggregation. Thus lethal toxicity for mice and pyrogenicity in rabbits decrease with increasing sedimentation coefficient, whereas lethality for rats, clearance from the blood, reactivity toward complement, and affinity to mammalian cells show a reverse dependence. Mitogenicity and activity in the Limulus lysate test, on the other hand, are not influenced by the state of aggregation. It is obvious that the distinctive physical state of a lipopolysaccharide influ-

S abortus --

LPS

Triet hylomine

(TEN1 Pyridine Ethanolomine No

K Putrescine Ca

Roteddcarcm Interaction with C ' from the Mood invivo and in vitro

jedirnentoticm Coefficient

1I

irnulus lysate gelation

I

230

partly insol

FIG. 13. Physicocheniical properties and biological activities of the lipopolysaccharide of Soltnorirllr ubortus c y r r i in different salt forms. Influence )f the nature of the cation neutralizing the acidic groups of the lipopolysaccharide (after Galanos et d., 197Ya). Arrows indicate direction of increasing stivity; (a) TEN form is completely inactive: (b) the different salt forms are of equal activity.

120

OTTO LUOERITZ ET AL.

TABLE V BIOLOGICAL ACTIVITIES OF COMPLtTt ~u/rnonc//u LIPID A. LIPIDA PRECURSOR 1 A N D CHFMtCAl.-DEGRADATION PRODUCTS" 0-

Type of activity

Free lipid A

Cross-reaction with anti-Su/rnonc//a lipid A anti serum Complement reactivity Mitogenicity Pyrogenicity Lethal toxicity Limulus activity

+ + + + + +

"

Lipid A precursor

Deacylated free lipid A

+

+

5

-

+ '' + +

0 , NDeacylated free lipid A

+

After Luderitz cf a / . 1978; R.G.McKenzie and C. Galanos, unpublished data.

ences the endotoxic activity. It is not surprising, therefore, that the same lipopolysaccharide, obtained by different methods, may differ in activity according to the cations present in the preparation, and certainly, for specific investigations it is important to know what kind of preparation is used with regard to the cations and amines present. The standardized S. aborrus equi lipopolysaccharide mentioned in Section II,A represents the sodium salt form.

2. BIOLOGICAL ACTIVITIES O F LIPIDA FRAGMENTS Complete free Salmonella lipid A, precursor I , and two chemically degraded Salmonella lipid A preparations were investigated comparatively for biological activity (Luderitz er al., 1978). The precursor lacks the head groups and the nonhydroxylated fatty acids (see Fig. IOa), whereas the 0-deacylated preparation obtained by alkali treatment contains only the amide-linked P-hydroxymyristoyl residues (and some phosphoryl and pyrophosphorylethanolamine units). The 0,N-deacylated product prepared by hydrazinolysis represents the glucosamine disaccharide partially substituted by phosphate groups. These preparations were tested in the following systems: cross-reactivity with anti-Salmonella lipid A antiserum, complement and Lirnulus reactivity, mitogenicity to B lymphocytes, pyrogenicity (rabbits), and lethality (adrenalectomized mice). As seen from Table V , the completely deacylated product is inactive in all tests, whereas the 0-deacylated lipid A and the precursor, both containing amide-bound fatty acids, cross-react with the lipid A antiserum, are mitogenic, and react in the Limulus gelation test. The precursor, in addition, exhibits lethal


121

toxicity, but is a weak pyrogen and shows decreased complement reactivity. A clear separation of the different endotoxic activities is thus apparent. 3 . BIOLOGICAL A C T I V I T IOEFSL I P O P O I . Y S A ~ C H A ARNI DI ) LIPID ES FAMILIES A’S FROM DIFFERENT BACTERIAL

In the past it was generally accepted that lipopolysaccharides, independent of their source, were equally active endotoxins, and this is still valid for lipopolysaccharides of Enterobacteriaceae and related families. This is of course due to the great structural similarities of their lipid A components. A few exceptions were known, however; for example, lipopolysaccharides from Brucellu had been found to exhibit very low toxicity in experimental animals (see Kreutzer et a l . , 1979). In recent years, lipid A’s derived from lipopolysaccharides of bacterial families remote from Enterobacteriaceae have been evaluated, and their structures have been discussed in Section 1I,D,5. In view of their structural differences when compared with Salmonella lipid A , it was not surprising lo find that differences also existed in the biological activities of these lipid A’s and of the corresponding lipopolysaccharides. Table VI shows some of the results obtained in various test systems with lipopolysaccharide and free lipid A from Chr. violaceurn, Rhsp. tenue, and Rhps. viridis and Rhps. pulustris, in comparison ; et a l . , 1978). with those from Sulrnonellu (Galanos et u l . , 1 9 7 7 ~ Luderitz Neither the lipopolysaccharide nor the free lipid A from Rhps. viridis and Rhps. palusrris exhibit any cross reaction with anti-Salmonellu lipid A antiserum. They are nontoxic and nonpyrogenic but react strongly with complement. In the Limulus gelation test, Rhps. pulustris is highly active. In this case pyrogenicity and Limulus activity do not run parallel. Chr. violaceurn and Rhsp. tenue free lipid A cross-react in the Salmonella lipid A system, whereas the corresponding lipopolysaccharides d o not. These preparations are inactive toward complement. The Chr. violaceurn lipopolysaccharide and free lipid A exhibit lethal toxicity and pyrogenicity, the free lipid A from Rhsp. tenue is toxic, but only a weak pyrogen; the lipopolysaccharide is also weakly pyrogenic, but of low lethality to mice. With our present knowledge, it is not possible to explain the different biological activities of the preparations on the basis of particular structural features. Differences in the serological behavior of free lipid A and the parent lipopolysaccharide may indicate a masking of the determinants of the Iipopolysaccharidebound lipid A by the 0 chains or by the polar head groups. These are usually removed (the head groups only partially) during the acid-catalyzed liberation of these lipid A ’ s . A similar reasoning may explain why even though the Rhsp. tenue free lipid A is toxic, the lipopolysaccharide is not. Chemical and biological

TABLE V1 B I O L O C I CACTILITIES ~L O F LIPOFOLYSACCHARIDES A N D CORRESFONDINC FREELIPIDA'S

Sulmonella Type of activity ~

LPS ~

Immunological cross-reactivity with anti-Solmonellu lipid A antiserum Reactivity with complement Pyrogenicity Lethal toxicity Limulus activity

~

Lipid A

Chromobacrerium violareurn LPS

Lipid A

-

-

FROM

DIFFERENT BACTERIAL GROUPS"

Rhodospirillum tenue LPS

Lipid A

-

+

Rhodopseudomonas viridus LPS

Rhodopseudomonas palustris

Lipid A

LPS

ND

~

+

+

+

+ +

+

+ + + +

+ + +

+

+ +

NDb

I

-

+

-

-

-

+

+

2

-

ND

-

+

ND

+

-

+

" After Liideritz et (11.. 1978; Galanos et al. 1977c; R. G . McKenzie and C. Galanos. unpublished data. Cross-reactivity of LPS with anti-lipid A antiserum was determined in vivo according to Rietschel and Galanos (1977). " N D . not determined.

123


studies must be conducted on further unusual lipid A systems before it will be possible to identify substructures in lipid A.

C. The Fate of Lipopolysaccharides (Lipid A) in Experimental Animals I . T H EROLEOF HIGH-DENSITY ~ I P O I ’ R O T E I N(HDL) LIPO PO L Y SAC c H A R I DE T R A N SPORI

IN

After intravenous administration to experimental animals, bacterial endotoxins persist in the blood for some time before they are cleared mainly into the liver (Braude, 1964). Since they are not found on circulating blood cells, the very early main targets must be constituents of the plasma. One early target of endotoxin after its intravenous administration is the complement system. The property of lipopolysaccharides to interact with the complement system in vitro has been known for a long time (Pillemer et a / . , 1955). Recently, using lipopolysaccharides in physicochemically defined forms, it could be shown that the property to interact with complement in Litro is expressed only by lipopolysaccharides that are present in a high degree of aggregation (Galanos and Luderitz, 1976). In a more recent investigation carried out in rats, it was demonstrated that a high aggregation is also required for the interaction of lipopolysaccharide with C’ iri viva (Freudenberg and Galanos, 1978). The in vivu anticomplementary activity of lipopolysaccharide, however, is not related to its toxic properties. Thus the Iipopolysaccharide of Rhps. virirfis, which was shown to be nontoxic (Galanos r t a / . , 1 9 7 7 ~ )in mice, was highly anticomplementary iri v i w following its adminstration in rats, however in the absence of toxic effects (Freudenberg and Galanos, 1978). Skarnes (1968) was the first to identify lipopolysaccharide complexes formed with lipoprotein from the plasma. Recently lipopolysaccharide-HDL complexes in circulating plasma (and serum) of rats have been identified by crossed immunoelectrophoresis (Freudenberg et a l . , 1980a). These complexes were formed within 3 minutes of the administration of lipopolysaccharide. The binding of lipopolysaccharide to HDL leads to a reduction in the rate of lipopolysaccharide clearance from the blood (Mathison and Ulevitch, 1979). Bound lipopolysaccharide circulates in low-density form (Ulevitch et a / . , 1979), and it has been reported that the endotoxic activity of the complexed lipopolysaccharide in this form is reduced (Skarnes, 1968; Ulevitch and Jonston, 1978). I t is the interaction with HDL, however, that probably determines the distribution of lipopolysaccharide in vivo. The affinity of lipopolysaccharide for HDL is very high, and coating of red blood cells with lipopolysaccharide is completely inhibited in the presence of HDL. The active principle in the binding reaction is lipid A, and incubation of

124


free lipid A with HDL immediately leads to complex formation (Freudenberg et a / . , 1980c). It should be noted that in other animal species lipoproteins other than HDL might also be involved in lipopolysaccharide transport. The clearance time for lipopolysaccharide from the blood depends on the immune status of the animal (host), the lipopolysaccharide structure (S, R form), and the degree of lipopolysaccharide aggregation as shown by investigating plasma clearance of [14C]lipopolysaccharidepresent in different salt forms. Under the conditions used, the half-time of clearance was less than 2 minutes and 30 minutes for Re lipopolysaccharide Na and triethylamine forms, respectively, and 6 hours and 7.5 hours for S lipopolysaccharide Na and triethylamine froms, respectively (M. A . Freudenberg and C. Galanos, unpublished data).

2. DISTRIBUTION OF LPS

IN

EXPERIMENTAL ANIMALS

Earlier studies had shown (see Braude, 1964) that endotoxin is cleared mainly into the liver, and to some extent into the spleen. This has since been confirmed in many laboratories. There is general agreement concerning the uptake of lipopolysaccharide by hepatic macrophages, and only a few authors have detected lipopolysaccharide in hepatocytes (Willerson et al., 1970; Zlydaszyk and Moon, 1976; G. Ramadori, C. Galanos, V. Hopf, and K . Meyer zum Buschenfelde, unpublished data). Recently immunohistochemical methods have been applied to study the kinetics of the distribution of Salrnorzellu S- and Re-form lipopolysaccharides in the rat. It was found (Freudenberg et nl., 1980b) that the first cellular targets for the S-form lipopolysaccharide were phagocytic cells, mainly Kupffer cells of the liver, macrophages of the spleen and some granulocytes in agreement with previous studies. Two to three days after injection, however, a redistribution of the lipopolysaccharide in the liver was observed, by this time the hepatocytes becoming strongly endotoxin positive. When Re-form lipopolysaccharide was used, it was from the beginning in Kupffer cells, granulocytes, and hepatocytes, indicating that hepatocytes may in principle also clear lipopolysaccharide from the blood. Both S- and R-form lipopolysaccharide preparations showed a decrease with time in the staining activity in the liver. Nine days after injection only very weak diffuse lipopolysaccharide staining was seen, indicating that part of the lipopolysaccharide was no longer present in the liver or that its antigenic specificity had been modified, or that both effects were taking place. Similar differences in the primary uptake of S- and Re-form lipopolysaccharides were detected in vivo and in witro in mice (Ramadori et ul., 1980). However, in a recent study in rabbits, association of the Re lipopolysaccharide with hepatocytes was not observed (Mathison and Ulevitch, 1979). There are considerable indications that the liver is an important organ of lipopolysaccharide excretion. In

125


rabbits high concentrations of [ I Z 5 I llipopolysaccharide were observed in the gallbladders (Mathison and Ulevitch, 1979); in rats injected with [14C/ :~H]lipopolysaccharide,about 60% of the label was found in the feces for u p to 3 weeks following lipopolysaccharide treatment (Kleine, 1981). An interesting observation was made concerning uptake of lipopolysaccharide by the lung (Freudenberg e / al., 1980b). During the first hours after injection this organ was free of tissue-bound endotoxin. However, a large number of endotoxin-positive cells appeared with a delay of several hours when compared to the presence of bound endotoxin in the liver. It is possible that lipopolysaccharide-carrying cells originate from the liver and (or) spleen and are excreted in the lung. The presence of a number of lipopolysaccharide-carrying alveolar and bronchiolar macrophages between 24 hours and 5 days after endotoxin treatment supports this assumption. PROSTAGLANDIN 3. LIPOPOLYSACCHARIDE-INDUCED

IN

SYNTHESIS

MACROPHAGES

It is known that a number of endotoxin effects such as fever, early hypotension, skin necrosis, and abortion can be suppressed by acetylsalicylic acid (aspirin) or indomethacin (i.e., drugs that are inhibitors of prostaglandin biosynthesis). Furthermore, certain of the prostaglandins induce typical endotoxic reactions. These observations have led to the hypothesis that prostaglandins may represent mediators of lipopolysaccharide effects. In fact, macrophages could be identified as the cell type that, on incubation with lipopolysaccharide, would be stimulated to synthesis and excretion of prostaglandins of the E, and F2a type (Rietschel e/ al., 1980a). Three lines of evidence exist indicating that macrophages are indeed the source of prostaglandins mediating the lipopolysaccharide effects in vivo. It has been demonstrated in several laboratories that the mouse strain C3H/HeJ, which is genetically resistant to a number of endotoxin effects including lethality (lipopolysaccharide nonresponder), produces macrophages that cannot be stimulated by lipopolysaccharide to secrete PGE, and F,a. As expected, cells from genetically related lipopolysaccharide-responder strains were stimulated to prostaglandin synthesis and excretion (Wahl et a / . , 1979; Rietschel e/ al., 1980b). Second, mice can be rendered nonresponsive (tolerant) to lipopolysaccharide by repeated administration of sublethal doses of lipopolysaccharide. Macrophages from tolerant mice are completely refractory to the action of lipopolysaccharide and do not secrete prostaglandins E, and F,a (Schade and Rietschel, 1980). Third, rats fed with a diet devoid of essential fatty acids, which are precursors of prostaglandin biosynthesis, are highly resistant to lipopolysaccharide lethality (Cook er a / . , 1979). Although these results are consistent with the hypothesis that macrophage-

126

Om0 LUDERITZ ET AL.

derived prostaglandins play a role i n the mediation of lipopolysaccharide effects, further studies are needed to elucidate possible etiology or regulatory functions of prostaglandins. Most data in the past have been obtained in in vitro experiments with mouse macrophages obtained from the peritoneum or from the bone marrow. Preliminary results show that Kupffer cells (rat) and alveolar macrophages (rabbit) are also stimulated by lipopolysaccharide to release prostaglandins E, and F p (B. Bhatnagar, K. Decker, K . Tanamoto, U . Schade, and E. T. Rietschel, unpublished data). 4 . GALACTOSAMINE-INDUCED SENSITIZATION TO T H E LE I’HAI. EFFECTSOF LIPOPOLYSACCHARIDES Several experimental models exist by which the natural sensitivity of animals toward lipopolysaccharide (endotoxin) may be increased. Examples are adrenalectomy, and treatment with Bacillus Calmette-Guerin, actinomycin D, lead tetraacetate, a-amanitin (Schlievert et al., 1980),galactosamine (Galanos et ul., 1979b), or antigen-antibody complexes (Galanos, 1979). Increased susceptibility to endotoxin may have a role to play under natural conditions. It is known that outer membrane proteins of gram-negative bacteria are immunogenic, and it has been shown that bacterial protein-antibody complexes will enhance the lethal toxicity of lipopolysaccharide. Since a wide serological cross-reactivity exists among proteins of different bacterial genera, and since animals and humans are permanently in contact with gram-negative bacteria, preexistent bacterial protein antibodies could play a role during infection by increasing susceptibility of the host to lipopolysaccharide (Galanos, 1979). The susceptibility-increasing models mentioned previously have multiple effects on the organism as a whole, but several are also known to be poisonous to the liver. Galactosamine, however, represents a highly specific hepatotoxic agent, and mechanisms underlying “galactosamine” hepatitis have been extensively elucidated (Decker and Keppler, 1974). Administration of galactosamine and lipopolysaccharide (either together or with the lipopolysaccharide within 1 hour later) will lead to a dramatic decrease (Galanos e f al., of the LD,, in mice (by factors lo-“ to lop5),or rabbits 1979b). It is the lipid A part of lipopolysaccharide to which the animals are sensitized. If uridine is applied within 3 hours after the galactosamine treatment, the effect of galactosarnine is negated. In this model the early metabolic effects induced by galactosamine are sufficient for sensitization to endotoxin. At a time when the animals are highly susceptible, typical liver enzymes have not yet been released into the blood, indicating the absence of cell injury. This shows that temporary metabolic changes without clinical symptoms may render an organism highly susceptible to very small amounts of endotoxin, and this may be of clinical importance.


127

It has been observed over many years that different animal species show large variations in their susceptibility to lipopolysaccharide. Mice and rats are relatively resistant, whereas rabbits and humans generally show a high susceptibility. Rabbits also exhibit a high degree of variation, however. On average, the LDIo0 is about 50 p g of endotoxin for a given strain, but in a single group of rabbits individuals are found that are very susceptible, whereas others can survive milligram amounts of lipopolysaccharide. It has now been shown that in the small percentage of rabbits that die with 0.1 to 1 p g of lipopolysaccharide, the kinetics of liver-enzyme release into the blood are similar to that observed in rabbits made sensitive to lipopolysaccharide by galactosamine. It might be concluded that in natural sensitization, biochemical alterations i n the liver, similar to those induced by galactosamine, are involved. The liver, apart from being the main organ of endotoxin clearance, could thus play a role in the initiation of endotoxic effects (C. Galanos and M . A . Freudenberg, unpublished data).

5. ROLE O F T H E POLYSACCHARIDE POKTION OF LIPOPOLYSACCHAKIDES I N CELLULAR RECOGNIT ION The fundamental importance of polysaccharides in recognition phenomena rests on their property of exhibiting compositional and structural diversity. Lipopolysaccharides are composed of a variety of different sugar residues that are linked in a number of different ways. They are thus extremely complex and diverse molecules with a high value of structural information. Inserted asymmetrically into the outer membrane with their carbohydrate determinants facing and extending outward, lipopolysaccharides are ideal recognition structures for phages, antibodies, or eukaryotic cells. It is this complexity and diversity of the polysaccharide moiety of lipopolysaccharides that make gram-negative bacteria useful probes for the detection and identification of carbohydrate-binding structures such as lectins o r lectin-like structures on eukaryotic cells. Evidence for the existence of lectin-like receptors on phagocytes comes from work on mouse peritoneal macrophages. Bacteria are bound to the macrophage membrane by a mechanism that appears to involve recognition of sugars derived from lipopolysaccharides by membrane-bound lectin receptors of the macrophage (Weir, 1980). The binding of bacteria by macrophages in monolayers can be inhibited by preexposure of the monolayer to a variety of monosaccharides. Glucose and galactose inhibit the binding of numerous bacterial species such as E . coli, Ps. ueruginosa, or S . typhimiirium. If one or both sugars are not present in mutant lipopolysaccharides, the missing sugars become noninhibitory, despite their ability to inhibit the binding of the wild-type organism. Thus S . f ~ ~ ~ ; core ~ mutants ~ z f of ~ the ; chemotype ~ ~ ~ Re, with no outer core or 0 chain, appear to bind via their inner-core components, and binding is not inhibited by any sugar that inhibits binding of the wild type. A similar picture

128


emerges with Klebsiellu uerogenes, in which galactose fails to inhibit binding of the galactose-deficient core mutant MlOB. in a more recent approach bacterial mutants from Safmone/lu strains were selected that are able to bind to lectin-like structures expressed on activated T lymphocytes and absent on resting T lymphocytes (Lehmann et af., 1980). The selection procedure included two consecutive steps: the mutagenesis of nonadhering smooth strains from Sulmonellu and the subsequent enrichment of mutants adhering to T lymphocytes that had been activated in a mixedlymphocyte culture. Enrichment of the desired mutants was achieved by I g-velocity sedimentation, a procedure that separates cells on the basis of size differences. Accordingly, T-cell blasts and adhering mutants were separated from nonresponsive small lymphocytes as well as from nonadhering bacteria. Adhering mutants were shown to belong to the class of rough mutants with the chemotype Ra or Rb (Fig. 4).These mutants exhibit specific binding properties. They bind to a T-cell subpopulation, tentatively identified as T helper cells, whose proliferation and differentiation is required for the generation of killer or suppressor cells. Binding of the mutants is mediated by lectin-like receptor sites on the T-cell subset, which recognizes a sugar portion of Ra or Rb lipopolysaccharides on the bacterial surface. The involvement of polysaccharides in the binding process rests on the observation that adherence is specifically inhibited by lipid A-free Rb polysaccharides. Marginal or no inhibition was observed with polysaccharides derived from the wild-type strain or from Sulmonellu Rc, Rd, or Re mutants. There is evidence that the Rb polysaccharide-recognizing structures on activated T lymphocytes have a functional role for the differentiation of T-cell subsets. This conclusion is based on the observation that polysaccharides of the chemotype Rb that inhibited adherence of the mutants to activated T cells suppress also the generation of killer T cells. It is believed, therefore, that the interaction between the bacterial polysaccharides and the lectin-like structures on T helper cells prevents differentiation of these cells, or alternatively, mediates the induction of suppressor cells that inhibit the generation of killer cells. So far only a limited number of lectin or lectin-like structures have been identified on phagocytic and nonphagocytic cells. Selected bacterial polysaccharides may reveal more of these structures on lymphoid and nonlymphoid cells and may eventually show that there are in nature similar numbers of carbohydrate and complementary structures, the lectin or lectin-like proteins, that interact and mediate fundamental functions in biology (see also Mayer and Teodorescu, 1980). 6. LIPOPOLYSACCHARIDE DEGRADATION IN NATURE

Several biological systems have been identified that will degrade lipopolysac~ R8, , and others), whose recharide. A number of bacteriophages (el5,E ~ P22,


129

ceptor site is lipopolysaccharide (from Salmonella, Escherichia, Shigella$exneri. Proreus), have been identified and shown to exhibit (end0)glycosidase activity (Iwashita and Kanegasaki, 1973; Reske et d.,1973; reviewed by Lindberg, 1977; Braun and Hantke, 1981). This activity resides in the spikes or tail parts, which consist of a single protein species. These enzymes recognize, absorb to, and cleave the 0 antigen, which is the specific phage receptor of the host. Interaction of phage or isolated enzyme with the respective lipopolysaccharide leads to the cleavage of specific linkages in the 0 chain and to the liberation of mono- and oligomeric repeating units (which may differ in their sugar sequence from the biological or chemical repeat unit). During this interaction and cleavage, core and lipid A remain intact. Slime molds constitute another system of biodegradation of lipopolysaccharides. These organisms, which are ubiquitous in soil, utilize bacteria for food. Dictyostelium discoideum has available acylesterases and amidases that degrade lipopolysaccharide by cleaving off the long-chain fatty acids from lipid A. The remaining polysaccharide-lipid A backbone is then secreted by the cells (Malchow et ul., 1969). Recently, Rosner ei d., ( I 979c) have succeeded in purifying two acyl amidases from D . discoideum. Reaction of amidase I with 0deacylated lipid A leads to the removal of the glucosamine I-linked 3-OH-14:O fatty acid. The reaction product formed (but not the original lipid A) then reacts as a substrate for amidase 11, which cleaves the glucosamine 11-linked 3-OH14:O fatty acid. This reaction sequence represents an elegant and specific way for preparing defined degradation products of lipid A , which may be useful in studies of structure-function relationships. Saddler et al. (1979a) have shown that the slime mold Physarium po/ycepha/um also reacts with lipopolysaccharide. Again, the lipid A part is degraded. In this case, however, only the nonhydroxylated fatty acids are removed. A similar degradation pathway for lipopolysaccharides was observed with the gut juice of the snail Helix pomatiu (Saddler et al., 1979b). A microorganism, Bacillus macerans, has been selectively grown from soil samples by culture on mineral medium supplemented by lipopolysaccharide. This organism is assumed to cleave the polysaccharide-lipid A linkage and to remove long-chain fatty acids from lipid A (Voets e t a / . , 1973; see also Saddler and Wardlaw, 1980). Whether the higher organism is capable of degrading lipopolysaccharide has hitherto not been established definitively. In connection with endotoxin inactivation by serum factors, it has been suggested that serum esterases might degrade lipopolysaccharide (Skarnes, 1968), but no further evidence has been presented for such reactions (see Johnson et a/.,1977). Current investigations (Kleine, 1981) on excretion products of rats and vervets injected with radiolabeled endotoxin indicated that more than 60% of the label is found in the feces and urine, mainly in the form of degradation products.

o n 0 LUDERITZ ET AL.

