Advances in Microbial Physiology Volume 35

Advances in

MICROBIAL PHYSIOLOGY

II

It is our sad duty to announce the death of the Editor of Advances in Microbial Physiology, Professor Anthony H . Rose, shortly before the publication of this volume. He will be widely missed by his friends, and scientific and publishing associates.

Advances in

MICROBIAL PHYSIOLOGY Edited by

A. H. ROSE School of Biological Sciences Bath University, UK

Volume 35

ACADEMIC PRESS Harcourt Brace & Company, Publishers London San Diego New York Boston Sydney Tokyo Toronto

ACADEMIC PRESS LIMITED 24-28 Oval Road London NWI 7DX US Edition published by ACADEMIC PRESS INC. San Diego CA 92101

Copyright 0 1993 by ACADEMIC PRESS LIMITED This book is printed on acid-free paper

All Rights Reserved

No part of this book may be reproduced in any form by photostat. microfilm, or any other means, without written permission from the publishers A CIP record for this book is available from the British Library

ISBN C12-027735-2 ISSN 0065-291 1

Typeset by J&L composition Ltd, Filey, North Yorkshire Printed in Great Britain by Hartnolls Ltd, Bodmin, Cornwall

Contributors J. Bailey Genetics Department, University of Leicester, Leicester, LE1 7RH, UK A. Bock Lehrstuhl fur Mikrobiologie der Universitat Munchen 19, Germany S. Bringer-Meyer Institut fur Biotechnologie, Forschungszentrum Julich, 5170 Julich, Germany T. G. Burland McArdle Laboratory, University of Wisconsin, 1400 University Avenue, Madison, WI 53706, USA D. B. Cunningham Unite de Genttique Moleculaire Murine, Institut Pasteur, 25 rue de Docteur Roux, 75015 Paris, France W. F. Dove McArdle Laboratory, University of Wisconsin, 1400 University Avenue, Madison, WI 53706, USA H. Fukuda Department of Applied Microbial Technology, Kumamoto Institute of Technology, Ikeda 4-22-1, Kumamoto 860, Japan J. Heider Lehrstuhl fur Mikrobiologie der Universitat Munchen 19, Germany J. A. Hoch Division of Cullular Biology, Department of Molecular and Experimental Medicine, The Scripps Research Institute, 10666N. Torrey Pines Road, La Jolla, C A 92037, USA T. Ogawa Department of Applied Microbial Technology, Kumamoto Institute of Technology, Ikeda 4-22-1, Kumamoto 860, Japan M. Rohmer Universite de Haute Alsace, Ecole Nationale Superieure de Chimie, 68093, Mulhousel, France H. Sahm Institut fur Biotechnologie, Forschungszentrum Julich, 5170 Julich, Germany L. Solnica-Krezel CVRC, Massachusetts General Hospital-East 4, Harvard Medical School, Thirteenth Street, Bldg 149, Charlestown, MA 02129, USA G . A. Sprenger Institut fur Biotechnologie, Forschungszentrum Julich, 5170 Julich, Germany S. Tanase Department of Biochemistry, Kumamoto University School of Medicine, Honjo 2-2-1, Kumamoto 860, Japan M. A. Valvano Department of Microbiology and Immunology, University of Western Ontario, London, Ontario, Canada N6A 5C1 R. Welle Institut fur Biotechnologie, Forschungszentrum Julich, 5170 Julich, Germany C. Whitfield Department of Microbiology, University of Guelph, Guelph, Ontario, Canada N l G 2W1

This Page Intentionally Left Blank

Contents Con t ribu t ors

V

Patterns of Inheritance, Development and the Mitotic Cycle in the Protist Physarum polycephalum TIMOTHY G . BURLAND, LILIANNA SOLNICA-KREZEL, JULIET BAILEY, DAVID B. CUNNINGHAM and WILLIAM F. DOVE I. Introduction 11. Life cycle 111. Genome organization

IV. V. VI. VII. VIII.

Cytoskeletal organization The mitotic cycle Expression of introduced molecules Concluding remarks Acknowledgements References

2 4 6 13 39 58 62 62 63

Selenium Metabolism in Micro-organisms JOHANN HEIDER and AUGUST BOCK I. Introduction 11. Selenium-containing enzymes 111. Selenium-containing tRNAs IV. Biosynthesis of selenoproteins and seleno-tRNAs

V. VI. VII. VIII. IX.

Selenium versus sulphur Transport and metabolism of selenium-containing compounds Geochemistry of selenium Conclusions Acknowledgements References

71 72 88 89 96 98 100 103 104 104

...

Vlll

CONTENTS

Regulation of the Onset of the Stationary Phase and Sporulation in Bacillus subtilis JAMES A. HOCH

I. Introduction 11. The phosphorelay 111. Control of the phosphorelay IV. Transition-state regulators V. Alternatives to sporulation VI. Initiation of sporulation VII. Acknowledgements References

111

113 120 126 129 130 132 132

Biosynthesis and Expression of Cell-Surface Polysaccharides in Gram-Negative Bacteria CHRIS WHITFIELD and MIGUEL A. VALVANO

1. 11. 111. IV. V. VI. VII. VIII.

List of abbreviations Introduction Structure and attachment of cell-surface polysaccharides Polysaccharide biosynthesis Export of polysaccharides and cell-surface assembly Genetics of polysaccharide biosynthesis Regulation of cell-surface polysaccharide synthesis Conclusions Acknowledgements References

136 136 138 154 171 188

212 229 230 23 1

Biochemistry and Physiology of Hopanoids in Bacteria HERMANN SAHM, MICHEL ROHMER, STEPHANIE BRINGERMEYER. GEORG A. SPRENGER and ROLAND WELLE

I. Introduction 11. Structural diversity of bacterial hopanoids 111. Detection and analysis of bacterial hopanoids

IV. Distribution and physiological role of hopanoids V. Biosynthesis and genetics VI. Conclusions References

247 250 253 254 259 270 270

CONTENTS

ix

Ethylene Production by Micro-organisms H. FUKUDA, T. OGAWA and S. TANASE I. Introduction 11. Production of ethylene by micro-organisms 111. Biosynthetic pathways to ethylene in micro-organisms and

higher plants IV. Mechanisms for formation of ethylene by Pseudornonussyringue V. Molecular cloning and expression of the gene for the ethylene-forming enzyme of Pseudomonus syringue VI. Comparison of the structure of the ethylene-forming enzyme from Pseudornonus syringue with that of related enzymes VII. Concluding remarks References Addendum added in proof Author index Subject index

275 277 28 1 292 292 295 302 303 307 309 333

This Page Intentionally Left Blank

Patterns of Inheritance, Development and the Mitotic Cycle in the Protist Physamm polycephalum TIMOTHY G. BURLAND," LILIANNA SOLNICAKREZEL,b JULIET BAILEY," DAVID B. CUNNINGHAMd and WILLIAM F. DOVE" McArdle Laboratory, University of Wisconsin, 1400 University A venue, Madison, W153706, USA, CVRC, Massachusetts General Hospital-East 4, Harvard Medical School, Thirteenth Street, Bldg 149, Charlestown, M A 02129, USA, Genetics Dept, University of Leicester, Leicester LEI 7RH, U K , and Unitk de Gtnktique Moleculaire Murine, lnstitut Pasteur, 25 rue du Docteur R o w , 75015 Paris, France

a

I. Introduction . . . . . . . . . . 11. Life cycle . . . . . . . . . . . A. Amoeba1 phase . . . . . . . . B. Plasmodia1 phase . . . . . . . . C. The sexual cycle and inheritance . . . . 111. Genome organization . . . . . . . . A. Nuclear chromosomal genome . . . . . B. Nucleolar D N A genome . . . . . . C. Mitochondria1 genome . . . . . . IV. Cytoskeletal organization . . . . . . . A. Microtubule organization . . . . . . B. Tubulin genes and polypeptides . . . . C. Tubulin utilization . . . . . . . D. Function of multiple tubulins . . . . . E. Microtubule-associated proteins . . . . F. The cytoskeleton in development . . . . G . Thecytoskeletonindevelopmentalmutants . . H. Other genesdifferentially expressed in development 1. Inferences . . . . . . . . . V. The mitotic cycle . . . . . . . . . A. The plasmodia1 mitoticcycle . . . . . B. Periodic variations . . . . . . . C. Chromosome replication . . . . . . D. Ribosomal D N A replication . . . . . E. Mitotic regulation . . . . . . . . ADVANCES IN MlCROBl AL PHYSIOLOGY. VOL. 35 ISBN lL12-027735-2

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

2 4 4

4 5 6 6 8 10

13 13 14

17 20 22 23 33 37 38 39 39 42 48 52 53

Copyright 01993, by Academic Press Limited All rightsof reproductioninany form reserved

2

T. C i . HURI.AND ET A / .

V1. Expression of introduced molecules A. Diffusion uptake . . . B . Macroinjection . . . C. DNA transformation . . VII. Concluding remarks . . . VIII. Acknowledgements. . . . References . . . . . .

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

.

.

.

.

.

.

.

.

.

.

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

58 58 58 59 62 62 63

I. Introduction Physarum polycephalum has commonly been cited in the GenBank sequence database as a plant, and in the Medline bibliographic database as a fungus; other descriptors such as “protostelid” have also been proposed (for a review, see Alexopoulous, 1982). Recent taxonomy places P. polycephalum convincingly in the protist kingdom (Margulis and Schwartz, 1982). but the unpretentious vernacular description “plasmodial slime mould” remains unchanged. The variety of cell types observed during the life cycle (Fig. 1) reveals how one might be led astray in trying to classify the organism: the amoeba and the plasmodium appear so different that they could easily be mistaken for two organisms from different kingdoms. The wealth of biological variation in this organism provides a broad array of opportunities for experimental analysis. The different cell types and developmental pathways of P. polycephalum provide abundant opportunities for analysis of problems in cellular and developmental biology; the natural mitotic synchrony of the plasmodium provides unique opportunities for experimental analysis of the unperturbed mitotic cycle; and genetic analysis is made possible by the sexual and meiotic alternations between amoebal and plasmodial stages of the life cycle. Beneath this veneer of variation lies a remarkably conserved array of fundamental biological structures and processes. The arnoebal microtubular cytoskeleton is built from conserved tubulins, orchestrated by organizing centres like the counterparts in the cells of animals. And amoebal mitosis follows the same basic pattern as mitosis in the animal cell. In the plasmodium, mitosis occurs inside t h e intact nuclear membrane, as it does in fungi, while the classical eukaryotic cycle of chromatin condensation and decondensation is preserved during the naturally synchronous mitotic cycle, based on a complement of histones and modifying enzymes similar to those of most other eukaryotes. Many other fundamental eukaryotic processes are conserved in this protean protist. In this review we summarize the biology of P. polycephalum, and give examples of how the organism has been utilized for analysis of patterns of inheritance, development and the mitotic cycle. We introduce recent

3

TIIF PROTIST PfiYSARUM POI.YCEPtIAI.UM

DORMANT CYSTS

FLAGELLATES

(DIPLOID)

DORMANT SCLEROTIUM

FIG. 1. Life cycle of Physarum polycephalum. The outer circuit summarizes the life cycle of heterothallic strains, typical of the species isolated from nature. The inner circuit summarizes the lil'e cycle of apogamic strains, such as the marA2gadAh mutant CL. Apogamic strains retain the ability to cross with heterothallic strains of appropriate mating types. Redrawn from Burland (1978).

4

T. G . BURI.AND ET

Al..

advances in DNA transformation and gene targeting in the organism, and point to definitive experiments now possible using the new technology with the inveterate biology of this plasmodial slime mould. 11. Life Cycle A. AMOEBAL PHASE

Among the distinct cell types of P. polycephalum, only the amoeba and plasmodium are capable of proliferation. The uninucleate amoebae are usually haploid, though some isolates of the close relative Didymium iridis are diploid (Collins and Betterley, 1982). Amoebae live primarily in the soil, feeding phagocytically on bacteria and other microbes (Olive, 1975). In the laboratory, amoebae are grown on lawns of Escherichia coli but, in order to grow amoebae of P. polycephalum axenically, mutant strains had to be selected (Dee ef al., 1989). On transfer to water, amoebae develop reversibly into flagellates (Fig. l ) , changing their mode of locomotion from amoeboid crawling to swimming. Upon return to a solid substrate or nutrient medium, flagellates revert to the amoeba1 cell type. Under adverse conditions, amoebae develop reversibly into cysts which, unlike amoebae and flagellates, possess a cell wall (Raub and Aldrich, 1982). Cysts germinate to form amoebae or flagellates upon exposure to favourable conditions. The third developmental option for amoebae is irreversible transition t o the plasmodium (Fig. 1). B. PLASMODIAL PHASE

The plasmodium of P. polycephalum is most commonly found in the litter of the forest floor. In this yellow-pigmented, multinucleate syncytium, growth and synchronous nuclear division continue in the absence of cytokinesis; the plasmodium increases in mass as long as nutrients are available. As far as we know, there is no limit to the size of a single plasmodial cell, although we have not grown a synchronous plasmodium larger than a 30 cm or so in diameter in the laboratory. The plasmodium can be grown axenically on a surface or in submerged, shaken liquid culture. In the latter situation, a large plasmodium breaks into pieces a millimetre or less in diameter; such “microplasmodia” grow to culture densities of over 100 mg wet weight ml-’, and without agitation readily settle to the bottom of the container at unit gravity. With these characteristics, in combination with emerging molecular technologies and absence of a cell

THE PROTlST PHYSARUM POLYCEPHALUM

5

wall, the plasmodium could serve as an efficient biofactory for research and commercial products. When two plasmodia carrying identical alleles of thefusA, f u s E and fusC loci come into contact, they fuse together (Poulter and Dee, 1%8). Nuclei and cytoplasm soon mix, although the nuclei do not fuse, so that a heterokaryon is formed. Plasmodia1 heterokaryons are useful for analysis of mitotic regulation, somatic compatibility and genetic complementation testing. The plasmodium has two developmental options, each requiring starvation for induction. In the dark, the starving plasmodium encysts, forming cell-walled structures known as sclerotia or spherules; these revert to active plasmodia when conditions become favourable. If starving plasmodia are illuminated, irreversible meiotic sporulation occurs. Each haploid spore cell is encased in a wall and, upon exposure to favourable conditions, spores “hatch” to release amoebae or, in moist conditions, flagellates, thereby completing the life cycle (Fig. 1). C. THE SEXUAL CYCLE AND INHERITANCE

Mating and meiotic recombination in P. polycephalum were first demonstrated by Dee (1960, 1966, 1982). Spores of wild-type diploid plasmodia hatch to yield haploid progeny amoebae, which carry alternate alleles, such as rnatAl or matA2, of the primary mating-type locus. Plasmodium development does not normally occur from amoebal clones of either mating type, but does occur in mixtures of matAZ and mafA2 amoebae, as a result of sexual fusion between haploid amoebae of different mating type, yielding diploid heterozygous plasmodia. Mating in P. polycephalum is thus heterothallic. Pairs of alleles o f two other mating-type loci, matB and marc, also segregate among the meiotic progeny of most plasmodia1 isolates. Allelic difference at matA is required for fusion of nuclei of different mating types and subsequent plasmodium development (Dee, 1982), while allelic difference at matB is required for efficient cell fusion (Youngman et al., 1981), and allelic difference at matC improves the efficiency of crossing at certain pH values (Kawano et al., 1987b). These requirements for genetic difference in amoebal mating contrast with the genetic similarity required for somatic fusion of plasmodia. From analysis of distinct natural isolates of P. polycephalum, at least 13 matA alleles (Collins and Tang, 1977) and 13mutB alleles (Kirouac-Brunet et a f . , 1981) have been identified, and it seems likely that outbreeding is rampant in natural populations. The multiplicity of mating-type alleles is reminiscent of mating types in basidiomycetes (e.g. Metzenberg, 1990), although cellular changes that occur under control of the mating loci appear completely different in the two groups of organisms.

6

T. C i BURI.AND ET AI.

To simplify genetic and cell-biological analyses, apogamic strains of P. polycephalum have been derived in which a haploid amoeba can develop into a haploid plasmodium without cell fusion and without change in ploidy (Fig. 1; Cooke and Dee, 1974). Such apogamy, or “selfing”, usually arises in the laboratory through mutation at or very close to matA, further implicating this locus as a major regulator of the development of amoebae into plasmodia. Apogamic strains retain the ability to cross with heterothallic strains (Fig. l), facilitating both isolation and genetic analysis of mutants. Although viability of spores from haploid apogamic plasmodia is generally low, it is adequate for laboratory analysis, and correlates well with the frequency of rare diploid nuclei found in the otherwise haploid plasmodia (Laffler and Dove, 1977). Somatic inheritance of specific markers appears to be very stable; for example, there is no known example of an amoeba changing mating specificity after extended subculture, nor have nuclear genome rearrsngements been observed between different cell types (Sweeney et al., 1987). Nevertheless, with continued extensive subculture of plasmodia in the laboratory, various sublines of the same original isolate have inherited distinct characteristics (Mohberg and Babcock, 1982). Of more concern, after extended subculture plasmodia can become heteroploid (Kubbies and Pierron, 1983; Kubbies et al., 1986), and this heteroploidy is associated with poor synchrony of DNA replication, which is potentially a problem for mitotic-cycle analysis. Therefore, the preferred strategy for maintaining strains in the laboratory is to generate plasmodia periodically from fresh crosses of various well-characterized pairs of amoebal stocks. There are no difficulties with heteroploidy in such plasmodia (Cunningham, 1992). Appropriate amoebal stocks can be stored frozen, recloned from time to time, and are publicly available; thus, to maintain precise, homogeneous ploidy, isogenicity and experimental consistency with P. polycephalum is a matter of elementary microbiological technique. 111. Genome Organization A. NUCLEAR CHROMOSOMAL GENOME

The nuclear genome is distributed among approximately 40 chromosomes (Mohberg, 1982b), which appear small and of similar size under the light microscope; a karyotype has not been determined. The haploid unreplicated DNA content of the nucleus in P. polycephalum is 0.3 pg (Mohberg, 1982b), corresponding to 2.7.10’ bp in each genome-about the same as in Drosophila melanogaster. Although P. polycephalum has a

Ttll- PRUI'ISI I'IIYSARUM /'OI.Y('EI't/AI.UM

7

variety of cell types in its life cycle, Dr. rnelanogaster has far more, and it would seem unlikely that all of the DNA in P. polycephalurn encodes protein o r RNA products. Gross analysis of nuclear DNA by reassociation kinetics suggests that no more than two-thirds of the genome is single copy. The repetitive sequences include both inverted-repeat and direct-repeat structures, the most abundant of which appear to be methylated (for a review, see Hardman, 1986). The predominant repeat structure, Tpl (transposon Physarurn l ) , is related to classical retrolransposon-like sequences such as copia in Dr. rnelanoguster (Rothnie er al., 1991), and the next most abundant repeat structure, Tp2, appears to be a member of the same group (McCurrach et al., 1990). Organization of I ' p l repeats into scrambled tracts of up to 50 kb suggests that, over evolutionary time, the elements have retrotransposed into already integrated transposons, although transposition of T p elements in P. polycephalurn has not been demonstrated in the laboratory. A single Tpl element consists of an 8343 bp sequence flanked by 277 bp LTRs (long terminal repeats) that are terminated by short inverted repeats (Rothnie et a l . , 1991). The LTRs contain putative transcriptional signals as well as sites analogous to initiation sites for DNA synthesis found in both retrotransposons and retroviruses. T p 1 elements include open-reading frames (ORFs) corresponding to homologues of the protease, endonuclease, reverse transcriptase and nucleic acid-binding products of copia (Rothnie er al., 1991). However, the transposon-like ORFs found so far appear incomplete, which is not surprising if indeed T p l tracts derive from multiple integrations of one transposon into another; only a minority of transposons might retain a complete structure. If any of the T p elements are functional retrotransposons, they may have practical value for recombinant DNA applications in P. polycephalurn, notably for integrative DNA transformation. And whether or not functional retrotransposons are present, elements like Tpl are of interest in elucidating the evolutionary origins of retrotransposons. Sequence homologies between Tpl and other retrotransposons, such as copia from an insect, Tyl and Ty3 from a yeast, and Tal and Tntl from plants, indicate that this group of transposons originated from a common ancestor. An interesting question is whether these related transposons populated genomes of the diverse group of hosts prior to the hosts' divergence, or afterwards by horizontal transmission. Comparisons of sequences of transposons in closely related species with those of distantly related species are beginning to address this issue (Rothnie et al., 1991). Cloning of structural genes from P. polycephalurn has sometimes been difficult, due to instability in bacteria of some of the repeated DNA structures and other unusual sequences (e.g. Nader et al., 1986). However,

8

1 G I3URI A N 0 I.T A /

sequences that present cloning problems tend to be regularly dispersed throughout the genome, typically with 5-10 kb of single-copy sequence interspersed. Structural genes of interest can be obtained by cloning DNA fragments of the order of 5-7 kb or less (e.g. Monteiro and Cox, 1987a; Gonzalez-y-Merchand and Cox, 1988; Adam et al., 1991). Structural genes examined so far typically contain several introns (e.g. Adam et al., 1991), but these are small relative to the size of introns in mammalian genes. Where detailed genomic structure is not needed, cDNA libraries have proved to be an efficient means of elucidating the coding potential of specific genes (e.g. Pallotta et al., 1986; Paul et al., 1992). H . NUCLEOLAR DNA GENOME

The genes for 26S, 19s and 5.8s rRNAs in P. polycephalum are encoded on 60 kb, linear extrachromosomal DNA molecules present at about 150 copies in each haploid nucleus. These rDNA molecules are located in the nucleolus, which greatly simplifies their purification from chromosomal and mitochondria1 DNA populations. Each rDN A molecule is a palindrome containing two sets of rRNA genes, each set being transcribed away from the centre (for a review, see Hardman, 1986). The two 13.3 kb transcription units are separated by a 23 kb central non-transcribed spacer, which contains a variety of repetitive elements and can vary in size even within one strain. In possessing multicopy extrachromosomal rDNA molecules, P. polycephalum is typical of many protists. A linear, palindromic structure with two transcription units is found in Dictyostelium discoideum (Welker et a l . , 1985) and Tetrahymena thermophila (Engberg and Nielsen, 1990), though D . iridis has only a single transcription unit (Johansen, 1991). In other protists, circular molecules with one or two transcription units have been found, and the presence of multiple copies of rDNA sequences is typical of genomes in a wide variety of organisms (Long and Dawid, 1980). The rDNA molecules in P. polycephalum have a repeated (T2AG3)" telomere structure (Forney et al., 1987), similar to telomeres in chromosomes of Trypanosoma spp., Neurospora spp., humans and other organisms (Zakian, 1989; Coren et al., 1991). Such sequences facilitate recognition of telomeres by the telomerase needed to complete replication (Blackburn, 1991) and probably also facilitate binding of other proteins to prevent exonuclease degradation of the ends of linear chromosomes, thereby stabilizing the linear topology. A protein, PPT, has been identified in P. polycephalum that specifically binds to the T2AG3 repeats of the rDNA molecule (Coren et al., 1991). Binding of the protein is resistant to ribonuclease, and thus PPT seems not to be a telomerase, which recognizes telomere sequences through its RNA moiety (Blackburn, 1991). It may

THF P H 0 7 ' I S l PHYSAR U M POL YCEPHA LUM

9

rather be a structural protein that protects replicated telomeres. Further study should yield information relevant not only for rDNA from P. pofycephafum, but also for the general nature of linear chromosome maintenance in eukaryotes. Unlike chromosomal genes, rDNA molecules are inherited in a non-Mendelian fashion. Using restriction fragment-length polymorphism (RFLP) markers for rDNA in crosses, Ferris et a f . (1983) found that amoebal-progeny clones carry either one or other parental rDNA species, but not both. However, the ratio of the two rDNAs among the progeny is biased in favour of one species; the older the plasmodium before sporulation and meiosis, the more biased is the distribution among the progeny. These observations suggest that replication efficiencies of different rDNA molecules are dissimilar, leading to an unequal proportion of the two rDNA molecules, and that a single master copy of rDNA is chosen randomly at meiosis to be passed on to the progeny (Ferris et a f . , 1983). One implication of this hypothesis is that no nuclear chromosomal master copy of rDNA need exist. Molecules of rDNA in P. pofycephafumcontain two or occasionally three so-called group-I introns (Muscarella and Vogt, 1989) in the gene encoding 26s rRNA. The third intron, intron-3, has so far been found in only one isolate, namely Carolina. Intron-3 occurs at the same position as another group-I intron in the highly conserved rRNA gene in Tetrahymena spp. (Muscarella and Vogt, 1989), and the 3' portion of intron-3 is remarkably similar to the intron found in Tetrahymena spp., including conserved sequences involved in self-splicing. One powerful feature of the biology of P. pofycephafumis the simplicity with which functions in the plasmodium can be compared between diploid heterozygotes, formed by crossing together amoebae of different mating types, and haploid heterokaryons, formed by fusing together plasmodia of identical fusion type. This feature was used to elucidate intron-3 action. When a strain of amoebae from the mould carrying intron-3 is crossed with a strain lacking intron-3, and the crossed plasmodium is analysed, the intron transposes site specifically into all of the rDNA molecules in the diploid heterozygous nuclei of the plasmodium (Muscarella and Vogt, 1989). In plasmodia1 heterokaryons formed by fusing together plasmodia carrying intron-3' and intron-3- nuclei, breakage of rDNA in some of the intron3- nuclei is detectable, consistent with synthesis of a specific, intron-3encoded endonuclease in the cytoplasm and subsequent entry into the nuclei. However, transposition is not observed in the heterokaryon, indicating as expected that intron-3 is nucleus limited. It is now clear that intron-3 encodes a site-specific endonuclease, called I-PpoI (Muscarella et a f . , 1990). It appears that I-PpoI is encoded from a transcript synthesized by RNA polymerase I (Muscarella and Vogt, 1989).

10

I G I3UHI.ANII I I A /

Endonuclease I-PpoI cuts rDNA at the site of integration of intron-3, both in vitro and in E. coli, indicating that intron-3 catalyses its own transposition. The recognition sequences for two other intron-encoded endonucleases, I-See1 and I-SceII, found in the mitochondrial rDNA of Sacch. cerevisiae are large compared with the more familiar bacterial restriction endonucleases (Delahodde et a l . , 1989). Likewise, the I-Ppol recognition sequence is 13-15 nucleotide pairs long, encompassing the sequence 5' CTCTCTTAA 4 GGTAGC 3 ' , where the arrow indicates the site of cleavage (E. L. Ellison and V. Vogt, personal communication; Lowery et al., 1992). Endonucleases of this type are expected to be useful for mapping large genomes, since the rare cleavage-site frequency allows a chromosome to be subdivided into large, discrete units for physical analysis. Endonuclease I-PpoI appears to have the catalytic efficiency and stability needed for in vitro analyses, and the enzyme is now commercially available (Lowery et al., 1992). In view of the efficiency with which transposition occurs in a cross between intron-3' and intron-3- strains, it is interesting that this intron has been found only in one strain of P. polycephalum. The Carolina isolate which carries intron-3 may be geographically isolated. Alternatively, the presence of intron-3 may be deleterious and selected against. Perhaps it is a recent invader of the Carolina isolate soon to spread to other strains. The nucleotide sequences of rRNA genes are among the most conserved through evolution, and sequences for both 19s rRNA (Johansen et al., 1988) and 26s rRNA (Hasegawa et al., 1985) have been used to estimate phylogenetic relationships between P. polycephalum and other eukaryotes (e.g. Baroin et al., 1988; Lenaers et al., 1988). These comparisons suggest that P. polycephalum represents a line of descent that was one of the earliest to diverge from other eukaryotic lines (Johansen et al., 1988). In this respect, it is remarkable that fundamental cellular processes like mitosis and the chromatin condensation-decondensation cycle in the slime mould are more similar to vertebrates than to those of some apparently laterdiverging fungi. This may reflect a decreased spectrum of motility functions in fungi compared with slime moulds and vertebrate cells, leading to loss of unnecessary functions in fungi. Other functions may also be more similar between slime moulds (and other protists) and vertebrates, since motility embraces so many aspects of cell structure and function. C. MITOCHONDKIAL GENOME

Mitrochondrial DNA (mtDNA) comprises 5-10% of cellular DNA in P. polycephalum. Estimates of size for the mitochondrial genome from a variety of strains range from 56-86 kb (Kawano et al., 1982, 1987a; Jones

THF PROTISI PHYSARUM POLYCEI’HALUM

11

et al., 1990; Takano et al., 1990). As with nuclear gene markers, different isolates display abundant mtDNA RFLP, and complete restriction maps are available for several mtDNAs. The 86 kb size estimate includes a map showing near-terminal duplications of 19.6 kb stretches (Takano et al., 1990), which could explain why many estimates are around 60 kb; either other mtDNAs lack this duplication or it was overlooked. However, mtDNAs can vary in size within several organisms, and it is possible that some of the size variation reported for mtDNA from P. polycephalurn is due to natural variation (e.g. Kawano er al., 1987a) rather than different experimental interpretations. Both linear and circular topologies have been proposed for mtDNA from this slime mould (Jones et a l . , 1990; Takano et al., 1990). The observation that a specific mtDNA restriction fragment that maps close to the proposed linear terminus is preferentially sensitive to exonuclease digestion strongly favours a linear rather than a circular topology, at least for the mtDNA from the Colonia genetic background (Takano et al., 1990). It remains possible, however, that both linear and circular mtDNAs could exist, perhaps related by recombination. Restriction fragment-length polymorphism markers were used to establish that mtDNA inheritance in P. polycephalurn is uniparental (Kawano et al., 1987a), as is so in a broad array of eukaryotes. However, uniparental inheritance in isogamous organisms such as P. polycephalurn is remarkable in that both gametes contribute mitochondria at sexual fusion. Therefore, a mechanism must exist to select or eliminate one genome. Meland et al. (1991) showed that one of the nitDNA genomes from this slime mould is specifically eliminated within two cell cycles after sexual fusion of amoebae. Interestingly, which mtDNA is eliminated depends upon the matA alleles carried by the parents; different rnatA alleles can be ranked in linear hierarchical order of dominance for determining loss of mtDNA (Meland et al., 1991). For example, in a matA7xrnarA2 cross, the mtDNA carried by the rnatA7 parent is inherited while the mtDNA from the matA2 parent is lost; in a rnatA2xrnatAIZ cross, the mtDNA of the rnatA2 parent is inherited; and from this pair of relationships, it can be deduced that, in a rnatA7xmatAlI cross, the rnatA7-associated mtDNA will be inherited (Meland et al., 1991). This is the first evidence for active degradation of mitochondria1 genomes in sexual crosses. The phenomenon bears a remarkable resemblance to degradation of chloroplast genomes in crosses in Chlarnydornonas reinhardtii (Kuroiwa, 1985). Thus, although rDNA and mtDNA are both inherited in a non-Mendelian fashion in P. polycephalum, as they are in D. iridis (Silliker and Collins, 1988), the mechanisms that distort segregation of rDNA and mtDNA genomes are distinct and act at different stages of development. Mitochondria in P. polycephalurn typically possess a single spherical mass

12

T G RUHI.ANI) b.7 A1

of material that stains with diamidinophenylindole, which is known as the mt nucleus and presumably represents the mitochondrial genome. Fusion of mitochondria in zygotes and during sporulation occurs in the Ng isolate of P. pofycephalum and most of its derivative strains, leading to larger, knotted, multinucleate mitochondria (Kawano et a [ . , 1991). Other strains tested do not exhibit this mitochondrial fusion. Following mitochondrial fusion, fusion of mt nuclei occurs. Remarkably, at spore germination, fused mitochondria and their mt nuclei divide to yield the original spherical mitochondrial-nuclear morphology (Kawano et al., 1991). Genetic evidence supports these morphological phenomena: rather than the usual uniparental inheritance of mtDNA, progeny of such crosses exhibit recombination of mtDNA RFLP markers. Kawano et af. (1991) suggest these observations reflect a mitochondrial meiotic cycle. Occurrence of this mitochondrial cycle correlates with the presence of a 16 kb linear Mif (mitochondrial fusion) plasmid in the mitochondrion (Kawano et a f . , 1991; Takano et al., 1991). Most progeny of the Ng isolate contain the plasmid and, when they are mated with another strain, whether it contains the plasmid or not, mitochondrial fusions ensue. The plasmid is inherited by nearly all of the progeny of such crosses, essentially displaying uniparental inheritance, and these progeny in turn transmit the Mif phenotype to their progeny. The Mif plasmid is thus acting like a selfish gene. However, one of the mtDNA species in each Mif cross benefits in that the mitochondrial fusions preserve the mtDNA that otherwise would have been eliminated according to the matA hierarchy (Meland et af., 1991). Hurst (1991) points out that this mechanism fits well with proposals as to how sex may have initially evolved. The DNA transformation system for P. pofycephafum (see Section VI .C) has been developed for integration into the nucleus but, if it can be extended to the mitochondrion, the specific detailed functions of the Mif plasmid would be open to investigation. Another remarkable characteristic of the mitochondrial genome in the slime mould is the editing of RNA for the a-subunit of ATP synthase (Mahendran et al., 1991). Insertion of cytosine residues at 54 sites is required to generate a functional reading frame for the mRNA from the gene. This is the first example of RNA editing by insertion of cytosine residues, contrasting. for example, with insertion of uridine residues in the RNA of Trypanosoma brucei (Feagin et af., 1988). This type of insertional editing observed in organelles of these and other protists appears to be distinct from substitutional editing found in plant mitochondria and in vertebrates (Scott, 1989; Benne, 1990). Given the apparent susceptibility of the nuclear and nucleolar genomes in P. pofycephalum to transposition, usually with no obvious benefit to the host, it is curious that an editing requirement for mtDNA genes has been maintained during evolution. If a transposon

THE PROTIST PHYSARUM POLYCEPIlAI.UM

13

could enter and function i n the mitochondrion, even if only rarely over evolutionary time, one might predict that it would be more efficient and therefore advantageous if the editing system were replaced by retrointegration of an edited transcript.

IV. Cytoskeletal Organization The cytoskeleton of P. polycephalurn has been a major focus of research, covering the roles of tubulin, actin, myosin, titin, profilin and other cytoskeletal proteins (Dove et al., 1986). The plasmodium is a particularly useful source of non-muscle actomyosin and related proteins, since it contains substantial quantities of cytoplasmic actin and myosin, and is easy to culture to the large mass needed for protein biochemistry. As Hatano (1986) adroitly phrased it, “Ten plastic buckets of 10 L each are used for cultivation of surface plasmodia in order to collect 100-200 g of material every two days”. Such prolific growth has spawned far more actomyosin biology than we can reasonably review here. Hence, we review principally the microtubular cytoskeleton, with only passing mention of other cytoskeletal elements. A . MICROTUBULE ORGANIZATION

Microtubules are fibres 25 nm in diameter that are constructed principally from heterodimers of a-tubulin and P-tubulin polypeptides. They are major components of several eukaryotic organelles, including mitotic and meiotic spindles, centrioles, axonemes and the cytoskeleton. Each of these structures is found in one or more cell types in P. polycephalurn. Microtubule organelles are usually organized by distinct organizing centres (MTOCs). The amoeba1 cytoskeletal microtubules in P. polycephalurn radiate from a single MTOC juxtaposed to a centriole pair beside the nucleus. The nucleus+entriole complex can be isolated structurally intact, and retains the capacity to nucleate microtubule assembly in vitro (Roobol et al., 1982). During mitosis in amoebae, cytoskeletal microtubules disappear, the nuclear membrane breaks down, the centriole pair separates and duplicates, and daughter centriole pairs migrate to opposite spindle poles, while the mitotic spindle and associated astral microtubule arrays assemble (Aldrich, 1969; Wright et al., 1980). Mitosis is accompanied by the usual cycle of chromatin condensation and decondensation. Following mitosis and cytokinesis, spindle microtubules disassemble and cytoskeletal microtubules reappear, remaining throughout the interphase. The pattern of amoeba1 mitosis is reminiscent of mitosis in animal cells. However,

14

'1'. (3. IilJHI.ANI) t? A / .

replication of centrioles during amoebal mitosis contrasts with that in animal cells, where centriole replication occurs throughout the cell cycle (Kochanski and Borisy, 1990). The fact that centriole duplication can be limited to mitosis in the amoebal cell cycle raises doubts about the hypothesis that the centrosome is a cog that helps t o couple cell growth with cell division (Railly and Bornens, 1992). Upon development of flagellates, the cytoskeleton reorganizes dramatically. The nucleus moves to the anterior o f the cell, with the associated centriole pair acting as basal bodies for the axonemes. Two flagella (one long and one short) assemble, while cytoskeletal microtubules form a cone around the nucleus and basal bodies (Havercroft and Gull, 1983). Flagellar and cone microtubules are highly organized, with five distinct MTOCs recognized, of which that designated MTOCl is considered to be the same structure which organizes the mitotic spindle (Wright et ul., 1988). While this reorganization of microtubules is occurring, the microfilament system is also substantially reorganized (Pagh and Adelman, 1988; Uyeda and Furuya, 1985). In the plasmodium, the most prominent microtubule structure is the mitotic spindle, present only during t h e synchronous mitoses. In contrast to extranuclear mitosis in amoebae, the plasmodial mitotic spindle is organized by an intranuclear MTOC, and has no astral microtubules (Aldrich, 1969; Tanaka, 1973; Havercroft and Gull, 1983). Plasmodia1 mitosis thus resembles mitosis in fungi, but the intranuclear MTOC in the plasmodium appears distinct from the spindle-plaque type of M'I'OC observed in fungi (Aldrich, 1969). Salles-Passador et ul. (1991) observed cytoskeletal microtubules in the mature plasmodium, contradicting previous failures t o detect cytoplasmic microtubules in this syncytium. So far it is not known whether these microtubules are nucleated by classic M'I'OCs. The function of plasmodial cytoplasmic microtubules is obscure, as the structure of the plasmodium is thought to be determined principally by the microfilament cytoskeleton, while vigorous protoplasmic streaming is thought to facilitate intracellular transport. H . TlJI3UI.IN GENES A N D POLYPEPTIDES

Given the multifunctional role of microtubules, detection of multiple forms of a-tubulin or p-tubulin polypeptides within a single cell type has raised interest in the multitubulin hypothesis, namely the question o f whether different tubulins provide distinct functions for microtubules (Fulton and Simpson, 1976). Physarum polycephalum was the first microbe from which assembly-competent tubulin was purified (Roobol el al., 1980, 1984), and multiple a-tubulin and P-tubulin polypeptides are found in different cell

TI

TABLE I . ~

Gene

PRO

risl

W Y S A R U M POI. Y ( W / / N .

Summary o f expression and utilization patterns of tubulins in f'hysarum polycephalum

~~~

~~

~

~

Expression pattern"

'Tubulin

Amoeba Flagellate alrA

cilA U3

altR(N) ultU(b,') hetA

hetA hetC

15

UM

(llB c12H SIA

PlB P2

+++ +

+++

-

-

+/-

++ -

+t+ -

+++ ++

Plasmodium

+ +++ + + ++ -

-

Utilizationh csk

msp (am) msp (pla)

+

+

-

-

+ + +

+

+ + + + ~

+

fla

+ + +

csk indicates cytoskeleton; msp. mitotic spindle (am. amoeba]; pla, plasmodial); tla, flagellar axonenie and cone. " Expression patterns are deduced from R N A levels o r for polypeptide levels o r for both. Utilization means detection in the specified structure; it docs not mean that the isotype is usually found there. For example, g2 tubulin has been detected in the flagellum on rare occasions. hut it is not normally expressed when flagella are present. Where a matrix element is left hlank, utilization o f the tubuliri has not been tested and cannot he deduced from present data.

'

types (Burland et al., 1983). These advances increased the attraction of the organism to examine the multitubulin hypothesis. Evidence that a- and 0-tubulins are encoded by multiple genes in eukaryotes (for a review, see Sullivan, 1988) also raised interest in the function of different tubulin gene products. Genetic mapping using RFLPs as markers and heterologous tubulin genes as probes reveals four loci in P. polycephalum for a-tubulin, namely altA, altB, altC and altD, and three loci for p-tubulin, namely betA, betB and betC (Burland, 1986). The altB locus comprises two tightly linked a-tubulin ; et al., 1987), but there is no evidence genes (Schedl et a l . , 1 9 8 4 ~Green for multiple sequences at any of the other tubulin loci. Monteiro and Cox (1987a) termed the two linked altB genes Ea-Tu and Na-Tu; we propose combining the original nomenclature of Schedl et al. (1984c), which follows published rules of genetic nomenclature for P. polycephalum (Anderson et al., 1986), with Monteiro and Cox's refinement so that the two alrB genes be referred to as altB(E) (Ea-Tu) and altB(N) (Na-Tu). The polypeptide products of altA, aftR(E),altR(N), betA, betB and betC have all been identified, and complete o r partial sequences deduced either from direct protein sequencing or from sequencing DNA clones. These genes show distinct patterns of expression in different cell types (Table 1). This has prompted various searches for clues as to the reasons for differential expression. Sequence differences among a1A, a l B and a2B

16

'I. Ci. HLIKI.ANI) El A1

TABLE 2.

PplA PplB Pp2B Ngl Cr I SI I Mni I I hI At1 Sp? sc I

Percentage identities of u-tubulin polypeptides in various organisms

PplB Pp2B Ngl

Crl

S11

M m l Dml

At1

Sp2

Scl

Sc3

Spl

91.5

90.6 86.6 87.2 93.1

X9.5 X5.3 X6 92.2 92.2

X6.6 82.1 82.3 85.7 X6.3 X5.3

X5.8 X4.4 84.4 86.6 Xh.9 85.7 79.4 78.9

71.6 71 70.X 71.7 71.9 69.6 76.5 76.3 6X.2

71.9 70.8 69.8 69.3 69.7 68.6 74.7 74.9 6h.X 75.5

71.2 69.X 68.9 69.3

70.3 69.5 69.4 70.4 70.5 69.4 76.3 75.x 67.9 x5.7 76.3 74.5

91.5 96.4

91.3 X7.3 X7.5

8S.X 81.7 X1.7 X5.5 85.8 84.5 96.9

sc3

70 68.2 73.4 74.3 67.1 73.9 90.8

PplA indicates Physarum polycephalum u l A tubulin; PplB, Physarum polycephalum u l B tubulin; PpZB, Physarum polycephulum u2B tubulin; Ngl , Naeglueria gruberi u l tubulin; Crl, Chlamydomonas reinhardtii a1 tubulin; S11, Stylonichia lemmue ul tubulin; Mml, Mus miuculus ul tubulin; Dml , hosophika melanogaster a1 tubulin; At1 Arabidopsis rhuliana ul tubulin; Sp2, Schizosaccharomyces pombe u2 tubulin; Scl, Saccharomyces cerevisiae a1 tubulin; Sc2, S1rccharrJmyce.s cerevisiae a2 tubulin. The sequences were obtained from thc GenBank database.

.

tubulins (Table 2) fall well within the typical range of a-tubulin sequence differences found within a single eukaryote (Singhofer-Wowra et al., 1986b; Cunningham et al., 1993). The a l B and a2B polypeptides, whose genes are closely linked, show greater sequence identity to one another than to ul A. Comparing across species, a-tubulins from P. polycephalum are more similar to a-tubulins from other protists, plants, vertebrates and insects than to those from known fungi (Table 2). The a l B - and a2B-tubulin polypeptides are distinct from other known a-tubulins in having a methionine residue at their C-termini, instead of the more usual glycine residue. In other organisms a terminal glycine residue is thought to be necessary for tyrosination (Gunderson et al., 1987), but neither tubulin tyrosine ligase nor Tyr-tubulin carboxypeptidase has been detected in P. polycephalum . In the 0-tubulin gene family in P . polycephalum, the betA and betB genes encode almost identical pl-tubulin polypeptides even though the two genes differ in 15% of their nucleotide residues (Werenskiold et al., 1988; Paul et al., 1992). By contrast, the 02-tubulin polypeptide encoded by betC differs from the 01 tubulins in 17% of its residues. Comparing across species, p l tubulins in P. polycephalum, like its a-tubulins, are more similar to those from protists, insects and vertebrates than to those from fungi (Fig. 2), but 82 tubulin stands out as being highly divergent in its sequence, showing n o particular similarity to any other P-tubulin (Burland et al., 1988).

THE PROTIST PHYSARUM POI.YCEPIIAI.UM

17

100

g

95

5 '0

90

I

0) c

c

g

85

80

0)

a

75 70

Ppl

Cr Mm5 Mml Gdl Dm2 Tb An112 An3 Sp

FIG. 2. Comparison of P-tubulin polypeptides. The shaded portion of each bar indicates the percentage identity between P2 tubulin from Physarum polycephalum and other P-tubulins. The lighter hatched of each bar indicates the percentage identity from Physarum polycephalum P I tubulin and other P-tubulins. The p2 polypeptide is more diverged in sequence than p l tubulin for all painvise comparisons. Ppl indicates Physarum polycephalum 01 tubulin; Cr, Chlamydomonas reinhardtii P-tubulin; Mm5, Mus musculus PS tubulin; Mml, M u s musculus divergent 01 tubulin; Gdl, Gallus domesticus Pl-tubulin; Dm2, Drosophila melanogaster P2 tubulin; Tb, Trypanosoma brucei P-tubulin; An112, Aspergillus nidulans PI12 tubulin; An3, Aspergillus nidulans divergent p3 tubulin; Sp, Schizosaccharomyces pombe 0-tubulin. The sequences were obtained from the GenBank database. C. TURULIN UTILIZATION

If different tubulin polypeptides possess distinct functional properties, discovering the location of tubulin isotypes in various microtubular organelles may elucidate their functional differences. In P. pofycephafum, it has been possible to distinguish several specific tubulins using a variety of methods.

1 . a l A and a3 Tubulins Although the a l A tubulin from P. polycephalum shows remarkable similarity to u-tubulins from organisms as evolutionarily distant as mammals, it appears to have at least one distinctive sequence characteristic, namely the presence of a lysine residue at position 40 in combination with a tyrosine residue at position 44. Birkett et al. (1985a,b) raised a monoclonal antibody to a-tubulin from P. polycephalum, namely KMP-1, specific for a l A tubulin from the slime mould. The KMP-1 epitope encompasses residues Lys4[, and TyrU in a l A tubulin (Walden et al., 1989a). Western and Northern blotting indicates that ul A tubulin is abundant in amoebae and flagellates, but only small amounts relative t o the other a-tubulins are present in the plasmodium (Birkett et al., 1985a,b; Cunningham et al., 1993).

18

I

(r

llllKl A N D I I A l

So far as we know, altA is the only a-tubulin gene expressed in amoebae and flagellates, and it seems likely that the u3 tubulin found in amoebae and flagellates is an acetylated form of the altA product (Table I ; Green and Dove, 1984; Cunningham etal., 1993). The monoclonal antibody KMP1 does not recognize amoebal a-tubulin when residue Lysa, is acetylated (Walden etal., 1989a), and can therefore detect non-acetylated a l A tubulin in microtubules also populated with acetylated a 3 tubulin. Using immunofluorescence microscopy, a1A tubulin has been found in microtubules in the amoebal and flagellate cytoskeleton, the flagellar axoneme, and amoebal and plasmodial mitotic spindles (Diggins and Dove, 1987; Sasse et al., 1987). Given that u l A tubulin is expressed in the plasmodium, and that expressed tubulin gene products are usually not excluded from any microtubular organelles, it is likely that a l A tubulin is also utilized in plasmodial microtubules. The a3 polypeptide can be distinguished by its unique electrophoretic mobility on two-dimensional gels (Burland et al., 1983). The antibody 6 11B-1 (Piperno and Fuller, 1985), which specifically recognizes an a-tubulin epitope only when residue Lysa) is acetylated, facilitates specific detection of acetylated a-tubulins such as a3 tubulin in microtubular organelles in individual fixed cells. The a3 tubulin in P. polycephalum is present principally in microtubules of the flagellate (Diggins and Dove, 1987; Sasse el a l . , 1987), both in cytoplasmic microtubules of the flagellar cone and in axonemal microtubules of flagella. In amoebae, a 3 tubulin has been detected only in centriole-associated MTOCs, and not in the microtubules of the cytoskeleton or mitotic spindle (Diggins and Dove, 1987; Sasse et al., 1987). Thus, u3 tubulin appears to be located in the more stable classes of microtubules found in P. polycephalum. When amoebae develop into plasmodia, a 3 tubulin is present at early stages in the MTOCs; at later stages of development and in mature plasmodia, a3 tubulin is n o longer detectable (Solnica-Krezel et al., 1990) either by Western blotting o f whole cell lysates or by immunofluorescence microscopy. Since the a1 A-tubulin substrate for acetylation, containing residue Lys4,,, is present in the plasmodium (Cunningham et al., 1993), it would seem either that acetylase is absent or that de-acetylation is highly efficient in the plasmodium. 2. u l B and a2B Tubulins

The a l B - and a2B-tubulin polypeptides, the products of altR(N) and altB(E), respectively, each have a glycine residue at position 44 which precludes KMP-1 reactivity (Walden el al., 1989a). The presence of a l B tubulin, which virtually co-electrophoreses with a1A tubulin, is conveniently inferred by the presence of an a-tubulin that does not react with the

1'11E PROTIST PHYSARUM P O L Y C E P t l A L U M

19

KMP-1 antibody (Birkett et al., 1985a,b), although this criterion runs the risk of mistaking for a l B tubulin the products of other genes, such as altC or altD, whose products are not known. The a2B polypeptide is readily detected by its unique electrophoretic mobility (Burland et al., 1983). Expression of a l B and a2B tubulins appears to be co-ordinate, and has so far been detected only in the plasmodium (Green et al., 1987; Monteiro and Cox, 1987b; Walden et al., 1989b). Isotype-specific polyclonal antibodies reveal the presence of a l B and a2B tubulins in plasmodia (Walden et al., 1989b), and analysis of the tubulins contained within plasmodia1 spindles confirms the presence of a2B tubulin as well as an a1 tubulin (Paul et al. , 1987). 3. (3IA and PIB Tubulins For wild-type amoebae, P1A and (31B tubulins co-migrate on twodimensional gels (Burland et al., 1984). However, the benD210 mutation, which confers resistance to antitubulin benzimidazole drugs, causes a structural alteration in the (3lB-tubulin polypeptide, giving it a unique, altered electrophoretic mobility (Burland et al., 1984). Using this distinction, and isolating cytoskeletons from amoebae and mitotic spindles from synchronous plasmodia, Paul etal. (1989) showed that PlB tubulin is utilized both in cytoskeletal microtubules of amoebae and intranuclear mitoticspindle microtubules of plasmodia. Thus, the betB gene product is found in two classes of microtubular organelles, and in two distinct cellular compartments. Hence, there is no evidence for a specific function for the betB gene product in one class of microtubules. The plA-tubulin isotype is also found in amoeba1 cytoskeletons (Paul et a l . , 1989) in a lower abundance than PlB, consistent with the lower expression of betA relative to betB in amoebae (Table 1). It is not known whether (31A or PlB tubulin is present in the flagellum, although both genes are expressed in flagellates so that both (31-tubulinsprobably have an opportunity to participate in this structure. When tubulin from amoebae from P. polycephalum is assembled in vitro, (31A and (31B tubulins assemble into microtubules with a stoichiometry similar to that in cells whence they were purified (Foster et al., 1987). Moreover, tubulin from amoebae carrying the hen0210 mutation assembles into microtubules in vitro in the presence of antitubulin benzimidazoles to which the mutant is resistant (Foster et al., 1987). Remarkably, the stoichiometry of the mutant (3lB-210and non-mutant (31A tubulins is similar when assembly in vitro occurs in the presence or absence of benzimidazoles (Foster et al., 1987). Thus, even under artificial strong selection in the laboratory, there seems to be no preferential association of one of these two (3-tubulins with assembled microtubules.

20

T C i . H1IRL.AND ET A /

4. P2 Tubulin

The P2-tubulin isotype is distinct in electrophoretic mobility and immunogenicity, allowing analysis of its distribution by both biochemical and immunological techniques. Expression of P2 tubulin, principally in plasmodia and not in amoebae or flagellates (Solnica-Krezel et al., 1990), normally restricts the spectrum of microtubular organelles in which this protein has the opportunity to function. Originally, we believed that this expression pattern restricted P2 tubulin mainly to mitotic spindles in plasmodia (Burland et al., 1988), but detection of p2 tubulin in cytoplasmic microtubules in developing plasmodia (Solnica-Krezel et al., 1990) leads us to question this view. Further, it is likely that P2 tubulin would be used in the cytoplasmic microtubules recently detected in mature plasmodia (Salles-Passador et a l . , 1991). Thus, there are opportunities for P2 tubulin to assemble into several classes of microtubule. Using the distinct electrophoretic mobility of P2 tubulin, Paul et al. (1987, 1989) demonstrated its presence in mitotic spindles isolated from plasmodia. A polyclonal antibody specific for P2 tubulin (Diggins-Gilicinski et a l . , 1989) revealed the distribution of the antigen using immunofluorescence microscopy and Western-blotting experiments. It is now clear that P2 tubulin can assemble in vivo into the microtubules of the cytoskeleton, the astral mitotic spindle, the MTOCs and even flagella (Diggins-Gilicinski et al., 1989; Solnica-Krezel et al., 1990, 1991), as well as into the intranuclear anastral mitotic spindle. Thus, there is no direct evidence that P2 tubulin has a specific role in a particular microtubular function.

D. FUNCTION OF MULTIPLE TURULINS

Acetylated a3 tubulin is the only tubulin isotype in P. polycephalum for which there is evidence of preferential utilization in particular organelles. Its association with more stable microtubules is dramatically preferential in amoebae, where a3 tubulin appears exclusively in the MTOC, and not in cytoskeletal microtubules. In contrast, this isotype is present in cytoskeletal microtubules of the flagellate, as well as in the flagellar axoneme. This distribution of acetylated a-tubulin in different microtubules of the slime mould is analogous to the distribution of acetylated a-tubulin in a wide variety of organisms, where this protein is found in more stable classes of microtubules (Ledizet and Piperno, 1991). It seems that acetylation of the L Y S residue ~ ~ ) is a characteristic that evolved before divergence of major eukaryotic groups, having been observed in members of both the kingdoms Protista and Animalia. The absence from fungi may reflect the restricted mobility functions of these organisms, notably absence of flagella.

THF PROTIST PHYSARUM POLYCEPHALUM

21

The lack of an obvious functional specificity for different tubulin gene products in P. pofycephafum is typical of a wide variety of organisms (Sullivan, 1988). And, as might be expected from the conserved nature of a l A and 81 tubulins from the slime mould, they can assemble in vivo into microtubules in mammalian cells (Prescott et a f . , 1989). However, experiments which examine only assembly of proteins into microtubules do not elucidate whether different microtubular organelles function normally; it is conceivable, for example, that 82 tubulin is functionally deleterious on the rare occasions when it is incorporated into the flagellum or the astral mitotic spindle during plasmodium development (cf. Hoyle and Raff, 1990). Regarding the function of multiple tubulin gene products in a single organism, it is possible to observe certain correlations for the gene family in P. pofycephafum.The betC gene exhibits a distinct sequence that places its product, p2 tubulin, into the class known as divergent p-tubulins; two others in this class are the murine pl tubulin and the p3 tubulin from Aspergiffus nidufans (see Fig. 2). These three (and other) divergent ptubulins are restricted in their pattern of expression to specific cell types, and thereby they are utilized in only a subset of microtubular organelles found in the respective organisms (Burland et a f . , 1988). The murine pl tubulin is expressed principally in erythrocytes, which limits its utilization to the marginal band of microtubules in these cells (Wang et a f . , 1986); the p3 tubulin from A. nidufans is expressed largely in conidiating tissues, where it is utilized mainly in mitotic spindles (Weatherbee et a f . , 1985); and the 02 tubulin from P. pof.ycephafum is expressed principally in the plasmodium, where it is limited to the mitotic spindle and presumably whatever cytoskeletal microtubules are present. Thus, fewer functional constraints may be placed on these divergent P-tubulins, allowing some , In Dr. melanogaster, divergent degree of neutral drift (Burland e f a f . 1988). 03 tubulin is normally expressed at only a low level in a restricted set of cell types (Kimble et al., 1989). When expressed beyond a 20% threshold level in the testis, assembly of microtubules in axonemes is disrupted, indicating that divergent p3 tubulin does not function properly in axonemal microtubules (Hoyle and Raff, 1990). For P. polycephafum,the expression pattern for 02 tubulin leaves this isotype normally absent from the flagellar axoneme, highlighting the view that utilization of a particular tubulin in the flagellar axoneme may be a powerful force for conservation of primary sequence (Little et a f . , 1986; Singhofer-Wowra et a f . , 1986a); this may reflect the greater number of specific protein-protein interactions in the flagellum than in other microtubular organelles. Thus, it may be specifically the absence from the flagellum that permits more neutral drift in 02 tubulin sequence than in tubulins used in the flagellum. Conservation of t h e p1 tubulins in P. polycephafum is consistent with this view. Despite their

22

'I' C i I3lIKI.ANI) F.7' A / . .

distinct expression patterns, both P I A and P1B tubulins are probably utilized in the flagellar axoneme, and the sequences of these two P-tubulins show much more similarity to tubulins in other organisms that possess flagella (or cilia) than they do to P2 tubulin in the slime mould (or t o other divergent P-tubulins). In fact, the a-tubulin gene family in P. pofycephufum mirrors the P-tubulin family, although in a less dramatic way. The more diverged a l B and a2B tubulins are expressed only in the plasmodium, where flagella have not been found, while the more conserved a1A tubulin is the principal, and possibly only, a-tubulin gene product in amoebae and flagellates (Cunningham et a f . , 1993). Conservation of primary sequence among tubulin gene families is notable among other protists that possess flagella (Silflow, 1991). Whether or not distinct tubulin gene products are used for specific functions, the fact remains that multiple a- and P-tubulin genes are the norm among eukaryotes. I t is clear from studies of the tubulin gene family in P. polycephalum, as well as from tubulin gene families from other eukaryotes, that multiple tubulin genes are expressed in distinct patterns in different cell types. The flexibility this arrangement offers for differential gene expression may be a characteristic that was selected for during evolution, resulting in multiple tubulin genes for most eukaryotes (Raff. 1984; Paul el al., 1992). f:.

MICROTURUI.E-ASSO<'IATEI) PROTEINS

The lack of evidence for distinct tubulin gene products determining specitic microtubular functions points to other potential molecules f o r this role, notably microtubule-associated proteins (MAPs; Olmsted, 19x6). Microtubule-associated proteins were originally defined as those that copurify when mammalian brain tubulins are purified by their capacity for self-assembly in vitro (Borisy ef a f . , 1075). When tubulins from P. polycephalum were first purified in this way, the high molecular-weight MAPs observed in brain-tubulin preparations were absent, although smaller proteins with molecular weights of 49,000,57,000 and 59,000 were observed (Roobol ef u f . , 1980). Phosphocellulose chromatography of the assemblypurified tubulin from P. polycephalum eliminates its self-assembly capacity, which can be restored by addition of brain MAPs, suggesting that those from the slime mould are present in the assembly-purified material and are removed, like brain MAPs would be, by chromatography (Roobol el a l . , 1980). Albertini et al. (1990) found six candidates for higher molecular-weight MAPs after adding the taxol analogue baccatine to the concentralcd amoeba1 extracts used for tubulin purification. One such MAP resembles

THE PROTIST PHYSARUM POLYCEPHALUM

23

kinesin, while another, a 125 kDa polypeptide, promotes microtubule assembly in vitro from mammalian brain tubulin and protects microtubules from disassembly upon dilution. The 125 kDa MAP is, like certain mammalian brain MAPs, heat-soluble, suggesting conservation of more than simply microtubule-binding capacity. However, this 125 kDa polypeptide does not cross-react with antibodies to the well-characterized mammalian brain MAP2 and r proteins (Albertini et al., 1990). Nevertheless, now that MAPs have been clearly identified in P. polycephalurn, more detailed analyses of their structures and functions will be possible, and elucidation of the evolutionary origins of these important proteins should follow. F. THE CYTOSKELETON IN DEVELOPMENT

The microtubule cytoskeleton is remarkably different in amoebae, flagellates and plasmodia. It is of interest to determine how cytoskeletal changes occur during developmental transitions between these cell types, not only for an understanding of cytoskeletal dynamics, but also to gain insights into fundamental developmental processes. Protists like P. polycephalurn are amenable to these studies because individual transitions generally occur by conversion of an entire population from one cell type to another; dissections or cell-separation schemes are not needed to determine specific developmental changes. With P . polycephalurn, simple manipulations of culture conditions can be used to induce either the amoeba-flagellate transition or the amoeba-plasmodium transition.

I . The Arnoeba-Flagellate Transition The amoeba-flagellate transition is induced in the laboratory by suspending amoebae in non-nutrient solutions, and can be reversed by returning cells to a nutrient medium or on to a solid substrate. The distinction between amoeba and flagellate in P. polycephalurn is well defined in terms of cytoskeletal architecture, as seen by both immunofluorescence and electron microscopy (e.g. Aldrich, 1968, 1969; Pagh and Adelman, 1988; Havercroft and Gull, 1983; Wright et al., 1988). Uyeda and Furuya (1985) made a detailed immunofluorescence study of cytoskeletal changes during this transition using antitubulin antibody to detect microtubules, and nitrobenzdiazole-phallacidin to detect microfilaments, in the same cells, simultaneously. They recognized seven distinct stages in the developmental sequence amoeba --+ flagellate-, amoeba, compiling the sequence of events from analysis of individual cells fixed for microscopy over the time-course of development (Fig. 3). In most cells, the first major events are migration

24

'I' C;

R1IRI.AND ET A /

FIG. 3. The amoeba-flagellate and flagellate-amoeba transitions. Shading in thc cytoplasm indicates regions showing immunofluorescence staining with nitrobenzdiazole-phallacidin (microfilaments) and curved lines indicate antitubulin antibody flurocscence (microtubules). MTA 5 indicates microtubular array 5 . Redrawn from Uyeda and Furuya (1985).

of the nucleus to the anterior end of the cell, concomitant with development of the anterior cone of microtubules; this is usually followed by extrusion of flagella (Fig. 3). The nuclear migration is sensitive to microfilament poisons but not to microtubule poisons, suggesting that migration is mediated by an actin-generated force (Ohta el al., 1991). This contrasts with nuclear migration in fungi, which is sensitive to microtubule poisons, and requires functional tubulin (Oakley and Morris, 1980). Analysis of P. polycephalum strains with abnormal flagellar organization suggests that the force for nuclear migration acts on the centrosome, which is connected to the nucleus, rather than on the nucleus itself (Ohta et al., 1991). Around the time of flagellar extrusion, a prominent ridge of microfilaments develops

TllE PRO’rIST PHYSARUM POI.YCBPHAI.UM

25

around the dorsal surface of the cell; the function of the ridge is not clear, and it disappears by stage 5, which Uyeda and Furuya (1985) consider the mature flagellate (Fig. 3). Another interesting structure observed during the amoeba-flagellate transition is the “backbone”, which on lightmicroscope resolution appears to be made up of coincident bundles of microtubules and microfilaments (Fig. 3). In contrast to the microfilament ridge, this backbone persists in the mature flagellate. Pharmacological and electron-microscopic evidence suggests that movement of the backbone structure involves interaction between microfilaments and microtubules, and Uyeda and Furuya (1989) propose that ATP induces reciprocal sliding between microtubules and microfilaments in the backbone. The precise function of the backbone structure, however, remains to be determined. Along with increases in acetylation of a l A tubulin to make a3 tubulin, substantial increases in levels of tubulin RNA occur during the amoebaflagellate transition in P. polycephalum (Green and Dove, 1984; Paul et al., 1992; Cunningham et al., 1993), as occurs in algae after loss of flagella (Lefebvre and Rosenbaum, 1986). For a- and (3-tubulin, a 5-7 fold increase in message occurs (Green and Dove, 1984); for a-tubulin, the increase is mostly or entirely in altA transcript, since altA is the principal or only atubulin gene expressed (Cunningham et d . , 1993). But for (3-tubulin, two genes are expressed, namely betA and betB (see Table l ) , with the level of betB transcript remaining approximately constant while that of betA transcript increases 100-fold from an initial low level (Paul et al., 1992). The burst of tubulin expression during the amoeba-flagellate transition is a pulse rather than a sustained increase. Before all of the cells in the population possess mature flagella, the level of altA transcript is already on the decline (Paul el al., 1992). The level of actin transcript declines only slightly during flagellate development, contrasting with flagellate development in the protist Naeglaeria gruberi, where a dramatic fall in the level of actin transcript occurs (Lai et al., 1984). The amoeba-flagellate transition is reversible, in that flagellates revert to amoebae, although the sequence of events in the flagellate-amoeba transition is not a simple reversal of amoeba-flagellate transition. Rather, distinct morphologies are detectable with respect to organization of microtubules and microfilaments (Fig. 3; Uyeda and Furuya, 1985). The flagellar axoneme retracts into the cytoplasm, a process that takes only a few seconds (Glyn and Gull, 1990), and appears to persist there for some time. Judging from immunolluorescence, de-acetylation of microtubules takes perhaps 45 minutes (Glyn and Gull, 1990). The amoeba-flagellate transition depicted in Fig. 3 shows a discrete pattern of development by progressive morphological change. However, Uyeda and Furuya (1985) noticed that, even after extended incubation in

26

r

G 13lJRI.AND E T A / .

a non-nutrient solution, a few cells in populations undergoing the transition were in stages 3 and 4, rather than the expected stage 5 (Fig. 3). They suggested that, while this could mean that some cells stay in these intermediate stages for longer times than others, it could instead reflect reversibility of the developmental sequence: stage 3 -+ stage 4 -+ stage 5. Further, in analysing individual cells for their microtubule and microfilament structures during the transition, Uyeda and Furuya (1%) noted that some cells did not show the expected characteristics for the developmental sequence they described. Pagh and Adelman (1988) similarly detected apparent plasticity in changes in the cytoskeletal pattern during the transition. For example, under optimum conditions for flagellate development, occasional cells can be found from which flagella are extruded but the microtubule cone around the nucleus is not formed, inconsistent with the sequence of events shown in Fig. 3 in which the cone is formed before flagellar extrusion (Uyeda and Furuya, 1985). One interpretation is artefactual loss of the cone during fixation. However, another explanation is that, although development progresses by the same sequence of events in most cells, it follows alternative sequences of events in a minority of cells. Where such alternatives are not substantially detrimental, variability in development may be a characteristic of natural populations. In the future, it would be interesting to ascertain expression patterns for other cytoskeletal-protein genes during the transition; probes for both myosin (Kobayashi et al., 1988) and profilin (Binette et al., 1990) are available and could be used to determine expression of these genes. Further elucidation of functional relationships between cytoskeletal changes and flagellar motility should result from analysis of conditional mutants defective for the transition (Mir et al., 1979), and from disruption of individual members of cytoskeletal gene families, a technique that has recently become feasible in P . polycephalum (see Section V1.C).

2. The Amoeba-Plasmodium Transition The development of amoebae into plasmodia, the amoeba-plasmodium transition, is a dramatic transition in cell type in the life cycle of P . polycephalum. The developmental changes in the amoeba-plasmodium transition are most readily observed in apogamic strains, where cinematographic analysis indicates that a single haploid amoeba can develop into a plasmodium without cell or nuclear fusion (Anderson et al., 1976; Bailey et al., 1987). The amoeba-plasmodium transition is induced in the laboratory by growing apogamic amoebae to a high cell density on lawns of bacteria.

27

IIIE PHOTIST PtfYSARUM POI.Y('EPtiA/ 11M

Developing cells can be fractionated on glass-bead columns to obtain a population of largely uninucleate cells committed to development (Blindt el a l . , 1986) but, under optimum culture conditions, synchrony of development is satisfactory for most applications (Solnica-Krezel el al. , 1988). A diffusible inducer of plasmodium development is released at a high cell density (Youngman el al., 1977; Nader et a l . , 1984), but it has not yet been characterized. The sequence of changes related to the cytoskeleton during the amoeba-

A i-

10h

LF -I

CC

UD

B

P

npfAl npfG

-ksssi

npfF1 npfMl npfLl npfK1 npfGlnpfFl npfK1npfLl

FIG. 4. The amoeba-plasmodium transition. The shaded boxes at the top indicate levels of a3 and 82 tubulins in microtubules as detected by immunofluorescence. Below are shown cellular events in the transition: A , amoeba, with nucleus and microtubules; LF, stage at which cells lose the ability to undergo the amoebaflagellate transition; CC, stage of cell commitment to plasmodium development; UD, uninucleate developing cell that has gained the ability to ingest amoebae and cysts; B, binucleate cell; P, young plasmodium. The double-headed arrow defines a time interval of 10 hours for development of strain CL, though timing of events in other strains may vary. The shaded boxes at the bottom indicate stages of development at which single and double mutants arrest development.

28

'I G. RURI.ANI) E?' AI.

plasmodium transition in the apogamic strain CL is summarized in Fig. 4. Initiation of apogamic development is thermosensitive, but, once initiated, development can proceed at any growth temperature. The first detectable change after the thermosensitive period is loss of the ability to undergo the amoeba-flagellate transition (Blindt et al., 1986), which occurs about one amoebal intermitotic time after the previous cell division. The next key event detected is commitment to development, the stage beyond which cells develop into plasmodia even when removed from the high-density conditions and plated individually (Youngman et al., 1977). Commitment occurs in uninucleate cells of strain CL (Burland et al., 1981) about 3-5 hours after loss of the ability to flagellate (Bailey et al., 1987), but before major changes in the cytoskeleton are evident (Solnica-Krezel et af., 1990, 1991). 'The timing of the commitment event places it on average around the middle of the extended cell cycle that distinguishes the first developmental cell cycle from the typical amoebal cell cycle (Bailey et al., 1987; Fig. 4). The time of commitment corresponds to the stage at which wildtype cells acquire the ability to grow axenically (Blindt et al., 1986), one of the first clear physiological characteristics of the plasmodium. Around the same time, uninucleate developing cells acquire other plasmodial characteristics, including ability to ingest amoebae (Solnica-Krezel el af., 1990). In heterothallic myxomycete development, cell fusion appears to define the point of commitment (Shipley and Holt, 1982), and flagellates as well as amoebae of compatible mating types are capable of cell fusion (Ross, 1957;Bailey et af.,1990). Thus, in heterothallic development, loss of ability to flagellate occurs no sooner than the time of commitment, rather than some hours before as in apogamic development of strain CL. Heterothallic amoebal cell fusion can occur at any stage in the cell cycle, and the two cells that fuse need not be at the same stage (Bailey et al., 1990). Nevertheless, an extended cell cycle is observed after cell fusion, with nuclei fusing about two hours after the amoebae fuse; subsequently, events are remarkably similar to apogamic development (Bailey et al., 1990). The extended developmental cell cycle is over twice as long as an amoebal cell cycle, for both apogamic and heterothallic development (Bailey et al., 1987, 1990). At the end of this cell cycle, the uninucleate developing cell, substantially larger than an amoeba, is ready for mitosis. A key question is when during the amoeba-plasmodium transition the pattern of mitosis switches from the amoebal to the plasmodial type, since this is an essential change for sustaining the multinucleate state. One difficulty in analysing this problem is that it is impossible to determine whether a cell is committed simply by looking at it under the microscope, since commitment is defined operationally as the point beyond which cells will develop into plasmodia,

THE PROTIST PHYSARUM POLYCEPHALUM

29

rather than amoebal colonies, when plated at a low density (Youngman et al., 1977). However, by correlative studies of strain CL, it is clear that p2 tubulin can, with minor exceptions, first be detected in microtubules just after the time of the commitment event (Fig. 4; Solnica-Krezel et al., 1990), so that the presence of p2 tubulin in microtubules can be used with reasonable reliability to determine whether a cell is committed to development. Thus, mitotic spindles in most committed cells should contain 82 tubulin. Furthermore, the extranuclear MTOC that nucleates the amoebal mitotic spindle contains a3 tubulin but the MTOC that nucleates the intranuclear mitotic spindle of the mature plasmodium does not (Diggins and Dove, 1987; Sasse et al., 1987). Thus, a3 and 82 tubulins can be used as developmental markers to investigate changes in mitosis during development (Fig. 4). Using these criteria, four types of mitosis can be observed in committed cells (Solnica-Krezel et al., 1991). These are: (a) Amoeba1 mitosis, which in developing cells is similar to mitosis in normal amoebae, except that p2 tubulin can be present. The spindle is organized by extranuclear, a3 tubulin-positive MTOCs, the spindle poles radiate astral microtubules, and cytokinesis usually follows mitosis. (b) Plasmodia1 mitosis occurs within the nuclear membrane, is organized by intranuclear MTOCs that lack a3 tubulin, and is anastral. However, in developing cells, extranuclear MTOCs may also be present, nucleating cytoplasmic microtubules, while some of these MTOCs divide in the usual manner during mitosis, even though they take n o part in the spindle. Cytokinesis does not occur. (c) Chimeric mitosis is characterized by an astral mitotic spindle nucleated by MTOCs that apparently lack a3 tubulin. Microtubuleorganizing centres that are a3 tubulin-positive are usually present outside the spindle, nucleating cytoplasmic microtubules; these additional MTOCs often appear to block cytoplasmic furrow development, leading to a cell with only a partial furrow, probably impairing cytokinesis. Chimeric mitosis probably results from initiation of intranuclear mitosis followed by nuclear-membrane breakdown during mitosis (Solnica-Krezel et al., 1991). (d) Star microtubule structures, which may be defective spindles, consist of a condensed mass of chromatin at the centre of a radial array of microtubules, and usually contain two a3 tubulin-positive MTOClike structures near the middle of the array. As a population of cells develops, the frequencies of the four different patterns of mitosis change (Solnica Krezel et al., 1991). Early in development, only amoebal mitoses are observed but, when committed cells first

30

T G HIJRI.AND ET A I .

appear in the population, plasmodial and star mitoses also become apparent. Chimeric mitoses have not been detected in amoebae, and they remain rare throughout development. The frequency of plasmodial and star mitoses continues to increase, while the frequency of amoebal mitoses decreases as the proportion of committed cells increases in the developing population. Eventually, when the binucleate stage of development is reached, all subsequent mitoses appear to be of the plasmodial type (Solnica-Krezel et al., 1991). It seems likely that both the plasmodial and amoebal types of mitosis, and probably also chimeric mitosis, produce viable somatic progeny, but the fate of cells undergoing star mitoses is unclear. Star mitoses comprise up to 20% of mitoses at one stage of the amoeba-plasmodium transition (Solnica-Krezel etal., 1991) so, if this is a non-viable event, it is a significant cost of development. Such loss may be analogous to elimination of abnormal nuclei during development of embryos of Dr. melanogaster (Sullivan et u l . , 1990). From the binucleate-cell stage onwards, the young plasmodia of apogamic strain CL increase in mass not only by growth but also by fusion with other multinucleate (and, rarely, large uninucleate) cells. Such cell fusions are a clear plasmodial characteristic, under control of three fus loci (Poulter and Dee, 1968). In contrast to the action of the matB gene, where allelic difference promotes cell fusion, similarity at the fus loci is required for plasmodial fusions. Although cytoplasmic microtubules are readily detectable in young plasmodia, the MTOCs that nucleate them no longer contain a 3 tubulin and, as the cells increase in size, the MTOCs are barely apparent, while cytoplasmic microtubules become more diffuse. The frequency changes and nature of the different classes of mitosis in developing populations of strain CL suggest that changes in organization of the mitotic spindle occur gradually during the amoeba-plasmodium transition, probably reflecting gradual loss of factors required for amoebal mitosis and gradual accumulation of factors required for plasmodial mitosis. The fact that different types of mitosis are found at the same time in a population of developing cells in part may reflect asynchrony of the transition, but biological variability also probably contributes to these differences. For example, if commitment does not always occur at the same point in the cell cycle, then commitment-dependent accumulation of plasmodial factors would vary between different cells by the time mitosis occurs. The switch from amoebal to plasmodial mitosis raises a key question as to the organization of the two types of spindles: are the spindle organizers modified forms of the same basic structure or are they independent structures? Wild-type haploid amoebae carry one major MTOC, termed

THE PROTISI' PHYSARUM POLYCEPHALUM

31

MTOC 1, which nucleates cytoskeletal microtubules during interphase and microtubules of the spindle during amoebal mitosis. Some diploid amoebal strains, constructed by abortive crossing between, for example, matAZmatBI and matA1-matB3 strains, carry two or three of these MTOCs (Wright et al., 1988). The presence of the extra MTOCs correlates with an increased frequency of normally rare multipolar mitoses, and the ranges of the frequencies are distinct for strains carrying one, two or three MTOCs. Akhavan-Niaki et al. (1991) crossed diploid amoebae carrying different numbers of MTOCs and examined mitoses in the resulting tetraploid plasmodia for multipolar frequency, using this frequency as a proxy for the number of spindle-organizing centres. These indirect data are consistent with the hypothesis that crossed plasmodia contain the same number of spindle-organizing centres present in one or other of the amoebal parents, and neither the sum nor the average of the two. Selfed, haploid wild-type amoebae with one MTOC also give rise to plasmodia that contain a single spindle-organizing centre, but Akhavan-Niaki et al. (1991) observed that one strain of amoeba with three MTOCs, when selfed, gave rise to plasmodia with frequencies of multipolar mitoses typical of strains with three MTOCs. These results are consistent with structural continuity of MTOCs throughout development. Alternatively, a similar number of MTOCs in amoebae and plasmodia could simply indicate that a mechanism that regulates their number is conserved in amoebae and plasmodia. This issue has also been addressed by light-microscope observation of MTOCs during the amoeba-plasmodium transition. For heterothallic development, fusion of two amoebae results in a cell with two nuclei and two major MTOCs (Bailey et al., 1990). Subsequently, the number of major cytoplasmic MTOCs observed in each nucleus falls, with most zygotes containing one or none. Among binucleate developing plasmodia, 20% contain two MTOCs (i.e. one per nucleus), 22% contain one MTOC and 58% contain none (Bailey et al., 1990). These data suggest that, at the light-microscope level, the major cytoplasmic MTOCs, which in amoebae reorganize to nucleate the spindle at mitosis, are lost gradually during development, and indicate that, in many cells, all of them are lost. Thus, structural continuity of amoebal and plasmodial spindle-organizing centres cannot be demonstrated from these data. Obviously, an MTOC is needed in these cells that apparently lack one, to organize the spindle at mitosis, but, as in the mature plasmodium, this would appear to be a structure which has not been detected during the interphase (Paul et al., 1987). Is it possible that the organizer for the intranuclear plasmodial spindle is synthesized de novo for each mitosis? Now that y-tubulin has been established as a molecular marker for MTOCs, it would be worth seeking the plasmodial spindle-organizing centre with anti-y-tubulin antibodies.

FIG. 5. Variable pathways of development in the amoebqlasmodium transition. Alternative pathways of development between the committed cell stage and binucleate cell development, deduced from SolnicaKrezel et al. ( 1 9 9 0 , 1 9 9 1 ) . Bold arrows indicate more common pathways. while dotted arrows indicate tentative pathways. Microtubuleorganizing centres (MTOCs) are indicated by the positions where microtubules (curved lines) focus.

The alternative pathways indicated can explain how binucleate cells with one, two or three MTOCs arise. A binucleate cell with no detectable MTOC would arise from plasmodia1 mitosis in a cell lacking a cytoplasmic MTOC. P indicates that plasmodium development is expected to be successful; ?, unknown fate. From SolnicaKrezel et al. ( 1 9 9 1 ) .

- P

- P

STAR

STRUCTURE

TIIE PROTIST PHYSARUM POLY(’EPHALUM

33

However, in a variety of cells, anti-y-tubulin immuno-staining of MTOCs is most intense at mitosis (Zheng et al., 1991), so the disappearing plasmodial organizing centre may yet remain cryptic. For apogamic strains, there appears to be only one major MTOC in each uninucleate developing cell but, in binucleate cells, there can be as many as three MTOCs, depending upon the type of mitosis that gives rise to the binucleate cell (Fig. 5 ; Solnica-Krezel et a f . , 1990, 1991). Clearly, control over the number of organizing centres is not tightly regulated during the amoeba-plasmodium transition. However, it is possible to detect a3 tubulin-containing, extranuclear MTOCs of the amoebal type in the same cell that contains intranuclear, a3 tubulin-negative spindle-organizing centres of the plasmodial type (Solnica-Krezel et a f . , 1991). This is difficult to reconcile with a proposal (Akhavan-Niaki etaf.,1991) that the plasmodial spindle-organizing centre arises from the major amoebal organizing centre MTOC1, unless only a subset of amoebal MTOCs act as progenitors for the plasmodial one. It is clear that development of amoebae into plasmodia can occur via different pathways (Fig. 5 ) . The pathway shown in Fig. 4 appears to be the most common, but other pathways are possible. For example, cytoplasmic MTOCs may or may not contain a3 tubulin by the binucleate cell stage of development. Perhaps more remarkable is the option for a developing cell to undergo both karyokinesis and cytokinesis before binucleate cell development occurs, rather than completing the extended cell cycle by forming a binucleate cell (Solnica-Krezel et a f . , 1991; Bailey et al., 1992). Whether the mitosis that gives rise to a binucleate cell is of the amoebal or plasmodial type appears to depend on the status of numerous developmental variables at the time mitosis is initiated. Variation in the length of the cell cycle itself may account for much of the variability observed in the pathway for cytoskeletal changes. Such variability is reminiscent of occasional variations seen during the amoeba-flagellate transition, implying that the asignment of a strict deterministic sequence of events to execute a specific developmental process may be inappropriately rigid. G . THE CYTOSKELETON IN DEVELOPMENTAL MUTANTS

1. Mutations Linked to matA

Isolation of haploid mutant apogamic (selfing) strains from heterothallic strains facilitates selection for mutants defective in plasmodium development (Wheals, 1973; Cooke and Dee, 1974). The most commonly used apogamic mutation, obtained from a matA2 strain, is gadAh (gad, greater asexual differentiation). The gadAh mutation has not been separated from

34

1 G HURI.AND E l A I .

matA by recombination (Anderson et al., 1989). Names of loci identified through mutants of gadA strains that no longer self are usually referred to as npf (no plasmodium formation). In early genetic screens, the most common developmental block detected resulted from mutation in one of two complementation groups, namely npfB or npfC, tightly linked to, or perhaps integral units of, matA (Anderson et al., 1989). The mating specificities in most npf mutants appear unchanged, despite the fact they were isolated through the two-step mutational pathway heteroihallic -+ apogamic -+ Npf. This suggests that npfB and npfC encode functions separate from the mating specificity function of matA (Anderson et a l . , 1989). Similar barriers to mutational changes in mating specificity are apparent in basidiomycetes (Metzenberg, 1990). Mutations identified in npfB and npfC appear to prevent initiation of development; cells of these strains cultured under conditions favourable to development remain indistinguishable from amoebae.

2. Mutations Unlinked to matA The first developmental mutation isolated from a gadAh strain was apt-I (Wheals, 1973), now called npfFl, which is unlinked to matA. Like most other developmental mutations, npfFZ is defective for development under all culture conditions. Since gadAh npf + development is thermosensitive, failing to occur at 30°C or above, it is difficult to isolate mutants thermosensitive for the amoeba-plasmodium transition. However, Anderson and Dee (1977) were able to isolate one mutation, npfAI, which is thermosensitive for the apogamic transition. This mutation is unlinked to matA and is the only mutant allele known for the npfA gene. Using modified screening methods to eliminate the original bias towards selecting mutants blocked early in development, several more genes essential for apogamic development have recently been detected, denoted npfD through npfM, all of which are unlinked to matA (R. W. Anderson, personal communication; Solnica-Krezel et al., 1992; Bailey et al., 1992). The changes that occur in the microtubule cytoskeleton during the amoeba-plasmodium transition are useful cellular markers for the stage of development at which mutants arrest or go wrong. Solnica-Krezel et a f . (1992) used phase-contrast and immunofluorescence microscopy to analyse the effects of mutations in the npfA, npfF, npfG, npfK, npfL and npfM genes on microtubule organization during the amoeba-plasmodium transition. The results are summarized in Fig. 4. a. The npfAZ Mutant This mutant arrests development at the uninucleate stage and does not lose the ability to form flagellates. Many cells encyst, while mitotic spindles

I H E PROllST PHYSARUM POLYCEPHALUM

35

among active cells all appear to be of the amoebal type. Moreover, p2 tubulin is only sporadically detected in microtubules, and its appearance probably reflects incomplete penetrance of the mutation (Solnica-Krezel et al., 1992). Clearly, npfAl is blocked at a very early stage of development. b. The npfGI, npfC2 and npfC3 Mutants These mutants also arrest development at the uninucleate stage, and retain the ability to flagellate. Star-microtubule arrays are occasionally observed in npfC cells during abortive development, although mitotic spindles formed are mainly of the amoebal type and rarely if ever contain detectable levels of p2 tubulin. Thus, npfC mutants appear to be blocked early in development. c. The npfFI Mutant Many npfFl cells encyst during the abortive development, but some enlarge and a few become binucleate. Some cells lose the ability to form flagellates, some express p2 tubulin, and some lose detectable a3 tubulin from MTOCs. A few of the mitoses observed are of the plasmodial type, indicating that npfFl cells arrest later in development than cells with npfA and npfC mutations. Curiously, occasional npfFI cells can be observed, under immunofluorescence optics, to contain thick, a3 tubulin-positive microtubules, a structure generally observed neither in amoebae nor in plasmodia. d. The npfMl Mutant Development in the npfMl mutant progresses as far as small multinucleate cells, many of which have lost a3-tubulin staining and express p2 tubulin. Mitotic spindles of the plasmodial type are readily detected in both uninucleate and binucleate cells, suggesting that development progresses to an advanced stage before arrest. Cinematographic analysis shows that developing npfM1 cells pass through an extended cell cycle, as d o npf* apogamic strains, although cell fusions have not been observed in npfMl cells, as are observed in development of npf cells by the stage at which the npfMI mutation arrests (Bailey et al., 1987). +

e. The npfLI Mutant This mutant also develops to the stage where a 3 tubulin is lost and p2 tubulin is present in microtubules, although abnormalities become evident

36

T Ci. MUHI.AND 6.Y Al..

at the uninucleate stage. Spindles either of the amoebal or plasmodial type can be detected, but many of the mitoses are abnormal, including not only the star microtubule arrays seen in the wild type but also some morphologies peculiar to the npfL1 mutation (Bailey et al., 1992). A remarkable nuclear abnormality involves the tight condensation of chromatin, which cinematography suggests occurs after npfLl cells pass through the extended cell cycle. Although this abnormality is seen occasionally in development of npf cells, it is much more common during development of npfLl cells. The abnormal cells can grow quite large, but nuclear organization is clearly defective (Bailey et al., 1992). Many npfL1 cells round up and begin to pulsate, a characteristic that appears as a “boiling” phenotype when viewed cinematographically at high speed (Bailey et al., 1992). Like some npfF1 cells, microtubules in npfLl developing cells often appear under immunofluorescence to thicken dramatically. These abnormal processes appear to result in cell death. This phenomenon may represent a protistan version of apoptosis, an active process long recognized in animal cells, where cell death proceeds through a distinct pathway of morphological changes, including the “boiling” phenotype and disintegration of the nucleus (Bailey et a l . , 1992). +

f. The npfKl Mutant Development in cells with the npfK1 mutation proceeds to the binucleate cell stage and beyond, with loss of ability to flagellate, loss of a3 tubulin, and appearance of 02 tubulin. Mitosis of the plasmodial type is found in both uninucleate and binucleate cells, and cells with many more nuclei can be detected. However, the cells are morphologically abnormal, and larger cells take on a stringy appearance. Thus, these cells progress farther in development than other mutants. In addition to focusing on the effects of lesions in specific genes, this work provides information on mechanisms of the transition from the amoebal to the plasmodia1 type of mitosis. Three aspects of cytoskeletal organization, namely activation of P2-tubulin expression, loss of a3 tubulin, and formation of plasmodial mitotic spindles, were not separated by any of the mutations studied. It is therefore possible that these three events are regulated co-ordinately. Mutants such as npfK1 and npfL1, which appear to execute several developmental functions successfully, indicate that development occurs on parallel pathways, some of which can continue when others fail. Analysis of double mutants is consistent with this interpretation. For example, cells of the npfFl npfCl double mutant arrest development at a stage characteristic of npfC cells (Fig. 4), suggesting that npfC function is

THE PROTIST PHYSARUM POLYCEPHALUM

37

executed on a pathway on which npfFfunction is dependent (Solnica-Krezel et al., 1992). By contrast, the npfK1 npfLZ double mutant arrests with characteristics of both npfKf and npfLl single mutants, suggesting that the two genes function on parallel rather than dependent developmental pathways. Much of the variation observed in the staging of events in the amoeba-plasmodium transition may reflect flexible timing of events that occur on parallel pathways, while events occurring on the same dependent pathway presumably occur in a stricter sequence. Control of development by both dependent and parallel pathways of events is analogous to mitotic-cycle control (Hartwell, 1991). It is thought that parallel pathways in the mitotic cycle must eventually be integrated at key check-points (e.g. Nurse, 1991). Variations on pathways of development during the amoeba-plasmodium transition appear to originate during the extended developmental cell cycle. This may reflect a paucity of developmental check-points during most of the cycle, with key checkpoint(s) being located later in development. H. OTHER GENES DIFFERENTIALLY EXPRESSED IN DEVELOPMENT

Amoebae and plasmodia show distinct patterns of expression of many genes. It has been suggested that amoebae and plasmodia largely express “different sets of genes” (Gingold et al., 1976; Wheals et al., 1976), based on the observation that plasmodia derived from thermosensitive mutant amoebae did not appear to be thermosensitive and, likewise, thermosensitive mutant plasmodia arose from amoebae that did not express a thermosensitive phenotype. More careful analyses indicate that, in fact, most thermosensitive mutations express a mutant phenotype in both amoebal and plasmodial cell types (DelCastillo et al., 1978; Burland and Dee, 1979, 1980), suggesting that only a minority of genes are expressed in a cell type-specific manner. Confirmation that only a minority of genes show distinct expression patterns between amoebal and plasmodial phases of the life cycle comes from analysis of cDNA libraries representing the more abundant amoebal or plasmodial mRNAs. Only 5% of the cDNAs from either cell type represent mRNAs specifically expressed in only one of the cell types (Pallotta et al., 1986; Sweeney et al., 1987). Switches to turn off amoeba-specific gene expression and to turn on plasmodiumspecific gene expression patterns occur during the first few cell cycles during the amoeba-plasmodium transition (Sweeney et al., 1987). Although there are five actin genes at four unlinked loci in P. polycephalum (Schedl and Dove, 1982), the ardA, ardB and ardC loci are isocoding, and most of the actin transcripts in amoebae and plasmodia derive from the ardB and ardC genes (Hamelin et al., 1988). However, a

38

T. G . BURLAND ET A1

fourth actin gene, urdD, encodes an actin that is only 84"/" identical to the ardB and ardC actin products, which is a low identity given that the isocoding urdB and ardC products are 95% identical to human cytoplasmic actin (Adam et al., 1991). Expression of the ardD gene has been detected only at a very low level in plasmodia, at a higher level in encysting plasmodia, and not at all in amoebae. In addition to genes for actin and tubulin, other cytoskeletal genes expressed in a cell type-specific pattern include profilins (Binette et ul., 1990) and myosins (Uyeda and Kohama, 1987). The sequences of the two profilin genes, proA (amoebal) and p r o p (plasmodial), are so distinct that it would be difficult to detect one with the other by nucleic acidhybridization techniques. This may explain why this was the first occasion on which cell type-specific differences in profilin genes were detected in any organism (Binette et al., 1990). The prevalence of cell type-specific cytoskeletal elements contrasts with the general interpretation of analyses of thermosensitive mutants, cDNA expression studies and two-dimensional electrophoresis of proteins (Turnock et al., 1981) taken to indicate that most genes expressed in amoebae are also expressed in plasmodia, and vice versa. Distinct expression patterns for cytoskeletal genes may be related to the dramatically different cellular organization and developmental potentials of amoebae and plasmodia. The approach of screening cDNA libraries made from a particular cell type of P. polycephalum has also been successful in identifying cell typespecific genes for sporulating plasmodia (Martel et al., 1988) and for encysting plasmodia (Bernier et al., 1986, 1987; Savard et al., 1989). I . INFERENCES

Different analyses of the amoeba-plasmodium transition have given several insights into protist development, some of which may also be relevant for mechanisms of development in species from other kingdoms. First, the maintenance of different cell types involves differential expression of a substantial number of genes, albeit a small fraction of the total number. Second, a key regulator of the developmental transition between amoeba and plasmodium is the matA gene, including the tightly linked complementation groups gadA, npfB and npfC. Third, the mat and npf genes usually act in a consistent sequential order: matB and matC appear to function before matA, while the npf genes execute their developmental functions after rnatA. Fourth, the fact that amoebae and plasmodia of the npfA, npfB and npfC mutants appear perfectly normal for all functions except selfing suggests that some key genes may function specifically during development. This suggestion will be directly testable when the npf genes

1 I IF P R O l I X T P f / YSARllM 1'01. Y C E Pf / A /.UM

39

in question are cloned. Fifth, the abilty to analyse many individual cells undergoing the same developmental transition, from one cell type to another, reveals that the order of events in development is not absolutely fixed. Some events in the amoeba-plasmodium transition, such as loss of ability to develop flagella or activation of a specific tubulin gene, may be executed out of the usual turn; some unusual sequences of events may be lethal but, for others, a viable plasmodium may be formed. This principle of variability on a developmental pathway may apply to other organisms, although it would be easy to overlook. On developmental pathways that are known principally from analysis of populations of cells, occasional variations could be undetectable. In other organisms, the number of cells analysed individually may not be large enough to detect the rarer, although legitimate, alternative paths, and there is a tendency to discount occasional experimental variation as artefacts. There may be more variation on other developmental pathways than we presently know. Even the mitotic cell cycle is not always the invariate sequence of events described in textbooks.

V. The Mitotic Cycle A . THE PLASMOLIIAL MITOTIC CYCLE

Mitotic cycle studies with P. pofycephafurnare typically carried out on the plasmodium to exploit its natural synchrony and large cellular mass. A key property of the plasmodium is the ability to remove multiple pieces, large or small, for analysis without disturbing the synchronous progression of t h e mitotic cycle (e.g. Schedl et al., 1984a,b). In addition, progression of the mitotic cycle can simply and immediately be monitored by phasecontrast microscopy of tiny biopsies of the plasmodium (Mohberg, 1982a). Together, these features allow accurate, simultaneous high-resolution analysis of multiple events at many stages of consecutive, unperturbed mitotic cycles. The plasmodial mitotic cycle differs from the amoeba1 cycle in its absence of cytokinesis, and occurrence of mitosis within the nuclear membrane, orchestrated by intranuclear MTOCs. This arrangement presumably prevents nuclear fusion during mitosis, which occurs in multinucleate amoebae (Burland e t a l . , 1981). In addition, there are no astral microtubules radiating from the plasmodial mitotic spindle poles (Aldrich, 1969; Tanaka, 1973). 'The occurrence of plasmodial mitosis without dissolution of the nuclear membrane is reminiscent of fungi (Heath, 1982) and some animal embryos (Hare1 et al., 1989). However, unlike some yeasts, the usual cycles of chromatin condensation4econdensation and mitotic-spindle assembly-

40 ardA replication ard6CD replication ard8C m RNA

Actin synthesis Actin level altA 7 replication allA2 replication alt61,Z replication

Tubulin mRNA Tubulin synthesis Tubulin level H41 replication H42 replication H4 mRNA H3,4 histone synthesis

H2 histone synthesis uH2 H1 phosphate level H1 kinase activity

Commitment to M ' Hobohm M factor

~

MPF ~

DNNnucleus S-phase Telophase Anaphase Metaphase Prophase M(n)

M(n+l) Time (min)

M(n+2)

THE PROTIST PHYSARUM POI.YCEPHALUM

41

disassembly do occur at mitosis. Synchrony in the plasmodium includes not only mitosis but also DNA replication, facilitating analysis of replication timing. As in amoebae, there appears to be no G I phase following mitosis in growing cells (Fry and Matthews, 1987; Dee et af., 1989), although a G I phase has been reported for pre-encystment amoebae (Fry and Matthews, 1987). The S-phase occupies 2-3 hours of plasmodial interphase, the balance of the 8-10 hour cell cycle consisting of a long G2 phase and 20-30 minute mitosis (Evans et al., 1982). Salles-Passador et af. (1991) report that cytoplasmic microtubules can be detected in the plasmodium (Salles-Passador et af., 1992) during interphase. These cytoplasmic microtubules were observed to disassemble during intranuclear mitosis of the plasmodium (Salles-Passador et af., 1992), as occurs during extranuclear mitosis in amoebae. This contrasts with some observations made on cells undergoing the amoeba-plasmodium transition, where assembled cytoplasmic microtubules were observed in occasional cells fixed during intranuclear mitosis (see Section IV.F.2; Solnica-Krezel et af., 1991). It will be interesting to learn what structures organize cytoplasmic microtubules in plasmodia. Some have questioned the use of the plasmodium to analyse the eukaryotic mitotic cycle, viewing the cycle as abnormal. In fact, the multinucleate state, intranuclear mitosis and absence of a G Iphase between mitosis and the S-phase are characteristics shared with organisms beyond the kingdom Protista. Early embryos of many animals are multinucleate syncytia and, in the case of embryos of Dr. mefanogaster, for example, mitosis is intranuclear for the first several divisions (Hare1 et af., 1989). The G I phase, defined by Howard and Pelc (1953) as a “gap”, has long been known as a dispensable time period in eukaryotes, including cultured FIG. 6. The plasmodial mitotic cycle schedule. The time-scale at the bottom spans two consecutive mitotic cycles, with the vertical dashed lines indicating the end of metaphase (M). The occurrence of an event, or presence of an activity or molecular species, is indicated by shading. MPF indicates maturation promotion factor or Mphase factor; Hobohm M factor, mitotic advancing activity detected by Hobohm et al. (1990), which occurs at the same time as the mitotic activity detected by Grobner and Loidl (1985); commitment to M,commitment to mitosis; HI, 2, 3,4, histones 1,2,3,4; uH2, ubiquitinated histone H2; altA1, altA2, altBI, altB2, alleles of the AltA and AltB a-tubulin genes; ardA, ardB, ardC, ardD, actin genes. Data are taken from the references indicated in the text, and normalized to a 10 hour mitotic cycle. Times of occurrence are plotted as accurately as possible, but the changes in amplitude are not intended to represent absolute values of the data; rather, they indicate the general nature of the changes. For example: the actin protein synthesis rate, protein level and mRNA level remain a proximately constant as functions of cell mass throughout the mitotic cycle; p34cdcykinaseactivity shows a sharp peak; replication of individual genes and alleles occurs in defined segments of the S-phase.

42

T (i BUR1 A N D I.

r A1

mammalian cells (Prescott, 1976). The fact that key events occur in the G I phase in many organisms does not mean that those events must occur between mitosis and the S-phase. And, as discussed later, it is now clear that fundamental aspects of mitotic regulation are common to most if not all eukaryotes. Summarizing, the plasmodium is far from abnormal and, as experimental evidence indicates, continues to be a useful system for mitotic cycle studies. Since Pierron (1986) and Laffler and Tyson (1986) reviewed the status of mitotic-cycle studies in P. polycephalum, significant progress has been made in this area. H. PERIODIC VARIATIONS

The most important criterion for establishing mitotic-cycle periodicity of specific events is to detect the same variations at the identical stage in consecutive mitotic cycles. One plasmodium can be grown to sufficient mass so that samples for biochemical analysis can be taken at many consecutive times, over consecutive cycles, without decay of synchrony, making this key criterion unusually straightforward to establish. Variations relating to tubulins and histones (Fig. 6) illustrate the practical value of the plasmodium. 1 . Tubulins

Tubulin expression varies dramatically over the cell cycle of the plasmodium, for both protein synthesis and transcript level (Fig. 6; Laffler et al., 1981; Schedl ei al., 1984b). This expression is low in the late S-phase and peaks 4(k80-fold higher at metaphase. The rise in the protein-synthesis rate is co-ordinate for all expressed tubulins, namely a l A , a l B , a2B, P1B and p2 (Schedl ei al., 1984b; Cunningham et al., 1993), although, after metaphase, synthesis of a2B and 02 tubulins decreases more rapidly than synthesis of a1 and PlB tubulins (Schedl et al., 1984b). The high level of expression of tubulin transcripts in the late G2phase can be suppressed by fusing a late G2-phase plasmodium with a plasmodium in the S-phase (Laffler, 1987), indicating that tubulin expression is negatively regulated during the S-phase. The stability of tubulin protein also varies over the cell cycle, with the tubulin protein level falling 50% or more during the S-phase (Ducommun and Wright, 1989), when the tubulin synthesis rate is falling (Fig. 6). Inhibition of DNA synthesis with hydroxyurea prevents tubulin degradation, suggesting some coupling between these events (Ducommun and Wright, 1989). The function of tubulin degradation is not clear. Inability to detect cytoplasmic microtubules in the plasmodium leads us to believe

I H E PROTIST PHYSARUM POLYCEPHAl.UM

43

that tubulin would not be utilized outside the mitotic phase of the cell cycle, but detection of cytoplasmic microtubules (Salles-Passador et al., 1991, 1992) suggests that tubulin is indeed utilized. It would be useful to know what fraction of available tubulin in the plasmodium is utilized in cytoplasmic microtubules. Despite some degradation of tubulin during the mitotic cycle, not all of the tubulin is lost. First, all tubulin isotypes are readily detectable in the S-phase, as well as in the G2 phase, by silver staining of proteins resolved on two-dimensional gels (Schedl et al., 1984a). Second, analysis of tubulin radiolabelled in vivo indicates that some of the tubulin protein synthesized in one mitotic cycle is used in the mitotic-spindle microtubules at the end of the following cycle (Paul et al., 1987). Although these experiments do not establish whether a tubulin molecule used in one mitotic spindle is reutilized in the spindle in the following mitosis, they clearly indicate that a significant fraction of the tubulin pool escapes degradation and remains available for use in more than one mitotic cycle. For tubulin transcripts, the variation in level appears to be regulated mainly by changes in the transcript-degradation rate, and to a lesser extent by the rate of synthesis (Green and Dove, 1988). Tubulin transcripts are substantially less stable after metaphase than before metaphase, and the lower stability correlates with a decrease in the length of the poly(A) tail on the mRNA (Green and Dove, 1988). The length of poly(A) tails correlates with mRNA stability for transcripts of several eukaryotic genes, both in vivo and in vitro (Bernstein and Ross, 1989). Just as tubulin expression increases in the flagellate shortly before a new microtubule structure, the axoneme, is assembled, tubulin expression increases on in the mitotic cycle of the plasmodium shortly before the mitotic spindle is assembled. It is nevertheless interesting that tubulin expression is already substantially rising over three hours before the metaphase. Given the apparent stability of tubulin proteins at this time, it seems unlikely that the upturn in tubulin expression is based on a negative-feedback loop wherein depletion of the tubulin pool by assembly into microtubules leads to increased tubulin-gene expression (Yen et a l . , 1988). The first visible sign of the mitotic spindle does not occur until 60 minutes before the metaphase, when a dot of tubulin-containing material can be detected in the nucleus by immunofluorescence microscopy (Paul et al., 1987). This structure is presumably related to the organizing centre for the mitotic spindle, and it increases in size for the next 30 minutes until the classical prophase can be discerned (Paul et al., 1987). Simultaneous measurements of tubulin-transcript levels and tubulinsynthesis rates, together with light- and electron-microscope structure determinations confirm that tubulin expression peaks at the metaphase in the plasmodium (Schedl et a l . , 1984b).

44

T . G RLIRI.AND E 7 A L

One might wonder why the dramatic variation in tubulin expression over the plasmodia1 mitotic cycle has not been observed in mitotic cycles in other organisms. However, evidence has been obtained for a 2-3-fold increase in tubulin expression late in the mitoticcycle for HeLa cells (Bravo and Celis, 1980). Since these experiments were performed on a population of imperfectly synchronous cells, it seems likely that the change in expression is greater for each individual cell. In a fascinating theoretical reconstruction experiment, Matthews and Bradbury (1982) compared the amplitude of changes in levels of H1 -histone phosphorylation observed during the mitotic cycle in a plasmodium with changes in phosphorylation levels that would be observed if the nuclei in the plasmodium were as asynchronous as a typical “synchronous” culture of mammalian cells used for the same experiment. They found that the imaginary imperfectly synchronous plasmodium showed a remarkably similar pattern of change in H1-phosphorylation level to that observed for mammalian cells, with a smaller amplitude of change than observed in a synchronous plasmodium. Thus, it would be unwise to conclude that because the plasmodium in the slime mould shows greater amplitudes for mitotic cycle changes than observed for other organisms, there is something odd about P. polycephalum. The analysis by Matthews and Bradbury highlights the power of the natural synchrony for revealing changes of small amplitude or of short period that may be overlooked when imperfectly synchronous populations of cells are analysed.

Histones Chromatin in P. polycephalum contains a complement of histones similar to those of most other eukaryotes, namely H1, H2A, H2B, H3 and H4 (Matthews and Bradbury, 1982; Hardman, 1986), as well as the more recently discovered H1” (Yasuda et al., 1986). In contrast to the coordinately expressed tubulins, different histones display distinct expression patterns during the mitotic cycle of the plasmodium. Estimating synthesis rates of pulse-labelled nuclear proteins resolved on gels, h i d l and Grobner (1987a) reported that: (a) H3 and H4 histones are synthesized primarily during the S-phase; (b) H2A and H2B histones are synthesized at a high rate during the S-phase and at a moderate rate in the G2 phase; and (c) H1-histone synthesis peaks towards the end of the Sphase (Fig. 6). When DNA replication is inhibited by hydroxyurea, synthesis of H3 and H4 histones is also inhibited, but synthesis of H2A and H2B histones is unaffected, indicating that the two types of histones are regulated by distinct mechanisms (Loidl and Grobner, 1987a). The two isocoding H6histone genes in P. polycephalurn show co-ordinate

TIIF PROTIST PHYSARUM POLYCEPHALLIM

45

expression through the mitotic cycle (Wilhelm and Wilhelm, 1989). The transcript level for histone H4 rises in the late G2phase, remains high when the S-phase initiates, then gradually falls during the S-phase (Fig. 6; Carrino et a f . , 1987). The lack of HChistone synthesis in the G2 phase, when transcript is present, appears to be due to sequestration of the message in ribonucleoprotein particles in the cytoplasm (Wilhelm er a f . , 1988). When DNA replication i s blocked by hydroxyurea, the H4-transcript level falls with a half-life of 15 minutes, in parallel with the hydroxyureainduced fall in the S-phase synthesis of the H4 protein. But the increased level of H4 transcript in the late G2 phase is unaffected by hydroxyurea (Carrino et a f . , 1987). The effects of hydroxyurea are confirmed when transcription of H4 genes is assayed by nuclear run-on in vitro (Carrino et a f . , 1987), suggesting that changes in transcription rates as well as stability are important in regulating HCtranscript levels. The 15 minute half-life of H4 transcript in the S-phase is comparable to the estimate of less than 19 minutes for the half-life of tubulin transcripts in the S-phase (Green and Dove, 1988), and H4 transcripts appear more stable in the late G2 phase, just as tubulin transcripts do. It has been postulated (Carrino and Laffler, 1986; Laffler and Carrino, 1986) that this reflects co-ordinate control of histone and tubulin expression during the mitotic cycle. However, this model does not account for the distinct patterns of regulation of the different histones. Histone modifications, such as phosphorylation, ubiquitination and acetylation, may be important for the transcriptional state of the chromatin, for the chromatin condensation-decondensation cycle or for chromosome replication. Of interest in the mitotic cycle of the plasmodium of P. pofycephafumis the parallel pattern of changes in the phosphate content of H1 histone, and activity of H1 kinase. Peak kinase activity is reached ahead of metaphase, falling to early interphase levels by the time division occurs (Matthews and Bradbury, 1982). Activity of H1 kinase thus peaks somewhat ahead of chromosome condensation at mitosis (Fig. 6), although the precise timing and any causal connection between H1 phosphorylation and chromosome condensation are the subject of debate. Nevertheless, phosphorylation of H1 histone in P. pofycephalum has demonstrable structural effects in virro (Jerzmanowski and Krezel, 1986); it seems likely that such profound changes would have biological significance. Jerzmanowski and Malaszewski (1985) detected individual H1-histone variants that differed stepwise from 0 to as many as 28 phosphate residues per polypeptide. During mitosis, H1 histone is present in variants with 14-28 phosphate residues per molecule, compared with 8-20 residues in the late S-phase, reaching a minimum in the mid-G2 phase of 6-14 phosphates residues per molecule. Thus, although H1-histone molecules with the

46

1' ti. HIII1I.ANI) El A l .

highest level of phosphorylation occur at mitosis, when chromatin is condensed, dephosphorylation does not minimize the phosphate content in H1 histone until well after decondensation of chromatin. Since the number of phosphate residues per H1 molecule is variable at any one time, Jerzmanowski and Malaszewski (1985) suggested that several phosphorylations may be needed to generate a quantum change in HI conformation. All of the core histones in P . polycephalum appear to be susceptible to acetylation, at multiple sites on each molecule. Waterborg and Matthews (1984) observed acetate turnover on H2A and H2R histones only in the Sphase, but H3 and H4 acetylation turned over in both the S- and Gz phases. However, the pattern of turnover of acetate on H3 and H4 histones differed in the S- and G z phases, and Waterborg and Matthews (1984) proposed a correlation between mono-acetylated H4 histone and inactive Chromatin. Meanwhile, H4 deacetylase activity changes only two-fold during the mitotic cycle (Waterborg and Matthews, 1982). Loidl and Grobner (1987b) found a correlation between the ability o f protamine to displace histone H4 from chromatin in vitro and the stage of the mitotic cycle whence the chromatin was isolated. Core histones were more readily released from S-phase chromatin than from Gz-phase chromatin, and the more highly acetylated H4 histones were more readily released by protamine than less highly acetylated molecules. Acetylation of core histones may serve to release them more readily from chromatin for the purposes of both DNA replication and transcription (Loidl, 1988). Perhaps the most dramatic change in histone modification during the cell cycle is ubiquitination of H2A and H2B histones (Fig. 6; Mueller el ul., 1985). Both histones are ubiquitinated during most of the mitotic cycle, but are de-ubiquitinated at prophase and metaphase, then re-ubiquitinated at anaphase. This modification pattern may be a general feature o f the eukaryotic mitotic cycle, as H2A histone in metaphase chromosomes of several eukaryotes lacks ubiquitin (Bradbury, 1902). Histidine kinase activity in extracts of nuclei from P. polycephalitm phosphorylates the His75 residue of free histone H4 in vitro (Huebner and Matthews, 1985), although not H4 in nucleosomes (Wei et a l . , 1989). Phosphohistidine is found in histone-like proteins in nuclei from P . polycephalirm (Pesis et a l . , 1988), but the precise substrate(s) of the histidine kinase and the biological significance o f phosphorylation of histidine in nuclear proteins remain to be determined. Purification of a protein histidine kinase from yeast that phosphorylates specifically His7s in H4 histone in vitro (Huang el a l . , 1991) suggests that this activity may have broad biological relevance. Considering the variety of covalent modifications that can be made to

1’HF P R O I I S I PliYSARUM 1’01.YCEPHA1.1IM

47

various histones, the possible combinations of modifications that might occur sometimes seem discouragingly high. Nevertheless, correlations between modifications and functional and structural changes are too remarkable, and occur in too broad an array of eukaryotes (e.g. Bradbury, 1992) to ignore. It would help greatly if chromatin structure could be analysed for a single gene over the course of consecutive synchronous plasmodial mitotic cycles, where both the replication time and expression pattern are accurately known. Such a test system may be feasible using plasmids carrying specific replication origins linked to various cloned genes. Given the progress being made in mapping replication origins (see Section V.C) and in DNA transformation (see Section V1.C) in P. polycephalum, this feasibility may also soon be at hand. Finally, conditional mutants defective in mitotic-cycle progression would help to provide the functional analysis needed to complement the biochemical progress. This last resource will be simple to achieve (Burland, 1978, 1986; Burland and Dee, 1980; Burland et a l . , 1981).

3. Significance of Periodicity Universally periodic events in the mitotic cycle, such as DNA synthesis and mitosis are presumably regulated by some underlying periodic changes, be they transcriptional, translational or post-translational (for a review, see Tyson, 1982). But some events, such as changes in thymidine kinase activity over the mitotic cycle (Sachsenmaier et al., 1970), can occur as a peak or “step” change, depending upon culture conditions (Wright and Tollon, 1979). Thus, observation of periodicity does not prove that the periodicity is necessary for orderly mitotic progression. An equally disconcerting problem is that events which show mitotic periodicity may not need to occur at the particular stage at which they are observed. This is another issue that mitotic mutants can help us understand better. Intermitotic time, which remains at a constant mean value for a given cell type under specific conditions, shows substantial variations between individual cells (e.g. Bailey et al., 1987). These variations in part reflect asymmetry of division, whereby daughter cells of unequal size, growing at a constant rate, take different times to reach the same DNA-to-mass ratio for the next division. However, mass at division also varies, further contributing to intermitotic time variation. In the plasmodium, natural mitotic synchrony occurs in billions of nuclei in the same cell, without cytokinesis. Thus, variation in cycle time due to variation in mass at division will tend to be averaged out as an experimental variable. This may contribute to the ready detection of periodicity of events in the plasmodial mitotic cycle compared with mitotic cycles in other organisms, particularly

48

T G HUHI.ANI) F.7 A l .

for events such as increased tubulin expression that occur later in the cycle. Analysis of cell-cycle mutants in a yeast shows that certain functions need to be carried out in a particular sequence for the orderly progression of the mitotic cycle (Hartwell, 1991), but the absolute stage of execution of certain functions often seems less critical than their sequence. However, precise timing of the execution of key functions is probably vital for events that occur during mitosis, and this is one area where the large-scale mitotic synchrony of the plasmodium offers powerful opportunities for experimental analysis.

C . CHROMOSOME REPLICATION

In the plasmodium, DNA synthesis begins immediately after mitosis (Nygaard et al., 1960). This feature provides a clear and simple cytological marker for the beginning of the S-phase. A fundamental principle of the temporal organization of chromosome replication, demonstrated in the plasmodium, is that DNA sequences replicated during a specific stage of the S-phase in one mitotic cycle are replicated at the same stage in the subsequent mitotic cycle (Braun et al., 1965). Thus, at the gross genome level, DNA replication occurs in a defined sequential order in each cycle. The size distribution of newly replicated DNA (i.e. pulse-labelled DNA, denatured and size fractionated) indicates that Okazaki fragments are synthesized throughout the S-phase. During the first hour of the S-phase, when replication is at its maximum rate, the size of the newly synthesized DNA increases to l.107-2.107 Da, the proposed size of a replicon (Funderud et al., 1978a). Subsequently, further replication and joining of replicons leads to discontinuous increases in the size of newly replicated DNA around 90 minutes and again at 120 minutes into the S-phase (Funderud et a l . , 1978a). These observations indicate that specific replicons are synchronous between all of the nuclei in the plasmodium, and that different replicons proceed with DNA replication at various times during the S-phase; isochronous replicons tend to lie in clusters. Funderud et al. (1978b) also obtained evidence that most if not all replicons initiate at the beginning of the S-phase. Thus, the initiation of replication and the elongation of replication forks are distinct events that can be separated by substantial intervals of time. It is remarkable that replication remains rather synchronous later in the S-phase, since many replicons proceed with DNA replication long after initiation has occurred. Flow cytometry of nuclei isolated from a plasmodium progressing through the S-phase confirms that replication begins very soon after mitosis and that individual nuclei within

THE PROTIST PHYSARUM POLYCEPHALUM

49

a single plasmodium increase their D N A content at approximately the same rate (Kubbies and Pierron, 1983; Kubbies et al., 1986). I . Replication Timing of Individual Genes Given that the genome in P. polycephalum replicates synchronously in all of the nuclei in a plasmodium, that specific portions of the genome replicate in a defined temporal sequence, and that replicons are substantially larger than structural genes, it should be possible to investigate the replication timing of individual genes during the S-phase in a plasmodium. If replication is allowed to progress in the presence of bromodeoxyuridine, newly replicated D N A can then be separated by isopycnic centrifugation from unreplicated DNA,on the basis of its higher density. If the separated D N A species are then analysed by Southern blotting, replicated D N A fragments will be detected in the heavier D N A while unreplicated fragments will be detected in the lighter D N A fraction. Using this technique to analyse replication of actin genes in the plasmodium of P. polycephalum, Pierron et al. (1984) found that three actin loci, ardB, ardC and ardD, are replicated early in the S-phase while the fourth, ardA, is replicated late in the S-phase (Fig. 6). The bromodeoxyuridine heavy labelling strategy has also been used to estimate the time of replication of the HChistone genes, H41 and H42, in the plasmodia1 mitotic cycle. Jalouzot et al. (1985) found that H41 replicates in the first 10 minutes of the S-phase while H42 replicates between 20 and 30 minutes into the S-phase (Fig. 6). One limitation to this technique is that bromodeoxyuridine labelling must be carried out for extended periods, usually beginning before the S-phase, to obtain a sufficient density increase for resolution of newly replicated and unreplicated DNA. To circumvent this problem, and to increase temporal resolution, Southern blots can be run for unlabelled D N A extracted at different stages of the S-phase, and the intensity of bands representing specific restriction fragments can be quantified; band intensity should double upon replication. To eliminate errors caused by inaccurate D N A loads in different lanes of a Southern blot, band intensities need to be normalized to other bands in the same lane; therefore, the time of replication can only be determined by this method if two or more fragments detected replicate at different times, since only then will the relative band intensities change. With appropriate care, relative band intensities can be estimated to within +lo%, well within the accuracy needed for this purpose (Cunningham, 1992). Using this strategy, Pierron et al. (1984) established that the actin genes ardB, ardC and ardD replicate in the first 10 minutes of the S-phase, when 10% of the genome is replicated; the ardA gene replicates much later, around 80-100 minutes into the S-phase, when 75%

50

I

Ci

BUHI A N D

t T A/.

of the genome is replicated (Fig. 6). Since the plasmodium used is heterozygous for restriction-fragment length at ardC and ardA , Pierron et al. (1984) were able to demonstrate for the first time contemporaneous replication of different alleles of the same gene. The use of the parameter “percentage of genome replicated” (“1’0 G) may prove more reliable as a measure of the stage of the S-phase than clock time into the S-phase. This value can be obtained accurately by flow cytometry of a sample of the nuclei used for replication analysis. While contemporaneous replication of a pair of alleles may be the norm, the exquisite synchrony of the S-phase in the plasmodium provides an opportunity t o observe significant differences in replication timing should any pair of alleles behave this way. Two alleles tested for the altB a-tubulin locus in P. polycephalum show contemporaneous replication in a diploid plasmodium heterozygous for restriction-fragment length. By contrast, two alleles of altA replicate at distinct times (Cunningham and Dove, 1993; Fig. 6). This is the first compelling example of distinct times of replication for alleles of the same locus. Thus, to obtain a clear picture of DNAreplication timing in any diploid cell, it is essential to recognize that the time of replication may differ for each allele.

2. Timing of Replication Relative to Timing of Transcription The idea that transcription and replication of a gene are coupled in eukaryotes has gained wide currency (e.g. Goldman, 1988), bolstered by some observations that expressed genes replicate early in the S-phase while silent genes replicate late. In P. polycephalum, electron-microscope examination of DNA samples from the S-phase plasmodium reveals that transcription units are often associated with replication “bubbles”, consistent with a coupling between replication and transcription (Pierron et al. , 1983). The establishment of reliable, high-resolution methods to determine the time of replication of individual genes in the plasmodium, together with isolation of cDNA clones of genes that show distinct patterns of expression (Pallotta et al., 1986), facilitated a rigorous test of the temporal patterns of replication and transcription of several genes. Ten cDNAs were chosen, representing single-copy genes; six are expressed in the plasmodium, while the other four are not. Of the four not expressed in the plasmodium, two are expressed in the amoeba, a third during spherulation, and the fourth during sporulation (Pallotta et al., 1986; Pierron et al., 1989). One of the six genes expressed in the plasmodium, LAV3-2, is also expressed in the amoeba, but the transcripts are of different size in the two cell types. Four of the six genes expressed in the plasmodium replicate early in the S-phase,

THE PROTIST PHYSARUM POLYCEPHAl.1IM

51

but the other two genes, LAV3-2 and the plasmodium-specific gene LAVl4, replicate late (Pierron et al., 1989). This clearly contradicts the general

rule that expressed genes are replicated early. Among the genes not expressed in the plasmodium, the spherulation-specific gene replicates late, the two amoeba-specific genes replicate in the mid-S-phase and the sporulation-specific gene replicates early. Again, the results argue against a general rule of coupling between replication and transcription.

3. Identifying Replication Origins The natural synchrony of DNA replication in the plasmodium renders detection of short-lived replication intermediates much simpler than in other organisms (e.g. Hamlin et al., 1991), circumventing the need for artificial synchronization or gene amplification to detect replication intermediates. Funderud et al. (1978a) showed that newly replicated DNA from the plasmodium can be detected initially as low molecular-weight single-stranded fragments, which increase in length per minute by about 1.5 kb, approximately the size of a typical gene. Benard and Pierron (1990) reasoned that synchronous replication of a specific gene in the plasmodium would generate nascent ssDNA fragments containing the gene which would increase in size during the S-phase. Very high molecular-weight parental ssDNA fragments bearing the gene would always be present. Therefore, if DNA is isolated from the plasmodium at consecutive stages of the Sphase, denatured, then resolved on agarose gels and blotted to filters, replication intermediates of a specific gene will be detectable by hybridizing the filters with a nucleic-acid probe for the gene. The smallest size of the replication intermediate first detected in the S-phase for a particular gene would depend upon the distance of the gene from the origin of replication; a gene containing a fixed origin of replication would show a replication intermediate at the lowest detectable size, while a gene 10 kb from a replicon would show a minimum replication intermediate of 20 kb, assuming bidirectional replication from a fixed origin with the two forks displaying the same replication rate. For the LAW-2 gene (Pallotta et al., 1986), Benard and Pierron (1990) found the first replication intermediate, only 5 minutes (less than 5% G) into the S-phase, with a minimum size of around 6 kb, indicating very early replication of this gene, and a replicon within 3 kb of the 1.2 kb gene, i.e. very close to the gene itself. This technique reveals the increasing size of the replication intermediates in samples taken from consecutive stages of the S-phase, and can thus reveal the rate of elongation of the replication fork. ‘ h e replication intermediate for LAVI-2 increases in size at about 1 kb each minute, in remarkably close agreement with the data of Funderud et al. (1978a) for average replicon growth rates in P. polycephalum.

52

T G RLIRI.AND ET A l .

Benard and Pierron (1990) also noted that the replication intermediates for LAV1-2 showed some size heterogeneity. One explanation for this heterogeneity is asynchrony. In fact, differences in timing of initiation of the LAVI-2 replicon in different nuclei could be at most 3-4 minutes for asynchrony alone to account for the heterogeneity, not bad synchrony for a 10 hour mitotic cycle. However, other factors may contribute to the size heterogeneity of ssDNA, including degradation during DNA isolation, variation in the rate of fork movement, and possibly initiation of replication from close but separate sites. Even with the size heterogeneity observed, the small, discrete size of the first replication intermediates detected provides a compelling case for a limited region of replication initiation in eukaryotes (cf. Hamlin et a f . , 1991). Further application and refinement of this strategy should lead to physical location of replication origins with respect to specific DNA sequences with greater accuracy, perhaps down to the 150-200-nucleotide size of Okazaki fragments. In parallel with this type of analysis, developments in DNA transformation in P. polycephalum provide a basis for functional assays for replication-origin activity of specific DNA fragments, both in vitro and in vivo. D. RIBOSOMAL

DNA

REPLICATION

I n the 60 kb palindromic extranuclear multicopy rDNA molecule from P. pofycephalum, two pairs of symmetrically arranged replication origins can be observed by electron microscopy, although only one replicon appears to be active on any one molecule (Vogt and Braun, 1977). As for chromosomal genomes, replication of rDNA is semi-conservative and bidirectional, although most replication occurs in the late S-phase and in the G 2 phase (Braun and Evans, 1969; Zellweger ef a f . . 1972; Newlon et a f . , 1973; Vogt and Braun, 1977). Regions of rDNA molecules around the four potential replication origins are heavily methylated (5methyldeoxycytidine), but the replicon that is active on any one molecule is hypomethylated (Cooney el a f . , 1988). Extracts of the plasmodium of P. pofycephafumcan selectively initiate replication of rDNA in vitro, at the same sites that replication initiates in vivo (Daniel and Johnson, 1989). Plasmid pPHR21 contains a segment of the rDNA that includes one of the replication origins (Ferris, 1985). Extracts taken from synchronous plasmodia catalyse replication of this plasmid in vitro, with highest catalytic activity coming from prophase plasmodia, when the nucleolus disperses. Moderate activity is found in the early S-phase but, by the time the nucleolus is fully reorganized later in the S-phase, little activity can be detected (Daniel and Johnson, 1989). Electron microscopy pinpoints the origin of rDNA replication in vifro

THE PROTIST PHYSAHUM POLYCEPHALUM

53

at the same position as estimated in vivo, corresponding to a region of 78 tandem repeats of an element with a 31 bp consensus sequence (Daniel and Johnson, 1989). It will be interesting to compare the rDNA origins with chromosomal replicons, which will soon be identified (Benard and Pierron, 1990), and to determine whether replication is controlled by the same or distinct factors for chromosomal and rDNA. This is particularly interesting given that chromosomal DNA replicates only once in each mitotic cycle, while some rDNA molecules can replicate more than once in a cycle in which others do not replicate at all (Vogt and Braun, 1977). E. MITOTIC REGULATION

I . Heterop loidic Fusion Plasmodia of P. polycephalum which carry identical alleles of the fusA, fusB and fmC loci fuse naturally when they come into contact. Fusion occurs regardless of the ploidy of the two plasmodia. Dee and Anderson (1984) fused together haploid and diploid plasmodia that carried distinct genetic markers. They observed that the phenotypes associated with the diploid nuclei were always lost from these heteroploid plasmodia, and by careful reciprocal experiments eliminated the possibility that residual genetic differences could account for directional loss of the nuclei observed (Dee and Anderson, 1984). It seems likely that this phenomenon is related to some fundamental aspect of mitotic-cycle regulation. Analysis of replication timing for haploid and diploid nuclei would be illuminating, and could be achieved by analysing replication times for two alleles of the same gene that normally replicate at the same time. For example, the heteroploidic heterokaryon could be made with haploid altB1 nuclei and diploid altB2/altB2 nuclei (see Section V.C). 2. Heterophasic Fusions The natural capacity for genetically similar plasmodia to fuse permits unusually straightforward analysis of mitotic averaging. For example, if plasmodium A is scheduled to divide at 2 p.m. and plasmodium B at 6 p.m., a plasmodium created by the fusion of A and B around noon will initiate mitosis between 2 p.m. and 6 p.m.; the greater the mass that plasmodium B contributes to the mixed plasmodium, the closer to 6 p.m. will be the initiation of mitosis (for a review, see Tyson, 1982). Such results implicate cytoplasmic, rather than nucleus-limited, factor(s) controlling the timing of mitosis. Plasmodia1 fusion experiments have been used to elucidate the time of

54

I

(0

HllKl AND f / A /

commitment to mitosis, defined as the stage beyond which fusion of plasmodium A , early in the mitotic cycle, to plasmodium €3, later in the cycle, can no longer retard initiation of mitosis in nuclei from plasmodium B. Commitment occurs typically about an hour (0.1 of a cycle time) before metaphase, although the time o f commitment in uiuo may correspond to the onset of prophase (15-30 minutes before metaphase), depending on the time correction made for plasmodial fusion and cytoplasmic mixing to occur (Loidl and Sachsenniaier, 1982). Beyond timing events, the technique of plasmodial fusion permits analycis of specific molecular events. For example, when plasmodial nuclei in the S-phase are introduced by brief fusion into a plasmodium in the G 2 phase, the S-phase nuclei continue t o replicate, but do not induce G2 nuclei to replicate (Guttes and Guttes, 1968). Nevertheless, after the G2 nuclei ultimately pass through nuclear division, they can then replicate nornmally, highlighting the now common knowledge that replicated nuclei cannot normally replicate again before nuclear division is initiated (Guttes and Guttes, 1968). Use o f fluorouracil t o inhibit DNA replication in plasmodia had already shown that mitosis is dependent o n prior DNA synthesis (Sachsenmaier and Rusch, 1964). Thus, heterophasic fusion experiments completed the picture of the now-familiar cycle of mutual dependency between DNA synthesis and mitosis in eukaryotes. These experiments paved the way for similar experiments with mammalian cells, which confirmed that this fundamental regulatory mechanism is a general feature of eukaryotic mitotic-cycle control (Rao and Johnson, 1970). The DNA synthesis-mitosis dependency cycle is perturbed only in rare cases in nature, for example to develop polytene chromosomes, although it can he interrupted in the laboratory not only by inhibitors but also by mutation (e.g. Broek el a l . , 1991).

3. Mitotic Factors A logical next step from fusing whole plasmodia at different phases of the mitotic cycle is to make extracts of plasmodia at specific points in t h e mitotic cycle and add these extracts to plasmodia at other stages in the cycle. Two features of the plasmodium facilitate this experimental approach. First, since one plasmodium can be cut into two or more pieces without detriment to its mitotic progression, one piece can be used as a robust control sample for other experimental pieces. Secondly, substances placed on top o f a plasmodium are usually taken up, obviating the need for injection. Since Oppenheim and Katzir (1971) showed that extracts from plasmodia in the late G2 phase could advance mitosis in plasmodia in t h e early Gz phase by at least an hour, many other studies have confirmed the presence o f mitotic

TtiE PROTIST PHYSARUM POLYCEPHALUM

55

factors in plasmodia about to undergo mitosis. Application of partially purified mammalian histone kinase to a plasmodium advances the time of mitosis by up to 40 minutes (Bradbury et al., 1974), strongly suggesting a role in mitotic regulation for histone kinase or a copurifying substance. Loidl and Grobner (1982) improved methods for uptake of extracts into plasmodia, and found that mitosis can be advanced by up to two hours by addition of extracts from a plasmodium in the late G,-phase. The proteasesensitive factor they assayed was not detected in extracts from other stages of t h e mitotic cycle (Loidl and Grobner, 1982). Once the presence of mitotic factors is established for a particular extract (which need not be from P. polycephalum), enrichment and purification are necessary. Grobner and Loidl(l985) reasoned that, if the late G,-phase mitotic factor were absent from other stages of the mitotic cycle, it should be possible t o enrich the factor by making antibodies to S-phase plasmodial extracts and, using the antibodies, to subtract S-phase proteins from late G,-phase extracts. They succeeded in decreasing the number of abundant proteins in a late G2-phase extract by a factor of 35, and increasing the specific activity of the mitotic stimulating factor 10-fold in the immunodepleted extract (Grobner and Loidl, 1985). However, they did not detect a novel protein in the G,-phase extract, although there are several reasons why the factor may be difficult to detect: the mitotic factor may be in low abundance; it may have properties that preclude its identification on the two-dimensional gels used; the active factor may be a complex mixture like M-phase factor (see below); and assay for the factor may not be linear. Hobohm et al. (1990) used direct injection of 3-5 pl volumes of plasmodial extracts to evaluate their capacity to advance the timing of the next mitosis. They found that only extracts from a 20-30 minute period just before prophase advanced the timing of mitosis in plasmodia, and then only when injected into plasmodia early in the mitotic cycle. The timing of the presence of this mitotic factor correlates rather well with the time of commitment of plasmodial nuclei to mitosis (Fig. 6; Loid1 and Sachsenmaier, 1982). Hobohm et al. (1990) then sought to purify the mitotic factor from late G2-phase extracts. They estimated that the factor has a mass of around 2500 Da, a very small protein. Such a protein would not have been detected on the twodimensional gels used by Grobner and h i d l (1985), so it could be the same factor. Following purification, the low molecular-weight factor retained its capacity to advance mitosis. Characterization of this protein is keenly awaited. 4. M P F / p 3 4 c d c 2 The temptation to compare mitotic factors detected in plasmodia from P. polycephalum with the universal mitotic regulators MPF and p34"""'