130

IV.

FINAL REMARKS

Lipopolysaccharides are still in fashion. It is tempting to believe, as many did 50, 20, or 10 years ago, that research during the next decade will solve all the remaining problems. Experience indicates, however, that with new answers new questions will arise. The past decade has furnished us with detailed chemical formulas of many lipopolysaccharides, and the structural principles of their architecture are now known for gram-negative bacteria of various families. Of equal importance has been the elucidation of their immunological principles, their pathways of biosynthesis, and their genetics. Using methods that have proved successful in the past, structures. biosynthesis, and genetics of lipopolysaccharides different from those hitherto studied will be investigated. Efforts in the research on lipopolysaccharides are now concentrated on their interaction with other outer membrane components. and their role in the molecular organization and biogenesis of the outer membrane. The function of the outer membrane and the communication of the cell surface with the environment are further topics of investigation. Readers interested in present highlights of research and perspectives in outer membrane biochemistry and physiology are advised to study the excellent reviews of Nikaido and Nakae (1979), Inouye (1979), Makela and Stocker (1981), Braun and Hantke (1981), and the recent papers of Osborn and her group (1980). Their discussions point to future research topics. Another central aspect in this field concerns the role of lipopolysaccharides in pathogenicity and protection against infection, as well as their physiological role in normal animal life. The reviews of Melchers (1980), and the other contributions at the Dahlem conference 1979 on “The Molecular Basis of Microbial Pathogenicity” provide facts and hypotheses in this field (Smith et ul., 1980). As a conseqeunce of the identification of lipopolysaccharide structures, a number of research groups have succeeded in synthesizing the immunodeterminant structures of the 0 chains of clinically interesting enterobacteria. Coupled to protein carriers, they provide potent, nontoxic immunogens that lead to specific high-titer antisera valuable for diagnostic purposes, and with the potential to protect against experimental infection (Svenungsson and Lindberg, 1978b; Josephson and Bundle, 1979). As immunization with R-mutant bacteria has been shown to induce cross protection against infection by a variety of bacterial pathogens based on the presence of a common (inner) core structure (Ziegler ef ul., 1979, 1981; McCabe erul., 1977; Diena et u l . , 1978), it certainly will be of importance that R-type-specific nontoxic immunogens (containing core-lipid A backbone coupled to proteins) and specific pure R antibodies (obtained by immunoadsorbants containing the antigen coupled to solid carriers) are now available. The same is true for studies on the role of anti-lipid A antibodies in

131


endotoxin pathogenesis; these antibodies can now be produced with nontoxic immunogens and highly purified by immunoadsorbants (Section III,A,2). In the vast field of endotoxin research, the old idea of Menkin (1953) that most endotoxin effects are mediated by secondary, endogenous factors, has been widely accepted and substantiated. A number of lipopolysaccharide-inducible mediators have been discovered (Schlessinger, 1980). Lipopolysaccharide nonresponder mouse strains have provided an important model in this line of research. Standardized pure lipopolysaccharide has been made available for comparative studies. But we are still only beginning to understand the molecular basis of endotoxin action (Berry, 1977). The evaluation of the lipid A structure has provided the possibility of its chemical synthesis. Carbohy,drate chemists experienced in the synthesis of amino sugar-containing natural antigens ( e . g . , bacterial 0 antigen or blood-group determinants) agree that by using available techniques, lipid A synthesis from the chemical point of view presents n o special problems. Just before completion of this article, Prof. Shiba from Osaka University, Japan, gave us a presentation on his approach to synthesize the “fundamental structure” of lipid A (Fig. 14; lnage rt al., 1980a,b) and his ideas to prepare the diphosphorylated derivatives. Also Kiso et al. ( 1 980, 1981a.b) and Nashed and Anderson ( I 98 1) have published recently the chemical synthesis of intermediates suitable for lipid A synthesis. Therefore, we can assume that a “lipid A ” synthesis will soon be achieved. Unfortunately, however, some reservations have to be made as to the lipid A structure. Even in the case of Salmondla, only the main structural features are proven; a number of minor details remain to be evaluated (Section lI,D,I). In spite of a lack of complete knowledge of the lipid A structure, we feel able to provide some predictions on substructures necessary for the activity of lipid A . This would make chemical synthesis more rational, as simpler molecules may be synthesized. From studies of structure-activity relationships (Section III,B) i t has become evident that the polar head groups of lipid A are not essential for the nHO HO

H,OH I

YoR1

R

I

R = R l = 1L.O

11

R =

14.0

to R1

R1 = 3 - O H - l L : O

FIG. 14. Synthetic partial lipid A

htructures

(Inage et d., 1980a,b)

132


biological activities; the phosphate group linked to glucosamine 1 may not even be necessary. It would follow that the acylated lipid A backbone represents the biologically effective structure of lipid A. The length of the fatty acids is not decisive and may vary within certain limits. That acylated hydroxy acyl esters are not a prerequisite for endotoxicity, follows from the finding that they are absent from certain active lipid A's (e.g., from Chr. violaceutn lipid A). Nothing is presently known about the role of the following factors for activity: Are the 3-hydroxy groups of amide- or ester-linked fatty acids essential'? Is their D-configuration important? Are 0-3-acylated hydroxy acyl amides essential? Is the (presently unknown) distribution of the ester-bound fatty acids on the backbone critical? An atomic model of lipid A, where the nonhydroxylated acyl units are attached directly to the backbone (not to the amide-bound hydroxy fatty acids as in Fig. 6), allows the acyl residues to be brought into parallel positions (like in Fig. 6). This may indicate that the fatty acid distribution on the lipid A backbone is not decisive for interaction of the molecules with cell membranes, which probably is a prerequisite for activity (Kabir et a / . , 1978). From our own investigations it would follow that the smallest lipid A substructure identified so far and endowed with some biological activities is represented by precursor I (Fig. IOa), containing the lipid A backbone (it seems that the presence of the glucosamine I-bound phosphate is not obligatory) and carrying two amide-bound and two ester-bound 3-hydroxy fatty acids (the latter in two unknown positions). This product exhibits antigenicity, mitogenicity, lethal toxicity, (weak) pyrogenicity, and (weak) complement reactivity, but strong Liinulus lysate activity. The two ester-bound acyl residues are important, since the completely 0-deacylated product has lost most activities. Substituents on the backbone other than acyl groups, such as glucosamine, and even core and 0 chains i n some of the cases (Rhsp. tenue), seem to reduce some endotoxic activities (Table VI). The chemical synthesis of the lipid A backbone substituted with two amidelinked 3-hydroxy fatty acids can be achieved according to the opinion of experts. A limited esterification with a suitable mixture of fatty acids should also be possible. Such synthetic products should be available in the near future and it will be of great interest to see whether they act as endotoxins. Alternatively, it may turn out that gram-negative bacteria have developed the synthesis of the one endotoxin structure, where the fatty acids are linked to defined positions. The synthetic product might then contain by chance the structurally correct active molecular species in a mixture with inactive ones. We must, however, also consider the prospect that the proposed structure of lipid A is still incorrect in one or more important details, or even incomplete. The de novo chemical synthesis of lipid A analogues or their resynthesis from lipid A degradation products represents a further promising approach. Table IV gives examples of such model substances all with lipid A-like activities, some

133

LIPOPOLYSACCHARIDESOF GRAM-NEGATIVE BACTERIA

Lipopolysocchoride (Endotoxin) l P e i j ICOO'lj I

6

Polysoccharide 0 Chain

Lipoprotein

I

r '

(Pel2 (FA16-7 I 1

I '

Lipid A

Core

(8-cell mitogen) IFAlj

L

l

-

Polypept d e

4

'

1

Lipid

Lipoteichoic Acid (Shwortzmon + I

Polyglycerol- P D

Peptidoglycon

1ipid

(Endotoxin-like1

Polysorcharide - Peptide

FIG. 15. acids.

Schematic structures and biological activities of bacterial cell-wall antigens. FA, fatty

structures being quite remote from lipid A . Korhana and his group are studying the effect of the nature of the amide-linked fatty acids on lipid A activity. With the aid of specific amidases from ameba, they are able to remove sequentially the amide-bound acyl residues from the lipid A backbone with the intent of replacing them by other acyl residues (Rosner el al., 1 9 7 9 ~ ) . In this context it should be emphasized that in gram-positive and gramnegative bacteria, cell wall components other than lipopolysaccharide may also be endowed with endotoxin-like activities (Wicken and Knox, 1980; Fig. 15). Like lipopolysaccharide, these constituents are amphipathic in nature (probably with the exception of murein, though it may contain lipoprotein). These findings indicate that at least some lipid A activities are not restricted to one specific structure, but are rather connected with general physicochemical properties. ACKNOWLEDGMENTS We would like to express our thanks to Dr. H. Formanek, University of Munchen, for his help in constructing the lipid A model (Fig. 6 ) . We thank our colleagues K. and B . Jann, H. Mayer, S. Schlecht, and G . Schmidt for critiques and suggestions. We also thank Miss Helga Kochanowski for preparing the drawings, Mrs. Ingrid Himmelspach and Mrs. Lore Lay for the photographs, and Mrs. Rosemary Schneider for typing the manuscript.

134

O n 0 LUDERITZ ET AL. REFERENCES

Ahamed. N. M.. Radziejewska-Lebrecht, J . , Widemann, C . , and Mayer. H. (1980). Reactivity of isolated lipopolysaccharides of enterohacterial R mutants with complete or incomplete core . I : Orig., structures with lectina. Zentru/b[. Bukteriol.. Porusitenkd. Infekrionskt-. H J ~ .Abt. Reihe A 247, 468-482. Agarwal, M. K . (ed.) (1980) “Bacterial Endotoxins and Host Response,” pp. 1-250. Elsevicr. Amsterdam. Bagdian, G . . Luderitr, 0.. and Staub, A. M. (1966). Immunocheniical studies on Sulmonellu. X I . Chemical modification correlated with conversion of group B Sulrnonrlltr by hacterophage 27. A W I . N . Y . Actrcl. P i . 133, 405-424. Bagdian, G . , Eticvant, M., and Staub, A . M. (1967). Etudes immunochimiques sur les Sulnionelltr. XIV. Analyse des polyosides specifiques extraits de S. /\phi T, sauvage et convertie par le phage 27. Colloy. lnt. C.N.R.S. 174, 161-187. Banerji, B., and Alving, C. R . (1979). Lipid A from endotoxin: Antigenic activities of putified fractions in liposomes. J . Immunol. 123, 2558-2562. Banerji, B., and Alving, C. R. (1981). Anti-liposome antibodies induced by lipid-A. 1 . lntlucnce of ceramide, glycosphingolipids. and phosphocholine on complement damage. J . Imnitnol. 126, IOXO- 1084. Bayer, M . E. (197.5). Role of adhesion zones in bacterial cell-surface function and biogenesis. I n “Membrane Biogenesis” (A. Tzagaloff, ed.), pp. 393-427. Plenum, New York. Beckmann, I., Subbaiah, T . V . , and Stocker, B. A. 0.(1964). Rough mutants of Salntone.llu typhimuriuni (2) Serological and chemical investigations. Nature (London) 201, 1299- 1301 Behling, U . H . , Campbell, B., Chang. C.-M., Rumpf, C . , and Nowotny, A. (1976). Synthetic glycolipid adjuvants. .I. Imrnunul. 117, 847-85 I . Berry, L. J . (1977). Bacterial toxins. Crit. Rcv. T o x i i d . 5, 239-318. Berst, M., Hellerqvist. C. G . , Lindberg, B . , Liideritz, O . , Svensson,’S., and Westphal, 0. (lY69). Structural investigations of TI lipopolysaccharides. E u r . J . Biochem. 11, 353-359. Blache, D., Bruneteau. M., et Michel, G. (1981). Structure de la region heptose du lipopolysaccharide de Escherichiu coli K-12 CR34. E u r . J . Biochrm. 113, 563-568. Blake. D., Hamlyn, A . N., Proctor, S . . and Wardle, E. N . (1980). Anti-endotoxin (anti-lipid A) antibodies. Experientu 36, 254-255. Boivin, A., Mesrobeanu, J . , and Mersobeanu, L. (1933). Technique pour la preparation des polyosides microbjens specifiques. C. R . Sennres Sor. B i d . SPS. Fil. 114, 307-312. Boman, H. G., and Monner, D. A. (1975). Characterization of lipopolysaccharides from Escherirhiu coli K-I2 mutants. J . Bucteriol. 121, 455-464. Bradley, S. G . (1976). Endotoxic activity of complexes of myristic acid and proteins. P r w Soc. Exp. B i d . Med. 151, 267-270. Bradley, S. G . (1979). Cellular and molecular mechanisms of action of bacterial endotoxins. Annn. Rev. Mirrobiol. 33, 67-94. Braude, A. I. (1964). Absorption, distribution and elimination of endotoxins and their derivatives. I n “Bacterial Endotoxins” (M. Landy and W. Braun, eds.), pp. 98-109. Rutgers Univ. Press, New Brunswick, New Jersey. Braun, V., and Hantke, K . (1981). Bacterial cell surface receptors. In “Organization of Procaryotic Cell Membranes” (B. K. Ghosh. ed.). Vol. 2 , pp. 1-73. CRC Press, Boca Raton, Florida. Broady, K . W., Rietschel, E. Th., and Liideritz, 0. (1981). The chemical structure of the lipid A component of lipopolysaccliarides from Vibrio cholerae. E u r . J . Biochem. 115, 463-468. Bruneteau, M., Volk, W. A . , Singh, P. P., and Liideritz, 0. (1974). Structural investigations on the Salmonella T2 lipopolysaccharide. Eur. J . Biochem. 43, 501 -508. Bryn, K., and Rietschel, E. T. (1978). L-2-hydroxytetradecanoicacid as a constituent of Salmonella lipopolysaccharides (lipid A). E u r . J . Biochem. 86, 31 1-315.


135

Chaby. R . . Moreau. M . , and Szabo. L. (1977). 2-U-(P-~-glucuronyl)-7-O-(2-amino-2-deoxyu-g1ucopyranosyl)-i .-glycero-D-manno-heptose: A constituent of the Bordi*td/uprrtirssi.\ endotoxin. E u r . J . Biochem. 76, 453-460. Chang, C . M . . and Nowotny, A. (1975). Relation of structure to function in bacterial 0-antigens. VII. Endotoxicity o f “lipid A . Immunochrrnistry 12, 19-28. Coleman, W . G . , and Leive, L. (1979). Two mutations which affect the barrier function of the E . coli K-12 outer membrane. J . Bocwriol. 139, 899-910. Cook, J. A., Wise. W. C., and Callihan, C. (1979). Resistance to endotoxin shock by m s deficient in essential fatty acids. Fed. Proc., F e d . A m . Soc. E-xp. Biol. 38, 1261 Creeger, E. S . , Chen, J . F . , and Rothfield, L . I . (1979). Cloning of genes for bacterial glycosyltransferases. 11. Selection of a hybrid plasmid carring the rfaH gene. J . B i d Chern. 254, ”

811-815.

Curvall, M . , Lindberg, B . , Lonngren. J . , Ruden, U., and Niinmich, W . (1973). Structural studiesof the Klehsiella 0 group 8 lipopolysaccharide. Actti Chem. Scund. 27, 4019-4021, Datta, D. B . , Arden. B . , and Henning. U . (1977). Major proteins of the Escherichiu coli outer cell envelope membrane as bacteriophage receptors. J . Buctrriol. 131, 821 -829. Decker. K . , and Keppler, D. (1974). Galactosamine hepatitis. Key role of the nucleotide deficiency period in the pathogenesis of cell injury and cell death. Rev. Physiol., Biochem. Phurinucol. 71, 77-106. Diena, B . B., Ashton, F. E.. Ryan, A , , and Wallace, R . (1978). The lipopolysaccharide ( R type) as a common antigen of Neisseriir gonorrhoecie. 1. Immunizing properties. Curl. J . Microhiol. 24, 117-123. Dmitriev, B. A . , Backinowsky, L. V . , Vjacheslav, L , Lvov. Y. A., Knirel, Y . A,, and Kochetkov, N. K. (1975a). The structure of the chemical repeating-unit of the 0-specific polysaccharide chain of Shigdla boydii 6 lipopolysaccharide. Ctrrhohdr. Rrs. 41, 329-333. Dmitriev, B . A., Knirel, Y . A., and Kochetkov, N . K . (197%). Selective cleavage of glycosidic linkages: Studies with the polysaccharide component of Shigeila clmenteriue type 6 lipopolysaccharide. Carhohydr. Res. 44, 77-85. Dmitriev, B . A., Lvov. V . L . , Kochetkov, N. K., Jann, B . . and Jann, K . (1976a). Cell wall lipopolysaccharide of the “Shigella-like” Eschrrichiu coli 0124. Structure of the polysaccharide chain. Eur. J . Biorhern. 64, 491-498. Dmitriev, B. A., Knirel, Y. A , , and Kochetkov. N. K . (1976b). Somatic antigens of Shigella. Structural investigation on the 0-specific polysaccharide chain of Shigella dysenteritre type I lipopolysaccharides. Eur. J . Biochern. 66, 559-566. Dmitriev, B. A , , Knirel, Y . A , , and Kochetkov, N. K . (1977a). Somatic antigens of ShigrlIa. Structural investigation on the 0-specific polysaccharide chain of Shigella c/ywn/rritre type-2 lipopolysaccharide. Eur. J . Rioclirm. 76, 433- 440. Dmitriev, B . A., Backinowsky, L. V., Knirel. Y. A , . and Kochetkov, N. K. (1977b). Somatic antigens of Shigella. The structure of the specific polysaccharide chain of Shigrllu dysentrriae type 5 lipopolysaccharide. Eur. J . B i o c h r r . 78, 1381-387. Dmitriev, B. A., Knirel, Y . A . . Kochetkov. N. K . , J a m , B., and J a m , K. ( 1 9 7 7 ~ ) Cell-wall . haridc of the “Shigellu-like” Eschrrichiu c d i 058. E i w J . Biorhem. 79, 1 I I 11s.

Dmitriev, B . A , , Lvov, V . L . , and Kochetkov, N . K . (1977d3. Complete structure of the repeating unit of the 0-specific polysaccharide chain of Shrgello clyscweriae type 3 l~popolysaccharide. Curhohydr. Res. 56, 207-209. Dmitriev, B. A., Knirel, Yu. A., Sheremet, 0. K . . Kochetkov, N . K . , and Hofman, I . L. (1977e). Antigenic polysaccharides of bacteria. V . Structure of the 0-specific polysaccharide chain of Shigrllu c/~se!iteriurtype 10 lipopolysaccharide. Bioorg. Khirn. 3, 1219-1225. Dmitriev. B . A., Lvov, V . L., Kochetkov, N . K . . and Hofman, I . L. (19770. Antigenic polysac-

136

OTTO LUDERlTZ ET AL.

charides of bacteria. VI. Structure of the 0-specific polysaccharide chain of Shigella dysenreriae type 4 lipopolysaccharide. Bioorg. Khim. 3, 1226-1233. Dmitriev. B. A . . Knirel. Y u . A , , Vinogradov. E. V., Kochetkov, N . K . , and Hofman, I. L. (197th). Bacterial antigenic polysaccharides. VII. Structure of the polysaccharide chain ofthe Shipellu dvsetiferiae type 9 lipopolysaccharide. Bioorg. Khim. 4, 40-46. Dmitriev, B. A . , Lvov, V . L., Ramos, E. L., Kochetkov, N . K., and Hofman, 1. L. (197%). Antigenic plysaccharides of bacteria. VIII. Structure of the specific polysaccharide chain of Shigella dysenteriue type 8 lipopolysaccharide. Bioorg. Khim. 4: 760-766. Dmitriev, B. A . , Knirel, Y . A . , Sheremet. 0. K . , Shashkov. A. A., Kochetkov, N . K., and Hofman, I. (1979). Somatic antigens of Shipella. The structure of the specific polysaccharide of Shigella newcusrle (Sh.Jlexneri type 6) lipopolysaccharide. Eur. J . Biuchem. 98, 309-316. Dmitriev. B. A , , Knirel, Y . A , , Kocharova, N. A , , and Kochetkov, N. K. (1980). Somatic antigens of Pwudomonrrs rrcwgimva. The structure of the polysaccharide chain of Ps. wruxinosu 0 sero-group 7 ( L m y i ) lipopolysaccharide. Eur. J . Binrhem. 106, 643-651. Drewry. D. T . , Symes, K. C., Gray. G . W., and Wilkinson, S . G . (1975). Studies of polysaccharide fractions from the lipopolysaccharide of Ps. armginmu N.C.T.C. 1999. Biuchcm. J . 149, 93-106. Drews, G . , Weckesser, J.. and Mayer, H. (1978). Cell envelopes. i n “The Photosynthetic Bacteria” (R. K . Clayton and W. R. Sistrom, eds.). pp. 61-77. Plenum, New York. Drdgc, W . , Lehmann, V.. Luderitz, O . , and Westphal, 0. (1970). Structural investigations on the KDO region of lipopolysaccharides. Eur. J . Biorhem. 14, 175-184. Edstrom. R . D., and Heath, E. C. (1967). Thc biosynthesis of cell wall lipopolysaccharide in Escherichia coli. V11. Studies on the structure of the 0-antigenic polysaccharide. J . Biol. Cham. 242,4125-4133. Erbing, C., Kenne, B., and Lindberg. B. (1977a). Structural studies of the 0-specific side-chains of the cell-wall lipopolysaccharide from Esrheric,hiu c d i 069. Carbohydr. Res. 56, 37 1-376. Erbing, C., Lindberg, B.. and Lonngren, J . (1977b). Structural studies on the Klebsiella 0 group 12 lipopolysaccharide. Carbuhvdr. Res. 56,377-38 I . Erbing, C ., Kenne, L., and Lindberg, B. (1978). Structure of the 0-specific side-chains of the Eschr>richiu c d i 075 lipopolysaccharide: a revision. Carbohydr. R e s . 60, 400-403. Fcige, U . (1977). Uber den Rezeptor des Bakteriophagen +X174. Ph.D. Thesis, Universitat Freiburg. Feige, U., and Stirm. S. (1976). On the structure of the Escherichirr coli C cell wall lipopolysaccharide core and on its 0 x 1 7 4 region. Biochem. Biophys. Rrs. Commun. 71, 566-573. Feige, U., Jann. B., Jann. K., Schmidt, G., and Stirm, S . (1977). On the primary structure of the Esd~erirhiuC O [ ~R4 cell wall lipopolysaccharide core. Biochem. Biuphys. Res. Cummun. 79, 88-95. Fink, P. C., and Galanos, C. (1981). Determination of anti-lipid A and lipid A by enzyme immunoassay. immunology 158, 380-390. Fink, P. C . , and Kozak, J . (1980). Determination of free lipid A by enzyme immunoassay. 2. A n d . Chem. 301, 117-1 18. Formanek, H. (1078). Berechnung der riumlichen Struktur des Kohlehydratanteils eines Glycolipids. s Z . Physiol. Chrm. 359, 1083-1088. Formanek, H., and Wcidner, H. (1980). Three dimensional structure of the carbohydrate moiety of a lipopolysaccharide. Computer calculations. Z. NufurfiJrsrh. 36 (in press). Freudenberg, M. A. (1975). Specific immune responses of different animals to lipid A. Absrr., Eur. fmmunol. M e e r . . 16fh Workshop B a r f . Endofoxins. lY75, pp. 166. Freudenberg, M . A . , and Galanos, C. (1978). Interaction of lipopolysaccharidcs and lipid A with complement in rats and its relation to endotoxicity. Injkcr. fmmun. 19, 875-882. Freudenberg, M. A., BogHansen, T . C . , Back, U . , and Galanos, C. (1980a). Interaction of lipopolysaccharides with plasma high-density lipoprotein. infc,rr. immun. 28, 373-3230,