56

'I. G . RLIRI.AND E?' A L

(Lewin, 1990; Nurse, 1990) is irresistable. Oocytes from X e n o p w faevis are naturally blocked in progression just before the prophase of meiosis 1. Maturation of oocytes, which begins with germinal vesicle breakdown and chromosome condensation, can be induced by extracts of cytoplasm from cells that have matured to the second meiotic metaphase (Masui and Markert, 1971). This activity is called MPF (M-phase factor, or maturationpromoting factor), and is now recognized as a complex protein involving the p34"""' protein kinase and cyclin (Gautier et a [ . , 1988; Dunphy et al., 1988). MPF can be detected in extracts of plasmodia in the late G2 phase, but not in extracts from other mitotic-cycle stages, and its activity peaks 10-20 minutes before metaphase in plasmodia (Adlakha et af., 1988). This timing is similar to the timing of the peak in mitotic-factor activity found by Hobohm et al. (1990) (Fig. 6). Clearly, the 2500 Da mitotic factor detected by Hobohm and his colleagues is too small to be MPF, but it will be interesting to determine whether it is a completely separate entity, a previously unknown component or a key activator. The MPF activity detected by Adlakha et al. (1988) shares several biochemical characteristics with similar factors from other sources, suggesting a strong evolutionary conservation of the structure and activity of this mitotic regulator. Inhibitors of MPF have been detected in the G I phase in various cell types (e.g. Adlakha ei af., 1983) while, in CHO V79-8 cells, which like plasmodia do not exhibit a G I phase, the inhibitors are found in the Sphase. Adlakha et af. (1988) found MPF-inhibitor activity in plasmodia in the S-phase but not in plasmodia in the G2 phase. The activity inhibited MPF from both plasmodia and from HeLa cells, again pointing to evolutionary conservation of mitotic regulators. , kinase that can An intriguing comparison is between ~ 3 4 ' ~ "a ~known phosphorylate H1 histone in vitro, and the H1-histone kinase that can advance timing of mitosis in plasmodia (Bradbury et al., 1974). Shipley and Sauer (1989) detected a homologue of ~34'~'.'in a plasmodium, and found constant levels of the protein throughout the mitotic cycle, as found in yeasts. Ducommun et al. (1990) assayed activity of the p34""' kinase over the mitotic cycle in a plasmodium, using the p13""'" protein of Schizosaccharomyces pamhe to purify the kinase, and histone H1 or SV40 large Tantigen as the substrate. They found that the kinase activity of the protein from P . polycephafum is low during most of the mitotic cycle, but that activity rises abruptly in the late G 2 phase, peaking in mitosis at early metaphase (Fig. 6). Thereafter, ~34""' kinase activity plunges dramatically. This experiment marks the most accurate estimate of the profile of p34'"'"' kinase activity in an unperturbed, ongoing mitotic cycle. Delaying mitosis with microtubule poisons or a DNA-synthesis inhibitor causes the peak in kinase activity to be delayed until mitosis occurs, while blocking a

T H E PROTIST PHYSARUM POl.YCEPHAl.UM

57

plasmodium in mitosis by treating with cycloheximide late in the G2 phase leads to a sustained increase in kinase activity; these results confirm that kinase activation and mitosis are tightly coupled (Ducommun ef a f . , 1990). The peak in ~34'~'' kinase activity occurs later in the mitotic cycle than the peak in H1-histone kinase activity described earlier for plasmodia (for a review, see Matthews and Bradbury, 1982). It therefore seems unlikely that the two activities are the same. Nevertheless, the H1 kinase that peaks earlier remains a candidate as a regulator of mitosis. The early H1 kinase could represent an alternative activity of p34'd'2, or the two activities could be distinct kinases acting either independently or in a dependent series. The timing and sharpness of the peak in ~34'~'' kinase activity at early metaphase in a plasmodium suggests that mitotic factors detected slightly earlier in the cycle by Loidl and Sachsenmaier (1982) and Hobohm et al. (1990) are distinct from ~34"""' kinase, although they may act on the same pathway, or in combination with ~ 3 4 " at ~ 'earlier ~ times. With respect to MPF, it would be useful to re-examine timing of its peak in activity with ~ ' ~ peak, with both activities determined from the timing of the ~ 3 4 " kinase same plasmodium. This would clarify whether the p34"d'2 kinase peak coincides with or follows the MPF peak, a matter of some concern for evaluating cause-and-effect relationships in this conserved mitotic regulatory pathway. And now that the structure and activity of MPF and its components are better understood, it should be possible to evaluate the other mitotic factors from P. polycephalum to determine whether they are part of active MPF o r contribute to its activation. One clear message from the studies of MPF, ~34""'~and other mitotic factors is the remarkable evolutionary conservation of key regulators. Homologous gene products regulate mitotic and meiotic cycles as diverse as the fast embryonic cycle of oocytes of X. faevis, the budding mitotic cycle o f Sacch. cerevisiae with its mitotic spindle assembly long before the classical M-phase, the textbook M-Gl-S-G2 mitotic cycle of mammalian cells cultured in v i m , and the GI-less mitotic cycle of the syncytial plasmodium of P. polycephalum (Nurse, 1990; Ducommun ef al., 1990), a remarkable variety of mitotic cycle strategies. From this perspective, it is logical to consider the natural mitotic-cycle synchrony of the plasmodium as a valid tool to elucidate key events in the unperturbed eukaryotic mitotic cycle. Now that utility of the plasmodium for analysing periodic molecular events has been amply demonstrated, and DNA-transformation and gene-targeting systems have been developed, use of plasmodia from P. polycephalum can be productive and informative. One resource still needed lor research on P. polycephalum is a set of mitotic mutants. Such mutants can be isolated by elementary microbialgenetic techniques (Burland, 1986) and, in addition to clarifying some of

58

T. (i 13UHI.ANI) ET A!.

the cause-effect relationships for known molecular events, should also identify new mitotic regulators. This strategy has worked admirably in research on A. nidulans,where analysis of mitotic mutants identified gene products that were not known from the collection of yeast cell-cycle mutants (Morris and Enos, 1992). For example, mitosis in A. nidulans requires not only ~ 3 4 ‘ ~kinase ‘ ~ activity, but also the activity of another protein kinase, ‘~ et a f . , 1991). Other NimA, which acts independently of ~ 3 4 ‘ ~(Osmani mitotic regulators probably await discovery (Hartwell, 1991), and P. polycephalum is an obvious organism in which to seek them.

VI. Expression of Introduced Molecules A . DIFFUSION UPTAKE

One of the most exciting possibilities for molecular analysis in the plasmodium of P. polycephalum is the introduction of exogenous molecules. Growth of the plasmodium on a surface, such as a filter paper laid on top of a nutrient medium, allows easy access to the biomass. Radiolabelled precursors are readily taken up by the plasmodium from the medium underneath, or, very efficiently, from droplets containing radiolabelled solutes placed on top. Passive uptake of exogenous molecules even occurs when substances are added to submerged liquid “microplasmodial” cultures. Perhaps the most remarkable demonstration of this phenomenon is the uptake into microplasmodia of derivatized H3 histone, which becomes incorporated into nucleosomes (Prior er al., 1980). This powerful capability is one of the most underutilized features of the plasmodium. Nevertheless, for molecules that are difficult to obtain in large quantities, more direct methods may be needed for uptake into this receptive giant cell. R . MACKOINJECTION

A mammalian cell has a volume of about 1 pl. The oocyte of X. laevis is useful for analysing biological activity of specific molecules in part because upwards of 50 nl of liquid can be injected into one cell, 50,000 times the volume of a mammalian cell. By contrast, several microlitres of liquid can be injected into a small plasmodium at each of several sites (Kazarinoff and Ruth, 1986), that is, about a million times the volume of one mammalian cell. This is enough material for analysis not only of biological activity but also for recovering the injected material to evaluate molecular changes or specific associations. Rapid mixing of injected liquid is aided by rapid protoplasmic streaming in the plasmodium. This technique is useful for

I H E PROTISI‘ PHYSARUM 1’01.YCEI’HAI.UM

59

reintroducing mitotic activators into the plasmodium after purification to verify biological activity, and may also be useful to text extracts or specific gene products for ability to complement mitotic and other mutations. Given the large volumes that can be introduced, it should eventually be possible to macro-inject not only purified proteins in quantities large enough for easy detection, but perhaps also multiprotein complexes and structural assemblies such as DNA-histone complexes. This seems to be another technique worthy of more intense exploitation. For example, MPF or one of its components could be isolated from a plasmodium at one stage of the mitotic cycle, labelled or derivatized in some way, and injected into plasmodia at other stages of the mitotic cycle. One could then analyse not just its biological activity, but also its stability, cellular localization and association with other proteins. C.

DNA

TRANSFORMATION

Were Jane Austen a late 20th-century molecular biologist, she might well write a novel, as she did nearly two centuries ago, entitled Pride and Prejudice, modifying the first sentence only slightly, to read It is a truth universally acknowledged that a single-celled organism in possession of good cell biology must be in want of DNA transformation.

Such is the scientific value of introducing heterologous or modified genes into a cell that no organism is considered worthy of the attention of researchers today unless DNA transformation can be applied routinely. Fortunately, a DNA-transformation system is a technique no longer absent from the toolkit of research into P. polycephalurn. 1 , Transient Expression

The hph (hygromycin phosphotransferase) gene confers selectable resistance to the antibiotic hygrornycin in both prokaryotes and eukaryotes (Gritz and Davies, 1983). The sequence immediately 5’ to the ardC actin gene from P. polycephalurn (PardC) drives expression of hph in both fission yeast and budding yeast, conferring hygromycin resistance on these ascomycetes (Burland et al., 1991). Thus, PardC can act as a functional promoter in a broad range of eukaryotes. To develop transformation systems for P. polycephalurn, PardC was tested in transient expression experiments. The promoter was fused to the cat (chloramphenicol acetyltransferase) gene and amoebae were electroporated in the presence of plasmids carrying the PardC-cat fusion. Under precise buffer and electrical conditions, achievable using the Bio-Rad Gene

I- G IlIlRI.ANI) E 7

80,000

3

111

60,000

Al.

r \ L r a b

pPardC-luc

40,000 20,000 0

pPardB-luc 1 J

10

I

20

1

30

I

I

40

50

T-

60

70

-1 80

Time (h)

FIG. 7. Transient expression of lucifcrasc in Physarum amoebae. Amoebae were clectroporated by standard procedures (Burland ef al., 1992a) in the presence of vectors carrying either PardB-luc, the promoter region of the ardR actin gcnc translationally fused to the firefly lucifcrase (luc) gcnc, or PardC-luc, the promoter region of the ardC actin gene translationally fused to luc. RLU indicates relative light units (background emission without vector addition is 500 RLU). Following electroporation (time zero), equal numbers of amoebae were harvested and lysed after incubation in the growth medium, and lucifcrase activity in the lysates was measured with a luminomcter, using the assay kit provided by Promega Inc. (Madison, Wisconsin, USA).

Pulser, expression of the cat gene can be detected (Burland el al., 1992a). The promoter region for the ardB actin gene (PurdB)from P. polycephalum also drives cat expression in amoebae, although at a lower level than for PardC, reflecting the relative in vivo activities of the actin genes (Burland et al., 1992a). This is not the first report of transient expression of cat in amoebae, but the plasmid originally described for expressing cat in amoebae, pElori-CAT (McCurrah et ul., 1988), does not express cat in our tests. The development of reproducible methods for transient expression now allows further optimization of transformation conditions. Expression peaks only a few hours after transformation with plasmid DNA, then falls steadily (Burland el al., 1992a). The same is true whether the PardC or PardB promoter is driving cat, and expression of the luc (firefly luciferase) gene under the control of PurdB or PurdC shows similar kinetics (Fig. 7; J . Bailey, W. F. Dove and T. G. Burland, unpublished findings). The sequence TurdC is the transcriptional terminator located downstream of the ardC gene, and includes the putative polyadenylation site. Therefore, TurdC was expected to make mRNA expressed from plasmids more stable, but it has no measurable effect on level or duration of plasmid-borne gene expression in amoebae (Burland et al., 1992a). It remains to be determined how expression of introduced genes can be

THE PROTISI PHYSARUM POLYCEPHALUM

61

sustained at high levels over time. However, the availability of luciferase expression vectors (J. Bailey, W. F. Dove and T. G. Burland, unpublished findings) greatly accelerates progress in refining transformation techniques, as the assays are simple, non-radioactive, quantitative and rapid (deWet et a f . , 1987). 2. Stable Expression

Once conditions were refined for transient expression, rational attempts could be made to transform amoebae of P. pofycephafurn to stable, selectable hygromycin resistance using vectors carrying the PardC element translationally fused to hph. Following electroporation of amoebae with vectors carrying PardC-hph, several hygromycin-resistant transformants have been isolated, and all of those analysed carry the PardC-hph fusion integrated into the genome at one copy in each nucleus (Burland et af., 1992b). Hygromycin resistance is heritable through the life cycle in crosses, a necessary characteristic for many gene-disruption experiments, since disruptions will need to be carried out on diploid amoebae for genes that might be essential in the amoebal phase of the life cycle. The frequency of stable hygromycin-resistant transformants in these first experiments was less than per treated cell, but further refinement of transformation conditions and efficiency are in progress. Identification of the protistan equivalent of yeast ars elements should increase transformation efficiency sharply. The high-resolution physical mapping of replicons of P. pofycephafurn made possible by the synchrony of the plasmodium (Benard and Pierron, 1990) may aid in this goal. 3. Gene Targeting The stable transformation system developed allows addition of specific genes to the nuclear genome of P. pofycephafurn. However, geneticists prefer also to subtract information from the genome and, with integrative DNA transformation available, this can be achieved by gene disruption. As a test system, we sought to disrupt the ardD actin gene, which encodes an actin protein of divergent sequence and unknown function (Adam et a f . , 1991). The ardD gene encodes a product that appears not to be expressed in amoebae (Adam et a f . , 1991), and is not expected to be essential for amoebal viability. Therefore, in trying to disrupt ardD, we used the same haploid strain of amoebae, LU352 (Dee et af., 1989), that was used in all of the other transformation experiments. A genomic clone for ardD already contains a deletion around the intronYexon-6 junction (Adam et a f . , 1991). We made a further 5’ deletion of

62

‘I

(i RlJHl AN11 I. f A /

the first 0.75 kb of the coding region, leaving a sequence of approximately 1.6 kb of unbroken homology with the wild-type ardD allele. We then cloned this doubly deleted allele into a vector containing PardC-hph alongside the ardD allele, and transformed amoebae to hygromycin resistance. Southern blotting indicated that, in a set of five transformants, the wild-type allele of ardD was replaced by the deleted allele in three, indicating that gene disruption will be readily feasible for any gene for which a DNA clone exists (Burland et al., 1992b). Considering that integrative DNA transformation was developed for P. polycephalum less than a year before the ardD targeting experiments, we can hardly be disappointed at this progress. The principal capability now sought is increasing transformation efficiency to a level that will permit DNAmediated complementation of mutants.

V11. Concluding Remarks The protists have been overlooked as experimental organisms among much of the research community. Some protists have been difficult to work with, partly due to their prodigious wealth of cell-biological features. Also, the genetic acrobatics possible with fungi has rightly attracted cell biologists to ascomycetes and basidiomycetes. Yet, despite the heavy weighting of research towards non-protists, key breakthroughs have been made with protists. Self-splicing RNA was discovered in the protist Tetrahyrnenu thermophilu (see Gold, 1990); acetylation of a-tubulin. now known to be widespread among eukaryotes, was discovered in the protist Chlamydornonas reinhardtii (L’Hernault and Rosenbaum, 1983; Piperno and Fuller, 1985); a defined temporal order of replication for individual genes and pairs of alleles was established in the protist P. polycephalurn (Pierron et al., 1984); and the exquisite mitotic synchrony of the plasmodium in P. polycephalum has now revealed that, for some genes, two alleles at a locus can replicate at distinct times in the S-phase (Cunningham and Dove, 1993). It is ironic that protists, named from the Greek protos, meaning “first”, should be last in line as research organisms. Now that sophisticated molecular approaches to cell-biological problems are possible in P. polycephalurn and in other protists, further studies will inevitably reveal more of the knowledge that has been awaiting discovery in these organisms for so many millenia. VIII. Acknowledgements

We thank our collaborators Dominick Pallotta and Jennifer Dee for providing clones of genes and strains that made much of our progress

THE PROTIST P H Y S A R U M P O L Y C E P H A L U M

63

possible, and Roger Anderson for communicating data on developmental mutants ahead of publication. We are grateful to Amy Moser for comments on the manuscript and Linda Clipson for preparation of figures. Our research has been supported by programme project grant CA23076 and core grant CA07175 from the National Cancer Institute. REFERENCES

Adam, L., Laroche, A , , Barden, A., Lemieux, G . and Pallotta, D . (1991). Gene 106, 79. Adlakha, R. C., Sahasrabuddhe, C. G., Wright, D. A. and Rao, P. N. (1983). Journal of Cell Biology 97, 1707. Adlakha, R. C . , Shipley, G . L., Zhao, J . Y., Jones, K. B., Wright, D. A , , Rao, P. N. and Sauer, H. W. (1988). Journal of Cell Biology 106, 1445. Akhavan-Niaki. H., Mir, L., Oustrin, M. L., Moisand, A. and Wright, M. (1991). Journal of Cell Science 99, 265. Albertini, C., Akhavan-Niaki, H. and Wright, M. (1990). Cell Motiliry and the Cytoskeleton 17, 267. Aldrich, H. C. (1968). Journal of General Microbiology 50, 217. Aldrich, H.C. (1069). American Journal of Botany 56, 290. Alexopoulos, C. J. (1982). In “Cell Biology of Physarum and Didymium” (H. C. Aldrich and J. W. Daniel, eds), vol. 1, p. 3. Academic Press, New York. Anderson, R. W. and Dee J. (1977). Genetical Research (Cambridge) 29, 21. Anderson, R. W., Cooke, D. J. and Dee, J. (1976). Genetical Research (Cambridge) 89, 29. Anderson, R. W.. Burland, T. G. and Dee, J. (1986). Physarum Newsletter 18(2), 2 . Anderson, R. W., Hutchins, G . , Gray, A , , Price, J. and Anderson, S. E. (1989). Journal of General Microbiology 135, 1347. Bai1ey.J.. Anderson, R. W. and Dec, J. (1987). Protoplasma 141, 101. Bailey, J., Anderson, R. W. and Dee, J. (1990). Journalof General Microbiology 136, 739. Bailey, J., Solnica-Krezel, L., Anderson, R. W. and Dee, J. (1992). Journal of General Microbiology 138, 2575. Bailly, E. and Bornens, M. (1992). Nature 355, 300. Baroin. A., Perasso, R., Qu, L.-H., Brugerolle, G., Bachellerie, J.-P. and Adoutte, A. (1988). Proceedings of the National Academy of Sciences of the United States of America 85, 3474. Benard, M. and Pierron, G. (19%)). Experimental Cell Research 186, 299. Benne, R. (1990). Trends in Genetics 6, 177. Bernier, F.. Pallotta, D. and Lemieux, G . (1986). Biochimica et Biophysica Acta 867, 234. Bernier, F.. Ixmieux, G . and Pallotta, D . (1987). Gene 59, 265. Bernstein, P. and Ross, J . (1989). Trends in Biochemical Science 14, 373. Binettc, F., Bernard, M.,Laroche, A,, Pierron, G . , Lemieux, G. and Pallotta, D. (1990). DNA and Cell Biology 9, 323. Birkett, C . R., Foster, K. E. and Gull, K. (1985a). In “Molecular Genetics of Filamentous Fungi” (W. E. Timberlake, ed.), p. 265. Alan R. Liss, New York. Birkett, C. R., Foster, K. E., Johnson, L. and Gull, K. (1985b). FEBS Letters 187, 211. Blackburn, E. H . (1991). Nature 350, 569. Blindt, A. B., Chainey, A. M., Dee, J. and Gull, K. (1986). Protoplasma 132, 149. Borisy, G . G., Marcum, J. M.,Olrnsted, J. B., Murphy, D. B. and Johnson, K. A. (1975). Annals of the New York Academy of Sciences 253, 107. Bradbury, E. M. (1992). Bioessays 14, 9. Bradbury, E. M., Inglis, R. J., Matthews, H. R. and Langan, T. A. (1974). Nature 249.5.53. Braun, R., Mittermayer, C. and Rusch, H. P. (1965). Proceedings of the National Academy of Sciences of the United States of America 53, 924. Braun, R . and Evans, T. E. (1969). Biochimica et Biophysica Acta 182, 511.

64

T. (i. RUR1.AND ET Al..

Bravo, R. and Celis, J. E. (1980). Journal of Cell Riology 84, 795. Broek, D., Bartlett, R., Crawford, K. and Nurse, P. (1991). Nature 349, 388. Burland, T. G . (1978). PhD Thesis: University of Leicester. Burland, T. G . (1986). In “The Molecular Biology of Physarumpolycephalum” (W. F. Dove, J. Dee, S . Hatano, F. B. Haugli and K.-E. Wohlfarth-Botterman, eds), p. 19. Plenum Press, Lmndon. Burland, T. G. and Dee, J. (1979). Genetical Research (Cambridge) 34, 33. Burland, T. G . and Dee, J. (1980). Molecular and General Genetics 179, 43. Burland. T. G . , Chainey, A. M., Dee, J. and Foxon, J . L. (1981). Developmental Biology 85, 26. Burland, T. G . . Gull, K., Schedl, T., Boston, R.S. and Dove, W. F. (1983). Journal o f Cell Riology 97, 1852. Burland, T. G . , Schcdl, T., Gull. K. and Dove, W. F. (1984). Generics 108, 123. Burland, T. Ci.. Paul, E. C., Oetliker, M. and Dove. W. F. (1988). Molecular and Cellular Biology 8 , 1275. Burland. T. G . , Pallotta. D.. Tardif, M. C., Lemieux, G. and Dove, W. F. (1991). Gene 100, 241. Burland, T. G . , Bailey, J., Adam, L., Mukhopadhyay, M.J . , Dove, W . F. and Pallotta, D. (1992a). Current Genetics 21. 393. Burland, T. G . , Bailey, J., Pallotta, D . and Dove, W. F. (l992b). Cell Biology International Reports 16. 1 I 1 1. Carrino, J. J. and Laffler, T . G . (1986). Journal of Cell Biofogy 102, 1666. Carrino, J . J . , Kueng, V . , Braun, R. and Laftler, T. G. (1987). MolecularandCellular Biology 7, 1933. Collins, 0 . R. and Bctterley, D. A . (1982). In “Cell biology of Physarurn and Didymium” (H. C. Aldrich and J . W . Daniel, eds), vol. I , p. 25. Academic Press, New York. Collins, 0. R. and Tang, H.-C. (1977). Mycologia 69, 421. Cooke, D. J . and Dee, J. (1974). Genetical Research (Cambridge) 23, 307. Cooney, C. A , . Eykholt, R. L. and Bradbury, E. M. (1988). Journal of Molecular Biology 204, 889.

Coren, J. S., Epstein, E. M. andVogt, V. M. (1991). Molecularand Cellular Riology 11.2282. Cunningham, D. B. (1992). PhD Thesis: University of Wisconsin. Cunningham, D . H . and Dove, W. F. (1903). Molecular and Cellular Riology 13, 449. Cunningham, D . B., Buchschacher, G . L.,Jr, Burland, T. G., Dove, W. F., Kessler, D . and Paul. E . C. A . (1993). Journal of General Microbiology, 139, 137. Daniel, D. C. and Johnson, E. M. (1989). Nucleic Acidr Research 17, 8343. Dee, J . (1960). Nature 185, 780. Dee, J. (1966). Journal of Protozoology 13, 610. Dee, J . (1982). I n ‘ T e l l Biology of Physarum and Didymium” ( H . C. Aldrich and J . W. Daniel, eds), vol. I , p. 212. Academic Press, New York. Dee. J. and Anderson, R . W. (1984). Journal of General Microbiology 131, 1167. Dee, J . , Foxon, J. 1.. and Anderson. R . W. (1989). Journalof General Microbiology 135, 1567. Delahodde, A , , Goguel, V . , Becam. A . M., Creusot, F., Perea. J., Banroques. J. and Jacq, C. (1980). Cell 56, 431. DclCastillo, L., Oustrin, M. L. and Wright, M. (1978). Molecular and General Genetics 164, 145. deWet, J . R..Wood. K. V., DeLuca, M.,Helinski. D . R. and Subramani, S. (1987). Molecular clnd C’ellular Rkilogy 7, 725. Diggins, M. A . and Dove, W. F. (1987). Journal of Cell Biology 104, 303. Diggins-Gilicinski, M . , Solnica-Krezel, L., Burland, T. G., Paul, E. C. and Dove, W. F. (1989). Journal of Cell Science 94. 217. Dove, W. F., Dce, J . , Hatano, S . , Haugli, F. B. and Wohlfarth-Botterman, K.-E. (eds) ( 1986). “The Molecular Biology of Physarum polycephalum”. Plenum Press, London. Ducommun, B. and Wright, M. (1989). European Journal of Cell Biology SO, 48. Ducommun, B . , Tollon, Y., Cares, M.,Beach, D . and Wright, M . ( 1 9 0 ) . Journal of Cell Sciencc 96. 683.

TllE PROTIST PHYSARUM POLYCEPHALUM

65

Dunphy, W. G., Brizuela, L., Beach, D. and Newport, J . (1988). Cell 54, 423. Engberg, J. and Nielsen, H . (1990). Nucleic Acids Research 18, 6915. Evans, 7‘.E . , Daniel, J. W., Evans, H. E., Matthews, H. R., Mohberg, J . M. and Tyson, J. T. (1982). In “Cell Biology of Physarum and Didymium” (H. C. Aldrich and J . W. Daniel, eds), vol. 11, p. 351. Academic Press, New York. Feagin, J. E . , Abraham, J . M. and Stuart, K. (1988). Cell53, 413. Ferris. P. J. (1985). Gene 39, 203. Ferris, P. J . , Vogt, V. M. and Truitt, C. L. (1983). Molecular and Cellular Riology 3, 635. Forney, J . , Henderson, E. R. and Blackbum, E. H. (1987). Nucleic Acidr Research 15,9143. Foster, K. E., Burland. T. G . and Gull, K. (1987). European Journalof Riochemistry 163,449. Fry, J. and Matthews, H. R. (1987). Experimental Cell Research 168, 173. Fulton, C. and Simpson. P. A . (1976). In “Cell Motility” (R. Goldman. T. Pollard and J. Rosenbaum, eds), p. 987. Cold Spring Harbor Press, Cold Spring Harbor. Funderud, S . . Andreassen, R. and Haugli, F. (1978a). Cell 15, 1519. Funderud, S., Andreassen, R. and Haugli, F. (1978b). Nucleic Acids Research 5, 3303. Gautier, J., Norbury, C., Lohka, M. and Maller, J. (1988). Cell54. 433. Gingold, E. C., Grant, W. D., Wheals, A. E. and Wren, M. (1976). Molecular and General Generics 149, 1 15. Glyn, M. and Gull, K. (1990). Protoplasma 158, 130. Gold, L. (1990). The New Riologist 2, 1. Goldman, M. A. (1988). Rioessays 9, 50. Gonzalez-y-Merchand, J. A. and Cox, R. A. (1988). Journalof Molecular Riology 202, 161. Green, L. L. and Dove, W. F. (1984). Molecular and Cellular Riology 4, 1706. Green, L. L. and Dove, W. F. (1988). Journal of Molecular Biology 200, 321. Green, L. L., Schroeder, M. M., Diggins, M. A. and Dove, W. F. (1987). Molecular and Cellular Biology 7, 3337. Gritz, L. and Davies, J . (1983). Gene 25, 179. Grobner, P. and Loidl, P. (1985). Journal of Cell Riology 100, 1930. Gunderson, G . G . , Khawaja, S. and Bulinski, J . C. (1987). Journal of Cell Biology 105,251, Guttes, S. and Guttes, E. (1968). Journal of Cell Riology 37, 761. Hamelin, M.,Adam, L., Lemieux, G . and Pallotta, D. (1988). DNA 7 , 317. Hamlin, J. L., Vaughn, J. P . , Dijkwel, P. A,, Leu, T.-Z. and Ma, C. (1991). Current Opinion in Cell Riology 3, 414. Hardman, N. (1986). In “The Molecular Biology of Physarum polycephalurn” (W. F. Dove, J. Dee, S. Hatano, F. B. Haugli and K.-E. Wohlfarth-Botterman, eds). p. 39. Plenum Press. London. Harel, A., Zlotkin, E . , Nainudel-Epszteyn, S. and Feinstein, N. (1989). Journal of Cell Science 94, 463. Hartwell, L. H. (1991). Genetics 129, 975. Hasegawa, M.,Iida,Y.,Yano,T.,Takaiwa, F. andlwabuchi, M. (1985). JournalofMolecular Evolution 22, 32. Hatano, S. (1986). In “The Molecular Biology of Physarum polycephalum” (W. F. Dove, J. Dee, S. Hatano, F. B. Haugli and K.-E. Wohlfarth-Botterman, eds), p. 165. Plenum Press, London. Havercroft, J . C. and Gull, K. (1983). European Journal of Cymlogy 32, 67. Heath, B. (1982). international Review of Cell Riology 64. 1. Heubner, V. D. and Matthews, H. R. (1985). Journal of Biological Chemistry 260, 16106 Hobohm, U., Hildebrandt, A. and Rensing, L. (1990). Experimental Cell Research 191, 332. Howard, A. and Pelc, S. (1953). Heredity 6(suppl.). 261. Hoyle, H. D. and Raff, E. C. (1990). Journal of Cell Biology 11 I , 1009. Huang, J.. Wei, Y., Kim, Y., Osterberg, L. and Matthews, H. R. (1991). Journalof Biological Chemistry 266, 9023. Hurst. L. D. (1991). Nature 354,23. Jalouzot, R., Toublan, B., Wilhelm, M. L. and Wilhelm, F. X. (1985). Proceedings of rhe National Academy of Sciences of the United Slates of America 82. 6475. Jerzmanowski, A. and Krezel, A. M. (1986). Riochembtry 25, 6495.

66

T. G . RIJRLAND 6 T Al..

Jerzmanowski. A . and Maleszewski, M . (1985). Biochemistry 24, 2360. Johansen, S. (1991). PhD Thesis: University of Tromsti, Norway. Johansen, T., Johansen, S. and Haugli, F. B. (1988). Current Genetics 14, 265. Jones, E . P . , Mahendrdn, R., Spottswood, M . R., Yang, Y. C . and Miller, D. L. (1990). Current Genetics 17. 331. Kawano, S., Suzuki, T . , and Kuroiwa, T. (1982). Riochimica et Riophysica Aria 696, 290. Kawano, S., Anderson, R . W., Nanba, T. and Kuroiwa, T. (1987a). Journal of General Microbiology 133, 3175. Kawano. S., Kuroiwa, T. and Anderson, R. W. (1987b). Journal of General Microbiology 133, 2539. Kawano, S., Takano, H., Mori, K. and Kuroiwa. T. (1991). Protoplasma 160, 167. Kazarinoff, M. N. and Ruth, D. C. (1986). In “The Molecular Biology of Physarum polycephalum” (W. F. Dove, J. Dee, S. Hatano, F. B. Haugli and K.-E. WohlfarthBotterman, eds), p. 287. Academic Press, New York. Kimble. M., Incardona, J. P. and Raff, E. C. (1989). Developmental Riology 131, 415. Kirouac-Brunet, J., Masson. S. and Pallotta, D. (1981). Canadian Journal of Genetics and Cytology 23, 9. Kobayashi, T., Tagaki, T., Konishi, K., Hamada, Y ., Kawaguchi, M .and Kohama, K. (1988). Journal of Riological Chemistry 263, 305. Kochanski, R . S. and Borisy, G . G . (1990). Journalof Cell Biology 110, 1599. Kubbies, M . and Pierron, G. (1983). Experimental Cell Research 149. 57. Kubbies, M., Wick, R., Hildebrandt, A . and Sauer, H. W . (1986). Cytometry 7, 481. Kuroiwa, T. (1985). Microbiological Science 2, 267. L’Hernault, S. W. and Rosenbaum, J. L. (1983). Journal of Cell Biology 97, 258. Laffler, T. G . (1987). European Journal of Cell Riology 43, 1 9. Laffler, T. G. and Carrino, J. J . (1986). Bioessays 5, 62. Laffler, T. G. and Dove, W. F. (1977). Journal of Bacteriology 131, 473. Laffler, T. G . and Tyson. J. T. (1986). In “The Molecular Biologyof Physarumpolycephalum” (W. F. Dove, J. Dee, S. Hatano, F. B. Haugli and K.-E. Wohlfarth-Botterman, eds), p. 79. Plenum Press. London. Laffler, T. G., Chang, M. T. and Dove, W. F. (1981). Proceedings of the National Academy of Sciences of the United States of America 78, SOOO. Lai. E. Y . , Remillard, S . P. and Fulton, C. (1984). In “Molecular Biologyof the Cytoskeleton” (G. G . Borisy, D. W. Cleveland, and D . B. Murphy, eds), p. 257. Cold Spring Harbor Laboratory, Cold Spring Harbor. Lcdizet. M. and Piperno, G . (1991). Methods in Enzymology I%, 264. Lcfebvre, P. A. and Rosenbaum, J. L. (1986). Annual Review of Cell Riology 2, 517. Lenaers, G . , Niclsen, H., Engberg, J . and Herzog, M. (1988). Riosystems 21, 215. Lcwin, B. (1990). Cell 61. 743. Little, M., Krammer, G . , Singhofcr-Wowrd, M. and Luducna, R. F. (1986). Annals of the New York Academy of Sciences 466,8 . Loidl, P. (1988). FERS Letters 227, 91. Loidl, P. and Grobner, P. (1982). Experimental Cell Research 137, 469. Loidl, P. and Grobncr, P. (1987a). Journal of Biological Chemistry 262, 10195. Loidl, P. and Grobner, P. (1987b). Nucleic Acids Research 15, 8351. Loidl, P. and Sachsenmaier, W. (1982). European Journal of Cell Biology 28, 175. Long, E. 0. and Dawid. I. B. (1980). Annual Review of Riochemistry 49, 727. Lowery, R., Hung, L., Knoche, K. and Bandziulis, R. (1992). In “Promega Notes 38”, p. 8. Promega, Madison, WI. McCurrach, K., Glover, L. A. and Hardman, N. (1988). Current Genetics 13, 71. McCurrach, K. J., Rothnie, H. M., Hardman, N. and Glover, I-. A . (1990). Current Genetics 17, 403.

Mahcndran, R., Spottswood, M. R. and Miller, D. L. (1991). Nature 349, 434. Margulis, L. and Schwartz, K. V. (1982). “Five Kingdoms: An Illustrated Guide to the Phyla of Life on Earth”. Freeman, San Francisco. Martel, R., Tessier, A., Pallotta, D. and Lemicux, G. (19M). Journal of Racteriology 170, 4784.

1 H E PROTISI PH YSARUM POI. YCEPHA LUM

67

Masui, Y. and Markert, C. L. (1971). Journal of Experimental Zoology 177, 129. Matthews, H. R. and Bradbury, E. M. (1982). In “Cell Biology of Physarum and Didymium” (H. C. Aldrich and J. W. Daniel, eds), vol. I , p. 317. Academic Press, New York. Meland, S., Johansen, S., Johansen, T., Haugli, K. and Haugli, F. (1991). Currenr Genetics 19, 55. Metzenberg, R. L. (1990). Genetics 125, 457. Mir, L., DelCastillo, L. and Wright, M. (1979). FEBS Microbiology Letters 5 , 43. Mohberg, J. (1982a). In “Cell Biology of Physarum and Didymiurn” (H. C. Aldrich and J. W. Daniel, eds), vol. 11. p. 273. Academic Press, New York. Mohberg, J. (1982b). In “Cell Biology of Physarum and Didymium” ( H . C. Aldrich and J. W. Daniel, eds), vol. I, p. 253. Academic Press, Ncw York. Mohberg, J. and Babcock, K. L. (1982). In “Cell Biology of Physarum and Didymium” (H. C. Aldrich and J. W. Daniel, eds), vol. I, p. 273. Acadmic Press, New York. Monteiro, M.J. and Cox, R.A. (1987a). Journal of Molecular Biology 193, 427 Monteiro, M. J. and Cox, R. A. (1987b). FEBS Letters 217, 260. Moms, N. R. and Enos, A . P. (1992). Trends in Genetics 8, 32. Mueller, R. D., Yasuda, H., Hatch, C. L., Bonner, W. M. and Bradbury, E . M. (1985). Journal of Biological Chemistry 260, 5147. Muscarella, D. E. and Vogt, V. M. (1989). Cell 56, 443. Muscarella, D. E., Ellison, E. L., Ruoff, B. M. and Vogt, V. M. (1990). Molecular and Cellular Biology 10, 3386. Nader, W. F., Shipley, G . L., Huetterrnan, A. and Holt, C. E. (1984). Developmental Biology 103, 504.

Nader, W. F., Isenberg, G . and Sauer, H. W. (1986). Gene 48, 133. Newlon, C . S., Sonenshein, G . E. and Holt, C. E. (1973). Biochemistry 12, 2338. Nurse, P. (1990). Nature 544, 503. Nurse, P. (1991). Nature 354,356. Nygaard, 0. F., Guttes, S. and Rusch, H. P. (1964)). Biochimica el Biophysica Acta 38, 298. Oakley, B. R. and Morris, N. R. (1980). Cell 19, 255. Ohta, T., Kawano, S. and Kuroiwa, T. (1991). Protoplasma 163, 114. Olive, L. S. (1975). “The Mycetozoans”. Academic Press, New York. Olmsted, J . B. (1986). Annual Review of Cell Biology 2 , 421. Oppenheim, A. and Katzir, N. (1971). Experimental Cell Research 68, 224. Osmani, A . H., McGuire, S. L. and Osmani, S. A. (1991). Cell 67, 283. Pagh, K. 1. and Adelman, M. R. (1988). Cell Motility and the Cytoskeleton 11, 223. Pallotta, D . , Laroche, A , , Tessier, A , , Shinnick, T. and Lemieux, G. (1986). Biochemistry and Cell Biology 64, 1294. Paul, E. C. A,, Roobol, A , , Foster, K. E . and Gull, K. (1987). Cell Motility and the Cytoskeleton 7, 272. Paul, E . C. A , , Burland, T . G. and Gull, K. (1989). Journalof General Microbiology 135,623. Paul, E. C . A , , Cunningham, D. 8 . . Buchschacher, G . L., Jr, Dove, W. F. and Burland, T. G. (1992). Journal of General Microbiology 138, 229. Pesis, K.. Wei, Y., Lewis, M. and Matthews, H. R. (1988). FEBS Letters 239, 151. Pierron. G . (1986). In “The Molecular Biology of Physarum polycephalum” (W. F. Dove, J, Dee, S. Hatano, F. B. Haugli, and K.-E. Wohlfarth-Botterman, eds), p. 67. Plenum Press, London. Pierron, G., Sauer, H. W.. Toublan, B. and Jalouzot, R. (1983). European Journal of Cell Biology 29, 104. Pierron, G . , Durica, D. S. and Sauer, H. W. (1984). Proceedings of the Narional Academy of Sciences of the United States of America 81, 6393. Pierron, G., Benard, M., Puvion, E.. Flanagan, R., Sauer, H. W. and Pallotta, D. (1989). Nucleic Acids Research 17, 553. Piperno, G . and Fuller, M. T. (1985). Journal of Cell Biology 101, 2085. Poulter. R. T. M. and Dee, J. (1968). Genetical Research (Cambridge) 12, 71. Prescott, A. R.. Foster, K. E.. Warn, R. M. and Gull, K. (1989). Journal of Cell Science 92. 595.

68

T. G HURI.ANI) b.’7 A1

Prescott, 11. M. (1976). In “Reproduction o f Eukaryotic Cells”, p. 36. Academic Press, New York. Prior. <1. P., Cantor. C. R . . Johnson, E. M. and Allfrcy, V. G . (1980). Cell 20, 597. Raff, E. C. (1984). Journal of Cell Biology 99, 1. Rao, P. N. and Johnson, R . T. (1970). Nature 225, 159. Raub, T. J . and Aldrich, H . C. (1982). In “Cell Biology of Physarum and Didymium” (H. C. Aldrich and J . W. Daniel, eds), vol. 11, p. 21, Academic Press, New York. Roobol, A , , Pogson. C . I . and Gull, K. (1980).Experimental Cell Research 130, 203. Roobol, A , . Havcrcroft, J. C. and Gull, K. (1982). Journal of Cell Science 55, 365. Roobol, A , , Wilcox. M., Paul, E. C . A. and Gull, K. (1984). European Journal of Cell Biology 33, 24. Ross. I . K. (1957). American Journal of Botany 10, 843 Rothnie, H . M., McCurrach. K. J . , Glover, L.. A. and Hardman. N. (1991). Nucleic Acids Research 19, 279. Sachsenmaier, W . and Rusch, H. P. (1964). Experimental Cell Research 36, 124. Sachsenmaier, W . , Bohnert, E . , Clausnizer, B. and Nygaard, 0.(1970). FEB.SLeffer.s10, 185. Salles-Passador, I . , Moisand, A , , Planques. V. and Wright, M. (1991).Journalof CellScience 100, 509.

Salles-Passador, I . , Moisand, A , , Planques, V. and Wright M. (1992). Publications of the University of Innsbruck 183, 31. Sassc. R . , Glyn, M. C.. Birkett, C. R. and Gull, K. (1987). Journal of Cell Biology 104, 41. S w a r d , L., Laroche, A , , Lxmieux, G. and Pallotta, D. (1989). Biochimica et Biophysica Acta 1007.264.

Schedl. T. and Dove, W . F.(IYX2). Journal of Molecular Biology 160, 41. Schedl, T . , Burland, T. G . . Dove, W. F., Roobol, A , , Paul, E. C. A , , Foster, K. I:. and Gull, K. (1984a). In “Molecular Biology o f the Cytoskeleton” (G. G . Borisy, D. W . Cleveland and D . B. Murphy, eds). p. 235. Cold Spring Harbor Laboratory, Cold Spring Harbor. Schedl,T.,Burland,T.G.,GuII, K. andDove, W. F. (1984b). JournalofCellBiology99, 155. Schedl, T., Owens, J . , Dove, W. F. and Burland. T. G. (1984~).Genetics 108, 143. Scott, J . (1989).Current Opinion in Cell Biology I , 1141. Shipley. G . I.. and Holt, C. E. (1982). Developmental Biology 90, 110. Shiplcy, G . L. and Sauer, H . W. (1989). European Journal of Cell Biology 48, 95. Siltlow, C. D. (1991). Protoplasma 164, 9. Silliker. M. E. and Collins, 0. R. (1988). Molecular and General Genetics 213, 370. Singhofer-Wowra, M., Clayton, L.. Dawson, P., Gull, K. and Little, M. (1986a). European Journal of Biochemistry 161. 669. Singhofer-Wowra, M., little, M., Clayton, L., Dawson, P. and Gull, K. (1986h). Journal of Molecular Biology 192, 919. Solnica-Krezel, LA., Dove. W. F. and Burland, T. G. (198X). Journalof General Microbiology 134, 1323.

Solnica-Krezel, L.., Diggins-Gilicinski, M., Burland. T. G . and Dove, W. F. (1990). Journal of Cell Science 96, 383. Solnica-Krezel, I . . , Burland, T. G . and Dove, W. F. (1991). Journalof Cell Biology 113, 591. Solnica-Krezel, L., Bailey, J., Dove, W . F., Dee, J. and Anderson, R. W. (1992). Submitted for publication. Sullivan, K. F. (1988).Annual Review of Cell Biology 4, 687. Sullivan, W . , Minden, J . and Alberts, B. (1990). Development 110, 311. Sweeney, G. E., Watts, D. 1. and Turnock, G. (1987). Nucleic Acids Research 15, 933. Takano, H.. Kawano, S., Suyama, Y. and Kuroiwa, T. (1990). Current Genetics 18, 125. Takano. H . , Kawano. S. and Kuroiwa, T. (IWI) Current Genetics 20, 315. Tanaka. K. (1973). Journal of Cell Biology 57, 220. Turnock, G . , Morris, S. R. and Dee, J. (1981). European Journal of Biochemistry 115, 533. Tyson, J. J . (1982). In “Cell Biology of Physarum and Didyrniurn” (H. C. Aldrich and J . W. Daniel, eds), vol. 1, p. 61. Academic Press, New York. Uyeda, -1. 0 . P. and Furuya, M. (1985). Proroplasma 126, 221.

69 Uyeda, T. 0. and Furuya, M. (19x9). Journal of ('ell Biology 108, 1727. Uyeda, T. Q. and Kohama, K. (19x7). Experimental Cell Research 169, 74. Vogt, V. M. and Braun, R . (1977). European Journal of Biochemistry 807. 557. Walden, P. D . , Blindt, A . B., Birkett, C. R., Cox, R. A . and Gull. K. (l989a). European Journal of Biochemistry 185. 383. Walden, P. D., Monteiro, M. J . , Gull, K. and Cox, R . A . (198Yb). European Journal oj Biochemistry 181, 583. Wang, D., Vaillasante, A , , Lewis, S. A . and Cowan, N. J. (1986). Journal of ('ell Bioloxy 103, 1903. Waterborg, J. H. and Matthews, H. R. (19x2). Experimenful Cell Reseurch 138, 462. Waterborg, J. H. and Matthews. H. R. (1984) European Journal of Biochemistry 142, 320. Weatherbee, J. A , , May. G . S.. Gambino. J. and Morris, N. R. (198.5). Journal of C'i4l Biology 101, 706. Wei, Y . , Morgan, J . E. and Matthews, H . R. (1989). Archives of Riochemistryand Riophysic..s 268. 546. Welker, D . L., Hirth, K. P. and Williams, K. L. (1985). Molecularand Cellular Biology 5,273. Werenskiold, A . K . , Poetsch, B. and Haugli, F. (1988). European Journal of Biochernisrry 174, 49 1. Wheals, A . E. (1973). Genetical Research (Cambridge) 21, 79. Wheals, A. E., Grant, W. D. and Jockusch, B. M. (1976). Molecular and General Genetics 149, 1 1 1 . Wilhelrn, M. L. and Wilhelm, F. X . (1989). Journal of Molecular Evolution 28, 322. Wilhelm, M. L., Toublan, B., Fujita, R . A . and Wilhelm, F. X. (1988). Biochemicul and Biophysical Research Communications 153, 162. Wright, M. and Tollon. Y. (1979). European Journal of Riochemistry 96. 177. Wright, M., Moisand, A. and Mir, L. (1980). Protoplasma 105, 149. Wright, M., Albertini, C., Planques, V., Salles, I., Ducommun, B . , Gely, C. AkhavanNiaki, H . , Mir, L . , Moisand, A . and Oustrin, M. L. (19x8). Biology of the ( P I 1 63, 239. Yasuda, H . , Mueller, R. D., Logan, K. A. and Bradbury, E. M. (1986). JournalofRioloRicul Chemistry 261, 2349. Yen, T. J . , Gay, D. A.. Pachtcr, J . S. and Cleveland, D. W. (1988). Molecular and C'eIIi4lur Biology 8, 1224. Youngman, P. J., Adler, P. N., Shinnick, T. M. and Holt, C. E. (IY77). Proceedings o f t h e National Academy of Sciences of the United Stutes of America 74, 1120. Youngman, P. J . , Anderson, R . W. and Holt, C. E. ( I Y X I ) . Genetics 97, 513. Zakian, V. A . (19x9). Annual Review of Genetics 23, 579. Zellweger, A . , Ryscr, U . and Braun. R. (1972). Journalof Molecular Biology 64, 681. Zheng, Y., Jung, M. K. and Oakley, B . R. (1991). C'ell65, 817.

This Page Intentionally Left Blank

Selenium Metabolism in Micro-organisms’ JOHANN HETDER and AUGUST BOCK Lrhrstuhl fiir Mikrobiologie iier Universitat Miinchen, Maria- Ward-Strasse l a , 0-80638 Munchen, Germany

1. 11. 111.

IV. V. VI. VI1. VIII. IX.

Introduction . . . . . . . . . . . . . . , Selenium-containing enzymes . . . . . . . . . . A. Biochemistry of prokaryotic selenoproteins . . . . . . Selenium-dependent enzymes not containingselenocysteineresidues B. . . Selcnium-containing tRNAs . . . . . . . . . Biosynthesisof selenoproteinsandseleno-tRNAs . . . . . . A . Selenoproteins . . . . . . . . . . . . B. Seleno-tRNAs . . . . . . . . . . . . Selenium versus sulphur , . . . . . . . . . . . A . In catalysis . . . . . , . . . . . . . B. Competition during incorporation . . . . . . . . Transport and metabolism of selenium-containingcompounds . . . Geochemistry of selenium . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . References . . . . . . . . . . . . . . .

71

72 73 86

88 89 89

95 96

96 97

98 100 103 104 104

I. Introduction Selenium is one of the “youngest” elements for which a biological function has been detected; it is, however, also the element for which the greatest progress has been achieved in terms of understanding its specific biochemical function. The discovery of selenium as a biological element dates back to the report of Pinsent in 1954 that formate oxidation by cell suspensions of Escherichia cofi requires growth of cells in a medium containing molybdate and selenite. It is now known that E . cofi possesses three formate dehydrogenase (FDH) isoenzymes and that they all contain

’ Dedicated to Professor M. H. Zenk on the occasion of his 60th birthday. AI)VANC‘IIS IN MlCROHlAl 1’HYSIOI.OGY. VOI.. 35 ISRN I& 12412773.S2

Copyright0 1993, by AcadcmicPressI.imi1ed All rights of reproduction in any formreserved

72

J

IIbI1)F.H A N I ) A I3O('K

a selenoprotein subunit (Cox et al., 1981; Sawers et al., 1991); their synthesis, therefore, requires medium supplementation with selenite. With the discovery that sclenium is an essential nutrient for rats (Schwartz and Foltz, 1957), emphasis in selenium biochemistry shifted to investigation of its physiological and biochemical role in animals and humans. However, research on selenium in the field of microbial biochemistry was revived with two central findings. First, the chemical structure of the organoselenium compound in a microbial enzyme, namely glycine reductase, was identified as selenocysteine (Cone et al., 1976), and secondly, it was demonstrated that selenocysteine incorporation into a bacterial enzyme, namely formate dehydrogenase H from E. cofi,occurs cotranslationally (Zinoni et al., 1987) and is directed by an in-frame TGA codon in the gene (Zinoni et al., 1986). At the same time, selenocysteine incorporation into a mammalian enzyme was reported to be determined by a TGA codon in the gene (Chambers et al., 1986). Because of their amenability to genetic and biochemical approaches, the microbial systems since then have given rise to a wealth of information on the unique molecular biology and biochemistry of this element. Selenium is a n element belonging to group VI of the periodic table. Like sulphur, selenium occurs in the oxidation states -2, 0, +2, +4 and +6, with the difference that the +4 state (selenite) is thermodynamically stable at normal environmental pH values and redox potential conditions whereas for sulphur it is the +6 state (sulphate). The form used in biological macromolecules (proteins, tRNAs), however, is the -2 oxidation state, mainly because selenols are more powerful nucleophiles than thiols, especially under physiological pH conditions. Nature has used this chemical property of the selenol group to incorporate it into the active site of oxidoreductases. In this review, we deal with microbial enzymes which are dependent in their catalysis on selenium and with the unique manner of incorporation of this element into proteins and tRNAs. Recent reviews o n this topic have been written by Stadtman (1990, 1991), Sunde (1990), Burk (1991) and Bock et al. (1991a,b). 11. Selenium-Containing Enzymes

Selenium is a constituent of several enzymes from prokaryotic and eukaryotic organisms. Depending on the particular enzyme, it is present either in the form of the amino-acid residue selenocysteine, specifically incorporated into the polypeptide backbone (Cone et a l . , 1976; Gunzler et al., 1984; Stadtman et a l . , 1991), or it appears in a form which can readily be detached by treatment with denaturing and/or reducing agents (Dilworth, 1983; Durre and Andreesen, 1983; Boursier el a f . , 1988). To date, several classes of selenocysteine-containing proteins (termed

9 - 1 I-NIUM METAROI ISM IN MICRO-ORGANISMS

73

selenoproteins in this review) are known. Selenoproteins from prokaryotes are represented by three groups of enzymes. These are glycine reductases from amino acid- and purine-fermenting bacteria, several FDH isoenzymes and some hydrogenases. Three different classes of selenoproteins of eukaryotic origin are known at present, namely gluthathione peroxidase isoenzymes (Flohe, 1989; Schuckelt et al., 1991), thyroxine de-iodinase (Berry et al., 1991a; Behne et al., 1990), and plasma protein P (Burk, 1991). At present, only vertebrate species are known to synthesize selenoproteins, but detection of selenocysteyl-tRNA in lower eukaryotes (Lee el al., 1990; Hatfield et al., 1991, 1992) as well as in vertebrates (Lee et al., 1989) provides compelling evidence for the presence of selenoproteins in many eukaryotic species. Synthesis of all these selenoproteins is wholly dependent o n the availability of selenium and on a functional biochemical pathway for biosynthesis and cotranslational incorporation of selenocysteine. On the other hand, selenium-containing enzymes in the second category do not contain seleno-amino acids in their peptide backbone. So far, several xanthine and nicotinic-acid dehydrogenases from anaerobic bacteria (Diirre and Andreesen, 1983; Dilworth, 1983), carbon-monoxide dehydrogenases from aerobic carboxydotrophic bacteria (Meyer et al., 1986) and a hydrogenase from Bradyrhizobium japonicum (Boursier et al., 1988) have been reported to contain selenium. Selenium availability usually exerts a stimulating effect on the activity of these enzymes. Depending on the particular enzyme, the demand for selenium may be more or less pronounced. A . BIOCHEMISTRY OF PROKARYOTIC SELENOPROTEINS

1 . Glycine Reductase

Glycine reductase is present in several anaerobic bacteria capable of fermenting amino acids through a Stickland-type reaction or of fermenting purines. Well-characterized “Stickland bacteria” include Clostridium sticklandii, C . litorale, C . histolyticum, C . sporogenes and Eubacterium acidaminophilum. Purine-fermenting bacteria containing glycine reductase are represented by Clostridium purinolyticum, C . acidiurici and C . cylindrosporurn (Durre and Andreesen, 1983, 1986; Dietrichs et al., 1991). Glycine is used by these bacteria for disposal of redox equivalents which are generated during fermentative degradation either of other amino acids or of purines (Gottschalk, 1986). The metabolic balance for a Stickland reaction involving the amino-acid pair alanine-glycine is shown in equation ( I ) , and t h e balance for degradation of the formiminoglycine intermediate in purine fermentation is given in equation (2).

74

J

IIFIIX-R AND A BOCK

+ 2glycine + 3ADP + 3P, -+ formiminoglycine + 2ADP + 2 P,

alanine

-P

+ C 0 2 + 3NH4+ + 3ATP ( I ) acetate + C 0 2 + 2NH4+ + 2ATP (2)

3acetate

Exergonic reductive cleavage of glycine into ammonia and acetate (AGO' = - 14.4 kcal mol-' with NADH as the reductant) is coupled to ATP synthesis in both fermentations (equation (3)). glycine

+ P,*- +ADP2- + 2e- + 3H+-+acetate- + NH4+ + ATP* + H 2 0

(3)

Growth of these bacteria on amino acids or purines is dependent on the availability of selenium in the medium (Diirre and Andreesen, 1983,1986). The only exception is growth at the expense of proline as an electron acceptor (e.g. by C . sticklandii) since the proline reductase does not contain selenocysteine and it can substitute for glycine reductase in removal of surplus redox equivalents (Gottschalk, 1986; Buckel, 1990). However, reduction of proline to 5-aminovalerate does not generate enough free energy for ATP synthesis. Among the species known to contain glycine reductase, Eu. acidaminophilum is clearly among the metabolically most versatile. Whereas the other species are restricted to glycine as the electron acceptor, Eu. acidaminophilum can also channel sarcosine and betaine into the glycine reductase pathway, and can even simultaneously utilize glycine as an electron donor and acceptor (Zindel el a f . , 1988; Hormann and Andreesen, 1989). Glycine reductase is a membrane-associated (Freudenberg et al., 1989) complex enzyme comprising three separate proteins, namely protein A (PA), protein B (PR)and protein C (Pc) (Turner and Stadtrnan, 1973; Tanaka and Stadtman, 1979). The selenoprotein PA and protein Pc represent the core enzyme. Selenoprotein PA is a small, acidic glycoprotein migrating at 12.5-18.0 kDa on protein gels (Turner and Stadtman, 1973; Dietrichs et al., 1991). It contains a selenocysteine residue coded by U (Cone et al., 1976) in a CxxU motif (Sliwkowski and Stadtrnan, 1988; Garcia and Stadtman, 1991; Dietrichs et al., 1991) identical to that of the cysteyl residues in thioredoxin (Gleason and Holmgren, 1988). Protein Pc is a large complex comprising two dissimilar subunits of 57 and 48 kDa in an a4P4arrangement. It has been shown recently that the 48 kDa subunit contains an acetyl-accepting active thiol group, whereas the hydrophobic 57 kDa subunit is assumed to be involved in formation of a multiprotein complex at the membrane and in generating an anhydrous reaction compartment (Schrader and Andreesen, 1992). Protein PH is probably required for initial activation of the substrate (Arkowitz and Abeles, 1989). In Eu. acidaminophilum, synthesis of three different PH proteins, which are specific for activation of either glycine, sarcosine or betaine, is induced

75

SEIENIUM METABOLISM IN MICRO-ORGANISMS

NADPH

Glycine

NADP+

Sarcosine

,coo-

,COO' . FH2 H3Nt

CH2 I H2N+ \CH3

1

1

,CW /cooCH2

Betaine

,cooCH2

I

(CH3)3N+

1 /coo-

I

v 2 H2r
Glycine

+

Sarcosin

(Pg)

FIG. 1. Mechanism of the glycine reductase reaction of Eubacterium acidaminophilum, as proposed by Schrader and Andreesen (1992). The proteins constituting the glycine reductase complex are indicated as PA, PB and Pc, respectively. TR indicates thioredoxin reductase; TRX, thioredoxin. The carboxymethyl groups of glycine, sarcosine and betaine which are transferred to PA are indicated in bold. R is either hydrogen or a methyl group attached to the released ammonium nitrogen.

76

J tltll)FK

A N D A HOCK

in response to the available substrate (Schrader and Andreesen, 1992; Fig. 1). So far, an activation pathway by P13has been proposed only for glycine. A covalently bound pyruvyl residue at P,,-glycine (Tanaka and Stadtman. 1979) may be involved through formation of a Schiff base (Arkowitz and Abeles, 1989). At the moment, no proposal exists for the mechanism of sarcosine and betaine activation by their specific PR proteins in Eu. acidaminophilum (Schrader and Andreesen, 1992). The activated u-position can subsequently be attacked by the selenocysteyl residue of PA to yield a selenium carboxymethylated derivative (Arkowitz and Abeles, 1990; Fig. 1 ) . Two possible pathways have been postulated for the next step on the pathway, namely translocation of the acetyl group from selenoprotein PA to the acetyl-accepting 48 kDa subunit of Pc. Arkowitz and Abeles (1991) suggested the formation of an acetyl thioester with Pc prior to release of the acetyl moiety from PA, whereas Buckel (1990) and Schriider and Andreesen (1992) propose release of a ketene intermediate from PA by formation of a selenodisulphide between the selenocysteyl and the neighbouring cysteyl residue as a first step (Fig. 1). The highly reactive ketene intermediate would readily react with the acetyl-accepting thiol group of the 48 kDa subunit of Pc and could be stabilized by water exclusion from the reaction compartment within the hydrophobic 57 kDa subunit of Pc (Schrader and Andreesen, 1992). The resulting enzyme-bound acetyl thioester is then cleaved phosphoroclastically and acetyl phosphate released (Fig. 1). It can be used for substrate-level phosphorylation by acetate kinase. The oxidized selenoprotein PA,generated in the reaction, is reduced by a redox chain involving NADPH, thioredoxin reductase and thioredoxin (Dietrichsetal., 1991; Schrader and Andreesen, 1992; Fig. 1 ) . Additionally, a lipoamide dehydrogenase side-reaction of thioredoxin reductase of Eu. acidaminophilurn was detected which may be used for direct redox-carrier regeneration (Dietrichs el al., 1991). 2. Formate Dehydrogenases

The occurrence of selenocysteine-containingFDHs has been demonstrated for several species of enterobacteria (Enoch and Lester, 1975; Cox et ul., 1981; Kramer and Ames, 1988; Heider et al., 1991), for Methanococcus vunnielii (Jones and Stadtman, 1981) and for a number of Clostridium spp. (Andreesen and Ljungdahl, 1973; Leonhardt and Andreesen, 1977; Wagner and Andreesen, 1977). All of these FDHs contain a molybdenum (or tungsten) cofactor and different amounts of iron-sulphur clusters. Analogous cysteine-containing enzymes have also been described for some other bacterial species (Schauer and Ferry, 1982; Kr6ger et ul., 1979).

TABLE 1.

Enzyme parameters of molybdenum cofactor-dependent formate dehydrogenases of different organisms. Values for k,, are given to the molvbdenum centres of the respective enzymes

Formate dehydrogenase

K , (formate) (mM) k,,, (min-')

Escherichia coli FDH, Wolinella succinogenes FDH Escherichia coli FDHH (Se) Escherichia coli FDHH (S) Methanobacterium formicicum FDH

0.12 1.5 26 9 Not known

170,ooO

Clostridium thermaceticum FDH Clostridium acidiurici FDH Clostridium pasteurianum FDH

0.23 3.1 Not known

Not known Not known 48,000

33,800 20.000 540 3.100

'

k,,,lK,,, ( m M ~ min-')

2.8.1 O5 1.3.10' 6.5.10' 6.0.10'

Reference Enoch and Lester (1975) Kroger ef al. (1979) Axley et al. (1990) Axley er al. (1991) Schauer and Ferry (1982, 1986) Yamamoto et al. (1983) Keamy and Sagers (1972) Scherer and Thauer (1978)

78

J IIEIDKR A N D A UOCK

Formate dehydogenases are involved in physiologically very diverse metabolic pathways; Table 1 gives a comparison of these enzymes. One class of FDH isoenzymes catalyses the first step in production of carbon dioxide and hydrogen from formate. In enterobacteria, the FDHH isoenzyme is part of the formate hydrogenlyase complex (Cox et al., 1981 ; Pecher et al., 1985). The FDHH isoenzyme of E. coli is encoded by the fdhF gene at 92.4' on the chromosome (Pecher et al., 1985) and consists of an 80 kDa selenopolypeptide which exhibits formate-benzylviologen oxidoreductase activity (Axley el al., 1990). Together with gene products of the hyc operon which encodes hydrogenase 3 and potential intermediate electron carriers (Bohm el al., 1990), FDHH forms the formate hydrogenlyase complex (Fig. 2). I n many methanogenic archaea, F42,,-reducing FDHs are required to generate the methanogenic substrates carbon dioxide and reduced Fjzo from formate (DiMarco et al., 1990). Methanococcus vannielii has been shown to contain two different forms of FDH, o n l y one of which contains selenium (Jones and Stadtman, 1981) whereas Methanococcus forrnicicurn synthesizes a selenium-independent FDH containing the molybdenum cofactor, iron-sulphur centres and FAD (Schauer and Ferry, 1983,1986). This FDH which is encoded by thefdhCAR operon (Shuber et al., 1986; Patel and Ferry, 1988) exhibits an a& structure of 177 kDa (Schauer and Ferry, 1986). The u-subunit is homologous to the fdhFgene product of E. coli (Zinoni et a l . , 1987) and the P-subunit contains iron-sulphur centres and exhibits homology with ferredoxin-like proteins encoded within the hyc operon of E. coli (Biihm et al., 1990). The FAD cofactor of FDH from M . forrniciurn is required for reduction of the physiological substrate F420(Schauer and Ferry, 1983; 1986). Both FDHH from E. coli and FDH from M . forrnicicum are associated with the membrane on the cytoplasmic side (Baron etal., 1989; Bohm, 1991; Fig. 2). A second class of FDH isoenzymes initiates respiration of formate by releasing carbon dioxide and channelling electrons into aerobic or anaerobic respiratory chains. This pathway has been characterized to date in enterobacteria (Pinsent, 1954; Enoch and Lester, 1975; Sawers el uf., 1991) and Wolinella succinogenes (Kriiger et al., 1990). In E. coli, two FDH isoenzymes are known to participate in formate respiration. One, namely FDHN encoded by thefdnGHIoperon at 32' (Bergetal., 199 l a ) , is required for anaerobic formate-to-nitrate respiration. It is a membrane-bound protein complex exhibiting an a4P4y4 structure of 600 kDa (Enoch and Lester, 1975). The 110 kDa a-subunit contains a selenocysteine residue. the M o cofactor and an iron-sulphur centre, the 30 kDa P-subunit is H ferredoxin-type electron-transfer protein, while the 20 kDa y-subunit is a membrane-internal cytochrome h,,, (Enoch and Lester, 1975; Berg ct 01.. 1991a). The K,, value for formate of FDHN is several orders of magnitude

FIG. 2. Function and localization of the formate dehydrogenase isoenzymes of Escherichia coli. Mo and -Se- designate, respectively, the molybdenum cofactor and the selenocysteine moiety present in the large subunit; Fe/S, iron-sulphur centres which transfer electrons t o acceptor sites; Ni, the active site of hydrogenase 3; 0.ubiquinone o r menaquinone. The existence of formate transporter protein(s) in the cytoplasmic membrane is purely speculative.

80

J IlI-ll)l-H ANI) A

HOCK

lower than that of FDHH (Table 1). The other isoenzyme, FDHo, is synthesized aerobically, and also anaerobically in the presence of nitrate, and it contains a 110 kDa selenocysteine residue-containing subunit (Sawers et al., 1991). The enzyme FDHo may also be involved in formate-to-nitrite respiration. Its genes are probably located at 88' between the fdhD and fdhE genes (Schlindwein et a l . , 1990) and its subunit structure appears to be very similar to that of FDHN. Both respiratory FDHs appear to feed electrons from formate into the quinone pool (Fig. 2; Enoch and Lester, 1975). An FDH homologous to FDHN, but lacking selenium, has been characterized from W. succinogenes (Kroger et al., 1979); it is required for growth of the bacterium through formate-to-fumarate or formate-tosulphur respiration (Kroger et al., 1990). Two identical but differently regulated operons code for this FDH in W. succinogenes (Bokranz ef NI., 1991; Kortner and Kroger, 1992). The enzyme has been shown to be exposed towards the periplasm (Kroger et al., 1990) and its large subunit contains a signal sequence typical for periplasmic proteins (Bokranz et ul., 1991). A very similar signal sequence was also detected for the 110 kDa selenopolypeptide of FDHN from E. coli (Berg et al., 199la), which may justify the notion that the membrane-bound respiratory FDHs of E. coli are located on the periplasmic side of the membrane, as shown in Fig. 2. Additional evidence for this assumption comes from the results of an analysis of the stoichiometry of proton translocation by the FDHN reaction which agrees with a periplasmic localization for the active site of FDHN (Jones, 1980). A third group of isoenzymes comprises the anabolic FDHs catalysing bidirectional interconversion of carbon dioxide and formate; they are constituents of the acetogenic pathway of Clostridium therrnaceticuni and C . formicaceticum (Leonhardt and Andreesen, 1977; Ljungdahl and Andreesen, 1978; Yamamoto et al., 1982). In C . thermaceticum, NADPH serves as the electron donor for reduction of carbon dioxide (Andreesen and Ljungdahl, 1973; Yamamotoetul., 1982). This FDH lacks molybdenum but contains equimolar amounts of tungsten, probably in the form o f a tungsten cofactor in the 96 kDa a-subunit. The FDH of C. formicaceticum also appears t o contain a tungsten cofactor and selenium. The electron donor for reduction of carbon dioxide in this species is unknown (Leonhardt and Andreesen, 1977). Bidirectional FDHs are also involved in purine fermentation by purinolytic clostridia. They serve either as redox equivalent-scavenging systems, reducing carbon dioxide to formate (Wagner and Andreesen, 1977), or they utilize formate derived from the intermediate formiminoglycine as an electron donor for glycine reductase (Vogels and Van der Drift, 1976; Gottschalk, 1986). To date, FDHs of purinolytic clostridia are poorly

SEIXNIUM METABOLISM IN MICRO-ORGANISMS

81

characterized. For C. acidiurici and C . cylindrosporum, stimulation of FDH activities by selenium supplementation has been reported. In addition, FDH from C. acidiurici seems to be dependent on tungstate, whereas the FDH from C. cylindrosporum requires molybdenum supplementation (Wagner and Andreesen, 1977). The first indication that selenium- and sulphur-containing isoforms of bidirectional FDHs may occur side by side came from the characterization of an anabolic FDH in Clostridium pastrurianum lacking selenium (Scherer and Thauer, 1978). The enzymes FDHN and FDHH from E . coli have been purified and subjected to detailed kinetic analysis (Enoch and Lester, 1975; Axley and Grahame, 1991). Formate-dependent reduction of the haem group in FDHN and coupling to quinone reduction were demonstrated in vitro (Enoch and Lester, 1975). In contrast, the reaction catalysed by FDH), does not involve two-electron carriers. In vitro, this isoenzyme exhibits ping-pong bi-bi kinetics for formate and the artificial one-electron acceptor benzylviologen (Axley and Grahame, 1991). Both isoenzymes are sensitive to oxygen, especially in the presence of formate, and they are inhibited by azide and cyanide (Enoch and Lester, 1975; Axley et al., 1990; Axley and Grahame, 1991). At pH values higher than 7, both enzymes are reported to become unstable (Enoch and Lester, 1975; Axley et al., 1990). Interestingly, the iodoacetamide-inhibition pattern of the two FDH isoenzymes differs significantly. The selenocysteyl residue in purified FDHH is not accessible to iodoacetamide, but it becomes sensitive when the enzyme is incubated with formate (Axley et ul., 1990; Axley and Grahame, 1991), whereas FDHN is partially inactivated in the absence of formate and incubation with formate appears to protect it from alkylation (Enoch and Lester, 1975). It is tempting to speculate that this difference is correlated with the observed channelling of electrons from formate to cytochrome bSshin FDHN (Enoch and Lester, 1975), which is not possible for FDHH. Thus, the redox-active centre of FDHN,probably the molybdenum cofactor (Fig. 3 ) , can reassemble to an oxidized state, protecting the selenol moiety against alkylation (Fig. 3 ) , whereas it remains reduced in FDHH,precluding covalent bonding of the selenol to, for example, the molybdenum cofactor (Fig. 3 ) . The availability of gene sequences for four molybdenum cofactordependent FDHs allows some conclusions to be made on the location of catalytically important centres. Besides the selenocysteine residue whose position is occupied by a cysteine residue in the FDHs lacking selenium, only four cysteine residues in the N-terminal domain are conserved in all four a-subunits (Fig. 4). Three of them constitute a CxxCxxxC motif which, together with a fourth conserved cysteine located not far away, probably correspond to the iron-sulphur centre. There is n o universally conserved

82

J HI-IDI-R AND A BOCK

S

\ VI/

S

r

-OH

II

Se'

/ c\

0

HC\

o-

A N o

H

H+

4

-OH

R=R S

FIG. 3 . Proposed reaction mechanism for molybdenum cofactor-containing formate dehydrogenases. Only the molybdenum co-ordination sphere of the molybdenum cofactor is shown. The hydroxyl group constituting a proposed fifth ligand to molybdenum may also be represented by a basic residue of the protein (e.g. histidine) which may help in proton abstraction from formate.

SELENIUM METABOLISM IN MICRO-ORGANISMS

83

cysteine residue in the vicinity of the selenocysteine-cysteine residue (Fig. 4). This would preclude a redox-cycling mechanism by formation of a selenodisulphide bridge as suggested previously (Stadtman, 1980). Therefore, it is reasonable to assume that the redox-active moiety in FDH is the molybdenum atom, as it is so in all other molybdenum cofactorcontaining oxidoreductases (Kaim and Schwederski, 1991). The reaction catalysed by FDH appears quite similar to that of xanthine dehydrogenase (see Section 1I.B). A X=CH-X-moiety is formally oxidized to X=COH-X. So, by analogy to the situation found in xanthine dehydrogenases with a free sulphide (or selenide) group co-ordinated to molybdenum (Fig. 5; Gardlik and Rajagopalan, 1990), the strongly nucleophilic selenium atom in selenocysteine may be directly co-ordinated to the molybdenum atom (see Fig. 3). The positive charge on molybdenum atom resulting from this model may be buffered by co-ordination of a fifth ligand providing a free electron pair (symbolized by the hydroxyl group in Fig. 3). This may explain why FDH is only active at slightly alkaline pH values, but stable at acidic pH values (Enoch and Lester, 1975; Axley et al., 1990). These considerations suggest WS EC EC MF

FDH FDHN FDHH FDH

WS EC EC MF

FDH FDHN FDHH FDH

111 0 0

n

54 SKKVKTI C TY C SGV c GIIAEVVDG------- VWVRQEVAQDHPISQGGH C CKG 43 AKEIRNT C TY C SVG c GLLMYSLGGDAKNAREAIYHIEGDPDHPVSRGAL c PKG 1 MKKVVTV C PY C ASG C KINLVVDNG-------KIVRAEA-AQGKTNQGTL C LKG 3 IKYVPTICPYCGVGCGMNLVVKDE-------KVVGVEPWKRHPVNEGKLCPKG

175 185 127 121

GTNNLDTIARIC HAPTVAGVSNTLGYGGMTNHLAD GMLAVDNQARVUHGPTVASLAPTFGRGAMTNHWVD GTNNVDCCARVU HGPSVAGLHQSVGNGAMSNAINE GTHNIDHCARLC HGPTVAGLAASFGSGAMTNSYAS

FIG. 4. Alignment of sequences of the large subunits of molybdenum cofactordependent formate dehydrogenases. The universally conserved cysteine residues (respectively, selenocysteine residues) are boxed. WS indicates Wolinella succinogenes; EC, Escherichia coli; MF, Methanobacterium formicum. Nitrate reductase

Xanthine dehydrogenases Eukaryotes

Anaerobic bacteria

FIG. 5. Co-ordination of molybdenum cofactors by oxide, sulphide or selenide ligands in different oxidoreductases.

84

J l l h l l ) F H A N D A BOCK

the reaction mechanism illustrated in Fig. 3. Binding of formate to the molybdenum atom activated by the selenol of the selenocysteine residue would displace the selenocysteyl residue which could then initiate formate oxidation by a nucleophilic attack on the molybdenum-bound formyl group. After carboxyl transfer to the selenocysteine residue, Mo4+ is generated by reduction (see Fig. 3). This corresponds well with the usual Moh+-+Mo4+ shift observed following reduction of molybdenum-cofactor oxidoreductases (Kaim and Schwederski, 1991). The electrons can then be transferred one by one to the iron-sulphur centres through an Mo” intermediate. An Moh+-+Mo4+-+Mo5++Moh+ redox cycle is characteristic of molybdenum cofactor-dependent oxidoreductases and is in accordance with the kinetic scheme proposed by Axley and Grahame (1991). 3, Hydrogenuses Selenocysteine residue-containing nickel-iron-selenium h ydrogenases have been detected until now in methanogenic archaea, namely in M. vannielii (Yamazaki, 1982), M . voltae (Muth et al., 1987) and in some sulphatereducing bacteria, namely Desulfomicrobium (formerly Desulfovibrio) baculatum (Teixeira et al., 1987), Desulfovibrio vulgaris and D . salexigens (Fauque ef al., 1988). The purified F420-reducingenzyme from M . voltae is composed of three subunits (Muth et ul., 1987) whereas other nickeliron-selenium hydrogenases that have been characterized exhibit the classic up structure o f hydrogenases with the larger u-subunit of 49-62 kDa carrying the selenocysteine residue in a 1 :I molar ratio with a nickel atom (Teixeira ef al., 1987). Different numbers of iron-sulphur centres are present in both subunits (Teixeira et a l . , 1987). The hydrogenases of Dm. baculatum have been analysed most thoroughly, and existing data indicate that up to three nickel-iron-selenium isoenzymes may be synthesized by this organism (leixeira et a/., 1987). Genes coding for one of them have been cloned and sequenced (Menon efal., 1987,1988). Recently, the sequences of the genes for the nickel-iron-selenium hydrogenase isoenzymes of the archaeon M . volfae have been published (Halboth and Klein, 1992). Genetic analysis revealed that this organism contains two gene clusters for different nickel-iron-selenium hydrogenases, one corresponding to the purified F420-reducingenzyme (Muth e f al., 1987) and the other one exhibiting homology to nickel-iron hydrogenases with unknown physiological electron acceptors. In addition, two gene clusters for nickel-iron isoenzymes have been detected in M. voltue which exhibit very extensive sequence similarity with isoenzymes containing selenocysteine residues (Halboth and Klein, 1992). From the gene sequences, a sclenocysteyl residue is predicted to be

SELENllJM METABOLISM IN MICRO-ORGANISMS

85

present in large subunits of nickel-iron-selenium hydrogenases at a position corresponding to a universally conserved cysteine residue in nickel-iron hydrogenases (Voordouw et a f . ,1989; Halboth and Klein, 1992). Moreover, it was demonstrated by electron paramagnetic-resonance (EPR) spectroscopy of 77Se-enrichedenzyme that the selenium atom in the nickel-ironselenium hydrogenase of D m . bacufatum is directly co-ordinated to the active-site nickel atom (Eidsness et a f . , 1989; He et a f . , 1989), thus supporting the notion that the active site of nickel-containing hydrogenases consists of a mixed nickelliron-sulphur cluster (Albracht, 1990; Fauque et a f . , 1988) and that at least one of the ligands of the nickel atom is a cysteyl or selenocysteyl residue in the large subunit. Comparative studies of the nickel-iron-selenium hydrogenase of Dm. bacufatum with the nickel-iron isoenzyme of Desuffovibrio gigas have indicated some possible physiological or biochemical functions of the selenocysteyl residue. Direct experimental support of them, however, is still missing. First, the selenocysteine residue-containing hydrogenase of Dm. bacufatum appears to be more efficient in hydrogen evolution in vitro, when compared with the nickel-iron hydrogenase of D . gigas. Therefore, nickel-iron-selenium hydrogenases are considered to possess a function in evolution rather than uptake of hydrogen (Fauque et a f . , 1988). Secondly, in contrast to nickel-iron hydrogenases, which need to be activated by hydrogen or carbon monoxide to attain a ready conformation of their nickel centres for reductive activation (Albracht, 1990), the active-site nickel atom of the nickel-iron-selenium hydrogenase of Dm. bacufatum displays an EPR-silent state (‘‘N?’ ready”) after isolation of the enzyme. This allows reductive activation more readily (Teixeira e t a f . ,1987; Fauque et a f . ,1988). Lastly, in contrast to the nickel-iron hydrogenase of D . gigas, the nickeliron-selenium hydrogenase of Dm. bacufatum possesses an increased ratio of double-exchange activity of both protons of a hydrogen molecule bound to the active-site nickel atom with deuterium from deuterium oxide. This has led to the assumption that the presence of the less electronegative selenium atom (compared to a cysteyl-sulphur group) co-ordinated to the active-site nickel atom decreases the stability of a nickel-bound hydride intermediate (Teixeira et a l . , 1987), and increases the probability of a double proton exchange (Fauque et al., 1988). At present, no bacterial species is known which synthesizes seleniumcontaining hydrogenases exclusively. The most complex isoenzyme pattern is found in D . vufgaris,which contains a periplasmic iron-only hydrogenase and membrane-bound nickel-iron (periplasmic side) and nickel-ironselenium (cytoplasmic side) hydrogenase isoenzymes at the same time (Fauque et al., 1988; Fig. 6). The subtle differences observed between the nickel-iron-selenium hydrogenase of Dm. bacufatum and the

86

J IiFIDl-R A N D A NOCK

FIG. 6. Proposed differential functions of the iron, nickel-iron and nickel-ironselenium hydrogenase isoenzymcs of Desulfovihrio vulgaris in transmemhranehydrogen cycling and sulphate reduction.

nickel-iron hydrogenase of D . gigas elicited the proposal that the different hydrogenases may serve different physiological functions in a proposed hydrogen-cycling metabolism feeding into sulphate reduction as depicted in Fig. 6 for D. vulgaris (Fauque et a l . , 1988). R. SELENIUM-DEPENDENT ENZYMES NOT CONTAINING SELENOCYSTEINE RESIDUES

1. Xanthine Dehydrogenases and Nicotinic-Acid Dehydrogenases

Xanthine dehydrogenases are involved in degradation of purine bases and have been characterized from a variety of prokaryotic and eukaryotic organisms (Vogels and Van der Drift, 1976).They are complex molybdenumironsulphur flavoproteins (Konig and Andreesen, 1992) and it is assumed that, in contrast to other molybdenum cofactor-dependent enzymes such as nitrate reductase, the cofactor in these enzymes is co-ordinated to a sulphide ligand and not to oxygen (see Fig. 5 ; Gardlik and Rajagopalan, 1990). In purine fermentation by anaerobic bacteria, xanthine serves as a starting substance for hydrolytic cleavage of the pyrimidine ring to give rise to 4-ureido-5-imidazol-carboxylate,which is further degraded lo formiminoglycine following deamination and decarboxylation reactions (Gottschalk, 1986). Interestingly, in some of these bacteria, namely C . acidiurici and C. cylindrosporum, xanthine dehydrogenase has been

SFLENIUM METABOI.ISM IN MICRO-ORGANISMS

87

demonstrated to be dependent on selenium supplementation during bacterial growth (Wagner and Andreesen, 1977, 1979). A similar stimulation of enzyme activity by addition of selenium to the medium was reported for the nicotinic-acid dehydrogenase from Clostridiurn barkeri (Imhoff and Andreesen, 1979), which catalyses a reaction very similar to that of xanthine dehydrogenase. Selenium is required by C. barkeri and Desulfobacterium niacini (Widdel and Hansen, 1991) for growth in the presence of nicotinic acid. Nicotinic-acid dehydrogenase, like xanthine dehydrogenase, is a molybdenum-iron-sulphur flavoprotein (Holcenberg and Stadtman, 1969; Dilworth, 1983). Selenium co-purifies with this enzyme but is not incorporated as a selenocysteine residue since autoradiographic detection of "Se-labelled enzyme was only possible on native gels and the label was completely lost upon denaturation (Dilworth, 1982, 1983). Moreover, the selenium moiety in these enzymes is released by heat or by chaotropic agents and is sensitive to cyanolysis (Dilworth, 1982; Diirre and Andreesen, 1983). Alkylation resulted in release of dialkyl selenide, which indicates that selenium is present as selenide (Dilworth, 1982). Co. migration of selenium and molybdenum upon denaturation of the protein in a low molecular-weight fraction of a gel-filtration column led to the assumption that selenium constitutes one of the ligands for the molybdenum cofactor in these enzymes (Dilworth, 1983; see Fig. 5 ) . A reaction mechanism has been postulated for xanthine dehydrogenase which involves initial binding of the substrate to the Mob+ centre through one of the ring nitrogen atoms and subsequent hydroxylation, leaving reduced Mo4+in the enzyme. This is then regenerated in two steps through Mo" by one-by-one transfer of electrons to the iron-sulphur centre (Kaim and Schwederski, 1991). Formally, the reactions catalysed by these heterocycle-hydroxylating enzymes are very similar to those involved in oxidation of formate by FDH. 2. Possible Involvement of Selenium in Other Metabolic Processes

Boursier et al. (1988) reported that a hydrogenase from B. japonicum appears to be activated when selenium is present in the medium. The enzyme co-purifies with selenium but the element is not attached covalently since it is lost completely upon denaturation (Boursier et al., 1988). It may well be that free selenide is incorporated into mixed (nickel)-iron-sulphur clusters that act in a similar way to the selenium in selenocysteine residues in co-ordination of nickel in the nickel-iron-selenium hydrogenase of Dm. baculutum. Likewise, Meyer and Rajagopalan (1981) reported in vitro formation of a covalent, selenium-containing adduct of carbon-monoxide

88

J l1FIIIER A N D A ROCK

dehydrogenase from Pseudornonas carboxydoflava when the enzyme was incubated in the presence of selenite. Also, "Se was incorporated into the enzyme in vivo when the bacterium was grown on [7sSe]selenitesupplemented medium (Meyer et af., 1986). Another example is selenium dependence of the ability of Desuffococcus rnultivorans to degrade benzoate (Widdel and Hansen, 1991). The biochemical basis of at least some of these phenomena may lie in non-specific insertion of selenium into iron-sulphur clusters, thereby replacing sulphur. Moulis et af. (1988), for example, have shown that this incorporation can change the properties of an enzyme in a subtle way.

111. Selenium-ContainingtRNAs

Selenium-containing tRNA species have been identified in several bacterial and archaeal organisms (Chen and Stadtman, 1980; Ching and Stadtman, 1982; Wittwer, 1983; Ching et af., 1984; Kramer and Ames, 1988; Politino et af., 1990; Heider et af., 1991) and also in mammals (Ching, 1984). In these tRNAs, a uridine residue at position 34 (the first anticodon nucleotide) is modified by selenium (Ching el af., 1985). The modified base is a 2-seleno-5-methylaminomethyl-uridineresidue (~e*-mam~-U'~) (Fig. 7) (Wittwer, 1983; Ching et af., 1984, 1985). Cfostridiurn sticklandii possesses only the selenium-modified isoform of tRNAG'" and this modification has been implicated as a possible determinant for recognition by glutamyl-tRNA synthetase (Ching and Stadtman, 1982). In contrast, about 50% of the tRNAG"' and tRNALyScomplement o f E. cofi contains the sulphur analogue ~ ~ - m a m ~(Sprinzl - - U ~ ~et af., 1991) instead of the selenium-containing residue (Wittwer and Ching, 1989). The residue ~ ~ - m a r n ' - Uis~ thought ~ to be involved in restricting "wobble"-base pairing of anticodons with glutamate (codons GAG/*) or se2-mam5-U

Phosphoribosilphosphate Phosphoridosylphosphate

FIG. 7. Two possible isomeric forms of the selenium-modified tRNA nucleotidc se2-mams-U.

SELENIUM METAROI.ISM IN MICRO-ORGANISMS

89

lysine (codons AA"/,) codons ending with G (Agris, 1991). Since E. coli is known to use glutamate and lysine codons ending with G at a high frequency (Wada et al., 1990), a possible function of the seleniumcontaining modification, at least in E. coli, could be to provide a set of tRNAs which are not restricted in recognition of "wobble"-position purine residues. This notion could indeed be proven by Wittwer and Ching (1989), who demonstrated that codon preference of purified selenium-containing tRNA"'" and tRNALdyS from E. coli is much less pronounced than that of the sulphur-containing isoforms. Similar results were also presented for Salmonella typhimurium by Kramer and Ames (1988).

IV. Biosynthesis of Selenoproteins and Seleno-tRNAs A . SELENOPKOTEINS

In all selenoproteins known to date, insertion of selenocysteine residues is directed by an in-frame TGA codon in the gene (Bock et al., 1991a,b). Colinearity of the TGA codon in the gene with selenocysteine residues in the polypeptide chain has been demonstrated for glutathione peroxidases (Giinzler et al., 1984; Chambers et al., 1986), for FDHHfrom E. coli (Zinoni et al., 1986; Stadtman et al., 1991) and for glycine reductase from clostridia (Sliwkowski and Stadtman, 1988; Dietrichs et al., 1991; Garcia and Stadtman, 1991). On the basis of these results, the view on the mechanism of incorporation of selenocysteine residues into proteins has shifted from that of a post-translational modification of precursor proteins (Stadtman, 1980) to a cotranslational mode of incorporation on the ribosome (Chambers et al., 1986; Zinoni et al., 1986). For FDHH from E. coli, translational incorporation could indeed be confirmed soon after the initial reports (Zinoni el al., 1987) and the proof that the translation system is involved in incorporation of this rare amino acid has immediately raised a number of interesting questions. First, a tRNA species must exist which is able to decode UGA codons. Secondly, a specific biosynthetic pathway should be present that provides selenocysteyl-charged UGA-decoding tRNA. Thirdly, UGA codons directing insertion of selenocysteine residues must be differentiated from UGA translational termination codons. Finally, termination of translation at the UGA codon determining selenocysteine incorporation must be prevented. The pathway of selenoprotein biosynthesis has been elucidated for FDHH from E. coli during the last few years, sometimes providing rather surprising information on solutions to the questions raised above (Bock et al., 1991a). It was accomplished by analysis of E. coli mutants pleiotropically deficient

90

J IIFII>l-H A N D A BOCK

N3) F; li; NI

N

wo -P-0

I

0-

HO

P-0

P-0

0-

0

l

l

OH

ATP

OH

I

3'

A C C

C

c

G 5'G-C

I=I I

tRNASe"

I

tRNASeC

A C C G 5'G C

1:1

I

-

I

tRNAS"

in FDHs. Four genes, termed selA, s e f B , selC and selD, were identified as being required for synthesis of selenoproteins (Leinfelder et al., 1988a).

I . The Gene for a UGA-Decoding tRNA: selC One of these genes, namely selC, was shown to code for an unusual form of tRNA (tRNAIJCA)that fulfils the requirements for decoding selenocysteine codons. It contains a UCA anticodon complementary to the UGA codon (Leinfelder et al., 1988b). I n addition, this tRNA molecule harbours

SF.I.ENIUM MFTABOLISM IN MICRO-ORCiANISMS

91

0

-1-0 b AMP

RSe--8-0

b-

ss

I c=o

\

HC-NHZ

ss

'0

I

C=O

I

,

tRNAS"

0 ' I

I

I

If I

Lao

'0

3'

A C C G 5G-C

i-CHZ

CH

A C C G 5'G -C .

.

II

I,

tRNA*

3'

A C C

G 5'G-C .

I

.

I

tRNASeC

FIG. 8. Diagram showing the mechanism of biosynthesis of selenocysteyl-tRNAS". SerS indicates seryl-tRNA synthetase; SS, selenocysteine synthase. Only a part of the aminoacyl acceptor stem of tRNASeCis shown. R represents either hydrogen or an as yet unknown carrier molecule which may bind to reduced selenium states.

primary sequence elements different from those of other elongator tRNAs (Leinfelder et al., 1988b; Schon et al., 1989; Baron et al., 1990); they are also present in the tRNAUCA sequence derived from the selC gene of Proteus vulgaris (Heider et al., 1989). The demonstration that tRNAucA carries selenocysteine residues in vivo proved its role as a selenocysteineinserting tRNA and prompted its designation as tRNASec (Leinfelder et

92

J HFIDI-H A N D A R ( X K

al., 1989). However, unlike any other tRNAs in E. coli, selenocysteine is

not charged directly to tRNASeC(Leinfelder e f al., 1988b), but it is formed from a serine residue attached by seryl-tRNA synthetase (Fig. 8) (Leinfelder et al., 1988b). 2. Biosynthesis of Selenocysteyl-tRNASe' from Seryl-tRNAS" These results and the properties of selC mutants implicated a role for tRNASec in biosynthesis of selenocysteine from serine (Leinfelder et al., 1988a,b). Indeed, it was shown that selenocysteyl-tRNASe' accumulated in vivo only when functional copies of selA and selD were present in the cell; in contrast, the selB gene was not required (Leinfelder et al., 1989). Subsequently, conversion of seryl-tRNASeCinto selenocysteyl-tRNASeCwas accomplished in v i m employing the purified proteins (SelA and SelD) encoded by the selA and selD genes (Leinfelder et al., 1990). The gene selA codes for the 50 kDa subunit of the 600 kDa enzyme selenocysteine synthase, which exhibits a decameric ring-shaped structure (Engelhardt et al., 1992). Each subunit contains a pyridoxal phosphate prosthetic group as part of the active centre (Forchharnrner et al., 1991a). The reaction (Fig. 8) catalysed by selenocysteine synthase involves specific binding of five seryl-tRNASeC molecules to the enzyme (Forchhammer et al., 1991b; Engelhardt el al., 1992). Binding involves specific recognition of tRNA determinants and covalent bonding of the a-amino group of the seryl residue to the carbonyl group of pyridoxal phosphate (Forchhammer et al., 1991a,b). The reaction proceeds through formation of an aminoacrylyltRNA intermediate by elimination of water from the activated seryl moiety and addition of an activated selenide species to yield selenocysteyl-tRNA, which is then released from the enzyme (Fig. 8; Forchhammer and Bock, 1991a). The protein SelD is only indirectly participating in this reaction; it synthesizes an activated low rnolecular-weight selenium donor from ATP and selenide (Forchhammer et al., 1991a). The protein has been characterized as being a selenium-dependent, AMP-releasing ATPase (Fig. 8; Ehrenreich et al., 1992; Veres et al., 1992). Interestingly, both of the anhydride bonds of ATP are cleaved by SelD, thus liberating the pphosphate as orthophosphate (Ehrenreich et al., 1992) and channelling the y-phosphate into the selenium donor (Ehrenreich et al., 1992; Fig. 8). The reaction product has been shown to contain a phosphorusselenium bond (Veres et al., 1992) and it is assumed to be a phosphoroselenoate (Ehrenreich et al., 1992; Veres et al., 1992). It is not known at present whether the highly reactive chemical species selenide and phosphoroselenoate occur in the free state in the cell or whether they are bound to

SELENIUM METABOLISM IN MICRO-ORGANISMS

93

redox-active substrates such as thioredoxin. Like phosphorothioate (Neumann et al., 1967), phosphoroselenoate should react readily with disulphide bonds. 3. A Unique Elongation Factor for Selenocysteyl-tRNASe': SelB

Because of its unique structural properties, tRNASe' does not bind to the elongation factor EF-Tu (Forster et al., 1990). Rather, the role of EF-Tu is taken over by an alternative elongation factor, the product of the selB gene (Forchhammer et al., 1989). The elongation factor is a 70 kDa protein displaying homology to translation factors EF-Tu and IF-2 in its N-terminal half; it binds GTP and selenocysteyl-tRNASe' stoichiometrically. It also discriminates tRNASecfrom other tRNA species by its 8 bp long aminoacyl acceptor stem (Baron and Bock, 1991). Furthermore, SelB recognizes the aminoacyl residue in tRNASeC,since neither seryl-tRNASeC(Forchhammer et al., 1989) nor alanyl-tRNASe' (Forchhammer et al., 1991b) is bound. The amino-acid specificity of the elongation-factor SelB provides an explanation for the exclusive incorporation of selenocysteine into proteins, even in the presence of the seryl-tRNASeCprecursor form. A similar situation exists in chloroplasts where glutamine is synthesized from glutamate in the tRNA-bound state (Schon et al., 1988). In this system, the plastidal EF-Tu was found to discriminate the precursor Glu-tRNAG'" from the end-product Gln-tRNA';'" (Stanzel et al., 1991). 4. Discrimination of Selenocysteine Codons from Stop Codons

One of the most intriguing questions in the field of selenium molecular biology is how the specific UGA codon in selenoprotein mRNAs is discriminated from a UGA translational stop codon. In a search for such determinants, it was found that an RNA secondary structure in the mRNAs for selenopolypeptides of FDHH and FDHN of E. coli immediately downstream of the UGA codon was indispensable for selenocysteine incorporation (Zinoni et al., 1990; Berg et al., 1991b). More specifically, the nucleotide sequence of the loop portion turned out to serve as a specific recognition element in this hairpin structure (Heider el al., 1992). It was shown that incorporation of selenocysteine into non-selenoproteins could be achieved when an mRNA structure similar to that in fdhFwas introduced (Heider and Bock, 1992). Thus, a recognition factor is probably involved in directing the ternary complex between SelB, GTP and selenocysteyltRNASecto the ribosomal A-site to decode selenocysteine codons. A likely candidate for this factor is the elongation factor SelB, because compared to EF-Tu an additional C-terminal domain lacking homology is present in

94

J llhll>i-,R A N D A HOC'K

the SelB protein, indicating an additional function (Forchhammer et a l . , 1989). Interestingly, a second site mutation in the 3' half of the se1R gene can revert a mutation in the recognition sequence of fdhF mRNA (A. Herzog, C. Baron and A. Bock, unpublished results). This supports the notion that Sell3 has a direct function in specific recognition of the mRNA context of selenocysteine codons (Heider et a l . , 1992).

5. Evolution of Selenocysteine Incorporation into Proteins Elucidation of the pathway for selenocysteine biosynthesis and its incorporation into proteins raises the question as to whether this system is phylogenetically old, a relic of evolution, or whether it represents a novel addition to the machinery of protein synthesis. There are a number of arguments in favour of either assumption. Evidence suggesting that selenocysteine may have entered protein synthesis early in evolution are its presence in proteins from all three lines of descent (Stadtman, 1990; Bock et al., 1991a) and similarities in the mode of its biosynthesis and incorporation within widely separated organisms. I t is now established that UGA codons are used in all three domains of life, namely Bacteria, Archaea and Eucarya, to code for selenocysteine Bock et al., 1991b; Halboth and Klein, 1992). Additionally, tRNA"" from eukaryotes (Hatfield et al., 1990) shares some important structural similarities with that from bacteria (Baron e t a l . , l992), such as the extended aminoacyl acceptor stem (Bock et a l . , 1991b) and aminoacylation with serine (Hatfield el al., 1990), which is subsequently converted to selenocysteine (Lee el u l . , 1989) through a pathway which probably proceeds in a similar way to that established in E . coli (Mizutani et al., 1991). The similarity of selenocysteine biosynthesis in organisms from different phylogenetic lines argues in favour of the existence of this pathway before separation of the organisms. This notion was confirmed by analysing the phylogenetic position of the elongation factor SelB from E. coli in relation to other translation factors (Bock et al., 1991b). The protein SelB branched off after EF-G and IF-2 had separated but before the branching point of elongation factors between archaea and bacteria, thus suggesting that it was already present in early forms of life (Bock et al., 1991b). Because of the scarcity of selenium in the biosphere, selenium-containing biomolecules were probably never synthesized abundantly, and development of a highly specific pathway for selenium incorporation may have been a prerequisite for use of this element in biomolecules (Forchhammer and Bock, 1991b). A fact which needs explanation in this connection is whether specific recognition of selenocysteine codons seems to proceed differently in various organisms. A recognition signal for selenocysteine

SF.I.ENIUM METABOI.ISM IN MICRO-ORGANISMS

95

incorporation within an mRNA secondary structure immediately 3' of the UGA codon was identified in E. coli (Heider et al., 1992) and similar signals may exist in other proteobacteria, such as Dm. baculatum (Zinoni et al., 1990). Possible mRNA secondary structures immediately 3' to the UGA codons of selenoprotein genes of other prokaryotic organisms such as the archaeon M. voltae (Halboth and Klein, 1992) and the Gram-positive bacterium C. purinolyticum (Garcia and Stadtman, 1991) were postulated. However, the calculated free energies of these structures (Garcia and Stadtman, 1991; Halboth and Klein, 1992) indicate that they may not be feasible thermodynamically and so the existence of different modes of selenocysteine-codon recognition in these organisms must be taken into account. An indication that such different modes may indeed exist comes from elucidation of the targeting signals for selenocysteine incorporation in eukaryotes. Incorporation into type-I thyroxine de-iodinase was shown to depend on the presence of a predicted mRNA secondary structure lying more than 1 kb away from the UGA codon for selenocysteine within the 3' untranslated region (Berry et al., 1991b). Thus, the present view of selective selenocysteine incorporation indicates that a variety of different recognition modes may exist in different organisms, which probably arose by individual co-evolution of mRNA signals and recognition factor(s). R.

SELENO-tRNAS

Incorporation of selenium into the modified tRNA base ~ e * - m a m ' - Uhas ~~ been shown to be independent of the products of the genes selA, selB and selC, but dependent on that of the gene selD (Leinfelder et al., 1988a; Kramer and Ames, 1988; Stadtman etal., 1989). Thus, phosphoroselenoate synthesized by SelD (Veres et al., 1992; Ehrenreich et al., 1992) delivers selenium for synthesis of the modified tRNA base and of selenocysteyl residues. The postulated involvement of a selenocysteyl residue as a potential selenium donor for tRNA modification (Veres et al., 1990) can therefore be excluded. The products of the genes rrmC and trmE (Bjork etal., 1987), which have a function in biosynthesis of the mam' modification, have been shown to be dispensible for selenium incorporation into tRNAs (Wittwer and Stadtman, 1986). In contrast, selenium-containing tRNAs are completely absent from an asuE mutant (Kramer and Ames, 1988) which is deficient in synthesizing s2-thiolated uridines (Sullivan et al., 1985). Results obtained with permeabilized E. coli by Wittwer and Stadtman (1986) led to the suggestion of a biosynthetic pathway for se2-mams-U (Fig. 8) biosynthesis from s2-mam'-U through a specific sulphur-selenium exchange. In view of present knowledge, the ATP dependence of this process (Wittwer and Stadtman, 1986) may reflect the requirement of SelD

96

J . HEIDKR A N D A. R&K

for synthesis of phosphoroselenoate (Ehrenreich et al., 1992; Veres et al., 1992) rather than the suggested activation of the sulphur moiety (Wittwer and Stadtman, 1986). Since phosphoroselenoate may be predicted to act as an energy-rich compound itself, sulphur activation of the s2-mam5-U precursor may not be required, and even a non-enzymic sulphur-selenium exchange seems chemically plausible. It is very interesting that, until now, the presence of selenium-modified tRNAs was only observed in bacteria which also synthesize selenoproteins (Stadtman, 1990; Heider et al., 1991). Therefore, the presence of selenium in se2-mam5-U may reflect a particular sensitivity of its precursor s2-mam5-Ufor reaction with phosphoroselenoate. V. Selenium versus Sulphur A.

IN CATALYSIS

Most selenium-containing enzymes and tRNAs are functional when selenium is replaced by sulphur. To date, glycine reductase is the only seleno-enzyme for which no sulphur-containing isoforms are known. This correlates with the strict demand for selenium supplementation by organisms employing this enzyme in their energy metabolism (Diirre and Andreesen, 1983,1986). For all other seleno-enzymes, close homologues lacking selenium which catalyse the same reaction are known. Examples of these selenium-sulphurenzyme couples are FDHH and FDHN of E. coli (selenium) and FDHs from M.formicicum and W. succinogenes (sulphur) (see Fig. 4), nickeliron and nickel-iron-selenium hydrogenases (Fauque et al., 1988) or typeI (selenium) and type-I1 (sulphur) thyroxin de-iodinases from mammals (Behne et al., 1990; Berry et al., 1991c; Safran el al., 1991). Sulphur- and selenium-containing isoenzymes may even be synthesized in the same organism. Comparisons of selenium- and sulphur-containing isoenzymes have shown that catalysis by the former is generally more effective, but a clear statement was only possible when the same enzymes containing residues of either selenocysteine or cysteine were compared. This was achieved with the FDHH isoenzyme of E. coli, which retained enzymic activity when the selenocysteyl residue was replaced by one of cysteine (Zinoni et al., 1987). A detailed analysis of the catalytic properties of the two variants has recently been performed (Axley et al., 1991). The sulphur and selenium enzymes did not exhibit significant differences in stability or sensitivity to inhibitors, but they differed in their catalytic properties. Interestingly, the K, value for formate of the sulphur enzyme turned out to even be slightly lower than that of selenium FDHH, but a drastic decrease in the rate of formate

SELENIUM METABOLISM IN MICRO-ORGANISMS

97

oxidation upon exchange of selenium by sulphur is responsible for a significantly lower catalytic efficiency of the sulphur enzyme (Table 1; Axley et al., 1991). This suggests direct participation of the selenium moiety of FDH in catalysis, but argues against its involvement in initial formate binding. B . COMPETITION DURING INCORPORATION

Cysteyl- and methionyl-tRNA synthetases from E. coli have been reported to charge tRNACys and tRNAMe' with the selenium analogues in vitro (Hoffman et a l . , 1970; Young and Kaiser, 1975). This contrasted with the finding that E. coli grown in a nutritionally rich medium in the presence of minute concentrations of radioactive selenite incorporates selenocysteine into distinct proteins, but not non-specifically into all proteins (Cox et al., 1981). This apparent discrepancy was resolved when the biosynthetic pathway for insertion of selenocysteine into proteins was elucidated (Bock et al., 1991a). According to the scheme shown in Fig. 8, all biosynthetic steps occur in tRNASe' in the ester-bonded state. Thus, neither free selenocysteine nor selenomethionine (derived from it) is required as an individual metabolite. However, selenium incorporation into proteins may also be accomplished by misacylated tRNACy"and tRNAMe'species under certain conditions such as the presence of high concentrations of selenite or selenate in the medium (Cowie and Cohen, 1957; Tuve and Williams, 1961). The main seleniumcontaining compound incorporated under these conditions has been characterized as selenomethionine (Huber and Criddle, 1967; Hartmannis and Stadtman, 1982; Frank et al., 1985). Recently, a method has been established which allows synthesis of proteins in which all methionyl residues are replaced by selenomethionine by overexpression of a cloned gene in a methionine-requiring strain of E. coli when grown in a medium supplemented with selenomethionine. It provides an elegant solution of the phase problem in determination of crystal structures of proteins (Hendrickson et al., 1990; Yang et al., 1990; Chen and Bahl, 1991). No deleterous effects on cell viability were observed in these experiments, which is in accord with earlier observations that selenomethionine can substitute for methionine during growth (Cowie and Cohen, 1957; Frank et al., 1985). It also suggests that the basis for the pronounced toxicity of selenite for many bacteria (Tuve and Williams, 1961; Weiss et al., 1965; Banffer, 1971) may reside solely in the non-specific substitution of cysteyl residues by those of selenocysteine. Unlike selenomethionine, selenocystine, which is readily taken up by cystine-transport systems in E. coli (Berger and Heppel, 1972), is very toxic for E. coli.

98

J HEIDbR AND A H O C K

VI. Transport and Metabolism of Selenium-Containing Compounds Since selenium is present in media in the redox state of +4 (in selenite) and of -2 in biological macromolecules, a reduction pathway must exist that converts selenite to selenide. At present, it is assumed that the thioredoxinglutaredoxin system is involved in this initial step (Holmgren and Kumar, 1989; Bjornstedt el al., 1992). This notion is supported by reported in vitro reduction of selenite by glutathione to selenotrisulphides (Nakagawa et al., 1988) which have been demonstrated to be substrates for further reduction through the thioredoxin system (Bjornstedt et al., 1992) or through glutathione reductase (Ganther, 1971). However, there is ample in vitro and in vivo evidence that seleniumcontaining compounds may readily enter pathways of sulphur metabolism. Formation of seleno-amino acids is catalysed in vitro by a number of pyridoxal phosphate-containing enzymes such as tryptophan synthase or O-acetylhomoserine sulphydrylase (Esaki and Soda, 1987a). In vivo, 0acetylserine sulphydrylase of E. coli is very likely to be the entry site for selenium into sulphur metabolism since selenite-resistant mutants were obtained which lack this enzyme (Fimmel and Loughlin, 1977). A few microbial enzymes involved in degradation of seleno-amino acids are known, the majority not exhibiting selenium-sulphur specificity (Soda, 1987). For example, a non-specific L-methionine y-lyase was detected in several bacterial species, degrading selenomethionine to ketobutyrate, ammonia and methylselenide (Esaki and Soda, 1987b)while a D-selenocystine a,&lyase, producing pyruvate, ammonia and elemental selenium, has been described (Soda el al., 1988). However, at least one catabolic enzyme which exhibits specificity for selenium is known. This is L-selenocysteine P-lyase from Citrobacter freundii and from mammals, generating elemental selenium and alanine from L-selenocysteine (Esaki and Soda, 1987~). Maximal rates of synthesis of selenium-containing biomolecules take place at concentrations of selenite in growth media as low as 1 PM (Cox et al., 1981) while selenocysteine incorporation into FDHH of E. coli has been shown to be saturated at a selenite concentration of 100 nM when expressed from a polycopy vector (Zinoni et al., 1987) and at a concentration of 20 n M when synthesized from a single chromosomal copy (B. Wollner and A. Bock, unpublished observation). Accumulation of selenium from the medium may amount to 50- to 100-fold, as for example in C. sticklandii (Stadtman, 1978). Although most selenium is probably incorporated into macromolecules, these accumulation ratios indicate that active transport systems for selenium must exist. Several attempts to characterize such uptake systems have been performed (Brown and Shrift, 1980, 1982; Hudman and Glenn, 1984; Lindblow-Kull et al., 1985; Bryant and Laishley,

SELENIUM METABOI.ISM IN MICRO-ORGANISMS

99

1989). It was found that transport of selenate into E. coli is intimately connected with that of sulphate, although selenate exhibits a somewhat lower affinity for the sulphate permease (Lindblow-Kull el al., 1985). Like sulphate transport, selenate uptake is repressed by the presence of cysteine in the medium (Brown and Shrift, 1982). Further metabolism of selenate also appears to proceed along the sulphate pathway since formation of stable adenosine phosphoselenate catalysed by ATP sulphurylase has been reported (Wilson and Bandurski, 1958) while growth of E. coli in minimal media containing selenate results in pronounced incorporation of selenomethionine into proteins (Huber et al., 1967). Selenite transport, in contrast, is not repressed in the presence of cysteine (Brown and Shrift, 1982). This suggests that a distinct transport system exists for uptake of selenite. Selenite appears to be transported on the sulphate carrier only when it is present at very high concentrations, which reflects the 50-fold lower affinity of the sulphate permease for selenite compared to the cognate substrate sulphate (Lindblow-Kull et al., 1985). The existence of specific uptake systems for selenite was also demonstrated for Selenomonas ruminantium, a species which cannot transport and metabolize sulphate or selenate (Hudman and Glenn, 1984) and for C. pasteurianum (Bryant and Laishley, 1989). In both organisms, selenite transport was shown to depend on an energized membrane. Although incorporation of seleno-amino acids into proteins has been demonstrated under these conditions, it has not been determined whether this was due to non-specific replacement of sulphur-amino acids or to formation of selenoproteins (Hudman and Glenn, 1984). Differences in non-specific uptake rates for selenite and selenate on the sulphate permease in different bacterial species may explain the surprisingly high levels of tolerance against selenite exhibited by some bacteria such as species of the genera Proteus and Salmonella (Weiss et al., 1965). These organisms are able to synthesize selenoproteins in response to low concentrations of selenium (Heider et al., 1991) and, therefore, must possess similar specific selenite-uptake systems as selenite-sensitive organisms. However, in the presence of very high concentrations of selenite, resistant species take up much less selenium, in comparison to selenite-sensitive species (Weiss et al., 1965). A biochemical basis for the differences in nonspecific selenite uptake may be provided by differences in the kinetics of sulphate uptake between species. Whereas a K, value of 3 PM has been reported for the sulphate permease in E. coli (Lindblow-Kull et al., 1985), the transport system in S. typhimurium has a K , value of 474 ~ L Mfor sulphate (Brown and Shrift, 1980).

J I 1 t l l ) t K AND A

UOC’K

VII. Geochemistry of Selenium

The geochemical cycle of selenium (Shrift, 1964) and the bacterial contributions to environmental selenium biotransformation have received major new insights in recent years (Oremland et al., 1989). Although selenium occurs in nature in forms resembling those of sulphur, differences between the elements are immediately visible from the Pourbaix diagrams (Fig. 9). Under aerobic conditions, e.g. in seawater, selenium is predominantly in the form of selenate, along with selenite, whereas organic selenides, mainly seleno-amino acids, prevail in anoxic water. In contrast to sulphide, free inorganic selenide cannot be detected in water samples (Cutter, 1982). Until recently, it was generally accepted that the metabolism of selenate proceeds by non-specific co-metabolism with sulphate (Doran 1082). Indeed, the capacity for selenate reduction to selenide was reported for sulphate-reducing bacteria pregrown on sulphate, probably using the same biochemical pathway as that used in sulphate respiration (Zehr and Oremland, 1987). However, this type of selenate reduction was found to be outcompeted when sulphate was present, indicating that it does not play a significant role under natural conditions (Zehr and Oremland, 1987). Organisms capable of dissimilatory selenate reduction to elemental selenium have been identified recently in high-selenium drainage water in seleniferous areas in California (Macy et al., 1989; Oremland el al., 1989; Steinberg and Oremland, 1990; Steinberg and Oremland, 1992); these isolates are described as Pseudornonas spp. (Macy et a l . , 1989) or as vibriolike bacteria (Steinberg and Oremland, 1992). Selenate reduction by these organisms is distinct from sulphate reduction since none of the isolated selenate-respiring organisms was capable of sulphate respiration. Rather, the organisms are generally able to perform anaerobic respiration with nitrate (Macy et a l . , 1989; Oremland et a l . , 1989; Steinberg and Oremland, 1992), which is used as a preferred electron acceptor over selenate (Steinberg and Oremland, 1992). Determinations of the contents of selenate and selenite in vertical sediment profiles have confirmed the premise of selenate reduction by organisms different from sulphate-reducing bacteria. Selenate and selenite reduction occur near the sediment surface, whereas sulphate reduction occurs only in deeper layers (Oremland et al., 1989). Thus, reductive processes involving selenium compounds take place at the site corresponding to their redox potential, +500 mV around pH 7 for Se042-/Se032-,in contrast to -200 mV for S042 /S2- (Fig. 9). Regarding FIG. 9. Pourbaix diagrams of sulphur- and selenium-containing compounds displaying thc thermodynamically stable rcdox states of the elements at given pH values and redox potential conditions (Kaim and Schwederski, 1991). The shaded areas represent the pH value and rcdox potential range which is usually accessiblc for neutrophilic organisms.

101

SELENIUM METABOLISM IN MICRO-OR<>ANISMS

L

-1.41

p

0.2

U

I

-1.0

Hs-

1 .o

1.41

2-

1 . 8 i

-2

0

2

4

6

8

10

12

14

16

PH

-1.8 -1.4-

HSe-

-1 .o h

L m .-c

Se:

H-Se

-0.6 -0.2

C

9)

R

0.2 0.6

1.o 1.4

h5e04

2.2 -2

0

2

4

6

8 PH

10

12

14

16

102

J iIFII)I-R ANI) A ROC’K

the low abundance of selenium in the environment, it is unlikely that a specific set of enzymes is synthesized for selenate reduction in these organisms, rather, as speculated by Oremland et al. (1989), a non-specific multicomponent terminal reductase such as the dimethyl-sulphoxide reductase in E. coli (Bilous and Weiner, 1985) may be responsible. Reduction of selenite to elemental selenium is a well-known feature of several organisms, such as Veillonella alcalescens and aerobically grown Salmonella spp. (McCready et al., 1966; Woolfolk and Whiteley, 1962), when they are confronted with selenite. The physiological basis for this reduction is not known, but it does not seem to be linked to anaerobic respiration since none of these species is able to grow at the expense of selenite reduction. Further reduction of selenium to selenide appears to be uncommon, except by biosynthetic pathways which result in selenium-containing biomolecules. Although tiny amounts of selenide produced from red selenium (Se,) by Thiobacillus ferrooxidans pre-grown under Fe3+ respiratory conditions at pH 3 have been reported (Racon and Ingledew, 1989), it is not known whether selenide can substitute for sulphide in growth of chemolithotropic organisms or whether it can be co-metabolized together with sulphide during leaching of sulphidic ores. Methyl selenides, especially the volatile dimethyl selenide, and methyl selenoxides, are common degradation and detoxification products of selenium (Reamer and Zoller, 1980). They may be formed directly during degradation of selenomethionine, as catalysed by methionine y-lyase (Esaki and Soda, 1987b), or by methylation of selenides at the expense of S-adenosylmethionine (Drotar et al., 1987). Dimethyl selenide can also be used as a carbon and energy source by some methanogenic and sulphate-reducing bacteria, releasing methane and carbon dioxide (Oremland and Zehr, 1986). Oxidative biological reactions involving selenium are probably rather common in aerobic water or soil. Although not much research has focused on this topic, a strain of Bacillus megaterium has been reported to convert elemental selenium into selenite. This strain even oxidizes the very inert grey modification of elemental selenium at a slow rate (Sarathchandra and Watkinson, 1981). Selenite is the species used for biosynthesis of selenium-containing biomolecules (Biick et al., 1991a; Fig. 10). Its uptake and reduction to a redox state of -2, as present in biological molecules, is unclear, but available data suggest a competitive situation with a specific pathway leading to selenoproteins and tRNAs and a non-specific one feeding into sulphur metabolism, depending on growth conditions and selenium concentration

SI;I.ENIUM MEI'ABOI.ISM IN MICRO-ORGANISMS

103

I

Anabolic sulphate-reduction pathway Cornetabolisrn in sulphate reducers ?

FIG. 10. Diagram showing the geochemical cycle of selenium.

(Bock el al., 1990; Stadtman, 1990). Biological degradation of selenoorganic compounds may result in different end-products, such as elemental selenium, selenide or methyl selenides (Esaki and Soda, 1987b,c; Soda et al., 1988; Fig. 10).

VIII. Conclusions Application of genetic and molecular-biological techniques have been indispensible tools in elucidation of the biochemical function of the trace element selenium. Determination of the nucleotide sequences of genes coding for selenoproteins revealed that the genetic code can be expanded to accommodate an additional amino acid. Isolation of mutants blocked in selenoprotein formation and cloning of the respective genes presented information on a hitherto unique pathway of amino-acid biosynthesis and incorporation. Overexpression of the genes, moreover, provided a means to purify enough of the gene products to enable detailed analysis of novel enzymic reactions. Recombinant-DNA techniques may even aid in construction of new selenoproteins. Despite progress in our knowledge, there are a number of areas in

104

J IIklI)EH A N D A ROCK

selenium biochemistry which need special o r renewed interest. It is unknown how selenite is taken up by micro-organisms or how it is reduced. What are the steps which prevent selenium flowing into sulphur metabolic routes at physiological concentrations and what is the biochemical basis of selenium toxicity? A deeper understanding of the mechanism by which coding specificity is brought about by mRNA will present information of general importance on the translational elongation and termination steps. Last, but not least, development of techniques to specifically incorporate selenocysteine into proteins will give nuclear magnetic-reasonance and X-ray analysis of proteins a new dimension. IX. Acknowledgements We wish to thank J. R. Andreesen and A . Klein for communicating results

prior to publication. This work was supported by the Bundesministerium fur Forschung und Technologie (via Genzentrum Munchen) and the Fonds der chemischen Industrie. REFERENCES

Agris, P. F. (1991). Biochimie 73, 1345. Albracht, S. P. J. (1y90). In “41. Colloquium Mosbach 1W):The Molecular Basis of Bacterial Metabolism” (G. Hauska and R. K. Thauer, eds), pp. 40-51, Springer-Verlag, Berlin. Andreesen, J . R. and Ljungdahl, L. G. (1973). Journalof Bacteriology 116, 867. Arkowitz, R. A. and Abeles, R. H. (1989). Biochemistry 28, 4639. Arkowitz, R. A. and Abeles, R. H. (1990).Journalofthe American ChemicalSociety 112,870. Arkowitz, R. A. and Abeles, R. H . (1991). Biochemistry 30,4W. Axley, M. J. and Grahame, D. A. (IWI). Journal of Biological Chemistry 246, 13731. Axley, M. J.. Grahame, D. A . and Shdtman, T. C. (1990). Journal of Biological Chemistry 265, 18213.

Axley, M. J., Bock, A. and Stadtman, T. C . (1991). Proceedings of the National Academy of Sciences of the United States of America 88, 8450. Bacon, M. and Ingledew, W. J. (1989). FEMS Microbiology Letters 58. 189. Banffer, J. R. (1971). Zentralblatt fur Bakteriologie 217, 35. Baron, C. and Bock, A. (1991). Journal of Biological Chemistry 246, 20375. Baron, C., Heider, J. and Bock, A. (1990). Nucfeic Acids Research 18, 6761. Baron, C., Westhof, E., Bock, A . and GiegC, R. (1992). Journal of Molecular Biology, in press. Baron, S. F.. Williams, D. S., May, H. D., Patel, P. S., Aldrich, H . C. and Ferry, J. G. (1989). Archives of Microbiology 151. 307. Behne, D.. Kyriakopoulos, A., Meinhold, H. and Kohrle, J . (19YO). Biochemical and Biophysical Research Communications 173, 1 143. Berg, B. L., Li. J.. Heider, J. and Stewart, V. (1991a). Journal of Biological Chemistry 266, 22380.

Berg, 9 . L., Baron. C. and Stewart, V. (1991b). Journal of Biological Chemistry 266, 22386.

Berger, E. A. and Heppel, L. A. (1972). Journal of Biological Chemistry 247, 7fiX4.

SELENIUM METAROI.ISM IN MICRO-ORGANISMS

105

Berry, M. J., Banu, L. and Larsen, P . R. (1991a). Nature 349, 438. Berry, M. J., Banu, L., Chen, Y . . Mandel, S. J., Kieffer, J. D . , Harney, J. W. and Larsen, P. R . (1991b). Nature 353, 273. Berry, M. J., Kieffer, J. D., Harney, J. W. and Larsen, P. R. (1991~).Journal of Biological Chemistry 266, 14155. Bilous, P. T. and Weiner, J. H. (1985). Journal of Bacteriology 163, 369. Bjork, G. R., Ericson, J. U., Gustafsson, C. E. D., Hagemall, T. G., Jonsson, Y. H. and Wikstrom, P. M. (1987). Annuul Reviews of Biochemistry 56, 263. Bjornstedt, M., Kumar, S. and Holmgren, A. (1992). Journal of Biological Chemistry 267, 8030. Biick, A , , Baron, C., Forchhammer, K., Heider, J., Leinfelder, W., Sawers, G., Veprek, B.. Zehelein, E . and Zinoni, F. (1990). In “41. Colloquium Mosbach 1990: The Molecular Basis of Bacterial Metabolism” (G. Hauska and R. K. Thauer, eds), pp. 61-68. SpringerVerlag, Berlin. Biick, A., Forchhammer, K., Heider, J., Leinfelder, W., Sawers, G., Veprek, B. and Zinoni. F. (1991a). Molecular Microbiology 5 , 515. Biick, A., Forchhammer, K., Heider, J. and Baron, C. (1991b). Trends in Biochemical Sciences 16, 463. Munchen. Bohm, R. (1991). Dissertation: Ludwig-Maximilians-Universitat, Bohm, R., Sauter, M. and Biick, A. (1990). Molecular Microbiology 4, 231. Bokranz, M., Gutmann, M., Kortner, C., Kojro, E., Fahrenholz, F., Lauterback, F. and Kroger. A. (1991). Archives of Microbiology 156, 119. Boursier, P., Hanus, F. J., Papen, H., Becker, M. M., Russell, S. A. and Evans, H. J. (1988). Journal of Bacteriology 170, 5594. Brown, T. A. and Shrift, A. (1980). Canadian Journal of Microbiology 26, 671. Brown, T. A. and Shrift, A. (1982). Canadian Journal of Microbiology 28, 307. Bryant, R. D. and Laishley, E. J. (1989). Canadian Journal of Microbiology 35, 481. Buckel, W. (1990). In “41. Colloquium Mosbach 1990: The Molecular Basis of Bacterial Metabolism” (G. Hauska and R. K. Thauer, eds), pp. 21-30, Springer-Verlag, Berlin. Burk, R. F. (1991). FASEB Journal 5 , 2274. Chambers, I., Frampton, J.. Goldfarb, P., Affara, N., McBain, W. and Hamson, P. R. (1986). EMBO Journal5, 1221. Chen, C . 4 . and Stadtman, T . C . (1980). Proceedings of the National Academy of Sciences of the United States of America 77, 1 4 3 . Chen, W. and Bahl, 0.P. (1991). Journal of Biological Chemisrry 266, 9355. Ching, W.-M. (1984). Proceedings of the National Academy of Sciences of the United States of America 81, 3010. Ching, W.-M. and Stadtman, T. C. (1982). Proceedings of the National Academy of Sciences of the United Stares of America 79, 374. Ching, W.-M., Wittwer, A. J., Tsai, L. and Stadtman, T. C. (1984). Proceedings of the National Academy of Sciences of the United Stares of America 81, 57. Ching, W.-M., Alzner-DeWeerd, B. and Stadtman, T . C. (1985). Proceedings of the National Academy of Sciences of the United Stares of America 82, 347. Cone, J. E., Martin del Rio, R. and Stadtman, T. C. (1976). Proceedings of the National Academy of Sciences of the United States of America 73, 2659. Cowie, D. B. and Cohen, G . N. (1957). Biochimica et Biophysica Acta 26, 252. Cox, J. C., Edwards, E. S. and DeMoss, J. A. (1981). Journal of Bacteriology 145, 508. Cutter, G. A. (1982). Science 217, 829. Dietrichs, D., Meyer, M., Rieth, M. and Andreesen, J. R. (1991). Journal of Bacteriology 173. 5983. Dilworth, G. L. (1982). Archives of Biochemistry and Biophysics 219, 30. Dilworth, G. L. (1983). Archives of Biochemistry and Biophysics 221, 565. DiMarco. A. A., Bobik, T. A. and Wolfe, R.S. (1990). Annual Reviews of Biochemistry 59, 355. Doran, J. W. (1982). Advances in Microbial Ecology 6, 1.

106

J

HElDER A N D A . ROCK

Drotar. A., Fall, L. R., Mishalanie. E. A , , Tavernier. J. E . and Fall, R. (1987). Applied and Environmental Microbiology 53, 21 1 1 . Diirre, P. and Andreesen, J. R. (1983). Journal of Bacteriology 154, 192. Diirre, P. and Andreesen, J. R. (1986). Biologie in Unserer Zeit 16, 12. Ehrenreich, A.. Forchhammer, K., Tormay, P., Veprek, B. and Biick, A. (1992). European Journal of Biochemistry 206, 767. Eidsness, M. K., Scott. R. A., Prickril, B. C., DerVartanian, D. V., LeGall, J . , Moura, I., Moura, J . J . G. and Peck, H . D. (1989). Proceedings of the National Academy of Sciences of the United States of America 86, 147. Engelhardt, H . , Forchhammer, K., Miiller, S.,Engel, A . and Bock, A . (1992). Molecular Microbiology, 6, 346 1. Enoch, H. G . and Lester, R. L. (1975). Journal of Biological Chemistry 250, 6693. Esaki, N. and Soda, K. (1987a). Methods in Enzymology 143, 291. Esaki, N . and Soda, K. (1987b). Methody in Enzymology 143, 459. Esaki, N. and Soda, K. (1987~).Methods in Enzymology 143, 493. Fauque, G.. Peck, H . D. Jr, Moura, J . J. G., Huynh, B. H.. Berlier. Y., DerVatranian, D. V., Teixeira, M., Przybyla, A. E . , Lespinat, P. A.. Moura. I. and LeGall, J. (1988). FEMS Microbiology Reviews 54, 299. Fimmel, A. L. and Loughlin. R. E. (1977). Journal of General Microbiology 103. 37. Floht, L. (1989). In “Glutathione: Chemical. Biochemical and Medical Aspects-Part A” (D. Dolphin, R. Poulson and 0. Avramovic, eds), pp. 643-731. Wiley, London. Forchhammer, K. and Bock, A. (I99la). Journal of Biological Chemisrry 266, 6324. Forchhammer, K. and Bock, A . (1991b). Naturwissenschaften 78. 4979. Forchhammer, K., Leinfelder, W. and Biick, A. (1989). Nature 342, 453. Forchhammer, K., Leinfelder, W., Boesmiller, K.,Veprek. B. and Bock, A. (I99la). Journal of Biological Chemistry 266, 6318. Forchhammer, K., Boesmiller, K. and Bock, A. (1991b). Riochimie 73, 1481. Forster, C . ,Ott, G.. Forchhammer, K. and Sprinzl, M. (1990). Nucleic Acids Research 18,487. Frank, P., Licht, A., Tullius,T. D., Hodgson, K.0.and Pecht, I. (1985). Journalof Biological Chemistry 260, 5518. Freudenberg. W., Mayer, F. and Andreesen, J. R. (1989). Archivesof Microhiology 152, 182. Ganther, H . E. (1971). Biochemistry 10, 4089. Garcia, G. E . and Stadtman, T. C. (1991). Journal of Bacteriology 173, 2093. Gardlik, S. and Rajagopalan, K. V. (1990). Journal of Biological Chemistry 265, 13047. Gleason, F. K. and Holmgren A. (1988) FEMS Microbiology Reviews 54, 271. Gottschalk, G . (1986). “Bacterial Metabolism”. Springer-Verlag, Berlin. Giinzler, W. A., Steffens, G. J., Grossman, A., Kim, S.-M. A., Otting, F., Wendel, A. and Floht. L. (19x4). Hoppe-Seyler’s Zeitschrift fur Physiologische Chemie 365, 195. Halboth, S. and Klein. A. (1992). Molecular and General Genetics 233, 217. Hartmannis, M. G. N . and Stadtman, T. C. (1982). Proceedings of the National Academy of Sciences of the United Stares of America 79, 4912. Hatfield, D., Choi, I . S., Mischke, S. and Owens, L. D. (1992). Binchemicaland Biophysical Research Communications 184, 254. Hatfield, D . L., Smith, D. W. E., Lee. B. J., Worland, P.J. and Oroszlan, S. (1990). Critical Reviews in Biochemistry and Molecular Biology 25, 71. Hatfield, D. L.. Lee, B. J., Price, N. M. and Stadtman, T. C. (1991). Molecular Microbiology 5, 1183.

He, S. H., Teixeira, M.,LeGall. J., Patil, D . S., Moura, I, Moura, J. J . G., DerVartanian, D.V.,Huynh,B. H . a n d P e c k , H . D..Jr(1989). JournalofBiologicalChemisrryM4,2678. Heider, J. and Bock, A. (1992). Journal of Bacteriology 174, 659. Heider, J . , Leinfelder, W. and Bock, A. (1989). Nucleic Acids Research 17, 2529. Heider, J., Forchhammer, K., Sawers, G. and Biick, A. (1991). Archives of Microbiology 155, 221.

Heider, J.. Baron, C. and Bock, A. (1992). EMBO Journal, 11, 3759. Hendrickson, W. A . , Horton, J. R. and LeMaster, D. M. (1990). E M B O Journal 9, 1665.

SELENIUM METABOLISM IN MICRO-ORGANISMS

107

Hoffman, J. L., McConell, K. P. and Carpenter, D. R. (1970). Biochimica er Biophysica Acra 199, 531. Holcenberg, J . S. and Stadtman, T. C. (1%9). Journal of Biological Chemistry 244, 1194. Holmgren, A. and Kumar, S. (1989). I n “Selenium in Biology and Medicine” (A. Wendel, ed.), pp. 47-51. Springer-Verlag, Berlin. Hormann, K. and Andreesen, J. R. (1989). Archives of Microbiology 153, 50. Huber, R. E. and Criddle, R. S. (1967). Biochimica er Biophysica Acta 141, 587.

Huber, R. E., Segel, I. H. and Criddle, R. S. (1%7). Biochimica er Biophysica Acra 141, 573. Hudman, J. F. and Glenn, A. R . (1984). Archives of Microbiology 140, 252. Imhoff, D. and Andreesen, J. R. (1979). FEMS Microbiology Letters 5 , 155. Jones, J. B. and Stadtman, T. C. (1981). Journal of Biological Chemistry 256, 656. Jones, R. W. (1980). FEMS Microbiology Lerters 8, 167. Kaim, W. and Schwederski, B. (1991). “Bioanorganische Chemie”. B. G. Teubner, Stuttgart. Kearny, J . J . and Sagers, R. D. (1972). Journal of Bacteriology 109, 152. Konig, K. and Andreesen, J . R. (1992). BioEngineering 8/2, 78. Kortner. C. and Kroger, A. (1992). BioEngineering 8, 38. Kramer, G. F. and Ames, B. N. (1988). Journal of Bacteriology 170, 736. Kroger, A,, Winkler, E., Innerhofer, A., Hackenberg, H. and Schrager, H. (1979). European Journal of Biochemistry 94, 465. Kroger, A,, Schrder, I . , Krems, B. and Klimmek, 0. (1990). In “41. Colloquium Mosbach 1990: The Molecular Basis of Bacterial Metabolism” (G. Hauska and R. K. Thauer, eds), pp. 128-133. Springer-Verlag, Berlin. Lee, 8. J., Worland, P. J., Davis, J . N., Stadtman, T . C. and Hatfield, D. L. (1989). Journal of Biological Chemisrry 264, 9724. Lee, B. J . , Rajagopalan, M.,Kim, Y. S., You, K. H., Jacobson, B. and Hatfield. D . (1990). Molecular and Cellular Biology 10, 1940. Leinfelder, W.. Forchhammer, K., Zinoni, F., Sawers, G., Mandrand-Berthelot, M.-A. and B&k, A. (1988a). Journal of Bacteriology 170, 540. Leinfelder, W., Zehelein, E., Mandrand-Berthelot, M.-A. and Bock, A. (1988b). Nature 331, 723.

Leinfelder, W., Stadtman, T. C. and Bock, A. (1989). Journal of Biological Chemistry 264, 9720.

Leinfelder, W., Forchhammer, K.,Veprek, B., Zehelein, E. andBiick, A. (1990). Proceedings of rhe National Academy of Sciences of the United Stares of America 87, 543. Leonhardt, U. and Andreesen, J . R. (1977). Archives of Microbiology 115, 277. Lindblow-Kull, C., Kull, F. J. and Shrift, A. (1985). Journalof Bacteriology 163, 1267. Ljungdahl, L. G. and Andreesen, J. R. (1978). Methods in Enzymology 53, 360. McCready, R. G., Campbell, J. N. and Payne, J . 1. (1%). Canadian Journalof Microbiology 12, 703.

Macy, J . M.,Michel, T. A. and Kirsch, D. G. (1989). FEMS Microbiology Lerrers 61, 195. Menon, N . K., Peck, H. D., Jr, LeGall, J . and Przybyla, A. E. (1987). Journalof Bacteriology 169, 5401.

Menon, N. K., Peck, H. D.,Jr, LeGall, J . and Przybyla, A. E. (1988). Journal of Bacteriology 170, 4429.

Meyer, 0 . and Rajagopalan, K. V . (1981). Journal of Biological Chemistry 259, 5612. Meyer, O., Jacobitz, S. and Kriiger, 8 . (1986). FEMS Microbiology Reviews 39, 161. Mizutani, T., Kurata, H. and Yamada, K. (1991). FEES Leners 289, 59. Moulis, J.-M., Lutz, M.,Gaillard, J. and Noodleman, L. (1988). Biochemistry 27, 8712. Muth, E., Morschel, E. and Klein, A. (1987). European Journal of Biochemistry 169, 571. Nakagawa, T., Aoyama, E., Kobayashi, N., Tanaka, H., Chikuma, M.,Sakurai, H. and Nakayama, M.(1988). Biochemical and Biophysical Research Communications 150. 1149. Neumann, H., Steinberg, I. Z., Brown, J. R., Goldberger, R. F. and Sela, M. (1967). European Journal of Biochemistry 3, 171. Oremland, R. S. and Zehr, J . P. (1986). Applied and Environmental Microbiology 52, 1031.

108

J IIEIIIFR A N D A .

HOCK

Oremland. R. S . , Hollihaugh, J . T . , Maest. A . S . , Presser, T. S . , Miller, I>. G . and Culbertson, C. W. (1989). Applied and Environmental Microbiology 55, 2333. Patel, P. S . and Ferry, J. G. (1988). Journal of Racteriology 170. 3390. Pecher, A , , Zinoni, F. and Bikk, A. (19x5). Archives of Microbiology 141, 359. Pinsent, J. (1954). Riochemical Journal 57, 10. Politino, M . , Tsai, l,., Veres, Z . and Stadtman, T. C. (1990). Proceedings of the Natiofial Academy of Sciences o f t h e United States of America 87. 6345. Reamer, D. C. and Zoller, W. H. (1980). Science 211, 600. Safran, M.. Farwell. A . P. and Leonard, J. L. (1091). Journal of Riological Chemistry 266, 13477. Sarathchandra, S. U . and Watkinson, J. H. (1981). Science 211, 600. Sawers, G . , Heider. J.. Zehelein, E. and Bock, A. (IY91). Journalof Bacteriology 173, 4083. Schauer, N. L. and Ferry, J . G . (1982). Journal of Racteriology 150, 1. Schauer, N. L. and Ferry, J. G . (1983). Journal of Racteriology 155, 467. Schauer, N. L. and Ferry, J . G. (1986). Journal of Racteriology 165. 405. Scherer, P. A . and Thauer. R. K. (1978). European Journal of Riochernistry 85, 125. Schlindwein, C., Giordano, G . , Santini, D.-L. and Mandrand, M.-A. (IWO). Journal of Racteriology 172, 61 12. Schiin, A,. Kannangara, C. G., Gough. S. and SOH, D. (1988). Nature 331, 187. Schon, A , , Bock, A , , Ott,G., Sprinzl, M. andSo11. D . (19x9). NucleicAcidsResearch 17,7159. Schrlder, T. and Andreesen, J . R . (1992). European Journal of Riochemisrry 206, 70. Schuckelt, R., Brigelius-FlohC, R., Maiorino, M . , Roveri, A , , Reumkens, J . , Strasshurgcr, W . , Ursini, F.. Wolf, B. and Flohe, L. (1991). Free Radicals Research Communications 14, 343. Schwartz, K. and Foltz, C. M. (1957). Journal of the American Chemical Society 79, 3292. Shrift, A . (1964). Nature 201. 1304. Shuher, A . P., Orr, E. C.. Recny, M. A , , Schendel. P. F., May, H. D . . Schauer, N. I-. and Ferry, J. G . (1986). Journal of Riological Chemistry 261. 12942. Sliwkowski. M. X. and Stadtman, T. C. (1088). Proceedings of the National Academy (~j’ Sciences of the United States of America 85, 368. Soda, K. (1987). Merhods in Enzymology 143, 453. Soda, K., Esaki, N.. Seraneeprakarn, V. andTanaka, H. (1988). Phosphorusand Sulfur38.25. Sprinzl. M.,Dank, N., Nock. S. and Schon, A. (1991). “Compilation o f tRNA and tRNA Gene Sequences”. Lahoratorium fur Biochemie, Universitlt Bayreuth. Stadtman, T. C. (1978). Methods in Enzymology 53, 373. Stadtman, T. C. (19x0). Annual Review of Riochemistry 49. 93. Stadtman, T. C. (1990). Annual Review of Riochemi.stry 59, I 1 1. Stadtman, T. C. (1991). Journal of Riological Chemistry 266. 16257. Stadtman, T. C., Davis, J. N., Zehelein. E. and Biick, A. (1989). RioFacrors 2, 35. Stadtman,T. C . , Davis, J . N.,Ching, W.-M., Zinoni,F. a n d B o c k , A . (1991). RioFuctor.s3, ? I . Stanzel, M., Schiin, A. and Sprinzl, M.(1991). In “Abstracts of the 14th International tRNA Workshop, Rydzyna-Poznan” (K. Mikolajczyk, ed.), p. 226. Steinherg, N. A . and Oremland. R . S. (l9YO). Applied and Environmental Microbiology 56, 3550. Steinherg, N. A . and Oremland, R. S. (1992). Applied and Environmental Microbiology 58, 426. Sullivan. M. A , , Cannon, J . F., Webb, F. H. and Bock, R. M. (19x5). Journaloj’RacteriokJgy 161. 368. Sunde, R. A . (l9Yo). Annual Reviews of Nutrition 10, 451. Tanaka, H . and Stadtman. T. C. (1979). Journal of Riological Chemistry 254, 447. Teixeira, M.. Fauque. G . , Moura, I., Lespinat, P. A , , Bcrlier, Y.. Prickril, B., Peck, H. D. Jr, Xavier, A. V., LeGall, J. andMoura, J. J. C . (1987). Europea~~/ournalofRiochcr~~istrv 167, 47. Turner. D . C. and Stadtman, T. C . (1973). Archives of Riochemistry and Biophysics 154,366. Tuve. T. and Williams. H . H. (1961). Journal of Riological Chemistry 236, 597.

SEI.ENIUM METAROI.ISM IN MICRO-ORGANISMS

109

Veres, Z., Tsai, L., Politino, M. and Stadtman, T. C . (1990). Proceedings o f t h e National Academy of Sciences of the United States of America 87. 6341. Veres, Z., Tsai, L., Scholz, T. D . , Politino, M . , Balaban, R. S. and Stadtrnan, T. C. (1992). Proceedings of the National Academy of Sciences of the United States of America 89, 2975. Vogels, G. D. and Van der Drift, C. (1976). Bacteriological Reviews 40,403. Voordouw, G., Menon, N . K., LeGall, J., Choi, E. S . , Peck, H. D. and Przybyla. A. E. (1989). Journal of Bacteriology 171, 2894. Wada, K., Aota, S., Tsuchiya, R., Ishibashi, F., Gojobori, T. and Ikemura, T. (1990). Nucleic Acids Research 18, 2367. Wagner. R. and Andreesen, J . R. (1977). Archives of Microbiology 114, 219. Wagner, R. and Andreesen, J . R . (1979). Archives of Microbiology 121, 255. Weiss. K. F., Ayres, J . C. and Kraft, A . A . (1965). Journal of Bacteriology 90. 857. Widdel. F. and Hansen, T. A. (1991). In “The Prokaryotes” (A. Balows, H. G. Triiper, M. Dworkin, W . Harder and K.-H. Schleifer, eds), pp. 583424. Springer-Verlag, New York. Wilson, L. G. and Bandursky, R. S . (1958). Journal of Biological Chemistry 233, 975. Wittwer, A . J . (1983). Journal of Biological Chemistry 258, 8637. Wittwer, A . J. and Ching, W.-M. (1989). BioFactors 2, 27. Wittwer, A. J. and Stadtman, T. C . (1986). Archives of Biochemistry and Biophysics 248,540. Woolfolk, C . A . and Whiteley, H. R. (1962). Journal of Bacteriology 84, 647. Yamamoto, I . , Saiki, T., Liu. S.-M. and Ljungdahl, L. G . (1982). Journal of Riological Chemistry 258, 1826. Yamazaki, S. (1982). Journal of Biological Chemistry 257. 7926. Yang, W., Hendrickson, W. A , , Kalman, E. T. and Crouch, R. J . (1990). Journalof Biological Chemistry 265, 13553. Young, P. A . and Kaiser, I. I. (1975). Archives of Biochemistry and Biophysics 171, 483. Zehr, J. A . and Oremland, R. S. (1987). Applied and Environmental Microbiology 53, 1365. Zindcl, U., Freudenberg. W., Rieth, M., Andreesen, J. R., Schnell, J. and Widdel, F. (1988). Archives of Microbiology 150, 254. Zinoni, F., Birkmann, A , , Stadtman, T. C. and Bock, A . (1986). Proceedingsofthe National Academy of Sciences of the United States of America 83, 4650. Zinoni, F.. Birkmann, A , , Leinfeldcr, W. and Bock, A. (1987). Proceedings ofthe National Academy of Sciences of the United States of America 84, 3156. Zinoni, F., Heider, J. and Bock, A . (1990). Proceedings ofthe National Academy ofSciences of the United States of America 87, 4660.

This Page Intentionally Left Blank

Regulation of the Onset of the Stationary Phase and Sporulation in Bacillus subtilis JAMES A. HOCH Division of Cellular Biology, Department of Molecukir and Experimental Medicine, The Scripps Research Institute, 10666 N . Torrey Pines Road, La Jolla, CA 92037, USA

. . A . FunctionsofthespoOgenes . . . B . lsolationofgenes for kinasesactivatingsporulation . C. Rolesof theotherspoOgenes . . . . . . 111. Control of thephosphorelay . . . . . . . , A . Control of phosphate flow . . . . . . . I.

Introduction

.

11. The phosphorelay

. .

.

.

. .

. .

. .

.

. .

. . .

V. VI.

V11.

.

. . .

. .

. . . . .

.

.

. . . . .

,

,

.

.

.

. .

Transcriptional regulation of genes for phosphorelay components . Transition-stateregulators . . . . . . . . . . . A . The AbrB protein . . . . . . . . . . . . H. TheHprprotein . . . . . . . . . . . . C. The Sin protein . . . . . . . . . . . . Alternatives tosporulation . , . . . . . . . , . Initiation of sporulation . . . . . . . . . . . . Acknowledgements. . . , . . . . . . . . . References , . . . . . . . . . . . . . .

B.

IV.

.

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

111 113 115 116 118 120 120 123 126 127 128 128 129 130 132 132

1. Introduction

Sporulation involves a complex series of intracellular morphological events occurring in a temporal sequence and resulting in an environmentally resistant body prepared to wait until better times to return to the cell from which it arose. In the natural environment, a cell has many means of ensuring its survival but, given a choice, it would rather grow and divide to maintain its position in its ecological niche. Sporulating organisms such ADVANCESIN MICRORIAI. PHYSIOLOGY, VOI.. 35 ISBN &-1242773.%2

Copynght0 1993. by Academic Press Limited All rights of reproductionin any form reserved

112

J

A

110~'11

as the aerobic bacilli are capable of competing successfully by maintaining high growth rates under conditions of nutrient sufficiency. When conditions are less favourable, such organisms retain the option to form a spore, thus allowing long-term survival under the most adverse natural circumstances. Formation of spores is an energy-intensive commitment on the part of the cell to produce a complex morphological structure. It should not be surprising, therefore, that the cell regulates initiation of sporulation carefully and only when further growth and division are not possible. How does a cell decide whether it will divide or sporulate? This simple question turns out to have a complex answer. Such a decision requires integration of the activities of a large number of synergistic and opposing regulatory activities, responding to different input signals. In early studies, it was found that a growing cell decided its fate during a small window of time in the cell cycle, and if it chose division it was committed to grow and divide before it could again initiate sporulation (Mandelstam and Higgs, 1974). Thus, conceptually the cell must sample its environment when presented with this window and compute from these signals its morphological fate (Fig. I). What are the signals, from where do they arise, and how are they transduced and integrated? Some answers to these questions have been obtained through studies of mutants that affect sporulation (Piggot and Coote, 1976). Since sporulation occurs through a defined series of temporal morphological events, mutants blocked at each step, or stage can be categorized. Those blocked at the earliest stage, stage 0, form none of the characteristic morphological

-

Sporulation Germination Cycle

Cell-Cycle Signals

FIG. 1. Schematic diagram showing options facing a bacterium capable of sporulation.

SPORULATION IN BACILLUS SUBTILIS

113

structures of sporulating cells. Those spo0 mutations were thought to define those genes whose products were involved in producing or recognizing signals that initiate the sporulation process (Hoch, 1976). Such spo0 mutants ultimately led to recognition of the spoOA gene as coding for the key transcription factor for sporulation initiation and to the pathway that activates its functions, the phosphorelay (Burbulys et al., 1991). Generally, sporulation is studied in a rich medium in which cells grow exponentially until they shut off growth and enter the stationary phase of growth. From the point at which exponential growth ceases (to) to the appearance of refractile spores (t6-t7), sporulation stages occur in an ordered sequence. The first hour or so after to is called the transition state between growth and sporulation, and this is the time that early stationary phase functions such as synthesis of proteases and antibiotics are produced. Initiation of sporulation is inexorably coupled to mechanisms that control gene expression during the early stationary phase of growth. Thus, spo0 mutants are blocked in transcription and expression of many early nonsporulation-related stationary phase functions as well as in initiation of sporulation. These mutants appear to be locked in exponential growth and, when they should be entering the transition state, they continue to grow and ultimately lyse without ever expressing such functions as synthesis of proteases or antibiotics. The spo0 mutants, therefore, lack the ability to shut down growth and division and cannot activate the transcription required to enter the classical stationary phase of growth. These phenotypes result from the inability of spo0 mutants to generate the active form of the SpoOA transcription factor because of defects in the signal-transduction pathway leading to its activation. 11. The Phosphorelay

The signal-transduction system for initiation of sporulation is a significant variation of two-component regulatory systems which function to interpret environmental signals in bacteria. Two-component systems are recently discovered mechanisms by which bacteria control transcription of a variety of genes in response to a wide variety of environmental and metabolic signals (Stock et al., 1989). The first component is a sensor kinase which receives the input environmental signal. The second component is a response regulator molecule which, in most cases, is a transcription factor specific for a number of genes whose products allow a response to the environmental signal. Figure 2 shows a general schematic diagram of the homologies observed between various sensor and response regulator proteins of two-component systems. Both components contain variable and

114

J

Regulation

A

110('11

Sensor protein NH; coo variable conserved

Regulatory protein NHj coo conserved' variable

Nitrogen

ntrC I

Phosphate

I

Osmotic

I

Chemotaxis

Sporulation

,

-

pchee

1-1

cheY

1-1

SPOOF

Unknown

FIG. 2. General structure o f sensor kinase and rcsponse-regulator proteins of twocomponent regulatory systems. Conserved and sariable regions are characteristic of each protein. The nitrogen and phosphate systems are activated by starvation for these nutrients whereas changes in osmolality control EnvZ and availability of attractants or repellants regulates CheA. Regulators of sporulation are unknown.

constant domains. The sensor kinases for a wide variety of systems of this type show homology in their carboxyl domain. The amino domains are unique to each protein and specific for the particular environmental signal that activates the sensor molecule. Conversely, the response-regulator proteins are all highly homologous in the amino-terminal 100-120 aminoacid residues, while the carboxyl portions of these molecules show less homology to one another. Although further combinations have now been recognized where the sensor and response regulator domains reside on a single polypeptide (Parkinson and Kofoid, 1992) in general, for the proteins studied here the homologies are as shown in Fig. 2. Among products of the sp00 gene, there are two proteins with responseregulator homologies, namely SpoOA and SpoOF (Trach et a l . , 1985; Stock et al., 1989). The gene product SpoOF is a protein consisting entirely of the observed homologous domain of response-regulator proteins (Trach et al., 1988), whereas SpoOA has the typical structure of response-regulator proteins that act as transcription factors (Kudoh et al., 1985; Ferrari el al., 1985). This pair is reminiscent of the CheY-CheB pair of proteins in the chemotaxis system but, in the case of chemotaxis, CheB is not a transcription activator (Stock et al., 1989). The homology observed between the sensor molecules and the homology of the response regulators suggests that all of these proteins work through a common mechanism (Fig. 3). An input signal serves to activate the kinase by promoting autophosphorylation of an internal histidine residue. This phosphorylation reaction is the actual signal-transduction event converting the concentration of a given effector

115

SPORUI.ATION IN RACILLUS SURTILIS

Autophosphorylation

I

KINASE

+

ATP

ACTlVATl0N SIGNAL

KINASE-PO4

I

+

ADP

(HIS-P04) Phosphotransfer

SPOOF - A 4 SPOOF-PO~ (ASPPO4)

Phosphatase

FIG. 3. Mechanism of phosphorylation of response-regulator proteins by sensor kinases in response to environmental signals during spore formation in Baciflus subtilis. All three enzymic functions may be properties of the kinase.

ligand to an activated protein molecule. The activated sensor kinase transfers this phosphate group to a response-regulator protein where it ultimately resides on the response regulator as a mixed anhydride of an aspartic acid residue. The histidine residue of the sensor kinase and the aspartic acid residue of the response regulators are conserved among all proteins of this type (Stock et al., 1990). The phosphorylated form of the response regulator is then activated to carry out its function, usually that of a positive transcription factor. In sporulation, however, the system works a little differently. A . FUNCTIONS OF THE SpOO GENES

The spoO genes were thought to code for the signal-recognition system for sporulation and, because of this, they were subjected to cloning and sequencing studies in the hope of deducing their function from their primary amino-acid sequences. Two of the important spoO genes coded for proteins with homology to a transcription factor, namely spoOA, and a gene of unknown function, spoOF, with homology to other two-component regulatory systems (Ferrari et a l . , 1985; Kudoh et a l . , 1985; Trach et a l . , 1985). Since response regulators were involved, i t was believed that one of the spo0 genes must code for a sensor kinase. When the third most common locus for spoO mutations, namely spoOB, was sequenced, the deduced product of this locus revealed a protein with no obvious homology to either component of two-component regulatory systems (Bouvier et a l . , 1984). Furthermore, cloning and sequencing of the spoOE, spoOH spoOJ and spoOK genes (already described) did not reveal a deduced protein from any of

116

J A HOCH

these loci with homology to sensor kinases of two-component systems. Because it was axiomatic that two-component systems work through phosphorylation of their response regulators to activate their functions, the apparent lack of kinases among spo0 genes deepened the mystery as to what sensor kinase was responsible for activating transcription functions of the spoOA protein. In addition, it was unclear what the role of SpoOF protein could be in this system. The CheY protein to which it is related, functions as a switch to reverse the direction of the flagellar motor but there turned out to be no relationship between SpoOF and chemotaxis or any other easily discernible process. As luck would have it, the sensor kinase gene was masquerading as a stage-I1 sporulation gene. B . ISOLATION OF GENES FOR KINASES ACTIVATING SPORULATION

Sequencing of a locus, namely spoltJ, in which mutations gave a stage-I1 sporulation phenotype, uncovered a gene for a sensor kinase with typical carboxy terminal homology to proteins of this class (Antoniewski ef af., 1990). Purification of this protein and studies of its activity in vitro showed that this kinase, KinA, was highly active in phosphorylation of the SpoOF molecule and weakly active in phosphorylation of SpoOA (Perego ef af., 1989). Studies of the kinetic parameters for these two substrates indicated that SpoOF was the kinetically preferred substrate over SpoOA in vitro, and presumably also in vivo (Burbulys et al., 1991). Thus, a kinase had been discovered to phosphorylate SpoOF, but the mechanism of phosphorylation of SpoOA was still inexplicable. Strains carrying the spolIJ mutation, or even deletions of the gene, were found to continue to sporulate, although their sporulation was delayed by several hours (Perego et al., 1989). This suggested that at least one other kinase must be able to carry out the KinA reaction, and also hinted why kinase deficiencies were not found among spo0 mutants because of the duplicity of the kinases involved. A solution to the identity of another kinase involved in sporulation was provided by studies in which kinase genes were cloned o n the basis of their homology to the common features of all kinases of this type (Trach and Hoch, 1992). A locus was identified, namely kinB, coding for a 47,774 Da kinase, with typical carboxyl-domain homology to sensor kinases and an amino-terminal region consisting of six membrane-spanning regions (Trach and Hoch, 1992). This kinase is in a small operon with another gene, kupB, coding for a 14,600 Da protein, with no homology to either sensory kinases or response regulators. Inactivation of the kinB gene or the kupB gene alone does not lead to a significant sporulation defect, but either of these mutations in combination with a kinA mutation lowers the residual sporulation frequency of a kinA mutant to almost zero. Thus, kinA and

SPORLII.ATION IN BACILLUS SUBTILIS

117

kinB represent two pathways for phosphorylation of either spoUFor spo0A. Unfortunately, the membrane location of the KinB protein has not allowed in uitro studies of the activity of this enzyme on the SpoOF or SpoOA proteins; however, from genetic results, it is clear that KinB cannot use SpoOA as a substrate for phosphorylation. This conclusion comes from several experiments. If the kinA gene is placed on a multicopy plasmid in spoUF and spoOA strains, and the product overproduced, the sporulation phenotype is suppressed in a spoOF strain but not in a spoUA strain, suggesting strongly that KinA, when overproduced, can phosphorylate the SpoOA protein. In contrast, KinB overproduction cannot suppress a AspoOF mutation, indicating that SpoOF is an obligate part of the pathway from KinB to SpoOA. This does not prove that the substrate for KinB is SpoOF, but it gives a strong indication that this might indeed be the case. If KinA and KinB act directly on spoOF, how does SpoOA become phosphorylated to activate transcription? This enigma was resolved by in vitro studies of phosphate transfer among the spo0-gene products (Burbulys et af., 1991). Gene products for the kinA, spoOA, spoOF and spoOB genes were overproduced in expression vectors and purified. These purified products were used to determine whether phosphorylation of SpoOA could be obtained by a combination of gene products. It was known that KinA was capable of transferring a phosphate group from ATP to the SpoOF protein, resulting in SpoOF-P (Perego et al., 1989). In the presence of SpoOA, this reaction proceeded unabated whereas, when SpoOA and SpoOB were added to such a reaction mixture, a phosphate group was transferred from SpoOF-P to SpoOA, resulting in SpoOA-P (Burbulys et af., 1991). These studies revealed that the enzymic function of the spoUB-gene product was simply to facilitate phosphate transfer between these two response regulators. Although SpoOB has no homology to kinases, it can carry out the phosphotransferase reaction characteristic of such kinases as long as the phosphate group is presented in the form of SpoOF-P. These reactions are summarized in Fig. 4. The amino-acid residue involved in autophosphorylation of sensory kinases is a histidine residue, and the properties of the phosphate on KinA are those of a phosphoramidate, which is consistent with this interpretation for KinA (Burbulys et af., 1991). This phosphate group is transferred to SpoOF, where it forms a mixed anhydride with an asFartic acid residue, probably Asp,,. Transfer of a phosphate group from SpoOF-P to SpoOA occurs through a SpoOB enzyme-bound intermediate with the properties of a histidine phosphoramidate. This group is then transferred directly to SpoOA, again phosphorylating an aspartic acid residue, in this case Asp,, (Burbulys et af., 1991). This series of reactions is called a phosphorelay and is unique from other two-component response regulator pairs in that

118

J A IIOCII

S/gna/ Transduction KinA

ATP

t

-b

ADP

t

KinA-P

(hisdo5)

KinA

t

OF-P

(asp 54)

Phosphotransfer KinA-P

OF

t

Phosphotransfer OF-P

t

06

OF +------

t

06-P

(his?)

OA

L OB +..----

t

OA-P

(asp56)

Phosphofransfer OB-P

t

FIG. 4. Enzymic reactions of the phosphorelay. The initial phosphorylation event is autophosphorylation of the kinase, KinA, brought about by activation with an unknown effector molecule. The phosphorylated KinA, KinA-P, transfers its phosphate group to OF, the product of the spo0Fgene. Phosphorylated OF, OF-P. is the substrate for the spo0R-gene product OB,which transfers phosphate to SpoOA (OA) via the enzyme intermediate OB-P. The amino-acid residues in parenthcses indicate the locations of the phosphate group on the specific residue of the product of each reaction. In thc case of OB-P the specific histidine is not known.

the SpoOB reaction has never been described in any other system. The rationale for this increased complexity of the signal-transduction system to activate sporulation compared with that of comparable pathways resides in the enormity of the cellular commitment to sporulation compared with that of simply activating a pathway. Thus, it has been postulated that the phosphorelay allows more levels of control than a two-component regulatory system (Burbulys et a l . , 1991).

c.

ROLES OF THE OTHER

spo0 genes

The original genetic studies identified more sp00 genes than could be accounted for in the phosphorelay. The spo0H gene, when cloned and sequenced, was found to code for a 27,447 Da protein with high homology to the a-factors of the RNA polymerase transcription complex (Dubnau et a l . , 1988). Transcription of accessory functions in Bacillus subtilis is highly dependent upon the activity of alternate a-factors. For example,

SPORULATION IN BACILLUS SUBTILIS

119

genes for flagellar biosynthesis and chemotaxis are controlled by the minor o-factor oD (Marquez et al., 1990; Mire1 et al., 1992). Genes requiring the o-factor for transcription are dependent upon the presence or absence of oD,which may form a second level of control on their expression. Similarly, the spoOH gene codes for oH,which is an essential sporulation gene required for the transcription of many genes, including those for the phosphorelay , spoOA, spoOF and kinA (Predich et al., 1992), and for genes involved in stage I1 of phosphorylation, including spollA (Wu et al., 1991). Regulation of spoOH transcription itself provides another level of control on these genes (Weir et al., 1991). In addition to sporulation genes, oH is required for transcription of other genes, such as citC, which are not required for sporulation as such, but whose expression is required during the transition state (Price et al., 1989). The spoOJ gene has been described and mapped very near to the origin of replication of the chromosome (Mysliwiec et al., 1991). Recently, it has been found that the spoOJ locus consists of a pair of overlapping genes, namely 0 r - 8 2 and orf253, homologous to pairs of genes such as korBincC, which are implicated in control of chromosome segregation during septation (Ogasawara and Yoshikawa, 1992). This exciting result shows a connection between DNA synthesis, septation and sporulation, and it will be of particular interest to determine in detail the functions of the proteins from the spoOJ locus. The spoOE locus has been found to code for a protein of 9791 Da. The spoOE gene is involved in some negative regulatory function and the phenotype of this locus is described later in this review (Perego and Hoch, 1991). The spoOK mutation has been found to reside within an operon of genes highly homologous to the opp operon of Salmonella typhirnuriurn (Perego et a l . , 1991; Rudner et al., 1991). The function of this operon is in uptake of oligopeptides of five amino acids or less. The operon consists of the oppA gene, coding for a lipoprotein on the external portion of the membrane, the oppB and oppC genes coding for membrane-spanning proteins, and the oppD and oppF genes both coding for ATP-binding domains presumably used to energize transport of peptides across the membrane. Curiously, mutation in only the first four genes results in a sporulation-defective phenotype, whereas deletion of the oppF protein does not result in sporulation deficiency. These oppF mutants, however, are deficient in competence (Rudner et al., 1991). It has been postulated that the role of the oligopeptide permease in peptide transport in competence is the transport of surfactin across the membrane to allow activation of competence genes (Hahn and Dubnau, 1991). The sporulation defect in opp mutants suggests that transport of some peptide from the

120

J A

11OCI~

external environment, perhaps produced extracellularly by the cell, may have some specific role in the activation of sporulation genes (Grossman and Losick, 1988). Whether this peptide acts directly on the phosphorelay or on some component that controls the phosphorelay is unclear; moreover, the nature of the peptide is still unknown. A dissident group believes that the o p p system serves to transport cell-wall peptides resulting from turnover and, by this means, communicates external structural information to the cytoplasm (Perego et al., 1991). Ill. Control of the Phosphorelay The role of the phosphorelay is to phosphorylate the SpoOA transcription factor. Since it appears that sporulation is a direct response to the cellular level of phosphorylated SpoOA, the pathway that produces it is subject to a surfeit of controls. These controls operate by regulation of the flow of phosphate through the phosphorelay and by repression of transcription of genes for the components of the phosphorelay. A . CONTROL OF PHOSPHATE FLOW

The initial source of phosphate for the phosphorelay arises from two kinases, namely KinA and KinB (Fig. 5 ) . These kinases are thought to be activated by different effector ligands presumably representing input signals from environmental stimuli or from metabolic pathways. However, no positive effectors have been identified for either kinase. One of the kinases, KinA, is inhibited by cis unsaturated fatty acids, whereas the homologous trans isomers are not inhibitory and saturated straight-chain or branchedchain fatty acids have little effect (Strauch el al., 1992b). The physiological significance of inhibition of this kinase by these fatty acids is open to speculation. It has been postulated that these rare fatty acids are involved with a structure that may have a specific spacial configuration in the cell. That is, such fatty acids may be part of a complex in which the KinA resides, or they could represent some metabolic signal which links sporulation to the status of membrane biosynthesis. Nothing is known of the nature of the effector molecules that influence activity of the KinB protein, although its membrane location might suggest that, it is involved in sensing external ligands, or perhaps in transport or other membrane-related phenomena (Trach and Hoch, 1992). There is no portion of the molecule that extends for any given length outside the membrane. This configuration suggests that the protein is not sensing an external ligand. This suggestion would be consistent with the potential role

121

SPORULA’IION IN BACILLUS SUBTILIS

Surfactin

4-

OB -P OA

-Obg? OA-P

Transcription Activator and Repressor

5’TGNCGAA31 1 1 0 ~

BOX”

FIG. 5. Schematic diagram showing known reactions and signal inputs into the phosphorelay. Phosphorylation of OF occurs by two kinases, KinA and KinB. The activity of KinB depends on the kapB-gene product, KapB, OK is the spoOK operon which codes for the oligopeptide permease, whose function is to transport peptides, and this activity is required for sporulation to initiate at high frequency. An essential G-protein, Obg, is postulated to affect the enzymic activity of OB. The activated transcription factor, OA-P, binds to promoters containing an “OA box”, whose nucleotide sequence is shown. Other symbols are as in Fig. 4.

of KapB, which appears to be essential for activity of KinB. The kinase KapB is a moderately charged protein with all of the characteristics of a soluble cytoplasmic protein. It is possible that it is the actual effector ligandbinding domain and that the effector molecule is cytoplasmic. This is one interpretation of the results. It is also possible that KapB is required for expression of the kin B operon, an interpretation which would certainly satisfy presently available results. Thus, the status of information on effector ligands for these kinases is meagre. Enzymic activity of KinA shows some unusual properties in vitro (C. E. Grimshaw, C. Hanstein, J. Grimsley, J. A . Hoch and J . M. Whitely, unpublished observation). The autophosphorylation activity of the kinase is stimulated by the presence of the SpoOF protein, suggesting that autophosphorylation in the absence of SpoOF is very slow. It is possible that it is simply the presence of SpoOF that activates the kinase to function. Thus, the presence of an effector molecule may not be required. Rather,

122

J

A

IIOCII

transcription of the spoOF gene and the increased cellular concentration of spoOF could be sufficient to activate the phosphorelay. Sensor kinases, in addition to phosphorylating a response regulator, also dephosphorylate such proteins. This phosphatase activity is not a simple reversal of the reaction, but is rather a hydrolysis reaction that produces inorganic phosphate. It is known that KinA has an overt phosphatase activity for SpoOF-P, although it has not been well characterized (C. E. Grimshaw, C. Hanstein, J . Grimley, J . A. Hoch and J . M. Whitely, unpublished results). It seems possible that specific effector ligands could be responsible for activating the phosphatase activity of KinA, particularly those ligands which play a negative role in control of this pathway. Although the SpoOF protein resembles CheY in size and structure, its phosphorylated form is highly stable in vitro (C. E. Grimshaw, C. Hanstein, J . Grimley, J. A. Hoch and J. M. Whitely, unpublished results; J . Cavanagh, N . Skelton, T . Tucker, J . A. Hoch and J . M. Whitely, unpublished results). This contrasts with CheY-P, which has a very short half-life, presumably due to a phosphatase activity inherent to the response regulator itself (Munoz el al., 1978). The protein SpoOF may also be the substrate for kinases other than KinA and KinB, but these additional kinases are not active enough under normal laboratory conditions to produce sufficient SpoOA-P for significant sporulation to occur. The SpoOA protein is not subject to phosphorylation by other kinases. A sensitive in vivo assay for SpoOA-P involves repression of abrB transcription. Using this assay, it has been found that mutations in either the spoOF or spoUR genes result in constitutive transcription of the abrB gene, suggesting strongly that very little, if any, SpoOA-P is formed in the absence of phosphorelay components in vivo (Trach and Hoch, 1993). Thus, cross-talk to SpoOA by heterologous kinases is not a significant producer of SpoOA-P. It is known that mutations in the spoOA gene can result in by-pass mutations for both the SpoOF or SpoOB proteins (Kawamura and Saito, 1983; Hoch et al., 1985). These mutations presumably result in an activated SpoOA that no longer requires phosphorylation for activity or modifies the specificity of the protein such that heterologous kinases can now act directly on the SpoOA protein itself (Spiegelman et a l . , 1990). Although the SpoOB protein is a phosphoprotein phosphotransferase that carries out the same reaction as a kinase, there is n o in vitro evidence for a phosphatase activity being an inherent property of the SpoOB enzyme. It is possible, however, that this could represent another level of control on the phosphorelay, which brings into question the role of the obg-gene product whose gene is in the same transcript as the spoUB gene (Trach and Hoch, 1989). This GTP-binding protein could affect activity of the spoOB

SPORUI ATION IN RACII.L.US SURTILIY

123

gene either by promoting or inhibiting its forward or reverse reactions, and perhaps activating or inhibiting a phosphatase-like activity of SpoOB. However, there is at present n o evidence to link obg to spoOB except the circumstantial happenstance that they are in the same transcription. The SpoOE protein plays a negative role in the phosphorelay. When the spoOE gene is deleted, cells appear to sporulate normally, but such strains accumulate mutations in one or more of the components of the phosphorelay, indicating that deletion of the spoOE gene results in overexpression of the phosphorelay and sporulation at inappropriate times (Perego and Hoch, 1991). Deletion of the spoOE gene also suppresses the sporulation defect of many of the SPOOFmis-sense mutations, but not of deletions of the SPOOFgene (J. A. Hoch, unpublished observation). This sparing affect can be explained if most of the spoOF mis-sense mutations result in an unstable SpoOF-P, which can be relieved by deletion of the spoOE gene. The easiest conclusion to come to from these results is that SpoOE is somehow involved in negative regulation of t h e phosphorelay and, when its activity is removed, even low levels of SpoOF-P are capable of being transferred to SpoOA to activate sporulation. Since carboxyltruncated spoOE proteins lead to a sporulation-defective phenotype (Perego and Hoch, 1987), the suggestion has been made that the carboxyl-terminal portion of the spo0E-gene product is a regulating domain for this protein, and that its negative activity can be controlled except in carboxyl-truncated non-sense fragments where a constitutive negative reaction must occur (Perego and Hoch, 1991). There are many points where restriction of information flow through the phosphorelay could occur and in some cases does. The effector ligands of such negative activity are yet to be discovered. H. TRANSCRIPTIONAL REGULATION OF GENES FOR PHOSPHORELAY

COMPONENTS

Transcriptional repression of certain components of the phosphorelay is a highly effective means of controlling flow of phosphate through the phosphorelay and, ultimately, production of SpoOA-P. Several growthpromoting conditions such as an excess of catabolites, bringing about catabolite repression and perhaps transition state regulators, work through this mechanism. Transcriptional control falls into two general categories. First, there are those genes that are little affected by different growth conditions such as spoOB (Ferrari et al., 1985b) and, secondly, those which show a moderate-to-strong induction at the end of exponential growth such as spoOE, SPOOF,kinA and spoOA (Perego and Hoch, 1987; Yamashita et al., 1986; Antoniewski et a l . , 1990). Of these latter genes, three of them,

124 Pv

Pv

J . A . tiorti

kinB

I

1

spoOB

obg

I

KinB

Pv A

t

OF

6 +

0

@

Ps I

1

H ATP

KinA

kapB

kinA

@

SPOOF

1

ADP

I

OB

u,OF-P

spoOA

1

OA-P

00

6

Pv A

spoOE

I

1

1

FIG. 6. Transcriptional interactions among genes for the components of the phosphorelay. Pv and Ps are vegetative and sporulation promoters. The black boxes are binding sitcs for SpoOA-P. The arrows from OA-P indicate sites of DNA interaction of the protein and promoters. The arrows from the abrB gene indicate promoters repressed by AbrB. @ and 8 indicate effects of AbrB or OA-P on transcription. Sporulation promoters, Ps, require the a-factor 0''. Other symbols are as in Figs 4 and 5.

namely kinA, spoUF and spoUA , are dependent on the product of the spoUH gene, oH,for induction at the end of growth (Predich et al., 1992). Although regulation of transcription of these components is very complex, in general the entire transcription regulatory scheme can be viewed as a complex system of autoregulation, where SpoOA-P is the product and the major controller (Fig. 6). A low level of the basic components of the phosphorelay, including SpoOF, SpoOB and SpoOA, is maintained through all phases of growth under all conditions by transcription from vegetative promoters. Thus, the bacterium is assured of the ability to respond to environmental conditions by maintaining the phosphorelay at this level. Both the spoOF and the spoOA genes are transcribed from tandem promoters with one vegetative promoter, G * , transcribing constitutively, and a sporulation promoter, oH, which is used for induction at the end of growth (Lewandoski et al., 1986; Yamashita et al., 1989). Under conditions conducive to vigorous growth, the level of SpoOA-P is very low. This allows high-level production of the

125

SPOHULATlON IN HA( ll.l.US SUHTIL.1.Y

repressor AbrB, which in turn represses synthesis of the spoOH and spoOE genes. Since induction of several key phosphorelay components requires crH, which is kept in check by the abrB repressor, the level of oHis controlled by the cellular level of SpoOA-P. When conditions degenerate and significant SpoOA-P levels begin to accumulate in the cell, the most sensitive promoter to these rising levels is the abrB promoter. This may be due to the presence of tandem OA boxes or OA-binding sites in this promoter (Strauch el al., 1990). Lowering of the levels of AbrB results in release from repression of the spoOH and spoOE genes. Production of crH allows high-level transcription of the spoOF gene, and concomitant induction of the spoOA gene from a oHpromoter. In addition, kinA is transcribed by cr" and its level in the cell rises (Predich el al., 1992). The gene product SpoOA-P is also a positive inducer of its own synthesis, causing both repression from the spoOA vegetative promoter and induction from the sporulation promoter (Strauch er al., 1992a). The end result of all this activity is to increase the relative amount of the components of the phosphorelay in the cell, which increases the cellular concentration of SpoOA-P. When this level reaches some critical concentration, SpoOA-P may repress further transcription from these genes (Strauch et al., 1992a). In the kinA gene, there is a SpoOA-binding site located just downstream from the transcription-start site, suggesting that SpoOA-P interacts at this site to turn off transcription (K. A. Trach, M. Strauch and J . A . Hoch, unpublished observation). Both the spoOF and the s p d A genes have more than one SpoOA-P-binding site. In spoOF (Fig. 7), the upstream binding site for spoOA serves as a means for this protein

-

I

GTATACAACAAAAGAGAAAATGCTCAGAAAA~GTCGTA~A AccI

P2+ I

GTAGACTATTATAATTAAAGGAAAAATCAAACAG Accl

P1-b I

I

I I

AATACATACAATACTGCTTACTTT~TGACGAAATCATAAT

* *

I

ATTGGGGTGTAAAATGATGAATGAAAAAA~TTTAATCGTT

M

M

N

E

K

l

L

l

V

FIG. 7 . Interaction of SpoOA with the spoOF promoter. This figure shows the 5'3' nucleotide sequence of the promoter region upstream of the spoOF gene and the initial translation product of the spoOF gene. The boxed nucleotides indicate the location of consensus OA-binding sites. Overbars show the extent of SpoOAprotected regions in DNasel footprints. PI and P2 are sporulation and vegetative promoters for SPOOF.

126

J A tlOCH

to enhance positively transcription from the aH promoter, whereas the downstream SpoOA-binding site appears to modulate this activity (Strauch et al., 1993). Thus, if the upstream binding site is deleted, induction does not occur whereas, when the downstream binding site is removed, there is an overproduction of SpoOF. A similar situation probably occurs in the spoUA promoter where there are both upstream SpoOA-binding sites and a binding site covering the -35 region of the a" promoter, which could be a repression site (Strauch et al., 1992a). This regulation scheme lacks many of the details that influence it under a variety of conditions. Catabolite repression has a strong influence on the spoUA aH promoter, preventing switchover from the vegetative promoter to the sporulation promoter (Yamashita et a l . , 1989; Chibazakura et a l . , 1991). Whether this occurs because of specific regulatory proteins binding at this site and preventing the switch has not been conclusively determined (Weickert and Chambliss, 1990). In addition, other regulatory proteins may play a role in control of these genes. We really do not know exactly how all of the transition-state regulators, such as AbrB, Hpr, Sin and others, work to control both transcription and flow of phosphate through this pathway, and it would not be surprising if one or more, o r others not yet discovered, would have a repressive or inactivating role on transcription of one or more of the genes of this phosphorelay. A common theme, however, is that several crucial components of the phosphorelay have promoters containing SpoOA-P-binding sites, and that these play a critical role in transcription of such promoters, suggesting that autoregulation by SpoOA-P is very important in transcription regulation.

IV. Transition-State Regulators There are several regulatory proteins in aggregate termed transition-state regulators that control expression of a large number of genes normally turned on during the transition state between exponential growth and the stationary phase of growth (Strauch and Hoch, 1993). Transition-state regulators are negative regulatory proteins that prevent transcription of the genes in question and cover a broad spectrum of genes for which individual gene products may have no obvious relationship to each other. There are no known effector ligands that control the activity of these regulators, but they certainly may exist, at least for some of them. Such regulators are of interest in this context because their negative regulation affects the ability of the cell to sporulate and, therefore, they must impinge somewhere in the phosphorelay to affect this control. We shall concentrate in this short review of their activities on three of the most

SPORUI.ATION IN BACILLUS SUBTILIS

127

well characterized of this class of regulators, the AbrB, Hpr and Sin regulatory proteins. A. THE

AbrB

PROTEIN

Mutations in the spoOA gene result in a wide variety of pleiotropic negative phenotypes for many different functions such as proteases and antibiotics normally expressed during the transition state. These phenotypes result from genes that are sensitive to AbrB regulation and remain repressed in a spoOA mutant. Transcription of the abrB gene is controlled by SpoOA-P such that, in its absence, the ahrB gene is constitutively expressed (Strauch el af., 1989a). The AbrB protein is a negative regulator of transcription that binds to a wide variety of promoters and prevents their expression (Zuber and Losick, 1987; Strauch et al., 1989b). It prevents transition-state genes being expressed during the exponential phase of growth, and does this in response to the level of SpoOA-P in the cell. Thus, its major role is not to regulate genes but to prevent their expression under conditions when most of the energy of the cell is directed toward growth and division. In this role, it serves as a “preventer” rather than as a repressor since most of the genes that it prevents from being transcribed have other regulatory controls that act as classical repressors. The AbrB protein can be thought of as a regulator that is unique to exponential growth and helps to promote exponential growth by preventing unnecessary protein synthesis for accessory functions occurring during this phase of rapid growth and division. The AbrB protein carries out its function by binding to promoters. In v i m footprint analyses of promoters controlled by this gene product show large regions of protection, and no apparent consensus binding site could be recognized from the primary sequence of the regions protected (Strauch et al., 1989b). One possibility for the lack of a common binding site is that promoters controlled by AbrB have a unique secondary or tertiary structure which can be generated by a wide variety of primary nucleic-acid sequences, and it is structure rather than sequence which is recognized by the AbrB regulator. The AbrB protein binds to some promoters that have important roles in the sporulation process, such as the spoOH promoter (Weir et af., 1991), and it binds to other promoters, such as aprE, whose gene product has no functional role in the sporulation process (Strauch et af., 1989b). Thus, in this context it is related to sporulation only in so much as some of the genes that it controls have effects on sporulation. There is no known effector ligand which controls activity of this regulatory protein, and it appears that the only control that the cell has on its activity is raising and lowering its concentration. Depending on the turnover rate of the protein, this could be a fast or a sluggish response to changing physiological

128

J A tlO(’H

conditions. However, the ahrB promoter is the most exquisitely sensitive promoter to low concentrations of SpoOA-P and, therefore, it may be shut down very early before turnoff of exponential growth is obvious. H . THE

Hpr

PROTEIN

Another transition-state regulator of some interest is the product of the hpr gene. This gene codes for a regulator first recognized as controlling production of subtilisin, and in which mutations give rise to proteaseoverproducing strains (Higerd et a f . , 1972). When the gene was cloned it was found that mutations causing excessive production of protease resulted in loss of the hpr-gene product, indicating that Hpr serves as a negative regulator (Perego and Hoch, 1988). This gene was also found to be the location of mutations originally described as cut, which relieve catabolite repression of sporulation under some conditions (Ito and Spizizen, 1973). Therefore, loss of the gene appears to have some role in catabolite repression, although this phenotype o r its basis has not been well characterized. When placed on a multicopy plasmid, the hpr gene results in a sporulation-defective phenotype, suggesting that there is at least one gene that Hpr controls which has a direct effect on the sporulation process (Perego and Hoch, 1988). The Hpr protein binds to promoters in vitro and, of those that have been tested, the binding sites appear to be found in multiples o f two (Kallio et a f . , 1991). The promoter at which Hpr binds to effect sporulation is unknown. Recently, it has been found that the Hpr protein binds to the promoter for the sinlgene (Kallio et a / .,1991), and may prevent production of the inhibitor for the action of Sin, a negative regulatory protein for sporulation. This could account for the sporulation-defective phenotype encountered when the gene product is overproduced, although no experiments have been carried out to test this possibility. There is no known effector ligand for Hpr that controls its activity and, therefore, the physiological status of the cell that causes Hpr to function is a mystery.

c.

THE

Sin

PROTEIN

A very interesting transition-state regulator is found in the product of the sin gene (Gaur et a / . , 1988). This protein, when overexpressed, inhibits sporulation and also the production of proteases (Gaur et al., 1986). Deletion of the sin gene results in a filamentous non-motile autolysinnegative cell which forms a colony with an extremely rough appearance in certain backgrounds. Thus, sin regulation has both positive and negative effects on many processes, some of which are involved in sporulation. In

SPORULATION

IN RAC'ILLUS .SURTII.I.S

129

addition, strains with sin deletions tend to overexpress spoll genes and, therefore, sin may affect sporulation by controlling some aspect of either their expression or of the phosphorelay. The purified sin-gene product is a DNA-binding protein that binds to the upstream region of the subtilisin promoter (Gaur et al., 1991). Most interestingly, there is a gene upstream of the sin gene, termed sinl, which acts as an inhibitor of the function of sin. This protein, known as Isin, probably complexes with sin to inhibit its activity (U. Bai, I. MandicMulec and I. Smith, unpublished observation). Transcription of I-sin is independent of the transcription of sin, and may be controlled by the Hpr, AbrB and SpoOA proteins (Gaur et al., 1988; Kallio et al., 1991; M. A. Strauch, unpublished observation). Thus, there is a complex interplay between sin and I-sin, and transcriptional regulation of sporulation genes. The transition-state regulators AbrB, Hpr and Sin work by controlling transcription of genes whose products promote sporulation. What metabolic or environmental factors are regulating the activity of these proteins? Certainly these factors which remain unknown must be conducive to growth or there would be no reason to inhibit sporulation.

V. Alternatives to Sporulation

Although sporulation is probably the most interesting morphological process of stationary phase bacteria, it is by no means the only process that can occur, nor is it inevitable. The transition-state regulators already described, as well as the phosphorelay, are an integrated package producing a global regulatory network that ultimately can control sporulation as part of its function. But they also can lead to other physiological states, such as competence, which can be viewed as an alternative to sporulation. Competence was defined originally as the state of a culture that was capable of taking up and integrating transforming DNA. Recently, it has become clear that formation of the competent state is in itself and interesting physiological process involving a cascade of regulatory genes (Dubnau, 1991b). Some of these regulatory functions are common to sporulation while others are not. This complicated pathway of genetic dependency has been reviewed recently and will not be explained in detail here (Dubnau, 1991a). The major concept that is evolving from studies of both sporulation and competence is that neither of the two processes is entirely exclusive of the other, and that they represent alternative pathways to different physiological states which may depend directly on the nutritional status of the culture at the time of entering the stationary phase of growth.

130

J . A . HOCH

Sporulation is clearly a last resort for the cell since, as already indicated, a cell would rather grow and divide to compete than spend this enormous amount of energy to produce spores. Competence may be part of an extended transition state in which the cell has sufficient nutritional requirements to maintain itself in the absence of sporulation but not enough to initiate growth. VI. Initiation of Sporulation

Conversion of a culture from exponential growth to the early stationary phase of growth is associated with production of low levels of SpoOA-P, release of AbrB repression and functioning of a large number of regulators that push the cell towards a number of different physiological states to take advantage of any residual nutrients in the medium. After having run this gauntlet of transition-state regulators and effector ligands that prevent onset of sporulation, the cell must produce higher levels of SpoOA-P in order to activate transcription of those genes that will carry it into the morphogenetic process that characterizes sporulation. The targets of this increased SpoOA-P level are the spoll genes, spollA (Trach et al., 1991), spolIE (York et a l . , 1992) and spollC (Satola et a l . , 1992), whose transcription requires positive activation by SpoOA-P. Promoters for these three genes contain SpoOA-binding sites, and SpoOA binds to these promoters in vitro. Only the spoIlA operon requires oH,whereas spollE and spoflc are both transcribed from a oA or vegetative promoter. The gene product SpoOA-P is thought to promote binding of the transcription complex to these promoters and probably serves to stabilize the ternary complex of RNA polymerase, SpoOA-P and DNA required for efficient initiation at these promoters (Satola et al., 1992; York el al., 1992). Regardless of the precise mechanism, the presence of sufficient SpoOA-P is essential, and the level is presumably much higher than that required for repression of ahrB by SpoOA-P. All of the available evidence suggests that it is only the cellular content of SpoOA-P which is the factor that determines whether spoll genes will be activated. Thus, there is little evidence that additional positive or negative regulators are required for spoll gene transcription, although it is possible that they exist and have not yet been discovered. Transcription o f the spollA and spoIIC operons represents the first stage of the so-called o-cascade (Stragier and Losick, 1990) in which different o-factors are used for transcription of genes in the two cellular compartments that characterize a sporulating cell, namely the forespore and the mother cell (Fig. 8). The spoffA operon codes for the o-factor 0’ (Stragier,

131

SPORUI ATION IN HACII.I.U.5 SUHTII.I.5

IIA

H 0-

f

'r,

I

I

- F

0-

----

I

FIG. 8. Diagram showing the relationship between SpoOA-P and compartmentalized gene expression during sporulation. The phosphorylated form of SpoOA, OA-P, is a transcription activator of the spolZA (HA) and spoIIC (IIG) operons aided by the o-factors oH and aA.One product of each of these operons is a ofactor used exclusively for transcription of genes whose products are used in the developing forespore (aF)or the mother cell (oE).

1986) that is required to produce the o-factor aG(Sun et al., 1991), which is exclusively used in the transcription of forespore genes (Nicholson et al., 1989; Sun el al., 1989). Transcription of the spoIfC operon activates synthesis of a p r o d a-factor, which, when cleaved to an active a-factor by the morphological event of sporulation septum formation, results in a o-factor whose activity is exclusively confined to the mother cell (Stragier et al., 1988). This exciting role for a-factors in compartmentalization of transcription is certainly one of the more interesting and unique events occurring during the sporulation process. Since this is an area of high investigative activity, the results of investigations to uncover the mechanism of the temporal control of gene activity during sporulation would require another chapter to this volume. In retrospect, when one looks at the many regulatory functions that exist to prevent sporulation, one wonders how a cell sporulates at all. Perhaps this results from our ignorance of the effector molecules that control all of the various regulatory proteins, for there may be a very simple pattern that underlies this seeming complexity. It will certainly be of interest to determine if a simple solution exists, or whether the more we investigate this system the more complex it becomes.

132

J. A. HOCII

VII. Acknowledgements

This research was supported by grant GM19416 from the National Institutes of General Medical Sciences, National Institutes of Health, US Public Health Service. This is manuscript 7625-MEM from the Department of Molecular and Experimental Medicine. REFERENCES

Antoniewski, C., Savelli, B. and Stragier, P. ( 1 W ) . Journal of Bacteriology 172, 86. Bouvier, J., Stragier, P., Bonamy, C. and Szulmajster, J. (1984). Proceedings ofthe National Academy of Sciences of the United States of America 81, 7012. Burbulys, D., Trach, K. A. and Hoch, J. A. (1991). Cell 64, 545. Chibazakura, T., Kawamura, F. and Takahashi, H.(1991). Journal of Bacteriology 173,2625. Dubnau, D. (1991a). Microbiological Reviews 55. 395. Dubnau, D. (1991b). Molecular Micro6iology 5, 11. Dubnau, E., Weir, J., Nair, G., Carter, L., Ill, Moran, C., Jr and Smith, I. (1988). Journal of Bacteriology 170. 1054. Ferrari, F.A., Trach, K., LeCoq, D., Spence, J., Ferrari, E. and Hoch, J. A. (1985a). Proceedings of the National Academy of Sciences of the United States of America 82,2647. Ferrari, F.A., Trach, K. and Hoch, J. A. (1985b). Journal of Bacteriology 161, 556. Gaur, N. K., Cabane, K. and Smith 1. (1988). Journal of Bacteriology 170, 1046. Gaur. N. K., Dubnau, E. and Smith, 1. (1986). Journal of Bacteriology 168, 860. Gaur, N. K., Oppenheim, J. and Smith, 1. (1991). Journal of Bacteriology 173. 678. Grossman. A. D. and Losick, R. (1988). Proceedings of the National Academy of Sciences of the United States of America 85, 4369. Hahn, J. and Dubnau, D. (1991). Journal of Bacteriology 173, 7275. Higerd, T. B., Hoch, J. A. and Spizizen, J. (1972). Journal of Bacteriology 112, 1026. Hoch, J. A . (1976). Advances in Generics 18, 69. Hoch, J. A , , Trach, K., Kawamura, F. and Saito, H. (1985). JournalofBacteriology 161,552. Ito, J. and Spizizen, J. (1973). Colloquim International CNRS 227, 81. Kallio, P. T., Fagelson, J. E., Hoch. J. A. and Strauch, M. A. (1991). Journal of Biological Chemistry 266, 1341 I . Kawamura, F. and Saito, H. (1983). MolGenGenet 192, 330. Kudoh, J., Ikeuchi, T . and Kurahashi. K. (1985). Proceedings of the National Academy of Sciences of the United States of America 82, 2665. Lewandoski, M., Dubnau, E. and Smith, I. (1986). Journal of Bacteriology 168, 870. Mandelstam, J. and Higgs, S. A. (1974). Journal of Bacteriology 120, 38. Marquez, L. M., Helmann, J. D., Ferrari, E., Parker, H. M.,Ordal, G. W. and Chamberlin, M. J. (1990). Journal of Bacteriology 172, 3435. Mirel, D . B.. Lustre, V. M. and Chamberlin, M. J. (1992). JournalofBacteriology 174,4197. Munoz, L. E., Nakayama, T. and Doi, R. H. (1978). In “SporesVII”, pp. 213-219. American Society of Microbiology, Washington, DC. Mysliwiec, T . H., Errington, J.. Vaidya, A. B. and Bramucci, M. G. (1991). Journal of Bacteriology 173, 191 1. Nicholson, W. L., Sun, D., Setlow, B. and Setlow, P. (1989). Journal of Bacteriology 171. 2708.

Ogasawara, N. and Yoshikawa. H. (19Y2). Molecular Microbiology 6, 629. Parkinson, J. S. and Kofoid, E. C. (1992). Annual Review of Genetics 26. Perego, M. and Hoch, J. A. (1987). Molecular Microbiology 1, 125. Perego, M. and Hoch, J . A. (1988). Journal of Bacteriology 170, 2560. Perego, M. and Hoch, J. A. (19Y1). Journal of Bacteriology 173, 2514.

SPORUI.ATION IN RAClLLUS SURTlLlS

133

Perego, M, Cole, S. P., Burbulys, D., Trach, K. and Hoch, J. A . (1989). Journal of Bacteriology 171, 6187. Perego, M., Higgins, C. F., Pearce, S. R . , Gallagher, M. P. and Hoch, J. A. (1991). Molecular Microbiology 5, 173. Piggot, P. J. and Coote, J. G . (1976). Bacteriological Reviews 40,908. Predich, M., Nair, G. and Smith, I. (1992). Journal of Bacteriology 174, 2771. Price, V. A., Feavers, I. M. and Moir, A. (1989). Journal of Racteriology 171, 5933. Rudner, D. Z., Ladeaux, J. R., Breton, K. and Grossman, A. D . (1991). Journal of Bacteriology 173, 1388. Satola, S. W., Baldus, J. M. and Moran, C. P. Jr (1992). Journal of Bacteriology 174, 1448. Spiegelman, G . B., Van Hoy, B., Perego, M., Day, J., Trach, K. and Hoch, J. A. (1990). Journal of Bacteriology 172, 501 1. Stock, J . B., Ninfa. A . J. and Stock, A . M. (1989). Microbiological Reviews 53, 450. Stock, J. B., Stock, A . M. and Mottonen, J. M. (1990). Nature 344,395. Stragier, P. (1986). FEBS Letters 195, 9. Stragier. P. and Losick, R. (1990). Molecular Microbiology 4, 1801. Stragier. P., Bonamy, C. and Karmazyn Campelli, C. (1988). Cell 52, 697. Strauch, M. A. and Hoch, J. A . (1993). Molecular Microbiology, in press. Strauch. M. A., Perego, M., Burbulys, D. and Hoch, J. A . (1989a). Molecular Microbiology 3, 1203. Strauch, M. A., Spiegelman, G. B., Perego, M., Johnson, W. C., Burbulys, D. and Hoch, J . A. (1989b). EMBO Journal 8, 1615. Strauch, M., Webb, V.,Spiegelman, G. and Hoch, J. A. (1990). Proceedings ofthe National Academy of Sciences of the United States of America 87, 1801. Strauch, M. A., Trach, K., Day, J. and Hoch, J. A. (1992a). Biochimie 74, 619. Strauch, M. A., de Mendoza, D. and Hoch, J. A. (1992b). Molecular Microbiology 6,2909. Strauch, M. A,, Wu,J.-J., Jonas, R. H . and Hoch, J. A. (1993). Molecular Microbiology, in press. Sun, D., Stragier, P. and Setlow, P. (1989). Genes und Development 3 , 141. Sun, D., Cabrera-Martinez, R. M. and Setlow, P. (1991). Journalof Bacteriology 173, 2977. Trach, K. and Hoch, J. A. (1989). Journal of Bacteriology 171, 1362. Trach, K. A. and Hoch, J. A. (193). Molecular Microbiology 8, 69. Trach, K . , Chapman, J. W., Piggot, P. J. and Hoch, J. A. (1985). Proceedingsof the National Academy of Sciences of the United States of America 82, 7260. Trach. K., Chapman, J . W.,Piggot, P., LeCoq, D . and Hoch, J. A. (1988). Journal of Bacteriology 170, 4194. Trach, K., Burbulys, D . , Strauch, M., Wu, J.-J., Dhillon, N., Jonas, R., Hanstein, C., Kallio, P., Perego, M., Bird, T., Spiegelman, G., Fogher, C. and Hoch, J. A . (1991). Research in Microbiology 142, 815. Weickert, M. J. and Chambliss, G. H. (1990). Proceedingsofthe NationalAcademyofSciences of the United States of America 87, 6238. Weir, J . , Predich, M., Dubnau, E., Nair, G. and Smith, I. (1991). Journal of Bacteriology 173, 52 I . Wu, J.-J., Piggot, P. J., Tatti, K. M. and Moran, C. P., Jr (1991). Gene 101, 113. Yamashita, S., Yoshikawa, H., Kawamura, F., Takahashi, H., Yamamoto, T., Kobayashi, Y. and Saito. H. (1986). MolGenCenet a 5 , 28. Yamashita, S., Kawamura, F., Yoshikawa, H., Takahashi, H., Kobayashi, Y. and Saito, H. (1989). Journal of General Microbiology 135, 1335. York, K., Kenney, T. J., Satola, S., Moran, C. P. Jr, Poth, H. and Youngman, P. (1992). Journal of Bacteriology 174, 2648. Zuber, P. and Losick, R. (1987). Journal of Bacteriology 169, 2223.

This Page Intentionally Left Blank

Biosynthesis and Expression of Cell-Surface Polysaccharides in Gram-Negative Bacteria CHRIS WHITFIELD“ and MIGUEL A . VALVANOb

.

‘Depurtment of Microbiology. University of Guelph. Guelph Ontario Canada. NI G 2WI. and ’Department of Microbiology and Immunology. University of Western Ontario. London. Ontario. Canada. N6A SCI

.

List of abbreviations . . . . . . . . . . . . . I . Introduction . . . . . . . . . . . . . . . I I . Structure and attachment of cell-surface polysaccharides . . . . . A. Surface association of bacterial polysaccharides . . . . . . H. Repeating unit structuresin cell-surface polysaccharides . . . . C. Expression of multiplc cell-surface polysaccharides . . . . . . . . . . . . . . . . 111. Polysaccharidc biosynthesis . A. Formation of undecaprenol-linked intermediates . . . . . R. Polymerization reactions . . . . . . . . . . . C. Polysaccharide-modification reactions: addition of side-chains and su bst i t uen ts . . . . . . . . . . . . . . IV . Export o f polysaccharides and cell-surface assembly . . . . . . . A. Location of biosyntheticcomplexes at the cytoplasmic membrane . R . Transport across the cytoplasmic membrane . . . . . . . (.. Translocation from thc cytoplasmic membrane to the cell surface . . V . <;eneticsofpolysaccharidebiosynthesis . . . . . . . . . A. Housekeeping and polysaccharide-biosynthesisgenes . . . . . Gcneticdeterminants for 0-polysaccharide biosynthesis . . . . . Geneticsof biosynthesisof extracellular polysaccharide . . . . Ll . Relationships between multiple polysaccharide-biosynthesis gene . . . . . . . . . . . . . . clusters . E. . Molecular basis for antigenic variation in cell-surface polysaccharides . v1 Regulation o f cell-surface polysaccharide synthesis . . . . . . A. Regulationoflipopolysaccharidesynthcsis . . . . . . . B . Regulation ofextraccllular-polysaccharidesynthesis . . . . . VII . Conclusions . . . . . . . . . . . . . . . VIII . Acknowledgements . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . .

AIIVANCISIN M I C R O H I A I . P I I Y S I O I . O ~ ~ Y . V..O35 I ISBN ILI 2412’7735-2

.

136 136 13x 138 144 149

154 154 159 16X

171 171 175 I82 188

188 190 198 2M 207 212 212 216 229 230 231

Copyright0 lW3 by Academic Pressl.imited All nghtsof reproduction in any form reserved

136

C'

WI1ITlIEI.I) A N 0 M A V A I V A N 0

List of Abbreviations

Cyclic diguanylic acid Capsular polysaccharide Extracellular polysaccharide, exopol ysaccharide Enterobacterial common antigen 1.-Glycerophosphatidyl-linkedform of enterobacterial common antigen Lipid A-core attached form of enterobacterial ECApG common antigen Lipid A-core attached form of K-antigen KIPS 3-Deoxy-~-manno-octulosonic acid KDO Lipopolysaccharide LPS Open-reading frame ORF Rough form of lipopolysaccharide R-LPS Sodium dodecyl sulphate-polyacrylamide gel SDS-PAGE electrophoresis Smooth form of lipopolysaccharide s-LPS a-Glucosyldiphosphorylundecaprenol Undecaprenol-PP-Glc The nomenclature for polysaccharide structures used in this review is based on the 1980 recommendations of the IUPAC-IUB (IUPAC-IUB Joint Commission on Biochemical Nomenclature (JCBN) (1982). European Journal of Biochemistry 126, 439). c-di-GMP CPS EPS

I. Introduction The architecture of the Gram-negative cell surface is depicted in Fig. 1. The cytoplasm of the bacterial cell is surrounded by the cytoplasmic membrane. This lipid bilayer is permeable to water and some small hydrophobic compounds, and it contains the machinery for electron transport, oxidative phosphorylation and solute transport. Biosynthesis of many cell-surface components, including polysaccharides, involves enzymes and enzyme complexes found in the cytoplasmic membrane. The peptidoglycan layer is located immediately external to the cytoplasmic membrane, and this layer is required for cell shape and rigidity. Gramnegative bacteria possess a periplasm which contains a variety of proteins and enzymes, including some involved in import and export of macromolecules. Although often called a periplasmic space, the periplasrn is frequently filled with an electron-dense periplasmic gel (Hobot et al., 1984; Graham el al., 1991a). The periplasm is contained by the outer

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

137

Slime Polysaccharide

Capsular Polysaccharide (CPS)

Lipopolysaccharide (LPS)

-

-

Outer Membrane Periplasm PeptidoQlycan Cytoplasmic Membrane

FIG. 1. Diagram showing the cell-surface architecture of Gram-negative bacteria.

membrane, which is composed of proteins, phospholipids and a unique molecule, namely lipopolysaccharide (LPS) (Lugtenberg and VanAlphen, 1983; Nikaido and Vaara, 1985). The presence of LPS in the outer leaflet of the outer membrane results in an atypical lipid bilayer and can provide a hydrophilic boundary on the cell surface. In many bacteria from diverse ecological niches, surface hydrophilicity is augmented by a layer of extracellular polysaccharide (EPS). The EPS can take the form of an adherent (often covalently bound) cohesive layer, forming the morphological entity termed the capsule (capsular polysaccharide, CPS). Alternatively, EPS can consist of a slime polysaccharide, with little or no cell association. The term “extracellular polysaccharide” is used here to describe both capsular and slime polysaccharides. Although EPSs are not essential for survival in the laboratory, most invasive pathogens produce EPS, and it is clear that the polysaccharides are equally prevalent in soil and aquatic isolates. Consequently, in many bacteria, the cell surface is predominantly composed of polysaccharide; exceptions include bacteria which have a proteinaceous surface array or S-layer (Beveridge and Graham, 1991). By virtue of their location, cell-surface polysaccharides influence the manner in which bacterial cells interact with the environment. The observation that polysaccharides play critical roles in antigenicity

138

C. WHITE’IE1.D A N D M A . V A I . V A N 0

provided an important thrust to polysaccharide research. Early studies were confined to structural and immunochemical analyses, principally in animal and human pathogens. Interest in the chemical and physical properties of polysaccharides, and their biotechnological applications, has now considerably broadened the scope of this research. This is particularly evident in the increased interest in studies of cell-surface polysaccharides in plantassociated bacteria (Leigh and Coplin, 1992). Over the past few years, several excellent reviews have been published on general aspects of structure, immunochemistry, function and production of bacterial polysaccharides (Sutherland, 1977b, 1982, 1985, 1988; Kenne and Lindberg, 1983) and two texts have been devoted to this subject (Sutherland, 1977a; Jann and Jann, 1990a). Biochemical (Sutherland, 1977b, 1982; Troy, 1979; Jann and Jann, 1984; Sutherland, 1985; Shibaev, 1986; Rick, 1987; Raetz, 1990) and genetic (Makela and Stocker, 1984; Rick, 1987) aspects of polysaccharide biosynthesis have also been covered. In addition, there are available a number of focused reviews and minireviews which concentrate on specific bacteria. Application of molecular biology to studies of bacterialpolysaccharide synthesis has provided an important stimulus to this research area and has resulted in significant recent findings. The objective of this article is to present a contemporary overview of the molecular mechanisms involved in synthesis and expression of cell-surface polysaccharides in Gram-negative bacteria. We bring together results obtained by biochemical and genetic approaches, but we limit the discussion to systems where information is available concerning mechanisms of synthesis and expression. Space limitations preclude inclusion of the many recent reports of genes which play uncharacterized roles in polysaccharide biosynthesis. Although much of this review leans towards enteric bacteria, where possible we examine components and reactions in diverse bacteria in an attempt to identify common concepts in polysaccharide expression. Although the nature of the association between EPS and LPS O-polysaccharide and the cell surface can differ, it is our contention that this does not affect many aspects of their synthesis, and we therefore examine these polysaccharides in a comparative way. 11. Structure and Attachment of Cell-Surface Polysaccharides A. SURFACE ASSOCIATION OF BACTERIAL POLYSACCHARIDES

There are several forms in which polysaccharides can be organized on the cell surface; these are summarized in Fig. 2. At the simplest level, some EPSs may be released directly from the cell without any terminal molecule

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

139

Slime polysaccharides(EPS) fscherichia coligroup-I K-antigens (uncertain)

ECbG Some Escherichia coli group-I1 K-antigens Neisseria meningitidis CPS Haemophilus influenzae CPS (uncertain)

Some Escherichia coli group-11 K-antigens

w

CORE REGION

LIPID A

LPS 0-polysaccharides EC4PS Some fscherichia coli group-I K-antigens (KLPS)

FIG.2. Cell-surface association of polysaccharides in Gram-negative bacteria. (a) No membrane anchor. (b) Attachment through diacylglycerolphosphate. The diacylglycerol moiety is linked to the polysaccharide through a phosphodiester linkage. (c) Attachment through lipid A. 3-Deoxy-~-manno-octulosonicacid and phosphate are represented by KDO and P, respectively. which would serve as a cell-surface anchor. These polymers often form slime EPS with minimal cell association. However, in most instances, the slime definition is an operational one; this is usually based on retention of EPS on the cell surface during centrifugation. It is inevitable that components are sloughed off the cell surface during growth, or following exposure to shear forces during centrifugation. Release of polymer would clearly be greater from cells that produce abundant EPS, and most heavily encapsulated bacteria could therefore be considered to be slime producers. Physical characteristics of the particular polymer may also play a role in determining the degree of cell association. Surprisingly, some EPS is retained at the cell surface, despite the absence of an identifiable anchor structure. For example, studies on the CPS of Aerobacter aerogenes DD45 failed to identify a non-polysaccharide anchor, yet 80% of the polysaccharide was associated with the cell surface and sedimented with cells during centrifugation (Troy et al., 1971).Whether ionic or other interactions are sufficient to maintain cell association remains to be established.

140

C WHITFIELD AND M A VALVANO

Two modes of covalent surface attachment have been proposed for cellsurface polysaccharides. These involve a hydrophobic anchor, provided by either phospholipid or lipid A. Both modes are seen in strains of Escherichia coli, whose CPSs are subdivided into two groups (designated I and 11) based on physical, chemical and genetic criteria (Jann and Jann, 1990b). Use of phospholipid anchor is a characteristic of group-I1 capsules in E. coli and a similar phospholipid is bound to the CPS on Neisseria meningitidis and, probably, on Haemophilus injiuenzae. Use of a common linkage mechanism for structurally diverse polysaccharides is reminiscent of attachment of teichoic acids to peptidoglycan in Gram-positive bacteria (Coley et al., 1978). A diacylglycerol residue, linked through a phosphodiester bridge to the reducing terminal sugar of the polysaccharide, was first observed in the CPSs from N. meningitidis groups A, B and C and E. coli K92 (group-11 capsules) (Gotschlich et al., 1981). Early studies on the CPSs of various serotypes of H. influenrue identified terminal phosphomonoesters and cyclic phosphoderivatives (Egan et al., 1982) and the presence of a lipid moiety was later confirmed for the type-b polysaccharide (Kuo et al., 1985). The most prevalent form of enterobacterial antigen (ECAPG)also contains a terminal L-glycerophosphatidyl residue (Kuhn ef al., 1988). A slightly different variation is seen in some other group41 Kantigens of E. coli. In the K12 and K82 CPSs, the L-glycerophosphatidyl residue is linked to the polysaccharide through a KDO (3-deoxy-~-mannooctulosonic acid) residue (Fig. 2). The KDO residue is located at the reducing terminus of the polysaccharide but does not form part of the repeating unit structure of the polysaccharide (Schmidt and Jann, 1982). However, KDO is not found at the end of the K92 (group-11) CPS (Gotschlich et al., 1981). The K92 polymer is comprised of N-acetylneuraminic acid (sialic acid) and it is possible that similarities in structure between sialic acid and KDO eliminate a requirement for KDO in some CPSs. This could be tested by examining the sialic acid-containing K1 (group-11) CPS in E. coli. The presence of an L-glycerophosphate residue modifies the physicochemical properties of the polysaccharide. This is evident in the ability of lipid-substituted polysaccharides to produce micelles, and has been implicated in surface-active phenomena, including interaction with membrane proteins (Gotschlich et al., 1981; Kuhn et al., 1988) and, presumably, attachment of polysaccharides to membranes. Removal of the Lglycerophosphatidyl residue eliminates micelle formation. Interestingly, in most group41 CPSs from E. coli, only 20-50% of the isolated CPS is substituted with lipid (Jann and Jann, 1990b). In some instances this may reflect the labile linkage (hydrolysed at pH 6, 100°C, two hours) created by KDO. Lability presents a significant problem in analysis of the anchor and interpretation of its function. It has been suggested that CPS which is

CEI.1. SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

141

not lipid substituted is retained at the cell surface by ionic interactions (Jann and Jann, 1990b). Conceivably, some polymers considered to be slime polysaccharides also result from loss of the anchor. The second and better known membrane anchor is LPS lipid A. Lipid A is involved in attaching the LPS O-polysaccharide, as well as some of the group-I CPSs of E. coli. In enteric bacteria, lipid A contains a diglucosamine structure asymmetrically substituted with six fatty-acyl residues and, potentially, with 4-amino-4-deoxy-~-arabinoseand phosphoethanolamine. In some photosynthetic bacteria (e.g. Rhodopseudomonas viridis), D-glucosamine is replaced by 2,3-diamino-2,3-dideoxy-~-glucose, to give a lipid A-like molecule (Raetz, 1990). The core oligosaccharide is attached to lipid A, and can be divided into an inner- and an outer-core domain. The inner region usually contains residues of KDO and L-glyceroD-manno-heptose; the outer core contains several hexose residues. Only one core chemotype has been detected in Salmonella spp. (Hellerqvist and Lindberg, 1971). The same may be true of Shigella spp. (Gamian and Romanowska, 1982). In contrast, wild-type E. coli strains contain several different cores, designated R1, R2, R3 or R4; E. coli K-12 has a unique core structure (Holst etal., 1991). Most of the differences in these structures are located in the outer core (Jann and Jann, 1984; Haishima el al., 1992). Systematic evaluation of core structures from multiple strains of other bacterial species has not been carried out. The rough form of LPS (R-LPS) contains lipid A and variable amounts of the core oligosaccharide; no O-polysaccharide is attached. In the smooth form of LPS (S-LPS), the O-polysaccharide (O-antigen) is attached at the distal end of the core oligosaccharide. The chain length of O-polysaccharides is typically variable and gives rise to the strain-characteristic ladder pattern seen following analysis of S-LPS by sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDSPAGE). Most strains that synthesize S-LPS also contain variable but significant amounts of R-LPS (Goldman and Leive, 1980;Palva and Makela, 1980; Hitchcock and Brown, 1983; Hitchcock et al., 1986). Precise structural analysis has facilitated molecular modelling of LPS. It is suggested that the lipid A-core structure is conformationally rigid, whereas the O-polysaccharide is flexible (Kastowsky et al., 1992). Earlier studies proposed that the O-polysaccharide adopts a heavily coiled state (Labischinski et al., 1985). However, recent results have predicted that extensive helical domains are absent and that flexibility in the 0polysaccharide is confined to bends at specific linkages in the polysaccharide (Kastowsky et al., 1992). The bend in the LPS molecule allows a portion of the O-polysaccharide to fold and lie on top of the outer membrane. The result is a protective hydrophilic meshwork which can be visualized in