137

Freudenberg, M. A,, Freudenberg, N., and Galanos, C. (1980b). lmmunohistochemischer Nachweis von Endotoxin in den Ausscheidungsorganen der Ratte. Verh. Disch. Grs. Pathol. 64, 437. Freudenberg, M. A,, Beg-Hansen, T. C., Back. U., Jirillo, E., and Galanos, C. ( 1 9 8 0 ~ ) Interaction . of lipopolysaccharides with plasma high density lipoprotein in rats. In ‘‘Natural Toxins” (D. Eaker and T. Wadstrom eds.), pp. 349-354. Pergamon, Oxford. Fuller, N. A., Wu, M. C.. Wilkinson. R. G . , and Heath, E. C. (1973). The biosynthesis of cell wall lipopolysaccharide in E.scherichicr c d i . VII. Characterization of heterogeneous “core” oligosaccharide structure. J . B i d . Chem. 248, 7938-7950. Fumahara, Y . , and Nikaido, H.( 1980).Asymmetric localization of lipopolysaccharides on the outer membrane of Salmonella typhimurium. J . Brccteriol. 141, 1463-1465. Galanos, C. ( 1975). Physical state and biological activity of lipopolysaccharides. Toxicity and immunogenicity of the lipid A component. Z . Immunitcretsfi,rsch., Exp. Klin. Immunol. 149, 2 14-229. Galanos, C. (1979). Biological properties of lipopolysaccharides (LPS) and lipid A . Toxicon 17, Suppl. I , 53. Galanos, C . , and Liideritz, 0. (1975). Electrodialysis of lipopolysaccharides and their conversion to uniform salt forms. Eur. J. Bioc,hem. 54, 603-610. Galanos, C., and Luderitz, 0. (1976). The role of the physical state of lipopolysaccharides in the interaction with complement. High molecular weight as prerequisite for the expression of anti-complementary activity. Eur. J . Biocham. 65, 403-408. Galanos. C., and Luderitz, 0. (1982). Isolation and purification of R-form lipopolysaccharides: Further applications of the PCP method. Methods Carbohydr. Chem. (in press). Galanos, C . , Liideritz, 0.. and Westphal, 0. (1969). A new method for the extraction of R lipopolysaccharides. Errr. J . Biochem. 9, 245-249. Galanos, C . , Rietschel, E. T . , Liideritz, O . , and Westphal, 0. (1971a). Interaction of lipopolysaccharides and lipid A with complement. Eur. J . Biochem. 19, 143-152. Galanos, C . , Liideritz, O . , and Westphal, 0. (1971b). Preparation and properties of antisera against the lipid A component of bacterial lipopolysaccharides. Eur. J . Biochem. 24, 116-122. Galanos, C . , Freudenberg, M., Hase, S . , Jay, F., and Ruschmann, E. (1977a). Biological activities and immunological properties of lipid A. I n “Microbiology-1977 ” (D. Schlessinger, ed.), pp. 269-276. Am. Soc. Microbiol., Washington, D. C. Galanos, C . , Liideritz, O., Rietschel, E. T., and Westphal, 0. (1977b). Newer aspects of the chemistry and biology of bacterial lipopolysaccharides, with special reference to their lipid A component. Int. Rcv. Biochem. 14, 239-335. Galanos, C., Roppel, J . , Weckesser, J . , Rietschel, E. T . , and Mayer, H. ( 1 9 7 7 ~ ) .Biological activities of lipopolysaccharides and lipid A from Rhodospirillaceae. Infect. Immun. 16, 407-412. Galanos, C., Freudenberg, M. A., Liideritz, O., Rietschel, E. T., and Westphal, 0. (1979a). Chemical, physicochemical and biological properties of bacterial lipopolysaccharides. I n “Biomedical Applications of the Horseshoe Crab (Limulidae)” (E. Cohen, ed.), pp. 321 -332. Alan R . Liss, Inc., New York. Galanos, C.. Freudenberg, M. A , , and Reutter, W. (l979b). Galactosamine-induced sensitization to the lethal effects of endotoxin. Proc.. Natl. Acad. Sri. U.S.A. 76, 5939-5943. Galanos, C., Liideritz, O., and Westphal, 0. ( 1 9 7 9 ~ ) Preparation . and properties of a standardized lipopolysaccharide from Salmonella abortus eyui. Zentralbl. Baktrriol.. Parasitenkd.. Inkktionskr. H y g . , Abt. 1: OriK., Reihr A 243, 226-244. Gentner, N.. and Berg, P. (1971). Occurrence of a glycyl-lipopolysaccharidestructure in E. coli and its enzymatic formation from glycyl-tRNA. Abstr.. Am. Soc. B i d . C h m . Meet.. /97/ p. 100. Gerisch, G . , Liideritz, O., and Ruschmann, E. (1967). Antikorper fordern die Phagocytose von Bakterien durch Amoben. 2. Narutfimch., 5: Anorg. Chem. , Org. Chem., Biochem., Biophys. B i d . 22B, 109.

138


Gmeiner, J . (1975). Glyceraldehyde phosphate at the reducing terminus of Sukorzellu 0 haptens. Eur. J . Biochem. 51, 449-457. Gmeiner, J . ( 1977). The ribitol-phosphate-containing lipopolysaccharide from Proreus mircrbilis, strain D 52. Eur. J . Biochem. 74, 171-180. Cmeiner, J . , and Schlecht, S. (1979). Molecular organization of the outer membrane of Salmonella ryphrmurium. Eur. J . Biochem. 93, 609-620. Gmeiner, J . . and Schlecht, S. (1980). Molecular composition of the outer membrane of Escherichtu coli and the importance of protein-Iipopolysaccharide interactions. Arch. Micwhiol. 127, 8 1-86, Gmeiner, J . , Luderiti.. O . , and Westphal, 0. (1969). Biochemical studies on lipopolysaccharides of Salmonellu R mutants 6. Investigations on the structure of the lipid A component. Eur. J . Biochern. 7, 370-379. Gmeiner, J . , Simon, M., and Luderitz, 0. (1971). The linkage of phosphate group and of 2ketu-3-deoxyoctonatc to the lipid A component in a Sult~ionelluminnesoru lipopolysaccharide. Erir. J . Uiochem. 21, 355-356. Goldman. R . C., and Leive. L. (1980). Heterogeneity of antigenic side chain length in lipopolysaccharide from Eschoric~hiucoli 01 I I and Salmonella vphimurium LT2. Eur. J . Biochern. 107, 145-1 5 3 . Gorbach, V . 1.. Krasikova, I . N., Lukyanov. P. A,, Razmakhina, 0. Y . , Soloveva, T . F . , and Ovodov, Y . S. (1979). Structural studies on the immunodominant group of lipid A from lipopolysaccharide of Yrrsinicr pscudotubrrcrilosis. Eur. J . Bioc.hem. 98, 83-86. Corshkova, R . P . , Komandova, N . A , , Kalinovsky. A. I..and Ovodov. Y . S . (1980). Structural studies on the 0-specific polysaccharide side-chains of Yrrsiniu pseudotuhcri,ulosis, type 111. lipopolysaccharide. Eur. ./. Rinc.hem. 107, 13 1 - 13.5. Hammerling, G., Luderitz. 0.. Wcstphal.0.. and Makela, P. H. (1971). Structural investigationson the core polysaccharide of Eschuichia c d i 0100. Eur. J . Biochem. 22, 331-344. Hartlcy, J . L., Adams, G. A . , and Tornabene, T. G. (1974). Chemical and physical properties of LPS of Yer.rinitr pestis. J . Bacreriol. 118, 848-854. Hase. S. (1982). Determination of thc structure of the carbohydrate backbone of lipid A. Merhods Curhohydr. Chem. 9 (in press). Hase, S., and Rietschel, E. Th. (1977). The chemical structure of the lipid A component of lipopolysaccharides from Chromfjhac.tc,rirrm r'ioluceum NCTC 9694. Fur. J . Bioc,hem. 75, 23-34. Heasley, F. A. (198 I).Reducing terminus of 0-hapten accumulated in a Sultnonrllu montevideo galE mutant. J . Bacreriol. 145, 624-627. Heath, E. C., Mayer. R. M.. Edstrcim, R. D., and Beaudreau, C . A. (1966). Structure and biosynthesis of the ccll-wall lipopolysaccharide of Ewherichitr c,oli. A n n . N . Y . A u ~ d Sci. . 133, 315-333. Hellerqvist, C. G., and Lindberg, A. A . (1971). Structural studies of the common-core polysaccharide of the cell wall lipopolysaccharide from Salmoncllu fyhimuritrrn. Carhohydr. Rrs. 16, 39-48. Hellerqvist, C. G . , Lindberg, B., and Svensson, S . (1968). Structural studies on the 0-specific side-chains of the cell wall lipopolysaccharide from Salmonella ryphimurim 395 MS. Curhohydr. Res. 8, 43-55. Hellerqvist. C. G . , Lindberg, B., and Svensson, S. (1969a). Structural studies on the 0-specific side chains of the cell wall lipopolysaccharides from .Snlrnoncl/a ?phi and S. cnreriridis. A w r Chtzm. S w n d . 23, 1588-1596. Hellerqvist, C. C . , Lam. 0.. Lindherg, B., Holme, T., and Lindberg, A. A. (1969b). Structural studies on the 0-specific side chains of the cell wall lipopolysaccharide from Salrnonrlla hredenev. A(,tti Chem. Scund. 23, 2217-2222.


139

Hellerqvist, C. G . , Lindberg, B., Svensson, S . , Holme, T., and Lindberg, A. A. ( 1 9 6 9 ~ )Structural . studies o n the 0-specific side-chains of the cell-wall lipopolysaccharide from Salmonello tvphimurium LT2. Curhohydr. R r s . 9, 237-241. Hellerqvist, C. G., Lindberg. B., and Pilotti, A . (l970a). Structural studies on the 0-specific side-chains of the cell-wall lipopolysaccharide from Sulrnonella strci.\hourg. Actcr Chrm. Scrrnd. 24, 1168-1 174. Hellerqvist, C. G . , Lindberg, B., and Svensson. S . (1970b). Structural studies of the 0-specific side-chains of the cell-wall lipopolysaccharides from Sulmonrllo ncwport and Salmonellu k r w r w k , v . Cnrhohydr. RPS. 14, 17-26. Hellerqvist, C. G., Lindberg, B., and Lonngren, J. (1971a). Fragmentation analysis of’the SQ/monellu nc’ivporr lipopolysaccharide, A u t r Chmt. S(.crnd. 25, 60 1-606. Hellerqvist, C. G., Lindberg, B.. Liinngren. J . , and Lindberg, A. A. (1971b). Structural studies on the 0-specific side-chains of the cell-wall lipopolysaccharide from Sulmonella neLvington. A c m Chem. Srund. 25, 939;944. Hellerqvist, C . 0.. Lindberg, B.. Samuelsson, K.. and Lindberg, A. A. (1971~).Structural studies o n the 0-specific side-chains of the cell-wall lipopolysaccharide from Sdmoncdlrr pwufyphi A vrir. durcrzzo. A r ~ c rChem. Sccrnd. 25, 955-961, Hellerqvist, C. G., Hoffman. J . . Lindberg. B . , Pilotti, A , , and Lindberg, A. A. (1971d). Anomeric nature of the D-mannose residues in the Sdmonella t ~ p h iand S. srrtrshourg lipopolysaccharides. Acru Chrm. Scond. 25, 1512-1513. Hellerqvist, C. G.. Lindberg, B., Lonngren. J . . and Lindberg, A . A. (1971e). Structural studies of the 0-specific side-chains of’the cell-wall lipopolysaccharide from Strlmondltr muensfcr-. Curhohydr. R e s . 16, 289-296. Hellerqvist, C. G., Lindberg. B.. and Pilotti, A . (lY7lf). Structural studies of the 0-specific sidechains of the cell-wall lipopolysaccharide from Strlmonellu scnfiunhrrg. Curhohydr. Res. 16, 297-302. Hellerqvist, C. G . . Lindberg. B., Samuelsson, K.. and Brubaker, R . R. (1972a). Structural studies on the 0-specific side-chains of the cell-wall lipopolysaccharide from Pcrsfr~ure//rr p.rr~rdotuhrrc~rrlosi.s group I I A . Aivrr Chem. S w n d . 26, 1389- 1393. Hellerqvist, C. G.. Lindberg, B., Samuelsson. K.. and Bruhakcr, R . R . (1972b). Structural studie\ on the 0-specific side-chains o f the cell-wall lipopolysaccharide from Pmtc~ureii~r p.teudoruhrrr.ult,srs group 118.Ac./tr C‘hrm. S u n d . 26, 1394- 1398. . Hellerqvist, C. G . , Hoffman, J . , Lindberg, B.. Lindberg. A. A , , and Svensson, S . ( 1 9 7 2 ~ ) Sequence analysis of the polysaccharides from Strlrnonrlllr r i w p o r t and Snlmonelltr kentuck!. A c w C h m . .Sr.and. 26, 3282-3286. Henning. U. (1975). Determination of cell shape in bacteria. Arinu. Reis. Miuohiol. 29, 46-60. Henning, U . . and J a m , K. (1979). Two-component nature of bacteriophage T4 receptor activity in Esr.htrichiu c d i K- 12. J . B t i c f r r i o l . 137, 664-666. Hoffman, J., Lindberg, B . . and Brubaker. R. R. (1980a). Structural studies on the 0-specific side chains of the lipopolysaccharide from Yrrsiniu ~,rttrror.olitic.rr.Ctrrhohydr. R a s . 78, 2 12-2 14. Hoffman, J.. Lindberg, B., Glowacka. M., Dcrylo, M., and Lorkiewicz, 2. (1980b). Structural studies of the lipopolysaccharide from Strlmonelltr fyphirnurium 902 (Collb drd2). Eur. J . Rior.hem. 105, 103-107. Humphreys. G . 0.. Hancock. J . C.. and Meadow, P. M . (1972). Synthesis of hydroxy acids in lipid A of /‘.scudomonas trrruginostr. J . G m . Microhiol. 71, 22 1-230. Hurlbert, R . E . . and Hurlbert, I . M. (1977). Biological and physicochcmical properties of the lipopolysaccharide of Chromtrtium vinosum. Irifect. !mmun. 16, 983-994. Ikawa. M . , Koepfli. J . B . , Mudd. S . G.. and Niemann, C. (1953). An agent from E . coli causing hemorrhage and regression of an experimental mouse tumor. 111. The component fatty acids of the phospholipid moiety. J . A m . Chem. SOC. 75, 1035-1038.

140


Inage, M . , Chaki, H., Kusumoto. S.. and Shiba, T. (1981). Chemical synthesis of phosphorylated fundamental structure of lipid A. Tetrahydron Lett. (in press). Inage. M.. Chaki. H., Kusumoto, S., and Shiba, T . (1980a). Synthesis of liposaccharide corresponding to fundamental structure of Salmonella-type lipid A. Terruhedron Lert. 21, 3889-3892. Inage. M..Chaki, H., Kusumoto. S., Shiba, T., Tai, A , , Nakahata, M . , Harada, T . , and Izumi, Y. (l980b). Chemical synthesis of bisdephospho lipid A of Salmoneila endoro-rin. Chem. Lett. pp. 1373-1376. Inouye. M . (ed.) (1979). “Bacterial Outer Membranes. Biogenesis and Functions,” pp. 1-534. John Wilcy & Sons, New York. Iwashita. S., and Kanegasaki, S. (1973). Smooth specific phage adsorption: Endorhamnosidase activity of tail parts of P22. Biochem. Biophys. Res. Commun. 55, 403-409. Jann, B., Jann, K.. Schmidt, G . . Orskov, I., and Orskov, F. (1970). lrnmunochemical studies of polyaaccharide surface antigens of Escherirhia coli 0100:K?(B):H2. Eur. J . Biochem. 15, 29-39. Jann. B., J a m , K., Schmidt. G., Brskov, F., and Brskov, J. (1971). Comparative immunochernical studies of the surface antigens of E . i d i strains 08:K87(B?):H19 and (032):K87(B?):H45. Eur. J . Biochem. 23, 515-522. Jann, B.. Jann, K.. and Beyaert, G. 0. ( 1973). 2-Amino-2.6-dideoxy-~-glucose (D-quinovosamine): A constituent of the lipopolysaccharides of Vibrio cholerae. Eur. J . Biochem. 37, 531-534. J a m , B.. Reske. K., and Jann, K. (1975). Heterogeneity of lipopolysaccharides. Analysis of pcdysdccharide chain length by sodium dodecylsulfate-polyacrylamide gel electrophoresis. Eur. J . Riochem. 60,239-246. Jann, B., Prehm. P., and Jann, K. (1978). Cirrohacter 0-antigens: Structure of the 0-antigenic polysaccharide from Citrobucter sp. 396. J . Bacteriol. 134, 462-469. Jann. K.. and J a m , B. (1977). Bacterial polysaccharide antigens. In “Surface Carbohydrates of the Prokaryotic Cell” (1. Sutherland,ed.), pp. 247-287. Academic Press, New York. Jann, K., and Jann, B . (1978). Bacterial polysaccharides. In “Methods in Enzymology” (V. Ginsburg, ed.), Vol. SO, Part C, pp. 251-272. Academic Press, New York. Jann, K., and Westphal, 0. (1975). Microbial polysaccharides. I n “The Antigens’’ (M. Sela, ed.) Vol. 3 , pp. 1-125. Academic Press, New York. Jann, K.. Jann, B., Brskov, F., and @rskov, I. (1966). Immunchemische Untersuchungen an K-Antigenen von E. coli. 111. lsolierung und Untersuchung der chemischen Struktur des sauren Polysaccharids aus E . coli 0141:K85(B):H4 (K84 Antigen). Bioehem. Z . 346, 368-385. J a m , K., Kanegasaki, S., Goldemann. G . , and Mlkela, P. H. (1979). On the effect of rfe mutation of the biosynthesis of the 08 and 09 antigens of E . r o l i . Biochem. Riophys. Res. Commun.86, 1185-1191. Jansson, P.-E., Lindberg, B., Bruse, G., Lindberg, A. A , , and Wollin, R. (1977). Structural studies on the hexose region of the lipopolysaccharide from Escherichia i v l i C . Corbohydr. Res. 54, 261-267. Jansson, P.-E., Lindberg, B., Lindberg, A. A . , and Wollin, R . (1979). Structural studies on the hexose region of the Ettterr~hacreriaccaetype R3 core polysaccharide. Carhohydr. Res. 68, 385-389. Jansson. P.-E., Lindberg, A. A . , Lindberg, B., and Wollin, R. (1981). Structural studies on the hexose region of the core in lipopolysaccharides from Enterobacteriaceae. Eur. J . Biochem. 115,571-577. Jantzen, E., Byrn, K . , Bergan, T., and B ~ v r e K. , (1975). Gas chromatography of bacterial whole cell methanolysate. VII. Fatty acid composition of Acinerobacrer in relation to the taxonomy of Neisseriaceae. Acra Pathol. Microbiol. Scand. Microbiol. Sect. B: 83, 569-580. Jantzen. E.. Bryn. K., and Bpvre, K. (1976). Cellular monosaccharide patterns of Neisseriaceae. Acra Pathol. Microbiol. Scand., Sect. B: Microbiol. 84B, 177-1 88.


141

Jay, F. A . (1978). Gram-negative lipopolysaccharide. Characterization of the immunodeterminant structure of lipid A . Ph.D. Thesis, 1978. University of Surrey, Guilford, England. Jennings, H. J . , and Smith, 1. G. (1980). Determination of polysaccharide structure with 13C NMR. Methods Carbohydr. Chem. 8. 97-105. Jirillo, E., and Furnarola, D. (1979). Properties and mechanisms of action of lymphokines (LK): Current state of studies of mediators of cellular immunity. Ann. Selavo 21, 573-620. Johnson, K. J.. Ward, P. A.. Goralnick. S., and Osborn, M. J. (1977). Isolation from human serum of an inactivator of bacterial Iipopolysaccharide. Am. J . Pathol. 88, 559-574. Johnston, J . H., Johnston, R. J., and Simmons, D. A. R. (1967). The irnmunochemistry of Shigellu Jrxneri 0-antigens. The biochemical basis of smooth to rough mutation. Eiochem. J . 105, 79-87, Jones, N. C., and Osborn, M . J . (1977a). Interaction of Sulmonellu fyphimurium with phospholipid vesicles. J . E i o l . C h w . 252, 7398-7404. Jones, N. C . , and Osborn, M . J . (1977b). Translocation of phospholipids between the outer and inner membranes of Salmunrllu fyphimurium. J . Eiol. Chem. 252, 7405-7412. Josephson, S . , and Bundle, D. R. (1979). Artificial carbohydrate antigens: The synthesis of the tetrasaccharide repeating unit of Shigellajlexneri 0 antigen. Can. J . Chem. 57, 3073-3097. Kabir, S . . Rosenstreich, D. L., and Mergenhagen, S. E. (1978). Bacterial endotoxins and cell membranes. I n “Bacterial Toxins and Cell Membranes” (J. Jelijaszewicz and T. Wadstriim, eds.), pp. 59-87. Academic Press, New York. Kadis, S . , Weinbaurn, G . , and Ajl, S. J . eds. (1971). “Microbial Toxins,” Vol. 5. Academic Press, New York. Kanegasaki, S . . and Jann, K . (1979). Demonstration by membrane reconstitution of a butanolsoluble intermediate in the biosynthesis of the 09 antigen of Escherichia coli. Eur. J . Biochrm. 95, 287-293. Kass, E. H., and Wolff, S. M., eds. (1973). “Bacterial Lipopolysaccharides, the Chemistry, Biology and Clinical Significance of Endotoxins. Univ. of Chicago Press, Chicago, Illinois. Katzenellenbogen, E . , and Romanowska, E. ( 1980). Structural studies on Shigellaflexneri serotype 6 core region. Eur. J . Eiochem. 113, 205-21 1 . Kauffmann, F. (1956). A new antigen of Sulmon~llupurury~~hi B and SalmotieNu typhimurium. Actu Purhol. Microhiol. Siund. 39, 299-304. Kawahara, K., Uchida, K., and Aida, K. (1979). Direct hydroxylation in the biosynthesis of hydroxy fatty acid in lipid A of Pseudomonus ovulis. Eiochim. Eiophys. Acra 572, 1-8. Kenne, L . , Lindberg, B., and Petersson. K . (1977a). Basic structure of the oligosaccharide repeating-unit of the Shigel/(/,fTexnrri0-antigens. Curbohydr. Res. 56, 363-370. Kenne, L., Lindberg, B., Petersson, K., Katzenellenbogen, E., and Romanowska, E. (l977b). Structural studies of the ShigrliaJ7rxnen‘ variant X , type 5a and type 5b 0-antigens. Eur. J . Eiochem. 76, 327-330. Keene, L., Lindberg, B., Peterson, K., Katzenellenbogen, E.. and Romanowska, E. (1978a). Structural studies of Shigellu flexneri 0-antigens. Eur. J . Eiochem. 91, 279-284. Kenne, L.. Lindberg, B., and Unger, P. (1978b). Structural studies of the Vibrio cholerue 0-antigen. Curhohydr. Res. 68, C14-CI6. Kenne, L . , Lindberg, B., Petersson, K., Katzenellenhogen, E . , and Romanowska, E. (1980). Structural studies of the 0-specific side chains of the Shigelfu sonnei phase I lipopolysaccharide. Curbohydr. Rex. 78, 119-126. Kent, J. L., and Osborn, M. J. (1968). Properties of the 0-specific hapten formed i n vivo by mutant strains of Salmonellu typhimurium. Biochemistry 7, 4396-4408. Kiso, M . , Nishiguchi, H.,and Hasegawa, A. (1980). Application of ferric chloride-catalyzed glycosylation to a synthesis of glycolipids. Carhohydr. Res. 81, C13-C 15. Kiso, M . , Nishiguchi, H., Murase, S., and Hasegawa, A. (1981a). A convenient synthesis of 2”

142


deoxy- 2- (D- 3-hydroxy- tetradecanoyl-amino)-I)-glucose: diastereoisomers of the monomeric lipid A component of the bacterial lipopolysaccharide. Carhohvdr. Rrs. 88, C5-C9. Kiso, M.. Nishiguchi, H.. Nishihori. K . , and Hasegawa, A . (1981b). Synthesis o f p - ~ - ( 1 , 6linked ) An approach to the lipid A disaccharides of N-fatty acylated 2-amino-2-deoxy-o-glucose: component of the bacterial lipopolysaccharide. Carhohydr. Res. 88, CIO-C13. Kleine, 8. [ 1981). Das Schicksal von baktaricllem Lipopolysaccharid in v i w . Untersuchungen zur .Qusscheidung von Lipopolysaccharid durch Meerkatzen und Ratten. Ph.D. Thesis, Universitat Freiburg. Knirel. Y. A.. Shashkov, A . S., Dmitriev. B. A , , Kochetkov. N. K . . Kasyanchuk, N. V.. and Zakharova. I . Y . ( 1980). Antigenic polysacharides of bacteria. I I, The structure and C- 13 NMR-spectrum of 0-specific polysaccharide from Pseudomonfrs r,rpucitr. Bioorg. Khim. 6 , 1851-1859.

Kochetkov, N . K.. and Chizhov, 0 . S . (1966). Mass spectromctry of carbohydrate derivatives. Ad\. Curhohydr. (‘hem. 21, 39-95. Kochetkov. N . K.,Dmitriev, B. A.. and Lvov, V . L. (1977). 4-O-(R)-l-carboxyethyl-~-glucose: A new acidic sugar from Shigellri dyscvzteriac. type 3 lipopolysaccharide. Carhohyrlr. Res. 54. 253-259. Koenig, S . , and Hoffmann. M. K. (1979). Bacterial lipopolysaccharide activates suppressor B lymphocytes. Pror. Nut/. Acrid. Sci. U . S . A . 76, 4608-4612. Kontrohr, T.,and Kocsis, B. (1978). Structure of the hexose region of Shigella sonnei phase 11. Lipopolysaccharide with 3-deoxy-D-munno-octulosonic acid as possible immunodeterminant and its relation to Eschcrichin coli R , core. Eur. ./. Biochem. 88, 267-273. Kontrohr. T.. and Kocsis. B. (1981). Isolation of adenosine 5’-diphosphate-~-glycero-~-mannoheptose. J . B i d . Chem. 256, 7715-7717. Kopecko, D. J . , Washington, O., and Formal. S. B. (1980). Genetic and physical evidence for plasmid control of Shigellrr sonnei form 1 cell surface antigen. Infect. Immun. 29, 207-214. Kotelko, K., Gromska, W., Papierz. M., Swr, K., Krajewska, D., and Sidorczyk, 2. (1974). The constitution of “core” in Proreus lipopolysaccharides. J . Hyg., Epidemiol., Microbiol., Immuno/. 18,405-410. Kreutzer. D. R., Buller, C. S . , and Robertson, D. C. (1979). Chemical characterization and biological properties of lipopolysaccharides isolated from smooth and rough strains of BruceNa abortus. Infect. Immun. 23, 81 1-818. Lehmann, V. (1977). Isolation, purification and properties of an intermediate in 3deoxy-D-manrro-octulosonic acid-lipid A hiosynthesis. Eur. J . Biochem. 75, 257-266. Lehmann, V., and Rupprecht, E. (1977). Microheterogeneity in lipid A demonstrated by a new intermediate in the biosynthesis of 3-deoxy-~-manno-octuloson~c acid-lipid A. E u r . J . Biochem. 81, 443-452. Lehmann, V . . Hammerling, G . , Nurminen, M., Minner, I . , Ruschmann, E., and Liideritz, 0. (1973). A new class of heptose-defective mutants of Sulmonella ryphimurium. Eur. J . Biochem. 32, 268-275. Lehmann, V . , Rupprecht, E., and Osborn, M. J. (1977). Isolation of mutants conditionally blocked in the biosynthesis of the 3-deoxy-~-munnu-octulosonic-acid-lipid A part of lipopolysaccharides derived from Salmonella typhimurium. Eur. J . Biochem. 76, 41 -49. Lehmann, V.,Redmond, J., Egan, A , , and Minner, 1. (1978). The acceptor for polar head groups of the lipid A component of Salmunellrc lipopolysaccharides. E u r . J . Biochem. 86, 487-496. Lehmann, V., Streck, H., Minner, I., Krammer, P. H., and Ruschmann, E. (1980). Selection of bacterial mutants from SalmoneIIa specifically recognizing determinants on the cell surface of activated T lymphocytes. A novel system to define cell surface structures. Eur. J . Immunol. 10, 685-693.


143

Leive, L. (1977). Domains involving nonrandom distribution of lipopolysaccharide in the outer membrane o f Escherichiir coli. /‘roc. Not/. A w d . Sci. U . S . A . 74, 5065-5068. Lindberg, A . A. (1977). Bacterial surface carbohydrates and bacteriophage adsorption. In “Surface Carbohydrates o f the Prokaryotic Cell” ( I . W. Sutherland, ed.). pp. 289-356. Academic Press, New York. Lindberg, B . (1979). Methods for the structural elucidation of complex carbohydrates. I n “Glycoconjugate Research” ( J . D. Gregory and R. W. Jeanloz, eds.). Vol. I , pp. 17-34. Academic Press. New York. Lindberg, B., Liinngren, J. Ruden, U . . and Simmons, D. A. R. (1973). A reinvestigation of Shigdlu flexnrri 0-antigens by methylation analysis. Eur. ./. Biochem. 32, 15- 18. Lindberg, B.. Lbnngren. J., and Svensson, S. ( 1975). Specific degradation of polysaccharides. A h . Carhohydr. Chern. Biochem. 31, 185-240. Liideritz, O . , Simmons. D. A. R . , and Westphal, 0. (1965). The immunochemistry of Sulmonellu chemotype VI 0-antigens. The structure of oligosaccharides from Snlmnnella group U (0 43) lipopolysaccharides. Biochefn. J . 97, 820-826. Liideritz, O . , Galanos, C.. Risse, H. J . , Ruschmann, E . , Schlecht, S . . Schmidt, G . . SchulteHolthausen, H . , Wheat, R.. Westphal. 0 . .and Schlosshardt, J. (1966a). Structural relationships of Sirlmont4o 0 and R antigens. Ann. N . Y . Acrrd. Sci. 133, 349-374. Liideritz, O., Staub, A . M., and Westphal, 0. (1966b). lmmunochemistry of 0 and R antigens of Sulmonella and related Enterobacteriaceae. Bucfrriol. Rcv. 30, 192-255. Luderitz, 0 . . DrDge, W . . Mhhlradt, P.. Ruschmann, E., and Westphal. 0. (1969). Biochemical investigations on the core of Salmonella lipopolysaccharides. Colloq. Int. C.N.R.S. 174, 95-107. Ldderitz, 0.. Westphal, O . , Staub, A. M., and Nikaido, H. (1971). Isolation and chemical and immunological characterization of bacterial Iipopolysaccharides. I n “Microbial Toxins” (G. Weinbaum, S. Kadis, and S. J . Ajl, eds.), Vol. 4, pp. 145-233. Academic Press, New York. Liideritz, O., Galanos, C., Lehmann, V., Nurminen, M . , Rietschel, E. T., Rosenfelder. G . . Simon. M., and Westphal, 0. (1973). Lipid A: Chemical structure and biological activity. J . Infr.cf. Dis. 128, Suppl. S17-S29. Lhderitz. O . , Galanos, C . , Lehmann, V . , Mayer, H . , Rietschel, E. T., and Weckesser, J . (1978). Chemical structure and biological activities of lipid A’s from various bacterial families. Nufurwissen.st~huffrn 65, 578-585. Lugowski. C . , and Romanowska, E. (1974). Biological properties of lipid A from ShigeNu wnnri. E u r . J . Biochw?. 48, 81-87. Lvov, V. L., Tochtamysheva, N. V . , Dmitriev. B. A., Kochetkov, N. K., and Hofman, I . L. (1980). Bacterial antigenic polysaccharides. 10. The structure of polysacciiaride chain of Shigella boydii type 4 lipopolyaaccharide. Bioorg. Khim. 6, 1842- 1850. McCabe, W. R . , Bruins, S. C., Graven, D. E., and Johns. M. (1977). Cross-reactive antigens: their potential for immunization-induced immunity to gram-negative bacteria. J . Infect. Dis. 136, I6 I - I 66. McChee, J . R., Kiyono, H., Miehalek, S . M., Bahb, J . L., Rosenstreich. D. L., and Mergenhagen, S. E. (1980). Lipopolysaccharide (LPS) regulation of the immune response: T lymphocytes from normal mice suppress mitogenic and immunogenic responses to LPS. J . Immunol. 124, 1603- 161 1 . Mikela. P. H.. and Mayer, H. (1976). The enterobacterial common antigen (ECA). Bucwriol. Rev. 40, 591-632. Makela. P. H . , and Stocker, B. A. D. (1981). Genetics ofthe bacterial cell surface. Symp. So(,. Gen. Microbial. 31 (in press). Malchow, D., Luderitz, 0.. Kickhbfen, 8..Westphal, 0.. and Gerisch, G. (1969). Polysaccharides in vegetative and aggregation-competent amoebae of Dirfymfelium di,rc.oideum. 2. Purification

144


and characterization of amoebae-degraded bacterial polysaccharides. Eur. J . Biochem. 7, 239-246. Mathison, J C., and Ulevitch, R. J . (1979). The clearance, tissue distribution, and cellular localization of intravenously injected lipopolysaccharide in rabbits. J . Immunol. 123, 21 33-2143. Mattsby-Baltzer, I . , and Kaijser, B . (1979). Lipid A and anti-lipid A. Infecr. Immun. 23, 758-763. Mayer, E. P., and Teodorescu, M. (1980). Bacterial adherence tn eucaryotic cells: Isolation of lymphocyte-binding mutants. Infect. Immun. 29, 66-69. Mayer, H . , and Schmidt, G. (1979). Chemistry and biology of the enterobacterial commnn antigen (ECA). Curr. Top. Micwhiul. Irnmunol. 85, 99-153. Mayer, H.. Rapin, A. M. C., Schmidt, G . , and Boman, H . G. (1976). Imrnunochemical studies on lipopolysaccharides from wild-type and mutants of Escherichiu tuli K- 12. Eur. J . Biochem. 66, 357-368. Melberg, S . (1979). Ph.D. Thesis, Technical University of Denmark, Copenhagen. Melberg, S . , and Rasmussen, K . (1979). Conformations of disaccharides by empirical force-field calculations. Part 1II.P-cellobiose. Curhohydr. Res. 71, 25-34. Melchers. F. (1980). Interactions of bacteria with the immune system. Lifi, S1.i. Res. R q . 6 , 285-305. Menkin. V . (1953). Modem views of inflammation. Int. Arch. Allerxy Appi. Immunol. 4, 131-16. Morgan, W. T. J . . and Partridge, S . M. (1941). Studies in immunochemistry. 6. The use of phenol and of alkali in the degradation of antigenic material isolated from Bacterium dyvsenreriur (Shigu). Biochem. J . 35. 1140-63. Morrison, D. C., and Ulevitch, R. J . (1978). The effects of bacterial endotoxins on host mediation systems. AAM. J . Pothol. 93, 526-618. Morton, J . J . . and Stewart, J. C. (1972). Structural studies on the lipopolysaccharide of a rough strain of Escherichiu c d i . Eur. 1. Biochem. 29, 308-318. Miihlradt, P. F . , Menzel, J . . Golecki, J . R., and Speth, V. (1973). Outer membrane of Salmonella. Sites of export of newly synthesized lipopolysaccharide on the bacterial surface. Eur. J . Biochem. 35, 47 1-48 1 . Muhlradt, P. F . , Menzel, J . , Golecki, J . R., and Speth, V . (1974). Lateral mobility and surface density of lipopolysaccharide in the outer membrane of Salmonella typhimurium. Eur. J . Biochem. 43, 533-539. Miihlradt. P. F . , Wray, V . , and Lehmann, V. (1977). A :"P-nuclear-magnetic-resonance study of the phosphate groups in lipopolysaccharide and lipid A from Sulmonrllu. Eur. J . Biochem. 81, 193-203. Miiller-Seitz, E., Jann, B . , and Jann, K . (1968). Degradation studies on the lipopolysaccharide from E . coli 071:K'?:H 12. Separation and investigation of 0-specific and core polysaccharides. FEES Lert. 1, 31 1-314. Munson. R . S . . Rasmussen, N. S . , and Osborn. M. J . (1978). Enzymatic incorporation of 3-deoxy-~-manno-octulosonate into a precursor of lipid A in Salmonella typhimurirrm. J . B i d . Chrm. 253, 1503-1511. Nashed. M. A., and Anderson, L. (1981). A convergent and flexible approach to the synthesis of enterobacterial lipid A. Fully substituted disaccharides having palmitoyl as fatty acyl moiety. Curhohydrut(JRrs. 92, C5-C9. Neal, D. J . . and Wilkinson, St. G . (1979). Lipopolysaccharides from Pseud~~monas rnultophiliu: Structural studies of the side-chain polysaccharide from strain N.C.T.C. 10257. Carhohvdr. Re.). 69, 191-201. Nghiem, H . O . , and Staub, A. M. (1975). Molecular immunological heterogeneity of the Solmoneilu iuerich ( I ,9.23,(46),27) cell-wall polysaccharides. Curbohydr. Res. 40, 153- 169. Nikaido, H . (1970). Lipopolysaccharide in the taxonomy of gram-negative bacteria. Inr. J . Sysr. Bacreriol. 20, 383-406.


145

Nikaido, H. (1979). Die Permeabilitat der lusseren Bakterienmembran. Angew. Chem. 91,394-407. Nikdido, H . , and Nakae, T. ( I 979). The outer membrane of gram-negative bacteria. Adv. Micruh. Physiol. 20, 163-250. Nishijima, M. (1980). Isolation and characterization of two lipid-A precursors in a phosphatidylglycerol deficient mutant of Eschericltia coli. Fed. Proc.. Fed. A m . SOC. Exp. B i d . 39, 1982. Nishijima, M., and Raetz. C. R. H. (1979). Membrane lipid biogenesis in Escherichiu coli: Identification of genetic loci for phosphatidylglycerophosphate synthetase and construction of mutants lacking phosphatidylglycerol. J . B i d . Chenr. 254, 7837-7844. Nishijima, M . , Bulawa, C. E.. and Raetz, C. R. H. (1981). Two interacting mutations causing temperature-sensitive phosphatidylglycerol synthesis in E . coli membranes. J . Bucteriol. 145, 113-121. Nixdorff, K. K . , and Schlecht. S. (1972). Heterogeneity of the hemagglutinin responses to Salmonella minnesoru R-antigens in rabbits. J . G e n . Microbiol. 71, 425-440. Nixdorff, K. K . , Schlecht, S., Rude, E., and Westphal, 0. (1975). Immunological responses to Salmonella R antigens. The bacterial cell and the protein edestin as carriers for R oligosaccharide determinants. Immunology 29, 87- 102. Nowotny, A. (1971). Chemical and biological heterogeneity of endotoxins. In “Microbial Toxins” (C. Weinbaum, S. Kadis, and S. J . Ajl,eds.), Vol. 4, pp. 309-329. Academic Press, New York. Ohkawa, T. (1980). On the structure of the lipopolysaccharide core in the cell wall of Escherirhia coli K-12 W2252-1 IU- and its Ter-mutant cells. Biochem. Biophys. Res. Commun. 95, 938-944. Brskov, I., 0rskov. F., Jann, B., and Jann, K. (1977). Serology, chemistry, and genetics of 0 and K antigens of Escherichiu c d i . Bucreriol. Rev. 41, 667-7 10. Osbom, M. J . (1966). Biosynthesis and structure of the core region of the lipopolysaccharide in Sulmonella ryphimurium. Ann. N . Y . Acad. Sci. 133, 375-383. Osbom, M. J . (1979). Biosynthesis and assembly of lipopolysaccharide of the outer membrane. In “Bacterial Outer Membranes, Biogenesis and Functions” (M. Inouye, ed.), pp. 15-34. John Wiley & Sons, New York. Osborn, M. J . , and Rothfield, L. I. ( I97 I ). Biosynthesis of the core region of lipopolysaccharides. In “Microbial Toxins” (G. Weinbaum, S. Kadis, and S. J. Ajil, eds.), Vol. 4, pp. 331-350. Academic Press, New York. Osborn, M. J . , Gander, J . E., Parisi, E., and Carson, J . (1972a). Mechanism of assembly of the outer membrane of Sulmonelln typlrimurium. J . Biol. Chem. 247, 3962-3972. Osborn, M. J . , Gander, J . E., Parisi, E., and Carson, J . (1972b). Mechanism of assembly of the outer membrane of Sulmonc4u /yphimurium. Site of synthesis of lipopolysaccharides. J . Biol. Chem. 247, 3973-3986. Osbom, M. J . , Rick, P. D., Lehmann, V., Rupprecht, E., and Singh, M. (1974). Structure and biogenesis of the cell envelope of gram-negative bacteria. Ann. N.Y.Arud. Sci. 235, 52-65. Osborn, M. J., Rick, P. D., and Rasmussen, N. S. (1980). Mechanism of assembly of the outer membrane of SalmonrIlu ryphimurium. Translocation and integration of an incomplete mutant lipid A into the outer membrane. J . B i d . Clzem. 255, 4246-4251. Overdth, P., Raufuss, E. M., Stoffel. W., and Ecker, W. (1967). The induction of the enzymes of fatty acid degradation in Escherichia coli. Bioch PC > PE. The biological relevance of these observations is complicated, though, by studies demonstrating that in mixtures of lipids that show a single broad transition, there does not appear to be a preferential association of choles-

284

DONALD C. MELCHIOR

terol with one lipid species over another (Calhoun and Shipley, 1979; Lange et ul., 1979). Because of a potential for preferential association, the presence of cholesterol may promote lipid heterogeneity within a bilayer by producing patches of cholesterol-rich regions containing specific lipids. Crystallization would not occur in such patches, and lipid-protein association might be altered. Not all types of fluidity-modulating lipids present in prokaryotes show the characteristic plasticizing effect of cholesterol. The effects of sterols, hopanoids, and carotenoids in microbial membranes are discussed in detail by Ourisson and Rohmer and by Razin in this volume.

V.

PATCHING

A. Fluid Bilayers Demixing of phospholipids is the rule during crystallization in bilayers, and although in some cases immiscibility or phase separation occurs in the solid state at low temperature (Fig. 4), such temperature-dependent demixing is the result of the crystallization process. In the liquid state, at temperatures above any orderdisorder transitions, the lipids in systems such as those illustrated in Figs. 3 and 4 are miscible in all proportions. Although transient associations no doubt occur between different lipids of miscible fluid systems, there are no actual phase separations into liquid domains. In membrane bilayers, a more permanent heterogeneity in the liquid state may be introduced by a variety of factors, including the interaction of cholesterol with specific phospholipid classes (Demel et a / . , 1977), by intrinsic protein immersed in the bilayer (Boggs et a / . , 1977), and by electrostatic binding of acidic phospholipids by ions and proteins (Galla and Sackman, 1975; Birrell and Griffith, 1976; Ohnishi and Ito, 1974; Papahadjopoulos et a / . , 1975a,b). However, at least two cases have been reported of separation of two liquid phases within bilayers, where phase separation is a result of the intrinsic properties of the lipids and not due to an added extrinsic factor. In one, fluid regions containing between 25 and 40% lecithin are formed in a fluid phosphatidic acid bilayer in the absence of ions (Galla and Sackman, 1975). In the second system a phase diagram has been worked out (Wu and McConnell, 1975). At temperatures higher than 50"C, the dielaidoyl lecithin-dipalmitoyl phosphatidylethanolamine binary system displays the behavior characteristic of partial immiscibility of liquids, such as butanol-water or phenol-water. Although both the lipids and the temperatures at which liquid-liquid phase separation occurs are rather unnatural in the dielaidoyl lecithin-dipalmitoyl phosphatidylethanolamine system, it is possible that the effect may take place in membranes. A puzzling effect occurs, for example, in the endoplasmic reticulum of the protozoan 7etrahjvnena pyrijorrnis (Wunderlich et al., 1975). At temperatures below 17"C, freeze-etch electron microscopy reveals the emergence of

285

MEMBRANE PHASE TRANSITIONS

smooth patches on the fracture faces. This change at about 17°C is accompanied by changes in the fluorescent intensity of 8-anilino-1-napthalene sulfonate, the motion of spin-labeled stearic acid, the partition of 4-doxyldecane, and the amplitude of the NMR signals arising from hydrocarbon chains. However, no hint of the usual order-disorder crystallization could be found by scanning calorimetry. Two environments of different fluidity were suggested, both by the physical studies and by the freeze-etch electron microscopy. An instance of separate coexisting liquid lipid phases occurs in the cholesterol-containing membrane of Mycoplusmu cupricolum (Melchior and Rottem, 1981, 1982). This prokaryote, in addition to several other mycoplasmas, has a membrane rich in long-chain cholesterol esters (CE) (Rottem, 1980). To understand the physical state of these membrane CE, it will be useful to discuss the interaction of CE, phospholipids, and cholesterol in terms of general phase behavior. Long-chain CE have a solubility in lipid bilayers of less than 5 mol%. When present in excess of this, they form a separate phase (Small, 1970). The relatively complex miscibility behavior of phospholipids, cholesterol, and cholesterol esters can be described by the type of phase diagram shown in Fig. 7 .

WATER

CE

I PHASE crvstal

I1 I PHASE. oily liquid

80

60

40

20

FIG. 7 . The three-component system of egg PCiCholicholesteryl linoleate at constant water content. The tetrahedron at the upper left shows the position of the section containing the fourcomponent system with 70% water by weight. This section is shown enlarged and is dealt with as the three-component system PCICholICE. In this illustration the three apexes are labeled PL for the phospholipid (egg phosphatidylcholine), C for cholesterol, and CE for cholesterol ester. Region I consists of one phase, PC bilayers containing varying amounts of Chol and CE (shown schematically in the upper right). Region I I is an oily CE phasc containing up to 8 weight percent Chol. In region 111 two phases are present, PC bilayers saturated with CE and Chol and an oily phase of CE. Region IV contains three invariant phases, PC bilayers saturated with CE and Chol, an oily CE phase saturated with Chol, and Chol crystals. (From Small and Shipley, 1974.)