142

C

WI11fflt.I I) A N D

M A VAI V A N O

electron microscopy, and which extends 20-40 nm from the surface of the outer membrane (Shands et al., 1967; Lam et al., 1987, 1992; McCallum el al., 1989). The 0-polysaccharide layer of Pseudomonas aeruginosa has been visualized by electron microscopy of cells which were prepared for examination by freeze substitution (Lam et a f . , 1992). This cryo-electron-microscopy technique involves dehydration and chemical fixation of specimens frozen at -80°C and more accurately than previous methods preserves cellular ultrastructure and molecular conformation. A 36 nm-wide layer results from an average 0-polysaccharide chain length of 50 repeating units, and the experimentally observed size equates very well with the size predicted from the model of Kastowsky et al. (1992). Lipid A-core is a versatile carrier for expression of multiple cell-surface polysaccharides (Fig. 2). It has been shown that one form of ECA (ECAi ps) can be attached to lipid A-core in E. coli.Unlike ECApc;, which seems to can only occur in O-polysaccharidebe present in all enteric bacteria, ECAIdPS deficient R-LPS E. coli mutants with R1, R4 and K-12 core chemotypes (Kuhn el al., 1988). There are no structural features of a polysaccharide as such, which would indicate it is an LPS 0-polysaccharide or a (non-LPS) EPS. Capsular polysaccharides (K-antigens in enteric bacteria) are often considered to differ from LPS 0-polysaccharides with respect to the mode of cell-surface linkage and size; CPSs tend to be longer and can form a layer on the cell surface which often masks underlying 0-antigen in serological (agglutination) tests. The width of this layer can vary from as little as 10 nm to several micrometres (Amako et al., 1988; Bayer, 1990; Beveridge and Graham, 1991). However, the distinction between CPSs and 0-antigens is marginal in some cases. This is particularly evident with group-J CPSs of E. cofi, where the same repeating unit structure can be found as an 0-polysaccharide in one strain of E. coli and as a group-I CPS K-antigen (co-expressed with a different 0-antigen) in another. Examples include the K87 (=032), K85 (=0141) and K9 (=0104) polysaccharides (Jann and Jann, 1990b). This phenomenon may be explained by the suggestion that some group-I K-antigens in E. coli are attached to lipid Acore (Jann et d.,1992; Fig. 2). We term this form of K-antigen K, ps, to indicate its attachment to lipid A-core and its serological designation as the K-antigen rather than 0-antigen. Group-I K-antigens are expressed with a limited range of neutral 0-antigens (08, 0 9 , 020) and the acidic K-antigen would therefore form an additional LPS. The acidic KI f,s would be assumed to be the 0-antigen in strains lacking 08, 0 0 or 020. Linkage of high molecular-weight E. coli group-I K40 CPS to lipid A-

Cf;l.l.-SCIRPACE POLYSACCHARIDES IN GRAM-NEGATIVE BACIERIA

143

core has been confirmed directly (Jann et al., 1992). However, a survey of the S D S P A G E LPS profiles of established group-I K-antigen wild-type strains has shown that attachment of more than two repeating units of K-antigen to lipid A-core is relatively rare; some strains appear not to attach any K-antigen in this way (P. R. MacLachlan, W. J. Keenleyside, C. Dodgson and Whitfield, unpublished data). In E. coli K30, short Koligosaccharides, primarily containing one repeating unit of K30 polysaccharide, were shown to be linked to lipid A-core as KLps (Homonylo et al., 1988; Whitfield et al., 1989). Mutants have been isolated with as many as eight repeating units of K30 polysaccharide attached to lipid Acore (P. R. MacLachlan et d . , unpublished data). However, the capsule structure seen under the microscope consists of high molecular-weight (over lo5) K30 polysaccharide; K30LPScannot form a capsule alone. The means by which high molecular-weight K30 polysaccharide is attached to the cell surface has been addressed in one of our laboratories (P. R. MacLachlan et al., unpublished data). Derivatives of E. coli K30 were constructed with deletions spanning four genes of the rfa (LPS-core biosynthesis) cluster. These strains contain a truncated LPS core and K3OLPs is not expressed due to lack of the attachment site for polysaccharides. However, a K30 capsule is still assembled in the absence of KI.PS, indicating that high moleculareight K30 CPS can be exported from the cell in an LPS-independent form. Data presented by Jann et al. (1992) also indicate that a substantial amount of the K40 CPS is not substituted by lipid A-core. In Fig. 2, this is shown as free polysaccharide but it is also possible that an alternative membrane anchor (conceivably a phospholipid) could be present. It is interesting that ECA also shows a mixture of short chains attached to LPS (ECALpS, approximately five repeating units) and LPSindependent longer chains with terminal phospholipids (ECApG,20 repeating units; Kuhn et al., 1988). However, unlike KLPS,which can be coexpressed with 0-polysaccharide, ECA12pSis only expressed in strains lacking 0-antigen. In many respects, the situation in the E. coli group-I CPSs also resembles the 0-antigen capsule of E. coli 0 1 11, originally reported by Goldman et al. (1982, 1984) and confirmed by others (Peterson and McGroarty, 1985). Polysaccharide of similar composition was identified attached to lipid Acore as a conventional 0 1 l l 0-polysaccharide and also in an unsubstituted form, comprising an 0 111 capsule. Unlinked 0-polysaccharide was subsequently found in E. coli 0 5 5 and 0127 (Peterson and McGroarty, 1985). Structurally identical CPS and 0-polysaccharide have been described in Proteus mirabilis 0 6 (Beynon et al., 1992), indicating that this

144

C W111WIEI.D A N D M

A

VAI.VANO

phenomenon is not unique to E. coli. Interestingly, strains of Salmonella enterica serovars typhimurium and minnesota do not appear to exhibit unlinked 0-polysaccharide (Peterson and McGroarty, 1985), and the full extent of this phenomenon is unclear. 8 . REPEATING UNIT STRUCTURES IN CELL-SURFACE POLYSACCHARIDES

The array of structures known to be present in bacterial cell-surface polysaccharides is large and is increasing as the techniques for structural analysis improve. We d o not attempt to cover structural aspects in anything other than cursory detail here, but an overview of structural themes is required to put biosynthetic processes into perspective. The examples chosen are ones which are relevant to the later sections of this review covering biosynthesis. For a broader coverage, the reader is referred to Kenne and Lindberg (1983). Bacterial polysaccharides are antigenically diverse. Changes in component monosaccharide residues, linkages, and substitution with noncarbohydrate residues, all influence the immunochemistry of the polymer. This results in the multiple serotypes established for many bacterial species. The component residues include pentoses, hexoses, heptoses, amino sugars, methylated sugars and uronic acids, together with non-carbohydrate residues such as phosphate, amino acids, glycerol and ribitol (Kenne and Lindberg, 1983). Monomers that were once thought to be unique to particular cell structures have now been found in several polysaccharides. For example, KDO was widely considered to be an indicator for LPS but its presence in the repeating units and linkage units of several E. coli group 11 CPSs indicates that it is not a reliable marker. Recently, muramic acid was identified in the 0-polysaccharide of Yersinia ruckeri serotype 0 2 (Banoub ef al., 1990); it was previously thought that this component was unique to peptidoglycan. Both EPSs and 0-polysaccharides include examples of acidic and neutral polymers, and homopolysaccharides and heteropolysaccharides are both found. Some cell-surface polysaccharides are widely distributed in a given species or in related species (Table 1). For example, enterobacterial common antigen (ECA) is found in all enteric bacteria, except Erwinia chrysanthemi (Kuhn et al., 1988), and is a true common antigen. The polyrhamnose Aband 0-polysaccharide of P . aeruginosa (Yokota et al., 1987; Arsenault et a l . , 1991) is present in strains which co-express a range of different serotypespecific B-band 0-polysaccharides (Sawada et al., 1985; Lam et al., 1989). A-Band 0-polysaccharide is also found in fseudomonas syringiae (Smith et al., 1985) and may also be a common antigen. Structurally identical polysaccharides can be produced by limited

TABLE 1. Some common Dolvsaccharide structures found in bacterial cell-surface layers Polysaccharide

Structure

Reference Garegg ef ul. (1971a.b)

OAc

Colanic acid in enteric bacteria

1 2l3 +4)-a-~-Fucp-(1+3)-p-~-Glcp-(I+3)-p~-Fucp-(I-+ 4

t

1 @~-Galp-( 1+4)-p-~-GlcpA-(1+3)-@~-Galp

II

Py' Enterobacterial common antigen (ECA) in enteric bacteria

+4)-pManNAcA-( I+4)-a-GlcNac-( 1+3)-a-Fuc4NAc-( 1-

Lugowski ef ul. (1983)

Vi antigen in some Sulmonellu and Citrobucter spp.

+4)-a-~-GalNAcA-(I+ 3

Heyns and Kiessling(1%7)

t

OAc Succinoglycan in Rhizobium meliloti (EPSI), Alcaligenes fuecufis and Agrobucterium spp.

,O-acetyl. or +4)-pD-Gicp-( 1+4)-p~-G&p-(1+4)-p~-Galp-(1+3)-B-D-Glcp( 16

,.'

'\,

t

1 @D-Glcp-(1+3)-pD-G,kp-( I+3)-pD-Gkp-( 1+6)-PDGlcp 46 *. or -.-cc* II 0-s'uccin91 Pvr

Aman er ul. (1981)

146

C WtiITFlEL.I> AND M A VALVANO

representative strains of several bacterial species, but these are not generally considered as common antigens even though they are sometimes members of the same family. For example, several members of the Enterobacteriaceae produce the slime EPS colanic acid (Markovitz, 1977), while the Vi antigen is restricted to certain Salmonella and Citrobacter spp. (Selander et al., 1992). Alternatively, similar polysaccharides can be produced by unrelated bacteria, as for example succinoglycan in Rhizobium meliloti (EPSI) and strains of Alcaligenes faecalis and Agrobacterium spp. These same bacteria also produce a range of similar low molecular-weight p-glucans (Sutherland, 1985). The p-( 1-+2)-glucans of A . tumefaciens and R . rneliloti are generally cyclic and can exist as neutral polymers containing between 17 and 40 glucose residues (Zevenhuizen et a l . , 1990; Koizumi et al., 1984). Some p-(1-+2)-glucans exist in an anionic state due to modification with 1-4 phosphodiester-linked glycerophosphate residues in each molecule (Batley et al., 1987; Miller et al., 1987). Linear mixed-linkage pglucans are found in E. coli, where they are known as membrane-derived oligosaccharides. p-Glucans are generally located in the periplasm (Abe et al., 1982; Kennedy, 1987) where they play a role in osmoregulation in low osmotic-strength media (Miller et a f . , 1986). In A . tumefuciens and R . meliloti, p-glucans also play a role in plant-microbe interactions (Dylan et al., 1986) and significant amounts are also excreted into the growth medium by these bacteria (Amemura et al., 1983). p-Glucans are not typical EPSs, although they are often considered alongside cell-surface polysaccharides. Alginate, an unusual linear copolymer comprised of D-mannuronic acid and D-guluronic acid residues, provides another example of a widespread polysaccharide. Pseudomonas aeruginosa shares the slime EPS alginate with strains of Azotobacter vinelandii and some brown algae (May et al., 1991). Many structures found in CPSs and 0-polysaccharides are considered to be serotype-specific. While this may be true when a given species is considered in isolation, it should be remembered that the same structure can occur as a serotype-specific polysaccharide in different bacterial species. For example, Serratia marcescens and Pasteurella haemolytica produce 0polysaccharides identical to the D-galactan I structure of Klebsiella pneumoniae serotype 0 1 (Table 2). Also, identical CPS structures are produced in N . meningitidis and E. coli (see Table 4) and in Klebsiella spp. and E. coli (see Table 5 ) . The presence of similar polysaccharides in diverse bacteria raises interesting questions concerning evolution of polysaccharide types (see Section V.E). Analysis of polysaccharides reveals recurrent structural themes in what could be considered as polysaccharide families. Addition of side groups to a common polysaccharide, or variation in one of the linkages, alters the

TABLE L. Structures of representative homopymeric 0-plysaccharides found in bacterial cell-surface layers Bacterium

Polysaccharide structure

Reference

Escherichio coli 0 8 Klebsiello pneumonioe 0 5

+3)-p~-Manp-(1+2)-a-~-Manp-( I+2)-a-D-Manp-( 1-

Jansson

Escherichio coli 0 9 Klebsiello pnewnonioe 0 3

+3)-a-~-Manp-(1+3)-a-~-Manp-( 1+2)-a-~-Manp-(1+2)a-~-Manp(1+2)-a-~-Manp-(l-

Prehm er 01. (1976) Curvall ef ul. (1973)

Escherichio coli 0 9 a

+3)-a-D-Manp-( 1+3)-a-D-Manp-( I+z)-a-D-Manp-( I+Z)-a-~-Manp-( I+

Parolis et al. (1986)

Klebsiella pneumonioe 01

-3)-pD-Galf-( 1+3)-a-D-Galp-( I+ +3)-U-D-Gdp-( 1+3)-P-D-G&?-( 1-

(D-Galactan I“) (D-Galactan 11“)

Whiffield er 01. (1991) Kol er 01. (1991)

(D-Galactan I)

Whiffield et al. (1992) Perry and Babiuk (1983) Richards and Leitch (1989) Oxley and Wilkinson (1989a)

Klebsiello pnewnonioe O h +3)-pD-Galf-( I+3)-a-D-Galp-( 1Pasreurello hoemolytico serotype 4 Pmreurella hoemolytico serotype 10 Serrafk mrcescem 0 1 6 and 020 Klebsiella pnewnonioe 0 2 a , 2f, 2g Serrotia nzorcescens 0 2 4

+3)-p-D-Galf-( 1+3)-a-D-Galp-( I+ 4

t

1

+3)-f!-D-Galf-(

al. (1985)

M. B. Perry, L. L. MacLean and C. Whiffield, unpublished results Oxley and Wilkinson (1989b)

a-D-Galp

Klebsiella pneumoniae 02a, 2e and 0 9

et

1+3)-a-D-Galp-( 1+3)-&D-GaIf-( 1+3)-a-D-Galp-( 1+ 2

MacLean er 01. (1993)

t

1 U-D-Galp (Acetylated with 1.7 nmol 0-acetyl groups per repeating unit) f The terms “D-gahCtan I” and “D-galactan 11” (Whitfield er ol., 1991) are used to describe the two distinct 0-plysaccharide structures found in the LPS of Klebsiella pneumonioe 01. a

148

C. Wt1ITFIEI.D AND M A . VAI.VAN0

structure and immunochemistry of the polysaccharide. For example, in E. coli 0 8 and 0 9 and Klebsiella spp 0 3 and 0 5 , similar mannose residuecontaining homopolysaccharides are differentiated by the nature of the linkages (Table 2). Several related D-galactose residue-containing 0polysaccharides are produced by Klebsiella spp. The 1,-galactan I polysaccharide of serotype 0 1 is modified in serotypes 0 2 a , 2f, 2g and 0 9 by addition of side-chain a-r>-galactosyl residues (Table 2). Another family of structures is found in S . enterica, where a common polymer backbone is substituted with different 3,6-dideoxyhexose residues to form serotypes A, B and D (Table 3). The O-acetyl and glucosyl residues in 0polysaccharides of S. enterica are non-stoichiometric substitutions and affect the immunochemistry of the polysaccharide. The group-I1 CPSs of E. coli include a number of very similar structures, highlighted by the K13, K20 and K23 family of CPSs. The K13 and K20 CPSs differ from K23 only in the presence or location of O-acetyl groups (Table 4). Some E. coli group41 CPSs have structures similar to those found in N. meningitidis and H . influenzae, with sialic acid, KDO, ribose and phosphate being common components (Table 4). The presence of ribitol and phosphate residues in some group-I1 CPSs also gives a resemblance to teichoic acids in Gram-positive bacteria (Kenne and Lindberg, 1983). Using the E. coli group41 CPSs as prototypes, we have termed these polysaccharides as “group 11-like’’ for the purposes of discussion. It is interesting that the group 11-like CPSs from Gram-negative bacteria are substituted with phospholipids and, as will be discussed later, may have some similarities in their biosynthetic pathways. Structures similar to the E. coli group-11 CPSs are found in the shared polysaccharides ECA and Vi antigen (see Table 1) and in some CPSs in streptococci (Kenne and Lindberg, 1983). Some group-I CPSs in E. coli resemble structures found in Klebsiella spp. The E . coli K30 and Klebsiella spp. K20 CPSs provide one of several examples of shared structures (Table 5 ) . Although the high molecularweight polymers are identical, no K,*pshas been detected in Klebsiella spp. K20 (C. Whitfield, unpublished results). Group-I CPSs from E. coli which contain amino sugars (e.g. K40) have no counterparts in Klebsiella spp. O-Acetyl and pyruvate groups are common in group I-like CPSs. In some instances, the position of these substituents (rather than the carbohydrate structure) defines the biological repeating unit of the polymer. For example, in Klebsiella spp. K54 and K70, the substituents are found on alternate carbohydrate repeating units. The structure of colanic acid (see Table 1) resembles that of group I-like CPSs of E. coli and Klebsiella spp., but the loose cell association of this polymer has led to it being considered as a slime polymer rather than a CPS. Genetic analyses (see Sections V.C and V1.B) indicate this may be a rather artificial distinction.

C'El.I.-SURFACE POI.YSA(:CHARIDES IN GRAM-NEGATIVE BACTERIA

149

C . EXPRESSION OF MULIIPLE CELL-SURFACE POLYSACCHARIDES

From the review already presented, it is clear that the cell surface of an individual bacterium can contain a range of different polysaccharides. This can result from the presence of structurally different polymers, o r from microheterogeneity in the structure of a single polysaccharide. Enteric bacteria, for example, produce 0-polysaccharides, K-antigen CPSs, ECAs and, in some cases, colanic acid. The 0- and K-antigens are detected serologically, but other polysaccharides may not be resolved by some serological techniques because they are present in small amounts o r are masked by other polysaccharides. A good example is the T (T1 and T2) 0-polysaccharides which are attached to lipid A-core in Salmonella spp. ; the T-antigens are normally masked by the longer-chain 0-antigens (Makelii and Stocker, 1984). The common A-band LPS of P. aeruginosa is masked by the serotype-specific B-band LPS in a similar way (Lam et al., 1989; Rivera and McGroarty, 1989). There are also examples where a particular polysaccharide is not synthesized under laboratory growth conditions and is detectable only under specific growth conditions, or following mutagenesis. Colanic acid in enteric bacteria (Markovitz, 1977), EPSII (EPSb) of Rhizobium sp. (Levery et al., 1991) and alginate from P. aeruginosa illustrate this point. The regulatory mechanisms involved in these phenomena are discussed in Section VI. Apparent multiplicity of cell-surface polysaccharides can also arise from non-stoichiometric substitution of polysaccharide backbones. This results in microheterogeneity in polysaccharide structure. For example, the serotype 0 1 0-polysaccharide of Klebsiella pneumoniae (see Table 1) contains two galactan structures (Kol et al., 1991; Whitfield et a l . , 1991). This was originally thought to be due to the presence of two discrete 0polysaccharides, each attached to independent lipid A-cores. However, recent data suggests that the [>-galactan I structure is attached to lipid Acore and the mgalactan I1 structure is covalently attached to the distal end of some [,-galactan I chains (Kol et al., 1992). Some I,-galactan I chains are not substituted, thereby providing heterogeneity in the 0polysaccharides. In the 0-polysaccharide of S . enterica serovar boecker, non-stoichiometric substitution of the backbone polymer with a-(1-3)glucosyl residues results in two glycan structures, one modified and the other not (Brisson and Perry, 1988). In S. enferica serovar madelia, a-(1~3)-glucosyland a-(I--~4)-glucosyl residues are attached to different sugars in the polysaccharide backbone. Three glycans containing two, one or no substituents are isolated from the same culture (DiFabio et al., 1989). It remains to be established whether these different forms are attached to individual lipid A-cores, or again reflect extensive domains within a single chain attached to the same lipid A-core.

TABLE 3 . Structures of representative heteropolymeric 0-polysaccharides found in bacterial cell-surface layers Bacterium Salmonella serogroup A

a-Parp 1

1

Polysaccharide structure

Reference

a-D-Gkp 1

Hellerqvist er al. (1971d)

1

3 416 -+2)-a-~-Manp-( 1+4)-a-i.-Rhap-( 1 +3)-a-~-Galp-( 1-

Salmonella serogroup B

a-Abep(2-OAc) 1

1

a-~-Glcp I

Hellerqvist et af. (1968, 1969a.b)

1

3 416 -2)-a-~-Manp-( 1-4)-a-i.-Rhap-( 1-3)-a-~-Galp-( I-

Salmonella serogroup DI

a-TYY

a-D-Gkf

1

1

1

1

3 416 -2)-a-~-Manp-( I+4)-a-[.-Rhap-( 1+3)-a-D-Gaip-( 1-

Hellerqvist er al. ( 1 9 6 9 ~ . 1970a. 1971a)

i

t

d

P

t

d

h

s m

Q

s:

9

m

h

t

w c

86

b

E n N

-t

V

2

$ 9 P t

h

N

?

9 6

b

c- 3 6

ss

I

TABLE 4. Structures of representative group 11-like plysaccharides found in bacterial cell-surface layers Bacterium

Polysaccharide structure

Reference

Escherichia coli K1 Neisseria meningitidis group B Moraxella nonliquefaciem Pasteurella haemolytica A2

-+8)-a-Neup5Ac-(2-

Escherichia coli K92

+8)-a-NeupSAc-( 2+9)-a-NeupSAc-(2-t

Lifely et al. (198.5)

Neisseria meningitidis group C

+9)-a-NeupSAc-(2718

Bhattacharjee et al. (1975)

McGuire and Binkley (1%)

Bhattacharjee er al. (197.5) Devi et al. (1991) Adlam et al. (1987)

t

OAc

Eschenchia coli K5

+4)-kD-GlcpA-( l-tJ)-a-~-GtcpNAc(1+

Vann et al. (1981)

Escherichia coli K7

-3)-k~-ManpNAcA-( 1-+4)-k~-Glcp-(1 4 6

Tsui er al. (1982)

T OAc

Eschenchia roll K13

+3)-P-r>-Ribf-( 1+7)-FKDOp-(?+ 4

Vann and Jann (1079)

t OAc

c 011 K20

Vann et al. (1983)

Eschenchia coli K73

V a n n cf a1 (19x3)

Eschmchra

0

Branefors-Helander el al (1976) Criwl e l a1 (1975

0

11 Escherichra coli K??

-?)-b-r,-Ribf-(

1-7)-r,-ribitoI-(5-0-P-O~

OH

Rodgriguez ef a / . (1988)

154

C . WHITFIF1.D AND M. A VALVANO

111. Polysaccharide Biosynthesis

Biosynthesis of bacterial cell-surface polysaccharides involves a series of sequential processes, the individual steps of which have been characterized to various extents. The initial step involves biosynthesis of activated precursors in the cytoplasm. This aspect has been covered in depth by Shibaev (1986) and will not be addressed in any detail here. The remaining steps involve: (a) formation of repeating units; (b) polymerization of repeating units; and (c) export of polysaccharides to the cell surface. In Gram-negative bacteria, the export phase must involve processes for traversing the cytoplasmic membrane, periplasm and outer membrane. Each of these stages in biosynthesis will be considered, using information available for both LPS 0-polysaccharide and EPS. Assembly of repeating units in bacterial polysaccharides and their subsequent polymerization have been studied by biochemical approaches. Membrane (particulate) preparations, or cells made permeable to precursors by treatment with EDTA, have been used as a source of biosynthetic enzymes. The in vitro synthesis of polymer is monitored by supplying radioactively labelled sugar nucleotide precursors. Many of these elegant studies were performed in the late 1960s. Recent molecular biology approaches have facilitated analysis of individual reactions in isolation, and detailed information regarding components and mechanisms is now becoming available. Most, or all, of the biosynthetic genes for some polysaccharides have been cloned and sequenced. A. FORMATION O F UNDECAPRENOL-LINKED INTERMEDIATES

Studies with two strains of Salmonella enterica clearly established that sugar nucleotides served as precursors for in vitro synthesis of the LPS 0polysaccharide. As was found in biosynthesis of peptidoglycan (Anderson et al., 1965), sugar nucleotide precursors were not transferred directly to a growing polymer chain but instead were assembled as oligosaccharide intermediates on a lipid carrier (Weiner et al., 1965; Wright et al., 1965; Dankert et al., 1966). The lipid (antigen-carrier lipid, ACL; glycosyl-carrier lipid, GCL) was subsequently identified by mass spectroscopy as a Cs5 polyisoprenoid alcohol derivative, undecaprenol phosphate (Wright el al., 1967); the same lipid was also found to be involved in peptidoglycan synthesis (Higashi et al., 1967). Several 0-polysaccharides of S. enterica (serogroups A , B, D and E; see Table 3) have the same (galactosyl-mannosyl-rhamnosyl) backbone, but differ in their side-chain dideoxyhexoses. Biosynthesis of 0-polysaccharide has been studied in detail in S . enterica serovars typhimurium (serogroup

C,-P-P

1,

d-Rha-Man

e['

LIPID A-CORE

SMOOTH LPS

ACCEPTOR

ELONGATED ACCEPTOR

FIG.3. Pathway for assembly of undecaprenol-linked oligosaccharide intermediates in biosynthesis of (a) 0-polysaccharide in Salmonella enterica serovar typhimuriurn, and (b) capsular polysaccharide in Aerobacter aerogenes DD45.Undecaprenol and phosphate are represented by Css and P, respectively.

156

C . WIIITFIEL.D AND

M . A. VALVANO

B) and anatum (serogroup El). As might be expected, the biosynthetic pathways for these structurally related polymers are very similar; that for the group-B 0-antigen is illustrated in Fig. 3(a). Assembly of both 0polysaccharide repeating units occurs through a series of interdependent and sequential reactions, each mediated by specific glycosyltransferases (Robbins et al., 1964; Nikaido, 1965; Nikaido and Nikaido, 1965; Zeleznick et al., 1965). The galactosyltransferase RfbP catalyses formation of galactosylpyrophosphorylundecaprenol (Osborn and Yuan Tze-Yuen, 1968). This reaction is reversible and differs from subsequent steps in that it involves transfer of galactose 1-phosphate from the precursor, the energy of the linkage in the precursor being fully conserved in the resulting intermediate. In subsequent reactions, the sugar rather than the sugar phosphate is transferred. Each of the undecaprenol-linked intermediates has been purified and characterized, and the sequential synthetic steps clearly demonstrated (Weiner el al., 1965; Wright el al., 1965; Dankert el al., 1966; Osborn and Yuan Tze-Yuen, 1968; Osborn and Weiner, 1968). Essentially similar pathways have been established for S. enterica serogroups C2 (newport), C3 (kentucky), and E4 (sefrenberg) (Shibaev, 1978, 1986; Shibaev el al., 1979). Biosynthesis of the CPS from Aerobacter (Klebsiella) aerogenes strain DD45 was studied in detail by Troy et al. (1971). The polymer comprises a tetrasaccharide repeating unit, the structure of which is shown in Table 5. The reaction sequence was established by isolation and characterization of each of the intermediates and, as seen in Fig. 3(b), the biosynthetic pathway is strikingly similar to that elucidated for the 0-polysaccharides of S. enterica serogroups B and E. The obligatory requirement for undecaprenol was shown by reconstitution of lipid-depleted membrane preparations with purified lipid fractions. Evidence has been reported for a similar reaction sequence in the CPS of K. aerogenes type 8 (Sutherland and Norval, 1970). Preliminary studies suggest that this type of mechanism also operates in assembly of colanic acid in E. coli K-12 (Johnson and Wilson, 1977). Based on similarities in CPS structure and location of genes involved in the biosynthesis of the polysaccharides (see Sections V.C.2 and VI.B.l), it would not be surprising if the assembly mechanism for group-I K-antigens Escherichia coli resembles that in Klebsiella spp. Studies on the biosynthesis of the EPS xanthan gum in Xanthomonas campestris have indicated that the general features of the biosynthetic reactions already described extend beyond enteric bacteria. In X . campestris, the EPS pentasaccharide repeating unit is synthesized by addition of sugars to a lipid fraction in the reaction sequence illustrated in Fig. 4. Each of the lipid-linked oligosaccharides has been identified. Although the identity of the lipid moiety was not determined, its properties

TABLE 5. Structures of representative group I-like polysaccharides found in bacterial cell-surface layers Polysaccharide structure

Bacterium Klebsiella spp. K54

+4)-a-~-GlcpA-(1+3)-a-~-Fucp-(1+6)-P-~-Glcp-(I+ 2 4

t

OAc

Reference Dutton and Memfield (1982)

t

1

P-D-GIc~ (OAc on alternate a-L-Fucp- residues) Klebsiella spp. K70 1-4)-fb~-GIcpA-( 1+4)-a-~-Rhap-(1+2)-a-~-Rhap(1+2)-a-~-Glcp-(1+3)-f&~-Galp-(1+2)-a-~-Rhap-( 3 4 (Pyruvate on alternate repeating units) 11 PYr Escherichia coli K40

+4)-&~-GlcpA-(1+4)-a-~-GlcpNAc-(I+6)-a-~-GlcpNAc-( I+ 6

Dutton and Mackie (1978)

Dengler ei al. (1986)

t

L-serine (amide) Escherichia coli K30

Klebsiella pneumonae K20

+2)-a-~-Manp-(1+3)-fbD-Galp-( 1-

3

Chakraborty et al. (1980) Choy and Dutton (1973)

t

I p-~-GlcpA-(1+3)-a-~-Galp Aerobacier aerogenes DD45

+3)-a-D-Galp-( 1+3)-a-~-Manp-(1+3)-a-~-Galp(12

t

1

&~-GlcpA

Yurewia ei al. (1971)

158

C. WHITFlELD AND M. A. VALVANO

C -,,

P-P-Glc-Glc

C,,-P-P-Glc-Glc

C,,-P-P-Glc-Cjlc Mm-OAc

JsqP acetylation

AC-COA

/

ketalation polymerizationpossibly at the

PEP

ACCEPTOR

ELONGATED ACCEPTOR

FIG. 4. Pathway for assembly of undecaprenol-linked oligosaccharide intermediates in biosynthesis of the extracellular polysaccharide xanthum gum in Xanthomonas campesrris. Undecaprenol and phosphate are represented by CSSand P, respectively.

were those expected of undecaprenol pyrophosphate (Ielpi ef al., 1981a). The O-acetyl and pyruvate groups are added to the undecaprenol-linked pentasaccharide (Ielpi et al., 1981b; Marzocca el al., 1991) and polymerization follows. A similar process is probably involved in assembly of EPSs in other plant-associated bacteria, including Rhizobium trifolii (Bossio el al., 1986; Gardiol and Dazzo, 1986) and R. melilofi (Tolmasky et al., 1980,

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

159

1982; Ugalde et al., 1986) and in formation of succinoglycan by Alculigenes fueculis var. myxogenes (Hisamatsu et al., 1978). In the biosynthesis of the group-I1 CPSs of E. coli K1 and Nekseriu meningitidis group B, a slightly different reaction sequence occurs (Troy et ul., 1975; Masson and Holbein, 1985). Residues of sialic acid are transferred from CMP-sialic acid to undecaprenol phosphate to form sialylmonophosphorylundecaprenol.The presence of the monophosphoryl linkage clearly differentiates this mechanism from that for biosynthensis of the serogroup-B 0-polysaccharide in S. entericu. B . POLYMERIZATION REACHONS

Two mechanisms have been described for polymerization of cell-surface polysaccharides. These involve either polymerization of preformed undecaprenol-linked repeating units in a blockwise manner, or a processive, sequential transfer of sugars to a growing polysaccharide attached to

C,,-P-Prn

C55-P-Pm

+ C,,-P-Prn

+

-

n(GDP-Man)

c 5 5 - P - P m

-+ c ~ ~ - P - P { Z H ~

FIG. 5. Mechanisms involved in polymerization of 0-polysaccharides: (a) rfeindependent 0-polysaccharides in Salmonella enterica; (b) rfe-dependent 0polysaccharides in Escherichia coli 0 8 and 09. Undecaprenol and phospate are represented by Cssand P, respectively.

160

C. WHI1'FIEI.D AND

M. A . VALVANO

undecaprenol. The prototypes for these models are rfe-independent 0polysaccharides of S. enterica and rfe-dependent O-polysaccharides of E. coli, respectively (Fig. 5 ) . The mechanism used may be correlated with the direction of polymer growth during polymerization. 1. Growth of O-Polysaccharide at the Reducing Terminus

Polymerization of the O-antigen of S. enterica serovar anatum (group E l ) occurs using undecaprenol-linked repeating-unit blocks, rather than sequential addition of sugars. The model formulated by Robbins et al. (1967), and shown in Fig. 5(a), is now generally accepted. The observation that preformed undecaprenol-linked trisaccharides can be polymerized without any requirement for de novo synthesis of lipid intermediates (Kanegasaki and Wright, 1970; Shibaev et al., 1979) is consistent with this blockwise assembly mechanism. A series of pulse-chase experiments unambiguously established that polymer growth occurred at the reducing terminus in the O-polysaccharide of S. enterica serogroup E (Bray and Robbins, 1967; Robbins et al., 1967). Polymerization of the CPS in A . aerogenes DD45 also occurs at the undecaprenol-intermediate level (Troy et al., 1971) and, although the direction of polymer growth was not examined in A. aerogenes, the two mechanisms appear to be very similar. The polypeptide antibiotic bacitracin inhibits recycling of undecaprenol by preventing dephosphorylation of spent undecaprenol pyrophosphate to regenerate the active monophosphoryl derivative (Siewart and Strominger, 1967; see Fig. 3). Consequently, in vitro assembly of the O-polysaccharide from S. enterica serogroup B (Jann and Jann, 1984) and type-8 CPS in K. aerogenes (Sutherland and Norval, 1970) is affected by bacitracin. If sensitivity to bacitracin can be interpreted as a clear indication of mechanism, colanic acid synthesis in E. cofi uses a similar process (Johnson and Wilson, 1977). Undecaprenol-linked intermediates containing two polysaccharide repeating units (degree of polymerization of two), have been isolated and characterized in both the O-polysaccharide of S. enterica serogroup B (Osborn and Weiner, 1968) and the CPS of A. aerogenes (Troy et al., 1971) systems. However, there are no reports of isolation of intact higher molecular-weight undecaprenol-linked intermediates. This may be due to solubility problems when isolating large lipid-linked oligosaccharide, together with lability of the pyrophosphate linkage. Regardless, it has become generally accepted that polymeric (high molecular-weight) polysaccharides are first assembled as undecaprenol-linked intermediates and then transferred en bloc to an acceptor molecule. Polymeric intermediates have been described as products of in vitro synthetic reactions in

CELL-SURFACE POLYSACCHARIDES IN CRAM-NEGATIVE BACTERIA

161

S. enterica serovar typhimurium (serogroup B) (Kent and Osborn, 1968b).

Similar molecules were also detected in vivo in strains with defective lipid A-core assembly (Kent and Osborn, 1968a). These polymers (up to 30 repeating units) are not linked to lipid A-core and are known as O-haptens. Serogroup-B O-hapten was retained at the cytoplasmic membrane and a precursor-product relationship between O-hapten and lipid A-core-linked O-polysaccharide could be demonstrated in vivo (Kent and Osborn, 1968c) and in vitro (Kent and Osborn, 1968b).The O-hapten contained a galactose l-phosphate residue at the reducing terminus, and was attached through an acid- and phenol-labile linkage. In all respects, the properties of the 0hapten were consistent with an O-polysaccharide attached to undecaprenol pyrophosphate (Kent and Osborn, 1968a). Polymerization of the O-polysaccharide from S. enterica serogroup B is dependent on the rfc-gene product. Mutations in rfc result in LPS comprising a single repeating unit of 0-polysaccharide attached to lipid Acore (Naide et al., 1965; Makela and Stocker, 1984; Collins and Hackett, 1991). Although a polymerase-defective phenotype is evident in rfc mutants, there is n o biochemical evidence that unequivocally identifies Rfc as a polymerase enzyme as such.

2. Growth of O-Polysaccharide at the Non-reducing Terminus The polymerization mechanism already described for polysaccharides in serogroups B and E of S. enterica is termed rfe-independent, i.e. there is no requirement for functions encoded by rfe. In contrast, rfe-dependent O-polysaccharide biosynthesis occurs in E. coli 0 8 and 0 9 (Jann er al., 1979), Salmonella spp. groups C1, L (Makela et af., 1970) and T1 (Makela and Stocker, 1984) and Klebsiella pneumoniae 0 1 (Clarke and Whitfield, 1992). Biosynthesis of ECA also requires rfe. The full extent of the requirement for rfe in polysaccharide biosynthesis is unknown, but pathways for group-I (Schmidt et al., 1976) and group-I1 (Meier-Dieter et al., 1990) CPSs of E. coli are rfe-independent. The involvement of rfe has been taken as an indication that rfe-dependent O-polysaccharide biosynthesis is fundamentally different from the rfe-independent process. However, extensive studies carried out in Jann’s laboratory on the mannose homopolymers in E. coli 0 8 and 0 9 (see Table 2) represent the only examination of an rfe-dependent O-polysaccharide-biosynthesis pathway to date. The O-polysaccharides in E. coli 0 8 (Flemming and Jann, 1978b) and 0 9 (Flemming and Jann, 1978a) grow at the non-reducing terminus, in the opposite direction to that described above for the O-polysaccharide of S. enterica serogroup E. The mannan polymer was found attached to a hydrophobic phosphorylated camer (Flemming and Jann, 1978a) through

162

C WI{ITI.IEI 1) AND M

A VAL V A N 0

a glucosyl residue at the reducing terminus (Flemming and Jann, 1978a.b). The carrier molecule was initially described as a glucosyl phospholipid but, after purification, it was unequivocally identified as a-glucosyldiphosphorylundecaprenol (undecaprenol-PP-Glc) (Weisgerber and Jann, 1982). a-Glucosyldiphosphorylundecaprenol(Jann et al., 1982; Weisgerber and Jann, 1982) or synthetic glucosyldiphosphorylpolyprenol derivatives (Jann et a f . , 1985) provided direct acceptors for mannose residues in membrane-reconstitution studies. In biosynthesis of the 0 9 polysaccharide, undecaprenol-PP-Gle serves as an acceptor for mannosyl chains consisting of at least 30 residues, equivalent to six pentasaccharide repeating units (Fig. 5 ; Jann et al., 1985). This is consistent with a biosynthetic model in which mannosyl residues are sequentially transferred to the growing chain from GDP-mannose, without participation of the type of blockwise assembly typical in biosynthesis of the 0-polysaccharide in S . enterica group E (Fig. 5). N o mutations equivalent to rfc have been reported in E. coli 0 8 and 0 9 (Jann and Jann, 1984) and the absence of a single repeating unit of 0-polysaccharide to lipid A-core might be predicted by this processive model. This mechanism also explains the absence of inhibition by bacitracin in vitro (Kopmann and Jann, 1975). Since one undecaprenyl phosphate molecule is used by each growing chain (rather than multiple carriers in the S. enterica rfe-independent process), the inhibitory effect of bacitracin would not occur until the chain is complete. The glucosyl residue in the lipid acceptor for 0 9 polymer is clearly not part of the 0-polysaccharide repeating unit and its presence is not a universal feature in rfe-dependent 0-polysaccharides. For example, 0hapten in the rfe-dependent S . enterica serovar montevideo (serogroup C1) is attached to undecaprenol pyrophosphate through a mannose residue, a component in the repeating unit (Heasley, 1981). Following polymerization of the 0 9 polysaccharide in E. cofi, the glucose residue is thought to be transferred with the 0 9 polymer to an acceptor molecule consisting of precore LPS. The precore lacks its terminal glucose residue (Weisgerber el al., 1984). In this respect, biosynthesis of 0 9 polysaccharide resembles assembly of teichuronic acid in Micrococcus lysodeikticus, in which polymer is assembled on an undecaprenol-linked trisaccharide (Stark et al., 1977). The trisaccharide forms part of the teichuronic acid linkage unit. The rfe-gene product also participates with products of the rff genes in assembly of ECA. Biosynthesis of the repeating unit of ECA involves a pathway resembling that involved in formation of undecaprenol-linked intermediates in 0-polysaccharide synthesis in S . enterica serogroup B (Rick et a f . , 1985, 1988; Barr and Rick, 1987; Barr et al., 1989; Meier-Dieter et al., 1990, 1992). Many of the genes participating in ECA synthesis have now been cloned and biosynthetic activities of the products identified

C I - I I SURFACE POLYSACCHAHIDI-S IN GRAM NEGAllVE BALTCRIA

163

(Meier-Dieter et al., 1990, 1992; Ohta et al., 1991). These studies clearly established that an rfe mutant lacks UDP-N-acetylglucosamine: :undecaprenylphosphate N-acetylglucosamine-l-phosphate transferase activity. Consequently, rfe strains are unable to form lipid I in the initial step of ECA assembly (Meier-Dieter et al., 1992). The reducing terminus of ECA is a residue of N-acetylglucosamine, part of the repeating unit structure (Barr and Rick, 1987). Multicopy rfe results in elevated N-acetylglucosamine transferase activity, consistent with rfe being the structural gene for the transferase. However, it is difficult to reconcile this activity with an effect o n biosynthesis of rfe-dependent O-polysaccharides, which, in most cases, lack N-acetylglucosamine residues. While it is conceivable that Rfe is a bifunctional transferase with relaxed specificity, it is also possible that Rfe acts in a regulatory capacity (Jann and Jann, 1984; Ohta et al., 1991; Meier-Dieter et al., 1992). The Rfe protein does have potential membrane-spanning domains but has no obvious consensus sequences for binding either DNA or dolichol (Ohta et al., 1991; MeierDieter et al., 1992). a-Glucosyldiphosphorylundecaprenolis absent from membranes prepared from rfe mutants of E. coli 0 9 . Synthesis of the mannan was restored in inactive (Rfe-) membranes by in vitro reconstitution using functional carrier extracted from Rfe' cells (Goldemann et al., 1979; Jann et al., 1979; Kanegasaki and Jann, 1979). However, undecaprenolPP-Glc can be produced in membranes from an rfe-defective E. coli 0 9 strain providing unusually high (and non-physiological) concentrations of UDP-glucose and magnesium ions are used in the reaction (Jann et al., 1982). The significance of this observation is unclear. The Rfe protein only acts in formation of undecaprenol-PP-Glc in synthesis of the 0 9 polymer and does not seem to influence mannosyltransferases (Jann et al., 1982). It is not clear how the regularity of the repeating-unit structure and sequence of linkages is maintained by the sequential process in 0 9 polymerization. It has been suggested that the 0 8 / 0 9 repeating units are established by the specificity of a complex of mannosyltransferases, which act in a co-ordinated fashion and are sensitive to fluctuations in the concentrations of available precursors (Jann and Jann, 1984). This model eliminates a formal requirement for a polymerase as such, but does require several different transferase activities. Although one might predict that one transferase would be required for formation of each linkage in the 0polysaccharide, there are several recent reports of transferases with more than one activity. The first bifunctional transferase described in bacteria was the KDO transferase of E. coli,required for biosynthesis of LPS core oligosaccharide (Clementz and Raetz, 1991). This enzyme catalyses formation of two linkages, whereas the KDO transferase (GseA) from Chlamydia trachomatis is capable of forming three linkages (Belunis et al., 1992). The

164

C . WIIIT1~IHI.I~ AN11 M. A . VAI.VANO

polysialyltransferase from E. coli K92 (Steenbergen et al., 1992; Vimr et al., 1992) is a bifunctional enzyme involved in assembly of K92 CPS (see Section III.C.3). It remains to be established whether transferase enzyme(s) involved in biosynthesis of the 0 9 polysaccharide is also multifunctional. The 09-polymer biosynthesis (rfb) gene cluster has been cloned (Kido el al., 1989), and a precise examination of the number of biosynthetic enzymes is underway (N. Kido, personal communication).

3. Polymerization of Group-II Capsular Polysaccharides in Escherichia coli The group-I1 CPSs of E. coli K1 (Kundig et al., 1971; Rohr and Troy, 1980) and K5 (Finke et al., 1991) grow at the non-reducing terminus. With K7 (group-TI) CPS, the polymerization mechanism is rfe-independent (Meier-Dieter et al., 1990). In vitro synthesis of polysialic acid by membrane preparations from E. coli K1 was first reported by Aminoff et al. (1963) and has been the subject of intense study at both genetic and biochemical levels (reviewed in Troy, 1992). Polysialyltransferase activity elongates a natural endogenous acceptor within the membrane or, alternatively, an exogenous acceptor consisting of polysialic acid (Kundig et al., 1971). A single polysialyltransferase is responsible for both reactions in E coli (Steenbergen and Vimr, 1990), and the same may be true for synthesis of polysialic acid in N. meningitidis type B (Frosch et al., 1989). The polysialyltransferase in E. coli K1 appears to operate in a processive fashion and forms a(2-+8)-linked polysialic acid (Steenbergen and Vimr, 1990; Steenbergen et al., 1992). Replacing the polysialyltransferase component of the K1 biosynthetic complex with the K92 polysialyltransferase results in the formation of K92 polymer (Steenbergen et al., 1992). Since K92 polymer has alternating a-(2-+8)la-(2+9) linkages, K92 polysialyltransferase must have dual-linkage specificity. Polysialyltransferases from different bacteria recognize specific exogenous acceptors. For example, the enzyme in E. coli K92 transfers sialic acid residues to an acceptor consisting of K92 polymer, but not to a K1 a-(2-+8) polymer (Vann and Troy, 1986). The CPS from N. meningitidis group C is a polysialic acid polymer with a(2-+9) linkages. The group-C polysialyltransferase elongates both K92 and group-C exogenous acceptors, but does not recognize the a-(2-+8)-linked polymer as an acceptor (Vann el al., 1978). The reaction which initiates synthesis of polysialic acid has not been resolved, but polysialyltransferase by itself is not sufficient for initiation (Steenbergen et al., 1992). The initial sialic acid residue is added to an acceptor molecule which does not contain one of these residues (Rohr and Troy, 1980). Although undecaprenol monophosphate-linked sialic acid residues have been isolated (Troy et al., 1975) and shown to transfer

CE1.I.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

165

sugar residues to both endogenous and exogenous acceptors (Troy and McCloskey, 1979), it is not clear whether growth of the polymer occurs while attached to undecaprenol. An alternative mechanism has undecaprenol phosphate acting as a carrier of single sialic acid residues, or short oligomers, between CMP-sialic acid and a protein-acceptor molecule. Proteins have been identified at the reducing terminus of polysialyl polymers in E. coli K1 (Troy and McCloskey, 1979; RodriguezAparicio et al., 1988; Weisgerber and Troy, 1990) and in N. meningitis group C (Vann et al., 1978). The protein moiety may play a role in either polymerization or export, or both. It is also possible that the acceptor for growing polymer is an initiase (Weisgerber and Troy, 1990), with polysialyltransferase catalysing subsequent elongation of polysialic acid (Steenbergen et al., 1992). Complexes containing functional NeuE protein but lacking polysialyltransferase incorporate low levels of sialic acid into high molecular-weight material (Steenbergen et al., 1992). This activity, together with the observation that NeuE contains membrane-spanning domains and a dolichol-binding consensus sequence (Steenbergen et a l . , 1992; Troy, 1992), suggested that NeuE may be involved in initiation of polysialic acid in E. coli K1. However, subsequent results demonstrated that intracellular polysialic acid can be synthesized in strains with a defined neuE defect, thereby precluding NeuE as an initiase (Vimr and Steenbergen, 1992). Termination of polymerization has been proposed as an alternative function for NeuE (Vimr and Steenbergen, 1992); meanwhile, the initiase awaits identification. Biosynthesis of a second group41 CPS has recently been studied in E. coli K5. Unlike K1 polymer, no evidence was obtained in support of undecaprenol-linked intermediates in K5-polysaccharide biosynthesis (Finke et al., 1991). Biosynthesis of K5 polymer is not inhibited by bacitracin in vitro, although the same is also true for synthesis of CPSs in E. coli K1 (Troy et al., 1975) and N. rneningitidis group C (Vann et al., 1978) and in the 0 9 polysaccharide of E. coli (Kopmann and Jann, 1975). These latter processes d o involve undecaprenol-linked intermediates. In addition, biosynthesis of K5 polymer was not stimulated by exogenous polyprenol lipid and no radiolabelled intermediates were extractable with lipid solvents. These results could suggest that an alternative mechanism operates in synthesis of CPS in E. coli K5, although this would be surprising given the conservation in the remaining biosynthetic components for group41 CPSs from E. coli (see Sections IV.B, 1V.Cand V.C). The polymer from strain K5 is an example of a group-I1 CPS with KDO at the reducing terminus (Finke et al., 1991) and this may be reflected in the biosynthetic mechanism. Most E. coli strains which produce group-I1 CPSs contain elevated levels of CMPD O synthetase activity when grown at 37°C; at 18°C no group-I1 CPSs are

166

C. WHITFIELD AND M. A. VALVANO

synthesized and CMP-KDO synthetase activity is lowered to levels detected in unencapsulated bacteria (Finke et al. , 1990). In temperature-upshift experiments involving E. coli K5, elevation of CMP-KDO synthetase activity is correlated with the appearance of K5 CPS (Finke et al., 1989). This led Finke er al. (1991) to propose that polymer synthesis is initiated by transfer of KDO to a carrier molecule, rather than KDO being added with the phospholipid anchor in a postpolymerization process. The KDOsubstituted carrier could then serve as an acceptor for the polymer, resulting in formation of a labile undecaprenol-linked intermediate. Alternatively, in virro polymerization of K5 CPS may proceed sequentially at a high rate, rapidly forming high molecular-weight undecaprenol-linked intermediates which are not readily extracted. Interestingly, K5 polysaccharide differs in only one linkage from the hyaluronic acid CPS produced by group-A streptococci (Kenne and Lindberg, 1983). Synthesis in vitro of streptococcal hyaluronic acid has been examined and no evidence of intermediates was obtained (for a review, see Markovitz, 1977). It remains to be established whether these similar structures are assembled by similar synthetic pathways. 4. Undecaprenol-Independent Polymerization Mechanism

Biosynthetic pathways involving undecaprenol have received much attention, and the depth of available information is reflected in the account already given. However, polysaccharide-biosynthetic systems which function by alternative mechanisms have also been reported. In synthesis of bacterial alginate by Azotobacrer vinelandii and Pseudomonas aeruginosa, no undecaprenol-linked intermediates have been identified by conventional techniques (Sutherland, 1982). Also, P-glucan synthesis by diverse bacteria unequivocally occurs without participation of isoprenoid lipid carriers. All of these polymers lack significant attachment to the cell surface. Alginate is a linear copolymer comprised of D-mannuronic acid and its 5' epimer D-guluronic acid, and is found in P. aeruginosa and Az. vinelandii, as well ascertain brown algae (May eral., 1991). In algae, GDP-mannuronic acid and GDP-guluronic acid are produced (Lin and Hassid, 1966a,b), and it is logical to assume that these serve as precursors for alginate synthesis. In contrast, alginate in Az. vinelandii is initially synthesized as a homopolymer of mannuronic acid residues (Pindar and Bucke, 1975); the polymerization process is uncharacterized. An extracellular calciumactivated C-5 epimerase is believed to convert some of the mannuronic acid residues to guluronic acid residues in a postpolymerization modification (Haug and Larsen, 1971; Larsen and Haug, 1971; Skjak-Braek and Larsen, 1985). The critical role played by alginate in pathogenesis of P. aeruginosa

CF.I.1.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

167

in the cystic-fibrosis lung has led to intense efforts to resolve the mechanism of its synthesis. Remarkably, despite these efforts, most of the steps remain unclear; the status of alginate synthesis by P. aeruginosa has been the subject of a recent comprehensive review (May et al., 1991). GDPmannuronic acid also serves as a precursor for alginate in P. aeruginosa, but the source of guluronic acid residues and the possible involvement of an epimerase have not been established (May et a f . , 1991). Similarly, the mechanism by which 0-acetyl groups are introduced into the alginate (Davidson et al., 1977; Skjak-Braek et al., 1986) has also not been resolved. Since these modifications may affect the physical properties of the alginate, they may have a substantial bearing on the role of the polymer in virulence (May et al., 1991). However, alginate provides an example of a biosynthetic system in which molecular-genetic approaches may resolve questions that have proved intractable to biochemical study. For example, it has been shown that an algC mutation in P . aeruginosa results in production of alginate lacking guluronic acid residue (Chitnis and Ohman, 1990). This mutation therefore identifies a component of the epimerization mechanism. Other genes, identified by the mutations ale-8, alg-44 and alg-76 (algE; Chu et al., 1991), form a biosynthetic gene cluster (Wang et a l . , 1987). The proteins Alg-8 and Alg-44 appear to be membrane bound, and are thought to play a role in polymerization and export of alginate (May et a l . , 1991). An alginate lyase (AlgL) which depolymerizes the polymer is found in P. aeruginosa (Linker and Evans, 1984; Dunne and Buckmire, 1985). The algL gene is located in the alginate-biosynthesis gene cluster (Schiller, 1992), suggesting a role for alginate lyase in biosynthesis. However, the precise function of the lyase in the physiology of P. aeruginosa remains to be established. As the components of the biosynthetic complex are systematically identified, and the structures of the gene products are resolved, the pathway should become amenable to biochemical analysis. Bacterial P-glucans are synthesized by pathways which do not involve undecaprenol-linked intermediates (Ross et al., 1991). The best-characterized example is bacterial cellulose, and the synthetic process has been the subject of a detailed recent review (Ross et al., 1991). Bacterial cellulose is synthesized from UDP-glucose by cellulose synthetase in Acetobacter xylinum (Swissa et a l . , 1980; Valla et a l . , 1989). Alkali-insoluble j3-(1+4)linked glucan is synthesized by a processive polymerization of glucose residues, and nascent cellulose remains attached to the synthetase during polymerization. Cellulose synthetase has been solubilized and the active form has been shown to be a complex of 420 kDa (Aloni et al., 1982) containing both catalytically active and inactive polypeptide subunits. Molecular analysis indicates that the catalytic subunit is a transmembrane protein, BcsB (bcs, bacterial cellulose synthesis; Wong et al., 1990). Three

168

C WIIITPIEI.I) AND M. A VALVANO

other components, BcsA, BcsC and BcsD, are also required for cellulose synthesis in vivo, but their role in the process has not been established (Wong et al., 1990). Cellulose synthetase is probably located in the cytoplasmic membrane (Amikam and Benziman, 1989; Bureau and Brown, 1987). Many plant-associated bacteria produce p-( 1+2)-glucans, which are cyclic polymers containing between 17 and 40 glucose residues (Koizumi et al., 1984; Zevenhuizen et al., 1990). The p-(1-+2)-glucan synthetase (NdvB) of R. rneliloti is a 319 kDa cytoplasmic-membrane protein with several transmembrane domains (Ielpi et af., 1990). Rhizobium fredi appears to contain a similar NdvB protein (Bhagwat ef al., 1992) while a homologue, ChvB, is produced by Agrobacterium sp. (Zorreguieta and Ugalde, 1986). The proteins NdvB and ChvB form stable intermediates when labelled with UDP-[ I4C] glucose, indicating that polymer synthesis may involve a processive mechanism similar t o that operating in synthesis of bacterial cellulose. Similarities in the p-( 1-+2)-glucan biosynthetic systems also extend to polymer export (see Section IV.B.2). Levans are homopolymers of fructose that are produced by Erwinia amylovora (Gross et al., 1989) and Aerobacrer levanicum (Sutherland, 1982). Although levan synthesis has been studied in some detail in Gram-positive bacteria (Sutherland, 1982), these polymers have received little attention in Gram-negative organisms. In Gram-positive bacteria, levans are synthesized by an extracellular enzyme, and this location by itself indicates a considerably different biosynthetic mechanism. Synthesis requires sucrose or a similar oligosaccharide as the glycosyl donor; neither sugar nucleotides nor undecaprenol-linked intermediates are involved in the process. Levansucrase splits sucrose and polymerizes fructose residues to form levans. C. POLYSACCHARIDE-MODIFICATION REACTIONS: ADDITION OF SIDE-CHAINS AND SUBSTITUENTS

In general, EPS and LPS 0-polysaccharides show regularity in their carbohydrate structures. In some exceptional cases, addition of polymer side-chains is not uniform. Addition of non-carbohydrate substituents can also occur with regularity, but in most cases the modifications are nonstoichiometric. The stoichiometry of substitution may reflect the stage in biosynthesis at which substitution occurs. Where the mechanisms have been described, side-chain additions and substitution reactions occur at the level of undecaprenol-linked intermediates. Examples of both pre- and postpolymerization substitutions reactions have been reported. In polysaccharides in enteric bacteria, the addition of polymer side-chains

CE1.I.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

169

is essential for polymerization. In vitro synthesis of the 0-plysaccharide of S . enterica serovar anaturn (group E) requires formation of the trisaccharide repeating unit (Man-Rha-Gal; see Fig. 3(a)); in the absence of the final mannose residue the backbone is not completed and no polymer is formed (Wright et al., 1965). The same trisaccharide 0-polysaccharide backbone in S . enterica serovar typhirnuriurn (group B) is not polymerized in vivo unless the abequose side-chains are added. Mutants which are unable to produce the precursor CDP-abequose d o not produce any 0polysaccharide (Yuasa et al., 1969). Optimal in vitro polymerization of the serogroup-B 0-polysaccharide also requires addition of the abequose sidechain. However, low levels o f polymerization of the trisaccharide were detected in vitro (Weiner et al., 1965; Osborn and Weiner, 1968), possibly reflecting relaxed specificity of the polymerase as a consequence of membrane perturbation (Osborn and Weiner, 1968). These results serve as a reminder that some in vitro data should be interpreted with caution. During synthesis of CPS by A. aerogenes, there is no in vitro polymerization of trisaccharides lacking the glucuronic acid side-chain (Troy et al., 1971). One reason for the different in vitro data between this and the group-B 0-polysaccharide system is apparent from the synthetic schemes shown in Fig. 3. During synthesis of CPS by strain DD45 of A. aerogenes, addition of the final galactose residue in the trisaccharide backbone is dependent on the preceding transfer of a glucuronic acid residue; the completed trisaccharide backbone is required for polymerization. In contrast, an abequose residue is added after formation of the trisaccharide backbone of the 0-polysaccharide in S . enterica group-B strains, potentially providing an alternate polymerization substrate. 0-Polysaccharides from S . enterica are modified non-stoichiometrically with 0-acetyl and glucosyl residues. The glucosylation reaction is the better characterized. a-Glucosylmonophosphorylundecaprenol is the direct donor for glucosylation, and this unusual lipid intermediate is synthesized by a membrane-associated enzyme (Nikaido and Nikaido, 1971; Takeshita and Makela, 1971; Makela, 1973). In S . enterica serovars typhirnuriurn and anatum, glucosyl residues are transferred to polymeric 0-antigen and it is assumed that the reaction occurs at the 0hapten level (Nikaido et al., 1971; Takeshita and Makela, 1971; Wright and Kanegasaki, 1971; Sasaki et al., 1974). Salmonella enterica serogroups C2 and C3 have similar glucosylation reactions using aglucosylmonophosphorylundecaprenol. However, in these bacteria, the substrate for in vitro glucosylation is unequivocally a single 0-repeating unit attached to undecaprenol pyrophosphate (Shibaev et al., 1979). The difference between the pathways used in serogroups C2/C3 and B may reflect the site of glucosylation. In serogroups C2 and C3, modification occurs o n the first (C3) or second (C2)sugar residue of the

170

C' WIil'r1~IEI.D A N D M. A . VALVANO

repeating unit whereas, in other examples, the site of modification is only available after polymerization (Wright and Kanegasaki, 1971). 0Acetylation occurs in the 0-polysaccharides of S . enterica serovar anatum (El; see Table 3). The acetyltransferase activity is membrane associated and uses acetyl-CoA as a donor. In vitro, a transacetylase modifies the 0hapten but not single repeating units attached to undecaprenol (Keller, 1966). However, this does not reflect the process in vivo because semirough LPS, containing a single repeating unit of 0-antigen, carries acetyl groups (Makela, 1966). Addition of 0-acetyl and pyruvate groups occurs at the lipid-intermediate stage during biosynthesis of the CPS in K aerogenes type 8 (Sutherland, 1977b). Optimal activity of the galactosyltransferase which forms the undecaprenylpyrophosphoryl (glucose-galactose) intermediate requires acetyl-CoA and phosphenolpyruvate, donors for 0-acetyl and pyruvate groups, respectively. In the presence of fosfomycin, a structural analogue of phosphoenolpyruvate, transfer of galactosyl residues was diminished. As expected, decreased formation of the undecaprenol-linked disaccharide influenced formation of the complete repeating unit and its subsequent incorporation into the polymer. These data suggest that, in this system at least, formation of the complete repeating unit, including non-carbohydrate substituents, is a prerequisite for polymer formation. An interesting unanswered question concerns assembly of repeating units of CPS, such as that in Kfebsieffaspp. K54, in which 0-acetyl groups occur on alternate repeating units, or that in Kfebsieffaspp. K70, which has alternate repeat units substituted by pyruvate (Table 5 ) . These structures presumably reflect the specificity of the modifying enzymes. It is not clear whether the biosynthetic repeating unit of these polysaccharides comprises two carbohydrate repeat units. 0-Acetylation also occurs in polysialic acid-containing group 11-like CPSs in which the donor is acetyl-CoA. In N. meningitidis group C, a membrane-bound 0-acetyltransferase modifies preformed oligosaccharides and acetylation is clearly a postpolymerization modification (Vann et a f . ,1978). The 0-acetyltransferase in E. coli K1 has been partially purified and also acts on preformed polymer containing more than 14 residues (Higa and Vaarki, 1988). Interestingly, 0-acetylation of the CPS from E. cofi K1 is switched on and off in a phase-variation process. Synthesis of EPS in X . campestris provides a most flexible polymerization system (see Fig. 4). It has been recognized for some time that the numbers of pyruvyl and 0-acetyl groups present in xanthan gum are dependent on growth conditions (for a review, see Sutherland, 1981). Also, mutants of X. campestris have been isolated in which the xanthan gum is nonpyruvylated (Smith et a f . , 1981a; Hassler and Doherty, 1990; Marzocca et a f . , 1991), or in which some of the trisaccharide side-chains are missing

CEI.1.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

171

(Whitfield et al., 1981). The ability to produce partial xanthan-gum structures is explained by genetic and biochemical studies on xanthan-gum biosynthesis (Ielpi et al., 1981a,b; Vanderslice et al., 1989). Vanderslice et al. (1989) cloned and sequenced the entire cluster of xanthan biosynthetic (gum) genes and introduced non-polar mutations into each of the transferase open-reading frames (ORFs). Polymers were extracted from cultures of the mutants and their structures determined. Biochemical defects in biosynthesis were established in vitro, and studies with mutant strains confirmed earlier analysis of the wild type (Ielpi et al., 1981a,b). Using a mutation affecting the first of the two mannosyltransferases (see Fig. 4), it was found that the lipid-linked cellobiose intermediate is not polymerized in vitro (Ielpi et al., 1981a; Vanderslice et al., 1989). Consequently, mutant strains of X. campestris are unable to produce a polymer consisting of cellulose (the xanthan-gum polysaccharide backbone) in vivo (Vanderslice et a l . , 1989). Bacteria harbouring mutations which eliminate the glucuronosyltransferase can produce a lipid-linked trisaccharide intermediate. Surprisingly, this truncated intermediate is polymerized at low levels in vitro (Ielpi et al., 1981a; Vanderslice et al., 1989). In vivo, these mutants produce a xanthan-based polymer with a trisaccharide structure; the amounts produced are 1-3% of that in the wild type (Vanderslice et al., 1989). Both the tetrasaccharide (lacking the terminal mannose residue of the side-chain) and the pentasaccharide repeating units are synthesized and polymerized efficiently in vivo and in v i m , without any requirement for addition of 0-acetyl and pyruvate groups (Ielpi et al., 1981a,b; Vanderslice et al., 1989; Marzocca et al., 1991). The molecular basis for the apparently relaxed specificity of the xanthan-gum polymerase remains an intriguing question for further investigation. Interestingly, genetic approaches which have helped to elucidate the biosynthetic mechanism may facilitate manipulation of the structure (and physical properties) of this industrially important polymer.

IV. Export of Polysaccharides and Cell-Surface Assembly A. LOCATION OF BIOSYNTHETIC COMPLEXES AT THE CYTOPLASMIC MEMBRANE

It is generally assumed that assembly of polysaccharide repeating units and subsequent polymerization reactions occur at the cytoplasmic membrane, using precursors synthesized in the cytoplasm. This has been confirmed by a number of membrane-fractionation studies, beginning with the elegant work from Osborn’s laboratory (Osborn et al., 1972b,c). Involvement of

172

C.

W I ~ I W I F I DA N D M A V A I V A N O

undecaprenol provides an obligatory requirement for the membrane in many systems. Sequential assembly of bacterial polysaccharides has led to the belief that the enzymes function in a co-ordinated complex with undecaprenol sequestered in the active complex (Anderson et al., 1972). A bacterial cell contains a limited amount of undecaprenol phosphate (10’ molecules in each cell; Wright and Kanegasaki, 1971). In vitro synthesis of LPS O-polysaccharide can be achieved in membrane preparations but not with solubilized enzymes. Glycosyltransferases have been solubilized usingdetergents (Osborn etal., 1972a; Shibaev, 1978; Danilov and Shibaev. 1991) and retain the ability to catalyse individual reactions in the formation of undecaprenol-linked intermediates. However, the integrity of the complex is affected and formation of polymerized repeating units does not occur. Precise localization of the biosynthetic complexes has been hampered by difficulties in purification of reasonable amounts of biosynthetic enzymes. With molecular-biology approaches, it is now possible to identify the enzymes and obtain information about the organization of the biosynthetic complexes. Interestingly, most transferases characterized to date do not show properties expected of transmembrane or integral membrane proteins. In O-polysaccharide biosynthesis by Salmonella enterica serovar typhimurium (group B), RfbP (the initial galactosyltransferase, see Fig. 3(a) contains five potential membrane-spanning segments (Jiang el al., 1991). Since this enzyme possibly interacts with undecaprenol, an intimate association with the membrane might be predicted. Other transferases have not yet been identified, but the Orf12.8 polypeptide is the only other possible integral membrane protein encoded by the rfb (O-polysaccharide biosynthesis) gene cluster. The protein O r f l 2 . 8 ~has 12 transmembrane segments but its function is not known (Jiang et al., 1991). Several of the remaining Rfb proteins have one or two potential hydrophobic segments, but many of them are cytoplasmic enzymes involved in precursor synthesis. These results suggest that the functional biosynthetic complex may be formed primarily from proteins which are loosely associated with the membrane. The polysialyltransferase enzymes involved in biosynthesis of polysialic acidcontaining CPS in Escherichia coli K1 and K92 and Neisseria meningitidis B also show no significant membrane-spanning domains (Frosch et a l . , 1991; Weisgerber et al., 1991; Steenbergen et al., 1992; Vimr et a l . , 1992). More importantly, it has been shown that the polysialyltransferase from E. coli K1 is sufficiently loosely membrane-associated that it can be transferred during mixing of membranes from polysialyltransferase-positive with polysialyltransferase-negative strains (Steenbergen et al., 1992). This property facilitates reconstitution of the complex in vitro.

CkiI.I.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE R A C T F R I A

173

How are the transferases associated with the membrane? It is possible that they interact directly with the membrane, since most transferases are basic proteins. Alternatively, transferases may interact with another protein. The protein O r f 1 2 . 8 ~ provides a potential candidate for interaction with 0-polysaccharide-biosynthesis(Rfb)proteins in S . enterica serogroups A , B and D (Jiang et al., 1091); an Rfb protein with similar secondary structure is found in serogroups El ( O r f 7 . 9 ~ Wang ; et al., 1992) and C2 ( O r f 1 2 . 6 ~Brown ; et af., 1992), and in E. coli 0 7 (C.L. Marolda and M.A. Valvano, unpublished observations). Troy (1992) suggested that undecaprenol may also provide a scaffold molecule to organize the biosynthetic complex. Two other proteins in the polysialic acid-biosynthesis complex from E. coli show consensus dolichol-binding domains; these are NeuE (Weisgerber et al., 1991; Steenbergen et al., 1992; Troy, 1992) and KpsM (Troy, 1992). Several lines of experimental evidence suggest that the biosynthetic complexes are located at the cytoplasm-cytoplasmic membrane interface. It is clear that sugar-nucleotide precursors are synthesized in the cytoplasm and, in the absence of detectable transport (export) systems that take these precursors across the cytoplasmic membrane, it is reasonable to assume that at least part of the biosynthetic process proceeds from the cytoplasmic face of the inner membrane. In membrane vesicles prepared from S . enterica serogroup B, UDP-galactose is only accessible to the initial galactosyltransferase (RfbP) when supplied at the cytoplasmic face of the membrane (Marino et al., 1991). The orientation of CPS transferase complexes in membrane vesicles of E. coli K1 (Troy, 1992) and K5 (Finke et al., 1991) has been assessed using membrane-impermeable probes and membrane vesicles with defined orientations. In both cases, the results supported location of at least some of the biosynthetic complex on the cytoplasmic face o f the membrane. Access to energy-generating processes is also presumably simpler with this orientation. With this membrane topology, it is obvious that surface assembly requires a mechanism to transport nascent polysaccharide across the cytoplasmic membrane, and two possible mechanisms have been proposed from available data. In rfe-independent 0-polysaccharide biosynthesis in S. enterica, the undecaprenol lipid cycle (Fig. 6(a)) may function as a transmembrane-assembly process, delivering completed polysaccharide to the periplasmic face of the membrane. In synthesis of group-I1 CPSs in E. coli (Fig. 6(b)), assembly occurs at the cytoplasmic face and a dedicated membrane-transport system translocates the nascent polymer across the cytoplasmic membrane. Evidence for these models is discussed in the following sections.

174 1 OUTER

Translocation and surface sly

PERIPLASM

~

~

Transfer and polyrnenzation Daphosphorylation and lipd recycling

CYTOPLASMIC

MEMBRANE

r/7 unit - Repeating synthesis

& I

CYTOPLASM

PRECURSORS

OU JER MEMBRANE

R

-

R

...._LI

Interaction with pariplasmic export

PERIPLASM

Repealing unit synthesis and polymerization possibly involving undecaprenol

J r a n s l mition and surfs,L x assemoiy

reaction

A JP-binding

CYTOPLASMIC MEMBRAh'E

cassene

J FIG. 6. Models for the topology of cell-surface polysaccharide biosynthesis in Gram-negative bacteria. The pathways are based on data for (a) rf'eindependent 0-polysaccharide biosynthesis in Salmonella enterica and (b) group-I1 capsular-polysaccharide biosynthesis in Escherichia coli. The solid circles signify repeating units of polysaccharide. Undecaprenol and phospate are represented by Cssand P, respectively.

Cf-LI S U R E A C F POI YSACCHARIDES IN G R A M Nf-GATIVF B A C T RIA

175

H. TRANSPORT ACROSS THE CYTOPLASMIC MEMBRANE

1 . Transmembrane Assembly of rfe-Independent 0-Polysaccharides of

Salmonella enterica Several key observations led to the proposal of a transmembrane-assembly system in assembly of 0-polysaccharides in S. enterica (McGrath and Osborn, 1991a). Undecaprenol-linked polymeric 0-polysaccharide intermediates (0-haptens) accumulate as products of in vitro synthetic reactions in S . enterica serogroup B (Kent and Osborn, 1968b). A precursor-product ) in vitro relationship was established in vivo (Kent and Osborn, 1 9 6 8 ~and (Kent and Osborn, 1968b) between 0-hapten and lipid A-core-linked 0polysaccharide. Subsequent immuno-electron microscopy studies localized stable accumulated 0-hapten molecules at the periplasmic face of the cytoplasmic membrane in a rough mutant of S. enterica serovar typhimurium (Mulford and Osborn, 1983). Preformed periplasmic 0-hapten could be ligated to nascent lipid A-core in conditional mutants (McGrath and Osborn, 1991a). Assuming that all preformed 0-hapten molecules in these cells are located at the periplasmic face of the cytoplasmic membrane, ligation of 0-polysaccharide to lipid A-core must occur on the periplasmic face (McCrath and Osborn, 1991a). Location of 0-hapten molecules at the periplasmic face of the cytoplasmic membrane also indicates that some 0-polysaccharide-modification reactions must occur at this location. Presumably, this is why the donor for glucosylation is aglucosylmonophosphorylundecaprenol, rather than UDP-glucose (see Section 1II.C). Components of the 0-polysaccharide ligase are identified by two mutations, namely rfaL and rfbT (Makela and Stocker, 1984). Little is known about the rfbT component, or indeed whether the rfbT mutations identify a single gene. The rfuL gene has been cloned from both E. coli K-12 (Klena et a l . , 1992) and S. enterica serovar typhimurium (MacLachlan et a l . , 1991). Defects in rfaL result in R-LPS and accumulation of 0-hapten in S . enterica. The attachment process for ECAIJPSalso requires participation of RfaL, suggesting that the mechanisms for attaching ECAI.PS and 0-polysaccharide are similar (Kuhn et al., 1988). The RfaL proteins of S . enrerica serogroup B (MacLachlan et al., 1991) and E. coli K-12 (Klena et al., 1992) have very similar hydropathy plots; both proteins have several transmembrane segments. Potential interaction of RfaL with undecaprenollinked intermediates is suggested by a putative dolichol-binding consensus sequence in the C-terminal region of the protein (Klena et al., 1992). The observation that heterologous 0-antigens are expressed in E. coli and S . enterica (see Section V.B) indicates that RfaL function is independent of

176

C WIIITIYFI I) A N D M A V A I V A N 0

polysaccharide structure. However, the efficiency of RfaL and its possible interaction with heterologous RfbT proteins may vary considerably (Klena el ul., 1992). Based on accumulated data, McGrath and Osborn (1991a) proposed that LPS assembly in serogroup B of S. enterica is a transmembrane event. I n Fig. 6(a), this is depicted with transfer of sugar residues to undecaprenol phosphate occurring at the cytoplasmic face of the cytoplasmic membrane, followed by translocation of undecaprenol-linked intermediates across the membrane. Since both undecaprenol-linked polymers and undecaprenollinked single repeating units (in rfc mutants) of 0-polysaccharide are efficiently linked to lipid A-core, both must be available for the ligation reaction at the periplasmic face. This has led to speculation that the polymerization reaction also occurs at the periplasmic face of the cytoplasmic membrane, providing a model in which only lipid-linked single repeating units must traverse the cytoplasmic membrane (McGrath and Osborn, 1991a). However, the authors concede that their data are also consistent with the alternative possibility that polymerization occurs at the cytoplasmic face, with lipid intermediates containing both single repeating units and 0-polysaccharide being translocated across the membrane. In either scenario, details of the translocation process are unclear. Kanegasaki and Wright (1970) have argued that undecaprenol-linked intermediates containing a single repeating unit are capable of lateral and transmembrane movement. However, studies performed with purified polyprenol lipids in model phospholipid bilayers indicate that the rate of transbilayer transposition is too slow to account for observed rates of polysaccharide synthesis (McCloskey and Troy, 1980). It is conceivable that a specific protein (which would be absent from model membranes) mediates transposition of lipidlinked intermediates across the membrane. To date, no cytoplasmic membrane-transport system has been identified for LPS. This type of transport protein would presumably be a transmembrane protein. If the transport protein is an rfb-gene product, O r f l 2 . 8 ~and its homologues are the only candidates in S. enterica (Jiang el al., 1991; Brown et al., 1992; Wang el al., 1992). Interestingly, the predicted structure of Rfc contains multiple membrane-spanning domains and, in many respects, resembles porin proteins (Collins and Hackett, 1991). This has led to speculation about a role for Rfc in translocation of 0-polysaccharides across the cytoplasmic membrane. If this is correct, the SR-LPS phenotype of rfc mutants can only be explained by a mechanism in which translocation of single repeating units attached to undecaprenol is +independent. The Rfc protein would only be required for translocation of undecaprenollinked polymers. This model accounts for the observed rfc phenotype while formally precluding for it a role as a polymerase. It remains to be established

CEI.1.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACI'ERIA

177

how wide is the disribution of the transbilayer-assembly model. Based on general similarities in the biosynthetic pathways, it is tempting to speculate that CPS assembly in Aerobacter aerogenes DD45 uses this system.

2. A Cytoplasmic-Membrune Transport System for Group-It-like Capsular Polysaccharides The biochemistry of the assembly pathway for synthesis of CPSs in E. coli K1 and K5 has been studied using defined mutations in cloned genes. Although it is conceivable that use of recombinant multicopy plasmids may introduce some artefacts which might not be apparent in mutations introduced o n to the chromosome, these experiments do provide a foundation for further study. Mutations in region 3 of the kps (group-I1 CPS biosynthesis) gene cluster in E. coli K5 eliminated formation of a capsular layer and resulted in cytoplasmic accumulation of the polysaccharide (Kronke et al., 1990a). In one mutant, the accumulated polymer was of lower molecular weight than the wild-type polysaccharide, although another mutant accumulated polymer with a normal chain length (Jann and Jann, 1992). The polymers in both mutants lacked the phospholipid anchor. Similar mutations have been identified in the kps genes of E . coli K1 (Boulnois et al., 1987; Pelkonen, 1990) and full-length chains of the polysaccharide have been detected in the cytoplasm of this bacterium (Troy et al., 1990). These results clearly reflect an orientation of the membrane-bound biosynthetic complex directed towards the cytoplasm, without a transmembrane topology. Furthermore, as with the 0polysaccharide of S. enterica serogroup B, coupling of synthesis and polymerization to export is not obligatory. This membrane topology is logical given the direction of polymer growth in synthesis of group-I1 CPSs in E. coli. Growth occurs at the non-reducing end of the polymer, that is, at the distal end from the undecaprenol (or other) carrier and the end which ultimately is distal to the cell. It was more difficult to reconcile this direction of chain growth using a membraneassociated transferase complex, if polymerization occurred at the periplasmic face. In contrast, growth of the 0-polysaccharide in S . enterica serogroup B occurs at the reducing end, nearest to the undecaprenol carrier. Proximity of the growing terminus to the cytoplasmic membrane can therefore be maintained during polymerization at the periplasmic face. It will be interesting to see if other polysaccharides with similar directions of polymer growth have similar assembly topology. Completion of polymerization at the cytoplasmic face of the membrane results in a requirement for a postpolymerization export mechanism. Similar export components have been identified in diverse bacteria, to transfer structurally different polysaccharides

178

C'

WtllTl~lElI> A N D M A V A I V A N 0

assembled by different polymerization mechanisms across the cytoplasmic membrane. Mutations in region 3 of the kps gene cluster in E. coli presumably define components which are required for translocation across the cytoplasmic membrane. The kps region 3 in this bacterium contains an operon of two genes termed kpsM and kpsT. The KpsM and KpsT proteins are highly conserved in serotypes K1 and K5 (Smith et al., 1990; Pavelka et al., 1991) and show homology with the ATP-binding cassette transporters defined by Higgins er al. (1986, 1990a). The proposed function of these proteins in polymer export relies heavily on this homology. ATP-binding cassette transporters are responsible for passage of diverse substrates into and out of prokaryotic and eukaryotic cells. The complex typically contains a transporter, comprised of two hydrophobic membrane proteins with five or six membrane-spanning segments, and two hydrophilic membrane proteins which couple ATP to transport. The latter can comprise a homodimer, or exist as a single protein with two domains. Periplasmic proteins or periplasmic domains are present. The KpsT protein is a peripheral cytoplasmic-membrane protein (Silver et al., 1992) with a consensus adenine nucleotide-binding fold. The same protein has been shown to bind ATP in vitro (Silver et al., 1992) while site-directed mutagenesis confirmed that the ATP-binding site is essential for function (Pavelka et al., 1991; Silver et al., 1992). Mutations in kpsT result in intracellular accumulation of polymer (Pavelka et al., 1991). The gene product KpsM is an integral membrane protein with several potential transmembrane segments (Smith et al., 1990; Pavelka et al., 1991). Conservation in region-3 genes (Boulnois and Roberts, 1990; I. Roberts et al., 1988) indicates that KpsM and KpsT functions are not influenced by the structure of the CPS. Proteins which are homologous to KpsM and KpsT are encoded by the CPS-biosynthesis gene clusters from Haemophilus influenzae type b (Bex; Kroll et al., 1990) and N . meningitidis group C (Ctr; Frosch et al., 1991). However, in these bacteria, four genes are arranged in the operon. These systems are summarized in Table 6. In H. influenzae, BexA appears to be the ATP-binding protein and it shares similar predicted secondary structure with KpsT (Smith et al., 1990; Kroll et al., 1990). The CtrD and BexA proteins show remarkable similarity, while BexB and CtrC share homology and appear to be integral membrane components; BexB has six potential transmembrane segments (Kroll et al., 1990; Frosch et al., 1991). Despite lower levels of homology, BexA and KpsT share almost identical hydropathy plots, indicating similarities in structure and, perhaps, function. Haemophilus influenzae with a mutation in bexA accumulates intracellular polysaccharide (Kroll et al., 1988). However, attempts to define clearly

TABLE 6. ATP-binding cassette transporters involved in export of polysaccharides across the bacterial cytoplasmic membrane Organism

Polymer

Components Comments

Escherichia coli

Group11 capsular polysaccharide

Haemophilur influenzae

Capsular polysaccharide

Neisseria meningitidis

Capsular polysaccharide

KpsM KpsT BexA BexB BexC BexD CtrA CtrB CtrC

Rhizobium meliloti p-( 1+2)-Glucan Agrobacterium rumefuciens &(1+2)-Glucan

CtrD NdvA ChvA

Homologues

Integral membrane protein BexB, CtrC, ChvA, NdvA ATP-binding protein BexA, CtrD ATP-binding protein KpsT. CtrD Integral membrane protein KpsM, CtrC Protein with periplasmic domainCtrB CtrA Outer-membrane lipoprotein BexD Integral membrane protein with periplasmic domain BexC Integral membrane protein with periplasmic domain BexB ATP-binding protein BexA Integral membrane protein KpsM, ChvA Integral membrane protein KpsM, NdvA

180

('

WIIITFIF.I.I) A N D

M

A

VAI VAN0

phenotypes for mutations in other bex genes have not been successful, and predicted functions are based on analogy to KpsM/KpsT and ATP-binding cassette transporters. Mutants with intracellular accumulation of CPS have been identified in N. rneningitidis group B (Frosch et a l . , 1989). In N. rneningitidis, CtrA is an outer-membrane lipoprotein which is conserved in several serotypes (Frosch el al., 1991, 1992), but its precise function in capsule expression is unknown. The significance of four proteins in the operons of H. infiuenzae and N . meningitidis, compared to two in E. coli strains with group-I1 CPSs, is not clear. Proteins similar to KpsM have been described in Rhizobiurn rneliloti (NdvA; Stanfield et al., 1988) and Agrobacterium turnefaciens (ChvA; Cangelosi et al., 1989). The NdvA and ChvA proteins are required for export of p-( 1+2)-glucans across the cytoplasmic membrane to their normal locations in the periplasm and outside the cell. The proteins ChvA and NdvA are functionally interchangeable (Cangelosi el al., 1989). Interestingly, NdvA is also remarkably similar to the HIyB protein involved in export of haemolysin in E. coli. A transporter for the similar periplasmic glucans (membrane-derived oligosaccharides) in E. coli has not yet been described, but comparison between the mechanisms of transport for these glucans and group-I1 CPSs in E. coli will be interesting. As indicated by Stanfield et al. (1988), it has not been shown whether NdvB, or the other putative transport-protein homologues already described, play a direct role in polymer export. It is equally possible that they are required for export of an intermediary component which carries the polymer. Alternatively, they could be required for an obligatory step involved in release of the nascent polysaccharide from a carrier molecule. The phenotype in each case would be cytoplasmic accumulation of CPS. Biosynthesis of group 11-like CPSs and p-( 1-+2)-glucansfollow radically different pathways (see Sections III.B.3 and 111.B.4) and offer no further clues to common functions. The step at which the phospholipid anchor is added to E. coli group-I1 CPSs has been inferred from mutant phenotypes. Cytoplasmic polysaccharide (accumulated in region3 mutants) lacks the terminal phospholipid. Bacteria mutation affecting a later stage in translocation to the cell surface accumulate with a polymer in the periplasm (see Section IV.C.l). The periplasmic polymer is lipid modified. It therefore appears that the lipidmodification step occurs either during translocation across the cytoplasmic membrane or at the periplasmic face following transport. It has been shown that KpsM has a dolichol-binding consensus sequence (Troy, 1992). This could suggest translocation of undecaprenol-linked polysaccharide across the cytoplasmic membrane, followed by addition of the diacylglycerol residue. Alternatively, KpsM may interact with undecaprenol in the formation of the active transferase complex.

W l , l -SIIRFA:ACE POI.YSA('CHARIDF.S IN GRAM-NEGATIVE BACTERIA

181

3. Energetics of transport A proton-motive force and ATP are required for translocation of both phospholipids (Donohue-Rolfe and Schaechter, 1980) and proteins (for a review, see Saier et al., 1989) across the cytoplasmic membrane, and this may be a common feature in transmembrane movement of large molecules. Several studies have described the requirement for a proton-motive force and ATP in assembly of cell-surface polysaccharides but a clear picture regarding the energization processes has not yet emerged. Assembly of t h e CPS in E. coli K1 requires a membrane potential (Whitfield et al., 1984a; Troy, 1992). It has been established that the in vitro polymerization of the CPS in E. coli K1 requires both a proton-motive force and ATP pools (Troy, 1992) but the precise nature of the energy requirement has not been resolved. An additional requirement exists for a proton-motive force and ATP pools in translocation of group-I1 CPSs across the cytoplasmic membrane, as might be predicted based on involvement of the ATP-binding cassette transporter. Inhibitors have been shown to disrupt vectorial translocation in membrane vesicles (Troy, 1992). These results are consistent with the observation that intact cells of E. coli require a proton-motive force for expression of CPSs in E. coli strains K1, K5 and K12. In the presence of cyanide rnchlorophenylhydrazone, both biosynthesis of CPS and its subsequent translocation across the cytoplasmic membrane were prevented (Kronke et al., 1990b). Energy inhibitors have several effects on assembly of LPS, and both a proton-motive force and ATP pools are required for expression of LPS on the cell surface (Bayer, 1979; Marino et a l . , 1985). At one level, translocation of nascent lipid A-core to the periplasmic face of the cytoplasmic membrane is energy-dependent in S. enterica serogroup B (McGrath and Osborn, 1991b). The 0-antigen-ligation reaction is not inhibited by the uncoupler of proton-motive force, dinitrophenol, but attachment of lipid-linked 0-polysaccharide cannot occur because of absence of the lipid A-core acceptor (McGrath and Osborn, 1991b). Dinitrophenol also inhibits synthesis of the 0-polysaccharide in S . enterica serogroup B, acting specifically on the galactosyltransferase (RfbP) reaction which initiates synthesis (see Fig. 3(a)); subsequent transferase and polymerase reactions are unaffected (Marino el al., 1991). The effect on RfbP galactosyltransferase activity is apparent in vivo, but not in vitro. This led to speculation that the energy requirement is associated with recycling of undecaprenol back to the cytoplasmic side of the cytoplasmic membrane.

182

C' Wl~l'l'lll-l I1 A N D M A V A I V A N 0

C. TRANSLOCATION FROM THE CYTOPLASMIC MEMBRANE TO THE CFLL SURFACE:

Once polysaccharides have been transported across the cytoplasmic membrane, subsequent reactions translocate the polymer across the periplasmic space and outer membrane. These represent the least well characterized steps in assembly of cell-surface polysaccharides. Nothing is known about the mechanism by which any bacterial polysaccharide traverses the outer membrane, although this remains one of the most intriguing questions in assembly. The width of the periplasm varies from 10.6 to 25.3 nm, depending on the bacterial species (Graham et al., 1991a,b) and, in most bacteria, the periplasm is filled with a gel consisting of peptidoglycan, enzymes and cellwall precursors (Hobot et al., 1984). Consequently, the rate of lateral diffusion of proteins (and presumably also polysaccharides) in the periplasm is remarkably low (Brass el a / . , 1986). This suggests that some kind of periplasmic translocation mechanism, perhaps involving binding proteins, may be required for export of polysaccharides. The periplasmic components could interact directly with components in the outer membrane to facilitate completion of translocation to the surface. Alternatively, the problem posed by the periplasmic space could be eliminated by transloc;iting polysaccharides at membrane-adhesion sites. These are regions where the cytoplasmic and outer membrane come into apposition and their existence is the subject of some controversy. As will be discussed later, these possibilities are not mutually exclusive.

I , Periplasmic Transport Systems Identification of a translocation apparatus for movement of some CPSs has come from studies on group-I1 polysaccharides in E. coli. Mutations in region 1 of the kps (CPS biosynthesis) clusters lead to accumulation of intracellular polysaccharide (Boulnois et al., 1987; Silver et al., 1987; I . Roberts et a l . , 1988) and, in E. coli K1 (Silver et al., 1987; Pelkonen. 1990) and K5 (Kronke et al., 1990a), the polysaccharide is located in the perip!asm. These results indicate that at least some of t h e products of region-] genes are involved in translocation across the periplasmic space. Conservation in region-1 genes (I. Roberts et a l . , 1986, 1988; Silver et a l . , 1987; M. Roberts et al., 1988; Boulnois and Roberts, 1990) and transcomplementation experiments with cloned region-1 genes (I. Roberts el a l . , 1986, 1988) indicate that the translocation processes are common for structurally diverse group-I1 CPSs. Five polypeptides are encoded by region-1 genes but, to date, only KpsD, a 60 kDa periplasmic protein, has

CEL1:SURFACE

POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

183

received detailed attention. Mutants of kpsD accumulate polymer in the periplasm (Silver et al., 1987) but the precise role of KpsD in translocation has not been resolved. In E. coli K5, periplasmic CPS showed a molecular weight comparable to wild-type CPS and was substituted with the phospholipid anchor (Kronke et a f . , 1990a). Components which may be involved in export of LPS have not been identified. Incomplete LPS with a KDO defect is poorly translocated, with a rate of about 20% of that in the wild type. The process is inefficient and 40% of the truncated LPS stays associated with the cytoplasmic membrane (Osborn et al., 1980). Since R-LPS with a complete core oligosaccharide is rapidly and efficiently translocated, it appears that the translocation machinery recognizes features of lipid A and inner core. This would provide a system which could operate independently of 0-polysaccharide structure (Osborn et af.,1980). The same could be achieved in group-I1 CPS of E. coli if the phospholipid moiety is the component recognized for translocation, but it is not known whether the presence of the phospholipid modification on the CPS is a prerequisite for translocation. Unlike CPS, LPS translocation is essential for viability in Gram-negative bacteria. Consequently, mutations identifying LPS-translocation components can only be isolated by searching among conditional lethal mutants. Presumably, the completed CPS and LPS must be released from the periplasmic face of the cytoplasmic membrane, prior to translocation. An analogous release step occurs for some exported proteins, where membrane association is not relieved until the leader peptide is cleaved (Koshland et al., 1982;Dalbey and Wickner, 1985). Periplasmicallylocated intermediates of some outer-membrane proteins have been reported (Stader and Silhavy, 1988) and Sen and Nikaido (1990) have presented arguments favouring a periplasmic export system for some porins. It is conceivable that some common mechanisms and/or components may be described.

2. Is There a Role for Outer-Membrane Proteins? There are no obvious features of LPS and CPS which could account for their precise location in the outer leaflet of the outer membrane, although this phenomenon is well studied in outer-membrane proteins (for reviews see Saier et al., 1989; Schatz and Beckwith, 1990). Coupling export of polysaccharides to an outer-membrane protein provides one attractive possible means of directing the polysaccharide specifically to the outer leaflet of the outer membrane. Porin proteins have been implicated in assembly of some CPSs. Porins are outer-membrane proteins that form water-filled channels across the outer membrane and facilitate transport of small hydrophilic solutes into the periplasm, where they can interact with

184

C . WHITF1EI.D AND M. A. V A I . V A N 0

transport systems. Protein K is a porin (Sutcliffe et al., 1983; Whitfield et al., 1983) found in different serotypes of encapsulated E. coli but rarely found in unencapsulated bacteria (Paakanen et al., 1979; Achtman et a l . , 1983; Van Alphen et al., 1983). Several observations suggested that protein K may play a role in surface expression of the CPS from E. coli K1, and perhaps in other E. coli group-I1 CPSs. Protein K is absent from outer membranes of E. coli K1 grown at 15°C a temperature which is nonpermissive for synthesis of group-I1 CPS in strains of E. coli (0rskov et al., 1984; Whitfield et a f . , 1985). Following the shift of E. coli K1 cells to the permissive temperature (37"C), appearance of protein K in the outer membrane is temporally correlated with appearance of K1 CPS on the cell surface (Whitfield et al., 1984b, 1985). A porin of some sort is essential for export of the latter polysaccharide since recipient strains devoid of known porins could not express the polysaccharide from E. coli K1 from cloned kps genes (Foulds and Aaronson, 1984). However, it is clear that K1 polymer is synthesized in E. coli K-12 strains carrying the kps cluster on a recombinant plasmid, despite the fact that these strains do not appear to contain protein K. This would tend to rule out any specific role for protein K as such. It is tempting to speculate that protein K may be the one of choice although other porins, which are structurally quite similar, may fulfil the role in recombinant strains. It would be interesting to assess the relative efficiency of synthesis of group-I1 polysaccharide in strains containing single, defined, porin types. Outer-membrane proteins have also been identified as components of the assembly system for EPS in other bacteria. It has been proposed that CtrA in N. rneningitidis is an outer-membrane protein, based on predictions of structure from DNA-sequence data and on immunological reaction of outer-membrane fractions with anti-CtrA antibodies (Frosch et al., 1992). The protein CtrA is highly conserved among N. meningitidis serotypes. Direct evidence linking CtrA with CPS synthesis is lacking, although location of ctrA within the region of the cps cluster associated with export (Frosch et al., 1991) argues for an involvement in export. The outermembrane protein AlgE is involved in assembly of alginate in Pseudomonas aeruginosa (May et al., 1991). It is exposed on the cell surface of many mucoid clinical strains, but is absent from non-mucoid strains (Grabert et al., 1990). The predicted amino acid-residue sequence of AlgE contains a cleavable leader sequence but does not possess recognizable transmembrane domains and bears little resemblance to porins (Chu et al., 1991). At present, t h e role played by outer-membrane proteins in EPS synthesis is not understood. It seems unlikely that porins provide a channel for export across the outer membrane, given the relatively small channel diameters facing a high molecular-weight helically wound polysaccharide chain.

C E l L S U R F A C E POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

185

Molecules of LPS are tightly associated with outer-membrane proteins, and this association may be responsible for the irreversibility of LPS assembly in the outer membrane (Osborn, 1979). However, it has been suggested that association of LPS with proteins is not essential for LPS translocation, since export of this molecule continues at normal rates in the absence of protein synthesis (Osborn, 1979).

3. Is There a Role for Zones of Adhesion? In the late 1960s Bayer first described structures known as zones of adhesion, or Bayer junctions, in electron micrographs of E. coli (Bayer, 1968). These structures reflected approximately 200-400 sites in each cell where the cytoplasmic and outer membranes came into close contact. Adhesion sites have now been implicated in export of CPS, LPS, phospholipids and some outer membrane proteins, in assembly of peptidoglycan and bacteriophages, and in bacteriophage binding and DNA injection. Interest in a possible role in membrane biogenesis for these membrane-adhesion sites began with studies by Miihlradt and his coworkers, who described sites of assembly of LPS on the cell surface in a conditional LPS mutant of S. enterica serovar typhimurium (Muhlradt et al., 1973). Lipopolysaccharide was seen to be inserted at approximately 220 discrete sites on the cell surface. Subsequently, Bayer used a phageconversion strategy to detect newly synthesized LPS molecules in S. enterica serovar anatum. Although only 20-40 export sites were identified in each cell during this work, the sites were located above membrane-adhesion sites (Bayer, 1979). Between 10 and 20 insertion sites were predicted by the studies of Kulpa and Leive (1976), who used membrane fractionation to follow newly synthesized S-LPS in a conditional LPS mutant derived from E. coli 0 1 11. Outer-membrane proteins are also inserted at discrete sites which may occupy up to 10% of the surface (deLeij et al., 1979) and may be located above adhesion sites (Smit and Nikaido, 1978). Insertion of CPS occurred at discrete sites located above membrane-adhesion sites in E. coli serotypes K29 (Bayer and Thurow, 1977), K1 (Whitfield et al., 1984b;Kronke etal., 1990a), K5 and K12 (Kronkeetal., 1990b), supporting the proposal that this is a common feature in biogenesis of the cell surface. A variety of bacteriophages have been seen to bind and inject DNA at sites above membrane-adhesion sites, suggesting that similar structural features may be required for both export and import of macromolecules (Bayer, 1979). Further evidence for a physiological role for adhesion sites has come from the observations that the numbers of sites can be influenced by the expression of bacteriophage MSZencoded lysis protein (Walderich et al., 1988; Walderich and Holtje, 1989), by

186

C. WHITFIE1.D AND M. A . VALVANO

incorporation of the gene-I product during assembly of bacteriophage f l (Lopez and Webster, 1985) and by modulation of levels of membranederived oligosaccharides (Holtje et al., 1988). A model for export of cell-surface components which relies on the involvement of membrane-adhesion sites is attractive because it conveniently addresses the problem of crossing the periplasm by simply eliminating it locally. With the two membranes in close juxtaposition, biosynthetic complexes in the cytoplasmic membrane could potentially interact directly with the outer membrane to facilitate co-ordination of synthesis with export. However, the organization and the existence of zones of adhesion is highly controversial. One of the major criticisms is that demonstration of these structures by electron microscopy has traditionally required chemical fixation, together with plasmolysis, to separate the two membranes. The existence of zones of adhesion has been challenged by Kellenberger (1990), based on results from electron-microscopy studies which utilize cryofixation to replace the chemical fixation procedures used by Bayer. In cryofixed samples, zones of adhesion were no longer evident. However, Bayer (1991) has more recently used a technical modification to demonstrate zones of adhesion in cryofixed samples. Furthermore, the number of adhesion sites could be increased in photo-cross-linked specimens, supporting the contention that the structures are fragile and may be disrupted during processing for microscopy. The debate concerning zones of adhesion will certainly continue as new electron-microscopy methods are developed. In electron micrographs, adhesion sites are seen as regions where the cytoplasmic and outer membranes come into apposition. Since the junctions remained after plasmolysis, it was initially suggested that the membranes were physically connected, perhaps by a fusion of the two membranes (Bayer, 1979). The suggestion of fusion has probably been the most contentious issue. Although fusion of eukaryotic membranes is well established, the interpretations for bacterial systems are based on electron microscopy data, and the precise molecular structure of the zones of adhesion has not been established. Indeed, most micrographs do not demonstrate unequivocally that the two membranes are fused (Bayer, 1991), but the limited space between the membranes at the adhesion sites (approximately 5 nm) would effectively exclude the presence of periplasm . Other studies have indicated that the peptidoglycan layer is continuous at the adhesion site (Leduc and Frehel, 1990), a situation that would seem to preclude localized membrane fusion involving anything but very thin membrane filaments. Despite continuing questions about the precise nature of adhesion sites and their possible significance, an increasing number of publications have described the isolation of membrane fractions which are thought to be

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

187

enriched in adhesion sites. When isolated membrane preparations from E. coli and S . enterica serovar typhimurium are subjected to sucrose densitygradient fractionation, the cytoplasmic and outer membranes are separated based on the significant difference in their buoyant densities. In addition, a membrane fraction with an intermediate buoyant density is consistently observed. The most detailed analysis of the intermediate membrane fractions was performed by Ishidate et al. (1986), who clearly demonstrated that the intermediate fraction contained markers of both cytoplasmic and outer membranes. Based on calculations using several reliable markers, intermediate (hybrid) fractions were shown to contain approximately 5067% outer membrane in a composite structure that could not be dismissed as a simple artefact resulting from the preparation procedures. The hybrids were shown to have the ability to synthesize authentic peptidoglycan from supplied precursors. While both the cytoplasmic membrane and hybrids could incorporate precursors into trichloroacetic acid-precipitable material, only the hybrid fraction was able to complete the process by forming SDSinsoluble peptidoglycan with appropriate cross-linking. Consistent with these results, two independent experimental approaches have localized penicillin-binding proteins at adhesion sites in E. coli (Barbas et al., 1986; Bayer et al., 1990); these proteins are required for peptidoglycan crosslinking. Ishidate and his coworkers also demonstrated that the hybrids played a role in translocation of nascent LPS from the cytoplasmic to the outer membrane (Ishidate el al., 1986). Significantly, this observation supported similar conclusions from Bayer’s laboratory, derived from a different experimental approach in studies involving S . enterica serovar anafum (Bayer et al., 1982). Membrane hybrids have been found to contain active polysialyltransferase complex, involved in biosynthesis of CPS in E. coli K1 (Vijay and Troy, 1975; Whitfield et al., 1984a). An additional lowdensity membrane fraction with sialyltransferase capability was also detected in E. coli K1 (Whitfield et al., 1984a), although its relationship to the activity found in hybrids remains obscure. Other functions ascribed to adhesion sites in isolated membrane hybrids include outer-membrane phospholipid translocation and metabolism (Bayer et al., 1982; Bayer and Bayer, 1985; Barbas et al., 1986), and translocation into the cell of DNA from bacteriophage T5 (Guihard et al., 1992). It is possible that some (or all) of the To1 proteins are required for the structural integrity of membrane-adhesion sites. The TolA protein is located in the periplasm and is anchored by its hydrophobic N-terminus in the cytoplasmic membrane (Levengood and Webster, 1989; Webster, 1991). The protein TolQ is located in membrane hybrids containing adhesion sites (Bourdineaud et al., 1989). The TolA and TolQ proteins are involved in export of group-A colicins and filamentous bacteriophages (Webster, 1991).

188

C. WIIITFIEI.1) A N D

M. A. V A I . V A N O

Cells with to1 mutations have pleiotropic defects, including leakage of periplasmic proteins and increased sensitivity to hydrophobic dyes and detergents (Webster, 1991). Recently, J . A. Thomas and M. A. Valvano (unpublished observation) have found that S-LPS is not expressed when the E. coli 0 7 biosynthesis genes are introduced into tolA and tolQ mutants of E. coli K-12, despite the fact that R-LPS is translocated to the outer membrane in these strains. While these studies all indicate that there are unique membrane fractions enriched in activities associated with cell-surface biogenesis, they do not offer further insight into the structure of the regions. In addition, the evidence that these active regions are related to structures observed by electron microscopy remains circumstantial. It is possible that zones of adhesion represent regions which are transiently stabilized by insertion of nascent cell-surface components, or by interaction of periplasmic translocation systems with both cytoplasmic and outer membranes. The junctions visible by electron microscopy would then take the form of slimy filaments, as described by Kellenberger (199O), possibly representing streaming of new material across the periplasm to the outer membrane. Resolution of the significance of membrane-adhesion sites will require electronmicroscopy images to be unequivocally corroborated by biochemical/ functional studies; at present the design of such experiments is not obvious. V. Genetics of Polysaccharide Biosynthesis A . HOUSEKEEPING AND POLYSACCHAKIDE-BIOSYNTHESIS GENES

Genes for biosynthesis of cell-surface polysaccharides are generally chromosomal and are usually arranged in clusters of one or more transcriptional units. There are some exceptions, as will be discussed later. The biosynthetic gene clusters contain information for assembly of the polysaccharide repeating units. Some clusters also contain genes coding for enzymes involved in polymerization of the repeating units, formation and attachment of the membrane anchor and export processes. Genes which regulate polysaccharide synthesis tend to be located at other loci. Enzymes responsible for synthesis of unique precursors are encoded within the biosynthetic gene clusters. However, where the precursors form part of metabolic pathways, housekeeping genes are generally used. Housekeeping genes are those whose products are involved in normal metabolism and are therefore not confined to polysaccharide synthesis. Biosynthesis of some polysaccharides involves housekeeping genes in early steps in production of a precursor, with polysaccharide-biosynthesis gene products completing

il

glucose

fructose 6-P

15

mannose 6-P I S

4

12

+ CDPglucose

+I13

glucose 6-P

glucose 1-P 12

M 1

dTDP-iF

CDP-4-ketod-deoxyglucose

1'.

dTDP-6-deoxy-Pxylo-4-hexulose

llo

dTDP-RHAMNOSE

CDP-4keto-3,6-dideoxyglucose

CDP-ABEQUOSE

CDP-PARATOSE

FIG. 7. Co-operation between the products of housekeeping and polysaccharide biosynthesis genes in formation of precursors of 0-polysaccharides in Salmonella enferica.The enzymes are as follows: 1, phosphoglucomutase (Pgm); 2, glucose-1-phosphate uridyltransferase (GalU); 3, UDP-galactose-4-epimerase (GalE); 4. glucosephosphate isomerase (Pgi); 5 , phosphomannose isomerase (Pmi); 6, phosphomannomutase (RfbK, CpsG); 7, mannose-1-phosphate guanidyltransferase (RfbM, CpsB); 8, glucose-1-phosphate thymidyltransferase (RfbA); 9, dTDP-glucose 4,Gdehydratase (RfbB); 10, dTDP-4-keto-~-rhamnose 3,Sepirnerase (RfbC); 11, dTDP-6-deoxy-~-lyxo-4-hexuloseCreductase (RfbD); 12, glucose-1-phosphate cytidyltransferase (RfbF); 13, CDP-glucose 4,Gdehydratase (RfbG); 14, RfbH; 15, abequose synthase (RfbJ); 16, paratose synthase (RfbS); 17, CDP-paratose 2-epimerase (RfbE). P indicates phosphate.

190

C' WIIITI.Il-I 1) A N D

M

A

VAI VAN0

the process. Interaction of specific and housekeeping gene products in formation of precursors for polysaccharides in Salmonella enterica serogroups A, B and D provides a good example (Fig. 7). Housekeeping genes are as a rule not linked to the polysaccharide-biosynthesis gene clusters and some of the reported, but uncharacterized, polysaccharide-biosynthesis genes may fall into this class once their precise functions are established. In some bacteria, location of a specific gene is not predictable, based on function. This is so with galE, the structural gene for the enzyme UDPgalactose 4-epimerase (GalE). This enzyme catalyses reversible conversion of UDP-glucose to UDP-galactose, providing the precursor for galactose residues which are common in bacterial cell-surface polysaccharides. It is also required for galactose metabolism through the Leloir pathway (Adhya, 1987) and is therefore considered to have a housekeeping function in enteric bacteria. The galE structural gene is found in the gal operon in both Escherichia coli and Salmonella typhimurium and is subject to complex regulation, involving two promoters regulated in an opposite fashion by CAMP,a gal repressor and translational coupling (Adhya, 1987). However, the location of galE differs in some bacteria. In Shigella dysenteriae serotype 1, the plasmid which carries the rfp gene required for 0-polysaccharide biosynthesis (see Section V.B.2) also carries a copy of galE. It is not clear whether this is the only galE gene, or a duplicate of a chromosomal function (Sturm et al., 1986b). In Haemophilus infiuenzae type b, galE is located at the lic3 locus, involved in biosynthesis of the LPS lipo-oligosaccharide (Maskell et al., 1991). In the plant pathogen Erwinia stewartii, galE is again separated from the remaining gal operon and is located adjacent to the cps (EPS biosynthesis) genes. As a result, galE is expressed constitutively in Er. stewartii (Dolph et al., 1988). A galE homologue, exoB, is found on the second megaplasmid of Rhizobium meliloti, with other exo (EPSI biosynthesis) genes. The gene exoB is also constitutively expressed (Buendia et al., 1991). It has been suggested that separation of galE is a common feature in plant-associated bacteria, and may reflect the requirement for the typically large proportions of galactose residues found in the EPS (Dolph et al., 1988). Interestingly, some Klebsiella spp. produce CPS and LPS 0-polysaccharides rich in galactose residues (Whitfield et al., 1991) using GalE activity determined by the gal operon (Clarke and Whitfield, 1992). €3. GENETIC DETERMINANTS FOR 0-POLYSACCHARIDE BIOSYNTHESIS

I . Chromosomal Genes f o r 0-Polysaccharide Biosynthesis The genetics of LPS biosynthesis has been studied extensively in S . enterica and E. coli (for a review, see Makela and Stocker, 1984) and much of the

CELI -SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

191

information for other bacteria is interpreted based on these prototypes. Early work by several groups showed that genes for biosynthesis of LPS are organized in several separate loci on the bacterial chromosome. LipidA biosynthesis is determined by the fpx genes (Raetz, 1990). Genes involved in synthesis of core oligosaccharide (rfa) are clustered together in the cysEpyrE region of the chromosome (Makela and Stocker, 1984). Mutations in any of the rfa genes lead to rough mutants with an incomplete LPS core, lacking the normal site for O-specific side-chain attachment. Many of the Rfa proteins are transferases, involved in sequential addition of KDO, heptoses and other carbohydrates in the core. Other functions modify the core with phosphoryl (RfaP) and ethanolamine groups and participate in the attachment of O-polysaccharide (RfaL) (Makela and Stocker, 1984; Rick, 1987). The rfa gene clusters of S. enterica serovar typhimurium and E. coli K-12 have been sequenced, and the organization of the gene clusters has recently been described (Schnaitman e t a f . ,1991). At least three operons are involved (Roncero and Casadaban, 1992). Some of the gene functions have been clearly established while others encode genes with undetermined roles in core assembly. The rfb genes determine biosynthesis of the 0polysaccharide repeating unit and are located near the his locus in enteric bacteria. Mutations in rfb block synthesis of O-polysaccharide and result in expression of R-LPS with a complete lipid A-core (Makela and Stocker, 1984). Escherichia cofi K-12 strains are defective in O-antigen biosynthesis but can support synthesis of a complete LPS molecule if functional rfb genes are provided (Jones et af., 1972; Schmidt, 1973). This ability has permitted molecular analysis of heterologous rfb genes. In the last few years, several laboratories have reported cloning and expression in E . cofi K-12 of O-polysaccharide-biosynthesis genes from a variety of bacteria (Table 7). In spite of the convenience of E. cofi K-12 as a recipient for cloning rfb genes, various problems have been experienced by a number of investigators (Valvano, 1992).These include instability and rearrangements in cloned DNA (Heuzenroeder et af., 1989; Kid0 et af., 1989), which can lead to altered structure in the polysaccharide product (Kido et af., 1989). Poor expression of the O-polysaccharide relative to that in the wild type (Valvano and Crosa, 1989; Haraguchi et af., 1989) may result from the structure of the lipid A-core acceptor, since these structures are not necessarily identical. This may alter the efficiency and/or specificities of ligation and translocation enzymes in E. cofi K-12 hybrids. A more serious problem results from rfb genes remaining on the chromosome of E. cofi K-12. These host functions may influence biosynthesis of O-polysaccharide directed by cloned DNA, giving modified structures with altered antigenic specificity in the recombinant product (Haraguchi et af., 1991). For

192 TABLE 7. Clusters of

C . WHITFIE1.D AND M. A VAI’VANO

rfh genes which have been cloned and expressed in Escherichia coli K-12

Organism Escherichia coli

Serotype

01 02 04

07 09 075 0101

Shigella dysenteriae

0111 01 B Cl c2 El Type 1

Shigella sonnei Shigella flexneri

Type 2a

Shigella hoydii

Type 3a Type 6 Type 12

Vibrio cholera Yersinia enferocolitica Yersinia pseudotuberculosis

01 0:3 IIA

Klebsiella pneumoniae Salmonella enterica

Reference Ding el al. (1991) Neal e f al. (1991) Haraguchi el al. (1989) Valvano and Crosa (1989) Kid0 ef a / . (1989) Batchelor el a / . (1991) Heuzenroeder el al. (1989) Bastin et al. (1991) Clarke and Whitfield (1992) Jiang et a / . (1991) Lee el a/. (1992) Brown e t a l . (1991, 1992) Wang et al. (1992) Sturm and Timmis (1986), Sturm et a / . (1986h) Yoshida el al. (1991) MacPherson et al. (1991), Yao ef a / . (1992) Yao el al. (1992) Cheah e f a / . (1991) M. Y. C. Handelsman and M. A . Valvano (unpublished observation) Manning ef al. (19%) Al-Hendy et al. (1991a) Kessler et al. (1991)

example, a gene in E. coli K-12 which maps within the rfb region encodes an 0-acetyltransferase activity, involved in the antigenic modification of 0antigens (Yao et al., 1992; Z . Yao and M. A. Valvano, unpublished observation). Strains of E. coli K-12 with a chromosomal deletion eliminating the rfb region have proved useful in the confirming that cloned genes for 0specific side-chains are capable of expressing the side-chain without participation of host rfb functions (Valvano and Crosa, 1989; Batchelor et al., 1991; Jiang et al., 1991; Macpherson et al., 1991; Clarke and Whitfield, 1992). Gene cloning has permitted detailed analysis of rfb-gene organization. The complexity of 0-polysaccharide biosynthesis is reflected in the number of rfb gene products. Different studies have identified six (E. coli 0101; Heuzenroeder el al., 1989), eight (E. coli 0 4 ; Haraguchi et al., 1989) and 16 (E. coli 0 7 ; Marolda et al., 1990; and S . enterica serogroup B; Jiang et al., 1991) Rfb polypeptides. Since some Rfb proteins may be poorly expressed, these results may underestimate the number of genes. The DNA

CELL SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

193

sequence and gene organization of the rfb regions of various serovars of S . enterica have been reported (Verma and Reeves, 1989; Jiang et al., 1991; Reeves, 1991; Brown et a l . , 1992; Wang et al., 1992; Fig. 8). In S . enterica serogroup B, 18 potential ORFs were identified in the rfb region. Of these, 16 are thought to be rfb genes (the status of ORFs 1.2 and 2.8 is unclear) and specific biosynthetic functions are assigned to 10 genes (Jiang et al., 1991; Fig. 8). Genes required for related functions, for example structural genes for sequential enzymes in precursor formation, are clustered together. All 16 ORFs in the rfb gene cluster in S. enterica group B are transcribed in the same direction, and potential promoters were found near the start of the rfa region indicating the possibility of a single operon (Jiang et a l . , 1991). An additional potential promoter exists within the rfb gene cluster in S . enterica serogroup El (Wang et al., 1992). In contrast, several lines of evidence, including transposon mutagenesis of cloned DNA, sitedirected mutagenesis in the wild-type strain, and complementation analyses, suggest that the rfb genes in E . coli 0 7 are organized as a complex gene cluster, rather than a simple operon. At least four transcriptional units are involved (Marolda et al., 1990; Valvano, 1992). The rfb clusters of Klebsiella pneumoniae 0 1 (B. R. Clarke, D . Bronner, C. Dodgson and C. Whitfield, unpublished results), Sh. dysenteriae type 1 (Sturm et al., 1986a) and Yersinia enterocolitica 0 3 (Al-Hendy et al., 1991a) also contain multiple transcriptional units. Confirmation of the number of transcriptional units requires analysis of mRNA transcripts; these data are not yet available for any rfa cluster. Although the majority of enzymes involved in biosynthesis of 0polysaccharide are encoded in the chromosomal rfb gene cluster in most enteric bacteria, some unlinked genes can be required. In S. enterica serovars A , B and D1, the unlinked rfc locus is required for polymerization of the 0-polysaccharide (Makela and Stocker, 1984). Homologous rfc sequences are present in serogroups A, B and D1, but are absent from serogroups C1, C2, C3, E2 and D2 and in E. coli K-12 (Collins and Hackett, 1991). These observations correlate with biosynthesis and expression data. Serogroup-B 0-antigen is polymerized in S. enterica serogroup D1 (Makela and Stocker 1984), but not in E. coli K-12 (Brown et a l . , 1991; Jiang et al., 1991). or in S . enterica serogroups C1, C2 (Naide et al., 1965; Makelii, 1966) and E (Nyman et al., 1979). The cloned rfb cluster of S . enterica serogroup B expressed in E. coli K-12 gives only SR-LPS (Brown et al., 1991; Jiang et al., 1991). In contrast, the cloned rfb clusters from E. coli, K . pneumoniae, S . enterica C1 and C2, and Shigella and Yersinia spp. (see Table 6) express polymerized 0-antigen in E. coli K-12. The possibility that E. coli K-12 contains an unlinked rfc function with no structural specificity is unlikely and the polymerase function is probably encoded by

serovar typhi (D) Man-Rha-Gal

serovar paratvphi (A)

I

Man-Rha-Gal

I

Tyv

Par

.

\

serovar typhimurium (B) Man-Rha-Gal

I

Abe

7.9

9.6

10.8

11.9

M

K

P

17.4

serovar rnuenchen (C2)

Abe

FIG. 8. Organization of rfb (0-polysaccharide biosynthesis) gene clusters in Salmonella entericu serovars. The serotype-B cluster is used as the prototype. The nomenclature for the known rfb genes (indicated by letters) is from Jiang et al. (1991) and Marumo et al. (1992). Uncharacterized open-reading frames are identified by map position. The serotype-A cluster contains an inactive rfbE* due to a frameshift mutation, and a triplicated region which fuses two open-reading frames, 15.4 and 12.8, to create a chimeric open-reading frame 15::12 (Verma et al., 1989). The physical maps are redrawn from Jiang et al. (1991), Reeves (1991), Verma et al. (1988), Verma and Reeves (1989), Wang et al. (1992) and Brown et al. (1992). N indicates non-rfb DNA; %I, CDP-3,6-dideoxyhexose synthesis; 0 , GDP-mannose synthesis; a, dTDP-rhamnose synthesis.

196

c

wwrr:ibi 11AND M A V A I VANO

the rfb gene cluster in these bacteria. Rfb proteins with similar hydrophobicity and secondary structure to the Rfc protein of S . enterica serogroup B have been predicted from analysis of the gene clusters identified in S. enterica serogroups C2 (Orf16.7~;Brown et af., 1992) and El (Orf17.4~;Wang et af., 1992). Furthermore, SR-LPS results from rfclike mutations in cloned rfb genes from E. cofi 0 7 (Marolda el af., 1990), E. cofi 0 7 5 (Batchelor et al., 1991) and in S. enterica serogroup C2 (Brown et af., 1992). The newly identified rof gene, which is involved in regulation of 0polysaccharide chain length in E. coli 075, is located near the his locus but is separated from rfb by gnd and an uncharacterized DNA region (Batchelor et af., 1992). A highly homologous protein (72%) is encoded by ruf in S . enterica serovar typhimurium (Batchelor et al., 1992). In comparison with the data available for animal pathogens, relatively little is known about the structure, composition and organization of LPS genes in plant-associated bacteria. However, involvement of cell-surface polysaccharides in plant-microbe interactions has stimulated recent activity in this area. Precise functions of most LPS genes in plant-associated bacteria are unclear, partly because of the complexity of LPS structures in these bacteria. Also, to date much of the research has focused on the role of LPS in plant-microbe interactions rather than LPS biosynthesis as such. Genes for LPS assembly have been identified in Bradyrhizobium japonicum (Staceyetaf., 1991), R . mefifuti(Cloveretaf., 1989; Williamsetal., 1990a,b; Brzoska and Signer, 1991) and R. feguminosarum (Brink et af., 1990; Calva et af., 1989). Available evidence suggests multiple LPS gene loci in these bacteria (Calva et af., 1989; Brink et af., 1990; Williams et al., 1990a). There is some overlap between the EPS and LPS biosynthetic functions in Pseudomunus sofanacearum (Kao and Sequeira, 1991); these genes most probably code for enzymes involved in the synthesis of precursors. Plantassociated bacteria provide many interesting problems in LPS biogenesis.

2. Pfasmid-Encoded Genes fur 0Pufysaccharide Biosynthesis Although genes involved in O-polysaccharide synthesis are usually chromosomal, plasmid-encoded functions are involved in some bacteria. For example, there have been several reports of plasmid-encoded LPS-associated functions in S . enterica. Kawahara et al. (1989) reported a plasmid-encoded rfc-like function in Safmunefladublin, which is involved in the polymerization of O-antigen. Expression of the 0 5 4 antigen in S . enterica involves functions carried on a 7.5 kb plasmid (Popoff and Le Minor, 1985). Recent research indicates that the 0 5 4 plasmid carries a complete functional rfb gene cluster, which is responsible for synthesis of

CEL1.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

197

an amino sugar-containing polysaccharide with a disaccharide repeating unit (C. Whitfield, M. Perry and C. Poppe, unpublished results). Further molecular analysis of this plasmid is in progress. The rfb gene cluster in E. coli 0 1 1 1 represents a curious situation because it has been reported to be on the chromosome in one strain (Bastin et al., 1991) and on a plasmid in another (Riley et al., 1987). The relationship between the two clusters has not been examined. In Shigella spp., plasmids are involved in O-polysaccharide biosynthesis to various extents. The form-1 O-antigen of Shigella sonnei is determined by genes encoded on a plasmid (Kopeck0 et al., 1980; Sansonetti et al., 1983; Yoshida et al., 1991). In contrast, the rfb genes in Shigella flexneri serotypes 2a (Macpherson et ul., 1991), 3a (Yao et al., 1992) and 6 (Cheah et al., 1991), and in Shigella hoydii type 12 (Marolda and Valvano, 1991) are chromosomal. Genes for O-antigen expression in Sh. dysenteriae serotype 1 are, interestingly, distributed between the chromosome and a plasmid (Hale et al., 1984). The plasmid-encoded rfp function from Sh. dysenteriae serotype 1 appears to add a galactose residue to the E. coli K12 LPS core (Sturm et al., 1986a). Galactose is the first sugar added during biosynthesis of O-polysaccharide, and the action of the Rfp galactosyltransferase is followed by sequential addition of the remaining sugar residues of the repeating unit (Sturm et al., 1986b). Subsequent steps in O-polysaccharide synthesis are mediated by products of the chromosomal his-linked rfb gene cluster (Hale et al., 1984; Sturm and Timmis, 1986; Sturm et al., 1986b). The possible advantage of having plasmid encoded O-antigen genes remains obscure and their significance in antigenic diversity is presently unclear. 3. Genes Involved in Modification of O-Polysaccharide Structure In enteric bacteria, most genes involved in modification of O-polysaccharide structure are not linked to the rfb gene cluster (for a review, see Makela and Stocker, 1984). The exception is O-acetyltransferase, which maps within rfb in E. coli K-12 (Yao et al., 1992, Z. Yao and M. A. Valvano, unpublished observation). The genes, required for glucosylation and 0acetylation, are generally encoded by lysogenic bacteriophages (Makela and Stocker, 1984). These modifications have a profound effect on polysaccharide immunochemistry . For example, serological specificity in 10 of 13 recognized O-serotypes of Sh. flexneri results from modification of the same tetrasaccharide repeating unit (Simmons and Romanowska, 1987). Bacteriophage Sf6 carries gene directing O-acetylation of 0-polysaccharide to form the group-3 antigen in Sh. flexneri (Gemski et al., 1975). The oac gene carried on the lysogenic bacteriophage Sf6 has been cloned

198

C‘ Wlll’l’FlEl.l> A N D

M A VAI VAN0

and sequenced (Clarke et al., 1991; Verma etal., 1991). The Oacsfhprotein has homology with NodX from R. legurninosarurn. The NodX protein is implicated in host-range specificity, possibly by catalysing the acetylation of a bacterial Nod factor (Fisher and Long, 1992). In S . enterica, bacteriophages and E~~ convert group serogroup El into E2 and E3, respectively (Jann and Jann, 1984; Makela and Stocker, 1984). Non-lysogenic cells of S. enterica serogroup E l synthesize 0acetylated repeating units linked by an a-linkage (Wright, 1971). Modification of the Rfc polymerase specificity by bacteriophage E” results instead in formation of a P-linkage (Robbins and Uchida, 1965; Robbins et al., 1965). Cells lysogenized with bacteriophages & I 5 and have a Plinkage and glucosyl substitution (Wright, 1971). Curing bacteriophage & I F reverses the linkage to a , indicating that bacteriophage functions exert a negative regulatory effect on host Rfc (Robbins et al., 1965). Bacteriophagemediated changes in 0-polysaccharide structure have also been reported in Pseudomonas aeruginosa (Kuzio and Kropinski, 1983). Presumably, examples will be described in other bacterial species, as the genetics of polysaccharide biosynthesis are described. C . GENETICS OF BIOSYNTHESIS OF EXTRACELLULAR POLYSACCHARIDE

Biosynthetic genes for EPS are arranged in clusters in most bacteria. This is true for both typical EPS and 0-glucans of Acetobacter xylinum (Wong et a l . , 1990; Ross et al., 1991), R . meliloti (Dylan et al., 1986) and Agrobacterium tumefaciens (Douglas et al., 1985). In only a few strains have EPS genes been expressed in heterologous hosts (e.g. E. coli K-12), to give an authentic polysaccharide product. Most of these are group-I1 like CPSs. In many other bacteria, the number and functions of genes involved in EPS synthesis are unknown. In these bacteria, the involvement of specific genes in EPS synthesis has been established by complementation of EPS mutations. Consequently, it is not clear whether the gene clusters reported in some bacteria contain the entire complement of essential biosynthetic genes. The alg genes involved in biosynthesis of alginate by P . aeruginosa provide a good example (for a review, see May et al., 1991). Of six complementation groups located at 34’, the precise functions of only algA (phosphomannose isomerase-GDP-mannose pyrophosphorylase) and algD (GDP-mannose dehydrogenase) are characterized. Functions of other biosynthetic genes identified in this region (and perhaps others yet to be identified) await clarification. An additional feature in EPS synthesis, not detected in synthesis of 0polysaccharides, is the occurrence of established regulatory genes. These regulatory genes tend not to be linked to clusters of biosynthetic gene. This

('I-I 1. SlJRFACF POI.YSA~'('liAH113ESI N (;RAM-NEGAI"VF BACTERIA

199

is true for group-I CPSs in E. i d ' (Gottesman and Stout, 1991), and EPSs in P. aeruginosu (May et al., 1991), Er. stewartii and several other plantassociated bacteria (Leigh and Coplin, 1992).

I . Group-ll Like Capsular-Polysaccharide Gene Clusters The CPS (K-antigen) genes of E. coli map to two different chromosomal locations, and the map positions provide an important criterion in t h e distinction between CPS groups I and 11 (Jann and Jann, 1990b). The kps locus, initially designated kpsA (0rskov and Nyman, 1974; 0rskov et a l . , 1976), is involved in synthesis of group-11 K-antigens. The kps locus maps near serA, and 64' on the E. coli linkage map, and the kps cluster from E. coli K1 provides the only example of a CPS gene cluster whose position on the chromosome is known with precision (Vimr, 1991). Molecular analyses of EPS genes began with the cloning of the 17 kb kps gene cluster from 1:. coli K1 (Silver et a l . , 1981). The organization of kps gene clusters is conserved in different serotypes, and this is consistent with early observations suggesting that the kps clusters are allelic (0rskov and Nyman, 1974). For detailed descriptions of the organization of kps clusters, the reader is referred to the review by Boulnois and Roberts (1990). In simple terms, the kps clusters are arranged in three functional regions (Silver et a l . , 1984; Boulnois ef a l . , 1987; Boulnois and Jann, 1989; Silver and Vimr, 1990; Vimr et a l . , 3989; Boulnois and Roberts, 1990) on a contiguous DNA segment (Fig. 9). The central region-2 genes encode functions related to the assembly and polymerization of the specific polymer. The size of region 2 reflects the complexity of the repeating units (Boulnois and Jann, 1989), suggesting the level of complexity in precursor synthesis and how many specific transferases are required. The K4 kps region 2 is the largest (14 kb) identified to date, and the polymeric product is comprised of a branched trisaccharide repeating unit with three different monomer residues (Drake et a / . , 1990; Fig. 9). In general, the phenotype of any mutation in a region-', gene is an inability to produce K-antigen. These functions include biosynthesis of CPS-specific precursors. For example, biosynthesis of the polymer in E . coli K1 requires a single precursor, namely CMP-sialic acid, while the genes neuA (Zapata et al., 1989), neuR (Vimr et al., 1989) and neuC (Zapata et al., 1992) are required for its formation. The genes neuA and neuC are part of the same transcriptional unit (Zapata et al., 1992). Other genes in region 2 of E. coli KI encode glycosyltransfer~ises which are involved in initiation and polymerization processes. Detailed information is only available for polysialyltransferases (NeuS) involved in biosynthesis of the polymers in strains K I (Weisgerber et al., 1991;Steenbergen et a l . , 1992) and K92 ( V i m

Surface expression Possible polymer modification Five genes

Precursor formation Polymerization

Export across the cytoplasmic membrane Two genes

al., 1992). l’he entire DNA sequence of kps region 2 in E. coli K5 has been determined and five region-2 genes are arranged in four transcriptional units (I. S. Roberts, personal communication). Region-2 genes are for the most part unique to a given serotype (Roberts ef al., 1988a). Exceptions can arise where CPS structures are very similar, as may be so in serotypes K l and K92 of E. coli. Both CPSs are homopolymers of sialic acid and differ only in linkage specificity. Consequently, precursor-forming enzymes and physical maps of the appropriate kps regions are conserved (Roberts ef al., 1986) but the polysialyltransferases, which define the linkage specificity, differ (Vimr et al., 1092). Regions I and 3 flank the central CPS-specific genes (Fig. 9) and encode

el

CELL-SURFACE POL.YSACCI1ARIDES IN CRAM-NEGATIVE HACTERIA

20 1

2 kb

[+ 8)-a-Neu5pAc-(2+8)-a-Neu5pAc-(2+]

[-b 8)-a-Neu5pAc-(2+9)u-NeuBpAc-(2+]

[+ ~ ) - ~ - D - M w I ~ N A c A - ( ~ + ~ ) - ~ - D - G ~ c @ ( ~ + ]

P

OAc

[ +3)-a-Rha-(1+2)-a-Rha-(l%)-p-KDO-(2+

1

y”

OAc

[ + 4)-p-D-GlcpA-(l-b 4)-a-D-GlcpNAc-(l+

]

[ + 4)-p-D-Gl~pA-(l+4)-p-D-GalpNAC-(l+

]

i

p-Fru

FIG. 9. Organization of kps (groupI1 capsular-polysaccharide biosynthesis) gene clusters in Escherichia coli. The nomenclature for genes in region 1 is based on the completed sequence from Escherichia coli K5 (I. S. Roberts, personal communication). The figure is modified from Boulnois and Roberts (1990).

functions which are conserved in biosynthesis of group-I1 K-antigens. Consequently, kps regions 1 and 3 from different serotypes are also conserved (I. Roberts et al., 1986, 1988; M. Roberts et al., 1988). Region 3 contains kpsM and kpsT, whose products are involved in transport across the cytoplasmic membrane. These genes have been characterized in serotypes K1 (Pavelka et al., 1991) and K5 (Smith et al., 1990). Region 1 encodes at least 5 proteins (Silver et al., 1984; Roberts et al., 1986; Boulnois and Jann, 1989) and region 1 from E. coli K5 has been sequenced (I. S. Roberts, personal communication). At least one region-1 protein (KpsD) is implicated in export across the periplasmic space to the outer membrane (Boulnoisetal., 1987; Silver etal., 1987). It is believed that region-1 proteins

202

c' WHlTFIEI.I) A N D M. A V A I . V A N 0

are also involved in postpolymerization modification reactions, such as addition of KDO to the reducing terminus of group-I1 CPS (Boulnois and Jann, 1989). The presence in region 1 of an additional functional copy of the structural gene for CMP-KDO synthetase (KpsU; I. S. Roberts, personal communication) probably gives rise to elevated levels of this enzyme characteristic of E. coli strains with group-I1 CPS (Finke et al., 1989, 1991). Gene clusters with similar functionally organized regions are involved in the assembly of the CPSs in H. injluenzae and Neisseria meningitidis, correlating with similarities among group 11-like CPS structures. Designations for the functional regions of cap in H. influenzae are reversed from those in kps in E. coli. Consequently, cap region 1 contains the four genes bexAbexD whose products are thought to be involved in transport of CPS across the cytoplasmic membrane (Kroll et al., 1988, 1990). Region 1 is conserved in the cap clusters of different H. influenzae serotypes (Kroll et al., 1989). A central variable region is thought to contain the serotype-specific information as in kps (Kroll et al., 1989). Flanking region 3 of cap is also conserved among different serotypes but a function analogous to region 1 in kps in E. coli has not been established. In N. meningitidis serogroup B, the cps gene cluster is divided into a central region (region A) containing genes required for precursor synthesis, repeating unit assembly and polymerization (Frosch et al., 1989). The polysialyltransferase structural gene is located in region A , and the predicted polypetide product shows significant homology with polysialyltransferases from E. coli K1 and K92 (Frosch et a l . , 1991; Weisgerber et al., 1991; Steenbergen et al., 1992; Vimr et al., 1992). Region A is flanked by region D, whose function is unclear, and region B which is necessary for transport across the cytoplasmic membrane. Region B contains ctrA, ctrB, ctrC and ctrD (homologues of bexD-bexA, respectively; see Table 6), whose products are required for export of CPS. To the other side of region A are regions C and E. Region C gene products play a role in translocation to the outer membrane while the functions of region E are not understood (Frosch et al., 1989). Homology was detected between some regions of the CPS cluster from N. meningitidis and DNA from other serotypes of N. meningitidis and N . gonorrhoeae, perhaps suggesting conservation in function (Frosch et al., 1989, 1991).

2. Group-I Like Capsular-Polysaccharide Clusters Genes for biosynthesis of group-I CPSs of E. coli, Erwinia spp. and K . pneumoniae share a common chromosomal location and are regulated in a similar fashion by products of the rcs (regulator of capsule synthesis)

CEI.1 -SURFACE POI.YSACCHARIDES IN GRAM-NEGATIVE BACTERIA

203

family of genes. The regulatory genes are not linked to those involved in assembly functions and are discussed in detail in Section V.B.l. It has been known for some time that the genes responsible for synthesis of group-I and group-I1 CPSs in strains of E. coli are not allelic (0rskov and 0rskov, 1962). Genes for synthesis of group-I CPSs are located near his and adjacent to rfb on the chromosome of E. coli (Schmidt et al., 1977; Laakso et al., 1988; Whitfield et al., 1989). An additional unlinked locus (near trp) was reported for CPS synthesis in E. coli K27 (Schmidt et al., 1977); the function of this locus was not examined biochemically, but strains lacking the locus produced less CPS. This, together with the location of the gene led to speculation that an unlinked polymerase (rfc-like) was required for synthesis of group4 CPSs in E. coli. A mutation with a similar phenotype (termed Ki for intermediate CPS production) has been reported in E. coli K30 (Whitfield et al., 1989). Synthesis of low molecular-weight KIaPSis not affected by the Ki mutation and it is not an rfc mutation in the classical sense (P. R. MacLachlan, W. J. Keenleyside, C. Dodgson and C. Whitfield, unpublished data). Genes for biosynthesis of CPS in E. coli K30, including that associated with the Ki phenotype, are all linked to his in E. coli K30; no requirement could be shown for a trp-linked locus (Laakso et al., 1988; Whitfield et al., 1989). The anomaly in the genetics of these two group4 CPS producers awaits clarification. At least some of the cps genes for colanic acid production in E. coli K-12 are also linked to his, and genetic analysis has revealed five complementation groups (cpsA-E) at this locus (Trisler and Gottesman, 1984; Gottesman and Stout, 1991). An additional unlinked locus (cpsF) is not essential for synthesis of colanic acid (Trisler and Gottesman, 1984). The only cps genes whose function has been determined are cpsB (mannosyl-l-phosphate guanidyltransferase) and cpsC (phosphomannomutase), both involved in formation of GDP-mannose and presumably part of the pathway leading to synthesis of GDP-fucose (Stevenson et al., 1991). Colanic acid is produced by many enteric bacteria. Recent studies have focused on distribution of colanic acid synthesis in E. coli strains which also produce serotype-specific K-antigen CPS (W. J. Keenleyside, D. Bronner, B. Jann, K. Jann and C. Whitfield, unpublished data). Strains of E. coli with group-I1 K-antigens are also capable of synthesizing colanic acid. In contrast, colanic acid is not produced by E. coli 09:K30, a prototype group I CPS producer (Keenleyside el al., 1992). This is consistent with preliminary results suggesting that cps genes in K30 strains and the cps genes for colanic acid are allelic (Keenleyside et al., 1992). Since little is known of the structure of these gene clusters at this time, hybridization

204

C . WHITFIEI.1) AND M. A. VALVANO

probes to resolve clearly this question have not been developed. When other group-I CPS-producing strains were examined, a fundamental split was observed among strains with group-I CPSs. (Jayaratne et al., 1993). All strains with group-I CPSs lacking amino sugars in their repeating unit (e.g. K30) were unable to synthesize colanic acid. However, strains with amino sugar-containing group-I K-antigens (e.g. K40; see Table 5 ) could simultaneously produce both colanic acid and the K-antigen. In the E. coli strains with group-I CPSs containing amino-sugar residues, the his region therefore contains a potentially large region of DNA devoted to polysaccharide synthesis, rfb genes for the O-polysaccharide and cps genes for both colanic acid and group-I CPS. The extent, if any, of interplay between these gene clusters remains to be established. The cps gene clusters from K . pneumoniae serotypes K20 (Laakso et al., 1988) and K2 (Arakawa et al., 1991) also map near his. Unlinked loci have not been detected and transfer of the his region of the chromosome in K. pneumoniae K20 to E. coli K-12 was sufficient for expression of the K20 CPS (Laakso et al., 1988). A plasmid containing a contiguous DNA fragment from the chromosome of K. pneumoniae K2 encoded all of the activities required for K2 CPS synthesis in E. coli K-12 hosts, providing the appropriate regulatory genes were present (Arakawa et al., 1991; Wacharotayankun et al., 1992). Although E. coli K-12 hosts may supply some precursors for biosynthesis, it is likely that the genes which are unique to CPS synthesis in K. pneumoniae are located in a single cluster. Some of the cps genes in Erwinia amylovora are also linked to his (McCammon and Coplin, 1982). At least five complementation groups have been detected, comprising 10 kb of contiguous DNA (Dolph et al., 1988; Coplin and Majerczak, 1990) and additional unlinked genes may also be involved (Leigh and Coplin, 1992). The ams (amylovoran synthesis) gene cluster in Er. amylovora contains five complementation groups within a 6 kb region, and there is some cross-complementation with the genes in Er. stewartii (Leigh and Coplin, 1992). Only the cps-associated galE gene has been characterized.

3. Genes for Biosynthesis of Extracellular Polysaccharide in Other Plant-Associated Bacteria Rhizobium meliloti is a symbiont associated with nitrogen-fixing nodules in roots of leguminous plants. Cell-surface polysaccharides play a role in nodulation process, and genes involved in their biosynthesis are being intensively studied (for a review, see Leigh and Coplin, 1992). Two structurally discrete EPSs are produced. The major polymer is

CE1I:SURI;ACE

POI YSACCHARIDES IN GRAM-NFGATIVE RALTERlA

205

succinoglycan (EPSI; see Table 1). The second polymer, EPSII (or EPSb), is structurally different (Levery et af., 1991) and is usually only produced when mutations eliminate synthesis of EPSI. The exo genes are responsible for synthesis of EPSI (Long et af., 1988; Reuber et af., 1991). Most of the exo genes are located within a 22 kb region and comprise multiple complementation groups with distinct transcriptional units; the entire 22 kb region has been sequenced (Arnold et af., 1990). The regulatory genes exoR and exoS (Doherty el af., 1988) are unlinked to the genes encoding enzymes involved in biosynthesis. The majority of the genes are located on the second endogenous megaplasmid pRmeSU47B, although chromosomal loci including exoC (Finan et af., 1986) and exoD (Reed and Walker, 1991) are also required. In R. feguminosarum, the genetic determinant for 0-acetylation of EPS is also chromosomal (Canter-Cremers et a f . , 1991). The ex0 genes are also responsible for succinoglycan synthesis in Agrobacterium radiobacter NCIB 11883. The exo genes in this bacterium may be distributed among several loci, since mutations are complemented by genes present on five non-overlapping cosmids (Aird el a f . , 1991). One exo mutation in Ag. radiobacter is complemented by the exoB ( g a f E )gene from R. mefifoti.The exp (or muc) genes encode specific functions in EPSII biosynthesis in R. mefifoti (Glazebrook and Walker, 1989; Zhan et af., 1989). As with the ex0 genes, exp is located on the second megaplasmid and consists of multiple complementation units (Glazebrook and Walker, 1989). In P. sofanacearum, the genes for biosynthesis of EPS are located in three regions. Some of them are probably involved in precursor synthesis because mutations simultaneously affect synthesis of both EPS and LPS. For example, the ops genes are involved in synthesis of both EPS and LPS and are arranged in seven complementation groups covering 6.5 kb (Kao and Sequeira. 1991). Two additional adjacent chromosomal loci are required for EPS synthesis. Region I (9 kb) is located 7 kb upstream from region 11, and these gene clusters are apparently regulated independently (Denny et a f . , 1988; Denny and Baek, 1991). The gum (also known as xgs or xan) genes required for the biosynthesis of xanthan gum in Xanrhomonas campestris have been cloned on a contiguous DNA fragment (Harding, et af., 1987; Thorne et af., 1987; Vanderslice et af., 1989). The entire gum cluster has been sequenced and analysed at the genetic and biochemical level (Vanderslice et a f . , 1989; see Section 1II.C). However, the DNA sequence data have not been published and is not available in current database releases, so comparison with other EPS genes is not possible at this time. Subcloning experiments performed by Harding et af. (1987) defined a minimum of five gum complementation groups. A subsequent study identified as many as 12 complementation

206

C WtIITFIbI I> AN11 M

A

VAI V A N 0

groups in a 35 kb region, and the directions of transcription were investigated using Tn5-lac mutagenesis (Hotte et a!. , 1990). The presence of transcriptional units arranged in opposing directions strongly argues for more than one promoter. The xanA and xanB genes forming complementation groups A and B have been cloned, sequenced and the activities o f the gene products characterized (Koplin et al., 1992). The proteins XanA and XanB are involved in precursor synthesis. The first of these proteins is required for formation of glucose 1-phosphate and mannose 1-phosphate, while XanB appears to be a bifunctional phosphomannose isomerase-GDPmannose pyrophosphorylase, with an activity similar to bifunctional AlgA in P. aeruginosa (Shinabarger et al., 1991). Curiously, reactions catalyzed by XanB and AlgA are not sequential steps in formation of GDP-mannose; participation o f phosphomannomutase is also required. In members of the Enterobacteriaceae, each of these reactions is mediated by different gene products. I). RELATIONSHIPS BETWEEN MlJLTIPLt..

t'OLYSACCHARI1)E-BIOSYNTHESIS GENE CLUSTERS

Many bacteria are capable of synthesizing several cell-surface plysaccharides and, as a result, have several biosynthetic gene clusters. For example, E . cofi strains producing group-I1 K-antigens potentially have loci involved in biosynthesis of ECA (rfe-rff, 85' on the chromosomal linkage map; Meier and Mayer, 1985), 0-polysaccharide (rfh, 42'; Makela and Stocker, 1984), group-I1 K-antigen ( k p s , 64'; Vimr, 1991) and colanic acid ( c p s , 42'; Trisler and Gottesman, 1984). This can lead to duplication of some activities. In E. cofi K7, UDP-N-acetylmannosaminuronic acid is a precursor for both ECA and group-I1 K7 CPS. This precursor is synthesized by products of the rffE and rffD genes, located in the ECA biosynthetic cluster. Since the cloned K7 kps biosynthetic gene cluster can complement rffD and rffE mutations, these activities may be duplicated in kps (Meier-Dieter et al., 1990). Salmonella enterica serovar typhiniurium has two distinct versions of phosphomannomutase (RfbK and CpsG) and mannose-1-phosphate guanidyltransferase (RfbM and CpsB) (Stevenson et a f . , 1991). These enzymes catalyze the sequential reactions: mannose 6-phosphatemannose 1-phosphate-+GDP-mannose, respectively (see Fig. 7). The proteins RfbK and RfbM are products of rfh genes, and synthesize GDPmannose for 0-antigen synthesis (Jiang et a l . , 1991). The proteins CpsG and CpsB are products of cps genes and are part of the pathway for GDP-fucose synthesis (via GDP-mannose) in biosynt hesis of colanic acid (Stevenson et al., 1991). In some bacteria there is clear co-operation between biosynthetic gene

('ElJ,-SLlRFACF POI.YSAC('HAHII>ES

IN GRAM-NEGATIVE H A C T F R I A

207

clusters. Participation of rfe and rfb functions in the rfe-dependent polymerization of 0-polysaccharides has already been described (see Section III.B.2). Another example is found in biosynthesis of GDP-mannose, a precursor for synthesis of both 0 9 polysaccharide and group-I K30 CPS in E. coli 09:K30. In contrast to S. enterica serovar typhimurium, GDPmannose is formed only by activites of RfbK and RfbM in E. coli 09:K30. In this organism, there is no duplication of GDP-mannose-synthesis functions in the cpsK3()gene cluster, so that an rfbM mutation eliminates synthesis of both 0 9 and K30 polymers (P. Jayaratne, P. R. MacLachlan, C. Dodgson ~ (K30 s and C. Whitfield, unpublished observation). The rfb,, and c CPS biosynthesis) clusters are located adjacent to one another on the chromosome (Laakso et af., 1988; Whitfield et ul., 1989). An interesting situation occurs in synthesis of TDP-N-acetylfucosamine, a precursor of ECA in strains of S. enterica. The rfbA and r P B genes in S. enterica serogroup B strains encode enzymes responsible for formation of TDP-4keto-6-deoxy-~-glucose(see Fig. 7), an intermediate on the pathway leading to TDP-rhamnose and TDP-N-acetylfucosamine. Rhamnose residues are present in the 0-polysaccharide while residues of N-acetylfucosamine are found in ECA. In S. enterica serogroup B, the ECA biosynthetic cluster does not carry analogues of rfbA and rfbB, and ECA biosynthesis is dependent on rfb function. Duplication of activities is therefore avoided. However, group C l S. enterica strains lack rhamnose residues in the 0polysaccharide and f54rfbA and rfbB are absent. Instead, N-acetylfucosamine precursor is synthesized by the products of additional genes in the rfe-rff locus (Lew et uf., 1986).

E. MOLECULARBASIS

FOR ANTIGENIC VARIATION IN CELL-SURFACE P0LYSACCHARII)ES

Selective pressures presumably led to the tremendous diversification in the structures of cell-surface polysaccharides. Application of molecular-biology techniques has made possible investigation of the molecular basis of this variation. The similarities in the group-I1 CPS gene clusters of E. coli suggest that some regions may have evolved from a common progenitor (Boulnois and Jann, 1989). Since kps clusters are found only in strains of E. coli expressing group-I1 CPSs, it is speculated that the kps cluster evolved independently from the DNA region in which it is located, and was added to the genome in E. coli by a transposition event (Vimr, 1991). A similar mechanism has been proposed for gene clusters for CPS synthesis in H. influenzae. The proposal is supported by identification of IS1016 sequences flanking the CPS (cap) genes in some strains of H. influenzae (Kroll et al., 1991; Kroll, 1992). There are two phylogenetic divisions (I and 11) of H. influenzae, but

~

~

208

C WIIITFIFI I) A N D M A V A I V A N 0

IS1016 sequences only flank division-I cap clusters. Transposition mediated by IS1016 may be responsible for the different chromosomal location of cap in division I-and division-I1 serotype-b strains (Kroll, 1992). Many of the genes in group-I1 CPS clusters are conserved (serotypeindependent). Recombination of genes within region 2 would therefore be sufficient to generate a new serotype (Kroll and Moxon, 1990). Haemophilus influenzae and N . meningitidis are naturally competent (Smith el al., 1981b). Both organisms can be transformed with heterologous DNA in vitro (Zwahlen et al., 1989; Frosch et al., 1991), suggesting the possibility that diversity in region 2 of these CPS gene clusters, leading to different serotypes, may have resulted from localized genetic exchange. Consensus DNA-uptake sequences have been identified at sites flanking region 2, providing a potential mechanism for this type of exchange in H . influenzae (Kroll, 1992). Recombination has also been proposed to explain similarities in polysialyltransferases from E. coli serotypes Kl and K92 (Vimr et al., 1992) and N . rneningitidis (Frosch el al., 1991; Weisgerber et al., 1991). Imperfect palindromic sequences in DNA at the junction of regions I and 2 in E. coli K1 kps may reflect these past recombination events (Steenbergen et al., 1992). The rfb gene clusters of S . enterica have been studied in Reeves’s laboratory as a model for molecular evolution. The proposal for a common progenitor has been made for some r - gene clusters in S. enterica and is supported by the observation of near-identity in DNA flanking rfh in S . enterica. The observation of regions of limited similarity suggests that part of the rfb cluster in S . enterica diverged into different forms over a long period of time. Three to five segments of DNA in the rfb region of S . enterica serovar typhimurium have abnormally low G+C content relative to the values for total DNA in this species, and atypical codon-usage indices (Jiang et al., 1991; Liu et a l . , 1991; Reeves, 1991). This has been interpreted as an indication that the rfb cluster in S. enterica was captured from another (low G + C ) species (Reeves, 1991). Significantly, several other genes involved in polysaccharide synthesis also demonstrate atypically low G+C content for the respective organism. Examples include rfc (Collins and Hackett, 1991) and some rfa genes (MacLachlan et al., 1991) in S . enterica serovar typhimurium, some rfa genes in E. coli K-12 (Klena et al., 1992), regions of the rfb cluster in E. coli 0 7 (Marolda and Valvano, 1993), kps region 2 in E. coli K1 (Steenbergen et al., 1992) and K5 (I. S. Roberts, personal communication), and rfbT of Vibrio cholerae (Stroeher et al., 1992). These results may indicate that all of these genes originated outside their current host. Alternatively, it is conceivable that the low G+C content may arise from a requirement for specific amino acids in the gene products. The use of atypical codons may reflect regulation by ensuring low levels of translation.

CE1.L-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

209

Remarkable similarity is seen in the rfb region of S. enterica strains with related 0-polysaccharides (Reeves, 1991). This is summarized in Fig. 8. Restriction-endonuclease mapping, hybridization experiments and DNAsequencing experiments show extensive homology in the rfb regions of S . enterica serovars typhimurium (serogroup B), typhi (serogroup D) and paratyphi (serogroup A) (Verma et al., 1988; Reeves, 1991). Structures of 0-polysaccharides from serogroups A, B and D differ in the nature of the 3,6-dideoxyhexose (see Table 3). Variation exists in the region of rfb responsible for biosynthesis of different 3,6-dideoxyhexoses, as expected since the final steps on the pathways to 3,6-dideoxyhexoses differ (Fig. 7). CDP-4-keto-3,6-dideoxyhexose is converted to CDP-abequose by RfbJ. The enzyme RfbS converts CDP-abequose to CDP-paratose while CDPtyvelose results from epimerization of CDP-paratose by RfbE. Serogroup B contains rfbJ but lacks rfbS and rfbE; its side-chain is therefore abequose. Serogroups D and A contain paratose and tyvelose, respectively, and serogroup D1 strains can be converted to serogroup A by a mutation in rfbE (Sasaki and Uchida, 1974). DNA sequence data shows that groups A and D both contain rfbE, but a frameshift mutation renders the gene non-functional in serogroup D (Verma and Reeves, 1989). If the remaining rfb DNA is identical, the implication is that the same transferase can add any available 3,6-dideoxyhexose and that 0-serotype specificity is mediated only by precursor formation. This is confirmed by the conversion of serogroup A and D strains to serogroup B following transformation with plasmids carrying cloned rfbJ (Wyk and Reeves, 1989). Mutation in rfbE and the distinction between S. enterica serogroups A and D provides perhaps the most simple cause of antigenic diversity in 0-polysaccharides. A similar mechanism has recently been reported in antigenic variation in the LPS of V. cholerae 0 1 . Conversion from serotypes Ogawa to Inaba occurs in vitro (Sack and Miller, 1969; Sakasaki and Tamura, 1971; Redmond et al., 1973) and in vivo (Sheehy et al., 1966; Gangarosa et al., 1967), and strains with the Inaba serotype may be selected by the immune response. The subtle immunochemical changes which result in 0polysaccharides in different serotypes have not been resolved. The rfb clusters of representatives of serotypes Inaba and Ogawa have been cloned and their DNA sequences have been determined (Stroeher et al., 1992). One gene, termed rfbT, but not related to the rfbT mutations giving rise to R-LPS in enteric bacteria, is involved in the conversion phenomenon. Both serotypes have rfbT sequences, but in Inaba strains rfbT is truncated by frameshift mutations. Residues of galactose, mannose and rhamnose are found in O-polysaccharides in S. enterica serogroups B and C2, but the arrangement of the repeating unit differs in these polysaccharides (see Table 3). Some

210

C'

W l l l l l ~ l h I) l AND M

A VALVAN0

homology between the rfb gene clusters in serogroup B and C2 might be expected due to the requirement for common precursors and, indeed, blocks of homologous genes were found to be arranged in the same order (see Fig. 8). A central region of low homology contains seven ORFs, presumably representing genes which encode transferases of different specificity (Brown et al., 1991, 1992). Genes for conserved functions which flank the variable region show some divergence in serogroups B and C2 (Brown et a l . , 1992). Homology is also evident among rfb clusters from O-serotypes of E. coli with structurally similar O-polysaccharides (Beger et al., 1989; Haraguchi et al., 1989). Restriction-site polymorphisms have been observed in the rfb regions of E. coli 0 2 and 018 isolates (Beger et al., 1989). This is consistent with the observation that a given O-serotype can be found in E. coli strains belonging to different clonal groups, based on multilocus enzyme electrophoresis (Caugant et al., 1985). The homologous rfb clusters from isolates of K . pneurnoniae 0 1 also show some restriction-site polymorphisms, indicating more than one clonal group (Clarke and Whitfield, 1992). Homology is not confined within a species. For example, O-polysaccharides in Yersinia pseudotuberculosis H A and S . enterica serogroup B both contain abequose residues and their rfb clusters show homology in rfbF and rfbC (Kessler et al., 1991). In contrast, some bacteria synthesize identical or very similar 0polysaccharides without any significant homology in most rfb genes. One example is found in rfb genes from Shigella boydii type 12, a strain which produces an O-polysaccharide very similar to that of E. coli 0 7 (L'Vov et al., 1984; see Table 3). When cloned rfb regions from E. coli 0 7 and Sh. boydii type-12 strains were compared, the only significant homology was confined to the rfbK homologues (see below) (M. Y.C. Handelsman and M. A. Valvano, unpublished observation; Valvano and Marolda, 1991). Furthermore, significant variation in rfb can also occur within a given species, in strains producing very similar O-polysaccharides. When K . pneurnoniae 0 1 rfb probes were used to examine the rfb clusters of other strains of K . pneurnoniae, synthesizing related O-polysaccharides (see Table 2), the interrelationships were complex (R. F. Kelly, B. R. Clarke, J . M. X. Tomas and C. Whitfield, unpublished results). Klebsiella pneumoniae 0 1 rfb functions direct the synthesis of the D-galactan I O-polysaccharide (Clarke and Whitfield, 1992). Significant homology was found in rfb from K. pneurnoniae strains which make 0 1 or 0 2 a antigen; 0 2 a is identical to 11-galactan I. However, 0 1 and 0 8 O-polysaccharides have identical carbohydrate structures, but surprisingly little homology was detected between the 01 and 0 8 rfb clusters. Klebsiella pneurnoniae serotypes formerly described as 0 2 a , 2e, 2h and 0 2 a , 2f, 2g produce identical polysaccharides (see Table 2), namely D-galactan I with an a-D-galactosyl

CtLL-SURFACE POLYSACCHARIDES IN GRAM-NFGATIVF BACTERIA

21 1

side-chain. Variation in the linkage and frequency of the a-galactosyl side-chain creates serotype 0 9 . Despite the common D-galactan I polysaccharide backbone in these polysaccharides, the rfb clusters in these strains share little homology with 0 1 rfb. These results clearly demonstrate that a similar chemical composition and structure of an 0-polysaccharide does not necessarily correlate with similarities at the genetic level. The differences between the levels of rfb conservation in K . pneumoniae and S . enterica may reflect the degree of genetic heterogeneity in these species. Comparative sequence date are now available for enzymes involved in GDP-mannose synthesis (RfbK, CpsG, RfbM, CpsB; see Fig. 7) from several strains of S. enterica and E. coli,and some interesting relationships were found. Enzymes of S . enterica serogroup B can be taken as the prototypes (Jiang et al., 1991). Two families of phosphomannomutase were detected. One family contains RfbK from serogroups B and El (Wang et al., 1992). The second family includes CpsG from S. enterica serogroup B, and RfbK proteins from S . enterica serogroup C1 (Lee et al., 1992), E. coli 0 7 (Marolda and Valvano, 1993), two strains of E. coli 0 9 (T. Sugiyama, N. Kido, T. Komatsu, M. Ohta, K. Jann, B. Jann and N. Kato, unpublished observation; P. Jayaratne, P. R. MacLachlan, C. Dodgson and C. Whitfield, unpublished observation) and Shigella boydii type 12 (M. Y.C. Handelsman and M. A. Valvano, unpublished data). The XanB enzyme from X . campestris (Koplin et al., 1992) is also related to this family, and all members contain a conserved catalase motif and a region with high homology to the active site of rabbit-muscle phosphoglucomutase (Koplin et al., 1992; Marolda and Valvano, 1993). These domains are absent from the RfbK protein of S . enterica serogroup B. Much less similarity is seen among RfbM proteins from different species and serotypes, and no consistent pattern has emerged. It is possible that relationships in RfbK reflect recombination between rfa and cps genes. It is notable that the rfbM-rfbK homologues in S. enterica strains and E. coli 0 7 are located near the gnd proximal end of the rfb cluster (see Fig. 8), with the same direction of transcription. In E. coli 0 9 , the direction of transcription of rfbK-rfaM is maintained, despite the fact that the two genes appear to be flipped as a block relative to gnd (P. Jayaratne, P. R. MacLachlan, C. Dodgson and C. Whitfield, unpublished observation). There is also duplication of the rfbM-rfbK region in E. coli 09:K30, with approximately 6 kb of DNA separating the two copies. The 3' ends of the copies of rfaK show sequence changes. Duplications of DNA also occur near rfbE in some strains of S . enrerica serovar paratyphi, where it gives rise to a triplicated region (Liu et al., 1991; see Fig. 8). These duplications may reflect pat recombination events involved in antigenic variation.

212

C. WII1TLlEI.D AND

M. A. V A I . V A N 0

VI. Regulation of Cell-Surface Polysaccharide Synthesis Synthesis of LPS may be subject to complex (fine tuning) regulation, but on-off switching is not possible due to the essential structural requirement for the lipid A-core LPS molecule. It is conceivable that regulatory systems which influence LPS synthesis are also essential in the cell, making their dissection and characterization difficult. Consequently, information regarding regulation of LPS synthesis is limited. Since EPSs may be required for bacterial survival only under certain circumstances, mechanisms which control the amounts of EPS synthesized occur frequently. Most bacteria use EPSs for protection, and many regulatory strategies are directed to modulating EPS synthesis in response to appropriate environmental cues. As with other cellular processes in bacteria from diverse habitats, there are remarkable similarities in strategies used by organisms to regulate EPS synthesis. The following discussion attempts to cover some of these common mechanisms and to indicate others which currently appear to be unique to specific bacteria. Regulatory mechanisms for EPS synthesis are broken down by strategy, rather than by organism, and some examples are summarized in Table 8. A . K E G U L A T I O N OF LIPOPOLYSACCHAKIDE SYNTHESIS

The spectrum of LPS molecules revealed by SDS-PAGE is characteristic for a given strain but can be affected by growth conditions (Al-Hendy et al., 1991b; Berry and Kropinski, 1986;Day and Marceau-Day, 1982; Dodds et al., 1987; McConnell and Wright, 1979; Weiss et al., 1986; Poole and Braun, 1988; McGroarty and Rivera, 1990; Nelson et al., 1991). Some of the mechanisms regulating LPS have multiple roles in cellular physiology. For example, in Escherichia coli and Salmonella enterica serovar typhimurium, LPS biosynthesis is influenced by amino-acid deprivation, and the control process involves relA and the stringent response (Ishiguro et al., 1986). The rfaH (sfrB) gene product also regulates LPS synthesis as well as several other functions in enteric bacteria. Mutants of S . enterica serovar typhimurium with an rfaH defect produce R-LPS (Wilkinson and Stocker, 1968) with heterogeneous lipid A-core molecules (Lindberg and Hellerqvist, 1980). The RfaH protein is involved in transcription of the Ffactor (tra) operon (Beutin and Achtman, 1979; Beutin et al., 1981) and regulates synthesis of core LPS in E. coli (Creeger et al., 1979; Sanderson and Stocker, 1981). The protein RfaH is analogous to HlyT, a positive regulator of the operon encoding haemolysin synthesis (hly) in E . coli (Bailey et al., 1992), indicating that RfaH/RfaH may interact with promoter-operator regions of diverse gene clusters. The Rfah protein was

CFI 1. SURFACE POI.YSAC<’f IARIDES IN GRAM-NEGATIVF BACTERIA

213

originally proposed to act as a transcription antiterminator (Beutin et al., 1981; Gaffney et al., 1983; Farewell et al., 1991). However, recent studies have utilized gene fusions to rfaCBIJ of S . enterica serovar typhirnuriurn (Brazas et al., 1991) and rfaQCPBI of E. coli (Pradel and Schnaitman, 1991) to show that RfaH most probably acts as a positive regulator for synthesis of core LPS. Prolonged passage of S . enterica serovar anaturn in the laboratory, in the absence of environmental selective pressures results in shorter average 0-polysaccharide chain lengths (McConnell and Schoelz, 1983). In contrast, selection for serum-resistance in E. coli 0111 results in strains which cap more R-LPS with 0-polysaccharide and show an increase in 0polysaccharide chain length (Goldman et al., 1984). The molecular processes underlying these phenomena are uncharacterized, although they have a particularly important bearing on the functional properties of LPS. One hypothesis for 0-polysaccharide chain-length control invokes competition between the reaction which polymerizes the 0-antigen and that which transfers (ligates) the 0-antigen to lipid A-core. The observation that SR-LPS (in rfc strains of S . enterica) is rapidly and efficiently ligated to lipid A-core indicates that the ligase function itself does not determine chain-length distribution (McGrath and Osborn, 1991a). Goldman and Hunt (1990) developed a mathematical approach to argue that heterogeneity in chain length results from an ability of the 0-antigen polymerase and/or ligase to generate specific 0-polysaccharide sizes for assembly in the outer membrane. However, this model is difficult to reconcile with data from mutants which causes a shift in average 0-polysaccharide chain length, such as those identified in S . enterica serovar anaturn (McConnell and Schoeltz, 1983; McConnell et al., 1986). These mutations could be explained by the recent discovery of a gene in E. coli 075, which appears to regulate distribution of 0-polysaccharide chain length in LPS. The gene is designated rol (regulator of 0 length) (Batchelor et al., 1991) and rol mutants express S-LPS with abnormal and unregulated 0polysaccharides. In E. coli 075, Rol is distinct from the polymerase although it is considered possible that Rol interacts with the polymerase to modulate its activity/specificity (Batchelor et al., 1991). The gene rol is present in S. enterica serovar typhirnuriurn (group B) (Batchelor et a l . , 1992) and mutants with phenotypes similar to rol defects have also been described in E. coli 0111 (Bastin et al., 1991) and S . enterica serogroup C2 (Brown et al., 1991). The Rol function may therefore be conserved in enteric bacteria. A different mechanism regulates the chain length of the 0:3 polysaccharide in Yersinia enterocolitica in response to growth temperature (AlHendy ef al., 1991b). In this bacterium, modulation of 0-polysaccharide synthesis involves decreased transcription of rfb genes, possibly due to

TABLE 8. Level of regulation Transcnption 1 Two-component regulatorq systems

2 . Histone-like proteins 3. Cap analogues 4. Other transcription effectors

DNA rearrangements

Examples of regulatory fac:ors affecting synthesis of bacterial cell-surface plysaccharides Organism

Pseudornonas aerugrnosa E s c h e n c h colr

Polymer

Components and commsnts

Alginate extracellular polysaccharide Colanic acid and group1 capsular polysaccharide Vi antigen

AlgRI-AlgR2. AlgB RcsB-RcsC, RnA-Lon. RcsF

Capsular polysaccharide Extracellular polysaccharide

RscA (possibly R n E R c s C ) R~~B-RCSC. RcsA

Extracellular plysaccharide Alginate extracellular polysaccharide Alginate extracellular polysaccharide Extracellular polysaccharide Extracellular polysaccharide 1 Extracellular plysaccharide 11 Extracellular polysaccharide

RpfC and other proteins AlgP (AlgR3). IHF CAP-like protein Clp (CAP-like protein) ExoS, ExoR ExpR PhcA-mediated global regulation

Hoemophrlus rnfluenzae

Type-h capsular polysaccharide

Pseudomonar aeruginosa

Alginate extracellular polysaccharide

ZoogIoeo ramigera

Extracellular polysaccharide

Curohac rer freundn Acetohucfer xdrnum Pseudornonas arlanticu

VI antigen Bactenal cellulose Ex1raccllular polysacchande

IslOl6-mediated amplification and reduction of cap DNA DNA regions affecting 018s. algT. algP Chromosomal rearrangements hy unknoun mechanisms IS insertion\ 1S10.71 insenions Unknown IS element

Salmonella spp and Cirrobacrer freundrr Klebsiella pneumonrne Envrnla amylororo and ErHrnra vlewartri Xanrhomonm cumpesrns Pseudornonas aeruginosa Pseudomonas aeruginosa Xanrhomonas carnpestns Rhrzobium rnelilon Rhrzobiurn mehloh Pseudornonas solamrearurn

R c s R R a C (possibly R a A )

Translational coupling

Escherichio coli Huemophilus injluenzoe Acetobocrer iylirium

KI capsular polysaccharide Type-b capsular polysaccharide Bacterial cellulose

kpsM-kpsT. neuE-neuS b e x A 4 e xB 4 e x C hcs genes

Proteiwprotein interactions

Rhizobium meliloh Rhizobium leguminosarum

Extracellular polysaccharide Extracellular plysaccharide

ExoX-ExoY in EPSl PsiA-PssA

Enzyme activity I . Feedback inhibition

Klebsiello oerogenes

Precursor synthesis

Salmonella enterico Salmonella enterico

Capsular polysaccharide 0-Pol ysaccharide 0-Pol ysaccharide

Acetobocter xylinum

Bacterial cellulose

Escherichio coli

K1 capsular polysaccharide

2. Modulation of activity

Precursor pools

Escherichiu coli

Precursor synthesis Modification of GalU by GalF in UDP-glucose formation Allosteric modification of synthetase by cyclic GMP Degradation of N-acetylneuraminic acid by lyase activity Degradation of UDP-Nacetylglucosamine by phosphatase activity

216

c' WIIITI'1EI.D A N D

M. A V A L V A N O

action of a repressor molecule. This is the only report of transcriptional regulation of rfb genes. In E. coli K-12, increased growth temperatures cause decreased transcription of some rfa genes (Pradel and Schnaitman, 1991). Temperature also influences the LPS profile in other bacteria (McConnell and Wright, 1979; Berry and Kropinski, 1986; Poole and Braun, 1988), but the regulatory processes have not been described. H. REGULATION OF EXTRACELLULAR-POLYSACCHARIDE SYNTHESIS

I . Two-Component Regulatory Systems and Regulation of Group-]-Like Extracellular Polysaccharides Many members of the Enterobacteriaceae can produce a slime polysaccharide called colanic acid (Markovitz, 1977). In most E. coli strains, including K-12, expression of colanic acid is very low at 37°C. Colanic acid production can be activated by growth of conditions including growth in chemically defined media with high concentrations of phosphate and high carbon:nitrogen ratios, and employing incubation at temperatures below 25°C (Markovitz, 1977). Activation also occurs in certain mutants, defining components of a regulatory system for colanic acid. Interaction between these components is relatively well understood in E. coli K-12 and is summarized in Fig. 10. It is now clear that some features of this regulatory system are used by a range of bacteria which produce structurally similar EPSs. These EPSs resemble the group-I CPSs of E. coli but, while it is convenient to include them within group I for the purposes of this discussion, it should be noted that other aspects which define group-I CPSs in E. coli may not be shared by all. It is possible that some of these features, such as the mode and degree of cell association of the polymer, may have no bearing on regulatory requirements for the polysaccharide. In 1964, Markovitz reported overproduction of colanic acid in capR (now known as lon) mutants (Markovitz, 1964); other cap loci were subsequently identified (reviewed in Markovtiz, 1977). The Lon protein is an ATPdependent protease which plays a role in ultraviolet sensitivity, induction of the SOS response and filamentation (for a review, see Gottesman, 1989). In the colanic acid system, the target of Lon is a positive regulator termed RcsA (rcs, regulator of capsule synthesis) (Torres-Cabassa and Gottesman, 1987). In Lon' cells, RcsA has a very short half-life and the amount of RcsA is low (Stout et al., 1991). Mutations in the lon gene increased levels of RcsA, resulting in colanic acid synthesis. This phenotype can be mimicked by introduction of multicopy plasmids carrying rcsA (TorresCabassa and Gottesman, 1987) or by a mutation in rcsA which results in a Lon-resistant RcsA* protein (Stout et al., 1991). The RcsA protein acts

217

CELLSURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BA(JTER1A

possible deactwation of response regulator by sensor/phosphatase

posslMe acbvabon of

and pcssibly other proteins L

I

L

cps promoter

l c p s biosynthetic genes

Fig.10. Proposed mode of action of the Rcs system in regulation of the group-llike extracellular polysaccharide colanic acid in Escherichia coli K-12. The model is modified from Gottesman and Stout (1990). to incorporate recently obtained data (Gervais and Drapeau, 1092; Gervais et af., 1992). Phosphate groups are represented by P.

on cps (capsular-polysaccharidesynthesis) genes by elevating the level of transcription (Gottesman et al., 1985). The protein has a helix-turn-helix DNA-binding motif and belongs to the LuxR family of regulatory proteins (Stout et al., 1991), consistent with its acting through direct interaction with cps DNA sequences. However, DNA binding has not been demonstrated directly and, as already indicated, relatively little is known about the precise structure and function of the cps gene cluster (Gottesman and Stout, 1991).

218

C. WHITF1BI.D AND M

A. VALVANO

Two other genes, rcsB and rcsC, also regulate transcription of cps genes (Gottesman et al., 1985; Brill et al., 1988). The RcsB and RcsC proteins share extensive homology with environmentally responsive two-component regulatory systems (Stout and Gottesman, 1990). These systems typically comprise a sensor protein (often a membrane-spanning protein), responsible for transmitting information from environmental stimuli to a cytoplasmic effector (or response regulator) protein. The response regulators mediate appropriate response in the cell and act at the level of transcription. I n general, the N-terminal conserved domain of the sensor receives the stimulus, resulting in autophosphorylation of a histidine residue in a conserved C-terminal domain. The sensor then acts as a kinase to phosphorylate a conserved N-terminal region of the response regulator; this activates the response regulator. A diverse array of physiological functions, including expression of virulence genes, respond to environmental stimuli using two-component systems (for a review, see Ronson et al., 1987; Albright el al., 1989; Bourret et al., 1989; Miller el al., 1989; Stock ef al., 1989; Mizuno and Mizushima, 1990). I n the colanic acidbiosynthesis system, RcsC is the transmembrane sensor and RcsB is the response regulator; RcsB affects transcription of cps (Stout and Gottesman, 1990). Although RcsC was initially thought to act as a kinase phosphorylating RcsB (Stout and Gottesman, 1990; Gottesman and Stout, 1991), new data suggest that RcsC may instead serve as a phosphatase, dephosphorylating RcsB (Gervais et al., 1992). One possible candidate for the kinase is the recently reported RcsF protein (Gervais and Drapeau, 1992). However, other kinases may also be involved. Phosphorylation of RcsB has not been directly demonstrated, nor has it been shown that RcsB binds DNA sequences, but the homology to proteins where these activities are well characterized is striking. Precise interactions, between Rcs proteins have not been resolved but a logical working model (Fig. lo), has been described, based on results from E. coli K-12. The RcsB protein is thought to play a central role in transcription activation, with RcsA acting in an accessory role. Genetic data suggest that the active form of the response regulator is an RcsARcsB dimer; where RcsA is limited by Lon activity, only low levels of activity can occur (Gottesman et al., 1985; Torres-Cabassa and Gottesman, 1987). Interestingly, multicopy RcsB can compensate for the absence of RcsA, possibly through formation of RcsB homodimers. However, RcsA cannot function without RcsB. Expression of rcsB is regulated (and autoregulated) in a complex fashion, involving sequences for d4binding, a LexA-binding site with a 070 promoter, and an uncharacterized upstream regulatory region (Stout and Gottesman, 1990; Gervais ef al., 1992). The expression shows no absolute

CE1.I.-SURFACE POLYSA('CHAR1DES IN GRAM-NEGATIVE BACTERIA

219

dependence on d4 and it was suggested that the promoter used may depend on both the prevailing environmental conditions and interaction of additional effectors (Gervais et a f . , 1992). Activation of RcsB (presumably through phosphorylation) may result from the activity of RcsF (Gervais and Drapeau, 1992) or occur as a result of cross-talk from other sensors (Gervais et a f . , 1992), affording a versatile system capable of responding to diverse environmental signals. The RcsF and RcsB proteins also interact with FtsZ in regulation of cell division (Gervais and Drapeau, 1992; Gervais et a f . , 1992), indicating that other important cellular functions, in addition to EPS synthesis, could be co-ordinated through a common regulatory element. However, since rcsB mutants are viable, RcsB is not essential for growth. The signal(s) which activates colanic acid synthesis in vivo is still unclear. Activation of this synthesis occurs in an RcsC-dependent fashion when the structure of the outer membrane is perturbed by rfa mutations resulting in deep-rough LPS (Parker et a f . , 1992). However, this is a drastic effect and it has yet to be established that it reflects signals recognized by Rcs proteins in the environment. The rcs genes do not positively regulate group-I1 CPSs in E. cofi,and both colanic acid and a group-I1 capsular K-antigen can be synthesized simultaneously (Jayaratne et a f . , 1993). In contrast, the genes for biosynthesis of colanic acid and group-I CPS in the E. cofi K30 appear to be allelic (Keenleyside et a f . , 1992; see Section V.C.2) and synthesis of K30 polysaccharide is regulated by the rcs system. The rcsA (Keenleyside et a f . , 1992), rcsB and rcsC (P. Jayaratne, W. Keenleyside, R. P. MacLachlan, C. Dodgson and C. Whitfield, unpublished observation) genes from E. cofi 09:K30 are almost identical with their counterparts in E. cofi K-12. The RcsA and RcsB proteins are not essential for synthesis of CPS and capsule formation in E. cofi K30, but are required for high levels of synthesis (Jayaratne el a f . , 1993). Increasing the copy number of R c s A K ~or ~ RcsBK, dramatically increases synthesis of K30 CPS, as do fon (Keenleyside et a f . , 1992) and rcsC (Jayaratne et a f . , 1993) chromosomal mutations. In E. cofi K-12, small amounts of colanic acid result from basal levels of cps transcription (Gottesman et a f . , 1985). Interestingly, all E. cofi strains producing group-I CPSs with structures devoid of amino-sugar residues (e.g. K30) showed an increase in synthesis of K-antigen CPS in response to increased dosages of RcsA and RcsB. These strains do not produce colanic acid. In contrast, strains with aminosugar residue-containing group-I CPS synthesized colanic acid in response to elevated amounts of RcsA and RcsB. These strains produce K-antigen

220

C

W111TFIl-II) A N D M A V A I V A N 0

and colanic acid simultaneously, but is is not clear whether synthesis of both polymers responds to the rcs regulatory system (Jayaratne et al., 1 993). The Vi polysaccharide (see Table 1) resembles group-I1 CPSs and is produced by Citrobacter freuridii and S . enterica serovars typhi, paratyphi C and dublin (Selander et al., 1992). Synthesis of Vi polysaccharide is intriguing because, despite the similarity to group-I1 like CPSs, the regulatory system is that used for group-I CPSs. Two chromosomal loci are involved in synthesis of Vi antigen. The polymer is synthesized by products of genes at the viaB locus (Johnson and Baron, 1969; Snellings et al., 1977, 1981). The viaB locus encodes polypeptides involved in assembly while some viaB mutants accumulate intracellular Vi polymer, resembling mutations in regions 1 and 3 of the group-TI CPS gene clusters in E. coli (Ou et al., 1988; Hashimoto et al., 1991; Kolyva et al., 1992). The unlinked viaA locus is involved in regulation and is present in E. coli K-12 and S. typhimurium, although these bacteria do not contain viaB, nor do they produce Vi antigen (Johnson and Baron, 1969; Snellings et al., 1977). The viaA locus in E. coli K-12 has now been identified as rcsB and, presumably, rcsC (Houng et al., 1992). Furthermore, synthesis of Vi polysaccharide is also rcsA-dependent. Mutants of S . typhi with viaA defects are complicated by the cloned rcsR gene from E. coli K-12 (Houng et a f . , 1992). Homologues of RcsA are found in K. pneumoniae (Allen et al., 1987; McCallum and Whitfield, 1991), Erwinia stewartii (Torres-Cabassa et al., 1987; Poetter and Coplin, 1991) and Erwinia amylovora (Bernhard et al., 1990; Chatterjee et al., 1990; Coleman et a l . , 1990). The RcsA proteins show a high degree of conservation (Poetter and Coplin, 1991) and each can functionally replace RcsAK-12 in E. coli. The cps genes of Er. stewartii are transcriptionally regulated by RcsAEs(Torres-Cabassa et al., 1987). In Er. amylovora, RcsAk., has a dual role; it transcriptionally activates cps genes involved in synthesis of the EPS amylovoran (Leigh and Coplin, 1992) and is also involved in levan synthesis (Bernhard et al., 1990). Homologues of RcsB-RcsC have been identified in Er. stewartii (Leigh and Coplin, 1992) and Er. amylovora (Roberts and Coleman, 1991). In K. pneumoniae, RcsAK, is involved in synthesis of the serotype-specific Kantigen (McCallum and Whitfield, 1991); RcsB and RcsC homologues have not yet been identified in K. pneumoniae. These diverse bacteria use EPSs as virulence determinants, and very different environmental stimuli would be expected to activate their expression. Stimuli recognized by these conserved regulatory systems in vivo remain unknown. It seems logical to expect that interaction of Rcs regulatory proteins proposed for E. coli will also occur in these bacteria.

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

22 1

In some K . pneumoniae strains there is an additional twist to regulation of CPS synthesis. Strains belonging to serotypes K1 and K2 carry rmpA and rmpB genes on a large virulence plasmid (Nassif et al., 1989a,b). The rmpA gene activates colanic acid synthesis when introduced into E. coli K-12 at 30°C, but activation at 37°C requires participation of both rmpA and rmpB. The RmpA protein can functionally replace RcsAK.12 and activates synthesis of colanic acid in E. coli K-12 using an RcsBK.12dependent pathway (Nassif et al., 1989b). The RmpB protein stabilizes activity of RmpA in an E. coli K-12 Lon' background (Vasselon el al., 1991). In K. pneumoniae K2, rmpA mutants still produce a K2 capsule, and RmpA-RmpB only seem to be required for high-level expression of a mucoid polysaccharide (Nassif et al., 1989a). Interpretation of the precise function of RmpA and RmpB is not yet possible since Nassif and his colleagues were unable to confirm that the extra mucoid polysaccharide was related to the K2 CPS; the product was definitely not colanic acid. Klebsiella pneumoniae K1 and K2 strains contain rcsA (Allen et al., 1987; McCallum and Whitfield, 1991). Furthermore, expression of cps genes of K . pneumoniae K2 in E. coli is stimulated by RmpA (Arakawa et al., 1991) and requires R C S A ~and - , ~RcsBK-12 (Wacharotayankun et al., 1992). These observations suggest that actvity of RmpA-RmpB is superimposed on the rcs system in an unknown fashion. 2. Two-Component Regulatory Systems and Synthesis of Alginate Extracellular Polysaccharides Pseudomonas aeruginosa Alginate production is characteristic of Pseudomonas aeruginosa isolates from lungs of cystic fibrosis patients. The mucoid phenotype of P. aeruginosa is enhanced by specific laboratory culture conditions, including nutrient limitation, dehydration, high osmolarity and presence of surfactants. These conditions reflect the environment predicted within the lung, suggesting that specific features of this ecological niche are important for expression of the mucoid phenotype in vivo (DeVault et al., 1989; Deretic et al., 1991; May et al., 1991). One target for regulation of alginate synthesis, algD, is well characterized. The AlgD protein (GDP-mannose dehydrogenase) catalyses conversion of GDP-mannose to GDP-mannuronic acid and represents the first committed step in alginate synthesis (Deretic et al., 1987a; Roychoudhary et al., 1989). Transcription of algD is influenced by a range of regulatory proteins, including a two-component regulatory system. One response regulator, AlgRl (Deretic et al., 1989a; DeVault et al., 1989), has been isolated in a complex with its protein kinase AlgR2 (Roychoudhary et al., 1992b). The AlgR2 protein is also known as AlgQ (Deretic and Konyecsni, 1989). It is

222

C. WHITFlE1.D A N D M A V A I . V A N 0

a small cytoplasmic protein that does not resemble other known twocomponent kinases (Kato et al., 1989). Moreover, the protein is unusual in being phosphorylated in vitro by GTP in addition to ATP, and it has been argued that this feature may facilitate sensing of the physiological status (e.g. nitrogen or phosporous starvation) in the cells (Roychoudhary et al., 1992a). The protein has an exceptionally high affinity for ATP, potentially allowing phosphorylation at very low intracellular ATP concentrations (Roychoudhary et al., 1992a). While it is clear that some alginate-promoting environmental conditions are sensed by AlgRI-AlgR2, others are not. For example, algR1 mutants show a limited response to osmolarity but no longer activate alginate synthesis in response to the concentration of the nitrogen source (Mohr et al., 1990). Mutations in a chromosomal region, termed muc, result in a stable mucoid phenotype (Fyfe and Govan, 1980). The muc mutations can result in constitutive transcription of algRl and algD genes, and the response of algRZ and algD to further environmental stimuli varies in a muc alleledependent fashion (Deretic et al., 1990). The algN, algS and algT genes map near muc and appear to be involved in the switch to the mucoid phenotype. Although their precise role is unclear, algS is thought to act on algT to induce alginate synthesis, with algN acting as a negative regulator (Flynn and Ohman, 1988; Ohman et al., 1990). One interpretation of the muc mutations (Deretic ef al., 1989b, 1990, 1991) is that they effectively deregulate interactions between sensors and response regulators. Alternatively, these mutations could influence superhelicity. It is known that expression of algD is affected by DNA supercoiling (see p. 221). The AlgB protein provides an additional response regulator for alginate synthesis (Wozniak and Ohman, 1991; Goldberg and Dahnke, 1992). Unlike non-mucoid mutants, afgB mutants synthesize small amounts of alginate; AlgB is therefore required only for high-level alginate expression (Goldberg and Ohman, 1987). Although AlgB and AlgRl share conserved N-terminal phosphorylation domains, phosphorylation of AlgB by a sensor/ kinase has not been demonstrated. Enhanced algD transcription requires the regulatory proteins AlgRl (Deretic et al., 1987b), AlgR2 (Kato et al., 1989) and AlgB (Wozniak and Ohman, 1991; Goldberg and Dahnke, 1992). It has been demonstrated that AlgRl binds to two 14 bp DNA sequences located at positions -380 and -457 upstream of algD (Mohr et al., 1990; Kato and Chakrabarty, 1991). The AlgB protein contains a helix-turnhelix DNA-binding motif, but direct binding to DNA has not been demonstrated (Wozniak and Ohman, 1991; Goldberg and Dahnke, 1992). The C-termini of AlgB and AIgRl differ, perhaps indicating binding to different DNA sequences (Goldberg and Dahnke, 1992).

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

223

3. Two-Component Systems and Transcriptional Regulators of Extracellular-Polysaccharide Synthesis in Plant-Associated Bacteria

Regulation of EPS synthesis is being investigated in a variety of plantassociated bacteria. At this stage, some components have been identified but, in most cases, targets for regulation and mechanism(s) of action have not been characterized. Regulatory elements influencing synthesis of EPS by Xunthomonas campestris appear to be part of global virulence regulons, and co-ordinately regulate exported enzymes with xanthan gum. As with many bacteria, X . campestris contains multiple two-component regulatory systems (Osbourn et al., 1990). Two xanthan-regulating genes have partial DNA sequences and predicted proteins resembling RcsB and RcsC. Furthermore, these two ORFs appeared to have convergent transcription, an organizational feature found in rcsB-rcsC (Stout and Gottesman, 1990). A cluster of negative regulatory genes has been identified (Tang et al., 1990) and an unlinked positive regulatory cluster termed rpf (regulation of pathogenicity factors) has also been described (Tang et al., 1991). There may be seven genes within rpf and their gene products possibly interact. Only rfpC has been studied in any depth, and the predicted RpfC protein contains conserved regions found in both sensors and response regulators in other systems. The RpfC protein contains transmembrane domains and the sensor region is similar to RcsC and EnvZ; the C-terminal domain resembles NtrC, OmpR and CheY. Other examples of dual sensor-response regulator proteins include BvgC, which regulates multiple virulence factors in Bordetella pertussis (Arico et al., 1989; Stibitz et al., 1989), VirA, which regulates tumour-induction genes in Agrobacterium turnefaciens (Leroux et al., 1987) and ArcB (Iuchi et al., 1990), which is involved in global repression of aerobic function operons. The phcA gene in Pseudomonas solanacearum provides an additional example of global regulation. Strains defective in phcA undergo phenotype conversion in which the amounts of EPS and exported enzymes decrease, while motility increases (Brumbley and Denny, 1990). The PhcA protein is therefore a positive effector for EPS synthesis. Recent studies have indicated that PhcA shares homology with NahR, a transcriptional activator involved in naphthalene degradation in Pseudomonas putida (Brumbley et al., 1991). Rhizobium meliloti produces two structurally distinct EPSs (EPSI and EPSIIEPSb) and synthesis of each polymer is determined by a separate biosynthesis gene cluster (Glazebrook and Walker, 1989; Zhan et al., 1989) with its own regulatory components. Both exoS and exoR are negative regulators of EPSI (Doherty et al., 1988; Reuber et al., 1991) and a mutation in exoR results in transcriptional activation of ex0 (EPSI biosynthesis) genes. The protein ExoR may be involved in regulation of ex0 genes in

224

C. WHITFlE1.D AND M. A. VALVANO

response to a nitrogen source, but it does not resemble other known regulatory proteins (Reed et al., 1991b). The EPSb polymer is not synthesized in EPSI-producing strains unless multicopy EPSb-biosynthesis genes are introduced (Glazebrook and Walker, 1989; Zhan et al., 1989) or a regulatory factor is mutated. The genes mucR and expR define negative regulatory elements for EPSb and mutants with defects in mucR (Zhan et al., 1989) or expR (Glazebrook and Walker, 1989) synthesize EPSb. The gene expR acts at the level of transcription (Glazebrook and Walker, 1989). Recently, it was shown that EPSb is also under transcriptional control of a phosphate regulon and is synthesized in response to phosphate limitation. Since an expR mutation eliminated phosphate repression of EPSb synthesis, this gene probably plays a role in an environmental sensory system (Zhan et al., 1991). 4. Involvement of Histone-Like Proteins in the Synthesis of Alginate in Pseudomonas aeruginosa

It has been suggested that local DNA topology is critical in expression of environmentally regulated genes (Higgins et al., 1990b). The observation that subinhibitory concentrations of DNA-gyrase inhibitors influence algD transcription (Berry et al., 1989; DeVault et al., 1989, 1990) is consistent with an important role for DNA supercoiling in synthesis of alginate by P. aeruginosa. The AlgP protein (Deretic and Konyecsni, 1990; Konyecsni and Deretic, 1990), also known as AlgR3 (Kato et al., 1990), is one regulatory element which may be involved at this level. The AlgP (AlgR3) protein resembles histone H1 in its structure, properties and cellular distribution (Deretic and Konyecsni, 1990; Deretic etal., 1991, 1992; Kato et al., 1990; Konyecsni and Deretic, 1990). The AlgP (AlgR3) protein influences alginate synthesis at the level of algD expression. It has been proposed that AlgP (AlgR3) binds upstream of the algD promoter and induces DNA bending to bring the bound response regulator, AlgRl (and perhaps AlgB), into closer proximity to the promoter (Deretic et al., 1991). The AlgR1 protein binds unusually far upstream of the algD promoter, and DNA bending would allow AlgRl to interact with, o r influence binding of, the o-factor. An analogue of the histone-like protein IHF increases algD transcription and may also contribute to DNA folding (Wozniak, 1992).

5 . Involvement of CAP-Like Proteins in Extracellular- Polysaccharicle Synthesis The clp (catabolite activator-like protein) gene from X. campestris partially complements crp mutations in E . coli (Dong and Ebright, 1992). As might

CELL.-SURFACE POI.YSACCHARIDES 1N GRAM-NEGATIVE BACTERIA

225

be expected, the Clp protein shares conserved regions with the CAP (CRP) protein from E. coli and the DNA-binding specificityof the two homologues appears to be similar (Dong and Ebright, 1992). The Clp protein is a regulator of several virulence factors in X. campestris (DeCrecy-Lagard et al., 1990).The protein is a positive regulator of synthesis of the EPS xanthan gum, and also acts as both a positive and negative effector for synthesis of several different exported enzymes and pigments. Mutations in clp of X. campestris produce smaller amounts of xanthan gum and the resulting polymer has a lower pyruvate content but normal degrees of O-acetylation. The target for Clp is presently unknown. A CAP-like activity has also been implicated in regulation of alginate biosynthesis in P. aeruginosa (DeVault et al., 1991). Transcription of algD was shown to be repressed by glucose in both E. coli and P. aeruginosa, while expression in E. coli requires cAMP-CAP. Complexes of CAMPand CAP were directly shown to bind to specific DNA sequences upstream of the algD promoter, and deletion of a DNA region containing the putative CAP-binding consensus sequence eliminated the glucose repression. The physiological response of P. aeruginosa to glucose differs from the typical catabolite repression seen in enteric bacteria, and DeVault et al. (1991) speculated that the CAP homologue in P. aeruginosa may be CAMPindependent. The CAP-binding consensus sequence upstream of algD is located very near to the AlgR1-binding site (DeVault et al., 1991) and it is possible that CAP-like proteins participate in DNA bending in a fashion similar to that proposed for AlgP (AlgR3) (Deretic et al., 1991). Whether AlgP (AlgR3), IHF homologues and CAP-like proteins work in a concerted fashion has not been resolved. A clp-gene probe from X. campestris hybridized to P . aeruginosa DNA at low stringency, suggesting the possibility of similar regulatory components in their respective EPSs.

6. IS Elements and DNA Rearrangements in Instability of Extracellular-Polysaccharide Synthesis Instability of EPS synthesis is a characteristic seen in many different bacteria, and a number of different underlying mechanisms have been reported. In strains of Huemophilus influenzae type b belonging to phylogenetic division I, instability results from the structure of the cap locus itself. In division-I strains belonging to serotypes a-d, the cup locus is flanked by an IS-like element (IS1016)(Kroll et al., 1991; Kroll, 1992). The genetic element IS1016 is also found in division-I1 strains, but its location does not flank cap. The cap locus was originally isolated from H. influenzae strain Egan (type b, division I) and contained a tandemly duplicated 18 kb region, arranged in a direct repeat (Hoiseth et al., 1986).

226

C' WtlITI~Il-II ) A N D M A VAL V A N 0

The duplication is not perfect, since the only functional copy of bexA is located between the two duplicated segments. Most of the second copy of bexA at the end of the region is missing, together with part of the adjacent flanking IS1016 (Kroll et al., 1991). A high-frequency conversion from Cap b to Cap- is associated with loss of cap DNA (Hoiseth et al., 1985) through a rec-dependent process (Hoiseth et a l . , 1986). The rearrangement is thought to involve a recombination event between tandem copies of cap, reducing the locus to a single copy. During this process, the only functional bexA gene can be disrupted, resulting in intracellular accumulation of CPS (Kroll et al., 1988). These strains grew either poorly, or not at all, and they were rescued by secondary mutations in cap which switch off CPS synthesis (Brophy et a l . , 1991). The requirement for functional BexA presumably provides the pressure to maintain the duplication in divisionI strains. Amplification of cap can also occur (Kroll and Moxon, 1988) through IS1016-mediated recombination events (Kroll, 1992) and multiple copies of cap can exist on the chromosome (Kroll et al., 1991; Hoiseth el al., 1992). There is a gene-dosage effect with cap and the amount of CPS increases with increased copy number (Kroll and Moxon, 1988). The advantage of the instability and gene-dosage phenomena in vivo is unclear since single-copy cap clusters of different serotypes in phylogenetic division I1 would not undergo this transition, and serotype-b strains with a functional single copy of cap are pathogenic (Ely et al., 1986; Kroll and Moxon, 1988). The CPS plays a protective role but it has been suggested that diminished amounts of CPS may promote adhesion to epithelia and invasion (St Geme and Falkow, 1902). In this sense, appropriate modulation of CPS synthesis at specific phases of infection would be an asset (Brophy et a l . , 1991). It has not been established whether Cap- cells can recover a Cap' phenotype by transformation in vivo. In vivo chromosomal DNA rearrangements in P. aeruginosa have been shown to influence synthesis of LPS and several other extracellular virulence factors, in addition to activating alginate synthesis (Woods et al., 1991). The precise changes which influence the alg system and the mechanism involved have not been resolved. Alterations in DNA structure have been detected in the algP (algR3) gene of non-mucoid strains (Deretic and Konyecsni, 1990). It was speculated that DNA rearrangements in repetitive regions within algP can generate deletions, possibly leading to defective versions of AlgP. This would result in a switch from the mucoid to the non-mucoid phenotype. Such deletions and rearrangements also result in instability in EPS synthesis in Zoogloea ramigera I-16-M (Easson et al., 1987). The mechanism has not been resolved although insertion of Tn5 in the EPS genes increased the frequency of deletion. Inactivation of EPS genes through IS-mediated insertion events has been +

CEI.I.-SURI'ACb POL.YSA(THARII)ES IN GRAM-NEGATIVE RACTERIA

227

reported in several systems. The best documented example is in C. freundii, where instability in the vial? locus for Vi-polysaccharide synthesis gives rise to a high-frequency reversible switch between Vi' and ViP cells (Snellings et a f . , 1981). Interestingly, the viaB genes in Salmonella spp. do not display this instability (Snellings et af., 1981). When the viaB locus from C. freundii was cloned i n E. coli K-12, an irreversible high-frequency switch from Vi' to Vi- resulted from the insertion of IS/-like element in a "hot-spot'' in viaB (Ou et af., 1988). Since the process in C. freundii is reversible, mechanisms that operate in wild-type C. freundii and in E. coli recombinants may be different. Irreversible inactivation of the chromosomal cellulose-synthesis gene cluster in Acetobacter xyfinumoccurs through insertion of IS1031 (Coucheron, 1991). In contrast, precise excision of a mobile genetic element from chromosomal genes has been proposed for reversible EPS' to EPS- switching in the marine bacterium Pseudomonas atfuntica (Bartlett et a f . , 1988). The element possessed characteristics typical of an IS element but showed no homology with known IS elements from fseudomonas spp. The element was detected in multiple copies o n a chromosome and appeared to insert using a duplicative event. 7. Regulation of Enzyme Activity in Precursor Formation

Production of precursors provides a convenient point for regulation of polymer synthesis. At the simplest level, Sutherland (1977b) suggested that using particular nucleotides for synthesis of specific polymers, for example ADP-glucose for glycogen and UDP-glucose for LPS and CPS, provides a way to separate different biosynthetic systems functionally. Kornfield and Ginsburg (1966) studied synthesis of GDP-mannose and GDP-fucose in enteric bacteria which produced cell-surface polysaccharides containing either mannose or fucose residues in isolation, or both residues simultaneously. Feedback inhibition was shown to control activities of the precursor synthetases appropriately. Feedback inhibition of UDP-glucose and TDP-glucose for LPS biosynthesis indicates that this may be a common strategy (Bernstein and Robbins, 1965). This strategy is not universal and other precursor synthetases, such as the CMP-sialic acid synthetase involved in CPS synthesis in E. coli K1, is not regulated by feedback inhibition. Instead, the intracellular levels of sialic acid appear to be controlled by the activity of the catabolic enzyme N-acylneuraminate-pyruvatelyase (adolase) (Vimr and Troy, 1985). A specific phosphatase may regulate levels of UDPN-acetylglucosamine in E. coli (Melo and Glaser, 1966), suggesting that this approach may also be widespread. An unusual situation exists with UDP-glucose pyrophosphorylase (GalU),

228

C. WIIITFIEI.D A N D M A . V A L V A N 0

the enzyme responsible for UDP-glucose synthesis in S . enterica serovar typhimurium. Three forms of the enzyme exist and each shows different catalytic and regulatory properties; all three forms contain a common GalU polypeptide (Nakae and Nikaido, 1971a). The interconversion between forms is thought to involve galF (Nakae and Nikaido, 1971b; Makela and Stocker, 1984). In the absence of GalF, GalU exists as a dimer. In the presence of GalF, GalU can exist as a modifiable monomer. Although galF is thought to be located near rfb, its identity is presently unknown.

8. Allosteric Activation of Cellulose Synthetase Synthesis of bacterial cellulose in Ac. xylinum is well characterized (reviewed in Ross et al., 1991). Optimal activity of purified cellulase synthase requires an unusual effector molecule. This effector has been identified as cyclic diguanylic acid (c-di-GMP) (Ross et al., 1985, 1986, 1987, 1990) and it is thought to act as an allosteric activator (Ross et al., 1991). The target is clearly the cellulase synthase complex, and c-di-GMP elevates the level of activity by as much as 200-fold. The effector is synthesized from pppGpG by diguanylate cyclase (Ross et al., 1987). As with adenylate cyclase, diguanylate cyclase exists in both soluble and membrane-bound forms, and this feature may allow the enzyme to respond to changing environmental cues (Ross etal., 1991). Factors which affect the guanylate cyclase are unknown. However, it was recently shown that the product of the bcsA gene (bcs, bacterial cellulose synthesis), a part of the biosynthetic gene cluster, is required for activity of both guanylate cyclase and cellulose synthase. Both of these proteins are made in an inactive form in bcsA cells (Wong er al., 1990). Attenuation of the effector signal is achieved by degradation of c-di-GMP by phosphodiesterases A and B (Ross et al., 1987). Phosphodiesterase A is highly specific for c-di-GMP (Ross et al., 1990). Ross et al. (1991) described several potential roles for the effector as a regulator of polymerization, but the precise role played by c-di-GMP is unclear. However, the observation that c-di-GMP also operates in cellulose synthesis in Ag. tumefaciens suggests that the function is conserved in different bacteria (Amikam and Benziman, 1989). 9. Protein-Protein Interactions and Translational Regulation of Extracellular-Polysaccharide Synthesis