286

DONALD L. MELCHIOR

This figure illustrates the egg PC/cholesteroI/cholesteryl linoleate/water system at 37°C and atmospheric pressure (Small and Shipley, 1974). Using this formalism, a proper representation of a four-component system requires a tetrahedron as shown in the upper left corner. Since we are concerned with lipid systems i n excess water, a simplification can be made by taking a triangular section parallel to the base of the tetrahedron at 70 weight percent water. According to the Gibbs phase rule, at a given temperature and pressure, I, = c - p , where v is the degree of freedom of the system, c is the number of components comprising the system, and p is the number of phrases present in the system. In this example, since the water content is fixed, we have in effect a three-component system and v = 3 - p . As shown in Fig. 7, the egg PC/cholesteroVcholesteryl linoleate system at 70 weight percent water is divided into four major regions. Region I contains only one phase, the bilayer, and therefore v = 2 . The composition of these bilayers has two degrees of freedom. As indicated in the phase diagram, the bilayers can incorporate up to about 68 weight percent Chol and up to a few percent CE. In region 111, two phases are present and v = 1. One phase is PC bilayers saturated with CE and Chol, and the other phase is an oily CE phase. Region I1 has only two components and one phase and, therefore, v = 1 . This region consists of an oily CE phase containing up to 8 weight percent Chol. Region IV contains three phases, v = 0, and there are no degrees of freedom. The composition of each of these phases is fixed. The phases are PC bilayers saturated with Chol and CE, oily CE saturated with Chol, and crystalline cholesterol. Although CE are not very soluble in lipid bilayers, they are, as mentioned, found in substantial quantities in the membranes of several of the mycoplasmas (Razin et al., 1980). Recent studies have shown these CE to be tightly associated with the mycoplasma membrane, but not intimately associated with the bulk of the membrane protein. Using DSC, it was demonstrated that the majority of these CE exist as fluid patches or “pockets” coexisting with the Chol/phospholipid membrane bilayer (Melchior and Rottem, 1981, 1982). The ratios of CE, Chol, and phospholipids found in these mycoplasma membranes fall into region 111 of Fig. 7, which predicts the coexistence of two lipid phases as found experimentally. The fluid CE pockets may be located in the hydrophobic core of the membrane bilayer or may be attached to either side of the bilayer. The CE in these pockets appear to be relatively pure, since they can crystallize upon low-temperature incubation in a manner characteristic of pure CE (Small, 1970; Tall and Robinson, 1979).

6. Membrane Proteins Although crystallization of bilayers may occasionally have little or no effect upon the random distribution of proteins, as a general rule intrinsic membrane


287

proteins are frozen out of the advancing crystalline regions produced during membrane crystallization. This effect is seen by freeze-fracture electron microscopy as the appearance of patches nearly or entirely free of intramembrane particles, when membranes are incubated before quenching at temperatures within or below the transition. Figure 8 shows this phenomenon in the cytoplasmic membrane of E . c d i W3110 (van Heerikhuizen et a / . , 1975). These membranes are fluid at 37°C and crystalline at 0°C. When quenched from 37°C (Fig. 8A), they display a random distribution of particles, whereas when quenched from 0°C (Fig. 8B) they display patches of aggregated membrane particles and particle-free patches. The temperatures at which patching occurs, as well as the ratio of the areas of smooth to particulate regions, have been shown in E.coli to correlate roughly with the cytoplasmic membrane transition, although the onset of aggregation can occur below the high-temperature end of the transition (Schechter et a l . , 1974). The correlation of protein patching with the bilayer transition of the inner membrane of E.coli has been studied by various physical techniques in addition to X-ray diffraction (Kleeman et al., 1974; Haest et ul., 1974; Verkleij and Ververgaert, 1975). Temperature-induced patching of intramembrane particles has been reported in

FIG.8. The effect of temperature on particle distribution in the cytoplasmic membrane of E . co/i W3110. (A) Incubation of membranes at 37°C before freezing results in a random distribution of particles. Bar = 0.5 pm. ( B ) Incubation of membranes at 0°C before freezing produces extensive particle-free patches. Bar = 0.5 p m . (C) Cytoplasmic membrane vesicles with low and high particle density obtained by breakage of EDTA-lysozyme spheroplasts at 0-4°C in a Ribi press. The sample was equilibrated and frozen from 25°C to ensure a random distribution of particles in all vesicles. Bar = 0.2 p m . (From van Heerikhuizen et d., 1975.)

288

DONALD L. MELCHIOR

FIG. 88 and C.

(See legend p. 287)

the membranes of many prokaryotes in addition to the inner and outer membranes of E . coli (Schechter et al., 1974; van Heerikhuizen e t a / . , 1975; Verkleij et al., 1976), for example, membranes of S. fueculis (Tsien and Higgins, 1974; Haest et al., 1974), Mycwplasma mycoides subsp. cupri (Rottem et al., 1973a), A . laidlawii (Tourtellote et al., 1970; Verkleij el al., 1972), V . parvula and


289

A . lipolyticu (Verleij et a l . , 1975), and the blue-green alga, Anucysfis niduluns (Verwer e f a l . , 1978). Although proteins are probably most frequently displaced

from ordered regions of the bilayer by moving laterally and accumulating at high concentration in aggregated regions, particle-free patches can occur without an obvious increase in particle concentration in the remaining areas (Tsien and Higgins, 1974; Duppel and Dahl, 1976). In this case, evidently particles are removed from the fracture faces by moving normal to the membrane surface. Thus the most common effect of bilayer crystallization is to produce protein-free regions with conservation of proteins into concentrated regions or to produce patching with loss of particles. That patching is a result of the order-disorder transition, and not a result of temperature itself, has been demonstrated in A . laidlawii by growing cells in media enriched in saturated fatty acids at constant temperature (Tourtellotte et d.,1970). Membranes from palmitate-supplemented cells, which have a transition extending well above growth temperature, showed patching at 37"C, whereas cells grown in oleate with transitions well below 0°C showed no patching. The formation of clear patches in membranes can be induced by factors other than temperature, for example, pH (Copps et ul., 1976). Thermally induced patching ordinarily implies lateral phase separation and free diffusion of both lipids and intrinsic proteins. Presumably the particles seen by freeze-fracture are excluded from the more solid portions of the membrane (which consists of lipids enriched in saturated fatty acids) and collect in fluid pools of lower melting point lipids. Dramatic evidence for the correctness of this presumption has been provided by the actual physical separation of particledepleted regions from particle-enriched areas (Fig. 8 C ) . By mechanical disruption of spheroplasts of wild-type E . coli W31 10 cells at ice temperatures followed by isopycnic-gradient centrifugation, van Heerikhuizen e f d . (1 975) were able to isolate a low-density population of vesicles devoid of intramembranous particles. The phospholipid/protein ratio of the protein-depleted membranes was four or five times greater than that of whole cytoplasmic membranes, whereas the fatty acids of their lipids were considerably more saturated. One protein with an apparent molecular weight of 26,000 was concentrated in the low-density fraction, where it comprised 50% of the total protein. NADH oxidase and succinic dehydrogenase were excluded from the smooth patches, but D-lactate dehydrogenase was not excluded and even appeared to be concentrated. By using a gentler method, osmotic lysis at 4"C, both particle-enriched and particle-depleted vesicles have been isolated from cytoplasmic membranes of an E . coli fatty acid auxotroph grown on linolenic acid (Letellier and Schechter, 1976; Letellier et ul., 1977). X-Ray diffraction showed that the particle-rich membranes, whose lipids were greatly enriched in unsaturated fatty acids, crystallized at lower temperatures than the smooth membranes. The preference of some proteins for the solid phase is quite remarkable, and offers evidence for specific lipid-protein association. These preferences can be

290

DONALD L. MELCHIOR

demonstrated in model systems (Kleeman et al., 1974). In mixed bilayers containing dielaidoyl and dipalmitoyl lecithin, or dimyristoyl and distearoyl lecithin, the erythrocyte protein glycophorin prefers fluid regions. However, glycophorin remains randomly distributed in pure dimyristoyl or dipalmitoyl lecithin both above and below the crystallization temperature. Magnesium -calcium ATPase from rabbit sarcoplasmic reticulum, on the other hand, shows a more pronounced incompatibility with the solid phase. It is excluded from pure dimyristoyl lecithin bilayers at low temperatures. Apparently the partition coefficients of proteins in membranes can depend upon rather subtle changes in protein conformation; bleached rhodopsin is randomly distributed in solid dimyristoyl lecithin bilayers, but unbleached rhodopsin is excluded (Chen and Hubbell, 1973). Thilo, Traiible, and Overath have considered the functional consequence of proteins partitioning between fluid and crystalline regions of the membrane (Thilo et al., 1977; Overath and Thilo, 1978). Using protein partitioning as a model, they tested the temperature dependence of sugar transport in E . coli. This work was the outcome of careful studies on the transport rates of P-glucoside and P-galactoside in the E . coli fatty acid auxotroph T105, whose membrane transition was varied by supplementation with different fatty acids. The membrane transition was characterized by fluorescence, and careful effort was expended to measure transport rates above and below the transition. According to this model (Fig. 9), individual carrier molecules sense changes in the state of their lipid environment, which influences their rate of transport. Transport proteins partition between fluid- and ordered-membrane regions, and the overall rate of transport is the sum of the transport rates of the carriers in the fluid and ordered domains. Thus the transport proteins act as membrane probes and their overall activity shows the first appearance of fluid membrane regions and the final disappearance of crystalline membrane regions. Not all membranes having order-disorder transitions show intramembrane protein patching. This is the case for S. aureus, B . subtilis, Bacillus cereus, and Bacillus megaterium, whose lipids contain almost exclusively branched chains (Haest et u / . , 1974). Lack of protein patching is definitely correlated with the presence of branched chains because A . laidlawii (Haest et a l . , 1974) and E . coli (Legendre et a / ., 1980) membranes, which ordinarily display protein patching during the membrane transition, do not do so if enriched in branched-chain fatty acids. From X-ray diffraction studies (Haest et al., 1974; Lengendre et al., 1980) it appears that fatty acids in membranes rich in branched-chain fatty acids are more loosely packed in the crystalline state than in membranes lacking branched-chain fatty acids. This looser packing is suggested as the reason that membrane particles are not squeezed out of crystalline regions of membranes rich in branched-chain fatty acids. Studies using phospholipase A, on A . laidlawii enriched with various fatty acids are in agreement with these X-ray observations

291


I

f

4

LO"

30"

31

32

'VC

200

h z l

I

-II

.

I

I

33

(TEMPERATURE).'

34 x

35

36

lo3 [OK-']

Fic,. 9. The distribution of carrier proteins between tluid- and ordered-membrane regions. The course of the membrane transition in E . c d i fatty acid auxotroph T I 0 5 supplemented with trans.19-16:I fatty acids is shown at the top of the figure as the ratio of fluid t o total membrane area. The solid curves at the bottom of the figure arc calculated p-glucoside transport rates for different distribution constants, K . of carrier proteins panitioning between fluid and crystalline bilayer regions (shown schematically in the center of the figure). The open circles in the lower pan ofthe figure show the experimentally determined irr vii.o temperature dependence of P -nitrophenyl pwglucopyranoside (NphGlu) hydrolysis. The best fit between theory and experiment was obtained for K = IS. (From Thilo cr ( I / . , 1977.)

292

DONALD L. MELCHIOR

(Bouvier e t a / ., 1981 ;Op den Kamp, 1982). Phospholipase A2 has no access to PG in palmitate- and elaidate-enriched A . luidlawii membranes in the crystalline state due to the tight packing of the membrane lipids. In contrast, the presence of branched-chain fatty acids in A . laidlawii results in a sufficiently loose packing of lipids below the membrane phase transition, so that even in the crystalline state, phospholipase A2 is able to penetrate into the bilayer and hydrolyze PG. Consistent with the action of cholesterol as a bilayer plasticizer, temperaturedependent protein aggregation does not occur in membranes rich in cholesterol (Rottem er ul., 1973b; Duppel and Dahl, 1976). Mycoplasma mycoides ordinarily incorporates sufficient amounts of cholesterol into its membrane to eliminate the order-disorder transition, and membrane proteins remain randomly dispersed upon low-temperature incubation. If the organism is adapted to grow on low levels of cholesterol, so that its membrane is almost devoid of the sterol, the membrane crystallizes at 4°C and extensive patching takes place (Rottem el a / ., 1973b). It is worthwhile noting that patching does not always occur in wild-type E . c-oli (Kleeman and McConnell, 1974). Since the organism contains neither cholesterol nor branched-chain fatty acids-both of which, as described, inhibit patching-the lack of patching in these studies must have been a result of the fatty acid distribution. Freeze-fracture is thought to split the membrane bilayer along its midplane, so that freeze-fracture electron microscopy reveals only intrinsic proteins that are deeply embedded in the lipid matrix. Little is known of the distribution of surface proteins, or of proteins that have limited bilayer penetration, during bilayer crystallization. A study on A . luidlawii (Wallace and Engelman, 1978) suggests that although the distribution of exposed protein is affected by lateral phase separation, the spatial distribution of some surface proteins may respond differently than intrinsic membrane proteins to the order-disorder transition. Ferritin-labeling was used to visualize surface proteins of A . tuidluwii in the electron microscope and the order-disorder transition was characterized by X-ray diffraction. As expected, intrinsic proteins visualized by freeze-fracture were found to be dispersed above the membrane transition and to patch progressively as the temperature was lowered through the transition. In contrast, the ferritinlabeled proteins appeared to form patches only at temperatures partially within the transition.

VI.

BIOLOGICAL CONSEQUENCES OF MEMBRANE STATE

There is no doubt that thermotropic transitions have given considerable insight into both structure and function of biomembranes and will continue to do so. It is also tempting to speculate that thermotropic order-disorder transitions play a direct role in the life of the cell. If growth temperature normally were to coincide


293

with the temperature range of a transition, the membrane would exist in a partly fluid and partly crystalline state, and lateral phase separation might serve a physiological purpose. However, the information currently available indicates that this is not the case. Although bulk transitions can sometimes occur at growth temperatures, there is evidently no physiological necessity for them to do so. Completion of the membrane melt below growth temperature is common in gram-positive bacteria, such as M . lysodeikticus. Yersiniu enterocoliticu grown at 37°C has its membrane fully melted by 8°C (Abbas and Card, 1980). In B . stearothermophilus, a thermophilic bacterium, the membrane can be completely melted for at least 20°C below growth temperature, and in the extreme thermophile, T . uquaticus, melting is completed about 40°C below growth temperature (McElhaney and Souza, 1976; Melchior and Steim, 1976). Although the bulk melt in wild-type E . coli W945 sometimes is not finished until a few degrees above growth temperature (see Fig. 5a; it begins at about -2O"C), in many cases it is completed 5-10°C below growth temperature. Escherichia coli fatty acid auxotrophs can be forced to a transition that terminates 50°C below growth temperature (Baldassare et ul., 1976). In A . luidluwii cells the transition temperature can be profoundly shifted without affecting the temperature coefficients of growth (Tourtellotte, 1972; McElhaney, 1974), or absolute growth rates at the optimal growth temperature, provided that the transition is not high enough to occur at the temperature of growth. If the transition is too high, growth ceases. The lack of evidence for a unique physiological role of bulk transitions at the temperature of growth does not imply that such phenomena might not be important in specialized regions. Microcrystalline regions might exist, for instance, to a very limited extent even well above the bulk melt of the membrane, and related transitions might be triggered by ions, pH changes, and so forth (Trauble, 1971). But the bulk thermotropic transition and lateral phase separation, as seen by experimental methods now employed, appear to be unnecessary for the life of the cell at growth temperature. On the contrary, it is evidently an effect to be avoided. It is accompanied by a variety of usually undesirable physiological events, and it is clear that living systems take pains to lower their transition range to acceptable temperatures. Physiologically, membrane transitions reveal themselves most obviously by their effects on growth. Cells do not proliferate at temperatures below their transitions (Steim rt ul., 1969; Overath et ul., 1970; Tourtellotte, 1972; McElhaney, 1974; Petit and Edidin, 1974; Thilo and Overath, 1976). At temperatures below the order-disorder transition, where ordinarily fluid membranes are converted to the solid state, viscoelastic properties are drastically altered. In this condition, mechanical compliance is greatly reduced and cells can become osmotically fragile (Tourtellotte, 1972; van Zoelen et d.,1975). In the ordered state, the passive permeability barrier provided by the bilayer can lose its integrity. Leaks of erythritol and intracellular potassium are produced in E . cot! fatty

294

DONALD L. MELCHIOR

acid auxotrophs either by quickly quenching the cells to low temperature or by the mechanical stress of filtration at low temperatures (Haest et ul., 1972). The temperature at which leakage begins depends on the fatty acid composition of the membrane lipids. This behavior is mimicked by liposomes prepared from cell lipids. Similar results have been reported for the passive leakage of o-nitrophenyl galactoside (ONPG) into E. coli unsaturated fatty acid auxotrophs (Steim, 1972). On quick quenching, increased leakage of ONPG into cells invariably begins around the low-temperature end of the calorimetrically observed transition. Both the temperature of the order-disorder transition and its attendant leak can be varied by varying the exogenous fatty acids supplied to the cells. This increased leakage, which reveals itself as an upward swing in Arrhenius plots at low temperature, is not affected by inhibitors of active transport, and occurs to the same extent in cells induced for permease, uninduced cells, and mutants free of permease. ONPG passes into the cells, since there is no release of P-galactosidase activity into the incubation medium. Again the lesion is transient and disappears within a few minutes. The effect in K1060 cells supplemented with oleate is illustrated in Fig. 10. In curve a, the cells were held at room temperature before adding ONPG at lower temperatures; in curve b, the same preparation of cells was first incubated for 3 minutes at 0°C before adding ONPG at higher temperatures. The transient increased leakage, which begins at about 15"C, coincides with the low-temperature end of the calorimeter peak. The fact

"C 25

5

5

15

2-

I-

u) 2

0

w

1-

.5-

3.3

3.4

I/T

3.5

3.6

x lo3

FIG. 10. Passive leakage of ONPG as a function of temperature (T = K ) through membranes of E . coli auxotroph K1060 cells supplemented with oleate. On quick quenching, increased passive leakage of ONPG into cells begins near the low-temperature end of the calorimetric transition (curve a). Curve b represents the same preparation of cells first incubated for 3 minutes at 0°C before adding ONPG at higher temperatures. This increased leakage is most likely a result of microscopic fissures able to heal themselves on a time scale of minutes. (From Steim. 1972.)


295

that this leakage occurs only below the bulk transition emphasizes that barrier integrity is maintained by only a small proportion of fluidity. Such leakage is most likely a result of microscopic fissures that can be induced by low-frequency mechanical deformation or simply shrinkage upon cooling. The cracks, which are not large enough to permit the passage of large proteins, are able to heal themselves on a time scale of minutes because the ordered membranes are not perfectly rigid. Some insight into the structure of crystalline membranes at temperatures just below their order-disorder transitions have come from studies of pure dipalmitoyl lecithin bilayers by Lee (1977a), who points out the likelihood of grain boundary defects. Such boundaries, which are commonly recognized to occur in other crystalline solids (Ubbelohde, 1965), occur at interfaces between differently oriented crystal domains, and necessarily give rise to disorder in those regions (Lawaczeck et al., 1975). In bilayers, grain boundaries might be expected to provide sites for increased permeation of ions or small molecules, although it seems unlikely that they are responsible for the cell leakages described earlier in this section because such leaks are transient and increase as temperature falls more and more below the order-disorder transition. Although it is very likely that grain boundaries or other lattice defects occur in crystalline lipid bilayers, it is not clear how extensive they may be, especially in naturally occurring mixtures of lipids. Another kind of boundary effect may occur within the temperature range of membrane melting, where both fluid and solid phases are present. The boundaries in this case are not between domains in the solid state, but at the interface between the fluid and solid regions within the bilayer. In these interfacial regions, the arrangement of lipid molecules would differ from either crystalline or fluid regions. A number of model systems (Papahadjopoulos et a!., 1973; van Dijck et a / . , 1975; Marsh et al., 1976; Blok et af., 1975; Nichols and Miller, 1974) show an increased permeability to small molecules in the neighborhood of the order-disorder transition, as might be expected if lipid conformation were deranged at the liquid-solid interface. Permeability decreases again at temperatures above and below the region of melting. Since increases in permeability in the neighborhood of the order-disorder transition occur in a number of different model systems, the leakage may arise from the same cause in all cases. The details differ, however, from system to system. In dipalmitoyl phosphatidylglycerol liposomes, sodium leakage reaches a maximum rate at the midpoint of the bilayer phase transition (Papahadjopoulos et al., 1973), whereas in equimolar mixtures of dimyristoyl phosphatidylglycerol and dimyristoyl lecithin, leakage rates of potassium peak at the beginning of the transition. In dimyristoyl lecithin liposomes prepared by sonication, Tempo leakage peaks at the upper end of the transition (Marsh et u l . , 1976), which is broadened and lowered in temperature by sonication (Sheetz and Chan, 1972; Melchior and Steim, 1976; Faucon and Lusson, 1973). Elevated permeability during melting does not seem

296

DONALD L. MELCHIOR

to occur ordinarily in cells (McElhaney et al., 1973). The heterogeneous mixture of lipids in natural membranes, which includes an appreciable amount of unsaturated species, may behave quite differently from the lipids used in model studies. Although a totally crystalline membrane will not support cell growth, cells are known to be capable of growing well within the temperature range of the transition, where much of the membrane is crystalline. The point within the transition at which cell division ceases has been carefully measured with A . laidlawii and E . coli. The most elegant work has been carried out with A . laidluwii, which is particularly well suited for such experiments because its transition temperature can be immensely varied by diet. Acholeplusmu luidlawii shows changes in gross morphology and growth that correlate with the melting point of the fatty acid incorporated into its membrane lipids (Razin et al., 1966; McElhaney and Tourtellotte, 1969; Tourtellotte, 1972). Raised on oleate or other low-melting acids, the cells are filamentous and growth is rapid, but a diet of palmitate or stearate can cause swelling and eventual lysis. Growth ceases even before lysis, but incorporation of saturated fatty acids into the lipids continues until lysis, so that the membrane transition in badly swollen cells can occur entirely above growth temperature. As temperature decreases, the melting points of the fatty acids required for growth also decrease. The swelling experienced by cells with high transitions is osmotic in origin (Tourtellotte, 1972), and at low temperatures cells enriched in high-melting fatty acids become fragile and no longer behave as osmometers (van Zoelen et al., 1975). Rough correlations of growth and morphology with calorimetry (Steim er al., 1969) reveal that filamentous shapes are associated with transitions that are complete below growth temperature, swollen cells with transitions that encompass growth temperatures, and lysis with transitions above growth temperature. As fatty acids of higher melting point are incorporated, the calorimetrically detectable membrane transition rises, but growth continues even when much of the membrane is in the ordered state. This conclusion has also been reached by deuteron-resonance experiments (Oldfield er al., 1972) and is consistent with ' T - N M R and Tempo partitioning (Metcalfe et ul., 1972). More precise correlations of growth with fatty acid composition and membrane transitions observed by differential thermal analysis (DTA) reveal that the absolute growth rates at optimal growth temperatures and the apparent temperature characteristics of growth (the slope of an Arrhenius-type plot) are independent of fatty acid supplement above the transition until the growth temperature is lowered to about the midpoint of the transition (McElhaney, 1974). At this point, the temperature characteristic changes abruptly until, close to the lowtemperature end of the melt, growth ceases entirely. Judged by areas under the thermal analysis peaks, regardless of the fatty acid supplement, growth does not stop until about 90% of the membrane is crystalline. Thus growth continues at a reduced rate at temperatures below the approximate midpoint of the melt until only about 10% of the membrane remains fluid.


297

The same results are obtained in E . coli (Steim, 1970; Melchior and Steim, 1976). Ordinarily, wild-type cells have a very broad transition entirely o r almost entirely below growth temperature. Figure 5 , curve a is typical. Grown in the presence of 3-decynoyl-N-acetylcysteamine (DNAC), which inhibits unsaturated fatty acid synthesis (Kass, 1968), the membranes undergo a sharpened transition at elevated temperatures, shown in Fig. 5 , curve c . By varying the concentration of DNAC and correlating growth with calorimetry, one can determine the extent of transition compatible with growth. For sublethal concentrations of DNAC, cell division continues, even though the majority of the calorimeter peak is above growth temperature, until the thermogram in Fig. 5, curve c is obtained. Judged from areas under peaks, the membrane is again about 5-10% fluid and 95-90% ordered. In the state characterized by Fig. 5 , curve c the cells have been maintained by serial passage for 100 generations, but higher concentrations of DNAC elevate the transition even more, and growth ceases. DNAC-treated cells grown in oleate grow normally and have lower transition temperatures. Identical results, with cell division ceasing at about 90% crystallinity, were obtained with fatty acid auxotrophs fed elaidate. During membrane assembly, newly synthesized lipids and proteins may be inserted into the fluid portions of the bilayer and, after lateral diffusion, take their position in the membrane. New membrane would no longer be formed when the transition temperature rises so high that fluid sites are no longer available (Tsukagoshi and Fox, 1973). This could happen before completion of the order-disorder transition seen by physical methods, since the fluid regions remaining when cell division stops could be distributed above and through islands of crowded protein aggregates, or could be sprinkled randomly about the membrane in many small patches in inappropriate places, or could even exist on one side of the bilayer. At temperatures below the bulk bilayer transition, ordinarily fluid membranes are converted to an ordered state. And as previously discussed, not only are membrane proteins put into abnormal environments, but entiremembrane viscoelastic properties are drastically altered. A concept related to transitions is the idea that cells may find it advantageous to control the viscosity of their membrane bilayers by proper choice of fatty acids, even above transitions when the membranes are in an entirely fluid state, so that constant viscosity is maintained at any growth temperature (Sinensky, 1974). For example, in unsaturated fatty acid auxotrophs of E . coli, the dependence of the activation energies of some membrane-bound enzymes on the fatty acids that the membranes contain (Mavis and Vagelos, 1972) could reflect a viscous effect. Although it is certainly true that under normal conditions cells manipulate the fatty acid composition of their membranes in order to suppress the temperature of transition and maintain a fluid state, and though it may be true that viscosities in fluid membranes of many organisms have similar magnitudes, a true homeostatic control of fluidity as such does not appear to occur or does not appear to be necessary. The data already discussed suggest that variations in

298

DONALD L. MELCHIOR

membrane viscosities over a wide range have little effect on the rate of growth of A . laidlawii. Absolute growth rates of A . laidlawii at optimal growth tempera-

tures and the apparent temperature characteristics of growth are independent of fatty acid supplement above the transition until the growth temperature is lowered to about the midpoint of the transition. Though absolute growth rates and temperature characteristics of growth are identical, it is unlikely that viscosities at 36°C are also identical in membranes greatly enriched in isostearate (where the transition begins at 20°C and terminates about 5°C above growth temperature), and in membranes rich in straight-chain heptadecanoic acid (where the transition starts at 0°C and terminates about 20°C below growth temperature) (McElhaney, 1974). Furthermore, the temperature characteristics of growth routinely remain unchanged well into the bulk transitions, where one might expect rather drastic viscosity changes. It has been shown (Sinensky, 1974) that apparent viscosities seen by methyl-1 2-nitroxylstearate in wild-type E . coli membranes are similar in magnitude, provided that the electron spin resonance (ESR) spectra are taken at the temperature of growth. This may be a special case, however, since transitions in wild-type E . coli are broad and ordinarily terminate in the neighborhood of growth temperature. At that temperature ESR patterns might indeed resemble one another and would be drastically different at any lower temperature, since at lower temperatures the membranes would be undergoing a transition. Thus what would appear to be a homeostatic control of viscosity might merely be a reflection of a more fundamental process, the cellular control of transition temperatures. Nevertheless, there is some evidence from A . laidlawii, B . stearorherinophilus, and Y . enterocolirica, indicating that although rigorous control of fluidity above a transition is not especially advantageous to growth, there might be an upper limit to the fluidity that cells will tolerate. In studies on A . laidlawii (McElhaney, 1974), optimal growth temperatures were 36°C for all fatty acid supplements, independent of the transition temperatures observed by DTA, except for oleate and linoleate. For these two supplements, with respective transition midpoints at -13 and -19"C, optimum growth occurred at 34 and 32"C, respectively. Stearate membranes were a special case, since their transition was so high that growth was abnormally slow even at 38°C. Thus at 37"C, slow growth in stearate correlates with a very high transition, whereas slow growth in oleate and linoleate correlates with a very low transition. For intermediate transitions, growth is faster and constant. Bacillus stearotherrnophilus wild-type cells (Reizer, 1978) increase the melting temperature of their membranes as growth temperature is increased. These cells adjust the membrane transition so that they grow near, but slightly above, its upper end. As in other organisms, an increased transition temperature is brought about by incorporation of higher melting fatty acids into membrane lipids. In the case of wild-type B . stearothermophilus, this is accomplished primarily by an increase in palmitate and stearate

299


relative to the lower melting branched-chain and unsaturated fatty acids. All these phenomena are illustrated by Fig. 11 (Reizer, 1978). A mutant of B . stearothermophilus, TS- 13, cannot increase the temperature of completion of its transition beyond 40°C (McElhaney and Souza, 1976). The membranes of TS-13 cells are fully melted at 40°C. From 42 to 52°C this mutant grew nearly as well as wild-type cells. When growth temperature was raised still further, cell growth ceased abruptly at about 6 0 T , 20" above the completion of the membrane transition. Since the wild-type cells grew normally in this temperature range, this may again represent an upper limit on the degree of membrane fluidity compatible with cell growth. Another possible example of an upper limit to membrane fluidity is seen in Y . enterocolitica (Abbas and Card, 1980). When grown at 3 7 T , this organism has a membrane transition extending from -18 to 8"C, whereas cells grown at 22°C have a transition extending from -24 to 4°C. When 37°C cultures were shifted to 45"C, good growth was observed. However, when 22°C cultures were shifted directly to 45"C, they failed to grow.

VII.

BIOLOGICAL CONTROL

A fluid or at least partially fluid lipid bilayer seems to be essential for cellular function. Abnormally high transition temperatures reflect abnormally crystalline membranes and are associated with cell leakage, changes in active transport and

0.50

0.75

M o l e F r a c t i o n (C,,:,+

0

C,8:o)

FIG. 1 I , The effect of growth temperature on the membrane fatty acid composition and phase transition of B . stearothertnr~pkilus.As growth temperature (@) is decreased, the temperature of the upper (X)and lower (0)ends of the transition is lowered. This results from a reduction in the mole fraction of high-melting-point fatty acids (palmitate and stearate) in favor of lower melting point fatty acids, mostly branched-chain. In all cases the transition is complete at growth temperature. (From Reizer, 1978.)

300

DONALD L. MELCHIOR

some membrane-associated enzymatic activities, prolonged generation times, and eventual loss of viability and even cell death. Since transition temperatures depend primarily on the fatty acid composition of membrane lipids, low transition temperatures are assured by the biosynthesis of appropriate fatty acids or their selection from exogenous sources. In addition to straight-chain saturated fatty acids, to maintain low transition temperatures membrane lipids must also contain some fatty acids possessing lower melting points. Furthermore, the composition must be responsive to temperature in such a way that the membrane is totally or almost totally fluid at the temperature of growth. This necessity requires a control mechanism, optimally one that senses temperature and the physical state of the membrane and directs the incorporation of proportionally more unsaturated or other low-melting fatty acids into membrane lipids as the temperature decreases. Such control is seen in higher organisms (Irving et a/., 1956; Johnston and Roots, 1964; Rose, 1967). but is particularly important in prokaryotes (Marr and Ingraham, 1962) and other microorganisms in which the membrane can crystallize near growth temperature. For example, Melchior et al. ( 1 970) found by calorimetry that the temperature range of melting in A . luidlawii B , grown i n ordinary tryptose medium at 37"C, was in the neighborhood of growth temperature. At 37°C the membranes were mostly fluid, but at 25°C the membranes of cells grown at 37°C became almost completely crystalline. However, if the same organism was grown at 25°C the melting range was shifted down so that again the membranes were mostly fluid at growth temperature. A similar phenomenon has been observed calorimetrically in other prokaryotes, such as E . ~ ~ (Steim, d i 1972), B . steurotherrnophilus (Reizer, 1978), and Y . enrerocolitica (Abbas and Card, 1980). In addition to actively maintaining a fluid bilayer, it has recently been proposed that cells regulate their membrane lipid class composition in order to maintain them in a stable bilayer conformation (Wieslander et d.,1980). Although the detailed mechanism or mechanisms for temperature modulation of membrane fatty acid composition have not been worked out in any organism (Fulco, 1973). it has become clear that control can take place at several, possibly interrelated, levels. In some cases, desaturase activity appears to be governed by the solubility of oxygen, which serves as an eventual electron acceptor (Brown and Rose, 1969). In others, such as B . rneguterium (Fulco, 1970), enzyme synthesis is affected by temperature. Fatty acid desaturase is not synthesized in this organism at 35°C but is strongly induced at 20°C. Temperature also has a direct effect on the desaturase protein itself, which, once synthesized at low temperature, undergoes rapid irreversible inactivation at higher temperatures. Direct temperature effects on activity have also been found in E . coli, which produces monoenoic fatty acids via dehydration of the growing acyl chain within the fatty acid synthetase system itself. Deprived of glycerol in order to uncouple phosphatidic acid synthesis from fatty acid synthesis, E . coli accumulates large


301

quantities of free fatty acids, which become proportionately more unsaturated as temperature is decreased (Cronan, 1975). A reversible dependence of P-hydroxy-decanoyl thioester dehydrase on temperature may be implicated. In addition to effects of temperature on the biosynthesis of unsaturated fatty acids, another level of control is apparently at the site of phospholipid synthesis in the membrane. Temperature-dependent selection of saturated and unsaturated fatty acid CoA by membrane-bound acyltransferase, which catalyzes the esterification of glycerophosphate, has been demonstrated by Sinensky (1971). Presented with a mixture of oleoyl- and palmitoyl-CoA, cell-free E . coli acylCoA:glycerophosphate acyltransferase produces increasingly greater proportions of unsaturated lysophosphatidic acid at lower temperature. The acyltransferase apparently possesses a “preprogrammed” selective temperature response. It is the molecular nature of this temperature “program,” and its possible interrelationships with fatty acid biosynthesis, that we shall consider in more detail. Crucial to the understanding of the temperature-dependent selection process at the membrane level is the realization that fatty acids seem to be selected on the basis of melting point, a thermodynamic property that only indirectly reflects molecular structure. Although unsaturation is the usual route to low melting point, the same goal is attained in many organisms by employing structural alternatives, such as branched chains in many gram-positive bacteria (Wakil, 1970). This is demonstrated in Fig. 11 for B . stearotherrnophilirs, which when grown at progressively lower temperatures reduces the mole fraction of its membrane straight-chain saturated fatty acid content. Saturated fatty acids, palmitate and stearate, are replaced by lower melting point fatty acids, primarily branched-chain. In this manner B . steurothermc~philusis able to keep the onset of its membrane transition below its growth temperature (Reizer, 1978). Another route for membranes to attain lower temperature transitions is to incorporate shorter chain fatty acids into their membrane lipids. A striking example of this occurs in the psychrophilic Microcvccus cryophilus (Russell, 197 1). This prokaryote has in its membrane a very high percentage of the monounsaturates, octadecenoic and hexadecenoic acids. When its growth temperature is reduced from 20 to O’C, the total membrane content of these fatty acids does not change but the ratio of the high melter (octadecenoic acid) to the low melter (hexadecenoic acid) goes down by a factor of 4. Yet another strategy used by prokaryotes to introduce lower melting point fatty acids into their membranes at reduced growth temperature is by the substitution of lower melting point unteiso-branched-chain fatty acids for higher melting point iso-branched-chain fatty acids. I n a study using four temperature-range variants of B . meguteriurn over a temperature span of 65”C, it was found that as growth temperature was lowered the dominant change in membrane fatty acid content was the substitution of anteiso- for iso-branched-chain fatty acids (Rilfors er ul., 1978). The ratio of the higher melters to lower melters decreased progressively with decreasing

302

DONALD L. MELCHIOR

growth temperature, and at 5°C was about 12 times lower than at 70°C. Chain shortening also occurred with decreasing growth temperature. If the fatty acids are classified into long- and short-chain categories, there is a progressive decrease in the long-chain to short-chain ratio with decreasing temperature, the ratio being five times smaller at 5°C than at 70°C. A convincing argument that the principal consideration of fatty acid selection is thermodynamic rather than structural is based on the fact that a given organism, if forced to do so, will choose any low-melting exogenously supplied fatty acid to accomplish its goal of lowering transition temperatures. The best illustration is again A . laidlawii (McElhaney and Tourtellotte, 1969; Melchior et al., 1970; Tourtellotte, 1972), which lacks desaturase activity and, as pointed out earlier, incorporates large amounts of exogenous fatty acids into its membrane lipids. Fatty acids of progressively lower melting points are required as the growth temperature is decreased. Cis-unsaturates serve the purpose admirably even at the lowest temperatures, but growth is also normal if cis-unsaturates are replaced by branched-chain or cyclopropane fatty acids or by elaidate, an unnatural trans-unsaturated compound. Unsaturated fatty acid auxotrophs of E . coli show similar behavior, and will accept elaidate or even bromostearate (Schairer and Overath, 1969; Fox e t a / . , 1970). If in fact the temperature-sensing selection mechanism within the membrane is thermodynamically determined and depends on melting point, which is a bulk phenomenon, rather than on the chemical structure of the lipid, it is difficult to imagine it to be based on enzyme specificity. The binding of substrates to enzymes reflects the molecular structure of the ligand, and interaction occurs on a one-to-one basis, so that strictly thermodynamic properties have no meaning in such interactions. In accord with this thermodynamic point of view, it is proposed that the temperature program of acyltransferase in A . Iuidlawii, and in some other organisms as well, is an innate property of the bilayer in which the enzyme is embedded rather than a property of the protein itself. Some rather strong evidence supports this novel suggestion. In A . laidlawii cells, the pattern of esterification of palmitate and oleate from the incubation medium into the membrane polar lipids closely parallels the physical state of the membrane bilayer as determined calorimetrically (Melchior and Steim, 1976, 1977). Furthermore, the physical binding of free fatty acids to lipid bilayers formed from total extracted membrane lipids shows the same temperature dependence shown by the enzymatic process in live cells. This effect is seen in Fig. 12. Calorimetry of membranes grown in tryptose at 37°C produce the profile seen in Fig. 12, curve b, which, as the integral of the raw calorimeter peak, is a rough measure of the extent of membrane melting. The membranes are mostly fluid, but not completely fluid, at growth temperatures. Figure 12, curve a is a plot of the ratio of palmitate to oleate, both exogenously supplied, which are incorporated into membrane lipids when aliquots of cells grown at 37°C are briefly exposed to lower temperatures.


303

0

c

0

.o

" Q)

c X

0

20

40

ternperature('C)

Frc. 12. The membrane bilayer as a selector of fatty acids. Correlations between (curve a) the palmitate/oleate ratio incorporated into membrane lipids of A . luidlawii as a function of temperature; (curve b) the extent of the membrane transition; and (curve c) the palmitate/oleate ratio of fatty acids physically bound to bilayers of extracted membrane lipids as a function of temperature. Both incorporation and binding curves reflect the state of the bilayer, and are identical within experimental error. (From Melchior and Steim, 1977.)

The close correlation of the two curves could suggest that the conformation of the acyltransferase protein is somehow affected by the extent of fluidity of the membrane so that its affinity for palmitate increases relative to oleate with increasing temperature. However, hypothetical changes in enzyme specificity need not be involved. Fatty acid binding by liposomes formed from total membrane lipids, shown in Fig. 12, curve c , mimics the temperature-dependent selectivity of real cells. The agreement between the extent of melt and the pattern of uptake is not fortuitous, since changing the temperature range of melting produces a parallel change in both incorporation and binding. The ability to act as a temperature sensor and selector may be a general property of any phospholipid bilayer. Although lecithin is not found in A . luidlawii, bilayers prepared from lecithin mixtures showed similar correlations between the extent of bilayer melt and selective fatty acid binding. The consistent correlations between the physical state of the membrane bilayer, the binding of fatty acids by extracted lipids, and the incorporation of fatty acids into membrane lipids in live cells provide a new mechanism, based upon thermodynamic principles, for the temperature program of acyltransferase

304

DONALD L. MELCHIOR

activity. As temperature is lowered, an increased amount of oleate relative to palmitate is accepted by the bilayer, where it is acted upon by the resident acyltransferase. If the cells are allowed to grow at the lower temperature, the membrane transition will be shifted to lower temperatures as well. The detailed response to temperature will depend upon the shape of the bilayer binding curve. The acyltransferase proteins, although membrane-bound in A . laidluwii (M. Tourtellotte, personal communication) as they are in E . coli, are not required to sense the physical state of the membrane or to distinguish between different fatty acids. They may accept and use essentially any fatty acid molecules presented to them. The temperature-programmed selectivity, and hence the control of membrane-transition temperatures, resides in the bilayer itself. From another point of view, the acyltransferase enzymes are not simply proteins but proteins embedded in a bilayer. The catalytic function is assumed by the protein and the selectivity by the lipids. Similar considerations might also account for positional specificity, since the melting point of the acid esterified to the @carbon is usually lower than that on the a-carbon of the phospholipid (Hildebrand and Law, 1969; McElhaney and Tourtellotte, 1970; Okuyama et ul., 1976). Although the model just suggested is based upon the uptake of exogenous fatty acids by A . fuidluwii, the same model might be generalized to explain at least some temperature effects on the composition of membrane lipids synthesized from endogenously biosynthesized fatty acids (Melchior and Steim, 1978, 1979). That an interrelationship exists between selectivity at the membrane level and the synthetase system is demonstrated in E . coli cells, which produce longer chain fatty acids when uncoupled from phospholipid synthesis than when lipid synthesis is allowed to take place normally (Cronan, 1975). The intermediary for linking synthesis and desaturation of fatty acids with lipid synthesis in the membrane may be the bilayer as a temperature-dependent selective sink. Sumper and Trauble (1973) have in fact verified that long-chain acyl-CoA molecules will bind to and dissolve in dimyristoyl lecithin bilayers and E . coli membranes, where they are free to diffuse about and encounter appropriate membrane-bound enzymes. Furthermore, the efficacy of bilayers as an acceptor for fatty acyl-CoA is known to depend upon environmental conditions. Phosphatidic acid becomes an increasingly effective acceptor as ionic strength is increased, presumably because of charge neutralization by counter ions (Sumper, 1974). Since environmental temperature also modifies the efficacy of bilayers as acceptors (Melchior and Steim, 1976, 1977), such modification may provide the basis for temperature-dependent control. Consider the regulation of desaturase activity, and suppose that a desaturase exists that competes with the bilayer for the saturated end product of fatty acid synthetase. If at low temperatures the bilayer is in a relatively crystalline state that does not readily bind saturated fatty acids, the substrate will be operated upon instead by the desaturase. The resulting unsaturated fatty acid can now


305

easily enter the bilayer and be incorporated into the membrane lipids. As the bilayer becomes more fluid, it more successfully competes with the desaturase for saturated fatty acid, and the spectrum of fatty acids entering the membrane reservoir shifts toward increasing saturation. A case of control of fatty acid unsaturation by the thermodynamic state of the membrane has been reported in T . pyrifr,rmis (Martin c'/ m l . , 1976; Kasai ot m l . , 1976). A second thermal effect that might be explained by the concept of the bilayer as a temperature-programmed sink is the shortening of biosynthesized fatty acids at lower growth temperatures. In addition to the previously described case of M . cryophilus, shortening occurs in many organisms, including A . IuidIm'ii deprived of exogenous sources of fatty acids (M. Tourtellotte, personal communication). The progressive tendency already demonstrated in A . lmicllaw~iifor the bilayer to accept relatively more unsaturated than saturated fatty acids as temperature is lowered is again central to the argument. If in fact thermodynamic properties rather than specific molecular structure are the predominant factor affecting acceptability of a fatty acid derivative by the bilayer, one might expect that shorter chain molecules, like unsaturated long-chain ones, would at low temperature be more acceptable than saturated long-chain ones. Long-chain saturated molecules that are not accepted could be desaturated before being accepted. An example of fatty acid synthesis linked to the properties of a bilayer acceptor is provided by in virro studies on the fatty acid synthetase of Mycobacteriutn smrgmatis (Odriozola and Bloch, 1977). This enzyme system is unusual in that it produces a bimodal product pattern of fatty acid acyl-CoAs, short-chain-length C16-C18 CoAs and longer chain length C,,-C2s CoAs. In experiments on the effect of added dimyristoyl lecithin bilayers upon the chain lengths of the fatty acids synthesized by this system, the effect of the bilayers was found to be slight in the temperature range below and near the transition temperature, but very marked above it. At higher temperatures, and with no added lecithin bilayers, a large portion of C24-C28fatty acids are ordinarily synthesized. However, in the presence of dimyristoyl lecithin bilayers at temperatures above the lipid's transition temperature, these long-chain acids were not synthesized. It appears that the fatty acid sink provided by the melted-lipid bilayers allows an earlier chain termination for the fatty acid synthetase end product. An alternative possibility for linking fatty acid chain length to the properties of a bilayer acceptor, this time through enzyme inhibition, is suggested by the work of Sumper (1974). Dimyristoy1 lecithin, acting as a fatty acid-CoA sink, reverses the inhibition of fatty acid synthesis from acetyl-CoA in a system containing fatty acid synthetase and acetyl-CoA carboxylase from yeast. The promotion of synthesis by added lecithin apparently arises from competitive reversible binding of palmitoyl- or stearoyl-CoA by the lipid bilayers and acetyl-CoA carboxylase. Furthermore, fatty acid chain length depended on inhibition of carboxylase by palmitoyl-CoA. Increased inhibition led to an increased rate of synthesis of fatty acids of shorter

306

DONALD L. MELCHIOR

chain lengths. Although Sumper directed his attention toward an explanation for the chain-shortening effect of anaerobiosis, the same point of view can be extended to thermal effects if the temperature-dependent properties of the bilayer sink are kept in mind. At lower temperatures, as relatively more longer-chain saturated CoA molecules are excluded from the bilayer and accumulate, acetylCoA carboxylase inhibition by the accumulated longer-chain compounds would cause a shift toward the biosynthesis of shorter chains. From the aforementioned studies it appears that in some cases the physical state of the membrane bilayer can provide a temperature-sensitive mechanism to control the types of fatty acids incorporated into membrane lipids. Thus it is postulated that membranes may possess the ability to “self-control” their physical state. From one point of view, it is proposed that the bilayer be considered a temperature-programmed acceptor. Accordingly, the types of fatty acids incorporated at any temperature into membrane lipids by resident membrane enzymes are proposed to be those that enter the membrane-bilayer phase. These could partition into the bilayer phase from free solution or micellar aggregates, from a cytoplasmic enzyme, or from a carrier protein. Once the fatty acid or its derivative has entered the bilayer, a transfer mechanism involving lateral diffusion in two dimensions within the plane of the membrane would carry it to the subsequent membrane-bound enzyme. Such lateral diffusion in two dimensions within the membrane, combined with free diffusion in three dimensions within the cytosol, permits much faster transfer of molecules from the cell cytoplasm to a small target on the cell membrane than is provided by free diffusion alone (Sumpcr and Trauble, 1973; Adam and Delbriick, 1968). If, conversely, attention is focused on fatty acids or fatty acid derivatives that are excluded from the membrane bilayer but play a role in regulating the biosynthesis of fatty acids by cytoplasmic enzymes, it is proposed that the bilayer can be looked on as a tPmperururr-programmed sink. In this type of regulation, control would reside in the temperature-sensitive ability of the bilayer to selectively remove end products from the fatty acid synthetase system or compete f o r these products with such enzymes as desaturases, chain-elongation enzymes, or acetyl-CoA carboxylases. This type of control, which involves a multienzyme complex and competing acceptors, clearly differs from classical feedback inhibition (Bloch, 1977). For the feedback mechanism postulated here, end products need not bind to the enzyme component catalyzing the committed or earliest step of the pathway. However, the wasteful accumulation of intermediates for which no alternative routes are available is avoided as effectively as by conventional feedback. The functions of the membrane bilayer as acceptors and sinks are interrelated, of course, and in a certain sense separating the two functions is an artificial imposition done for the sake of clarifying this proposed mechanism. Taken together, the two functions-properly emphasized for the appropriate case at

307


hand-provide a sensitive, unified means to control the fatty acid composition and hence the physical state of biological membranes, whether fatty acids are supplied exogenously in the growth medium or are endogenously biosynthesized by the organism. ACKNOWLEDGMENTS The author is grateful to Professor J . M. Steim for reading this article in manuscript. He would like to thank Genevieve D. Goditt for her help in preparing this manuscript and gratefully acknowledges the permission of other authors for use of their illustrations. D. L. M. is the recipient of an American Diabetes Association, Inc., Research and Development Award. REFERENCES Abbas, C . A , , and Card, G. L. (1980). The relationship between growth temperature, fatty acid composition, and the physical state and fluidity of membrane lipids in Yersinia enterocolitica. Biochim. Biophys. Actu 602, 469-477. Adam, G., and Delbruck, M . (1968). Reduction of dimensionality in biological diffusion processes. In “Structural Chemistry and Molecular Biology” (N. Davidson and A. Rich, eds.), pp. 198-215. Freeman. San Francisco. Ashe, G . B., and Steim, J . M. (1971). Membrane transitions in gram-positive bacteria. Biochim. Biophvs. Actu 233, 810-814. Baldassare, J . J . , McAfee, A. G., and Ho, C . (1973). A spin label study of E . coli membrane vesicles. Biochem. Biophys. Res. Commuri. 53, 617-623. Baldassare, J . J . , Rhinehan, K. B., and Silbert, D. F. (1976). Modification of membrane lipid: Physical properties in relation to fatty acid structure. Biochemistry 15, 2986-2994. Bevers, E. M., Singal, S . A., Op den Kamp, J. A. F., and van Deenen, L. L. M . (1977). Recognition of different pools of phosphatidylglycerol in intact cells and isolated membranes of Achofeplrrsma Luidluwii. Biochemistry 16, 1290- 1295. Bevers, E . M., Op den Kamp. J . A.F., and van Deenen, L. L. M. (1978). Physical Chemical Properties of Phosphatidylglycerol in Membranes of Acholeplusma laidlawii. E u r . J . Biochem. 84, 35-42. Bevers, E. M . , Wang, H. H . , Op den Kamp, J . A. F., and van Deenen, L. L. M . (1979). On the interaction between intrinsic proteins and phosphatidylglycerol in the membranes of Acholeplasma laidlnwii. Arch. Biochem. Biophys. 193, 502-508. Birrell, G . B., and Griffith, 0. H. (1976). Cytochrome c induced lateral phase separation in a diphosphatidylglycerol-steroidspin label model membrane. Biochemisrry 15, 2925-2929. Bittman, R., and Blau, L. (1972). The phospholipid-cholesterol interaction. Kinetics of water permeability in liposomes. Biochemistry 11, 4831 -4839. Blazyk, J . F., Melchior, D. L., and Steim. J . M. (1975). An automated differential scanning dilatometer. Anal. Biochem. 68, 586-599. Bloch, K. (1977). Control mechanisms for fatty acid synthesis in Mycobacteriurn smegmaris. Adv. Enzymol. 45, 1-84. Blok, M. C . , van der Neut-kok, E . C. M., van Deenen, L. L. M . , and de Gier, J . (1975). The effect of chain length and lipid phase transitions on the selective permeability properties of liposomes. Biochim. Biophys. Actu 406, 187-196. Boggs, J . M., Wood, D. D.. Moscarello, M. A . , and Papahadjopoulos, D. (1977). Lipid phase separation induced by a hydrophobic protein in phosphatidylserine-phosphatidylcholinevesicles. Biochemr.\t~y 16, 2325-2329.