Polysaccharide-biosynthesis enzymes are arranged in multicomponent complexes. Activities of the various components are interdependent and it is possible that the stoichiometry of the individual components is important. For example, certain kps mutations affecting export of the CPS

CELLSURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

229

in E. coli K1 lower the activity of polysialyltransferase and ultimately affect the amount of polymer synthesized (Vimr et al., 1989). In contrast, mutations in regions D and E of the gene cluster regulating synthesis of the CPS in Neisseria meningitidis group B show increased synthesis and have been suggested to be regulatory regions (Frosch et al., 1989). It is possible that these regulatory phenomena result simply from an imbalance in an important stoichiometry of the components within the functional biosynthetic complex. The importance of balance is supported by the frequent occurrence of translational coupling in biosynthetic genes. This is seen in kpsM and kpsT (Pavelka et al., 1991) and neuE and neuS (Steenbergen er a f . , 1992) in E. coli K1, in bexA, bexB and bexC of H. influenzae (Kroll et al., 1990), and in the bcs genes of Ac. xylinum (Wong et al., 1990). Stoichiometry requirements could give rise to problems in cloning portions of gene clusters, particularly with high copy-number vectors and may explain some observations in EPSI synthesis in Rhizobium spp. The ExoX protein from R. meliloti and its counterpart in R. leguminosarum, PsiA, are described as negative regulators of synthesis of EPSI. Mutations in enoX and psiA result in overproduction of EPSI: EPSb is not affected (Zhan et al., 1991). The gene products ExoX and PsiA appear to act at a point beyond translation of exo-specific mRNA (Zhan and Leigh, 1990; Latchford et al., 1991; Reed et al., 1991a). Inhibition of EPSI synthesis caused by multicopy exoxcan be overcome by multicopy exoY in R. meliloti (Gray et al., 1990; Zhan and Leigh, 1990). The protein ExoY has homologues in other Rhizobium spp. and shows some homology with GumD, the glucosyltransferase that initiates xanthan-gum synthesis in X. campestris (see Fig. 4 ) (Reed et a f . , 1991a). The effects of ExoX and PsiA are also counterbalanced by ExoF (Zhan and Leigh, 1990) and PssA (Borthakur et al., 1988), respectively. The ExoF and PssA proteins are both essential for EPSI synthesis and are thought to be part of the biosynthetic complex (Borthakur and Johnson, 1987; Gray et al., 1990). Taken together, these observations suggest that ExoX and PsiA function by interacting with various components of the EPSI biosynthetic complex. Possibilities include translocation of EPSI , or steps leading to formation of the undecaprenol-linked repeating unit (Reed et al., 1991a). In this respect, it is particularly intriguing that mutations in exoC and exoJ (Long et al., 1988), now known to be located in exoX and in the intergenic region between exoX and exoY (Reed et al., 1991a), result in release of a low molecular-weight form of EPSI in R. meliloti. VII. Conclusions Application of genetic and biochemcial approaches has facilitated detailed analysis of complex, multicomponent systems, such as those involved

230

C. WHI’1PIEI.D A N D M. A . V A L V A N 0

in synthesis of cell-surface polysaccharides. An understanding of the mechanisms of synthesis first requires that the number of components be established, and this information is now available for some systems. Once individual proteins have been characterized at the biochemical level, the next challenge will be to assemble the biosynthetic complex from purified component parts. There are several intriguing questions which are currently unanswered, but towards which progress is being made: (a) What is the precise role played by undecaprenol lipids? Possibilities include their acting as molecular scaffolds for assembly of the enzyme complex, as transport molecules for oligosaccharides, or as regulatory molecules which, by their limited amounts, control activities of different polysaccharide-biosynthesis systems. (b) How are polysaccharides transported across the cytoplasmic membrane, periplasm and outer membrane? Some hypotheses have been proposed for transport across the cytoplasmic membrane and, for some polysaccharides, periplasmic components are identified. However, the mechanisms remain to be elucidated. The process(es) involved in transport across the outer membrane is also unknown. (c) How are the biosynthesis systems regulated? Many regulatory components have been identified but the range of targets for regulatory molecules is often not established. Although precise details of the interaction between the regulatory and biosynthetic machinery are known for some systems, the signals to which the systems respond have not been resolved. While this review was at proof stage, a number of research advances have occurred which are particularly relevant to our discussion (see the Addendum on p. 307).

VIII. Acknowledgements The authors wish to acknowledge stimulating discussions held with many friends, colleagues and with members of our own laboratories. Many of these discussions have found their way into this review. We are especially grateful to W. J . Keenleyside, R . P. MacLachlan, C. L. Marolda and E. R. Vimr for their helpful suggestions during the preparation of this review. We thank those colleagues who provided information prior to publication; the list is too long to include here. Research in C. Whitfield’s laboratory is supported by funding from the Medical Research Council, Natural Sciences and Engineering Research Council, and the Canadian Bacterial Diseases Network. M. A . Valvano acknowledges financial support from the Medical Research Council and the National Sciences and Engineering Council.

C'EI.I.-StJRI~AC'FPOI.YSAC('tIARIDES IN GRAM-NEGATIVE BACTERIA

23 1

REFERENCES Abe, M.. Amemura, A . and Higashi, A . (1982).Plant and Soil64, 315. Achtman, M.. Mercer. A., Kusecek, B.. Pohl, A , , Heuzenroeder, M., Aaronson, W . , Sutton, A. and Silver, R. P (1083). Infection and Immunity 39, 315. Adhya, S. (1987).In "Escherichia colr and Salmonella typhimurium, Cellular and Molecular Biology" (F. C. Neidhardt, J . L.. Ingraham, K. B. Low, B. Magasanik, M. Schaechter and H. E. Umharger. eds). pp. 1503-IS12. American Society for Microbiology. Washington, DC. Adlam, C.. Knights, J . M., Mugridgc. A , , Williams. J . M . and Lindon, J. C. (1987).FEMS Microbiology Letters 42, 23. Aird. E . L.. H . , Brightwell, G . , Jones, M. A . and Johnson, A. W. B. (1991). Journal of General Microbiology 137, 2287. Al-Hendy. A , , Tovianen, P. and Skurnik, M. (1991a). Microbial Pathogenesis 10,47. Al-Hendy, A. , Toivanen, P. and Skurnik, M. (IWIb). Microbial Pathogenesis 10, 81. Albright. L. M., Huala, E . and Ausubel, F. M . (1989). Annual Reviews in Generics 23, 311. Allen, P., Hart, C. A and Saunders. J . R. (1987).Journalof General Microbiology 133, 331. Aloni. Y . , Delmer. D. P. and Benziman, S. (1982).Proceedings of the National Academy of Sciences of the United States of America 77. 6448. Amako. K., Meno. Y. and Takade, A . (1988). Journal of Bacteriology 170, 4960. Aman, P., McNeil, M., Franzen, L.-E., Darvill, A . G. and Albersheirn, P. (1981). Carbohydrate Research 95, 263. Amemura. A , , Hisamatsu, M. and Mitani, H . (1983). Carbohydrate Research 114, 277. Amikani, D. and Benziman, M. (l9W). Journal of Bacteriology 171. 6649. Aminofl', D.. Dodyk, F. andRoseman. S. (1963). JournalofBiologicalChemisrryU8,PC1177. Anderson, J . S., Matsuhashi, M., Haskin, M. A. and Strominger, J . L. (1965). Proceedings of the National Academy of Sciences of the United States of America 53, 881. Anderson, R. G . , Hussey. H . and Baddiley, J. (1972). Biochemical Journal, 127, 11. Arakawa, Y.,Ohta, M., Wacharotayankun, R . , Mori, M., Kido, N., Ito, H., Komatsu, T . , Sugiyania, T. and Kato, N. (1991). Infection and Immunity 59, 2043. Arico, B.. Stihitz, S., Royo, C.. Miller, J . and Falkow, S. (1989). Proceedings of the National Academy of Sciences of the United States of America 86, 6671. Arnold, W . , Keller. M., Becker, A , , Roxlau, A , , Niehaus, K., Quandt, J . . Weng, W. M. and Puhler, A . (19%)). Endocytohiosis and Cell Research 8 , 197. Arsenault, T. L., Hughes, D. W., MacLean, D. B., Szarek, W. A , , Kropinski, A . M. B. and Lam, J . S. (199l). Canadian Journal o f Chemistry 69, 1273. Bailey. M. J . A , , Koronakis, V . , Schrnoll, T. and Hughes, C. (1992).Molecular Microbiology 6, 1003. Banouh, J . H . , Shaw, D. H., Pang, H . , Krepinsky, J . J . , Nakhla, N. A. and Patel, T. R. (19%)). Biomedical and Environmental Mass Spectrometry 19, 787. Barbas, J . A , , Diaz, J., Rodgriguez-Tehar, A . and Vazquez, D. (1986). Journalof Bacteriology 165, 269. Barr, K . and Rick, P. D. (1987).Journal of Biological Chemistry 262, 7142. Barr, K.. Nunes-Edwards, P. and Rick, P. D. (1989). Journal of Bacteriology 171, 1326. Bartlett, D.H.. Wright. M . E. andSilverman, M. (1988). Proceedingsofthe NationalAcademy of Sciences of the United States of America 85. 3923. Bastin. D. A . , Romana, L. K. and Reeves, P. R. (1991). Molecular Microbiology 5, 2223. Batchelor, R . A , , Haraguchi, G . E.. Hull, R . A. and Hull, S. 1. (1991). Journalof Bacteriology 173, 56W.

Batchelor. R. A . , Alifano. P., Biffali, E., Hull, S. I. and Hull, R. A. (lW2).Journal o f Racteriology 174, 5228. Batley. M., Redmond, J . W.. Djordjevic, S. R. and Rolfe, B. G. (1987). Biochimica et Riophysica Acta 901, 119. Baycr. M. E . (1968).Journal of General Microbiology 53, 395. Bayer. M. E. (1979).In "Bacterial Outer Membranes: Biogenesis and Function" (M. Inouye, cd.). pp. 167-202. Wiley, New York.

232

C . WHlTFIEI.I> A N D M. A. V A L V A N 0

Bayer, M. E. (1990). Current Topics in Microbiology and Immunology 150, 129. Bayer, M. E. ( I W I ) . Journal of Structural Biology 107, 268. Bayer, M. H. and Bayer, M. E . (1985). Journal of Bacteriology 162, 50. Bayer, M. E . and Thurow, H . (1977). Journal of Racteriology 130, 911. Rayer. M. H., Costello, G . P. and Bayer, M. E. (1982). Journal of Bacteriology 149, 758. Bayer, M. H., Keck, W. and Bayer. M. E . (1990). Journal of Bacteriology 172, 125. Beger, D. W., Heuzenroeder, M. W. and Manning, P. A. (1989). FEMS Microbiology Leners 57, 317.

Belunis. C. J.. Mdlui, K. E., Raetz, C . R. H. and Nano, F. E. ( 1 9 2 ) . Journal of Riological Chemistry 267, 18702.

Bernhard, F., Poetter, K., Geider, K. and Coplin, D. L. (1990). Molecular Plant-Microbe Interactions 3. 429.

Bernstein, R. L. and Robbins, P. W. (1965). Journal of Biological Chemistry 240, 391. Berry, A. DeVault, J. D. and Chakrabarty, A. M. (1989). Journalof Bacteriology 171,2312. Berry, D. and Kropinski, A . M. (1986). Canadian Journal of Microbiology 32, 436. Beutin, L. and Achtman, M. (1979). Journal of Bacteriology 139, 730. Beutin, L., Manning, P. A . , Achtman, M. and Willetts, N. (1981). Journal of Bacteriology 145, 840.

Beynon, L. M., Dumanski, A. J., McLean, R. J. C., MacLean, L. L., Richards, J. C. and Perry, M. B. (1992). Journal of Racteriology 174, 2172. Beveridge, T. J. and Graham, L. L. (1991). Microbiological Reviews 55, 684. Bhagwat, A. A , , Tulley, R . E . and Keister, D. L. (1992). Molecular Microbiology 6, 2159. Bhattacharjee, A. K., Jennings, H. J., Kenny, C. P., Martin, A. and Smith, I. C. P. (1975). Journal of Biological Chemistry 250, 1926.

Borthakur, D. and Johnson, A . W. B. (1987). Molecular and General Genetics 207. 149. Borthakur, D., Barker, R. F., Latchford, J. W., Rossen, L. and Johnson, A. W. B. (1988). Molecular and General Genetics 213, 155.

Bossio, J . C., delannino, N. 1. and Dankert, M. A . (1986). Biochemical and Biophysical Research Communications 134. 205.

Boulnois, G . J. and Jann, K. (1989). Molecular Microbiology 3, 1819. Boulnois, G. J . and Roberts, 1. S. (1990). Current Topics in Microbiology and Immunology 150, I . Boulnois, G . J.. Roberts, I. S . , Hodge, R., Hardy, K. R., J a m , K. and Timmis, K. N. (1987). Molecular and General Genetics 208, 242.

Bourdineaud, J. P., Howard, S. P. and Lazdunski, C . (1989). Journal of Bacteriology 171. 2458.

Bourret, R . B., Hess. J. F., Borkovitch, K. A , , Pakula, A. A. and Simon, M. 1. (1989). Journal of Biological Chemistry 264, 7085.

Branefors-Helander, P., Erbing, C., Kenne, L. and Lindberg, B. (1976). Acta Chemica Scandinavicu Series B 30,276. Brass, J . M.,Higgins, C. F., Foley, M., Rugman, P. A , , Birmingham, J. and Garland, P. B. (1986). Journal of Bacteriology 165, 787. Bray. D. and Robbins, P. W. (1967). Biochemical and Biophysical Research Communications 28,334.

Brazas, R.. Davie. E., Farewell, A. and Rothfield, L. I. (1991). Journalof Bacteriology 173, 6168.

Brill, J . A.. Ouinlan-Walshe, C. and Gottesman. S. (1988). Journalof Bacteriology 170,2590. Brink, B. A., Miller, J . , Carlson, R. W.and Noel, K. D . (1990). Journal of Bacteriology 172, 548.

Brisson, J.-R. and Perry, M. B. (1988). Biochemistry and Cell Biology 66, 1066. Brophy, L. N., Kroll, S. J . , Ferguson, D . J. P. and Moxon, E. R. (1991). Journal of General Microbiology 137, 2571.

Brown, P. K., Romana, L. K. and Reeves, P. R . (1991). Molecular Microbiology 5, 1873. Brown, P. K., Romana, L. K. and Reeves, P. R. (1992). Molecular Microbiology 6 , 1385. Brumbley, S. M. and Denny, T . P. (1990). Journal of Bacteriology 172, 5677.

CELL-SURFACE POLYSACC‘HARIDES IN GRAM-NEGATIVE BACTERIA

233

Brumbley, S. M., Carney, B. F. and Denny, T. (1991). Phyroparhology 81, 1145. Brzoska, P. M. and Signer, E. R. (19Y1). Journal of Bacteriology 173,3235. Buendia. A. M., Enenkel, B., Koplin, R., Niehaus, K., Arnold, W. and Puhler, A. (1991). Molecular Microbiology 5 , 1519-1530. Bureau, T. E. and Brown, R. M. Jr (1987). Proceedings of the National Academy of Sciences of the United States of America 84, 6985. Calva, J . R., Elias, P. M., Turowski, D. A. and Noel, K. D. (1989). Journal of Bacteriology 172,8. Cangelosi, G . A., Martinetti, G . , Leigh, J. A., Lee, C. C., Theines, C. and Nester, E . W. (1989). Journal of Bacteriology 171. 1609. Canter Cremers, H. C. J., Batley, M., Redmond, J. W., Wijfyes, A. H. M., Lugtenberg, B. J. J . and Wijffelman, C. A. (1991). Journal of Biological Chemistry 266, 9556. Caugant, D. A., Levin, B. R., Orskov, I . , Orskov, F., Svanborg-Eden, C. and Selander, R. K. (1985). Infection and Immunity 49,407. Chakraborty, A. K., Friebolin, H. and Stirm, S. (1980). Journal of Bacteriology 141,971. Chatterjee, A , , Chun, W. and Chatterjee, A. K. (1990). Molecular Plant-Microbe Interactions 3,144.

Cheah, K.-C., Beger, D. W. and Manning, P. A. (1991). FEMS Microbiology Leners 83,213. Chitnis, C. and Ohman, D. E. (1990). Journal of Bacteriology 172, 2894. Choy, Y.-M. and Dutton, G . G . S. (1973). Canadian Journal of Chemistry 51, 3015. Chu, L., May, T . B., Chakrabarty, A. M. and Misra, T. K. (1991). Gene 107, 1. Clarke, B. R. and Whitfield, C. (1992). Journal of Bacteriology 174,4614. Clarke, C. A., Beltrame, J. and Manning, P. A. (1991). Gene 107,43. Clementz, T. and Raetz, C. R. H. (1991). Journal of Biological Chemistry 266, 9687. Clover, R. H., Kieber, J. and Signer, E. R. (1989). Journal of Bacteriology 171,3%1. Coleman, M., Pearce, R., Hitchin, E., Busfield, F., Mansfield, J. W. and Roberts, 1. S. (IM).Journal of General Microbiology 136, 1799. Coley. J . , Tarelli, E . , Archibald, A. R. and Baddiley, J . (1978). FEES Letters 88, 1. Collins, L. V. and Hackett, J. (1991). Journalof Bacteriology 173, 2521. Coplin, D. L. and Majerczak, D. R. (1990). Molecular Plant-Microbe Inreructions 3,286. Coucheron, D. (1991). Journal of Bacteriology 173,5723. Creeger, E. S., Chen, J. F. and Rothfield, L. 1. (1979). Journal ofBiologica1 Chemistry 254, 811. Crisel. R. M., Baker, R. S. and Dorman, D. E. (1975). JournalofBiological Chemistry 250, 4926. Curvall. M., Lindberg, B., Lhnngren, J. and Nimmich, W. (1973). Acta ChemicaScandinavica 27,264s. Dalbey, R. E. and Wickner, W. (1985). Journal of Biological Chemistry 260, 15925. Danilov, L. L. and Shibaev, V. N. (1991). In “Studies in Natural Products Chemistry” (AttaurRahman, ed.), vol. 8, pp. 63-114. Elsevier, Amsterdam. Dankert, M., Wright, A , , Kelley, W. S. and Robbins, R. W. (1969). Archivesof Biochemistry and Biophysics 116,425. Davidscm, I. W . , Sutherland, 1. W. and Lawson, C. J. (1977). Journalof General Microbiology 98,603. Day, D . F. and Marceau-Day, M. L. (1982). Current Microbiology 7,93. De Crecy-Legard, V . , Glaser, P., Lejeune, P., Sismeiro, 0..Barber, C. E., Daniels, M. J . and Danchin, A. ( I M ) . Journal of Bacteriology 172,5877. deLeij, I.., Kingma, J. and Witholt, B. (1979). Riochemica et Biophysica Acta 553, 224. Dengler, T., Jann, B. and Jann, K. (1986). Carbohydrate Research 150,223. Denny, T . P. and Baek, S. R. (1991). Molecular Plant-Microbe Interactions 4, 198. Denny, T. P., Makini, F. W. and Brumbley, S. (1988). Molecular Plant-Microbe Interactions 1, 215. Deretic, V . and Konyecsni, W. M. (1989). Journal of Bacteriology 171, 3680. Deretic, V. and Konyecsni, W. M. (1990). Journal of Bacteriology 172, 5544. Deretic, V.,Gill, J. F. and Chakrabarty, A. M. (1987a). Journal of Bacteriology 169,351.

234

C.

WHITFlE1.I) A N D

M A. VALVA N O

Deretic, V., Gill, J. F. and Chakrabarty, A. M. (1987b). Nucleic Acids Research IS, 4567. Deretic, V., Dikshit, R., Konyecsni, W. M., Chakarabarty, A . M. and Misra, T. K. (1989a). Journal of Bacteriology 171, 1278. Deretic, V., Konyecsni, W. M., Mohr, C. D., Martin, D. W. and Hibler, N. S. (1989b). Biol Technology 7, 1249. Deretic, V., Govan, J . R. W., Konyenni, W. M. and Martin, D. W. (1990). Molecular Microbiology 4, 189, Deretic, V., Mohr, C. D. and Martin, D. W. (1991). Molecular Microbiology 5, 1577. Deretic, V.. Hibler, N.S. and Holt, S. C. (1992). Journalof Bacteriology 174, 824. DeVault, J. D., Berry, A., Misra, T. K., Darzins, A . and Chakrabarty, A. M. (1989). Biol Technology 7 , 352. DeVault, J. D., Kimbara, K. and Chakrabarty, A. M. (1990). Molecular Microbiology 4,737. DeVault, J. D., Hendrickson, W., Kato, J. and Chakrabarty, A. M. (1991). Molecular Microbiology 5 , 2503. Devi, S. J . N.. Schneerson, R., Egan, W., Vann, W. F., Robbins, J. B. and Shiloach, J. (1991). Infection and Immunity 59, 732. DiFabio, J. L., Brisson, J. R. and Perry, M. B. (1989). Biochemistry and Cell Biology 67, 78. Ding, M.-J., Svanborg, C., Haraguchi, G. E., Hull, R. A. and Hull, S. 1. (1991). Molecular Pathogenesis 11, 379. Dodds, K. L., Perry, M. B. and McDonald, I. J. (1987). Canadian Journal of Microbiology 33, 452. Doherty, D., Leigh, J. A., Glazebrook, J. and Walker, G. C. (1988). Journalof Bacteriology 170, 4249. Dolph. P. J., Majerczak, D . R. and Coplin, D . L. (1988). Journal of Bacteriology 170, 865. Donahue-Rolfe, A. and Schaechter, M. (1980). Proceedings of the National Academy of Sciences of the United States of America 77, 1867. Dong, Q . and Ebright, R. H. (1992). Journal of Bacteriology 174, 5457. Douglas, D. J.. Staneloni, R. J., Rubin, R. A. and Nester, E. W. (1985). Journal of Bacteriology 161, 850. Drake, C. R., Roberts, I . S., Jann, B., Jann, K. and Boulnois, G. J. (1990). FEMS Microbiology Letters 66, 227. Dunne, W. M., Jr and Buckmire, F. L. A . (1985). Applied and Environmental Microbiology 50, 562. Dutton, G. G. S. and Mackie, K. L. (1978). Carbohydrate Research 62, 321. Dutton, G. G. S. and Merrifield, E. H. (1982). Carbohydrate Research 105, 189. Dylan, T., lelpi, L., Stanfield, S., Kashyap, L., Douglas, C., Yanofsky, M., Nester, E., Helsinki, D. R. and Ditta, G. (1986). Proceedings of the National Academy of Sciences of the United States of America 83, 4403. Easson, D. D., Sinskey, A. J. and Peoples, 0. P. (1987). Journalof Bacteriology 169, 4518. Egan, W., Schneerson, R., Werner, K. E. and Zon, G. (1982). Journal of the American Chemical Society 104, 2898-2910. Ely, S . , Tippett, J., Kroll, J. S. and Moxon, E. R. (1986). Journal of Bacteriology 167, 44. Farewell, A., Brazas, R., Davie, E., Mason, J. and Rothfield. L. I. (1991). Journal of Bacteriolony 173, 5188. Finan, T. MY; Kunkel, B., DeVos, G. F. and Signer, E. R. (1986). Journal of Bacteriology 167, 66.

Finke, A,. Roberts, I., Boulnois, G., Pazzani, C. and Jann, K. (1989). Journalof Bacteriology 171, 3074. Finke, A., Jann, B. and J a m , K. (1990). FEMS Microbiology Letters 69, 129. Finke, A., Bronner, D., Nikolaev, A. J., Jann, B. and Jann, K. (1991). Journalof Bacteriology 173, 4088. Fisher, R. F. and Long, S. R . (1992). Nature 357, 655. Flemming, H.-C. and Jann, K. (1978a). European Journal of Biochemistry 83, 47. Flcmming, H.-C. and Jann, K. (1978b). FEMS Microbiology Letters 4, 203. Flynn, J. L. and Ohman, D. E. (1988). Journal of Bacteriology 170, 3228.

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE B A C E R I A

235

Foulds, J. and Aaronson, A. (1984). In “Abstracts of the 84th General Meeting of the American Society for Microbiology”, p. 54, abstract D21. Frosch, M., Weisgerber, C. and Meyer, T. F. (1989). Proceedings of the National Academy of Sciences of the United Sfares of America 86, 1669. Frosch. M., Edwards, U.,Bousset, K., Krausse, B. and Weisgerber. C. (1991). Molecular Microbiology 5 , 125 1 . Frosch, M. Miiller, D., Bousset, K. and Miiller, A. (1992). Infection and lmmunify 60, 798. Fyfe, J. A. M. and Govan, J. R. W. (1980). Journal of General Microbiology 119, 443. Gaffney, D., Skurray, R. and Willetts, N. (1983). Journal of Molecular Biology 168, 94. Gamian, A. and Romanowska, E. (1982). European Journal of Biochemistry 129, 105. Gangarosa, E. J., Sonati, A., Saghari. H. and Freeley, T. C. (1967), Lancer I, 646. Gardiol, A . E. and Dazzo, F. B. (1986). Journal of Bacteriology 168, 1459. Garegg, P. J., Lindberg, B., Onn, T. and Holme, T. (1971a). Acfa Chemica Scandinavica 25, 1185. Garegg, P. J., Lindberg, B., Onn, T. and Sutherland, I. W. (1971b). Acto Chemica Scandinavica 25, 2103. Gemski, P., Koeltzow, D. E. and Formal, S. B. (1975). Infection and Immunity 11, 685. Gervais, F. G . and Drapeau, G. R. (1992). Journal of Bacteriology 174, 8016. Gervais, F. G., Phoenix, P. and Drapeau, G. R. (1992). Journalof Bacferiology 174, 3964. Glazebrook, J . and Walker, G. C. (1989). CeN 56, MI. Goldberg, J. B. and Dahnke, T. (1992). Molecular Microbiology 6, 59. Goldberg, J. B. and Ohman, D. E. (1987). Journal of Bacferiology 169, 1593. Goldemann, G.. Kanegasaki, S. and Jann, K. (1979). FEMS Microbiology Lefters 5 , 443. Goldman, R. C. and Hunt, F. (1990). Journal of Bacteriology 172, 5352. Goldman, R. C. and Leive, L. (1980). European Journal of Biochemistry 107. 145. Goldman, R. C., White, D., Brskov, I . , Rick, P. D., Lewis, M. S., Bhattacharjee, A. K. and Leive, L. (1982). Journal of Bacferiology 151, 1210. Goldman, R. C., Joiner, K. and Leive, L. (1984). Journal of Bacteriology 159, 877. Robbins, J. B. and Liu, T.-Y. (1981). Gotschlich, E. C., Fraser, B. A,, Nishimura, 0.. Journal of Biological Chemistry, 256, 8915. Gottesman, S. (1989). Annual Reviews in Generics 23, 163. Gottesman, S. and Stout, V. (1991). Molecular Microbiology 5 , 1599. Gottesrnan, S . , Trisler, P. and Torres-Cabassa, A. (1985). Journal of Bacteriology 162, 111 1. Graben, E., Wigender, J. and Winkler, U.K. (1990). FEMS Microbiology Lefters 68, 83. Graham, L. L., Beveridge, T. J. and Nanninga, N. (1991a). Trends in Biochemical Sciences 16, 328. Graham, L. L., Harris, R., Villiger, W. and Beveridge, T. J. (1991b). Journalof Bacferiology 173, 1623. Gray, J. X., Djordjevic, M. A. and Rolfe, B. G. (1990). Journal of Bacteriology 172, 193. Gross, M., Geier, G., Geider, K. and Rudolph, K. (1989). In “Proceedings of the 7th International Conference on Plant Pathogenic Bacteria, Budapest”, pp. 81-84. Academiai Kiado. Guihard, G., Boulanger, P. and Letellier, L. (1992). Journal of Biological Chemistry 267, 3173. Haishima, Y., Host, 0. and Brade, L. (1992). European Journal of Biochemisfry 203, 127. Hale, T. L., Guerry, P., Seid, R. C., Kapfer, C., Winglieid, M. E., Reaves, C. B., Baron, L. S. and Formal, S. B. (1984). Infection and Immunity 46,470. Haraguchi, G. E.. Hull, R. A., Krallman-Wenzel, U. and Hull, S. I. (1989). Microbial Pathogenesis 6 , 123. Haraguchi, G . E., Zahringer, U.,Jann, B., Jann, K., Hull, R. A. and Hull, S. I. (1991). Microbial Pathogenesis 10, 351. Harding, N . E., Cleary, J. M., Cabanas, D. K., Rosen, I. G. and Kang, K. S. (1987). Journal of Bactvriology 169, 2854.

236

C. WHIlTIE1.I) AND M. A. VALVANO

Hashimoto, Y., Ezaki, T., Li, N. and Yamamoto, H. (1991). FEMS Microbiology Leners. 90, 53. Hassler, R. A. and Doherty, D. H. (1990). Biotechnofogy Progress 6 , 182. Haug. A. and Larsen, B. (1971). Carbohydrate Research 17, 297. Heasley, F. A. (1981). Journal of Bacteriology 145, 624. Hellerqvist, C.-G. and Lindberg, A. A. (1971). Carbohydrare Research 16, 39. Hellerqvist. C.-G.. Lindberg, B., Svensson, S., Holme, T. and Lindberg, A. A. (1968). Carbohydrate Research 8 , 43. Hellerqvist, C . - G . ,Larm, O., Lindberg, B., Svensson, S., Holme, T. and Lindberg, A. A. (1969a). Acra Chemica Scandinavica 23, 2217. Hellerqvist. C . G . , Lindberg, B., Svensson, S., Holme, T. and Lindberg, A. A. (1069b). Carbohydrate Research 9, 237. Hellerqvist, C.-G., Lindberg, B., Svensson, S., Holme, T. and Lindberg, A. A. ( 1969~). Acra Chemica Scandinavica 23, 1588. Hellerqvist, C.-G., Lindberg, B., Pilotti, A. and Lindberg, A. A. (1970a). Acta Chemica Scandinavica 24, 1168. Hellerqvist, C.-G.,Lindberg, B., Svensson, S., Holme, T. and Lindberg, A. A. (1970b). Carbohydrate Research 14, 17. Hellerqvist, C . - G . ,Hoffman, J., Lindberg, B.. Pilotti, A. and Lindberg, A. A. (1971a). Acta Chemica Scandinavica 25, 1512. Hellerqvist, C . G . , Lindberg, B., Ldnngren, J. and Lindberg, A. A. (1971b). Carbohydrare Research 16, 289. Hellerqvist, C . - G . ,Lindberg, B., Mnngren, J. and Lindberg, A. A. (1971~).Acra Chemica Scandinavica 25. 601. Hellerqvist, C.-G., Lindberg, B., Samuelson, K. and Lindberg, A. A. (1971d). Acta Chemiru Scandinavica 25, 955. Hellerqvist, C.-G.,Hoffman, J.. Lindberg, A. A., Lindberg, B. and Svensson, S. (1972). Acta Chemica Scandinavica 26, 3282. Heuzenroeder, M. W., Beger, D. W., Thomas. C. J. and Manning, P. A. (1989). Molecular Microbiology 3, 295. Heyns, K. and Kiessling, G. (1967). Carbohydrate Research 3, 340. Higa, H. H. and Vaarki, A. (1988). Journal of Biological Chemistry 263, 8872. Higashi. Y., Strominger, J. L. and Sweeley, C. C. (1967). Proceedingsof the National Academy of Sciences of the United States of America 57, 1878. Higgins, C. F., Hiles. 1. D., Salmond, G. P. C., Gill, D. R., Downie, J. A,, Evans, I. J.. Holland, 1. B., Gray, L., Buckel, S. D., Bell, A. W. and Hermondson, M. A. (1986). Nature 323, 448. Higgins, C. F., Gallagher, M. P., Hyde, S. C . , Mimmack, M. L. and Pearce, S. R. (1990a). Philosophical Transactions of rhe Royal Sociery of London B 326, 353. Higgins, C. F., Hinton, J. C. D., Hulton, C. S. J., Owen-Hughes, T., Pavitt, G . D. and Seraifi, A. (1990b). Molecular Microbiology 4, 2007. Hisamatsu, M.. Abe, J., Amemura, A. and Harada,T. (1978). Carbohydrate Research 66.280. Hitchcock, P. J. and Brown, T. M. (1983). Journal of Bocteriology 154, 269. Hitchcock, P. J., Leive. L., Makela, P. H., Rietschel, E. T., Strittmatter, W. and Morrison, D. C. (1986). Journal of Bacteriology 166, 699. Hobot, J. A., Carlemalm, E., Villiger, W. and Kellenberger, E. (1984). Journul of Bacteriology 160, 143. Hoiseth, S . K., Connelly, C. J. and Moxon, E. R. (1985). Infection and Imrnuniry 49. 389. Hoiseth, S. K., Moxon, E. R. and Silver, R. P. (1986). Proceedings of the National Arudemy of Sciences of rhe United Srares of America 83, 1106.

Hoiseth, S . K.,Corn, P. G. and Anders, J. (1992). Journalof Infectious Diseuses 165, S114. Holst, 0..Zahringer, U., Brade, H. and Zamojski, A. (1991). Carbohydrate Research 215. 323. Hbltje, J.-V.. Fiedler. W., Rotering, H., Walderich, B. and vanDuin, J. (1988). Journal of Biological Chembrry 263, 3539.

CELL-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

237

Homonylo, M. K., Wilmot, S. J., Lam, J. S., MacDonald, L. A. and Whitfield, C. (1988). Canadian Journal of Microbiology 34, 1159. Hotte, B., Rath-Arnold, I. R., Piihler, A. and Simon, R . (1990). Journal of Bacteriology 172, 2804.

Houng, H . 4 . H., Noon, K. F., Ou, J. T . and Baron, L. S. (1992). Journal of Bacteriology 174, 5910.

lelpi, L., Couso, R. 0. and Dankert, M. (1981a). FEBS Letters 130, 253. lelpi, L., Couso, R. 0. and Dankert, M. (1981b). Biochemical and Biophysical Research Communications 102. 1400. lelpi, L., Dylan, T., Ditta, G. S., Helsinki, D. R. and Stanfield, S. W. (1990). Journal of Biological Chemistry 265, 2843. Ishidate, K., Creeger, E. S., Zrike, J . , Deb, S.. Glauner, B., MacAlister, T. J . and Rothfield, L. 1. (1986). Journal of Biological Chemistry 261, 428. Ishiguro, E. E . , Vandenvel, D. and Kusser, W. (1986). Journal of Bacteriology 168, 328. Iuchi. S., Matsuda, Z., Fujiwara, T . and Lin, E. C. C. (1990). Molecular Microbiology 4,715. J a m , K. and J a m , B. (1984). In “Handbook of Endotoxin. Chemistry of Endotoxin” (E. T. Rietschel, ed.), vol. I, pp. 13S186. Elsevier, Amsterdam. Jann, K. and Jann, B. (1990a). Current Topics in Microbiology and Immunology 150. Jann, K. and Jann, B. (1990b). Current Topics in Microbiology and Immunology 150. 19. Jann, K. and J a m , B. (1992). Canadian Journal of Microbiology 38, 705. J a m , K.. Kanegasaki, S., Goldemann, G. and Makela, P. H. (1979). Biochemical and Biophysical Research Communications 86, 1185. J a m , K.. Goldemann, G., Weisgerber, C., Wolf-Ullisch, C. and Kanegasaki, S. (1982). European Journal of Biochemistry 127, 157. J a m , K., Pillat, M., Weisgerber, C., Shibaev, V. N. and Torgov, V. I. (1985). European Journal of Biochemistry 151, 393. J a m , K . , Dengler, T. and J a m , B. (1992). Zentrablanfur Bakteriologie 276, 196. Jansson, P.-E., Lonngren, J., Widmalm, G., Leontein, K., Slettengren, K., Svenson, S. B., Wrangsell, G., Dell, A. and Tiller, P. R. (1985). Carbohydrate Research 145, 59. Jayaratne, P., Keenleyside, W. J . , MacLachlan, P. R., Dodgson, C. and Whitfield, C. (1993). Journal of Bacteriology 173. in press. Jiang, X.-M., Neal, B., Santiago, F., Lee, S. J., Romana, L. K. and Reeves, P. R. (1991). Molecular Microbiology 5 , 695. Johnson, E . M. and Baron, L. S. (1969). Journal of Bacteriology 99, 355. Johnson. J. G. and Wilson, D. B. (1977). Journal of Bacteriology 129, 225. Jones, R. T., Koeltzow, E. E . and Stocker, B. A. D. (1972). Journal of Bacteriology 111,763. Kanegasaki, S. and Jann, K. (1979). European Journal of Biochemistry 95, 287. Kanegasaki, S. and Wright, A. (1970). Proceedings of the National Academy of Sciences of the United States of America 67, 951. Kao, C. C . and Sequeira, L. (1991). Journal of Bacteriology 173, 7841. Kastowsky, M., Gutberlet, T . and Bradaczek, H . (1992). Journal of Bacteriology 174, 4798. Kato, J . and Chakrabarty, A . M. (1991). Proceedings of the National Academy of Sciences of the United States of America 88, 1760. Kato, J . , Chu, L., Kitano, K., DeVault, J. D., Kimbara, K., Chakrabarty, A. M. and Misra, T. K. (1989). Gene 84, 31. Kato, J., Misra, T. K. and Chakrabarty, A. M.(1990). Proceedings of the National Academy of Sciences of the United States of America 87, 2887. Kawahara, K., Hamaoka, T., Suzuki, S., Nakamura, M., Murayama, S. Y.. Arai, T., Terakado, N. and Danbara, H. (1989). Microbial Pathogenesis 7, 195. Keenleyside, W. J . , Jayaratne, P., MacLachlan, R. P. and Whitfield, C. (1992). Journal of Bacteriology 174, 8. Kellenberger, E. (1990). Molecular Microbiology 4, 697. Keller, J. M. (1966). Ph.D. Thesis: Massachussetts Institute of Technology. Kenne, L. and Lindberg, B. (1983). In “Bacterial Polysaccharides” (G. 0. Aspinall, ed.), vol. 2, pp. 287-363. Academic Press, New York. Kennedy, E. P. (1987). In “Escherichia coli and Salmonella typhimurium, Cellular and Molecular Biology” (F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik,

238

C . WHI'I'FlKl.ll A N D M . A. V A I . V A N O

M. Schaechter and H . E. Umbarger, eds), pp. 672-679). American Society for Microbiology. Washington, DC. Kent, J. L. and Osborn, M. J . (1968a). Biochemistry 7, 4396. Kent, J. L. and Osborn. M. J . (1968b). Biochemistry 7, W . Kent, J. L. and Osborn, M. J . (1968~).Biochemistry 7. 4419. Kessler, A. C., Brown, P. K., Romana, L. K. and Reeves, P. R. (IY91). Journalof t i e n e r d Microbiology 137. 2689. Kido, N., Ohta. M., lida, K.-L., Hasegdwa, T., Ito, H . , Arakawa, Y . , Komatsu, I . and Kato, N. (1989). Journal of Bacteriology 171, 3629. Klena, J. D . , Pradel, E. and Schnaitman, C . A. (1992). Journal of Bacteriology 174, 4746. Koizumi, K., Okada, Y . , Utamara, T., Hisamatsu, M. and Amcmura, A . (IYX4). Jourriul o f Chromatography 299, 21.5. Kol, 0.. Wieruszeski, J. M . , Streckcr, G . , Montreuil, J . , Fournet, B.. Zalisz, R . and Smets. P. (19YI). Carbohydrate Research 217, 117. Kol, 0..Wieruszeski. J. M., Strecker, G . , Fournet, B . , Zalisz, R. and Smets, P. (1992). Carbohydrate Research 236, 339. Kolyva, S., Waxin, H. and Popoff, M. Y. (192). Journal of General Microbiology 138. 2Y7. Konyecsni, W. M. and Deretic, V. (1990). Journal of Bacteriology 172, 251 1 . Kopecko, D. J., Washington. 0. and Formal, S. B. (1980). Infection and Immunity 29, 207. KAplin, R.. Arnold, W . , Hotte, B., Simon, R.,Wang, G . and Piihler. A. (1992). Journal of Bacteriology 174, 191. Kopmann. H . J . and Jann, K. (1975). European Journal of Biochemistry 60. 5x7. Kornfeld, R. H. and Ginsburg, V. (1966). Biochimica et Riophysica Acta 117, 79. Koshland, D., Sauer, R. T. and Botstein, D . (1982). Cell 30,YO3. Kroll, J . S. (1992). Journal of Infectious Diseases 165, S93. Kroll, J . S. and Moxon, E. R. (1988). Journal of Bacteriology 170, X59. Kroll, J . S. and Moxon, E. R. (IYYO). Journal of Bacteriology 172, 1374. Kroll, J . S., Hopkins, I. and Moxon, E. R. (198X). Cell 53, 347. Kroll, J . S., Zamze, S., Loynds, B. and Moxon, E. R. (1989). Journalof Bacteriology 6.3343. Kroll, J . S., Loynds, B . , Brophy. L. N. and Moxon, E. R. (19%)). Molecular Microbio1og.v 4, 1853.

Kroll. J . S., Loynds, B. M. and Moxon, E. R. (lY9l). Molecular Microbiology 5, 154Y. Kronke, K. D., Boulnois, G . , Roberts, I . , Bitter-Suermann. D . , Golecki. J . R., Jann, R . and Jann, K. (19YOa). Journal of Bacteriology 172, 108.5. Kronke, K. D . , Golecki, J. R. and Jann, K. (19YOb). Journal of Racterio1og.v 172. 3469. Kuhn, H.-M., Meier-Dieter, U. and Mayer, H. (1988). FEMS Microbiology Reviews 54. 105. Kulpd, C. F. and Leive, L. (1Y76). Journal of Bacteriology 126, 467. Kundig, F. D . , Aminoff, D . and Roseman, S. (1971). Journal of Biological Chemistry 246, 2543.

Kuo, J. S.-C., Doclling, V. W., Graveline, J . F. and McCoy, D. W. (1985). Journal o f Bacteriology 163, 769. Kuzio, J . and Kropinski. A. M. (1983). Journal o f Bacteriology 155, 203. Laakso, D. H . , Homonylo, M . K., Wilmot. S. J . and Whitfield. C. (19HX). Canadian Jorrriiul of Microbiology 34, 987. Labischinski, H., Barnickel, G., Bradaczek, H . , Naumann, D . , Rictschel. E. -1. and Giesbrecht. P. (198.5). Journal of Bacteriology 162, 9. Lam, M. Y. C., McGroarty, E. J., Kropinski, A . M., MacDonald, L. A , , Pcderson, S. S., Hoiby. N. and Lam, J . S. (1989). Journal of Clinical Microbiology 27, 962. Lam, J. S,. Lam, M. Y.C., MacDonald, 1.. A. and Hancock, R. E. W . (1987). Journal o f Racteriology 169, 353 1 . Lam, J. S., Graham. L. l.., Lightfoot. J . , Dasgupta. T. and Hcvcridge. T. J . (1992). Jorrrriul of Bacteriology 174, 71.59. Larsen, B. and Haug, A. (1971). Carbohydrate Research 20. 225. Latchford, J. W . , Borthakur, D . and Johnson. A . W . B. (1991). Molecular Microbiology 5. 2107.

(’FI.I.-SURFACE POI.YSACCIIARIDES IN GRAM-NEGATIVE HACTERIA

239

Leduc, M. and Frehel, C. (IYW). FEMS Microbiology Ixtters 67, 39. Lee, S. J . , Romana, L . K. and Reeves, P. R. (1992). Journalof General Microbiology 138,305. Leigh, J . A . and Coplin, D. L. (1992). Annual Reviews in Microbiology 46, 307. Leroux, B . , Yanofsky. M. F., Winans, S. C., Ward. J. F., Ziegler, S. F. and Nester, E. W. (1987). EMBO Journal 6, 849. Levengood, S . K. and Wehster, R . E. (1989). Journal of Bacteriology 171, 6600. Levery, S. B., Zhan, H., IAX, C. C., I.eigh, J. A . and Hakamori, S. (1991). Carbohydrate Research 210, 339. Lew. H. C . , MLkelL, P. H.. Kuhn, H.-M., Mayer, H. and Nikaido, H. (1986). Journal of Bacteriology 168, 7 15. Lifely, M. R.. Lindon, J. C., Williams, J . M. and Moreno, C. (1985). Carbohdyrate Research 143, 191.

Lin, T.-Y. and Hassid, W. 2. (1966a). Journal of Biological Chemistry 241, 3283. Lin, T.-Y. and Hassid, W . Z. (1966h). Journal of Biological Chemistry 241, 5284. Lindhcrg, A. A. and Hellerqvist, C.-G. (1980) Journal of General Microbiology 116, 2 5 . Lindherg. A . A. and Lc Minor, L. (1984). Methods in Microbiology 15, I . Linker. A . and Evans, L. R . (1984). Journal of Bacteriology 159, 958. Liu, D. E . , Verman, N . K., Romana, 1. K. and Reeves, P. R. (lY9l). Journalof Bacteriology 173. 4814.

Long, S., Reed, J. W . , Himawan, J . and Walker, G . C. (1988). Journalof Bacteriology 170, 4239.

Lopez, J . and Webster, R . E. (1985). Journal of Bucteriology 163, 270. Lugowski. C . , Romanowska, E., Kennc, L. and Lindherg, B. (1983). Carbohydrate Research 118, 173. Lugtenberg. B. and Van Alphen, L. (1983). Riochemica el Biophysica Acta 737, 51. L’vov, V . I.., Shashkov, A . S., Dimitriev, B. A., Kochtov, N. K., Jann B. and Jann, K. (1984). Carbohydrate Research 126, 249. McCallum, K. L. and Whitfield, C. (1991). Infection and Immunity 59, 494. McCallum. K. I.., Schoenhals. G . . Laakso, D., Clarke, B . and Whitfield, C. (1989). Infection and Immunity 57, 3816. McCammmon, S . L. and Coplin. D. 1.. (1982). Phytopathology 72, 1001. McCloskey, M. A. and Troy. F. A. (IYXO). Biochemistry 19. 2061. McConnell, M. and Wright, A . (1979). Journal of Bacteriology 137, 746. McConnell, M. and Schoelz, J . E. (1983). Journal of General Microbiology 129, 3177. McConncll. M. R . , Foster, B. D., Davis, D. P., Kat, B., Blair, J. G., Long, R. A . and Steed. M. M. (1086). Microbios 48, 135. McCrath, B. C. and Oshorn, M. J. (IYYIa). Journal of Bacteriology 173, 649. McGrath. B. C . and Osborn, M. J . (IY91b). JournalofBacteriobgy 173, 3134. MeGroarty, E. J. and Rivera, M. (1990). Infection and lmmunity 57, 1030. McGuire, E. J . and Binkley, S. B. (1964). Biochemistry 3, 247. MacLachlan, P. R . . Kadam, S. K. and Sanderson, K. E. (1991). Journalof Racteriology 173. 7151.

MacLean, L. L., Whitfield, C. and Perry, M. B. (1993). Carbohydrate Research 239, 325. Macpherson. D. F . . Morona, R., Beger, D. W., Cheah, K.-C. and Manning, P. A. (1991). Molecular Microbiology 5 , 149I . Mlkela. P. H. (1966). Journal of Bacteriology 91, 1 115. Makela, P. H. (1973). Journal of Bacteriology 116, 847. Makell, P. H. and Stocker, B. A. 11. (1984). In “Handbook of endotoxin. Chemistry of Endotoxin” (E. T. Rietschel. ed.). vol. I , pp. 5Y-137. Elsevier, Amsterdam. Makela, P. H . ,Jankola, M. and Luderitz, 0 . (1970). Journalof General Microbiology60,Yl. Manning, P. A , , Heuzenroeder, M. W., Yeadon, J . , Leaveslay, D. !., Reeves, P. R. and Rowlcy, D. (1986). Infection and Immunity 53, 272. Marino, P. A., Phan. K. A . and Osborn. M. J. (1985). Journalof Biological Chemistry 260, 14965.

Marino. P. A , . McGrath. B. C. and Oshorn, M. J. (1991). Journalof Bacteriology 173,3128.

240

C . Wt1ITFlF.I.D AND

M. A. VAI.VANO

Markovitz. A. (1964). Proceedings -~of the National Academy. of. Sciences of the United States of America, 5 1 , 239. Markovitz. A. (1977). In “Surface Carbohvdrates of the Prokarvotic Cell” (1. W. Sutherland, ed.), pp: 4 1 h 2 . ’ Academic Press, Loidon. Marolda, C. L. and Valvano, M. A. (1993). Journal of Bacteriology 175, 148. Marolda, C. L., Welsh, J., Dafoe, L. and Valvano, M. A. (1990). Journal of Bacteriology 172, 35%.

Marumo, K . , Linqvist, L., Verma, N., Weintraub, A,, Reeves, P. R. and Lindberg, A. A. (1992). European Journal of Biochemistry, 204, 539. Marzocca, M. P., Harding, N. E., Pctroni, E. A,, Cleary, J . M. and lelpi, L. (1991) Journal of Bacteriology 173, 7519. Maskell, D. J., Szabo, M. J . . Butler, P. D., Williams, A. E. and Moxon, E. R. (1991). Molecular Microbiology 5, 1013. Masson, L. and Holbein, B. E. (1985). Journal of Bacteriology 161, 861. May, T. B., Shinabarger, D., Maharaj, R., Kato, J . , Chu, L., DeVault, J . D., Roychoudhary, S . , Zielinksi, N. A,, Berry, A,, Rothmel, R. K., Misra, T. K. and Chakrabarty, A. M. (1991). Clinical Microbiology Reviews 4, 191. Meier, U. and Mayer, H. (1985). Journal of Bacteriology 163, 756. Meier-Dieter, U . , Starman, R., Barr, K., Mayer, H. and Rick, P. D. (1990). Journal of Biological Chemistry 265, 13490. Meier-Dieter, U . , Barr, K., Starman, R.. Hatch, L. and Rick, P. D. (1992). Journal of Biological Chemistry 267, 746. Melo, A. and Glaser, L. (1966). Biochemistry and Biophysics Research Communications 22, 524.

Miller, J . F., Mekalanos. J . J . and Falkow, S. (1989). Science 243, 916. Miller, K. J., Kennedy, E. P. and Reinhold, V. N. (1986). Science 231, 48. Miller, K. J., Reinhold, V. N., Weissborn, A. C. and Kennedy, E. P. (1987). Biochimica et Eiophysica Acta 901, 112. Mizuno, T . and Mizushima, S. (1990). Molecular Microbiology 4, 1077. Mohr, C. D., Martin, D. W., Konyecsni, W. M., Govan, J . R. W., Lory, S. and Deretic. V. (19yo). Journal of Bacteriology 172, 6576. Mulford, C. A. and Osborn. M. J. (1983). Proceedings of the National Academy of Sciences of the United States of America 80, 1159. Miihlradt, P. F., Menzel, J., Golecki, J. R. and Speth, V. (1973). European Journal of Biochemistry 35, 471. Naide, Y.,Nikaido. H., Makela, P. H., Wilkinson, R. G. and Stocker, B. A. D. (1965). Proceedings of the National Academy of Sciences of the United States of America 53, 147.

Nakae, T. and Nikaido, H. (1971a). Journal of Biological Chemistry 246, 4386. Nakae, T. and Nikaido, H. (1971b). Journal of Biological Chemistry 246, 4397. Nassif, X . , Fournier, J . - M . , Arondel, J . and Sansonetti, P. J . (1989a). Infection and lmmunity 57, 546. Nassif, X.,HonorC, N., Vasselon, T., Cole, S. T. and Sansonetti, P. J . (1989b). Molecular Microbiology 3, 1349. Neal, B. L., Tsiolis, G. C., Heuzenroeder, M. W., Manning, P. A. and Reeves, P. R. (1991). FEMS Microbiology Letters 82, 345. Nelson, D., Bathgate, A. J. and Poxton, 1. R. (1991). Journal of General Microbiology 137, 2741.

Nikaido, H . (1965). Biochemistry 4, 1550. Nikaido, H. and Nikaido, K. (1965). Biochemical and Biophysical Research Communications 19, 322.

Nikaido, K. and Nikaido. H. (1971). Journal of Biological Chemistry 246, 3912. Nikaido, H. and Vaara, M. (1985). Microbiological Reviews 49, 1. Ni kaido, H., Ni kaido, K., Nakae, T. and Makela, P. H. ( 197I ) .Journal of Biological Chemistry 246.3902.

CF.l.1.-SURFACE POLYSACCHARIDES IN GRAM-NEGATIVE B A C E R I A

241

Nyman, K., Plosila, M., Howden, L. and Makela, P. H. (1979). Zenrrablarrfur Bakreriologie 243, 355. Ohman, D. E., Goldberg, J. B. and Flynn, J . L. (1990). In “PseudomonusBiotransformations, Pathogeneisis and Evolving Technology” (S. Silver, A. M. Chakrabarty, B., lglewski and S. Kaplan, eds), pp. 2X-35. American Society for Microbiology, Washington, DC. Ohta, M., h a . K., Kusuzaki, K., Kido, N., Arakawa, Y. and Kato, N. (1991). Molecular Microbiology 5, 1853. Brskov, F. and Brskov. I. (1962). Acfa Parhologica et Microbiologica Scandinavica 55, 99. Brskov, I. and Nyman, K. (1974). Journal of Bacreriology 120, 43. Brskov, I., Sharma, S. and Brskov, F. (1976). Acra Parhologica er MicrobiologicaScandinavica Section B 84, 125. Brskov, F., Sharma, F. and Brskov, I. (1984). Journal of General Microbiology 130, 2681. Osborn, M. J. (1979). In “Bacterial Outer Membranes: Biogenesis and Functions” (M. Inouye, ed.), pp. 15-34. Wiley, New York. Osborn, M. J. and Weiner, 1. M. (1968). Journal of Biological Chemistry 243, 2631. Osborn, M. J. and Yuan Tze-Yuen. R. (1%8). Journal of Biological Chemisfry 243, 5145. Osborn, M. J., Cynkin, M. A,, Gilbert, J. M., Miiller, L. and Singh, M. (1972a). Merhodr in Enzymology 28, 583. Osborn, M. J . , Gander, J. E., Parisi, E. and Carson, J. (1972b). Journal of Biological Chemistry 247, 3962. Osborn, M. J., Gander, J. E. and Parisi, E. (1972~).Journalof BiologicalChemisfry247,3973. Osborn, M. J., Rick, P. D. and Rasmussen, N. S. (1980). Journal of Biological Chemistry 255,4246. Osbourn, A. E., Clarke, B. R., Stevens, B. J. H. and Daniels, M. J. (1990). Molecular and General Generics 222, 145. Ou, J. T., Baron, L. S., Rubin, F. A . and Kopecko, D. J . (1988). Proceedings of rhe Narional Academy of Sciences of rhe Unired Sfares of America 85,4402. Oxley, D. and Wilkinson, S. G. (1989a). Carbohydrate Research 193,241. Oxley, D. and Wilkinson, S. G. (1989b). Carbohydrate Research 195. 117. Paakanen, J . , Gotschlich, E. C. and Makela, P. H. (1979). Journalof Bacteriology 139,835. Palva, E. T. and Makela, P. H. (1980). European Journal of Biochemisfry 107, 137. Parker, C. T., Kloser, A . W., Schnaitman, C. A,, Stein, M. A., Gottesman, S. and Gibson, B. W. (1992). Journal of Bacteriology 174,2525. Parolis. L. A. S., Parolis, H. and Dutton, G. G. S. (1986). Carbohydrate Research 155,272. Pavelka, M., Wright, L. F. and Silver, R. P. (1991). Journal of Bacreriology 173,4603. Pelkonen, S. (1990). Currenf Microbiology 21,23. Perry, M. B. and Babiuk, L. A. (1984). Canadian Journal of Biochemisfry and Cell Biology 62, 108. Peterson, A. A. and McGroarty, E. J . (1985). Journal of Bacteriology 162,738. Pindar, D. F. and Bucke, C. (1975). Biochemical Journal 152,617. Poetter, K. and Coplin, D. L. (1991). Molecular and General Genetics 229, 155. Popoff, M. Y. and Le Minor, L. (1985). Annales of the Instirure PusreurlMicrobiology 1368, 169. Poole, K. and Braun, V. (1988). Journal of Bacteriology, 170,5146. Pradel, E. and Schnaitman, C. (1991). Journal of Bacreriology 173,6428. Prehm, P., Jann, B. and Jann, K. (1976). European Journal of Biochemistry 67,53. Raetz, C. R. H. (1990). Annual Reviews in Biochemistry 59, 129. Redmond, J. W., Korsch, M. J . and Jackson, G. D. F. (1973). Ausrralian Journal of Experimenral Biology and Medical Science 83, 535. Reed, J. W. and Walker, G. C. (1991). Journal of Bacteriology 173,664. Reed, J. W., Capage, M. J. and Walker, G. C. (1991a). Joumul of Bacreriology 173,3776. Reed, J. W., Glazebrook, J . and Walker, G. C. (1991b). Journalof Bacreriology 173,3789. Reeves, P. (1991). Today’s Life Science 3, 30. Richards, J. C. and h i t c h , R. A. (1989). Carbohydrate Research 186, 275. Reuber. T. L.. Long, S. and Walker, G. C. (1991). Journal of Bacteriology 173,426.

242

C WHITFIEI.I) AND

M. A.

VALVANO

Rick, P. D. (1987). In “Escherichia coli and Salmonella fyphimurium, Cellular and Molecular Biology” (F. C. Neidhardt, J. L. Ingraham, K. B. L o w , B. Magasanik, M. Schaechter, and H . E. Umbarger, eds), pp. 648462. American Society for Microbiology, Washington, DC. Rick, P. D., Mayer, H.. Neumeyer, B. A,, Wolski, S. and Bitter-Suermann. D . (198s). Journal of Bacteriology 162, 494. Rick, P. D., Wolski, S., B a n , K.,Ward, S. and Ramsay-Sharer, L. (1988). Journal of Bacteriology 170, 4008. Riley, L. W.. Junio, L. M., Libaek, L. B. and Schoolnik, G. K. (1987). Infection and Immuniry 55, 2052.

Rivera, M. and McGroarty, E. J. (1989). Journal of Bacteriology 171. 2244. Robbins, P. W. and Uchida, 1.. (1965). Journal of Biological Chemistry 240, 375. Robbins, P. W., Wright, A. and Bellows, J. L. (1964). Proceedings ofthe Narional Academy of Sciences of the United States of America 52, 1302. Robbins, P. W., Keller, J . M.,Wright, A. and Bernstein, R. L. (1965). Journal of Biological Chemistry 240, 384. Robbins, P. W., Bray, D., Dankert, M.and Wright, A. (1967). Science 158, 1536. Roberts, I., Mountford, R., High, N., Bitter-Suermann, D., Jann, K., Timmis, K . and Boulnois, G . (1986). Journal of Bacteriology 168, 1228. Roberts, I.. Mountford, R., Hodge, R., Jann, K. B. and Boulnois, G. (1988). Journal of Bacteriology 170, 1305. Roberts. I. S . and Coleman, M. J. (1991). Journal of General Microbiology 137, 1453. Roberts, M.,Roberts, I . , Korhonen, T. K., Jann, K.. Bitter-Suermann. D., Boulnois, G . J. and Williams, P. H. (1988). Journal of Clinical Microbiology 26,385. Rodgriguez-Aparicio, L. B., Reglero, A.. Ortiz, A. I. and Luengo, J . M. (1988). Biochemical Journal 251, 589. Rodriguez, M.-L., Jann, B. and J a m , K. (1988). Carbohydrate Research 173, 243. Rohr, T . E. and Troy, F. A. (1980) Journal of Biological Chemistry 255, 2332. Roncero, C . and Casadaban, M. J. (1992). Journal of Bacteriology 174, 3250. Ronson, C. W., Nixon, T. B. and Ausubel, F. M. (1987). Cell 49, 579. Ross, P. Y.,Aloni, H., Weinhouse, D., Michaeli. D., Ohana, P., Mayer, R. and Benziman, M. (1985). FEBS Letters 186, 191. Ross, P. Y.,Aloni, H., Weinhouse, D., Michaeli, D., Ohana, P., Mayer, R. and Benziman, M. (1986). Carbohydrate Research 149, 101. Ross, P., Weinhouse. H., Aloni, Y.,Michaeli, D., Weinberger-Ohana, P., Mayer, R., Braun, S., deVroom, E . , van der Marel, G . A , , van Boom, J. H . and Benziman M. (1987). Nature 325,279.

Ross, P. Y.,Mayer, R., Amikam, D., Hugirat, Y.,Weinhouse, H.,Benziman, M., deVroom, E., Fiddler, A. dePaus, P., Sliedregt, L. A. J. M.,van der Marel, G. A. and van Boom, J. H. (1990). Journal of Biological Chemistry 265, 18933. Ross, P., Mayer, R. and Benziman, M. (1991). Microbiological Reviews 55, 35. Roychoudhary, S., May, T. B., Gill, J. F., Singh, S. K., Feingold, D. S. and Chakrabarty, A. M. (1989). Journal of Biological Chemistry 264, 9380. Roychoudhary, S., Sakai, K. and Chakrabarty, A. M.(1992a). Proceedings ofthe National Academy of Sciences of the United States of America 89. 2659. Roychoudhary, S . , Sakai, K., Schlictman, D. andchakrabarty, A. M. (1992b). Gene 112,45. Sack, R . B. and Miller, L. E. (1969). Journal of Bacteriology 99, 688. Saier, M. H. Jr, Werner, P. K. and Muller, M.(1989). Microbiological Reviews 53, 333. Sakasaki, R. andTamura, K. (1971). Japanese Journalof MedicalScience and Biology 24, 93. Sanderson, K. E. and Stocker, B. A. D. (1981). Journal of Bacteriology 1 4 6 , 535. Sansonetti, P. J., Hale, T . L., Dammin, G . J., Kapfer, C., Collins, H. H.. Jr and Formal, S. B. (1983). Infection and Immunity 39, 1392. Sasaki, T. and Uchida. T . (1974). Journal of Bacteriology 117, 13. Sasaki, T., Uchida, T. and Kurahashi, K. (1974). Journal of Biological Chemistry 249. 761.

CELL-SURFACE POLYSACCHARIDFS IN GRAM-NEGATIVE BACTERIA

243

Sawada. S . , Kawamura, T., Masuho, Y. andTomibe, K. (1985). Journalof Infectious Diseases 152, 1290.

Schatz, P. J. and Beckwith, J. (1990). Annual Reviews in Genetics 24, 215. Schiller, N. H. (1992). In “Abstracts of the 92nd General Meeting of the American Society for Microbiology”, p. 134, abstract D-233. Schmidt, G . (1973). Journal of General Microbiology 77, 157. Schmidt, G., Mayer, H. and Makela, P. H . (1976). Journal of Bacteriology 127, 755. Schmidt, G., J a m , B., J a m , K. Brskov, I. and Brskov, F. (1977). Journal of General Microbiology 100, 355. Schmidt, M. A. and Jann, K. (1982). FEMS Microbiology Letters 14, 69. Schnaitman, C . A., Parker, C. T., Klena, J. D . , Pradel, E. L. Pearson, N. B., Sanderson, K. E. and MacLachlan, P. R. (1991). Journal of Bacteriology 173, 7410. Selander, R. K., Smith, N. H., Li, J., Beltran, P., Ferris, K. E., Kopecko, D. J. and Rubin, R. A . (1992). Journal of Bacteriology 174, 3587. Sen, K. and Nikaido, H. (1990). Proceedings of the National Academy of Sciences of the United States of America 87, 743. Shands, J. W., Jr, Graham, J. A. and Nath, K. (1967). Journalof Molecular Biology 25, 15. Sheehy, T. W., Sprintz, H., Augerson, W. S. and Formal, S. B. (1966). Journal of the American Medical Association 197, 321. Shibaev, V. N. (1978). Pure and Applied Chemistry 50, 1421. Shibaev, V. N. (1986). Advances in Carbohydrate Chemisrry 44, 277. Shibaev, V. N., Druzhinina, R. N., Popova, A. N., Rozhinova, S. S. and Kilesso, V. A. (1979). European Journal of Biochemistry 101, 309. Shinabarger, D . , Berry, A., May, T. B., Rothmel, R., Fialho, A. and Chakrabarty, A. M. (1991). Journal of Biological Chemistry 266, 2080. Siewart, G. and Strominger, J. L. (1967.) Proceedings of the National Academy of Sciences of the United States of America 57. 767. Silver, R. P. and Vimr, E. R. (1990). In “The Bacteria. Molecular Basis of Bacterial Pathogenesis” (B. lglewski and V. Miller, eds), vol. XI, pp. 39-60. Academic Press, New

York.

Silver, R. P., Finn, C. W., Vann, W. F., Aaronson, W., Schneerson, R., Kretschmer, P.J. and Garon, C.F. (1981). Nature 289, 696. Silver, R. P., Vann, W. F. and Aaronson, W. (1984). Journal of Bacteriology 157, 568. Silver, R. P., Aaronson, A. and Vann, W. F. (1987). Journal of Bacteriology 169, 5489. Silver, R. P., Annunziato, P., Pavelka, M. S., Pigeon, R. P., Wright, L. F. and Wunder, D . E. (1993). In “Polysialic Acid: From Microbes to Man” (J. Roth, U. Rudishauer and F. A. Troy 11. eds), pp. 59-71. Simmons, D. A. R. and Romanowska, E. (1987). Journal of Medical Microbiology 23, 289. Skjak-Braek, G. and Larsen, B. (1985). Carbohydrate Research 139, 273. Skjak-Braek, G . , Grasdalen, H. and Larsen, B. (1986). Carbohydrate Research 154, 239. Smit, J . and Nikaido, H. (1978). Journal of Bacteriology 135, 687. Smith, A. N., Boulnois, G. J. and Roberts, I . S. (1990). Molecular Microbiology 4, 1863. Smith, A. R. W., Zamze, S. E., Munro, S. M.,Carter, K. T. and Hignett, R. C. (1985). European Journal of Biochemistry 149,73. Smith, I . H . , Symes, K. C., Lawson, C . J. and Morris, E. R. (1981a). lnternational Journal of Biological Macromolecules 3, 129. Smith, H. 0.. Danner, D. B. and Deich, R. A. (1981b). Annual Reviews in Biochemistry 50.41, Snellings, N. J., Johnson, E. M. and Baron, L. S. (1977). Journal of Bacteriology 131, 57. Snellings, N. J., Johnson, E. M., Kopecko, D. J . , Collins, H. H. and Baron, L. S. (1981). Journal of Bacieriology 145, 1010. Stacey, G . , So, J.-S., Roth, L. E., Bhagya Lakshmi, S. K. and Carlson. R. W. (1991). Molecular Plant-Microbe Interactions 4, 332. Stader, J . and Silhavy, T. J. (1988). Journal of Bacteriology 170, 1973. Stanfield, S. W., lelpi, L., O’Brochta, D . , Helinski, D. R . and Ditta, G. S. (1988). Journal of Bacteriology 170, 3523.

244

C. WIIITIIELD A N D M. A. V A L V A N 0

Stark, J. W., Levy, G. N., Rohr, T. E. and Anderson, J. S. (1977). Journal of Biological Chemistry 252, 3466. Steenbergen, S. M.and Vimr, E. R. (1990). Molecular Microbiology 4, 603. Steenbergen, S. M., Wrona, T. J. and Vimr, E. R. (1992). Journalof Bacteriology 174, 1099. Stevenson, G., Lee, S. J., Romana, L. K.and Reeves, P. R. (1991). Molecular and General Genetics 277, 173. St Geme, J . W., 111 and Falkow, S. (1992). Journal of Infectious Diseases 165, S117. Stibitz, S., Aaronson, W., Monack, D. and Falkow, S. (1989). Nature 338, 2 M . Stock, J. B., Ninfa, A. J. and Stock, A . M. (1989). Microbiological Reviews 53, 450. Stout, V. and Gottesman, S . (1990). Journal of Bacteriology 172, 659. Stout, V., Torres-Cabassa, A., Maurizi. M. R., Gutnick, D. and Gottesman, S. (1991). Journal of Bacteriology 173, 1738. Stroeher, U. W., Karageorgos, L. E.. Morona, R. and Manning, P. A. (1992). Proceedings of the National Academy of Sciences of the United States of America 89, 2566. Sturm, S. B. and Timmis, K. N. (1986). Mircrobial Pathogenesis 1 289. Sturm, S. B., Jann, B., Jann, K.,Fortnagel, P. and Timmis, K. N. (1986a). Microbial Pathogenesis 1, 299. Sturm, S. B., Jann, B., Jann, K.,Fortnagel, P. and Timmis, K. N. (1986b). Microbial Pathogenesis 1, 307. Sutcliffe, J., Blumenthal. R . , Walter, A,, and Foulds, J. (1983). Journalof Bacteriology 156, 867.

Sutherland, 1. W. (ed.) (1977a). “Surface Carbohydrates of the Prokaryotic Cell”. Academic Press, London. Sutherland, I. W. (1977b) In “Surface Carbohydrates of the Prokaryotic Cell” ( I . W. Sutherland, ed.), pp. 27-96. Academic Press, London. Sutherland, 1. W. (1981). Carbohydrate Polymers 1, 107. Sutherland, I. W. (1982). Advances in Microbial Physiology 23, 79. Sutherland, I. W. (1985). Annual Reviews in Microbiology 39, 243. Sutherland, 1. W. (1988). Infernational Reviews of Cytology 113, 187. Sutherland, 1. W. and Norval, M. (1970). Biochemical Journal 120, 567. Swissa. M.,Aloni, Y., Weinhouse, H. and Benziman, M. (1980). Journal of Bacteriology 153, 173.

Takeshita, M. and Makela, P. H. (1971). Journal of Biological Chemistry 246, 3920. Tang, J.-L., Gough, C. L. and Daniels, M. J. (1990). MolecularandGeneralGenerics222. 157. Tang, J.-L., Liu, Y.-N., Barber, C. E., Dow, J. M., Wooton, J. C. and Daniels, M. J. (1991). Molecular and General Genetics 226, 409. Thome, L., Tansey, L. and Pollock, T. J. (1987). Journal of Bacteriology 169, 3593. Tolmasky, M. E., Staneloni, R. J., Ugalde, R. A. and Leloir, L. F. (1980). Archives of Biochemistry and Biophysics 203, 358. Tolmasky, M. E., Staneloni, R. J. and Leloir, L. F. (1982). Journal of Biological Chemistry, 257. 6751.

Torres-Cabassa. A. and Gottesman, S. (1987). Journal of Bacteriology 169, 981. Torres-Cabassa, A., Gottesman, S . , Frederick, R. D., Dolph, P. J . and Coplin, D. L. (1987). Journal of Bacteriology 169, 4525. Trisler, P. and Gottesman, S. (1984). Journal of Bacteriology 160, 184. Troy, F. A., I1 (1979). Annual Reviews in Microbiology 33, 519. Troy, F. A., I1 (1992). Glycobiology 2. 5. Troy, F. A. and McCloskey, M. A. (1979). Journal of Biological Chemistry 254, 7377. Troy, F. A., Frerman. F. E. and Heath, E. C. (1971). Journalof BiologicalChemisrry246.118. Troy, F. A., Vijay. I. K. and Tesche, N. (1975). Journal of Biological Chemistry 250, 156. Troy, F. A.. Janas, T., Janas, T. and Merker, R. 1. (1990). Glycoconjugate Journal 8 , 153. Tsui, F.-P., Boykins, R. A. and Egan, W. (1982). Carbohydrate Research 102, 263. Ugalde, R. A., Coira, J. A. and Brill, W. J . (1986). Journal of Bacteriology 168, 270. Valla, S., Coucheron. D. H., Fjaervik, E., Kjosbakken, J., Weinhouse, H., Ross, P., Amikam, D. and Benziman M. (1989). Molecular and General Genetics 217, 26.

CELLSURFACE POLYSACCHARIDES IN GRAM-NEGATIVE BACTERIA

245

Valvano, M. A. (1992). Canadian Journal of Microbiology 38, 71 1. Valvano, M.A. and Crosa, J. H. (1989). Infection and Immunity 57, 937. Valvano, M. A. and Marolda, C. L. (1991). Infection and Immunity 59, 3917. Van Alphen, L. F., Van Kampen-de Troye, F. and Zanen, H. C. (1983). FEMS Microbiology t5etter.s 16, 261. Vanderslice, R. W., Doherty, D. H., Capage, M. A., Betlach, M. R., Hassler, R . A , , Henderson, N. M.. Ryan-Graniero, J . and Tecklenburg, M. (1989). In “Recent Developments in Industrial Polysaccharides: Biomedical and Biotechnological Advances” (I. C. M .Dea and S. S. Stivola, eds), pp. 145-156. Gordon and Breach, New York. Vann, W. F. and Jann. K. (1979). Infection and Immunity 25, 85. Vann, W. F. and Troy, F. A. (1986). In “Abstracts of the XlIlth International Carbohydrate Symposium, Cornell University, Ithaca, New York”, p. 360, abstract 0 9 . Vann, W. F., Liu, T.-Y. and Robbins, J. B. (1978). Journal of Bacteriology 133, 1300. Vann, W. F., Schmidt, M.A., Jann, B. and Jann, K. (1981). European Journalof Biochemistry 116, 359.

Vann, W. F., Soderstrom, T., Egan, W., Tsui. F.-P., Schneerson, R., Brskov, I. and Brskov, F. (1983). Infection and Immunity 39, 623. Vasselon, T., Sansonetti, P. J . and Nassif, X. (1991). Research in Microbiology 142, 47. Verma, N. and Reeves, P. (1989). Journal of Bacteriology 171. 5694. Verma, N. K.. Quigley, N. B. and Reeves, P. R. (1988). Journal of Bacteriology 170, 103. Verma, N. K., Brandt, J . B., Verma, D . J. and Lindberg, A. A. (1991). Molecular Microbiology 5 , 71. Vijay, 1. K. and Troy, F. A . (1975). Journal of Biological Chemistry 250, 164. Vimr, E. R. (1991). Journal of Bacteriology 173, 1335. Vimr, E. R. and Steenbergen, S. M. (1993). In “Polysialic Acid: From Microbes to Man” (J. Roth, U. Rudishauer and F. A. Troy 11, eds), pp. 7S91. Birkauser-Verlag. Basel. V i m , E . R. and Troy, F. A. (1985). Journal of Bacteriology 164, 854. V i m , E . R., Aaronson, W. and Silver, R . P. (1989). Journal of Bacteriology 171, 1106. Vimr, E. R., Bergstrom, R.,Steenbergen, S. M.,Boulnois, G. and Roberts, 1. S. (1992). Journal o f Bacteriology 174, 5127. Wacharotayankun, R., Arakawa, Y., Ohta, M.,Hasegawa, T., Mori,M.,Horii, T. and Kato, N. (1992). Journal of Bacteriology 174, 1063. Walderich, B. and Holtje, J.-V. (198Y). Journal of Bacteriology 171, 3331. Walderich, B., Ursinus-Wossner, A.. van Duin, J . and Holtje, J.-V. (1988). Journal of Bacteriology 170, 5027. Wang, L., Romana, L. and Reeves, P. R. (1992). Genetics 130, 429. Wang, S.-K., Sa’Correia, I., Darzins, A. and Chakrabarty, A. M . (1987). Journal of General Microbiology 133, 2303. Webster. R. E. (1991). Molecular Microbiology 5 , 1005. Weiner, I . M.,Higuchi, T., Rothfield, J., Saltmarsh-Andrew, M.,Osborn, M. J . and Horecker. B. L. (1965). Proceeding of the National Academy of Sciences of the United Stares of America 54, 228. Weisgerber, C . and Jann, K. (1982). European Journal of Biochemistry 127, 165. Weisgerher, C. and Troy, F. A. (1990). Journal of Biological Chemistry 265, 1578. Weisgerber, C., Jann, B. and J a m , K. (1984). European Journal of Biochemistry 140, 553. Weisgerber, C., Hansen, A. and Frosch, M. (1991). Glycobiology 1. 357. Weiss, J., Hutzler, M. and Kao, L. (1986). Infection and Immunity 51, 594. Whitfield, C., Sutherland, I. W. and Cripps, R . E. (1981). Journal of General Microbiology 124 385.

Whitfield, C., Hancock, R. E. W. and Costerton, J. W. (1983). Journal of Bacteriology 156, 873.

Whitfield, C . , Adams, D. A. and Troy, F. A. (1984a). Journal of Biological Chemistry 259, 12769.

Whitfield, C., Vimr, E. R, Costerton, J . W. and Troy, F. A. (1984h). Journalof Bacteriology 159, 321.

246

C , WIIITFIEI.I) A N D M . A V A I . V A N O

Whitfield, C . , Vimr, E. R., Costerton. J . W . and Troy. F. A. (19x5). Journalof Bacteriology 161, 743. Whitfield, C . , Schoenhals. G . and Graham, L. L. (1989). Journal of General Microbiology 135. 2589. Whitfield, C . , Richards. J . C.. Perry, M. B., Clarke, B. R. and MacLean, L. L. (1091). Journal of Bacteriology 173, 1420. Whitfield, C . , Perry. M. B.. MacLean, L. L. and Yu, S. H. (1992). Journal of Uacterio1og.y 174, 4913. Wilkinson. R. G . and Stocker, B. A. D. (1968). Nature 217, 955. Williams, M. N. V., Hollingsworth, R . I., Brzoska, P. M . and Signer. E. R . ( IWOa). Journal of Bacteriology 172, 6596. Williams, M. N. V., Hollingsworth, R. I., Klein, S . and Signer, E. R . (IWOb). Journal of Uacteriology 172, 2622. Wong, H . C., Fear, A. L., Cdlhoon, R. D., Eichinger. G . H., Mayer, R., Amikam, D.. Benziman, M., Gelfand, D. H., Meade, J . H., Emerick, A . W., Bruner, R., Ben-Bassat, A. and Tal, R. (1990). Proceedings of the National Academy of Sciences of the United States of America 87, 8130. Woods, D. E., Sokol, P. A , . Bryan, L. E., Storey, D . G., Mattingley, S . J.. Vogel, H. J. and Ceri, H. (1991). Journal of lnfectious Diseases 163, 143. Wozniak, D. J. (1992). In "Abstracts of the 92nd General Meeting o f the American Society for Microbiology", p. 134, abstract D-232. Wozniak, D . J . and Ohman, D. E. (1991). Journalof Bacteriology 173, 1406. Wright, A. (1971). Journal of Bacteriology 105, 1406. Wright, A. and Kanegasaki, S . (1971). Physiological Reviews 51, 748. Wright, A., Dankert, M. and Robbins, P. W. (1965). Proceedings of the Nationul Acurlemy of Sciences of the United Sfates of America 54, 235. Wright, A., Dankert, M., Fennessey, P. and Robbins. P. W. (1967). Proceedings of the National Academy of Sciences of the United States of America 57, 1798. Wyk, P. and Reeves, P. (1989). Journal of Bacteriology 171. 5687. Yao, Z . , Lui, H. and Valvano, M. A . (1992). Journal of Bacteriology 174. 7500. Yokota, S., Kaya, S . , Sawada, S . , Kawamura, T., Araki, Y. and Ito, E. (1987). European Journal o f Biochemistry 167. 203. Yoshida. Y . , Okamura, N., Kato, J . and Watanabe, H. (1991). Journal of Generul Microbiology 137, 867. Yuasa, R., Levinthal, M. and Nikaido, H . (1969). Journal of Racteriology 100, 433. Yurewicz, E. C . , Ghalambor, M . A. and Heath, E. C. (1971). Journalof RiobKiculChemisfry 246,5596.

Zapata, G . , Vann, W. F., Aaronson, W., Lewis, M. S . and Moos, M. (1989). Journal C J Biological Chemistry 264, 14769. Zapata, G . , Crowley, J . M. and Vann, W . F. (1992). Journal of Bacteriology 174, 31s. Zeleznick. L. D . , Rosen, S. M., Saltmarsh-Andrew, M . , Osborn. M. J . and Horeckcr, B. 1.. (1965). Proceedings of the National Academy of Sciences o f the United States of Americrr 53, 207. Zevenhuizen. L. P. T. M., Van Veldhuizen, A . and Fokkcns, R. (1990). Journal o f ' Microbiology and Serology 57, 173. Zhan, H . and Leigh, J . A. (1990). Journal of Bacteriology 172, 5254. Zhan, H . , Levcry, S. B., Lee, C. C. and Leigh, J . A . (1989). Proceedings of the Nutionul Academy of Sciences of the United States of America 86, 3055. Zhan. H., Lee, C. C. and Leigh, J. A . (1991). Journalof Bacteriology 173, 7391. Zorreguieta, A. and Ugalde. R . (1986). Journal o f Bacteriology 167, 947. Zwahlen, A . , Kroll, J. S . , Rubin. L. G . and Moxon, E. R.(1989). Microbial Pathogenicifies 7, 22s.

~

Biochemistry and Physiology of Hopanoids in Bacteria HERMANN SAHMa, MICHEL ROHMERb, STEPHANIE BRINGER-MEYER", GEORG A. SPRENGER" and ROLAND WELLE" a

lnstitut fur Biolechnologie, Forschungszentrum Jiilich, 5170 Jiilich. Germany, and UniversitC de Haute Alsace, Ecole Nationale SupCrieure de Chimie, 68093 Mulhouse. France

. . . . . . . . . . . . . I Introduction I 1 Structuraldiversityof bacterial hopanoids . . . . . . . . . . 111. Detection and analysisof bacterial hopanoids . . . IV. Distribution and physiological role of hopanoids . . . . . . V. Biosynthesis and genetics . . . . . . . . . . . A. Synthesis of isoperileiiyl pyrophosphate . . . . . . . B. Formation of the hopane pentacyclic skeleton from isopentenyl pyrophosphate . . . . . . . . . . . C. Geneticsof isoprenoid and hopanoid biosynthesis . . . . VI. Conclusions . . . . . . . . . . . . . . References . . . . . . . . . . . . . . .

. .

. . . . .

. . .

247 250 253 254 259 259

261 268 270 270

I. Introduction

For many decades, the bacterial plasma membrane has been thought to consist of a mixture of proteins and lipids, and that the lipids are a mixture of phospholipids and glycolipids. The membrane is also known to lack sterols, a class of lipids which is found in virtually all plasma membranes in eukaryotic micro-organisms. This view of bacterial plasma-membrane lipid composition has recently had to be revised following the discovery of a new class of lipids, namely hopanoids. Pentacyclic triterpenoids of the hopane series are probably the most CopyrlghtQ 1W3. by Academic I'rc\s I imited All rightsof reproduction in any form r e s r v e d

NHZ

V

(9) X = -NH,, Y = -CHZOH (10) X = -NHCO-(CH&-cyclohexyl. Y (1 1) X = -OH, Y = -CHZNH2 (12) X = -OH, Y = -COOH

HO =

-CH20H

(13) Z = -H (14) Z = -C = NH

I NHz

(26) 22R (27) 225

FIG. 1. Structural variations in hopanoids from eubacteria. Hopanoids are listed in this review according to their sidechain structures, characterized by a number as found in this figure, and to their pentacyclic skeleton, characterized by a letter (a, saturated skeleton; b, A6; c, A"; d, A6*"; e, 2&CH3; f, 2a-CH3; g, 3bCH3; h, 3B-CH3-A6; i, 3f3-CH3-A"; j , 38CH3-A6."). The symbols A6, A l l and A6," used in the characterization of the pentacyclic skeleton refer to double-bonds at C-6 andor C-11. Aminobacteriohopanetriol with a saturated pentacyclic skeleton will thus be designed by (3a). Sidechains (26)-(30) do not correspond to those of intact hopanoids as found in cells, but to those of the derivatives obtained after periodatesodium borohydride degradation.

250

H SAHM E / A l

abundant natural products of defined structure (Ourisson et al., 1987). Geohopanoids are ubiquitously present in the organic matter of sediments and petroleum deposits (Ourisson ef ul., 1987; Ourisson and Albrecht, 1992). They have been recognized as the molecular fossils of a widespread lipid family from bacteria (biohopanoids). Hopanoids with a C30triterpenic skeleton are also found in several scattered eukaryotic organisms (e.g. higher plants, ferns, mosses, lichens and fungi) but in these the occurrence is rather rare (Ourisson et al., 1987; Ourisson and Albrecht, 1992). Hopanoids are amphipathic components of bacterial cell membranes and play a role in maintaining membrane stability by increasing rigidity in the lipid matrix. Thus, the function of hopanoids in bacteria is analogous to that of membrane sterols in eukaryotes. Biosynthesis of hopanoids, in contrast to that of sterols, does not require oxygen during squalene cyclization and other reactions (Poralla, 1982). This review summarizes the state of our knowledge on these cornpounds from the viewpoint of their structures, distribution and physiological role as membrane stabilizers. During recent studies o n bacterial-hopanoid biosynthesis, interest in the early steps of isoprenoid formation, which have for a long time been neglected, were reinvestigated and produced unexpected findings that could well lead to a new view on bacterialisoprenoid biosynthesis. This will also be addressed herein.

11. Structural Diversity of Bacterial Hopanoids

The presence of diploptene ( l a ) and diplopterol (2a) (Fig. 1) was already recorded from a few bacteria and cyanobacteria in the early 1970s (Gelpi el al., 1970; Bird el a l . , 1971; de Rosa el al., 1971; Bouvier, 1974). These simple hopanoids with a normal Cm triterpenic framework are indeed always present, at least in traces, in nearly all hopanoid-synthesizing bacteria. The principal compounds are, however, always C3s bacteriohopane derivatives, first detected in Acetobacter xylinum by Forster el al. (1973) and characterized by an additional C5polyhydroxylated n-alkyl unit linked to the hopane isopropyl group. Correct insertion of this unit and the resulting correlation between geo- and biohopanoids have been independently reported by Langworthy el ul. (1976) and Rohmer and Ourisson ( 1976a). These bacteriohopane derivates were the first representatives of a much more varied bacterial lipid family. Thorough investigation of the triterpenoid content of bacterial strains (Table 1) revealed numerous structural variations affecting the triterpenic moiety, including additional methyl groups at C-3p (Rohmer and Ourisson, 1 9 7 6 ~Zundel ; and Rohmer,

25 1

HOPANOIDS IN R A f l E R I A

TABLE 1 , Complex hopanoids detected in eubacteria" Strainb

Cyanobacteria Nostoc sp. PCC 6720 Synechocystis sp. PCC 6714 Microcystis aeruginosa PCC 7806

Hopanoid structure' (la). ( 4 4 %(4g), (184, (18gL (19a), (1%) (Ila) ( 3 4 , (4a)

Rhodospirillaceae Rhodomicrobium vanniehi ATCC 17100 (la), (2a), ( 3 4 , (7% ( 8 4 Rhodopseudomonus palusiris DSM 123 (la,, ( 2 4 , ( 3 4 Rhodopseudomonas acidophila DSM 145 ( l a ) , (4a), (9a), (13a), (16a). (17a), (22a). (23a) Rhodo.spiril1um rubrum DSM 467 and 468 (la), (2a), (I2ai Obligate methyiotrophs Methylosinus trichosporium NCIMB I 1 131 Methylocystis parvus NClMB 11 120 Methylomonus methanica NCIMB 11 130 Methylococcus capsulatus NCIMB 11 132 Methvlococcus luteus NCIMR I1914

(la), (2a), (3a), (20a) (2a). (3a), (4a), (20a) (34, (204, ( l a ) ( l a ) , (2a), (204, (20g). (21a), (2lg) (la,. P a ) , (3% (20~1).(20g), (2la), (21gJ

Other Gram-negative heterotrophs Acetohacter aceti spp. xylinum R 2277

( l a ) , (2a), (2g). (4a), (4b), (4d), (4g), (4h), (41). (41). (5a). (lsa), (I%), (15). (IS]), (24bj. ( 2 4 a (24h), (249, (24,) Acetohacter pasteurianus ssp. pasteurianus (la), (2a), (2g), (4a), (4b), (4d), (4g). (4h), NClMB 6249 (4j). (Sa). (ISa). (I%), (15g). (151) Zymomonm mobilis ATCC 29101 ( l a ) , (2a), (4a), (9a). (13a). (25a) Methylobacterium organophilum NClMB 11278 (1% ( 2 4 , (2e). (2f). ( 4 4 , (4th ( 1 0 4 , (l3a). (144 Arotobacter vinelandii DSM 2289 (2aL (13aJ Beijerinckia mobilis DSM 2326 (la). ( 2 4 , (3a) Beijerinckia indica NCIMB 8712 (la), (2a). ( 2 ~ ) (3a). . (4aJ

Gram-positive heterotrophs Alicyclobacillus (Racillus) acidocaldarius ATCC Alicyclobacillu~(Bacillus) acidoterrestru DSM 2498 Streptomyces tendae Tu 901

(la). ( 4 4 , ( 9 4 , ( 1 0 4 (4a). ( 9 4 , ( W (34

All data, with the exception of those on Alicyclobacillus acidocaldarius (Langworthy et al., 1976) are from Rohmer and his coworkers (references given in the text or unpublished results). More complete lists of hopanoid producers can be found in Rohrner et al. (1984) and Ourisson et at. (l987), usually, however, without indications concerning the structures of the intact side-chains ATCC indicates American Type Culture Collection; DSM, Deutsche Sammlung von Mikroorganisrnen; NCIMB, National Collection of Industrial and Marine Bacteria; PCC. Pasteur Culture Collection; R, Hoffmann-La Roche, Basel; Tii, Universitat Tiibingen lnstitut fur Mikrobiologie. '- See Fig. 1.

252

t i S A H M E T Al.

1985), C-2p (Bisseret et al., 1985) or C-2a (Stampf el al., 1991), double bonds at C-6 and/or C-11 (Rohmer and Ourisson, 1976b, 1986), the presence of two diastereoisomers at C-22 (Rohmer and Ourisson, 1976a) and the polar side-chain (Fig. 1). The most common side-chains are derived from aminobacteriohopanetrioi (3) or from bacteriohopanetetrols (4) and (15) with two different side-chain configurations (Bisseret and Rohmer, 1989; Peiseler and Rohmer, 1992). Structural variations affect mostly the C-35 functional group and include: (a) Introduction of a polar group at C-35, e.g. carbohydrate derivatives linked through a glycosidic or an ether bond to the C-35 hydroxyl group (Langworthy et a l . , 1976; Renoux and Rohmer, 1985; Flesch and Rohmer, 1989; Llopiz et a l . , 1992; Simonin et al., 1992) as in side-chains (9)-(14) and (25) or amino acids linked through a peptidic bond to the C-35 amino group (Neunlist et al., 1985) as in sidechains (7) and (8). (b) Methylation of the C-35 hydroxyl group (side-chain (6); Kogan rt al., 1991). (c) Carbamoylation of the C-35 and C-24 hydroxyl groups (side-chains (16) and (17); Neunlist et a l . , 1988) (d) The presence of additional hydroxyl groups at C-30 and/or C-31 (side-chains (18)-(21); Bisseret et al., 1985; Neunlist and Rohmer, 1985a,b) . (e) Nucleoside analogue-shaped side-chains of the two diastereo-isomeric adenosylhopanes (22a) and (23a) (Neunlist and Rohmer, 198%; Neunlist et ul., 1988). Further, CD3 diols (side-chain (24), Fig. 1) have been found in two Acetobacter spp. (Peiseler and Rohmer, 1991). Whether these hopanoids with shortened chains derive from a parallel biosynthetic route or arise as a result of biodegradation of C35 bacterio-ho-panetetrols has still to be determined. Combination of all the known structural variations will already lead to an enormous number of different compounds. However, the amphiphilic character permitting membrane stabilization (see Section VII) is nearly always withheld. Structural variations might thus be allowed as far as this common feature resulting from linkage between a hydrophobic triterpenic moiety and a polar side-chain is present. Until now there has been no explanation for this large structural variability which is unusual for compounds involved in primary metabolism.

IIOPANOIDS IN BACTERIA

253

111. Detection and Analysis of Bacterial Hopanoids

The simplest hopanoids, diplopterol (2) and diploptene ( I ) , are readily extractable from freeze-dried cells by organic solvents and can be characterized by gas-liquid chromatography (GLC) and by combined CLC-mass spectroscopy (Rohmer et af., 1984). However, it has been observed that diplopterol sometimes undergoes partial decomposition on the GLC column which results in formation of a mixture of diplopterol, diploptene and hop-21-ene. This can be avoided by silylation of the lipids prior to GLC analysis (Rohmer et af., 1984; Schulenberg-Schell et af., 1989). With more complex hopanoids bearing extended side-chains, extraction and analysis are hampered by their high polarities and low solubilities in organic solvents. For quantitative determinations of these hopanoids, special care has to be taken with the choice of solvent mixtures for extraction from cells. Mixtures of methanol, chloroform and water (2: 1:O.S) according to the method of Bligh and Dyer (1059) have been successfully applied in some experiments (Schulenberg-Schell et af., 1989; Herrnans et uf., 1991; Horbach et al., 1991), or variations of this method (Folch et af., 1957; Rohmer et af., 1984). A standard procedure for analysis of complex hopanoids was developed by Rohmer et af. (1984) which involves degradation of the polyhydroxylated side-chains. The lipid extract is oxidized with periodic acid, leading to aldehydes which are reduced by sodium borohydride into the primary alcohols (26)-(29) (Fig. 1) or eventually into the secondary alcohols (30). These alcohols can be isolated and purified by thin-layer chromatography and the corresponding acetates can be analysed by GLC (Rohmer and Ourisson, 1976b; Rohmer et af., 1984; Poralla et uf., 1984). Recently, another method for analysis of periodateoxidized and borohydride-reduced hopanoids was described by Barrow and Chuck (1990) in which C3* and C3, hopanoid alcohols are converted into benzoate or naphthoate derivatives, yielding chromogenic compounds which can be separated with a reversed-phase high-pressure liquid chromatography (HPLC) system using ultraviolet-radiation detection. The systems described so far allow analysis of diplopterol, diploptene and of the sum of the hopanoids with polyhydroxylated side-chains present in the organism under investigation. Studies on the distribution and the quantities of different hopanoids and their derivatives in an organism obviously require a method for analysis of intact hopanoids. This can be done after acetylation and separation by chromatography of peracetylated derivatives as described, for instance, by Renoux and Rohmer (1985) or Neunlist et al. (1985, 1988). Hopanoid-containing fractions can usually be detected only by proton nuclear magnetic-resonance (NMR) spectroscopy

254

II \ A H M I I . ( I

exploiting the characteristic pattern of methyl singlets in the hopanc skeleton. As the hopanoid composition of most bacteria so far analyscd is completely different (Table l ) , direct qualitative and quantitative determination has to be adapted for each bacterium. Such a method was developed for the extended hopanoids occurring in Zymomonas mohilis, i.e. bacteriohopanetetrol (4a), bacteriohopanetetrol ether (174 and bacteriohopanetetrol glycoside (9a). Lipids are acetylated with acetic anhydride:pyridine and are separated on a reversed-phase HPLC column with a gradient of acetonitrile in methanol, and subsequent monitoring of the eluant at a wavelength of 206 nm (Schulenberg-Schell et al., 1989). The simplest and most rapid procedure for determination of the total hopanoid content of a bacterial population appears to be the side-chain cleavage method followed by G I X , as this method abolishes the structural heterogeneity arising from the numerous different bacteriohopane sidechains (25 different types identified presently). However, the presence of two vicinal hydroxyl groups is required. Racteriohopanepentol (19) and adenosylhopanes (22) and (23) are examples of hopanoids which cannot be detected by this method. Furthermore, for quantitative determination of the hopanoid content of a given organism it is essential that the hopanoids are solubilized during the lipid-extraction step. This should also be considered for those strains of bacteria in which hopanoids could not he detected (Rohmer er al., 1984). In addition, it may be that hopanoids are lost during periodate-borohydride treatment. Such a situation was encountered with hopanoids of the ethanologenic bacterium Z. mohilis when its hopanoid content was analysed, in parallel, by the acetylationHPLC method and the side-chain cleavage-GL,C method starting from identical lipid extracts (Hermans el al., 1991). It turned out that the hopanoid values for cells grown with low endogenous ethanol concentrations were not reproducible when determined by the periodateborohydride method. The mechanism by which hopanoids could be lost during this derivatization is not clear, but this observation may explain the large variability between reported data on the hopanoid content of Z. mobilis determined after periodate oxidation (Bringer ef al., 1985; Tahara ef al., 1988a; Barrow and Chuck, 1990); Clearly, direct analysis of intact hopanoids by HPLL is most desirable but this method has not yet been extended to a large variety of bacteriohopanoids.

1V. Distribution and Physiological Role of Hopanoids About 200 different bacterial strains have been investigated for the possible presence of hopanoids using mainly the side-chain degradation

tIOPANOIDS IN BACTERIA

255

procedure (Rohmer et al., 1984; M. Rohmer et al., unpublished results). Bacteriohopanepolyols are typical eubacterial metabolites and they could be detected neither in eukaryotes nor in archaebacteria. Traces found in some higher plants (e.g. Viola odorata) or mosses (e.g. Abietinella abietina) reflect the presence of hopanoid-synthesizing bacteria on the phylloplane, such as pink-pigmented facultative methylotrophs belonging to the genus Methylobacteriurn (M. Knani and M. Rohmer, unpublished results). Common biochemical potentialities most probably reflect phylogenetic relationships. Hopanoids might thus be useful indicators in taxonomy. Although hopanoids have been found in numerous species. no clear-cut conclusions can be drawn, as pointed out by the following examples: (a) Hopanoids are present in nearly all cyanobacteria. Synechocyslis and Synechococcus are thought to be two closely related genera, but species in the former genus synthesize hopanoids, while those in the latter do not. (b) There are hopanoid producers among species belonging to the family Rhodospirillaceae (Khodopseudornonas palustris, R. acidophila, Rhodornicrobiurn vannielii and Rhodospirillurn rubrurn) as well as bacteria without hopanoids (Rhodobacter capsulatus and Rhodocyclus gelatinosus). (c) The previous comment also applies to nitrogen-fixing species in the family Azotobacteraceae. Two Beijerinckia spp., an Azotobacter vinelandii and an A . chroococcurn strain are capable of synthesizing hopanoids. These triterpenoids could not be detected in another A . chroococcurn strain or in species in other genera belonging to this family (C. Vilcheze and M. Rohmer, unpublished results). (d) In some Streptornyces strains, hopanoids have been regularly detected. In others they are absent or present in minute amounts only. The hopanoid content of most bacteria studied so far is in the range 0.15.0 mg (g cell dry weight) as determined by the periodate-borohydride method (Rohmer el al., 1984). Exceptionally high contents of bactenohopanes were found in the ethanol-producing bacterium Zyrnornonas rnobilis (30 mg (g cell dry weight) i.e. 4&50% of the total lipid content), in the thermoacidophilic Bacillus acidocaldarius (up to 16% of the total lipid content), and in Frankia spp., bacteria which can fix nitrogen (30-50 mg (g total lipids)-’) (Poralla et al., 1984; Schulenberg-Schell et al., 1989; Berry etal., 1991; Hermans et al., 1991). Zyrnornonas rnobilis, an obligately fermentative bacterium, can produce up to 13”/0 (w/v) ethanol. Thus, its ethanol tolerance is remarkably high and comparable to that of Saccharornyces cerevisiae (Rogers et al., 1982). The high content of

’

’,

256 TABLE 2

H PAHM t T A1

Influence of growth conditions on composition of lipids from Zvrnornonas rnohibs ATCC 29191 grown in rontinuous culture Mean concentration f. SEM

Ethanol

Hopanoid: Phospholipids Extracted total phospholipid ratio (mg g-')b lipids (mg g') (mol mol-I)

(g I-')

Growth (h-')

Hopanoids ( m g g ')o

4.9 4.8

0.08 03

390 k 70 430 SO

*

606+8 so4 4

10.1 9.8 9.8

0.08 0.25 0.w

450 k 30 400 5 M 490 70

*

591 rt 4 so4 ? 7

35

0.15 0.3Zd

420 k SO 460 ? 40

0.08 0.19"

430 rt_ 30 370 f 40

34

60 5s

*

990 80 930 f 60

0.67 0.93

40 990 f 70 990 80

* *

0.78 0.72 1.03

642 ?. 7 482 ? 7

1070 f 60 940 so

0.70

613 f IS 613 f 7

1040 ?

so

0.77 0.68

*

613 ? 7

1060

*

960

* so

1.os

From Hermans et al. (1991). "n=8. 'n=4.

Maximum growth rate for 10 g ethanol I" and 1 12 g glucose I-' as steady-state concentrations. Maximum growth rate at the indicated ethanol concentrations.

hopanoids in this organism (Table 2) may establish optimal membrane stability and fluidity over a wide range of ethanol concentrations. In R . acidocaldorius, growth temperature strongly influences the hopanoid level of the cells (Poralla et al., 1984). At a growth temperature of 6 5 T , the hopanoid content of the cells was seven-fold higher compared with that of cells grown at 60°C, and amounted to 16% of the total lipid content. In addition, there was a pH-value effect of the medium on the hopanoid content which was highest under acid conditions (pH 3). In this organism, hopanoids counteract the fluidizing effects of high temperatures and probably play a role in increasing the "tightness" of the membrane for protons (Poralla et al., 1984). Frankia spp. can fix nitrogen in the presence of atmospheric oxygen. It was suggested that the vesicle envelope, a lipidcontaining structure, provides an oxygen-diffusion barrier to protect nitrogenase activity. The large, amounts of bacteriohopanetetrol present in this organism may contribute to the impermeability of the vesicle for oxygen (Berry et al., 1991). The occurrence of hopanoids in these specialized organisms exemplifies their general role as membrane reinforcers in

257

HOPANOIDS IN BACTERIA

eubacteria. Their ample presence counteracts the fluidizing effects of high concentrations of ethanol and high temperatures, and contributes to the “tightness” of membranes for protons and oxygen. The presence of hopanoids is essential for survival of those bacteria which contain them (Flesch and Rohmer, 1987). This was demonstrated using various inhibitors of squalene-hopene cyclase, e.g. 2,3-dihydro-2azasqualene, Hopanoid-synthesizing bacteria did not grow in the presence of these inhibitors at a concentration of 1 p ~ whereas , the growth rate of hopanoid non-producers was not changed in the presence of an inhibitor . effect of azasqualene on the growth and concentration of 200 p ~ The hopanoid content of 2. mobilis has also been studied (Horbach et al., 1991). Anaerobically grown cells of 2. mobilis showed a retarded growth rate in the presence of 4 PM inhibitor and failed to grow in the presence of concentrations of azasqualene above 10 ~ L M(Horbach et al., 1991). When grown in the presence of 3% (w/w) ethanol and 7 PM azasqualene, the hopanoid content of the cells was only 60% of that of cells grown in the absence of inhibitors. Under these conditions, Z. mobilis accumulated

0

20

40

60

i

80

Ethanol (g/l) FIG. 2. Effect of ethanol concentration on growth rate of Zymomonas mobilb with reduced hopanoid content. Cells grown in the presence of 2% glucose and 7 FM azasqualene with 3% (w/v) ethanol (m), and without ethanol (A), were harvested at an optical density value of 0.8, washed twice and transferred to fresh medium containing different concentrations of ethanol.

258

I1 SAI1M E9' A I .

22 pg of squalene in each milligram of total cell lipid, confirming that squalene-hopene cyclase was the in vivo target of azasqualene. The effect of azasqualene offered the possibility of studying the dependence of the ethanol tolerance of the bacterium on the hopanoid contents of cells. Zymomonas mobilk with a lower hopanoid content was obtained by growing cultures in the presence of azasqualene followed by transfer of cells to media containing 0-70 g ethanol 1.' (Fig. 2). Growth inhibition of cultures caused by increasing concentrations of ethanol was much stronger for cells with a decreased hopanoid content, showing that hopanoids are directly involved in ethanol tolerance in 2. mobilis. Sterols and hopanoids are amphipathic molecules possessing a rigid, hydrophobic, quasi-planar polycyclic skeleton and a polar head-group corresponding to the 3P-hydroxyl group of sterols or the polyhydroxylated side-chain of hopanoids. Their similar molecular shapes and dimensions allow their insertion into phospholipid bilayers and van der Waals interactions with acyl side-chains. From these common structural features corresponding to those required for sterols as efficient membrane stabilizers (Demel and De Kruyff, 1976), identical structural roles have been postulated for sterols and hopanoids (Rohmer et al., 1979). Indeed, hopanoids have recently been localized in the plasma membrane and in the outer membrane of the Gram-negative bacterium 2. mobilis (Taharaelal., 1988b) as well as in thylakoids and the outer membrane of the cyanobacterium Synechocysfis sp. (Jiirgens ef al., 1992). The functional equivalence of sterols and hopanoids as membrane stabilizers was directly tested in two biological systems. The first evidence was obtained with the ciliate Tefrahymenapyriformis. When grown in the absence of sterols, this organism synthesizes de novo terahymanol, a quasihopanoid isomer of diplopterol (2), and incorporates it mainly into the plasma membrane (Thompson et al., 1971). When sterols are present in the culture medium, tetrahymanol biosynthesis is inhibited, and sterols are found in membranes (Conner et al., 1968; Ferguson et al., 1975). Furthermore, in the presence of hypocholesterolemic drug triparanol, acting in this protozoon as a squalene cyclase inhibitor (Sipe and Holmlund, 1972), growth was completely inhibited (Aaronson ef al., 1962). Cell division could be restored by addition to media of sterols, sterol precursors such as cycloartenol or lanosterol, as well as by addition of the natural bacteriohopanetetrol mixture from an Acefobacter sp. (Peiseler and Rohmer, 1992) o r carotenoids, suggesting to some extent the physiological equivalence of these polyterpenoid compounds (Raederstorff and Rohmer, 1988). The functional equivalence of hopanoids and membrane sterols was also verified in a prokaryote, namely Mycoplasma mycoides, which requires cholesterol for growth. This micro-organism could be grown in the absence

HOPANOIDS I N BACTFRIA

259

of sterols, provided that a hopanoid, in this instance diplopterol (2), was added to the culture medium (Kannenburg and Poralla, 1982). Further confirmation of the membrane-reinforcing role of hopanoids has been obtained with artificial membranes. Addition of bacteriohopanetetrol (4a) or of its glycoside (loa), isolated from B. acidocalduriur, to a phospholipid monolayer resulted in condensation of the saturated palmitoyl n-acyl chains and in a decrease in sharpness of the phase transition between the gel-like phase and the liquid-crystal phase, analogous to the effect of cholesterol (Kannenberg et ul., 1980; Poralla er al., 1980). The condensing effect of bacteriohopanetetrol was shown to be equal to that of cholesterol when monolayers of phosphatidylcholine with branched ante-iso chains were used, glycerophospholipids often found in bacteria (Kannenberg ef a l . , 1983, 1985, 1986). This effect was also observed with monolayers obtained using phospholipids from Z. mobilis after inclusion of a mixture of total polar lipids containing high concentrations of bacteriohopanetetrol glycoside (9a) and ether (13a) (Hermans, 1992). Similar results were obtained with phospholipid bilayers. Complete suppression of the phase transition was achieved by a concentration of 40 mol% of bacteriohopanetetrol as observed by microcalorimetry (Kannenberg ef ul., 1983). Incorporation of 2P-methyldiplopterol (2e) or bacteriohopanetetrol (4a) into a black lipid membrane, a phospholipid bilayer system used to estimate membrane permeability, decreased permeability (Renz et al., 1983). When incorporated into unilamellar phospholipid vesicles, a natural mixture of bacteriohopanetetrols from an Acerobacfer sp. improved their rigidity and barrier properties as determined by swelling under an osmotic shock followed by stopped-flow transmittance measurements (Risseret et a l . , 1983). The same bacteriohopanetetrol mixture was further shown by deuterium-NMR spectroscopy to be capable of ordering (less efficiently than cholesterol, however) oriented bilayers of dimyristoylphosphatidylcholinewith perdeuterated sn-2 chains, whereas diplopterol (2a) had only a small effect (Krajewski-Bertrand, 1991).

V. Biosynthesis and Genetics

A . SYNTHESIS OF ISOPENTENYL PYROPHOSPHATE

In eukaryotes, isopentenyl pyrophosphate (IPP) is the biosynthetic precursor for all isoprenoid compounds (Lynen et a l . , 1958a, 1959; Nes and McKean, 1977). According to the generally accepted scheme, the socalled acetoacetate or Bloch-Lynen pathway, IPP is synthesized from three

260

t l . SAHM t?'A I

2 Acetyl- CoA

1 Aceto-acetyl

- coA

A

u 0

- CoA

1

Hydroxymethylglutaryl - CoA

Mevalonate

1 Mevalonate PP

1 lsopentenyl PP

0

SCoA

0

OH 0

AAA SCoA

HO

HO

OH

rn 0

Mevalonate P

SCoA

OH

HO

OP

* 0

OH

OPP

HO

AA

OPP

FIG. 3. Biosynthesis of isopentenyl pyrophosphate in eukaryotic cells following the acetoacetate pathway. P indicates phosphate, PP pyrophosphate.

molecules of acety-CoA, via acetoacetyl-CoA, hydroxymethylglutarylCoA, mevalonate, mevalonate phosphate and mevalonate pyrophosphate (Lynenelul., 1958b;Bloch, 1965;Nesand McKean, 1977;Fig. 3). However, so far it has not been shown in detail that the same pathway is operating in prokaryotes. Several lines of evidence point to a long-overlooked problem concerning the first steps of isoprenoid biosynthesis in prokaryotes. First, with archaebacteria, at least three reports established the overall operation of the classical route from acetate to isoprene units, showing either incorporation of [ 1-I3C]or [2-'4C]acetate into biphytanyl components

HOPANOIDS

A

IN HACTEKIA

26 I

B

FIG 4. Origin of carbon atoms in isoprenic units synthesized by Halobacterium cutirubrum and Hulobucterium hlobium alter feeding cultures with [ 1,2-"CJacetate or [U-'3C]lysine. Shown in the diagram are (A) the expected labelling pattern from the eukaryotic pathway and (B) the observed labelling pattern. 0 indicates arnino-acid the methyl group of acetate; @, the carboxyl group of acetate; 0, residues, particularly lysine.

of extreme thermoacidophilic bacteria of the genus Caldariella (de Rosa er ul., 1977a, 1980) o r incorporation of "C from Il-"C]- and [2'*C]acetate into the isoprenoid side-chain of caldariellaquinone from Culdariellu acidophilo (de Rosa et a l . , 1977b). In contrast, Ekiel etal. (1986) observed that the labelling patrern after feeding with [ 1,2-l3C2]acetate, of both squalene and the phytanyl side-chain of Halobacterium cutirubrum and H . hulobium lipids, could not be explained in the classical manner (Fig. 4). The authors proposed the involvement of two molecules of acetate, probably through acetyl-CoA, together with a degradative product provided by amino acids, particularly lysine. Therefore, it seems possible that, in the species the genus Culdurieffu,the classical route is followed, while in Halobacterium spp. a slightly different isoprenoid pathway (two acetate residues and one unknown carbon source) is realized. Secondly, in 1981 Pandian et al. (1981) reported that [l-''C]- or [2-I4C]acetate, in the presence of glucose (2%), labelled the side-chain of ubiquinones from Escherichia coli, Azotobacter vinelandii, Pseudomonas sesami and Rhodopseudomonas palustris in a non-classical way. The ratio of incorporation of C-1 to that of C-2 of acetate into the isoprenic unit was much less than the ratio that would be expected (2:3) from the wellknown pathway through acetoacetate and hydroxymethylglutaryl-CoA (see Fig. 4, labelling pattern A). Consequently, the authors ruled out the acetoacetate pathway in favour of an acetolactate pathway with pyruvate and acetaldehyde as initial substrates leading to mevalonate, which is further converted to IPP through the Bloch-Lynen pathway. Zhou and White (1991) excluded operation of the acetolactate pathway in Escherichia coli proposed by Pandian et al. (1981) because intermediates on the pathway did not label ubiquinone (UQ-8). A second contradiction

262

II

SAIIM 1 7 A /

of the data of Pandian et al. (1981) came from incorporation of I3C2units into fatty acids but not into tJO-8 by cclls of E. coli grown on [ 1,2-''C7]acctate and non-labelled glucose (0.4%). Using [U-'3C]glucose and acetatc. however, fatty acids and UQ-8 carried the 'k7label. Zhou and White (1991) presented further data which indicated that the C-2 and C-3 carbon atoms of pyruvate. but not free acctyl-CoA, were incorporated into the C-3 and C-5 positions of TPP in E. coli. The authors proposed enzyrncbound intermediates on the route to IPP. these being acetylated protein4 with acyl groups derived directly from pyruvate by dccarboxylation. Thcsc acetylated proteins could be analogous to acyl-carrier proteins involved in fatty-acid biosynthesis (Wakil, 1989). Thirdly, Streptomyces hygroscopicus, S. griseus and S . noursei failed to incorporate [U-I4C]acetic acid into squalene and related hydrocarbons in pulse-labelling experiments, in contrast to incorporation into the triglyceride fraction (Grafe et a l . , 19x5). This report as well as biosynthctic studies on isoprenoid pentalenolactone with [6-*H2Jglucose using Streptomyces strain UC5319 (Cane et a l . , 1981) are opposed to the classical acetoacetate scheme for isoprenoid biosynthesis. However, biospnthetic studies with a Kitasatosporia sp. (strain MF73O-N6), Chinia rubra and Streptomyces aeriouvifer revealed incorporation of [ I .2-"C2]- and/or [?I3C]acetate into isoprenoid derivatives following the acetoacetate route (Isshiki et al., 1986; Shiomi et al., 1987: Shin-ya ef al., 1990). Fourthly, several bacteria are known to incorporate mevalonate, ;in intermediate on the acetoacetate pathway in eukaryotes, into distinct isoprenoids. These bacteria include Staphylococcus spp., I,actobacil/u.s plantarum, Streptococcus mutans, St. faecium, H . cutirubrum, Micrrmxxx.s luteus and Myxococcus fulvus (Suzue et al., 1964; Gough et al., 1970; Thorne, 1973; Kushwaha and Kates, 1976; Taylor and Davies, 1981; Evans and Prebble, 1982; Kleinig, 1975). In contast, incorporation of ['4C]mevalonate and [I4C]acetate could not be achieved into the various homologues of coenzyme Q by Agrobacterium tumefaciens, A . vinelandii, a Aeudomonas sp. or E . coli, although significant amounts of radioactivity were present in cellular lipids (Raman et al., 1965). Fujisaki et al. (1986n) were unable to demonstrate incorporation of [ 14C]mevalonateinto UQ-8, in contrast to ['4C]IPP, using lyophilized cells of E. coli. Finally, further results originated from a series of feeding experiments using [l-"C]- and [2-I3C]acetate and Methylobacterium organophilum, R. palustris and R . acidophila (Flesch and Rohmer, 1988). Although poly-phydroxybutyrate was normally labelled, there was no direct incorporation of acetate into terpenoids, added to which no scrambling took place. The most conclusive hints originated from incorporation of [ I3C]glucose into the hopanoids of Zymomonas mohilis (M. Rohmer, M. Knani, P. Simonin,

263

IIOP4NOIDS IN H 4 C I ' E H I A

R. Sutter and 14. Sahm, unpublished observation). This bacterium is especially useful f o r this kind of experiment because it contains substantial amounts of hopanoids. uses glucose as a sole carbon source via the EntnerDoudoroff route (Gibbs and IIeMoss, 1954; Swings and DeLey, 1977) and lacks other pathways for glucose catabolism, while finally it is incapable of gluconeogenesis and has an incomplete tricarboxylic acid cycle (Dawes CI a l . , 1970: Swings and DeLey, 1977). A complete tricarboxylic acid cycle could cause rearrangements o f labelled carbon atoms during metabolism. Using [l-"'C]-, [2-"C]-, [S-"C]-, [5-I7C]- and [6-I3C] glucose, the origin of all of the carbon atoms i n isoprene units forming the triterpenic ring system of t h e hopanoids has been determined (M. Rohmer, M. Knani, P. Simonin, R . Sutter and H. Sahm. unpublished results). The pattern of labelling in 1PP obtained by these experiments was compared to the distribution of labels expected if IPP synthesis occurred through the acetoacetate pathway (Fig. 5 ) . Obviously, the labelling pattern of IPP in Z.mobifis is incompatible with operation of the acetoacetate pathway as it operates in eukaryotes. This suggests a novel pathway for biosynthesis of the isoprcnoid unit in Z.mobilis o n which involvement of two different compounds must be postulated. One of these precursors (constituting C-1, C - 2 , C-4 of IPP with 100% labelling intensity) is derived from an intermcdiatc of the Entner-Doudoroff pathway of glucose catabolism at a A

B

Eukaryotes

I 5

Zymomonas mobilis 15

2

4 lsoprenic unit

lsoprenic unit

.C 1 . C-2

from C-2 C-5 of glucose (50 %i from C-3 C-6 of glucose (50 I)

- C-1: from C-6 of glucose (100 %) - C-2: from C-5 of glucose (100 %)

- C-3

from C-2 C-5 of glucose (50 %)

.C-3: from C-2iC-5 of glucose (50 %)

C-4 from C-3 C-6 of glucose (50 %)

- C-5

from C-3C-6 of glucose (50 %)

- C-4: from C-4 of glucose (100 %) ~

C-5: from C-3C-6 of glucose (50 %)

FIG. 5 . Origin-of carbon atoms in isoprenic units of Zyrnornonas rnobilis after feeding with ["C]glucose. Shown in the diagram are (A) the labelling patterns, after glucose catabolism through the Entner-Doudoroff route, expected on the pathwily followcd by eukaryotes and (B) that observed in Zyrnarnanas mabilis. The percentages indicate the relative labelling intensities.

264

H S A I I M b T Af.

stage prior to pyruvate. 'The second precursor (constituting C-3 and C-S of IPP with 50% labelling intensity) is pyruvate itself or a derivative of i t . In E. coli, C-2 and C-3 of pyruvate were incorporated into the C-3 and (2-5 positions of IPP, respectively (Zhou and White, 1991). Llnfortunarely , labels in the C-3 and C-5 positions of IPP synthesized by Z. mobilis are indistinguishable from those stemming from a biosynthetic pathway involving acetoacetate and therefore cannot be used for differentiation (Fig. 5). Thus, results obtained by Zhou and White (1991) with E. coli canriot be directly applied to Z . mobrfis. Evidence available at present leads to the conclusion that different pathways for IPP synthesis are realized in eubacteria For Staphvbcoccw curnosus, H . cutirubrum, M . fulvus, I , . pluniurum knowii to incorporate mevalonate into isoprenoids, we recently demonstrated in v i m erlzyrriic conversion of acetyl-CoA to IPP, following the same pathway as eukaryotes (S. Horbach, H. Sahm and R . Welle, unpublished results). It IS therefore possible that all bacteria incorporating mevalonate into isoprenoids follow the acetoacetate route. However, for 2. mobilis, in vivo labelling data froin glucose are clear-cut and show that the IPP pathway differs from that i n eukaryotes to a large extent. Recent data suggest that the pathway to IPP as found in Z.mobilis is also realized in other bacteria including Me. orgunophilum, Me. jujisuwaense, Rhodopseudomonus sp. and E . c d i (M. Rohmer, M. Knani, P. Sirnonin, B. Sutter and H . Sahm, unpublished results). 'I'his question needs to be addressed further, and is currently under investigation. B . FORMATION OF 'THE HOPANE I'ENI'ACYCLIC' SKEl-WON FKOM ISOPFNTENYL PY KOPHOSPHA I'E

'I'he second stage in isoprenoid synthesis leads to formation o f compounds which are based on (C,),, skeletons in which C5 is an isoprene residue. 'l'he reaction sequence in eukaryotes leading to squalene (composed of six isoprene residues) is shown in Fig. 6. Does this pathway also operate in prokaryotes? It is well established in E. coli (Fujisaki et ul., 1986a,b) and Z.mobilis (Shigeri et a l . , 1991) that IPP serves as the precursor tor specific isoprenoid compounds. For these two bacteria, corresponding enzyme activities have been demonstrated in vitro. The gene encoding farnesylpyrophosphate synthase (ispA) from E. coli was cloned (Fujisaki el u l . , 1990). Furthermore, glucose-feeding experiments with Z.mobilis (Rohmer et ul., 1989) proved that the hopanoid framework consisted of six identically labelled isoprene units. Thus, in Z . mobilis arid E . coli, IPP is converted to dimethylallyl pyrophosphate by isopentenyl-pyrophosphate isomerase, and subsequently to geranyl pyrophosphate and farnesyl pyrophosphate involving farnesyl-pyrophosphate synthase. Squalene, an acyclic, symmetrical

DMAPP

Lapp-

OPP

A

OPF

ipp

GPP

-

lsopentenyl Adenine (t RNA)

Monoterpenes

Dolichol Haem A Ubiquinone Carotenoids Chlorophyll Farnesylated Proteins

AAAAAA

OPP

S q ualene FIG. 6. Pathway for biosynthesis of squalene from isopentenyl pyrophosphate (IPP) as observed in eukaryotes and prokaryotes. DMAPP indicates dimethylallyl pyrophosphate; GPP, geranyl pyrophosphate; FPP, farnesyl pyrophosphate; 1, isopentenylpyrophosphate isomerase; 2, farnesyl-pyrophosphate synthase; 3, squalene synthase.

266

t I SAHM ET A L .