308

DONALD L. MELCHIOR

Bouvier, P., Op den Kamp, J . A . F.. and van Deenen, L. L. M. (1981). Studies on Acholeplasma laidlawii grown on branched-chain fatty acids. Arch. Biochem. Biophps. 208, 242-247. Brown, C . M . , and Rose, A . H. (1969). Fatty-acid composition of Candiclu utilis as affected by growth temperature and dissolved oxygen tension. J . Bucteriol. 99, 371 -378. Bfilet, P., and McConnell, H. M . (1976). Kinetics of phase equilibrium in a binary mixture of phospholipids. J . Am. Chern. Soc. 98, 1314-1318. Calhoun, W . I . , and Shipley, G . G . (1979). Sphingomyelin-lecithin bilayers and their interaction with cholesterol. Biochemistq 18, 1717-1722. Chapman, D., Williams, R . M . , and Ladbrooke, B. D. (1967). Physical studies of phospholipids. Chem. Phjs. Lipids 1, 445-476. Chen, Y . S . , and Hubbell, W . L. (1973). Temperature- and light-dependent structural changes in rhodospin-lipid membranes. Exp. Eye Res. 17, 5 17-532. Cheng, S . , Thomas, J . K., and Kulpa, C . F. (1974). Dynamics of pyrene fluorescence in Escherichiu coli. Biochemistry 13, I 135- 1339. Cheng, W . H. (1980). A theoretical description of phase diagrams for nonideal lipid mixtures. Biochirn. Biophjs. Acru 600, 358-366. Copps, T. P., Chelock, W. S . . and Pethan, A. (1976). Variation in distribution of membrane particles in Acholepkustnu lcrickluwii B with pH. J . Ulrrusrrucr. Re.\. 55, 1-3. Costerson, J . W . , Ingraham, J . M . L., and Cheng, K. J . (1974). Structure and function of the cell envelope of gram-negative bacteria. Bacteriol. Rev. 38, 87-1 10. Cronan. J . E. (1975). Thermal regulation of the membrane lipid composition of Escherichia coli. J . B i d . Chetn. 250, 7074-7077. Cullis, P. R . , and de Kruyff, B. (1978a). Polymorphic phase behavior of lipid mixtures as detected by "'P NMR. Biochim. Biopltys. Actu 507, 207-218. Cullis, P. R., and de Kruyff, B. (1978b). The polymorphic phase behavior of phosphatidylethanolamines of natural and synthetic origin. Biochirn. Biopphys. Acra 513, 3 1-42. Cullis, P. R.. and de Kruyff, B. (1979). Lipid polymorphism and the functional roles of lipids in biological membranes. Bioc,hinj. Biopl7js. Actu 559, 399-420. Cullis, P. R . , and Hope. M. J . (1980).The bilayer stabilizing rote of sphingomyelin in the presence of cholesterol. Biochirn. Bic~phys.Acru 597, 533-542. Cullis, P. R. and Verkleij, A . J . (1979). Modulation of membrane structure by Ca++ and dibucaine as detected by '"P NMR. Biochim. Biophps. Actci 552, 546-551. Cullis, P. R . , de Kruyff, B., and Richards, R. E. (1976). Factors affecting the motion of the polar headgroup in phospholipid bilayers. Biochim. Biophys. Acru 426, 433-446. Cullis, P. R., van Dijck, P. W. M . , de Kruyff, B . , and de Gier. J . (1978). Effects of cholesterol on the properties of equimolar mixtures of synthetic phosphatidyl-ethanolamine and phosphatidylcholine. Biochim. Biophjs. Acm 513, 21 -30. Davis, J . H.. Nichol, C . P . , Weeks, G . , and Bloom, M . (1979). Study of the cytoplasmic and outer membrane of Escherichio c d i by deuterium magnetic resonance. Biochemistn 18, 21 0321 12. de Kruyff, B., Demel, R . A., Slatbooni, A . J . , van Deenen, L. L. M . , and Rosenthal, A . F. (1973a). The effect of the polar headgroup on the lipid-cholesterol interaction. Biochini. Biophjs. Acto 307, 1-19. de Kruyff. B . , van Dijck, P. W . M., Goldbach, R . W.. Demel, R. A . , and van Deenen, L. L. M. (1973b). Influence of fatty acid and sterol composition on the lipid phase transition and activity of membrane-bound enzymes in Acholeplusmcr laidluwii. Biochim. Biophys. Actu 330, 269282. ile Kruyff, B., van Dijck, P. W . M . , Demel, R. A., Schuijff, A ., Branto, F., and van Deenen, L. L. M . ( 1974). Non-random distribution of cholesterol in phosphatidylcholine bilayers. Biochim. Biophj.s. Actu 356, 1-7.


309

Demel, R. A., and de Kruyff, B . (1976). The function of sterols in membranes. Biochirn. Biophys. Acru 457, 109-132. Demel, R. A., van Kessel, W. S. M. G . , Zwaal, R. F. A . , Roelofsen, B . , and van Deenen, L. L. M. (1975). Relation between various phospholipaae actions on human red cell membranes and interfacial phospholipid pressure in monolayers. Biochirn. Biophys. Acta 406, 97-107. Demel, R. A., Jansen, J . W . C. M., van Dijck, P. W. M., and van Deenen. L. L. M. (1977). The preferential interaction of cholesterol with different classes of phospholipids. B i o c h n . Biophys. Acru 465, 1-10, Devaux. P . , and McConnell, H. M. ( 1972). Lateral diffusion in spin-labeled phosphatidylcholine multilayers. J . A m . Cheni. So(,. 94, 4475-4481. DiRienzo, J . M . , Nakamura, K.. and Inouye, M. (1978). The outer membrane proteins of gramnegative bacteria. Annu. Rev. Biochern. 17, 481 -532. Dupont, G., Gabriel, A., Chabre, M.. Gulik-Krzywicki T . , and Schechter, E. (1972). Use of a new detcctor for X-ray diffraction and kinetics of the ordering of the lipids of E . Coli: Membranes and model systems. Notitrt~(London) 238, 33 1-333 Duppel, W., and Dahl, C. (1976). Effect of phase transition on the distribution of membraneassociated particles in microsomea. Biorhirti. Biophys. Acra 426, 408-417. Edidin, M . (1974). Rotational and translational diffusion in membranes. Arrriir. Re\,. Biophys. Bioetig. 3, 179-201. Engelman, D. M. (1971). Lipid bilayer structure in the membrane of Mycoplrrsmcr /uid/ciwsii, J . Mol. B i d . 58, 153-165. Faucon, J.-F., and Lusson, C . (1973). Aliphatic chain transitions of phospholipid vesicles and phospholipid dispersions determined by polarization fluorescence. Biochirn. Biophys. Actcl 307, 459-466. Findlay, A . (revised by Campbell, A . N . . and Smith, N . 0.)(1951). “The Phase Rule.” Dover, New York. Fox, C . F., Law, J . H., Tsukagoshi. N . . and Wilson, G . (1970). A density label for membranes. Pro(,. N u / . A ~ ~ r c%i. l . U . S . A . 67, 598-605. Fulco, A . J. (1970). Temperature-mediated hyperinduction of fatty acid desaturating enzyme. Biochirn. Biophys. A ~ c r218, 558-560. Fulco. A . J . (1973). Metabolic alterations of fatty acids. Aririic. Rev. B i o c h w ~ .43, 215-241. Galla, J . H., and Sackman. E. (1975). Chemically induced lipid phase separation in model membranes containing charged lipids: A spin label study. Biochrin. Biophys. Ac,tu 401, 509-529. Gally. H. U . , Pluschke. G . , Overath, P., and Seelig. J . (1979). Structure of Esc./ieric./riu coli membranes. 6lioc.hrmisrr:y 18, 5605 -56 10. Grant, C. W . M . , Wu, S. H.-W., and McConnell, H. M. (1974). Lateral phase separations in binary lipid mixtures. Biochirn. Biophys. A ( . m 363, 151-158. Haest, C. W. M . , de Gier, J . , van Es, C . A , , Verkleij. A . J . . and van Deenen, L. L. M. (1972). Fragility of the permeability barrier of Esdierichitr c d i . Biodiitn. Biophys. At,io 288, 43-53. Haest, C. W . M., Verkleij. A . J . , de Gier, J . , Scheek, R . , Ververgaert, P. H. J.. and van Deenen, L. L. M . (1974). The effect of lipid phase transitions on the architecture of bacterial membranes. Biocliirn. Biophy.\. A(,rtr 356, 17-26, Hauser, H., and Phillips, M. C. (1979). Interactions of the polar groups of phospholipid bilayer membranes. PrcZy. Surf. Mrrnhr..%i. 13, 297-404. Hildebrand, J . G . , and Law, J . H. (1964). Fatty acid distribution in bacterial phospholipids. The specificity of the cyclopropane synthetase reaction. Hiochctnistrv 3, 1304- 1308. Inouye, M., ed. (1979). “Bacterial Outer Membranes: Biogenesis and Functions.” Wiley, New York. Irving, L . , Schmidt-Nielson. K . , and Abrahamson. N. S . B. (1956). On the melting points ofanimal fats i n cold climates. Phv.sio/. Z o o / . 30, 93-105.

31 0

DONALD L. MELCHIOR

Ishinaga, M., Kanamoto, R., and Kito, M . (1979). Distribution of phospholipid molecular species in outer and cytoplasmic membranes of E . coli. J . Biochetn. (Tookw) 86, 161-165. Israelachvili, J . N., Mitchell, J . . and Ninham, B. W. (1977). Theory of self-assembly of lipid bilayers and vesicles. Biochim. Biophys. Acta 470, 185-201. Israelachvili, J . N., Marcelja, S . , and Horn, R. G. (1981). Physical principles of membrane organization. Q. Rev. Biophys. 13, 121-200. Johnston, P. V . , and Roots, B. I. (1964). Brain lipid fatty acids and temperature acclimination. Cornp. Biochern. Physiol. 11, 303-309. Kang, S . Y . , Gutowsky, H. S . , and Oldfield, E. (1979). Spectroscopic studies of specifically deuterium labelled membrane systems. Biorhemistry 18, 3268-3272. Kang, S . - Y Kinsey, R. A , , Rajan, S . , Gutowsky, H. S . , Gabridge, M. G., and Oldfield, E. (1981). Protein-lipid interactions in biological and model membrane systems. J . B i d . Chern. 256, 1155-1 159. Kasai, R., Kitajima, Y . , Martin, C. E . , Nozawa. Y.,Skriver, L . , and Thompson, G . A , , Jr. (1976). Molecular control of membrane properties during temperature acclimination. Membrane fluidity regulation of fatty acid desaturase action’? Biochetnisfry 15, 5228-5233. Inhibition in vivo of Kass, L. R . ( 1968). The antibacterial activity of 3-Decynoyl-N-acetylamine. p-hydroxydecanoyl thioester dehydrase. J . B i d . Chem. 243, 3224-3228. Kleeman, W . , and McConnell, H. M. (1974). Lateral phase separations in Escherichiu Coli membranes. Biorhirn. Biophvs. Artu 345, 220-230. Kleeman, W., Grant, C. W. M., and McConnell, H. M. (1974). Lipid phase separations and protein distribution in membranes. J . Suprumol. Strucr. 2, 609-616. Ladbrooke, D. B., Williams, R . M., and Chapman, D. (1968). Studies on lecithin-cholesterol-water interactions by differential scanning calorimetry. Biochirn. Biophvs. Actu 150, 333-340. Ladbrooke, M. D., and Chapman. D. (1969). Thermal analysis of lipids, proteins, and biological membranes. Chetn. Phys. Lipids 3, 304-367. Lange, Y . , D’Alessandro, J . S . . and Small, D. M. (1979). The affinity of cholesterol for phosphatidylcholine and sphingomyelin. Biochirn. Biophys. Acru 556, 388-398. Lawaczeck, R., Kainosho, M.. Giradet, J . L., and Chan, S . I . (1975). Effectsof structural defects in sonicated phospholipid vesicles on fusion and ion permeability. Ncrtrtre (LondonJ 256, 584586. Lee, A. G . (1977a). Lipid phase transitions and phase diagrams. I . Lipid phase transitions. Biochirn. Biophys. Acra 472, 237-28 1. Lee, A. G. (1977b). Lipid phase transitions and phase diagrams. 11. Mixtures involving lipids. Biochirr7. Biophys. Arta 472, 285-344. Lee, A. G., Birdsall, N . J . M., and Metcalfe, J . C. (1973). Measurement of fast lateral diffusion of lipids in vesicles and biological membranes by ‘H nuclear magnetic resonance. Biochernisfry 12, 1650-1659. Legendre, S., Letellier, L., and Shechter, E. (1980). Influence of lipids with branched-chain fatty acids on the physical, morphological and functional properties of the Escherirhiu coli cytoplasmic membrane. Biochim. Biophys. Actu 602, 491-505. Leive, L . , ed. (1973). “Bacterial Membranes and Walls.” Dekker, New York. Letellier, L., and Schechter, E. (1976). Evidence for a lipid and protein segregation in cytoplasmic membranes isolated from an Escherichiu coli fatty acid auxotroph: A chemical and morphological study of the different cytoplasmic membrane fractions. J . Microsc. Biol. Cell. 25, 191-195. Letellier, L., Moudden, H., and Schechter, E. (1977). Lipid and protein segregation in Escherichiu coii membrane: Different cytoplasmic membrane fractions. /‘roc. Narl. Arad. Sci. U.S.A. 74, 452-456. Linden, C. D., Keith, A. D., and Fox, C . F. (1973a). Correlations between fatty acid distribution in

,.


31 1

phospholipids and the temperature dependence of membrane physical state. 1. Suprarnol. Sfrucr. 1, 523-534. Linden, C . D., Wright, K . L . , McConnell, H . M., and Fox, C. F. (1973b). Lateral phase separations in membrane lipids and the mechanism of sugar transport in Escherichiu coli. Proc. Nut/. Acud. Sci. U . S . A . 70, 2271-2275. Loomis, C. R . , Shipley, G. G . , and Small, D. M. (1979). The phase behavior of hydrated cholesterol. J . Lipid Re.7. 20, 525-535. Lugtenberg, E. J . J . , and Peters, R. (1976). Distribution of lipids in cytoplasmic and outer membranes of Escherichiu coli K12. Biochim. Biophys. Acta 441, 38-47. Luna, E. J . , and McConnell, H. M. (1978). Multiple phase equilibria in binary mixtures of phospholipids. Biochim. Biophy. A(,tu 509, 462-473. Luzzati, V. (1968). X-ray diffraction studies of lipid-water systems.In “Biological Membranes” (D. Chapman, ed.), pp. 71-124. Academic Press, New York. Mabrey, S . , and Sturtevant, J . M. (1976). Investigations of phase transitions of lipids and lipid mixtures by high sensitivity differential scanning calorimetry. Proc. Natl. Aead. Sci. U.S.A. 73, 3862-3866, McConnell, H. M. (1976). Molecular motion in biological membranes. I n “Spin Labeling” (L. J . Berliner, ed.), pp. 525-560. Academic Press, New York. McConnell, H. M. (1978). Dynamic properties of membranes. “Membrane Immunochemisrry, ’. Adv. Chem. Phys. 39, 249-285. Macdonald, A. G . (1978). A dilatometric investigation of the effects of general anaesthetics, alcohols, and hydrostatic pressure on the phase transition in smectic mesophases of dipalmitoyl phosphatidylcholine. Biochirn. Biophys. Actu 507, 26-37. McElhaney, R. N. (1974). The effect of alterations in the physical state of the membrane lipids on the ability of Achok~p/us~iu luidlawii B to grow at various temperatures. J . Mol. Bin/. 84, 145157. McElhaney, R. N . , and Souza, K. A. (1976). The relationship between environmental temperature, cell growth and the ffuidity and physical state of the membrane lipids of Bacillus srcurothermophilus. Biochim. Biuphys. Actu 443, 348-359. McElhaney, R. N., and Tourtellotte, M. E. (1969). Mycoplasma membrane lipids: Variations in fatty acid composition. Science 164, 433-434. McElhaney, R. N . , and Tourtellotte, M. E. (1970). The relationship between fatty acid structure and the positional distribution of esterified fatty acids in phosphatidyl glycerol from Mycoplnsma laidlawii B . Biochim. Biophys. Actu 202, 120-128. McElhaney, R . N., de Gier, J . , and van der Neut-kok, E. C . M. (1973). The effect of alterations in fatty acid coniposition and cholesterol content on the nonelectrolyte permeability of Acholeplasrnu laidluwii B cellc and derived liposomes. Biwhirn. Biophys. Acra 298, 500-512. Marr, A. G . , and Ingraham, J . L. (1962). Effect of temperature on the composition of fatty acids in Escherichiu coli. J . Buc-teriol. 84, 1260- 1267. Marsh, D., Watts, A., and Knowles, P. F. (1976). Evidence for phase boundary lipid. Permeability of tempo-choline into dimyristoylphosphatidylcholine vesicles at the phase transition. Biochernisrn 15, 3570-3578. Martin, C. E . , Hiramitsu, K . , Kitajima, Y . , Nozawa, Y . , Skriver, L., and Thompson, G . A , , Jr. (1 976). Molecular control of membrane properties during temperature acclimination. Fatty acid desaturase regulation of membrane fluidity in acclimating Tetruhyrnena cells. Biochemistr): 15, 5218-5227. Marvis, R. D., and Vagelos, P. R. (1972). The effect of phospholipid fatty acid composition on membrane enzymes in Escherichia c.oli. J . Biol. Chem. 247, 652-659. Melchior, D. L., and Czech, M. P. (1979). Sensitivity of the adipocyte transport system to membrane fluidity in reconstituted vesicles. J . B i d . Chern. 254, 8744-8747.

DONALD L. MELCHIOR

312

Melchior, D. L., and Rottem, S. (1981a). The organization of cholesterol esters in the Mycoplusmu cupricolum membrane. Eur. J. Bioc,hem. 117, 147-153. Melchior. D. L., and Rottem, S. (1982). Physical state of mycoplasma membrane lipids. Re\,. I n f i ~ t Dis. . 4, Supplement 3. Melchior, D. L . , and Steim, J . M. (1976). Thermotropic transitions in biomembranes. Annu. Rev. Biophys. Biorng. 5 , 205-238. Melchior, D. L., and Steim, J . M. (1977). Control of fatty acid composition of Acholeplusma Itritlltnvii membranes. Bioc/iim. Biophys. A u u 466, 148-159. Melchior, D. L . , and Steim, J . M. (1978). The self control of lipid bilayers. Liq. Crysr. Or&rrr/ Fluids 3, 113-150. Melchior, D. L.. and Steim, J . M. (1979). Lipid-associated thermal events in biomembranes. P m g . Surf: Mcmbr. S(,i. 13, 2 I 1-296. Melchior, D. L., Morowitz, H. J . , Sturtevant, J . M.. and Tsong, T.-Y. (1970). Phase transitions of membrane lipids. Binrhi~7.Biophys. Acki 219, 114- 122. Melchior, D. L., Scavitto. F. J . , Walsh, M. T . , and Steim, J . M . (1977). Thermal techniques in biomenibrane and lipoprotein research. Tliermochim. Actrr 18, 43-71 . Melchior, D . L . , Scavitto, F. J . , and Steim. J . M. (1980). Dilatometry of dipalmitoyllecithincholesterol bilayers. Biochrmistry 19, 4828-4834. Metcalfe, J . C., Birdsall. N. J. M., and Lee, A. G. (1972). ':'C NMR spectra of Acko/op/trsmrr luirl/a~siimembranes containing '"C labeled phospholipids. FEES Lerr. 21, 335-340. Nagel, J . F. (1973). Lipid bilayer phase transition: Density measurements and theory. Proc. Nut/. Acucl. Sci. U . S . A . 70, 3443-3444.

Nakayama, H., Mitsui, T., Nishihara, M., and Kito, M . (1980). Relation between growth temperature of E . c d i and phase transition temperatures of its cytoplasmic and outer membranes. Biochim. Biophys. Acru 601, I 10. Nichol, C . P., Davis, J . H., Week, G . , and Bloom, M. (1980). Quantitative study of the fluidity of Escherichitr coii membranes using deuterium magnetic resonance. Biochn~isrry19, 45 1-457. Nichols. P., and Miller, N. (1974). Chloride diffusion from liposomes. Biochim. Biophys. Acrtr 356, -

184- 198.

Nikaido, H., and Nakae, T . (1979). The outer membrane of Gram-negative bacteria. A t h . Mic.rob. Physiol. 20, 164-250. Odriozola, J. M., and Bloch, K . (1977). Effects of phosphatidylcholine liposomes on the fatty acid synthetase complex from Mvcohnctrrirrm smegmutis. Biochim. Biophys. Acru 488, 198-206. Okuyama, H., Yamada, K . , Ikezawa, and Wakil, S. J . (1976). Factors affecting the acyl selectivities of actyltransferases i n Eschrrichiu coli. J . Biol. Ckrm. 251, 2487-2492. Oldfield, E., Chapman, 0 . .and Derbyshire, W . (1972). Lipid mobility in A(holepla.sma membranes using deuterium magnetic resonance. Chem. Phys. Lipids 9, 69-81. Onishi, S . , and Ito, T . (1974). Calcium-induced phase separations in phosphatidyl-serine - phosphatidylcholine membranes. Riochemistn 13, 881 -887. O p den Kamp. J . A. F. (1982). The lipid phase transition of phosphatidylglycerol in Acho/ep/usmu l ~ i d l a ~membranes ii studied with phospholipase A2. R r v . Infrcf. Dis. 4, Supplement 3 Overath, P., and Thilo, L. (197X). Structural and functional aspects of biological membranes revealed by lipid phase transitions. MTP I r i t . Rev. Scr.: Biochem., S r r . Two 19, 1-44. Overath, P . , and Trauble, H . (1973). Phase transitions in cells, membranes, and lipids of Eschrrichio c d i . Detection by fluorescent probes, light scattering, and dilatometry. Biochemisrn 12, 2625-2634.

Overath, P., Shairer, H. V., and Stoffel. W. (1970). Correlation of in iivu and iti virro phase transitions of membrane lipids in 6sc.herichiu c,o/i. Proc. Natl. Acad. Sci. U . S . A . 67, 606-612.

Overath, P., Hill, F. F., and Lammek-Hirsch, 1. (1971). Biogenesis of E. c d i membrane: Evidence for randomization of lipid phase. Nurure (London), NPW B i d . 24, 264-267.


313

Overath, P., Brenner, M.. Gulik-Krzywicki, T., Schechter. E., and Letellier, L. (1975). Lipid phase transitions in cytoplasmic and outer membranes of Eschcr-ichia coli. Biochim. Biophys. A C ~ U 389, 358-369. Papahadjopoulos, D.,Jacobson, K.. Nir, S., and Isac. T (1973). Phase transitions in phospholipid vesicles. Fluorescence polarization and permeability measurements concerning the effect of temperature and cholesterol. Biochim. Bioph\s. Acru 311, 330-348. Papahadjopoulos, D., Moscarello. M.. Eylar. E. H., and Isac, T. (1975a). Effects of proteins on thermotropic phase transitions of phospholipid membranes. Bioc.hiru. Biophy.5. A m r 401, 317-335. Papahadjopoulos, D., Vail, W . J . , and Moscarello, M. (1975h). Interaction of a purified hydrophobic protein from myelin with phospholipid membranes. Studies on ultrastructure, phase transitions and permeability. J . M e m h . Biol. 22, 143-164. Petit, V. A , , and Edidin, M. (1974). Lateral phase separation of lipids in plasma membranes: Effect of temperature on the mobility of membrane antigens. Science 184, 1183-1 185. Phillips, M. C., Ladhrooke, B. D., and Chapman, D. (1970). Molecular interactions in mixed lecithin systems. Eiochitn. Biophw. Acrrr 196, 35-44. Phillips, M. C . , Hauser. H . , and Paltauf. F. (1972). The inter- and intramolecular mixing of hydrocarbon chains in lecithiniwater systems. Chetn. Ph\.s. Lipids 8, 127-133. Poo, M.-M.. and Cone, R. A. (1974). Lateral diffusion of rhodospin in the photoreceptor membrane. Nature (Londoir) 247, 438-441. Razin, S . (1969). Structure and function of mycoplasma. Atinu. Re\’. Micruhiol. 23, 317-356. Razin, S. (1975). The mycoplasma membrane. Prog. Sut$ Menrbr, Sci. 9, 257-312. Razin, S. (1978). The mycoplasmas. Micruhiol. Rev. 42, 414-470. Razin, S . , Tourtellotte, M. E . , McElhaney, R. N . , and Pollack, J . D. (1966). Influence of lipid components of Myc~opI“r.smuluiclluwii membranes on osmetic fragility of cells. J . Bac[eriul. 91, 609-616. Razin, S . , Kutner. S., Ephrati, H., and Rottem, S. (1980). Phospholipid and cholesterol uptake by mycoplasma cells and membranes. Biochim. B i o p h ~ sActa 598, 628-640. Reinert, J . C . , and Steim, J . M . (1970). Calorimetric detection of a membrane-lipid phase transition in living cells. S(~ierrce168, 1580-1582. Reizer, J . (1978). Thermotropic properties and function of the membrane of a therniophilic bacillus. Ph.D. Thesis, Hebrew University, Jerusalem, Israel. Rilfors, L . , Wieslander, A . , and Stahl. S . (1978). Lipid and protein composiiion of membranes of Bacil1u.c megurerium. Variants in the temperature range 5 to 70°C. J . Bacteriol. 135, 10431052. Rose, A. H , , ed. (1967). “Thermobiology.” Academic Press, New York. Rottem, S . (1980). Membrane lipids of mycoplasmas. Biochim. Biophys. Acru 604, 65-90, Rottem, S . , and Samuni, A. (1973). Effect of proteins on the motion of spin-labeled fatty acids in mycoplasma membranes. BkJChittl. Biophvs. Acru 298, 32-38. Rottem, S . , Hubbell, W . L., Hayflick, L., and McConnell, H. M . (1970). Motion of fatty acid spin labels i n the plasma membrane of mycoplasma. Biochim. Biuphys. Acm 219, 104-1 13. Rottem, S . , Yashouv, J . , Ne’enian, Z . . and Razin, S. (1973a). Cholesterol in mycoplasma membranes. Composition, ultra-structure and biological properties of membranes from M w u plasma Mycoides var. m p r i cells adapted to grow with low cholesterol concentrations. Biochim. Biophys. Actu 323, 495-508. Rottem, S . , Cirillo, V . P.. de Kruyff, B.. Shinitzky, M., and Razin, S . (1973b). Cholesterol in mycoplasma membranes, correlation of enzymic end transport activities with physical state of lipids in membranes of Myc~op/a,st~ru rnycnides var. cupri adapted to grow with low cholesterol concentrations. Riochitrr. Biophys. Actu 323, 509-5 19. Russell, N. J . (1971). Alteration in fatty acid chain length in Micrococcus cryophilus grown at different temperatures. Biochim. Bioph\s. Acra 231, 254-256.

314

DONALD L. MELCHIOR

Sackmann, E., Trauble, H., Galla, H . , and Overath, P. (1973). Lateral diffusion, protein mobility, and phase transitions in Escherichiu coli membranes. A spin label study. Biochemistry 12, 5360-5369. Schairer, H. V . , and Overath, P. (1969). Lipids containing rruns-unsaturated fatty acids change the temperature characteristic of thiomethylgalactoside accumulation in Escherichia c d i . J . M o l . B i d . 44, 209-214. Schechter, E . , Gulik-Krzywicki, T . , and Kaback, H. R. (1972). Correlations between fluorescence, X-ray diffraction, and physiological properties in cytoplasmic membrane vesicles isolated from Escherichiu coli. Biochim. Biophys. Aetu 274, 466-477. Schechter, E., Letellier, L., and Gulik-Krzywicki, T . (1974). Relations between structure and function in cytoplasmic membrane vesicles isolated from an Escherichiu culi fatty-acid auxotroph. Eur. J . Biochem. 49, 61-76. Shah, D. O., and Schulman, J . H. (1967). Influence of calcium, cholesterol, and unsaturation on lecithin monolayers. J . Lipid Res. 8 , 215-226. Sheetz, M. P., and Chan, S. 1. (1972). Effect of sonication on the structure of lecithin bilayers. Biochemisrry 11, 4573-4581. Shimshick, E. J . , and McConnell, H. J . (1973). Lateral phase separation in phospholipid membranes. Biochemisrn 12, 2351 -2360. Shipley, G . G . , Green, J . P., and Nichols, B. W . (1973). The phase behavior of monogalactosyl, digalactosyl, and sulphoquinovosyl diglycerides. Biochim. Biophys. Acru 311, 53 1-544. Shipley, G . G . , Avecilla, L. S . , and Small, D. M. (1974). Phase behavior and structure of aqueous dispersions of sphingomyelin. J . Lipid Res. 15, 124-131. Sinensky, M. (1971). Temperature control of phospholipid biosynthesis in Escherichiu coli. J . Bacteriol. 106, 449-455. Sinensky, M. ( 1974). Homeoviscous adaptation-a homeostatic process that regulates the viscosity of membrane lipids in Escherichiu coli. Proc. Nutl. Acud. Sci. U.S.A. 71, 523-525. Small, D. M. (1967). Phase equilibria and structure of dry and hydrated egg lecithin. 1. Lipid Res. 8, 551-557. Small, D. M. (1970). The physical state of lipids of biological importance: Cholesterol esters, cholesterol, triglycerides. Adv. Exp. Met/. B i d . 7, 55-83. Small, D. M., and Shipley, G. G . (1974). Physical-chemical basis of lipid deposition in atherosclerosis. Science 185, 222-229. Smith, I. C. P., Butler, K. W . , Tulloch, A. P., Davis, J . H . , and Bloom, M. (1979). The properties of gel state lipid in membranes of Achnlep/usma laidlutvii as observed by ‘H NMR. FEBS Lett. 100, 57-61. Sreim, J . M. (1968). Spectroscopic and calorimetric studies of biological membrane structure. In “Molecular Association in Biological and Related Systems’’ (R. F. Could, ed.), pp. 259-302. Am. Chem. Soc., Washington, D.C. Steim, J . M. (1970). Thermal phase transitions in biomembranes. Liq. Cryst. Ordered Fluids 1, 1-11. Steim, J . M. (1972). Membrane transitions: Some aspects of structure and function. In “MitochondriaiBiomembranes” (S. A. van den Berg, P. Brost, L. L. M. van Deenen, J . C. Riemersma, E. C. Slater, and J . M. Tager, eds.), pp. 185-196. North-Holland Publ., Amsterdam. Steim, J . M., Tourtellotte, M. E., Reinert. J . C., McElhaney, R. N . , and Rader. R. L. (1969). Calorimetric evidence for the liquid-crystalline state of lipids in a biomembrane. P r w . Natl. Acud. Sci. U . S . A . 63, 104-109. Stockton, G . W., Johnson, K . C., Butler, K. W., Polnaszek, C. F., Cyr, R., and Smith, I . C. P. (1975). Molecular order in Acholeplusmu luidluwii membranes as determined by deuterium magnetic resonance of biosynthetically incorporated specifically labelled lipids. Biochim. Biophys. Aclu 401, 535-539.


315

Sumper, M. (1974). Control of fatty-acid biosynthesis by long-chain acyl CoAs and by lipid membranes. Eur. J . Biochetir. 49, 469-475. Sumper, M . , and Traiible, H. (1973). Membranes as acceptors for palniitoyl CoA in fatty acid biosynthesis. FEBS Lett. 30, 29-34. Tall, A. R., and Robinson. A. (1979). Absence of liquid crystalline transitions of cholesterol in reconstituted low-density lipoproteins. FEBS Lrrt. 107, 222-226. Tanford, C . (1973). "The Hydrophobic Effect: Formation of Micelles and Biological Membranes. '' Wiley, New York. Tecoma, E. S . , Sklar, L. A , . Simoni, R . D . , and Hudson, B. S . (1977). Conjugated polyene fatty acids as fluorescent probes: Biosynthetic incorporation of parinaric acid by Escherichitr c,o/i and studies of phaae transition$. Bioi.hi,i)iiury 16, 829-835. Templin, P. R. (1956). Coefficient of volume expansion for petroleum waxes and pure n-paraffins. /rid. E n g . Cheni. 48, 154-161. Thilo, L . , and Overath, P. ( 1976). Randomization of membrane lipids in relation to transport system assembly in Esc.hericliiu coli. Biochctnistry 15, 328-334. Thilo, L. Trauble, H., and Overath, P. (1977). Mechanistic interpretation of the influence of lipid phase transitions on transport functions. Brot~hetnistry16, 1283- 1290. Tourtellotte. M. E. (1972). Mycoplasina membranes, structure and function. Ir, "Membrane Molecular Biology" (C. F. Fox and A. 0. Keith, eds.), pp. 439-470. Sinauer Assoc., Stanford, Connecticut. Tourtellotte, M. E.. Branton, D., and Keith, A. (1970). Membrane structure: Spin labeling and freeze etching of Mye~op/us~uo ltridluwii. Proc. Nirtl. Acud. Sci. U.S.A. 66, 9OY-916. Trauble, H. (1971). Phase transitions in lipids. Biornrmhrcuirs 3, 197-227. Triuble, H., and Overath, P. (1973). The structure of E d w r i c A i r i i d i membranes studied by fluorescence measurements of lipid phase transitions. Biochitn. Biophys. A r m 307, 491 -5 12. Triuble, H . , and Sackmann. E. (1972). Studies of the crystalline-liquid crystalline phase transition of lipid model membranes. 111. Structure of a steroid-lecithin system below and above the lipidphase transition. J . A m . Chcwi. SO(,.94, 4499-45 10. Tsien, H . C . , and Higgins. M. L. (1974). Effect of temperature on the distribution of membrane particles in StrL.l"o(,oc.i.ic.\ ./iircdis as seen by the freeze-fracture technique. J . Bu(,trrio/. 118, 125-734.

Tsukagoshi, N.. and Fox, C. F. (1973). Transport system assembly and the mobility of membrane lipids in E.sc~/ierichiucoli. Biot.hrwtisrr;y 12, 2820-2822. Ubbelohde, A. R . (1965). "Melting and Crystal Structure.'' p. 325. Oxford Univ. Press. London and New York. Ulmius. J . , WennerstrAm. H.,Lindblom, G . , and Avidson, G . (1977). Deuteron nuclear magnetic resonance'studies o f phase equilibria in a lecithin-water system. Biochernr.\try 16. 57425745.

Untracht. S . H . , and Shipley, G . G . (1977). Molecular interactions between lecithin and sphingomyelin. J . Bi$. ChivJi. 252, 4449-4457. van Dijck, P. W. M . . Ververgaert, P. H . J . T . , Verkleij. A. J . , van Deenen, L. L. M . , and d c Gier, J . (1975). Influence of CaY' and Mg'+ on the thermotropic behavior and permeability properties of Iiposomes prepared from dirnyristoyl phosphatidylglycrrul and mixtures of dimyristoy1 phosphatidylcholine. Hioc.hirrr. B h p h k r t r 406, 465-478. van Dijck, P. W . M.. de Kruyff. B., van Deenen, L. L. M., de Grier, J . , and Demel. R. A . (1976). The preference o f cholesterol for phosphatidylcholine in mixed phosphatidylcholinephosphatidylethanolamine bilayers. E i o ~ ~ h i r iRi .i o / ~ h , ~k. tt. u 455, 576-587. van Heerikhuizen, H., Kwak, E . , van Bruggen, F. J . . and Witholt, B. (1975). Characterization of a low density cytoplasmic membrane subfraction isolated from Eschrrichiu coli. Biochi~rr. Biophys. Ac,tu 413, 177-191. van Zoelen, E. J. J., van der Neut-kok, E . C, M., de Gier, 1 . . and van Deenen, L. L. M . (lY75).

31 6

DONALD L. MELCHIOR

Osmotic behavior of Acholeplasma laitllawii E cells with membrane lipids in liquid-crystalline and gel state. Eiochim. Eiophys. Acto 394, 463-469. Verkleij, A. J . , and Ververgaert, P. H. J . T . (1975). The architecture of biological and artificial membranes as visualized by freeze etching. Annu. Rev. Phys. Chem. 26, 101-122. Verkleij, A. J., Ververgaert, P. H. J. T., van Deenen, L. L. M., and Elbers, P. F. (1972). Phase transitions of phospholipid bilayers and membranes of Acholepfusma laidlawii B visualized by freeze fracturing microscopy. Eiochinr. Eiophys. Acta 288, 326-332. Verkleij, A. J . , Ververgaert, P. H. J. T., de Kruyff, B . , and van Deenen, L. L. M. (1974). The distribution of cholesterol in bilayers of phosphatidylcholines as visualized by freeze fracturing. Eiochirn. Eiophys. Acru 373, 495-501. Verkleij, A. J., Ververgaert, P. H. J . T . , Prins, R. A . , and van Golde, L. M. C. (1975). Lipid phase transitions of the strictly anaerobic bacteria V e i l l o n e h Purr& and Atiuerovibrio lipolyrico. J . Eucteriol. 124, 1522- 1528. Verkleij, A. J . , Lugtenberg, E. J . J., and Ververgaert, P. H. J. T. (1976). Freezeetch molphology of outer membrane mutants of Escherichici coli K 1 2 . Eiochim. A i o p h ~ s Arta . 426, 581 -586. Ververgaert, P. H. J . T . , Verkleij, A. J . , Elberts, P. F., and van Deenen, L. L. M. (1973). The distribution of cholesterol in bilayers of phosphatidylcholines as visualized by freeze fracturing. Biochim. Eiophys. Acru 373, 495-501, Verwer, W., Ververgaert, P. H. J. T . , Leunissen-Bijvolt, J . , and Verkleij. A. J . (1978). Particle aggregation in photosynthetic membranes of the blue-green alga Aiiucystis nidulurrs. Eiochirn. E i o p h y ~ ACIU . 504, 23 1-234. Wakil, S. J., ed. (1970). “Lipid Metabolism.” Academic Press, New York. Wallace, B. A , , and Engelman, D. M . (1978). The planar distribution of surface proteins and intramembrane particles in Acholeplasma laidlawii are differentially affected by the physical state of membrane lipids. Eiochim. Eiophys. A m 508, 431 -449. Wieslander, A , , Ulmius, J . , Lindblom. G . . and Fontell, K. (1978). Water binding and phase structures for different Acholeplasma laidlawii membrane lipids studied by deuteron nuclear magnetic resonance and X-ray diffraction. Biochim. Eioph?/.s. Acta 512, 241 -253. Wieslander, A , , Christiansson, A , , Wlater, H., and Weibull, C. (1979). Fractionation of membranes from Acholeplasmci laidlawii A on the basis of their surface properties by partition in twopolymer aqueous phase systems. Eiochim. Eiophys. Actu 550, 1-15. Wieslander, A,, Christiansson, A , , Rilfors, L . , and Lindblom, G . (1980). Lipid bilayer stability in membranes. Regulation of lipid composition in Acholephsma laiclluwii as governed by molecular shape. Eioch~mistry19, 3650-2655. Wu, S. H . , and McConnell, H. M. (1975). Phase separation in phospholipid membranes. Eiochemistry 14, 847-854. Wunderlich, F., Ronai, A., Speth, V., Seeling, J . , and Blume, A. (1975). Thermotropic lipid clustering in Tetrcihymena membranes. Biochemistry 14, 3730-3735. Yeagle, P. L., Hutton, W. C., and Martin, R. B. (1978). Sphingomyelin multiple phase behavior as revealed by multinuclear magnetic resonance spectroscopy. Eiochemistv 17, 5745-5750.

CURRENT TOPICS IN MEMBRANES AND TRANSPORT. VOLUME 17

Effects of Membrane Lipids on Transport and Enzymic Activities RONALD N . McELHANEY Depurrment of’Biochrmistr.v Universitj of Alhertu Edmonton, Alhrrtu, Ctrrttrtlu

Introduction . . . . . . . . . . . . . . . . . . . . . Relevant Properties of Membrane Constituents . . . . . . . A , Membrane Lipids . . . . . . . . . . . . . . . . B . Membrane Proteins . . . . . . . . . . . . . . . . 111. Arrheniua Plots of Membrane Transport Systems and Enzymes IV. Studies of Cells and Membranes . . . . . . . . . . . . A . CellGrowth . . . . . . . . . . . . . . . . . . . B. Chemotaxis . . . . . . . . . . . . . . . . . . . C. DNA Synthesis . . . . . . . . . . . . . . . . . D. Protein-Mediated Transport Processes . . . . . . . . E. Membrane-Associated Enzyme Activities . . . . . . . V. Studies of Isolated Membrane-Bound Enzymes . . . . . . VI. Conclusions . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . I.

11.

1.

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . .

317 318 . . 318 . . 319 . . 320 . . 323 . . 324 . . 331 . . 331 . . 331 . . 347 . . 362 . . 369 . . 369

INTRODUCTION

The generally accepted fluid-mosaic model of membrane structure proposed by Singer and Nicolson ( 1 972) regards biological membranes essentially as two-dimensional solutions of oriented globular proteins in a fluid lipid bilayer phase. Thus proteins are free to diffuse laterally in the plane of the membrane unless constrained by their interactions with other proteins, either within the membrane itself or outside the membrane proper, including proteins functioning as cytoskeletal structural elements. The differential lateral mobility of proteins within the lipid bilayer can lead to a locally heterogeneous or “mosaic” two31 7

Copyright 0 1982 by Academic Press. Inc All nghts of reproduction in any form reserved. ISBN 0-12-153317.4

318

RONALD N. McELHANEY

dimensional organization of the membrane. Moreover, Israelachvili ( 1 977) has pointed out that thermodynamic and packing considerations suggest a coupled organization of lipids and proteins within the membrane, in which both lipid and protein may deform or cluster in order to accommodate each other most favorably. Although a refined fluid-mosaic membrane model seems applicable to most membrane systems, certain membranes, which have a low lipid/protein ratio and very extensive interactions between membrane proteins, may be more accurately characterized as two-dimensional, quasi-crystalline arrays of proteins with lipids filling the interstices between the protein molecules. In such membranes, almost all the lipid present undergoes continuous interaction with the membrane protein, and the lateral mobility of both constituents is quite restricted. A microbial example of such a membrane is the purple membrane from Halobacterium halobium (for review, see Stoe7I77 of prokaryotcs, 158- I67 Prokaryotes, .see ~ r l s oBacteria biological consequence5 of membrane state, 292-299 distribution of lipids in. 12-13 actinomycetes and related organisms. 28-3 1 budding and/or appendagcd bacteria, 16- I7 cyanobacteria, 14 endospore-forming hacteria. 27-28 gliding and sheathcd bacteria, 16 Gram-negative aerobic rods and cocci, 18-20 Gram-negative anaerobic bacteria, 21 -23 Gram-negative anaerobic cocci, 23 Gram-negative chemolithotrophic bacteria. 23-24 Gram-negative cocci and coccobacilli, 23 Gram-negative facultatively anaerobic rods, 20-21 Gram-positive cocci, 25-27 Gram-positive, non-spore-forming rods. 28 methane-producing bacleria, 24 mycoplasma, 3 1 phototrophic bacteria, 14- IS rickettsia, 3 1 spiral and curved bacteria, 18-20 spirochetes, 17- I8 lipids and phylogeny. 3 1-34 membrane biological control of, 299-307 fluidity modulating lipids, 282-284 membrane bilayer transitions examples, 275-282 general properties, 267-269 lateral phase separation, 269-274 membrane patching fluid bilayers, 284-286 membrane proteins, 286-292 polyterpenoids as phylogenetic precursors ol sterols archaehacterial polyterpenes. 176- 177

carotenoids as precursors of cyclic polyterpenoids, 175- I76 general features of common polyterpene biogenetic pathway, 167-170 hopanoids as precursors of sterols, 170 I75 polyterpenoids of, IS8 distribution of, 159-162 functional equivalence to sterols, 166- 167 structural regularities in membrane polyterpenoids, 162-166 regulation of membrane fluidity, lipid phases and. 264-267 sterols of absence of, 155- I56 apparent exceptions, 156- I57 case of unicellular eukaryotes, 157-158 structure of lipids apolar chains, 2-5 nonextrdctable lipids, I2 nonpolar lipids, 11-12 other polar lipids, 9- 1 I polar lipids with I ,2-diradylsn-glyceroI backbone, 5-9 Protein(s), membrane, relevant properties of, 3 19-320

S Stcrols o f prokaryotes absence of, 155- I56 apparent exceptions, 156- 157 case of unicellular eukaryores, 157-158 role in mycoplasma membrane, 19 1-200

T Transport processes, protein-mediated amino acids, 340-343 ions, 343-346 other, 346-347 sugars. 33 1-340 Tricyclohexaprenol, as putative prokaryotic triterpene and putative phylogenetic precursor of hopanoids, 177-178


No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

Recommend Documents