C30triterpenoid hydrocarbon, is finally formed by tail-to-tail condensation of two molecules of farnesyl pyrophosphate involving squalene synthase (Fig. 6). Bacterial squalene cyclase, a membrane-bound enzyme, catalyses cyclization of squalene to the pentacyclic hopanoid skeleton, yielding diploptene and diplopterol (Fig. 7). Using enzyme preparations from Bacillus acidocaldarius, diploptene and diplopterol are formed in a molar ratio of 5 4 1 , irrespective of reaction conditions and degree of enzyme purification (Seckler and Poralla, 1986; Neumann and Simon, 1986). The hopanol is thought to be formed by an alternative reaction of the cyclase favoured under in vitro conditions whereas hopene is the presumed biosynthetic precursor of bacteriohopanetetrol. So far, squalene cyclase is the only enzyme of the hopanoid biosynthetic pathway which has been studied in detail. It has been characterized in several bacteria (Anding et al., 1976; Rohmer et al., 1980a,b; Flesch and Rohmer, 1987) and was purified to homogeneity from the thermoacidophile (Alicyc1o)lBacillus acidocaldarius (Neumann and Simon, 1986; Seckler and Poralla, 1986; Ochs el al., 1990). Values for molecular weight and K, were determined to be 69,473 Da (627 amino-acid residues) and 3 PM, respectively. Compared with squalene-epoxide cyclases from eukaryotes there are two major differences (Rohmer et al., 1979). First, biosynthesis of the pentacyclic triterpenoid skeleton of the bacterial hopanoids and of tetrahymanol from Tetrahymena pyriformis is independent of oxygen and therefore an anaerobic process (Fig. 7). It proceeds by direct cyclization of the squalene molecule. In contrast, biosynthesis of sterols is an aerobic process. The squalene molecule is first converted into squalene 2,3-oxide through a mono-oxygenase. The epoxide is then cyclized, to form either the tetracyclic triterpenoid lanosteral (in animals and fungi) or the pentacyclic triterpenoid cycloartenol (in photosynthetic organisms). Secondly, substrate specificity is lower. Cyclases in eukaryotes act only on the (3s) enantiomer of squalene oxide whereas the enzymes in prokaryotes attack, besides their normal substrate squalene, both enantiomers of squalene oxide which normally are absent from bacterial cells. Common properties of the enzyme from B. acidocaldarius compared with epoxysqualene-cyclizing enzymes are inhibition by sulphydryl reagents, such as p-chloromercuribenzenesulphonic acid, indicating the presence of an essential sulphydryl group in the enzymes (Seckler and Poralla, 1986). Squalene analogues (e.g. azasqualene), mimicking putative high-energy intermediates of the transition state, and n-alkyltrimethylammonium halides are particularly potent inhibitors of bacterial cyclases at concentrations in the micromolar range. Furthermore, growth of hopanoid-producing bacteria is efficiently inhibited by these compounds

Squalene

(c30)

I I

I

I I I

8

4

Tetrahymanol

1

I

(S) - squalene 2,3 - oxide (Cs0)

I I

@

I I

I Diploptene ( c 3 0 ) 0%

Diplopterol (c30)

I

I I

*.

I

Bacteriohopanepolyols

(C35

1

Lanosterol (cs0)

I (animals and fungi) I I I

J-*

I

Cyclization

Cyclo-artenol (C30) (higher plants and algae)

Further Sterols

I

FIG. 7. Pathways showing formation of (A) hopanoids and tetrahymanol in prokaryotes and Tetrahymena pyriformis, respectively and (B) sterols in eukaryotes. Enzymes involved are squalene cyclases ( l ) , protozoon cyclase (la), bacterial cyclase (1 b), squalene mono-oxygenase (2). squalene-2.3-oxide cyclases (3). lanosterol cyclase (3a). and cycloartenol cyclase (3b).

268

H S A H M ET AI.

whereas hopanoid-less bacteria are not affected at concentrations up to 200 PM (Flesch and Rohmer, 1987). A unique and most striking feature of C35bacteriohopane derivatives is the carbonxarbon linkage between the pentacyclic triterpenoid hopane skeleton and a Cs polyhydroxylated n-alkyl chain. Feeding experiments using R. palustris, R. acidophila, Me. organophilum (Flesch and Rohmer, 1988) and 2. mohilis (Rohmer et al., 1989) indicated unambiguously that the precursor of the side-chain is a n-pentose arising from the non-oxidative pentose phosphate pathway linked via its C-5 carbon atom to the hopane skeleton. From the stereochemistry, this percursor is most probably a D-ribose derivative in the biosynthesis of tetrol (4a) and aminotriol(2a) (Neunlist and Rohmer, 1988; Neunlist et al., 1988; Bisseret and Rohmer, 1989) or possibly a D-arabinose derivative in the production of tetrol(l5a) (Peiseler and Rohmer, 1991). At present, neither the enzymic reaction involved in the coupling nor the precise structure of the triterpenic precursor is known. C. GENETICS OF ISOPRENOID AND HOPANOID BIOSYNTHESIS

Up to now, only few data are available on the genetics of bacterial isoprenoid and hopanoid biosynthesis. Early steps in hopanoid biosynthesis still need to be revealed and, therefore, no bacterial genes or mutant strains in this area of metabolism have been described yet. From Pseudomonas mevalonii (utilizing mevalonate as a carbon source), the gene mvaA for an NAD+-dependent enzyme HMG-CoA reductase has been isolated and characterized (Beach and Rodwell, 1989). The enzyme, however, is probably mainly involved in mevalonate catabolism rather than in isoprenoid biosynthesis. Combined biochemical and genetic evidence for the absence of the mevalonate pathway in E. coli was presented by Oulmouden and Karst (1990). They found that the gene ERG12 (for mevalonate kinase) from Saccharomyces cerevisiae was expressed in E. coli, whereas the host strain did not contain mevalonate kinase activity. Mutants of isoprene biosynthesis in E. coli were selected by their resistance to aminoglycoside antibiotics (e.g. kanamycin). The rationale, therefore, was that uptake of aminoglycosides is energy-dependent and is hampered when respiratory ubiquinones (carrying isoprene side-chains) are lacking. A number of mutants with decreased activities of IPP isomerase, farnesyl-pyrophosphate F'PP synthase or prenyltransferase were found but profound genetic analysis was prevented by relatively high residual enzyme activities (Sherman et al., 1989). An E. coli mutant with a temperature-sensitive farnesyl-pyrophosphate synthase (gene ispA) was found to possess 21% residual activity in vitro at 30°C and 5% at 42°C.

HOIANOIDS IN B A C E R I A

269

The effect on the ubiquinone content of the mutant strains, however, was low. An almost equal content was found at 30°C and 66% of normal at 42°C (Fujisaki ct af., 1989). This may be a hint that farnesyl-pyrophosphate synthase is either not a rate-limiting enzyme in isoprenoid biosynthesis in E. coli or that the enzyme in the temperature-sensitive mutant may be stabilized in vivo by the intracellular environment (Fujisaki et af.. 1989). Cationic detergents, such as octadecyltrimethylammonium chloride, inhibit at micromolar concentrations only the gowth of hopanoidcontaining bacteria (Flesch and Rohmer, 1987). Mutants of Z. mohilis resistant to octadecyltrimethylammonium chloride were isolated but were not altered in their hopanoid content (Michel et al., 1992). Tahara et al. (1988a) reported isolation of Z. mohifis mutants that showed 6% residual tetrahydroxybacteriohopanol but a three-fold higher hopane-22-01 content. Enrichment for mutant isolation was with nystatin, a macrolide antibiotic active against sterol-containing cells. No further reports on the mutant have been given Two reports on bacterial genes involved in farnesyl pyrophosphate (gene ispA from E. rofi) and hopene synthesis (squalene-hopene cyclase from B. acidocafduriw) were recently given (Fujisaki et a l . , 1990; Ochs et a l . , 1992). The ispA gene of F . cofi (for farnesyl-pyrophosphate synthase) was cloned and analysed. T h e enzyme (calculated M, 32,158) activity in E. cofistrains carrying ispA on a plasmid was raised up to 16-fold as compared with the wild type. The derived amino acid-residue sequence of the enzyme from E. cofi showed prominent boxes of identical residues with farnesylpyrophosphate synthases of eukaryotic origin (rat and Saccharomyces cerevisiae) and, moreover, to phytoene synthase (gene crtE) an enzyme involved in carotenoid biosynthesis in Rhodohacter capsufatus (Fujisaki et af., 1990) A stretch of 51 amino-acid residues contained 28 (55%) identical residues in the enzyme from E. cofi and in the crtE gene product. Another box of 22 amino-acid residues was common to all four sequences with the pattern: GxxFQxQDDxL~DxxGxxxxxGKwhere x denotes non-conserved residues. These highly conserved amino-acid residues may constitute an important site (possibly substrate binding) for the activity of prenyltransferases (Fujisaki et af., 1990). The gene for squalene-hopene cyclase was cloned from Bacillus acidocaldarius and expressed in E. coli (Ochs et al., 1992) with a four-times higher activity compared with B. acidocafdurius. The cyclase is a dimer consisting of two identical subunits. The DNA-derived amino acid-residue sequence gave a subunit size of 69,473 Da, corresponding well with the size estimated for the punfied enzyme subunits (Ochs et af., 1990). Cyclase activity was found in the membrane fraction of recombinant E. cofi. Since no transmembrane (hydrophobic) portion could be detected from the

270

I 1 SAHM 61 A I

derived amino acid-residue sequence, the cyclase was suggested to be anchored in the membrane by hydrophobic interactions (Ochs el al., 1992).

VI. Conclusions Hopanoids are abundant natural products of eubacteria. 'The lipid family of hopanoids consists of numerous structural variants, a fact that can lead to pitfalls in their detection and analysis. These pentacyclic triterpenoids act as structural reinforcers of bacterial membranes and, thus, are structurally and functionally related to sterols found in membranes of eukaryotes. As a result of feeding experiments with 13C- and ''C-labelled substrates, conducted with different hopanoid-producing bacteria, deviations from the isoprenoid biosynthetic pathway used by eukaryotes have been observed. This apparently new pathway and the nature of the linkage of the glycosidic side-chains to the hopanoid skeleton remain to be revealed by combined biochemical and genetic analyses. KEFEKENCES

Aaronson, S., Bensky. B.. Shifrine, M. and Baker, B. (1962). Proceedings ofthe Society of Experimental Biology and Medicine 109, 130. Anding, C., Rohmer, M. and Ourisson, G . (1976) Journal of rhe ,4merican Chemical Society 98, 1274. Barrow, K. and Chuck, J. A. (1990). Analytical Biochemutry 184, 395. Beach, M. J . and Rodwell, V. W. (19x9). Journal of Bacteriology 171, 2994. Benz, R.. Hallmann, D . , Poralla, K.and Eibl, H. (1983). Chemistrvand Physics of lipids 3 4 , 7 . Berry, A. M., Moreau, R. A. and Jones, A . D. (1991). Plant Physiology 95. 111. Bird, C. W.. Lynch, J . M., Pirt, S. J . , Reid, W. W.. Brooks. C. J. W. and Middleditch. B. S. (1971). Nature 230, 473. Bisseret, P. and Rohmer, M. (1989). Journal of Organic Chemistry 54, 2958. Bisseret, P., Wolff, G., Albrecht, A.-M., Tanaka, T., Nakatani, Y.and Ourisson. G . (1983). Biochemical and Biophysical Research Communications 110, 320. Bisseret. P., Zundel, M.and Rohmer, M. (1985). European Journal ofhiochcmistry 150.29. Bligh, E. G. andDyer. W. J. (1959). UanadianJournalofRiochemistryand Physiology37,911. Bloch, K. (1965). Science 150. 19. Bouvier, P. (1974). These de 38me Cycle: Universite Louis Pasteur, Strasbourg. Bringer, S., Harntcr.T., Poralla, K.and Sahm, H. (1985). ArchivesofMicrobir,bgy 140,312. Cane, D. E., Rossi. T.. Tillrnann. A. M. and Pachlatko. J. P. (1981). Journalof the .4mc,ricun Chemical Sociery 103, 183X. Conner. R. L., Landrey, J. R., Burns, C. H. and Mallory. F. B. (Iyhx). J O I J ~ R Uof / Protozoology 15, 600. Dawes, E. A., Medgley, M. and Ishaq, M. (IY70). Final technical report (Dcc. 1970) for Contract No. DAJA 37-67-C-0567, AD 726033. European Research Oftice, US Army. de Rosa, M., Gambacorta, A., Minale, L. and Bu'lock, J . (1971). Journal ofthe Chemical Society, Chemical Communications 619. de Rosa, M., de Rosa S. and Gambacorta, A. (1977a). Phytochemistry 16, 1909. De Rosa, M., de Rosa, S . , Gambacorta, A . , Minale, L., Thompson, R. H. and Worthirigton, R. D. (1977b). Journal of the Chemical Society, Perkin Transactions 1, 653.

IIOPi\NOIDS IN R A ( T E R 1 A

271

d e Rosa. M.. Gamhacorta A . and Nicolaus, B (1980). Phytochemisrry 79, 791. Demcl. R . A . and De Kruyff. €3 (1976) Riochimica et Riophysica Acra 457, 109. Ekiel, I . , Sprotl. G D. and Smith, I. C. P. (1986). Journal of Racreriology 166. 559. Evans, J . A . and Prrhblc. J . N. (1982) Microbios /,erters 21, 149. Ferguson, K . A , Davis F M . Conner. R K , Landrey, J R . and Mallory F. B (1975). Journal of Riologicul (‘hemivtrv 250. 6998. Flesch. C i . and Rohmcr. M . (1987) Archives of Microbiology 147, 100. Flesch. G. and Rohmcr. M .(19x8). European Journal of Biochemistry 175. 405 Flesch. G . and Rohmer, M . (1989). Biochemical Journal 262, 673. Folch. J . . Ixes, M. and Stanley. G. H . S. (1057). Journal ofRiologica1 Chemistry 226, 497. Forster, H. J . Bicmann. K . . Haigh. W . C i . , Tattrie, N. H. and Colvin, J . K. (1973). Riochemical Journal 135. 133. Fujisaki, S . . Nishino. T. and Katsuki. H. (1986a). Journal of Biochemistry 99, 1137. Fujisaki. S , Nishino. T. and Katsuki, H . (1086b). Journal of Biochemistry 99, 1327. Fujisaki, S , Nishino, T , Katsuki. H . , Hara. H , Nishimura, Y . and Hirota, Y . (1989). Journal of RacwrioloKy 171, 5654. Fujisaki. S Hara. H . . Nishimura, Y . . Horiuchi, K. and Nishino, 7’. (1990). Journal of Rivchem~stry108, W.5. Gelpi, E , Schncider. H . , Mann, J . and CJro, J . (1070) Phytochemistry 9, 603. Gibhs. M and DeMoss. R . D (1054). Journul of Biological Chemistry 2M,689. Chugh, D. P , Kirbv. A L., Richards. J . B. and Hemming. F. W. (1970). Biochemical Jaurncil 1 18. I67 Grafe I1 , Reinhardt G Hlnel. F Schadc, W . and Ciumptert, J . (1085). Journal of Basic Microbiology 25, 503. Hermans. M. (1992). Thesis, Hcinrich-Heinc-l!niversit~t, Diisseldorf. Hermans, M. A . F , Neuss. B and Sahm, H (1991). Journal of Bacteriology 173, 5592. Ncuss. R and Sahm, H (1991). FEMS Microbiological Letters 79, 347. Isshiki, K , Tiamamura, 7 . Sawa, T Yaganawa. H , Tdkeuchi, T. and Umezawa, H . (1986). The Journal of Antibiotics XXXIX, 1634. Jiirgens. I-‘. R‘.. Simonin. P . and Rohmer. M. (1992). FEMS Microbiological Letters 92,285. Kanncnberg. E. and Poralla, K (1982). Archives of Microbiology 533, l(KJ. Kannenbcrg, E . , Poralla, K. and Blume, A . (1980). Nuturwissenschaften 67, 458. Kannenberg, E . , Blume, A , , McElhancy, R. N and Poralla, K. (1983). Riochimica er Ririphy.sit.u Arm 733, I 1 I . Kanncnberg, E , Blumc. A , Gcckclrr. K. and Poralla, K. (1985). Binchirnir.a et Biophysicu Acra 814. 170. Kanncnhcrg. E.. Blume, A . . McElharicy, R . N . and Poralla, K. (1986). Chemistryand Physics of />i[iid.s9 , 145. Klcinig, H. (1975). Europtun Journal of Biochemistry 57, 301. Kogan, L. M., K o z l ~ v a 1. , V.. Filipova, T. M . , Obolnikova, E. A . , Orekh, G T., Skriabina. S. V . and Kazantsev, Y.Y . (1991j. Khimiko-FarmatsevticheskiiZhurnal SSSR 25, 83. Krajcwski-Bcrtrand, M.-A. (1991). Thesis: Universite Louis Pasteur, Strasbourg. Kushwaha, S. C. and Kates, M. (1976). Cunudian Journul of Riochemistry 54, 816. Langworthv, 1 ,A , . Mayherry, W. R . and Smith, P . F. (1976). Riochimica er Riophysica Acra

.

.

431, 550.

Llopiz. P , Neunlist. S. and Rohmer M. (1992). Riochemical Journal 237, 159. Lynen, F . . Eggercr, H . , Henning, U and Kessel, I. (1958a). Angewandte Chemie 70, 739. Lynen, F., Henning, U . , Bublitz, C . ,Sorbo, €3. and Kroeplir-Rueff, L. (1958b). Riochemische Zeirung 330, 260. Lyncn, F.. Agranofl, €3. W., Eggcrer, H . , Hcnning, U . and Moslein, E. M. (1959). Angewundre C‘hemie 71, 657. Michcl, Ci. P. F . , Neuss, B . , Tappe, C. H . and Raratti, J . (1992). Archives of Microbiology 157, 116. Ncs, W .R. and McKean. M . L. (1977). In “Biochemistry of Steroids and Other Isoprenoids”, pp. 147-324. University Park Press, Baltimore

272

ti \ A H M

Al

Neumann, S . and Simon, H . (1986). Hiologische Chemie Hoppe-Seyler 367,723. Neunlist. S. and Rohmer, M . (1985a). Journul of Generul Microbiology 131, 1363. Neunlist, S. and Rohmer, M. (1985b). Biochemical Juurnal23l. 635. Neunlist. S. and Rohmer, M . ( 1 9 8 5 ~ ) .Biochemical Journal 228, 769. Neunlist. S. and Rohmer, M . (1988). Journal of the Chemical Society. Chemical Communicution 830. Neunlist, S., Hoist, 0. and Kohmer, M. (1985).European Journuluf Biochemistry 147, 561. Neunlist, S., Bisseret, P. and Rohmer, M . (1988). European Journalof Biochemistry 171.245. Ochs, D . , Tappe, C . H . , Gartner, P., Kellner, R. and Poralla, K (1990). European Journul of Biochemistry 194, 75. Ochs, D . , Kaletta, C . , Entian. K - D . , Beck-Sickinger, A and Poralla. K. (1992). Journal of Hucreriology 174, 298. Oulmouden, A . and Karst, F. (1990). Gene 88, 253. Ourisson, G. and Albrecht, P. (1992) Accounts of Chemicul Reseurch 25, 398. Ourisson, G . . Kohmer, M. and Poralla, K. (1987).Annual Review of Microbiology 41, 301. Pandian, S., Saengchjan, S. and Kaman, 1 . S. (1981). Biochemicuf Journal 1%. 675. Pciseler, €3. and Kohmer, M. (1991). Journal of the Chemical Society, Perkin 'liunsucrions I , 2449. Peiseler, B. and Rohmer, M. (IWZ). Journal of Chemical Research (S) 298, (M) 27.53. Poralla. K. (1982). FEMS Microhiological Letters 13, 131. Poralla, K.. Kannenberg, E. and Blume, A . (1980). FEES Letters 113, 107. Poralla, K.,HArtner, T. and Kannenberg, E. (1984). FEMS Microbiological Lxrteru 23. 253. Kaederstorft, D . and Rohmer, M. (1988). Eiochimicu el Biophvsicu Acta 960,190. Kaman. 7'.S., Sharma, B . V. S . , Jayaraman, J . and Kainasarma, 1. (1975). Archives o f Biochemistry und Biophysics 110, 75. Renoux. J . M. and Kohmer, M . (1985). European Journul of Biochemistry 154, 405. Rogers, P. L., Lee, K. J . , Skotnicki, M. L. and Tribe, D. E. (1982). Advances in Biochemical Engineering 23, 37. Rohmer, M. and Ourisson, G. (I476a). Tetrahedron Leffers, 3633. Rohmer, M . and Ourisson, G. (l976b). Tetrahedron Letrer.s, 363'7. Rohmer. M. and Ourisson, G. (lY7hc).Tetrahedron Letters. 3641. Rohmtx. M. and Ourisson, G . (19116). Journal of Chemical Reseurch (5') 356, (M)3037. Rohmer, M . , Bouvier, P. and Ourisson, C ; . (1979).Proceedings o f t h e Narionul Acudemy C J f Sciences of the United Stares of America 76, U47. Kohmer, M . , Anding, C. and Ourisson, G . (1980a). European Journul of Biochemlstry 112, 541.

Rohmer, M . , Bouvier, P., Berger, Y. and Ourisson, G . (tY80b). European lournul o f Eiochemistry 112. 557 Rohmer, M., Bouvier-Nave, P . and Ourisson, G . (1984). Journal of General Microbiulogy 130, 1137. Rohmer, M.,Sutter, B. and Sahm, H . (1989). Journal of the Chemicul Society, Chemicul Communications 19, 1471. Schulenberg-Schell, H., Neuss, B. and Sahm, H.(1989). Analytical Biochemistry 181, 120. Seckler, B . and Poralla, K. (19%). Riochimica er Biophysica Acra 11111. 356. Sherman, M. M., Petersen, L. A . and Poulter, C . I). (1989). /ournulofBacteri~logy171,3619. Shigeri. Y . , Nishino, T., Yumoto, N. and Tokushige, M . (1991). Agriculntrul and Biological Chemistry 55, 589. Shin-ya, K., Furihata. K., Hayakawa, Y. and Seto, H . 11090). Tetrahedron Letters 31. ho2.5. Shiomi, K., linuma, H . , Naganawa, H., Isshiki, K., Takeuchi, -1'. and Umezawa, H. (1987). Journal of Antibiotics XL, 1740. Simonin, P . , Jurgens. U J . and Rohmer, M. (1992). Tetrahedron Letters 33, 3629. Sipe, J. D.and Holmlund, C. E. (1972). Biochimica et Biuphysica A d a 280, 145. Stampf, P.. Herrmann, 11,. Bisseret, P . and Rohmer, M (1991). Tetrahedron Leners47,7081. Suzue. G . , Orihara, K.. Morishima, H. and Tanaka, S. (1964). Radioisotopes 13, 300. Swings, J . and DeLey, J (1977).Bacteriological Review 41 I . ~

HOPANOIDS IN BACTERIA

273

Tahara, Y ., Yamashita, T.,Kondo, M. and Yamada, Y. (1988a). Agricultural and Biological Chemistrv 52. 3189.

Tahara, Y.: Yuhara, H . and Yamada, Y. (1988b). Agricultural and Biological Chemistry 52, h07.

Taylor, R. F. and Davies, B. H. (1981). Canadian Journal of Biochemistry 60, 675. Thompson, G. A., Bambemy, R. J . and Nozawa, Y. (1971). Biochemistry 20,4441. Thorne, K J . I . (1973). Biochemical Journal 135, 567. Wakil, S. J. (1989). Biochemistry 28, 4523. Zhou, D . and White, K. H. (1991). Biochemical Journal 273, 627. Zundel. M. and Rohmer, M. (1985). European Journal of Biochemistry 150, 23.

This Page Intentionally Left Blank

Ethylene Production by Micro-organisms H. FUKUDA”, T. OGAWA“ and S. TANASEb a

Department of Applied Microbial Technology, Kumamoto Institute of Technology, Ikeda 422-1, Kumamoto 860, Japan, and Department of Biochemistry, Kumamoto University School of Medicine, Honjo 2-2-1, Kumamoto 860, Japan

I . Introduction

.

.

.

.

.

.

.

.

11. Productionofethylenebymicro-organisms .

. .

. .

. .

. .

. .

. .

. .

Ill. Biosynthetic pathways to ethylene in micro-organisms and higher plants . . A . The 2-keto-4-methylthiobutyric acid pathway in micro-organisms . . B. The 2-oxoglutarate pathway in micro-organisms . . . . . . C. The I-aminocyclopropane-I-carboxylic acid pathway in higher plants . IV. Mechanisms for formation of ethylene by Pseudomonussyringue . . . V. Molecular cloning and expression of the gene for the ethylene-forming . . . . . . . . . . enzyme of Pseudomonussyringue VI . Comparison of the structure of the ethylene-forming enzyme from Pseudomonussyringuewiththatofrelatedenzymes . . . . . . . . . . . . . . . . . . . VII. Concluding remarks References . . . . . . . . . . . . . . . .

275 277 281 282 284 287 288 292 295 302 303

1. Introduction

Gaseous unsaturated hydrocarbons, such as ethylene, propylene and butene, as well as gaseous saturated hydrocarbons, such as methane, ethane, propane and butane, are produced from fossil sources such as natural gas and crude oil. Available fossil sources are, however, limited. It has been said that the earth’s supply of crude oil will be exhausted in about 30 years. Although this time-frame may be incorrect, supplies of crude oil will eventually be exhausted and, thus, gaseous hydrocarbons from fossil resources will also become unavailable. “If only we were able to produce crude oil or natural gas artificially” sighs everyone from the politician to the driver. In fact, petroleum-producing plants, such as species in the families Euphorbiaceae (Nielsen et al., 1977), Pittosporaceae A D L A k C t S l N MICROBIAI.PHYSIOLOGY.VOL 35 ISBN ~ 1 L 4 1 i : n 3 . C 2

CopynghrO 1993, by Academic Press Limited All nghts of reproduction in any form reserved

276

n

FUKUDA F T AI.

(Nemethy and Calvin, 1982) and Asclepidaceae (Adams, 1984), have already been investigated in earnest with this possibility in mind. Ingredients of crude oil or natural gas might be produced economically if they could be produced by utilizing the action of micro-organisms. Ethylene, which is the simplest of unsaturated hydrocarbons and the most important starting material in petroleum chemistry, is also a plant hormone that is involved in a number of physiological processes, such as fruit ripening and plant senescence (McKeon and Yang, 1987), and it has been known since 1934 that ethylene is produced by plants themselves (Gane, 1934). Biale was the first person to show, in 1940, that microorganisms also produce ethylene. However, most available reports concern research carried out in relation to the physiological activity of ethylene in plants, although no reports can be found describing production of ethylene by micro-organisms with a view to establishing a controlled fermentation. It has been reported that gaseous hydrocarbons, such as ethane, ethylene. and C4-C7 hydrocarbons, can be detected as by-products of methane fermentation (Davis and Squires, 1954; Hunt et a l . , 1980; Oremland, 1981; Gollakota and Jayalakshmi, 1983; Oremland and Marais 1983; Relay and Daniels, 1987). However, the amounts of these hydrocarbons produced were very small and the fermentation required many days because it was anaerobic. Accordingly, at the beginning of our research directed towards efficient production of ethylene by micro-organisms, we set as our target the eventual use of raw materials that are easily available, are regarded generally as useless, and are constantly produced, for example faeces and urine of cows or pigs, and excess sludge. We hoped to grow cultures aerobically so as to shorten cultivation time, and we wanted to identify micro-organisms capable of producing ethylene efficiently in high yield (Fukuda, 1984). Eighty genera, 150 species and 178 strains of microorganisms with clear taxonomic identifications were tested (Fukuda et al. , 1984). Each of these strains was cultured aerobically in a liquid medium after which a portion of the culture was transferred to a test-tube, which was tightly sealed with a rubber stopper and shaken for several hours, and the gas formed in the gaseous phase analysed by gas chromatography. Any strain that produced ethylene at a rate of not less than 0.1 nl (mi liquid culture broth)-' h-' was defined as an ethylene-producing strain. As the result of this screening, we found that 30% of the tested strains were indeed producers of ethylene. Among them, 62% were moulds, 20°/" were yeasts, 21% were bacteria and 6% were actinomycetes. Since there are many ethylene-producing microbes in nature, we are attempting to clarify the ethylene biosynthetic pathways exploited by micro-organisms. Details of the biosynthetic pathway to ethylene in micro-organisms should give us a reasonable basis for breeding optimal ethylene-producing strains

FTHYI.EN€ I’Ri)l)U(TTION BY MICRO-ORGANISMS

277

and should also contribute to more effective control of the fermentation process. In this review, we summarize the biosynthetic pathways to ethylene found in micro-organisms and compare them with results from higher plants. We discuss characteristics of the reaction catalysed by the ethylene-forming enzyme, which IS t h e final step in biosynthesis of ethylene, as well as the molecular cloning of the gene for the enzyme from Pseudomonas syringae pv. phawoliccda PK2

11. Production of Ethylene by Micro-organisms

Since 1940 (Riale, 1940. Miller et a f . . 1940) there have been many reports of production of ethylene by aerobic heterotrophs. Table 1 lists microorganisms that have been reported in the literature to be ethylene producers. I t should be noted first that ethylenogenic micro-organisms are overwhelmingly bacteria and moulds and, secondly, there are many pathogenic microbes among these micro-organisms. Freebairn and Buddenhagen ( I 964) reported that a pathogenic bacterium, Pseudomonas sofanacearum, produced ethylene using peptone. glucuronic acid, glutamic acid or fumaric acid as the substrate. Theirs was the first report of an axenic bacterial ethvlene-production system. Subsequently, Primrose (1976a,b, 1977) and Primrose and Dilwnrth (1976) showed that many soil bacteria have an ethylene-producing ability added 10 which they demonstrated that Eschenchia cofi could produce ethylene using methionine as the substrate. Ince and Knowles (1985,1986) developed an ethylene-forming system using cell-free extracts from E. coli. Efficient production of ethylene by Ps. syringar pv phaseoficola, a pathogen specific for Pueraria fnbata (Willd) Ohwi (common name “Kudzu“) (Goto et al.. 1985; Golo and Hyodo, 1987) and P.Y syringac pv. gfycinea which causes halo blight in soya bean plants (Sato e l a l . , 1987) has been reported. Recently, Shipston and Bunch (1989) examined the physiology of the catabolism of L-methionine in an ethyleneproducing strain o f E. cofi, and Mansouri and Bunch (1989) studied the ethylenogenic capabilities of a number of selected bacteria during growth in media supplemented with 1.-methionine and 2-keto-4-methylthiobutyric acid (KMRA). Nagahama et al. (1991a.b) reported the construction of an ethylene-forming system in v i m and purification of the ethylene-forming enzyme from Ps. syrrngae pv. phaseoficola PK2 (see p. 288). Hahm et a f . (1992) reported that oxygen tension affected production of ethylene by Ps. syringac. According to their report. production of ethylene was restricted at dissolved-oxygen tensions below 4 ppm and was nearly constant over the broad range of 4.5 -8.0 ppm a t all dilutions. They also found that the

278

I I l+IIKIJI>A I. I A /

TABL.E I

Micro-urganisins reported to po\resr ethyleiiogrriic dclivity

Bacteria

Acinetohacter calcoaceticus" Aeromonas hydrophiluh Arthrcibacter sp. Bacillus mycoides" Bacillus subti1i.s' Rrevihacterium linens' Chromohactrriirm violaceurn' C'itrohacrer sp ~'orynehucterrumuyuuticum" Enlerohacter oerogmes' Enterohucter cloaceaeh Erwinia herhicoluh Escherichiu coli' Klehsiella pneumoniaeh .'I

Micrococcus /uteuS' Pseudomonas firiores~.ens' P.seudomonas indigofivu" Pseudomonus syringue pv . glycinrd Pseudomonus syringur pv. mori' Pcc~udomona.ssyringar pv. pha.sri~~Icold P.seud~monussolanuc~earitrd Hhizobium irifdit' Serrutia Iiyuefaciens" Serrutia murcrsens' Staphylococcus uureus' Throhacillru n o velluv" Throhucillus frrroxrduns" Xunlhomonas cumpemis'

Yeasts

cryplfJClJCC1I.S albirlush C'ryptococcus /uirrentiiJ

Moulds Agarius hisporu9 Alternuria soluni" Asc-orhyra imperfecfi' Aspergillus rundidid Aspergilh clavutus' Aspergilliu fiavus' Aspergillus itsti4.s' Aspergillus variecoki# Botrytis spertuhilis' Cephalosporium grumineumh Chaetomium chlumuloides' Chaetomium glohosum' C'oriolus hirsutus"' C'oriolus ver.siculor'" Duedolea dickirtsii"' Drmatium pullrtluns' Fistulina heputicu"' Flammulina veltipes"' Fomitopsis pinrcolu"' Fusarium oxysporum f . sp. tulipae" Gloeriphyllum suepiarium'" Gloeophyllum truheum"' Hirshioporus uhietinusm Hymenochuete tahacina"

Iypex lucteus"' I .aetiporw srrlphureuv"' Ixnzites betulinu"' I .entinus Iepideus"' Mucor hiemolis" Myrothecium roridum' Neurospnrc~crussu' P c n i c i h m corylophiIum' PeniciIIium cyclopium' PeniciIIium digiratum' PeniciIIitrm Iuteumk PeniciIIium patulum' Phanerochaete chry.sosporiuni"' Phaeolus sch weinitzii"' Pholiotu adiposu"' Phycomyces nitrns' Phycoporus roccineus"' Schizophyllum commune' Sclerotiniu laxa' .Scoiii~luriii~i.subrwrcuulisk Thumnrdium ekegun:;' Thiela viu ulutuli Tyromyces pulustris"'

279

ETtiYIENE PRODUCTION BY MICRO-ORGANISMS

TABLE 1. Continued Actinornycetes Streptomyces sp .

'

Algae Codium l a t u d Padina a r b o r e s c e d

Porphyra tenerd'

References: " Billington el al. (1979), Primrose (1976b). ' Mansouri and Bunch (1989). Fukuda et al. (1984). Primrose (1976a), Sato er al. (1987), * Nagahama et a/. (1992). Fukuda el al. (1989b), ' Thomas and Spences (1977), Turner et a/. (1975), llag and Curtis (1968). Fukuda and Ogawa (1991). Tanaka et a/. (1986). " Hottiger and Boller (1991). " Lynch and Harper (1974) and P Watanabe and Kondo (1976).

'

'

maximum specific rate of ethylene production within the range of optimal dissolved-oxygen tensions (4.5-8 ppm) in a chemostat increased linearly with increasing dilution rate. These observations indicate that production of ethylene by Ps. syringae is typically growth associated. Apart from reports on Saccharornyces cerevisiae (Thomas and Spences, 1977), Schizosaccharornyces octosporus (Fukuda et al., 1984b), Cryptococcus albidus (Fukuda et al., 1984b, 1989b,c; Ogawa et al., 1990) and Cr. laurentii (Fukuda et a l . , 1984b), there are no other reports of ethylene production by yeasts. Production of ethylene by moulds was first reported in 1940 from experiments with Penicilliurn digitaturn, the green mould of citrus fruit (Biale, 1940; Miller et al., 1940), after which similar results were obtained with Blastornyces derrnatitidis (Nickerson, 1948). Ilag and Curtis (1968) reported that 58 of 228 species of moulds examined produced ethylene, and they concluded that the hydrocarbon is a common metabolic product of moulds. Chalutz and his colleagues observed activation of ethylene production under phosphate-limiting conditions in shaken cultures of P . digitatum (Chalutz et al., 1977,1978; Chalutz and Lieberman, 1978; Mattoo el al., 1979). They also followed production of ethylene by P . digitatum when the mould was cultivated on citrus-peel media (Chalutz et al., 1983). They optimized conditions for production of ethylene, and the rate of production was increased to approximately 60 nl (g fresh weight of peel)-' h-'. In 1979, Lieberman summarized the available data on production of ethylene by P . digitaturn and other ethylenogenic microbes. Kutsuki and Gold (1982) found that the assay of ethylene produced by micro-organisms is suitable for monitoring their ligninolytic activity because there is a positive correlation between production of hydroxyl radicals, as determined by production of ethylene from KMBA, and production of 14C02by ligninolytic cultures of Phanerochaete chrysosporium. Tanaka et al. (1986) also

280

I{ FUKIIDA t7 A /

found a good correlation between production of ethylene from KMBA and degradation of dimeric model-lignin compounds by wood-dwelling fungi. Fukuda and Ogawa (1991) examined the cellulase activity of P. digitaturn I F 0 9372 and investigated production of ethylene from homogenates of recycled paper and Unshiu fruit peel. Recovery of carbon as ethylene from carbon sources that consisted of 10 g recycled paper I-' and 40 g Unshiu peel I-' were I . 1 0 and 0.48%, respectively. Fukuda et al. (1988) attempted to control the amount of carbon consumed by growth and respiration of P. digitaturn I F 0 9372. In order to achieve this objective, they tried to obtain mutants that formed smaller colonies than the parent on agar plates. A methionine-requiring auxotroph, P. digitaturn I F 0 9372-1040 (Met-), was induced by treatment of the wild-type strain I F 0 9372 with N-methylN'-nitro-N-nitrosoguanidine.Forty-six growth-suppressed mutants, which formed smaller colonies on agar plates, were isolated from the parent strain. Penicilliurn digitaturn I F 0 9372-1040-S-6 (Met-) was the greatest producer of ethylene among the strains tested. Fukuda et al. (1988) calculated the carbon-recovery data under different culture conditions for the three strains of P. digitaturn ( I F 0 9372, 1040 (Met-) and S-6 (Met-)) of interest. The rate of conversion of carbon in glucose to carbon in ethylene by S-6 (Met-) was four-fold and 2.3-fold higher than that by I F 0 9372 and 1040 (Met-), respectively. If the S-6 (Met-) strain was fed batchwise with glucose in culture, the extent of conversion of carbon in glucose to carbon in ethylene increased to 2.1%).However, if we compared the conversion in terms of ethylene produced with that during methane fermentation, the carbon yield of ethylene produced should increase to 10 times more than the value found using S-6 (Met-). IIag and Curtis (1968) examined 20 unidentified strains of streptomycetes that had been routinely isolated from soil, and demonstrated the presence of ethylene in the atmosphere that surrounded some of the cultures. However, the presence of ethylene was verified for only one sample. Watanabe and Kondo (1976) reported that two marine algae, namely Codium laturn (a green alga) and Porphyra tenera (a red alga), evolved a significant amount of ethylene when indole-3-acetic acid was applied exogenously, while Padina arborescens (a brown alga) evolved only a little. Fujii et al. (1985) developed a screening system for isolation of gaseous olefin-producing microbes using olefin-utilizing microbes that belonged to the genus Mycobacterium, in combination with a colour reaction in which molybdenum reacts with olefins. Several excellent ethyleneproducing microbes were identified from among 296 isolates by use of their system.

28 1

ETHYLENE PRODUrnION B Y MICRO-ORGANISMS

Precursor (1) H3C-S

Substrate for ethylene-forming enzyme

-CH-COOH+

0 2

NH2

2-Keto-4-methylthiobutyric acid

L-Methionine

(2)

CH -COOH

HOOC

7

CO-COOH

H3C-S

I

Product

-+

HOOC-

Ethylene

- C O - C O O H I

I

o*

NH2 L-Glutarnic acid

2-Oxoglutarate

Ethylene

L-Methionine

1-Aminocyclopropane-1-carboxylic acid

Ethylene

FIG. 1. Biosynthetic pathways t o ethylene in micro-organisms a n d higher plants.

111. Biosynthetic Pathways to Ethylene in Micro-organisms and Higher

Plants Figure 1 shows the typical biosynthetic pathways to ethylene in microbes; it also includes the biosynthetic pathway to ethylene in higher plants for comparison. Micro-organisms and higher plants synthesize ethylene by different biochemical pathways. In micro-organisms, there are two biosynthetic pathways to ethylene. In one pathway, ethylene is produced via KMBA, a transaminated derivative of methionine, as, for example, in E. coli (Ince and Knowles, 1986) and Cr. albidus (Fukuda et al., 1989b,c; Ogawa et al., 1990), and probably also in most ethylenogenic microorganisms. In the other pathway, ethylene is produced via 2-oxoglutarate as, for example, in P. digitaturn (Chou and Yang, 1973; Fukuda el al., 1986,1989a) and in Ps. syringae (Goto et al., 1985; Goto and Hyodo, 1987; Nagahama et al., 1991a,b, 1992; Fukuda et al., 1992a,b). Nagahama et al. (1992) classified 229 strains of ethylene-producing bacteria. Two hundred and twenty-five methionine-dependent strains were identified, while the only 2-oxoglutarate-dependent strain was Ps. syringae pv. phuseolicola PK2. Three strains of chemolithotrophs had ethylene-forming capacity, while Thiobacillus novellus I F 0 12443 had a novel ethylene-forming system that was dependent upon addition of meat extract to the culture medium. This pathway has not yet been characterized. By contrast, in higher plants, ethylene is produced from S-adenosylmethionine via the pathway:

282

H. FUKIJDA ET AI..

S-adenosylmethionine.+l-aminocyclopropane-1-carboxylic acid + ethylene. I n this view, we shall discuss these three pathways for biosynthesis of ethylene, with emphasis on the ethylene-forming enzymes. A.

THE 2-KETO-4-METHYLTHlOBUTYRlC ACID PATHWAY IN MICRO-ORGANISMS

From a series of studies on the production of ethylene by E. coli, Primrose (1977) proposed that KMBA might be an intermediate in the production of ethylene by E. coli, and she suggested two possible mechanisms for its conversion to ethylene: (a) by the action of light and excreted flavin; (b) by an enzymic process, possible via a peroxidase. Billington et al. (1979) described methods for identifying KMBA and showed that this compound is a common metabolic product of micro-organisms such as E. coli, Pseudomonas phi, Bacillus mycoides, Acinetobacter calcoaceticus, Aeromonas hydrophila, Rhizobium trifolii and Corynebacterium sp., when cells are grown in the presence of methionine. Ince and Knowles (1986) developed an ethylene-forming system using a cell-free extract of E. coli B SPAO. Their system consists of KMBA, NAD(P)H, Fe3+ chelated to EDTA and oxygen, and they discussed the possibility that production of ethylene by many bacteria might follow the route identified in E. coli. However, no ethylene-forming enzyme has yet been isolated. Fukuda et al. (1989b) selected Cr. albidus I F 0 0939 as a typical methionine-dependent ethylene-forming microbe and constructed a cell-free ethylene-forming system. This system resembled the cell-free system of Ince and Knowles (1986). The standard reaction mixture consisted of 3 mM KMBA, 0.5 mM NADH, 0.4 mM Fe” EDTA, and a cell-free extract in 20 mM potassium phosphate buffer (pH 6.5). Fukuda et al. (1989~)purified the ethyleneforming enzyme of Cr. albidus about 1600-fold from a cell-free extract, to a specific activity of 8250 pl (mg protein)-’ h-’, with a yield of 1.7%. The relative mass of the ethylene-forming enzyme of Cr. albidus was estimated to be 56,000 by gel filtration and 62,000 by sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDSPAGE). These results showed that the enzyme from Cr. albidus I F 0 0939 is a monomeric protein. The properties of the purified ethylene-forming enzyme were also studied by us (Fukudu et al., 1989c; Ogawa et al., 1990) and we concluded that this enzyme is an NADH:Fe3+-EDTA oxidoreductase, which reduces 2 mol of Fe3+ EDTA with 1 mol of NADH to give 2 mol of Fe” EDTA and 1 mol of NAD’ under anaerobic conditions at pH 6.5. From these results, Fukuda ei al. (1989~)and Ogawa et al. (1990) proposed a mechanism for formation of ethylene from methionine, as shown in Fig. 2. The first step involves an NADH-dependent reduction of Fe3+ to Fe2+, which is catalysed by an NADH:Fe”+-EDTA oxidoreductase. Oxidation of Fez+ EDTA by

ETHYI.ENE PRODUflION

NAD’

B Y MICRO-ORGANISMS

INAD+ 3NADH

NADH

283

Methionme Methionine aarninotransferase

oxidoreductase

~

OH-

02

FIG. 2. Pathway for formation of ethylene from 2-keto-4-methylthiobutyric acid (KMBA) by the ethylene-forming enzyme from Crypfococcus albidus I F 0 0939.

molecular oxygen (equation (1)) yields the superoxide radical anion (O*), which can undergo a dismutation reaction (equation (2)) to form hydrogen peroxide. Hydrogen peroxide, in turn, can react with Fez+ via the Fenton reaction (equation (3)), generating the hydroxyl radical (OH.).

+ +

+ +

Fe7+ O2 + Fe3+ 05 205 2H+ + H202 O2 Fe2+ H 2 0 2 + Fe3+ OH. OH-

+

+

+

(1) (2) (3)

It had previously been established that hydroxyl radicals can serve as oxidizing agents in oxidation of KMBA to ethylene (Diguiseppi and Fridovich, 1980; Tauber and Babior, 1980; Kutsuki and Gold, 1982). Although some enzymes that catalyse conversion of Fe3+to Fez+are known (Halliwell and Gutteridge, 1984), all such enzymes known are membrane bound and of the oxidase type. Since the purified ethylene-forming enzymes of Cr. albidus I F 0 0939 is neither membrane bound nor an oxidase, it seems to be a new type of enzyme. The NADH:Fe3+-EDTAoxidoreductase may act as an Fe3+-Fe2+recycling enzyme in vivo. It should be noted that the production of ethylene requires the presence of a specific transaminase that catalyses formation of KMBA from rnethionine. One might ask here whether or not an NADH:Fe”-EDTA oxidoreductase is present in microbes that do not produce ethylene even after addition of methionine to the medium. Ogawa et al. (1990) measured the activities both of transaminases and of NADH:Fe3+-EDTAoxidoreductasesin bacteria unable to produce ethylene. All strains tested had transaminase activity; while some had NADH:Fe”-EDTA oxidoreductase activity others did not. The former type may be negative for methionine-uptake activity. It seems that formation of ethylene by micro-organisms through a methionine-dependent

284

II FUKUDA E7 A / .

ethylene-forming system, as found in most ethylenogenic micro-organisms, may occur through the route identified in Cr. afbidus,as suggested by Ince and Knowles (1986). B. THE

2-OXOGLUTARATE PATHWAY

IN MICRO-ORGANISMS

Micro-organisms which have been reported as exploiting the 2-oxoglutarate pathway are P . digitatum, P . cyclopium (Fukuda and Ogawa, 1991), Ps. syringae, Chaetomium globosum (Fukuda and Ogawa, 1991), Phycomyces nitens (Fukuda and Ogawa, 1991) and Fusarium oxysporum (Hottiger and Boller, 1991). Penicillium digitatum produces large amounts of ethylene. Chou and Yang (1973) elegantly demonstrated, using ''C-labelling experiments that the carbon atoms in ethylene are derived from C-3 and C-4 of glutamate or the corresponding 2-oxoglutarate. However, they failed to establish whether 2-oxoglutarate or glutarate was the direct precursor of ethylene. In 1986, Fukuda et al. reported partial purification of an ethyleneforming enzyme from P. digitatum I F 0 9372 by ion-exchange chromatography on DEAE-Sepharose C L d B . They succeeded in constructing an in vitro system using the partially purified enzyme from P. digitatum I F 0 9372. They found that the immediate precursor of ethylene was 2oxoglutarate. The in vitro system required L-arginine, Fe2+ions maintained in the reduced form (by reducing agents such as dithiothreitol) and oxygen as essential factors. Since the time-course of formation of ethylene by the cell-free system (in vitro) and that by the living cells (in vivo) of the fungus were similar, Fukuda et al. (1986) concluded that the analogue of the cellfree ethylene-forming system operates in living cells. The in vitro system consisted of 1 mM 2-oxoglutarate, 0.5 mM L-arginine, 0.075 mM ferrous sulphate, 1 mM dithiothreitol, 20 mM HEPES buffer (pH 8.0) and the partially purified enzyme fraction, and incubation was performed at 25°C in a sealed test-tube. The ethylene-forming enzyme was subsequently purified from P . digitatum I F 0 9372 (Fukuda el al., 1989a). The preparation of the purified enzyme was shown to be homogenous by polyacrylamide disc-gel electrophoresis and the relative mass of the enzyme was estimated to be 42,00(00 by gel filtration and SDS-PAGE, suggesting that the enzymic activity is associated with a single monomeric protein. The specific substrate and cosubstrate for the purified enzyme were 2-oxoglutarate and L-arginine, respectively. The enzyme was strongly inhibited by the chelating reagents EDTA and Tiron. This result suggests that some kind of Fe2+ complex might be required for catalytic activity. The divalent transition metal ions Co2+,Cu2+and Mn2+inhibited ethylene-forming activity to a certain extent, probably by competing with Fe2+ for formation of such a complex. Some sulphydryl groups in the enzyme may play an important role in the catalytic

285

ETHYLENE PRODUCTION BY MICRO-ORGANISMS

activity since the enzyme was inhibited by 5,5'-dithio-bis(2-nitrobenzoate). Superoxide and hydrogen peroxide did not seem to be involved in the enzymic reaction since superoxide dismutase and catalase failed to suppress the reaction. Involvement of hydroxyl radicals also seems unlikely since mannitol and sodium benzoate had minimal inhibitory effects. These results imply that activation of molecular oxygen occurs by direct co-ordination of Fe2+,which may be loosely bound to the enzyme. Other inhibitors of the activities of free radicals, such as propyl gallate, diazibicyclo-octane and hydroquinone, suppressed the reaction considerably. These effects may not necessarily, however, be attributable to the radical-scavenging properties

I

fc o o - p q

(COO

/ -

.' ! '

H H

t4

O+

'

Arginine

-(

L,-',-

OH I

E6

i2

E7

(CH2COO-)2.H-Gua' ,

(P5C)

FIG. 3. A dual-circuit mechanism proposed for the simultaneous formation of ethylene and succinate from 2-oxoglutarate by the enzymes from Penicilliurn digitaturn and Pseudomonas syringae pv. phaseolicola PK2. L1 and L2 are ligands on the enzyme and sites I and I1 are binding sites on the enzyme. For simplicity, co-ordinated water molecules are not shown. All of the iron ions are assumed to be hexa co-ordinated. Gua+ indicates a protonated guanidine group; P5C, L-AIpyrroline-5-carboxylate.

286

I I I~UKLIIIA ET A1

of these agents. The compounds may instead act as chelating agents for Fe2+ and, thus, suppress the reaction in a similar manner to EDTA. A model for the proposed intermediates, which consists of the ethyleneforming enzyme, 2-oxoglutarate, L-arginine, an Fe2+ ion and oxygen, is shown in constructs El and E2 of Fig. 3. Under optimum conditions, the concentrations of L-arginine, Fe2+and 2-oxoglutarate are too low for these chemical species to interact with each other in bulk solution. They must be concentrated within a certain domain of the enzyme in some special way. Thus, the active site of the enzyme can be considered hypothetically to be an Fe2+ complex bound to the enzyme through appropriate ligand atoms L1 and L2 (see also Section VI). The complex also involves reaction of L-arginine and 2-oxoglutarate as a Schiff-base structure to form an intermediate. This intermediate reacts with oxygen to form an unstable Fe4+ complex of the peroxo type (see construct E2 in Fig. 3), the 2oxoglutarate moiety of which is decomposed to ethylene and three molecules of carbon dioxide. The ethylene-forming enzyme of Ps. syringae pv. phaseoficolu PK2 may also form the same type of intermediate (see Section IV). Ogawa et a f . (1992) reported that the relationship between the rate of formation of ethylene and the concentration of the purified ethylene-forming enzyme of P. digitaturn I F 0 9372 was not linear. When catalase and bovine serum albumin were added to the reaction mixture, the rate of formation of ethylene was directly proportional to the concentration of enzyme. The non-linearity of the reaction, in the absence of these additives, is probably due to the hydroxyl radical ions produced by the Fenton reaction, which occurs in the reaction mixture when Fe2+and oxygen are present together under reducing conditions. To investigate whether or not there are micro-organisms other than P. digitaturn that produce ethylene from 2-oxoglutarate, 2-oxoglutaratedependent ethylene-forming fungi were screened and one strain of P. cycfopiurn KIT 0229, which was newly isolated from soil, was found to exhibit the highest 2-oxoglutarate-dependentethylene-forming activity. The ethylene-forming enzyme from P. cyclopiurn KIT 0229 was partially purified and its properties were found to be similar to those of the enzyme from P. digitaturn (Fukuda and Ogawa, 1991). Goto et a f . (1985) and Goto and Hyodo (1987) studied the production of ethylene from many strains of plant-pathogenic bacteria. A significant amount of ethylene was produced by Ps. sofunacearum, Ps. syringae pv. phaseoficola and Erwinia rhapontici. Among these strains, the Kudzu strain of Ps. syringae pv. phuseolicola was selected and its ethylene-producing activity was studied in vivo. Goto and Hyodo (1987) constructed a cellfree ethylene-forming system, in which 2-oxoglutarate served as the substrate and Fe2+ions, dithiothreitol, histidine and oxygen were essential

ETHYLENE PHODUCTION BY MICRO-ORGANISMS

287

for formation of ethylene. In this bacterial system, L-arginine had no stimulatory effect on production of ethylene. However, Nagahama et al. (1991a) studied a reconstruction of the in vitro system from Ps. syringae pv. phaseolicola PK2 and found that L-arginine was an essential factor (see p. 288). Recently, Hottiger and Boller (1991) reported biosynthesis in vivo of ethylene by F. oxysporum f. sp. tulipae, a tulip pathogen, and they found that this strain has a similar enzyme system to that of P. digitaturn. They suggested that the ethylene-forming enzyme isolated by Fukuda et al. (1989~)is the natural enzyme in fungi. C. T H E 1-AMINOCYCLOPROPANE-1-CARROXYLIC ACID PATHWAY I N HIGHER

PLANTS

Ethylene is synthesized from methionine via S-adenosylmethionine and 1aminocyclopropane-1-carboxylicacid (ACC) in higher plants (Adams and Yang, 1979). The ethylene-forming enzyme that is responsible for oxidation of ACC to ethylene was not extensively characterized in vitro until 1989 (Kende, 1989) because the activity disappears completely when tissues are homogenized (Yang and Hoffman, 1984). A study of the products of the reaction in vivo revealed that the ethylene-forming enzyme catalyses the following reaction (Peiser el al., 1984). ACC

+

4 0 2 -+

C2H4

+ HCN + COZ + H2O

(4)

Earlier reports have shown that the activity of the ethylene-forming enzyme is dependent on membrane integrity (Guy and Kende, 1984; Mayne and Kende, 1986). Many artifactual in vitro systems have been reported because ACC is readily oxidized to ethylene by chemical oxidants. Such systems lack high affinity for ACC and display no stereospecificity for 2-ethylACC stereo-isomers (McKeon and Yang, 1984; Venis, 1984). Hoffman et al. (1982) described a test that permits a distinction to be made between artifactual and natural ACC-dependent ethylene-forming activity. Recently, Ververidis and John (1991) made the important discovery that authentic activity of the ethylene-forming enzyme from melon fruit can be fully recovered if it is extracted and assayed under conditions required for extraction of active flavanone-3-hydroxylase (EC 1.14.11.9). Their in vitro assay system contained 1 mM ACC, 30 mM sodium ascorbate and 0.1 mM ferrous sulphate in 0.1 M Tris-hydrochloric acid buffer (pH 7.2) prepared with 10% glycerol. There followed two reports on ACC oxidase from avocado fruit and apples (McGarvey and Christoffersen, 1992; Fernandez-Maculet and Yang, 1992). The solubilized enzyme systems resembled in vivo systems in that the enzymes had low K,,, values (17 PM)for the substrate ACC, they were stereospecific with respect to stereo-isomers of 2-ethyl-ACC for production

288

If

b’UKUI)A ET A l

of 1-butene, and they were inhibited by cobalt ions and a-aminoisobutyric acid. However, unfortunately, attempts at purification of ethylene-forming enzymes from higher plants have not yet been successful. Following preparation of this review, purification of ACC oxidase from ripening apple fruit was briefly reported by Dong et al. (1992b). Table 2 is a summary of the various in vitro systems with active ethyleneforming enzymes from micro-organisms and higher plants. There are many similarities between these in vitro systems, for example iron ions and oxygen are essential for all the systems, and all the systems need cosubstrates, although the cosubstrates differ between systems.

1V. Mechanisms for Formation of Ethylene by Pseudomonas syringae Pseudomonas syringae pv. phaseolicola PK2 (Kudzu strain), which in Japan causes halo blight of the viny weed Pueraria lobata (Willd) Ohwi (common name “Kudzu”), is known to produce large amounts of ethylene (Goto et al., 1985). Goto and Hyodo (1987) reported that this bacterium produced ethylene under aerobic conditions in the presence of 0.5 mM 2-oxoglutarate, 0.5 mM ferrous sulphate, 10 mM L-histidine and 5 mM dithiothreitol in 50 mM HEPES sodium hydroxide (pH 7.0). However, the ethylene-forming activity of this system was completely lost when the cell-free extract was dialysed against 10 mM potassium phosphate buffer (pH 7.0) for 24 hours at 4°C. Loss of activity during dialysis indicates that some factor(s) is essential for enzymic activity. Nagahama et al. (1991a) reported that, when the fraction of the cell-free extract with a relative mass of less than 10,OOO (called SupI) was added back to the enzyme fraction after gel filtration of the cell-free extract, enzymic activity increased to about four times that of the gel-filtered crude enzyme. The action of SupI could be reproduced by addition of L-arginine. The complete system for formation of ethylene by the enzyme from Ps. syringae under aerobic conditions in vitro required 0.25 mM 2-oxoglutarate, 0.2 mM ferrous sulphate, 2 mM dithiothreitol, 10 mM L-histidine and 0.2 mM L-arginine. The cosubstrate specificity was examined by replacing L-arginine or L-histidine with various analogues, but none of these was effective. The components of this system, with the exception of L-histidine, are similar to those of a system derived from the ethylene-producing, plant-pathogenic fungus P. digitaturn, which also produces ethylene in vitro in a reaction that is dependent on 2-oxoglutarate. Nagahama et al. (1991b) subsequently succeeded in purifying an ethyleneforming enzyme from a cell-free extract of Ps. syringue pv. phaseolicola PK2. It was purified about 2800-fold with an overall yield of 53%, and gave a single band of protein after SDS-PAGE. The purified enzyme had a

TABLE 2. Comparison of in vitro systems with ethylene-forming enzymes from microbes and higher plants Microbe or plant

Substrate

Cosubstrate

Metal ion

Oxygen

Other factor

Penicillium digitatum

2-Oxoglutarate

L- Arginine

Fe2+

+

Dithiothreitol

Pseudomonm syringae

2-Oxoglutarate

L- Arginine

Fe2+

+

L-Histidine

Cryptococcus albidus

2-Keto-4methylthiobutyric acid

NADH

Fe'++EDTA

+

None

I-Aminocyclopropane-1carboxylic acid

Ascorbate

Fez+

+

None

Cucumis melo (melon fruit)

+ indicates oxygen is essential.

290

II FUK1II)A El' A / . .

TABLE 3. Comparison of some properties of the ethylene-forming enzymes from Pseudomonas syringae and Penicillium digitatum Property

Pseudomonas syringae

Penicillium digitatum

36 42 5.9

42 42 5.9

7.G7.5 6.04.0

7.0-7.5 6.0-8.0

20-25 <30"

<30"

Molecular mass (kDa) Gel filtration SDS-PAGE

PI PH

Optimum Stability at 5°C Temperature ("C) Optimum Stability K, value (M) Fez+ L- Arginine 2-Oxoglutarate N-Terminal sequence

25

5.9 ' 10P 1.8 . 1 . 9 . lo-< M T N L Q T F E L P-

4 . 0 . 1w5 6.0 . 10P

3.8 . lo--s L A P P A P S N L G-

specific activity of 660 nmol ethylene (mg protein)-'min-'. 'The relative mass of the enzyme was approximately 36,000 by gel filtration and 42,000 by SDS-PAGE. These results indicate that the enzyme from Ps. syringae is a monomeric protein. Table 3 lists a comparison of some properties of the ethylene-forming enzymes from Ps. syringae and P . digifarum. The properties of the enzyme from P. digifarum are almost the same as those of the enzyme from Ps. syringae. There was no homology between the Nterminal amino acid-residue sequence of the two enzymes, even though they have similar properties. Recently, we found that purified ethylene-forming enzyme isolated from Ps. syringae pv. phaseolicola PK2 simultaneously catalysed two reactions, namely formation of ethylene and formation of succinate From 2-oxoglutarate, at a molar ratio of 2: 1 (Fukuda et al., 1992a). In the main reaction (the ethylene-forming reaction; see equation ( 5 ) ) , 2-oxoglutarate was dioxygenated to produce one molecule of ethylene and three molecules of carbon dioxide. In the subreaction (the succinate-forming and L-argininedecomposing reaction; equations (6) and (7)), both 2-oxoglutarate and Larginine were mono-oxygenated to yield succinate together with carbon dioxide and L-hydro-oxyarginine, respectively, the latter being further transformed to guanidine and ~-A'-pyrroline-5-carboxylate (P5C). We have proposed a dual-circuit mechanism for the entire reaction, in which binding of L-arginine and 2-oxoglutarate in a Schiff-base structure generates a common intermediate for the two reactions (see Fig. 3). CSHh05

+ 0,

-+

CZH4

+ 3CO2 + H20

(5)

FTHYLENE PRODUCTION BY MICRO-ORGANISMS

29 1

+ 4 0 2 -+ C4H604 + CO;! -+ CHsN3 + CSH7N02 + HzO

CSH605 40 C,HI4N,O,

(6) (7) Since formation of P5C and guanidine from L-arginine involves hydroxylation at the &carbon of L-arginine, the latter two reactions appear similar to the 2-oxoglutarate-dependentdi-oxygenase reaction (Hayaishi el ul., 1975). Therefore, equations (6) and (7) are considered to represent a succinate-forming reaction. Since the ratio of the amounts of ethylene, succinate and P5C (or guanidine) produced was 2:l: 1, the overall reaction for the ethylene-forming enzyme from Ps syringue can be represented by equations (5) X 2 (6) (7), which can be summarized as equation (8).

+

+

3CSH605

2

+

+ 302 + CnHI4N402 + 2C2H4 + 7 c o 2 + c4H(,o4 + 3 H 2 0 + CHsN3 + CSH7N02

(8)

Figure 3 shows the reaction mechanism for formation of ethylene in Ps. syringue. At the active site, the enzyme is assumed to act as a bidentate ligand (L1 and L2) and it forms a complex with Fe2+, as shown in E l . In Fig. 2, L1 and L2 may be histidine residues (Fukuda et ul., 1992a; see Section V). The Fe2+ ion is further co-ordinated to a tridentate Schiff's base of 2-oxoglutarate and L-arginine, whose terminal carboxylate and guanidido groups are trapped, respectively, by binding sites I and I1 on the enzyme. Then, E l reacts with dioxygen to form a peroxo complex, E2. Following this event, E2 decomposes irreversibly, with cleavage of the unstable oxygen-oxygen bond and simultaneous cleavage of two bonds in the 2-oxoglutarate moiety by separate mechanisms. Thus, one reaction involves simultaneous cleavages of C - 2 4 - 3 and C-4-C-5 bonds to generate ethylene, carbon dioxide and E3. The complex E3 comprises a strongly oxidizing Fe4+ ion with a strongly reducing oxalyl-like ligand. Thus, an intramolecular redox reaction readily occurs to generate E4. The sequence E l -+ E2 + E3 + E4 (+ E l ) constitutes a major catalytic cycle for formation of ethylene. The other mode of decomposition of E2 involves cleavage of C - 1 4 - 2 bond of the 2-oxoglutarate moiety, to generate the Fe4+ complex E5 that contains succinyl arginate as a ligand. The highly reactive Fe4+ion of E5 hydroxylates the &carbon of the arginine moiety of the ligand. The hydroxylated configuration in E6 is then spontaneously degraded to succinate, guanidine and P5C. The sequence E l + E2 -+ E5 + E6 + E7 + E4 (+ E l ) constitutes a minor catalylic cycle for formation of succinate. Complexes E4, E l and E2 are common to the two catalytic cycles, and it is chemically impossible to determine the specific cycle in which they are involved. The ratio of the activities of the two reactions is determined at the stage of decomposition of E2 since this process is essentially irreversible. Formation of ethylene is favoured when all of the bonds in the 0 - 0 - C - 2 4 - X - 4 - C - 5 configuration in E2 are stretched into a W-like

292

ti R J K U D A E 7 A / .

shape, which maximizes the overlap of orbitals that are produced during the decomposition process. By contrast, when the bonds are not stretched out, formation o f ethylene does not occur and succinate is formed from the 2-oxoglutarate moiety of E2, in a reaction that is controlled by the spatial alignment of binding site 1, that is to say, by the nature of the enzyme itself. V. Molecular Cloning and Expression of the Gene for the Ethylene-Forming Enzyme of Pseudomonas syringae

Important advances in the molecular biology of ethylene-forming enzymes of higher plants have been made recently. Hamilton et al. (1990) suggested that the product of pTOM13 (‘TOMETHYBR) is related to the ethyleneforming enzyme because enzyme activity in tomato fruit was greatly decreased by introduction of a pTOM13 antisense gene. The product of pTOMl3 has ethylene-forming enzyme activity when expressed in Sacch. cerevisiae (Hamilton et al., 1991) and in oocytes from Xenopus laevis (Spanu et al., 1991). Dong et al. (1992a) constructed two degenerate oligonucleotide primers, which corresponded to the conserved amino acid-residue sequences of ACENWGF and KFQAKEP and generated a full-length cDNA designated pAE12 (MAURRP) from apple fruit. The sequence of pAE12 is 1199 bp long and contains an open-reading frame that encodes 3 14 amino-acid residues. The nucleotide and deduced amino acid-residue sequences are highly homologous to those of the cDNA for ACC oxidase from the tomato. However, there were no reports of isolation of a gene that encodes an ethylene-forming enzyme in micro-organisms until 1992. Fukuda et al. (1992b) described the molecular cloning of the ethyleneforming enzyme of Ps. syringae. Caesium chloride-ethidium bromide density gradient centrifugation of indigenous plasmid DNA, which had been extracted from cells with alkali, showed the presence of cccDNA in cells of Ps. syringae. Fukuda et al. (1992b) constructed oligonucleotides (17-mer) as probes that were complementary to the possible coding sequences derived from the six-amino-acid sequence Met-Thr-Asn-LeuGln-Thr of the N-terminus of the enzyme. Southern-blot analysis revealed that mixtures of the oligonucleotide probes hybridized strongly with 2.5 kb Hind11 fragments, 9.5 kb EcoRI fragments and 23 kb BarnHI fragments. These results suggest that the indigenous plasmid DNA (named for pPSP1) encodes the gene for the ethylene-forming enzyme of the bacterium. Fragments of DNA obtained by digestion with Hind111 were separated by electrophoresis on a 0.7% agarose gel. A fragment of 2.5 kb was recovered and ligated with pUC19 that had been cleaved by H i n d I I . Escherichia coli

ETHYIXNE PRODUCTION BY MICRO-ORGANISMS

293

JM109 was transformed with the resultant plasmids, and transformants were selected as white colonies on a modified Luria-Bertani (LB) medium (composition (g I-'): peptone 10.0. Bacto yeast extract 5.0, sodium chloride 10.0) that contained ampicillin, 5-bromo-4-chloro-3-indolyl-~-~galactopyranoside and isopropyl-p-D-thiogalactoside. Clones of E. coli harbouring the gene for the ethylene-forming enzyme were screened by Southern hybridization of the alkaline-extracted plasmid DNAs using mixtures of oligonucleotide probes. Among the 40 colonies selected, two positive clones were obtained. Fukuda et al. (1992b) selected one of these, E. coli JM109 (pEFEOl), because its ethylene-forming activity was much higher than that of the other clone. The enzyme activity was 230 nl ethylene (ml culture medium)-' h-', while E. coli JM109, the original host, did not produce any ethylene. The activity of the enzyme was about one-fifth to one-tenth of that of Ps. syringae at the same density of cells. Antibody raised against the ethylene-forming enzyme reacted with the cell-free extract of E. coli JM109 cells that harboured pEFEOl, and the immunoreactive protein showed identical electrophoretic mobility with that of purified enzyme. The amount of enzyme protein as a percentage of the total protein in E. coli JM109 (pEFEO1) was estimated to be 0.3%. These results show that the gene for the ethylene-forming enzyme in Ps. syringae is encoded by the indigenous plasmid pPSP1. Fukuda et al. (1992b) constructed a series of subclones, designated pEFE02 through pEFE 10, by deletions from the 5' end of the plasmid pEFEOl. The nucleotide sequence of pEFE02, which contains a fragment of 1809 bp from Ps. syringae, was determined. The sequence pEFE02 contains a 1053 bp openreading frame that encodes a protein of 350 amino-acid residues (Mr 39,444). The calculated relative mass is similar to the values of 36,000 obtained by gel filtration and 42,000 obtained by SDS-PAGE (Nagahama et al., 1991b). The N-terminal amino acid-residue sequence deduced from the cloned DNA was identical with that of purified enzyme, as determined by Edman degradation up to the 29th residue (Nagahama et al., 1991). In the 5' noncoding region, Fukuda et al. (1992b) found consensus sequences of typical prokaryotic promoter regions, such as -35 and -10 regions, and a ShineDalgarno sequence. The ethylene-forming activity of E. coli JM109 (pEFEO1) was not affected by addition of isopropyl-P-D-thiogalactoside,and both clones that harboured pEFEO1 and those that harboured a plasmid with the reverse orientation of the gene had the same ethylene-forming activity. Therefore, the ethylene-forming enzyme activity of E. coli JM109 (pEFEO1) may be controlled by the promoter from Ps. syringae, and not by the lac promoter of pUC19. The sequence of pEFE02 contains an inverted repeat in the 3' non-coding region. The role of this inverted repeat is not clear, but it may function as a terminator (Freidman et al., 1987) of the transcription of mRNA from the gene coding for the ethylene-forming enzyme.

EFEPS TOMElHYBR

TOMETH1 LEEFEMR TOMGTOMA PETEFE MAURRP DINCARSR BNAEFEMR AVOAVOE3 HYSH6H

84 1411 1494 1585

TOME8 A2 BLYFL3DOX PETAN3 DAOCS-SC

56

47

47

57

61

59

64

61

102

161

161

169

156

140

146

165

-

65

90

46

143

181

133

-

13

138

94

91 1516

DAOCS-CA PECIPS STMIPNSA

IPNSSL

105

140

108

85 1031 1247 1671

I-THY1 FNF PRODUCTION BY MICRO-ORGANISMS

295

VI. Comparison of the Structure of the Ethylene-Forming Enzyme from Pseudomonas syringae with that of Related Enzymes

W e found partial homology betwen the ethylene-forming enzyme of Ps. syringae and counterpart enzymes from plants, ripening-related proteins, 2-oxoglutarate-dependentdioxygenases, and isopenicillin-N synthases. T h e amino acid-residue sequence is highly conserved among the ethyleneforming enzymes of plants, and middling scores are obtained between 2oxoglutarate-dependent oxygenases (Fig. 4). In t h e comparison of the ethylene-forming enzyme from Ps. syringae with ethylene-forming enzymes from plants, the homology score was low (15% sequence identity), but several clusters of invariant residues were found in the middle part of the FIG. 4. Similarity among ethylene-forming enzymes, ripening-related enzymes, 2oxoglutarate-dependent enzymes and isopenicillin-N synthases. Homology scores, considered with gaps that were inserted for optimization of sequence alignment, were calculated by the program HOMOGAPP of GENETYX (SDC Inc., Tokyo, Japan) using the 250PAM parameter (Schwartz and Dayhoff, 1979) and the method of Lipman and Pearson (1985). In the recent edition of GeneBank (release 76), EFEPS. TOMETHYBY, TOMETH1, TOMGTOMA, BNAEFEMR, TMOE8, BLYFL3DOX and PETAN3 are registered as PSEETH, LEETHYBR, LEETHl, LEGTOMA. BJEFEMR, LEES, HVFL3DOX and PHAN3, respectively. EFEPS indicates the ethylene-forming enzyme of Pseudomonas syringae pv. phaseolicola PK2; TOMETHYBR, the ethylene-forming enzyme related to the pTOM13 gene product of Lycopersicon esculentum (Holdsworth et a l . , 1987a); TOMETHl , the ETH 1 gene product involved in ethylene biosynthesis in Lycopersicon esculentum (Kock et al., 1991); LEEFEMR, the ethylene-forming enzyme of Lycoperisicon esculentum (Spanu et al., 1991); TOMGTOMA, the ethylene-forming enzyme related to the GTOME gene product of Lycopersicon esculentum (Holdsworth et a l . , 1087b); PETEFE, the ethylene-forming enzyme of Petunia hydrida L. (Wang and Woodson, 1992a); MAURKP, the ripening-related protein of Malus sylvestris (Dong et a l . , 1992a); DINCARSR, the senescence-related protein of Dianthus caryophyllus (Wang and Woodson, 1992b); BNAEFEMR, the ethylene-forming enzyme of Rrassica jucea (Pua et al., 1992); AVOAVOE3, the ripening-related protein of Persea americana (McGarvey et al., 1990); HYSH6H. the hyocyamine6-P-hydroxylase of Hyoscyamus niger (Matsuda et al. , 1991); TOME8, the fruit ripening-related E8 gene product of Lycopersicon esculentum (Deikman and Fischer, 1988); A2, the 2-oxoglutaratc-dependent dioxygenase-related A2 gene product of Zea mays (Messen et al., 1990); BLYFWDOX, the flavanone-3-Phydroxylase of Hordeum vulgare (Meldgaard, 1991); PETAN3, the flavanone-3-Phydroxylase of Petinia hybrida (Britsch et al., 1991); DAOCS-SC, the deacetoxycephalosporin-C synthetase of Streptomyces clavuligerus (Kovacevic et a l . , 1989); DAOCS-CA, the de-acetoxycephalosporin-Csynthetase of Cephalosporium acremonium (Samson et al. , 1087); PECIPS, the isopenicillin-N synthetase gene product of Penicillium chrysogenum 23XNL26Y-37-2 (Cam et al., 1986); STMIPNSA, the isopenicillin-N synthetase gene product of Streptomyces lipmanni (Weigel et a l . , 1988); IPNSSL, the isopenicillin-N synthetase ( ~ c b Cgene ) product of Nocardia lactamdurans (Conquc et al., 1091).

EFEPS TOMETHYBR TOMETHl LEEFEMR TOMGTOMA PETEFE MAURRP

-

51

-

51 51

1 1

52

I 1

51 37 35 1

5

4

DINCARSR BANAEFEMR AVOAVOE3

51

------------ ----- --

KLN---- -GDERANTME NLN---KLN---KVN- - - - -GVERAATME LVN---- -GEERAATLE MAN Ia FP I ID M E K L N N Y N - - G V E MEKNIKFPVVDOS K L I- - - - -GEE -- 4 D S F P V IN H E K L E - - - - - G Q E V P V G N D V P I I D - - - L Q Q - - - - - -H C E T H F V F P V I D - - - L Q G I D ED P I K D H T A P R I P V V D I S P F L D S S SQQQQ DRFSDAVPLISRH GIDGA-GIDDE-N .L..G.. PLFG--D L M P S A ~ V P D~I I sPLFG--.

31

51

-

HHL

N s NHGLP N H ~PI NHGFP NHGIP GHGIP

ADL I A D M T RL a v Is QM TTF D.DH.SA.DT V.RLS.KT.E VQ.LQDVVNE

51 57

54 51

79 108 111

91 91 51

62 61

52

109

52

88 108

32 52

108

52

108 108 108

52

52 58 55

52

EFEPS TOMETHYBR TOMETHI. LEEFEMR TOMGTOMA PETEFE MAURRP

E T -~- EPV T G

1

TOME8

EFEPS TOMETHYBR TOMETHl LEEFEMR TOMGTOMA PETEFE MAURRP DINCARSR BANAEFEMR AVOAVOE3

~

1

1

1 29

BLYFL3DOX PETAN3 C-DAOCS C1-IPNS C2-IPNS

Q

-

DINCARSR BANAEFEMR AVOAVOE3 HYSH6H A2

- -L

1 1

DM-LKSKGLENLERE~~J~DWESTFYLRHLPQSE L - M A S K - V E ~ ~ V D A N D M D W E S T F ~ I R H L P ~ S -

114 111

107 168

141 161 161

161

115

112 108

161 161

167

164

162

226 178 198 198 198 198 198 204 201 199 231 256 214 240 240 215 240 240

EFEP S TOMETHYBR TOMETH1 LEEFEMR TOMGTOMA PETEFE MAURRP

DINCARSR BANAEFEMR AVOAVOE3 HYSH6H TOME8 A2 BLYFL3DOX PETAN3 C-DAOCS C1-IPNS C2-IPNS EFEPS TOMETHYBR TOMETHl LEEFEMR TOMGTOMA PETEFE MAURRP DINCARSR BANAEFEMR AVOAVOE3 HYSH6H TOME8 A2 BLYFL3DOX PETAN3 C-DAOCS C1-IPNS C2-IPNS

P Q

~ I

fllrpw T T

PPMRHSIV

~~DII[FP

NLGDO

EVI?~GK

FKNADHQAVV---FKNADHQAVV----

Fig. 5 . Continued on p. 298.

278 225 245 245 245 245 245 251 248 246 285 303 321 289 289 2 63 282 282

EFEPS TOHETHYBR TOMETHl LEEEEMB TOMGTOMA PETEFE MAVRRP

DINCARSR BANAEEEUR AVOAVOE3 HYSH6H TOME8 A2 BLYFL3DOX PETAN3 C-DAOCS C1-IPNS C2-IPNS EFEPS TOMETHYBR TOMETHl

LEEFEWR

TOMGTOMA PETEEE MAURRP

DINCARSR BANAEFEMR AVOAVOE3

219 226 246 246 246 246 246 252 249 241 286 304 322 290 2 90 264 283 283 338 284 304 300 305 305 303 309 302 305

YGEHFTNMFMRCYPDRITTQRINKEURLAHL --STQVYPKEVFDDYMKLYAGLKFQAKEPRE --STQVYPKEVEDDYMKLYAGLKFQAKEPRE --SKQVYPKEVEDDYMKLYAGLKFQPKEPRF -HNKQVYPKEMEDDYMKLYANLKFQAKEPRF -KNKQVYPKFVFDDYMKLYAGLKFQAKEPRF APT---YPKFVEDDYMKLYSGLKFQAKEPRF --KCRAYPKFVFEDYMNLYLKLKFQEKEPRF ------YPSFVEDDYHKLYAGVKFQPKEPRE --KKEVYPKEVEEDYMNLYAGLKFQAKEPRF

337 283 303 298 304 304 302 308 301 304

FIG. 5. Comparison of the amino acid-residue sequence of the ethylene-forming enzyme of Pseudomonas syringae pv. phaseolicola PK2 (EFEPS) with that of the protein of the Fez+/axorbate superfamily. Alignments are conducted according to the program Gene Works (IntelliGenetics Inc., CA, USA) and modified manually. Sequences of HYSH6H, TOMES, A2, BLYFL3DOX, PENTAN3, C-DAOCS, Cl-IPNS and C2-IPNS showed only similar regions to the ethylene-forming enzymes. Residues identical to EFEPS in the sequences are boxed. Gaps(-) are represented by dashes that were introduced to optimize the alignment. Periods (.) indicate any amino-acid residues in the consensus sequence. C-DAOCS indicates the consensus sequence of de-acetoxycephalosporin-Csynthetases, DAOCS-SC the de-acetoxycephalosporin-C synthetase of Streptomyces clavuligerus (Rovacevic et al., 1989), DAOCS-CA the de-acetoxycephalosporin C synthetase of Cephalosporiurn acremonium (Samson et al., 1987), STMCEFF the p-lactam hydroxylase cefF gene product of Streptornyces clavuligerus (Kovacevic and Miller, 1991); Cl-IPNS the consensus sequence of isopenicillin-N synthetase group 1, EMEIPNS the isopenicillin-N synthetase gene product of Emericella nidulans FGSC-4 (Weigel et al., 1988), EMEIPS the isopenicillin-N synthetase gene product of Emericella nidulans blAl (Ramon et a f . , 1987). PECIPS the isopenicillin-N synthetase gene product of Penicilfium chrysogenum 23x40-269-37-2 (Carr et al. , 1986), PECISPG the isopenicillin-N synthetase gene product of Peniciflium chrysogenum AS-P-78 (Barredo et al., 1989), APEIPNS the isopenicillin-N synthetase pcbC gene product of Acremonum chrysogenum (Ramsden et al., 1989), YSSIPSG the isopenicillin-N synthetase gene product of Cephalosporiurn acremonium (Samson et al., 1985); C2-IPNS the consensus sequence of isopenicillin-N synthetase group 2,. STMIPNS the isopenicillin-N synthetase gene product of Streptomyces clavuligerus (Leskiw et al., 1988). STMIPNSSJ the isopenicillin-N synthetase gene product of Streptomyces jumonjinesis (Shiffman et nl., 1988), FVBPCBC the isopenicillin-N synthetase gene product of Flavobacterium sp. (Shiffman et al., 1990), STMIPNSA the isopenicillin-N synthetase gene product of Streptomyces lipmanni (Weigel el a f . , 1988), and IPNSSL the isopenicillin-N synthetase pcbC gene product of Nocardia lactamdurans (Conque et al., 1991). Other abbreviations are listed in the legend to Fig. 4. In the recent edition of GeneBank (release 76), YSSIPSG, FVBPCBC and IPNSSL are registered as CAIPSG, FSPCBC and NLPCBABC, respectively.

300

n

FUKUDA ET AL.

sequences (Fig. 5 ) . Comparison of these sequence segments revealed that the ethylene-forming enzyme from Ps. syringae has a relationship with the counterpart enzyme from plants. When ethylene-forming enzyme was compared with 2-oxoglutarate-dependentdioxygenases, homology scores were also low. A search for proteins homologous to the ethylene-forming enzyme from Ps. syringae using the NBRF database revealed that isopenicillin-N synthases have homology to the iron-binding site described later. This enzyme is a component of the pathway for biosynthesis of plactam antibiotics and produces a precursor of the substrate of ringexpansion hydroxylase, which utilizes 2-oxoglutarate as one of the substrates. All proteins that have partial homology to the ethylene-forming enzyme from Ps. syringae are members of an Fe*+/ascorbate oxidase superfamily noted by McGarvey et al. (1992). We could not find significant similarity to the other protein sequence deposited in GenBank and the NBRF database. Only six invariant amino-acid residues ( G I Y ~Hislx9, ~, Asp191,Gly,,,, Arg277)are observed in the optimized alignment of amino acid-residue sequences in the Fe2+/ascorbate oxidase superfamily (Fig. 5 ) . Matsuda et al. (1991) suggested that there are three regions with a relatively high degree of sequence conservation among 2-oxoglutaratedependent dioxygenases, and that these regions might function as 2oxoglutarate-binding sites and as sites for binding of Fe2+to two histidine residues. The ethylene-forming enzyme from Ps. syringae also contains the sequences for the iron-binding sites, including two histidine residues (HislS9 and but only Gly3, is conserved at another homologous site. The function of the N-terminal segment around Gly,, in catalysis is not clear because of its low degree of homology. Two amino-acid residues (Aspl9, and Arg277) are conserved in all sequences examined. These charged residues may perform significant roles in the catalytic reaction (e.g. substrate-binding or intermediate trap). The residue Thr2s9 is replaced by a serine residue in four 2-oxoglutarate-dependentenzymes and is conserved in the other enzymes examined. The hydroxyl side-chains seem to be important in catalysis. The hydropathy plot of the ethylene-forming enzyme from Ps. syringae, calculated by the method of Kyte and Doolittle (1982), is similar to those of corresponding enzymes from plants and ripening-related proteins, with the exception of two extra regions (Fig. 6). One extra region is relatively hydrophobic (residues 93-1 14) and the other (residues 209-226) is extremely hydrophilic. L-Arginine is essential for the formation of ethylene by the enzyme from Ps. syringae, as are 2-oxoglutarate, Fe2+and oxygen. However, L-arginine is not a cosubstrate for other 2-oxoglutarate-dependent enzymes. Sequence alignment (see Fig. 5 ) revealed that amino-acid residues from Tyrzln to GILI~~, represent an insertion sequence between two highly

301

ETHYLENE PRODUCTION BY MICRO-ORGANISMS

2.0 1.o 0

-1.o -2.0 -3.0

h

I

I

I

I

I

I

50

100

150

200

250

300

3E 3

z

; 2.0

a El a h z

Q)

bD

1.0

0 -1.0 -2.0

Q)

1

86

87

177 178

295

2.0 1.o 0

-1 .o -2.0

-3.01 1

.. ..

. .

106 107

197 198

.

I

.

.

314

Residue number FIG. 6. Comparison of hydropathy profiles of the ethylene-forming enzymes from (a) Pseudomonas syringae pv. phaseolicola PK2 (EFEPS) and (b), (c) those from higher plants. Hydropathy profiles ( n = 9) were calculated by the method of Kyte and Doolittle (1982) using Geneworks (IntelliGenetics, CA, USA). Hydropathy profiles of (b) TOMETHYBR, the ethylene-forming enzyme related to the pTOM13 gene product of Lycopersicon esculenturn (Holdsworth et al., 1987a), and (c) MAURRP, the ripening-related protein of Malus sylvestris (Dong et a l . , 1992a), are presented separately in the comparison. Dashed lines indicate boundary divisions.

302

II FIIKUI)A f i I A1

conserved domains, which include important histidine residues, and it provides an extra segment within the enzyme. In this region, there are four negatively charged glutamic acid residues that may be related to the binding of L-arginine. Attempts at overproduction of the ethylene-forming enzyme from P s . syringae by manipulation of the gene in pEFE03 are now in progress, and it will be of interest to determine the site for binding o f I arginine to the enzyme. Hydrophobic regions are present near the conserved histidine residues in all 2-oxoglutarate-dependentoxidases, but significant structural similarities are not found from predictions of the secondary structure of these regions. The low degree of homology of the entire sequence compared with those of the family of ethylene-forming enzymes o f higher plants and the similarities between the hydropathy profiles lead us to wonder about the evolutionary relationship between the enzyme from Ps. syringae and the other proteins.

VII. Concluding Remarks

I n this review, we have presented a summary of the recent progress in characterizing biosynthesis of ethylene by micro-organisms. There are two pathways for formation of ethylene in micro-organisms, namely the KMBA and the 2-oxoglutarate pathways. Examples of microbes that exploit these two pathways are Cr. albidus and E. coli for the first pathway, and P. digitaturn and Ps. syringae for the 2-oxoglutarate pathway. By contrast, the ACC pathway is exploited by higher plants. However, the KMBA pathway is not of great interest with respect to microbial production of ethylene because the final step on this pathway in most ethylenogenic microorganisms is a chemical reaction. Therefore, the ethylene-forming enzyme of the 2-oxoglutarate pathway has been analysed from an enzymological and molecular-genetic perspective. 'The ethylene-forming enzyme of Ps. syringae was described as an example. First, an in vitro system was constructed and then the enzyme was purified to homogeneity. Using the purified enzyme, we surveyed the ethylene-forming reactions and found that the purified enzyme simultaneously catalyses two reactions, namely formation of ethylene and of succinate from 2-oxoglutarate, with a molar ratio of 2:l. In the main reaction, 2-oxoglutarate is dioxygenated to produce one molecule of ethylene and three molecules of carbon dioxide. In the subreaction, both 2-oxoglutarate and L-arginine are mono-oxygenated to yield succinate together with carbon dioxide and L-hydroxyarginine, respectively, the latter being further transformed to guanidine and L-A'-pyrroline5-carboyxlate. A dual-circuit mechanism has been proposed for the entire reaction, in which binding of i*-arginine and 2-oxoglutarate in a Schiff-base

F I H Y L ~ N PRODUCTION F BY MICRO-ORGANISMS

303

structure generates a common intermediate for the two reactions. The gene for the ethylene-forming enzyme in fs. syringae was found to be encoded by an indigenous plasmid, designated pPSP1, of fs. syringae. The gene for the enzyme was cloned and expressed in E. coli JM109. Nucleotidesequence analysis of the clone revealed an open-reading frame that encodes a putative protein of 350 amino-acid residues (M, 39,444). In a comparison with the ethylene-forming enzymes of higher plants, the homology score for the entire amino acid-residue sequence of the enzyme from Ps. syringae compared with the enzymes from plants, or with 2-oxoglutarate-dependent dioxygenases, was low. However, functionally significant regions appear to be conserved. We also discussed the genes for the ethylene-forming enzyme of Ps. syringae and higher plants from the evolutionary point of view. We expect that the gene in fs. syringae will be very useful for production of ethylene, and we also expect that, following further research, practicable new processes for production of ethylene will be established and will herald a new age free from the threat of shortages of petroleum resources. HEFEKENCES

Adams, R. P. (1984). Biomass 4, 82. Adams, D. 0. and Yang, S. F. (1979). Proceedings of National Academy of Sciences of the United States of America 76, 170. Barredo. J . L . , Cantoral, J . M., Alvarez, E., Dicz, B. and Martin, J . F. (1989). Molecular and General Genetics 216. 91. Belay, N . and Daniels, L. (1987). Applied and Environmental Microbiology 53, 1604. Biale. J . B. (1940). Science 91, 458. Billington. D. C., Golding, B. 1’. and Primrose, S . B. (1979). Biochemical Journal 182, 827. Britsch, L.. Ruhnau. B. and Forkmann, G. (l991), GenBank, X60512. Carr, L. G., Skatrud. P I>.. Scheetz, M. E. 11. Queener, S . W. and Ingolia, T. D. (1986). Gene 48. 257. Chalutz, E. and I.ieberman, M. (1978). Plant Physiology 61, 1 1 1 . Chalutz, E . , Lieberman, M . and Sister, H. (1977). Plant Physiology 60.402. Chalutz, E . , Mattoo. A. K . , Anderson. J. D. and Lieberman, M. (1978). Plant and Cell Physiology 19, 189. Chalutz. E . , Kapulnik, E. and Chet. 1. (1983). European Journal of Applied Microbiology Biotechnology 18, 293. Chou, T. W. and Yang. S. F. (1973). Archives of Biochemistry and Biophysics 157, 73. Conque. J . J . R.,Martin, J . F.. Calzada, J . G. and Liras, P. (1991). Molecular Microbiology 5, 1125. Davis. J. B. and Squires, R. M . (1954). Science 119, 381. Deikman, J . and Fischer. R . L. (1988). E M B O Journal 7, 3315. Diguiseppi, J . and Fridovich, I . (1980). Archives of Biochemistry and Biophysics 205, 323. Dong, J . G . , Olson, D.. Silverstone, A . and Yang, S. F. (1992a). Plant Physiology 98, 1530. Dong, J . G.. Fernandez-Maculet, J. C., Olson, D. C . and Yang, S. F. (1992b). Plant Phvsiology (suppl.),22 (abstracts of papers for the Annual meeting of the American Society of Plant Physiologists). Fernandcz-Maculet, J . C. and Yang. S. F. (1992). Plant Physiology 99, 751. Frecbairn. H . T. and Buddenhagen. I. W. (1964). Nature 202, 313.

304

II. FIJKIJIIA b;T Al..

Freidman. D. J . , Imperiale, M. J. and Adhya, S. L. (1987). Annual Review of Genetics 21, 453. Fukuda, H. (1984). Chemical Economy and Engineering Review 16. 18. Fukuda, H. and Ogawa, T . (1991). In “The Plant Hormone Ethylene” (A. K. Mattoo and J. C. Suttle, eds), pp. 27Y-292. C R C Press, Boca Raton. Fukuda, H . , Fujii, T. and Ogawa, T . (1984). Agriculturaland Biological Chemistry 48, 1363. Fukuda, H . , Fujii. T. and Ogawa. T. (1986). Agricultural and Biological Chemistry 50, 977. Fukuda, H . , Fujii, T. and Ogawa, T . (1988). Biotechnology and Bioengineering 31, 620. Fukuda. H., Kitajima, H., Fujii, T., Tazaki, M. and Ogawa, T. (1989a). FEMS Microbiology Letters 59, 1. Fukuda. H . , Takahashi, M., Fujii, T. and Ogawa, T. (1989b). Journal of Fermentation and Bioengineering 67, 173. Fukuda, H.,Takahashi, M., Fujii,T.,Tazaki, M. andOgawa,T. (1989~).FEMS Microbiology Letters 60. 107. Fukuda, H., Ogawa, T., Tazaki, M., Nagahama, K., Fujii, T., Tanase, S. and Morino, Y. (1992a). Biochemical and Biophysical Research Communications 188, 483. Fukuda, H., Ogawa, T., Ishihara, K., Fujii, T., Nagahama, K., Omata, T., Inouc, Y., Tanase, S. and Morino, Y. (1Y92b). Biochemical and Biophysical Research Communications 188, 826. Fujii, T., Ogawa, T. and Fukuda, H . (1985). Agricultural and Biological Chemistry 49, 651. Cane, R. (1934). Nature 134, 1008. Gollakota, K. G . and Jayalakshmi, B. (1983). Biochemical and Biophysical Research Communications 110, 32. Goto, M. and Hyodo, H. (1987). Plant and Cell Physiology 28, 405. Goto, M., Ishida, Y.,Takikawa, Y. and Hyodo, H. (1985). PlantandCell Physiology26, 141. Guy, M. and Kende, H. (1984). Planta 160, 281. Hahm, D . H., Kwak, M. Y.,Bae, M. and Rhee, J. S. (1992). Bioscience, Biotechnology and Biochemistry 56, 1146. Halliwell, B. and Gutteridge, J. M. C. (1984). Methoah in Enzymology 105, 47. Hamilton, A. J . , Lycett, G. W. and Grierson, D. (1090). Nature 346,284. Hamilton, A. J.. Bouzayen, M. and Grierson, D. (1991). Proceedingsofthe NationalAcademy of Sciences of the United States of America 88. 7434. Hayaishi, 0.. Nozaki, M. and Abbott, M. T. (1975). In “The Enzymes” (P.D. Boycr, ed.), vol. 12, pp. 119-189. Academic Press, New York. Hoffmann, N. E., Yang, S. F., Ichihara, A. and Sakamura, S. (1982). PlantPhysiology70,195. Holdsworth, M. J., Bird, C. R.. Ray, J.. Schuch, W. and Grierson, D. (1Y87a). Nucleic Acids Research 15, 731. Holdsworth, M. J . , Schuch, W. and Grierson, D. (1987b). Nucleic Acids Research 15, 1o600. Hottiger, T. and Boller, T. (1991). Archives of Microbiology 157, 18. Hunt, J . M., Miller, R. J. and Whelan, J. K. (1980). Nature 288, 577. Hag, L. and Curtis, R. (1968). Science 159, 1357. Incc, J. E. and Knowles, C. J. (1985). Archives of Microbiology 141, 209. Ince, J . E. and Knowles, C. J. (1986). Archives of Microbiology 146, 151. Kende, H. (1989). Plant Physiology 91. 1 . Kock, M., Hamilton, A. J. and Grierson, D. (1991). Plant Molecular Biology 17. 141. Kovacevic, S . and Miller, J. R. (1991). Journal of Bacteriology 173. 398. Kovacevic, S . , Wcigel, B. J., Tobin, M. B., Ingolia, T. D. and Miller, J. R. (1989). Journal of Bacteriology 171, 754. Kutsuki, H. and Gold, H . (1982). Biochemical and Biophysical Research Communications 109, 320. Kyte, L. and Doolittle, R. F. (1982). Journal of Molecular Biology 157, 105. Leskiw, B. W., Aharonowitz, Y., Mevarech, M., Wolfe, S . , Vining, L. C., Westlake, D. W. and Jensen. S. E. (1988). Gene 62, 187. Lieberman. M. (1979). Annual Review of Plant Physiology 30. 533. Lipman, D. J . and Pearson, W. R. (1985). Science 227, 1435.

ETHYLENE PRODUCTION B Y MICRO-ORGANISMS

305

Lynch, J . M. and Harper, S. H . T. (1974) Journal of General Microbiology 80, 187. McGarvey, D. J . and Christoffersen, R. E. (1992). Journalof Biological Chemistry 267,5964. McGarvey, D. J . , Yu, H. and Christoffersen, R.E. (1990). Plant Molecular Biology 15, 165. McGarvey. D. J . , Sirevag, R. and Christoffersen, R. E. (1992). Plant Physiology 98, 554. McKeon, T. A. and Yang, S . F. (1984). Planta 160, 84. McKeon, T. A. and Yang, S. F. (1987). In “Plant Hormones and Their Roles in Plant Growth and Devclopment” (P. J. Davies, ed.), pp. 94-112. Martinus Nijhoff, Dordrecht. Mansouri, S. and Bunch, A. W. (1989). Journal of General Microbiology 135, 2819. Matsuda, J . , Okabe, S., Hashimoto, T . and Yamada, Y. (1991). Journal of Biological Chemistry 266, 9460. Mattoo, A . K., Chalutz, E., Anderson, J. D. and Lieberman, M. (1979). Plant Physiology 64, 5s. Mayne, R. G. and Kende, H. (1986). Plant 167, 159. Meldgaard, M. (1991). GenBank, X58138. Messen, A , , Hohman, S., Martin, W., Schnable, P. S., Petersom, P. A., Saedler, H. and Greril, A. (1990). E M B O Journal 9, 3051. Miller, E. V., Winston, J . R. and Fisher, D. F. (1940). JournalofAgricultural Research 60,269. Nagahama, K . , Ogawa, T . , Fujii, T., Tazaki, M., Goto, M. and Fukuda, H. (1991a). Journal of General Microbiology 137, 1641. Nagahama, K., Ogawa, T., Fujii, T.. Tazaki, M., Tanase, S . , Morino, Y. and Fukuda, H. (1991 b). Journal of General Microbiology 137, 2281. Nagahama, K . , Ogawa, T., Fujii, T. and Fukuda, H. (1992). Journal of Fermentation and Bioengineering 73, 1. Nemethy, E. K. and Calvin, M. (1982). Phytochemistry 21, 2981. Nickerson, W. J . (1948). Archives of Biochemistry 17, 225. Nielsen, P. E., Nishimura, H., Otvos, J. W. and Calvin, M. (1977). Science 198, 942, Ogawa, T., Takahashi, M., Fujii, T., Tazaki, M. and Fukuda, H . (1990). Journal of Fermentation and Bioengineering 69, 287. Ogawa, T., Murakami, H., Yamashita, K., Tazaki, M., Fujii, T . and Fukuda, H. (1992). Journal of Fermentation and Bioengineering 73, 58. Oremland, R . S. (1981). Applied and Environmental Microbiology 42, 122. Oremland, R . S. and Marais, D. J. D. (1983). Geochimica el Cosmochimica Acta 47, 2107. Peiser, G. D., Wang, T. T . , Hoffmann, N. E., Yang, S. F., Liu, H. W. and Walsh, C. T. (1984). Proceedings of the National Academy of Sciences of the United States of America 81, 3M9. Primrose, S. B. (1976a). Journal of General Microbiology 95, 159. Primrose, S. B. (1976b). Journal of General Microbiology 97, 343. Primrose, S. B. (1977). Journal of General Microbiology 98, 519. Primrose, S. B . and Dilworth, M. J. (1976). Journal of General Microbiology 93, 177. Pua, E., Sim, G. K. and Chye, M. L. (1992). Plant Molecular Biology 19, 541. Ramon, D., Carramolino, L., Patino. C., Sanchez, F. and Penalva, M. A. (1987). Gene 57, 171. Ramsden, M., McQuade, B. A , , Saunders, K., Turner, M. K. and Harford, S. (1989). Gene 85, 267. Samson, S. M., Belagaje, R., Blankcnship, D . T., Chapman, J . L., Perry, D., Skatrud, P. L., VanFrank, R. M., Abraham, E. P., Baldwin, J . E., Queener, S. W. and Ingolia, T. D. (1985). Nature 318, 191. Samson, S. M., Dotzlaf, J . E., Slisz, M. L., Becker, G. W., Van Frank, R. M., Veal, L. E., Yeh. W. K.,Miller, J . R . , Queener, S. W. and Ingolia, T. D. (1987). BiolTechnology 5, 1207. Sato, M., Urushizaki, S., Nishiyama. K., Sakai, F. and Ohba, Y. (1987). Agricultural and Biological Chemistry 51, 1177. Schwartz, R. M. and Dayhoff, M. 0. (1979). In “Atlas of Protein Sequence and Structure” (M. 0. Dayhoff, ed.), vol. 5, suppl. 3, pp. 352-368. National Biochemical Research Foundation, Washington, DC.

306

I I FUKUDA hT Al..

Shiffman, D., Mevarech, M., Jensen, S. E.. Cohen, G. and Aharonowitz, Y.(1988). Molecular and General Genetics 214, 562. Shiffman, D., Cohen, G . , Aharonowitz, Y.,von Dohren. H . and Kleinkauf, H. (1990). Nucleic Acids Research 18. 660. Shipston, N. and Bunch, A . W. (1989). Journal of General Microbiology 135, 1489. Spanu, P., Reinhardt, D.and Boller, T . (1991). EMEO Journal 10, 2007. Tanaka, H., Enoki, A. and Fuse, G . (1986), Mokuzai Gakkaishi 32, 125. Tauber, A. and Babior, B . M. (1980). Journal of Clinical Investigation 60,374. Thomas, K. C. and Spences, M. (1977). Canadian Journal of Microbiology 23, 1669. Turner, E. M., Ward, M. W. T . and Osborne, D.J. (1975). Journalof GeneralMicrohiology 91, 167. Venis, M. A. (1984). Planta 162, 85. Ververidis, P. and John, P. (1991). Phorochernistry 30,725. Wang, H. and Woodson, W. R. (1992a). GenBank, M90294. Wang, H. and Woodson, W. R. (1992b). GenBank, M62380. Watanabe, T. and Kondo, N. (1976). Plant and Cell Physiology 17, 1159. Weigel, B. J., Burgett, S. G . , Chen, V. J., Skatrud, P. L., Frolik, C. A,, Queener, S. W. and Ingolia, T. D. (1988). Journal of Bacteriology 170, 3817. Yang, S . F. and Hoffman, N. E. (1984). Annual Review of Plant Physiology 35, 155.

Addendum Added in Proof

“Biosynthesis and Expression of Cell-Surface Polysaccharides in GramNegative Bacteria” by Chris Whitfield and Miguel A. Valvano. While this review was at the proof stage, a number of research advances have occurred. Several affect a n d o r support conclusions made here and are particularly relevant to our discussion: (1) Bastin etal. (Bastin, D. A., Stevenson, G . , Brown, P. K. and Reeves, P.R. (1993). Molecular Mictohiology 7 , 725) have reported the sequence of the cld (also termed roo gene from E. coli 0 1 11 and S. enterica serovar Typhimurium LT2. A model is proposed whereby CldRol interacts with Rfc to generate a polymerase complex with two states: “E”, which facilitates extension, and “T”, which is involved in transfer of nascent 0-polysaccharide to LPS lipid A-core. Interaction between the two proteins in the complex is suggested to determine the heterogeneity of 0-polysaccharide chain lengths found in LPS. (2) Recent studies have shown that rfe participates with the rfb gene cluster in biosynthesis of the E. coli 0 7 0-polysaccharide; no 07 polysaccharide is synthesized in rfe mutants. The 0 7 rfb cluster lacks a gene similar to rfbP in S. enferica (C. L. Marolda and M. A. Valvano, unpublished results), which encodes the initiating sugar-transferase enzyme in the assembly of the 0-polysaccharide repeating unit. It is proposed that Rfe provides the N-acetylglucosaminyltransferase required for synthesis of the initial undecaprenylpyrophosphoryl-N-acetylglucosamineintermediate in 0 7 biosynthesis. This further extends the role of Rfe in E. coli and provides another important example of interaction between genes in different clusters in the biosynthesis of cell-surface polysaccharides. (3) Kogan et al. (Kogan, G . , Haraguchi, G., Hull, S. I. Hull, R . A., Shashkov, A. S., Jann, B. and Jann, K. (1993). European Journal of Biochemistry 214,259) have described a series of modified 0-plysaccharides formed by E. coli K-12 derivatives containing the rfb region transferred from E. coli 0 4 . Hybrid strains were constructed by conjugation, P1 transduction, and transformation with plasmids carrying the cloned 0 4 rfb genes. The 0-polysaccharide modifications included alterations in the sites, glycosyl group substituents, new linkage configurations, and replacement of one sugar component with another not normally seen in the 0 4 repeating unit. The data is consistent with co-operation between host and donor rfb gene products in generation of novel 0-polysaccharides. These studies have important implications for all future studies on cloned rfb genes (and perhaps also other polysaccharide biosynthesis genes) expressed in E. coli K-12.

308

A1)I)ENI~UM

( 4 ) Frosch et a f . (Frosch, M. and Miiller, A. (1993). Molecular Micmbiology 8, 483) recently described two region-B cps genes (lipA and lib@ required for biosynthesis of Neisseria meningitidis group-BCPS. These genes are implicated in the phospholipid substitution of the group TI-like CPS. Mutants were identified which accumulate lipid-modified polymer in the cytoplasm, an observation which clearly indicates that lipid modification precedes, rather than follows, transport across the cytoplasmic membrane. The authors propose that lipid modification is essential for subsequent transport. These results imply that the ATP-binding cassette transporters which translocate CPS across the cytoplasmic membrane do not function as hydrophilic pores. It remains to be established whether identical pathways occur in other bacteria with group 11-like CPSs, but the similarity in the components (discussed in Section IV.B.2) is consistent with a common mechanism.

Author Index

Numbers in bold refer to pages on which references are listed at the end of each chapter Aaronson, A , , 184, 235 Aaronson, S., 258, 270 Aaronson, W., 184, 199, 201, 223, 228, 231, 243. 245. 246 Abbott, M. T., 291, 304 Abe, J., 159, 236 Abe, M., 146, 231 Abeles, R. H., 74, 76, 104 Abraham, E. P., 2Y9, 305 Abraham, J. M., 12, 65 Achtman, M., 184. 212, 213. 231, 232 Adam, L., 8, 37, 38, 60,61, 63, 64, 65 Adams, D. A , , 181, 187, 245 Adams, D. 0..287, 303 Adams, R. P., 276, 303 Adelman, M . R., 14, 23, 26, 67 Adhaya, S . L., 293, 304 Adhya, S . , 190, 231 Adlakha, R. C., 56, 63 Adlam, C., 152, 231 Adler, P. N., 27, 28, 29, 69 Adoutte. A , , 10, 63 Affara, N., 72, 89, 105 Agranoff, B. W . , 259, 271 Agris, P. F., 89, 104 Aharonowitz, Y.,2Y9, 304,306 Aird, E. L. H., 205, 231 Akhavan-Niaki, H., 14, 22, 23, 31, 33, 63, 69

Al-Hendy, A , , 192. 193, 212, 213, 231 Albershcim, P., 145, 231 Albertini, C., 14, 22, 23, 31, 63, 69 Alberts, B., 30, 68 Albracht, S . P. J., 85, 104

Albrecht, A . -M., 259, 270 Albrecht, P., 250, 272 Albright, L. M., 218, 231 Aldrich, H. C., 4, 13, 14, 23, 39, 63, 68, 78, 104

Alexopoulos, C. J., 2, 63 Alifano, P., 196, 213, 231 Allen, P., 220, 221, 231 Allfrey, V. G., 58, 68 Aloni, H . , 228, 242 Aloni, Y.,167, 231, 244 Alvarez, E., 299, 303 Alzner-DeWeerd, B., 88, 105 Amako, K., 142, 231 Aman, P., 145, 231 Amemura, A., 146, 159, 168, 231, 236, 238

Ames, B. N., 76, 88, 89, 95, 107 Amikam, D., 167, 168, 198, 228, 229, 231, 242,244,246

Aminoff, D . , 164, 231, 238 Anders, J., 226, 236 Anderson, J. D., 279, 303, 305 Anderson, J. S . , 154, 162, 231. 244 Anderson, R. G., 172, 231 Anderson, R. W . , 4, 5, 10, 11, 15, 26, 28, 31, 33, 34, 35, 36, 37, 41, 47. 53, 61, 63, 64,66,68,69

Anderson, S. E., 34, 63 Anding, C . , 266, 270. 272 Andreassen, R., 48, 51, 65 Andreesen, J . R., 72, 73, 74. 75, 76, 80, 81, 86.87, 89, 96, 104. 105, 106, 107, 108, 109 Annunziato, P., 243

310

AlI’l‘IIOH I N l ) t X

Antoniewski, C., 116, 123. 132 Aota, S., 89, 109 Aoyama. E., 98, 107 Arai, T., 196, 237 Arakawa, Y., 163, 164, 191, 192, 204, 221, 231, 238. 241, 245

Araki, Y . , 144, 246 Archibald, A . R . , 140, 233 Arico. B., 223, 231 Arkowitz, R. A , , 74, 76. 104 Arnold, W., 190,205,2M, 21 I , 231,233,238 Arondel, J . , 221, 240 Arsenault, T. L.. , 144, 231 Augerson, W. S., 209, 243 Ausubel, F. M., 218, 231, 242 Axley, M. J . , 77, 78, 81. 83, 84, 96, 97, 104 Ayres, J. C., 97, 99, 109 Babcock, K. L., 6, 67 Babior. B. M., 283. 306 Babiuk, L. A , , 147, 241 Bac, M., 277, 304 Bachellerie, J. -P., 10. 63 Bacon, M., 102, 104 Baddiley, J . , 140, 172, 231, 233 Baek, S. R., 205, 233 Bahl, 0. P., 97, 105 Bailey, J . , 26, 28, 31, 33, 34, 35, 36. 37, 47, 60,61, 62, 63, 64,68 Bailey, M. J. A , , 212, 231 Bailly, E., 14, 63 Baker, B., 258, 270 Baker, R. S., 153. 233 Balaban, R. S., 92, 95, 96, 108 Baldus, J . M., 130. I33 Baldwin. J. E.. 299. 305 Bamberry, R. J . , 258. 273 Bandursky, R. S., 99, 109 Bandziulis. R., 10. 66 Banffer, J. R., 97. 104 Banoub, J. H., 144, 231 Banroques. J . , 10, 64 Banu. L.. 73, 95, I04 Baratti, J., 269, 271 Barbas, J. A , , 1x7, 231 Barber, C. E., 223, 225. 233, 244 Barden, A . , 8, 38, 61, 63 Barker. R. F., 229, 232 Barnickel, G., 141, 238 Baroin, A , , 10, 63 Baron. C., 72, 89, 91, 93, 94, 95, 102, 104, 105, 106

Baron, L. S., IY7,220,227,235,237,241,243 Baron, S. F., 78. 104 Barr, K., 138, 161, 162, 163. 164, 191. 206. 231. 240, 242

Barredo, J . L,., 299. 303 Barrow, K . . 253. 254. 270 Bartlett, D. H . 227, 231 Bartlett, R.. 54, 64 Bastin, D. A . , 192, 197, 213, 231 Batchelor, R. A , . 192, 196, 213. 231 Bathgate, A . J . 212, 240 Batley, M.,146, 205, 231, 233 Bayer, M. E., 142, 1x1, 1x5, 186. 187, 231, 232

Bayer, M . H . 185, 187, 232 Beach, D., 56. 57. 64,65 Beach, M. J . , 268, 270 Becam, A . M., 10, 64 Beck-Sickinger, A , , 269, 270, 272 Becker, A , , 205, 231 Becker. G. W . , 295. 299, 305 Becker, M. M.. 72, 73, 87, I05 Beckwith, J., 183, 243 Beger, D. W., 191. 192. 197, 210, 232, 233. 236, 239

Behne. D., 73, 96, 104 Belagaje, R., 299, 305 Belay, N.. 276. 303 Bell, A. W., 178, 236 Bellows, J. L., 156. 242 Beltrame, J., 198, 233 Beltran, P., 146, 220, 243 Belunis, C. J . , 163, 232 Ben-Bassat, A.. 167, 168, 198, 228,229.246 Renard, M.. 26,3X, 50. 5 1 , 5 2 , 5 3 , 6 l . 63,67 Benne, R., 12, 63 Bcnsky. B., 258, 270 Benz, R., 259, 270 Benziman, M., 167, 168, 198, 228, 229, 231, 242,244,246 Benziman, S.. 167. 231 Berg, B. I>.. 78, 80,93, 104 Berger, E. A . , 97. 104 Berger, Y., 266, 272 Bergstrom, H.,164, 172, 200, 202, 208, 245 Berlier. Y . , 84, 85,X6, 96, 106, 108 Bernhard, F., 220, 232 Bcrnier, F., 3X. 63 Rcrnstein, P., 43, 63 Bernstein, R . I-., 198, 227, 232, 242 Berry. A . , 146, 166, 167, 184. IYX, 190, 206, 221, 224, 232, 234, 240. 243 Berry. A . M . , 255, 256. 270 Berry, D., 212, 216, 232 Berry, M.,73. 95, 36, 104 Betlach, M. R., 171. 205, 245 Bctterley, D. A , , 4, 64 Beutin, L,., 212, 213, 232 Beveridge, T. J., 136, 137, 142, I X2.232,235, 238

AUTHOR INDEX

Beynon, L. M..143, 232 Bhagwat, A. A, , 168, 232 Bhagya Lakshmi, S. K., 196, 243 Bhattacharjee, A. K., 143, 152, 232, 235 Biale, J. B., 277, 279, 303 Biemann, K., 250, 271 Biffali, E., 196, 213, 231 Billington, D. C., 279, 282, 303 Bilous, P. T., 100, 104 Binette, F., 26, 38, 63 Binkley, S. B.. 152, 239 Bird, C. R., 295, 301, 304 Bird, C. W . , 250, 270 Bird, T.. 130, 133 Birkett, C. R., 17, 18, 19, 29, 63,68, 69 Birkmann, A., 72, 78, 89, 96, 98, 109 Birmingham, J., 182, 232 Bisseret, P., 252, 253, 259, 268, 270. 272 Bitter-Suermann. D., 138,162,177, 182,183, 185, 191, 200, 201, 238,242 Bjork, G. R., 95, 105 Bjornstedt. M., 98, 105 Blackburn, E. H., 8, 63,65 Blair, J. G., 213, 239 Blankenship, D. T., 299, 305 Bligh, E. G., 253, 270 Blindt. A. B., 17, 18, 27, 28, 63. 69 Bloch, K., 260, 270 Blume, A , , 259, 271, 272 Blumenthal, R., 184, 244 Bobik. T. A , , 78, 105 B&k, A., 71, 72, 76, 77, 78, 80,88, 89, 90, 91. 92,93,94,95,96,97,98, 99, 102, 104, 105. 106, 107, 108, 109

Bock, R. M.. 95, 108 Boesmiller, K., 92, 93, 106 Bohm, R., 78, 105 Bohnert. E., 47.68 Bokranz, M., 80, 105 Boller, T., 279, 284, 287, 292, 295, 304,306 Bonamy, C . , 115. 131, 132, 133 Bonner, W. M., 46, 67 Borisy, G. G., 14, 22, 63, 66 Borkovitch, K. A , , 218, 232 Bornens, M., 14, 63 Borthakur, D., 229, 232, 238 Bossio, J. C . , 158, 232 Boston, R. S., 15, 18, 19, 64 Botstein, D., 183, 238 Boulanger, P., 187, 235 Boulnois, G., 164, 177, 178, 182, 183, 185, 172, 178, 182,200,201,202,208,234,238, 242,245

Boulnois. G . J., 177, 178, 182, 199, 201,207, 232,234,24

Bourdineaud, J . P., 187, 232

31 1

Bourret, R. B., 218, 232 Boursier, P., 72, 73, 87, 105 Bousset, K., 172, 178,180, 184,202,208,235 Bouvier, J., 115, 132 Bouvier, P., 250, 258, 266, 270, 272 Bouvier-Nave, P., 251, 253, 254, 255, 272 Bouzayen, M., 292, 304 Boykins, R. A., 152, 244 Bradaczek, H., 141, 142, 237, 238 Bradbury, E. M., 44,45, 46,47, 52, 55, 56, 57, 63, 67, 69 Brade, H., 141, 236 Brade, L., 141, 235 Bramucci, M. G., 119, 132 Brandt, J. B., 198, 245 Branefors-Helander, P., 153, 232 Brass, J. M., 182, 232 Braun, R., 45, 48, 52, 53, 63,64, 69 Braun, S., 228, 242 Braun, V., 212, 216, 241 Bravo, R., 44, 64 Bray, D., 160, 232, 242 Brazas, R., 213, 232, 234 Breton, K., 119, 133 Brigelius-Flohe, R., 73, 108 Brightwell, G., 205, 231 Brill, 3. A,, 218, 232 Brill, W. J., 159, 244 Bringer, S., 254, 270 Brink, B. A,, 196, 232 Brisson, J. R., 149, 232, 234 Britsch, L., 295, 303 Brizuela, L., 56, 57, 65 Broek, D., 54, 64 Bronner, D., 164, 165, 166, 173, 202, 234 Brooks, C. J . W., 250, 270 Brophy, L. N., 178, 202, 226, 229, 232, 238 Brown, J. R.,93, 107 Brown, P. K., 173, 176, 192, 193, 195, 196, 210, 213, 232, 238 Brown, R. M., Jr 168, 233 Brown, T. A., 98, 99, 105 Brown, T. M., 141, 236 Brugerolle, G., 10, 63 Brumbley, S . , 205, 233 Brumbley, S. M., 223, 232, 233 Bruner, R., 167, 168, 198, 228, 229, 246 Bryan, L. E., 226, 246 Bryant, R. D., 98-9, 105 Brzoska, P. M., 196, 233, 246 Bu’lock, J., 250, 270 Bublitz, C., 260, 271 Buchschacher, G. L., Jr 8, 16, 17, 18,22,25, 42, 64,67 Bucke, C., 166, 241 Buckel, S. D., 178, 236

312

AUTHOR INDEX

Buckel, W., 74, 76, 105 Buckmire, F. L. A , , 167, 234 Budenhagen, I . W., 277, 303 Buendia, A. M., 190, 233 Bulinski. J. C., 16, 65 Bunch, A . W.. 277, 279. 305, 306 Burbulys, D., 113. 116, 117, 118, 127, 130. 132, 133 Bureau, T. E., 168. 233 Burgett, S. G., 295, 299, 306 Burk, R. F., 72, 73, 105 Burland, T. G., 3, 8, 15, 16, 17, 18, 19, 20. 21. 22. 25, 27, 28, 29, 30, 32, 33, 37, 39, 41, 42, 43, 47, 57, 59, 60,61, 62, 63, 64, 65, 67, 68 Bums, C. H., 258, 270 Busficld, F., 220, 233 Butler, P. D., 190, 240 Cabanas, D. K.,205, 235 Cabane. K.,128, 129, 132 Cabrera-Martinez, R. M., 131, 133 Calhoon, R. D. 167. 168. 198, 228, 229, 246 Calva, J. R., 196, 233 Calvin, M., 275, 276, 305 Calzada, J. G., 295. 299, 303 Campbell. J . N., 102, 107 Cane, D. E., 262, 270 Cangelosi, G. A,, 180, 233 Cannon, J. F., 95, 108 Canter Cremers, H. C. J., 205, 233 Cantor, C. R., 58, 68 Cantoral, J. M., 299. 303 Capagc, M. A., 171. 205, 245 Capage, M. J . , 229, 241 Carlernalm, E . . 136. 182, 236 Carlson, R. W., 196. 232, 243 Carney. B. F.. 223, 233 Carpenter, D. R.. 97, 106 Carr, L. G . , 295, 299, 303 Carramolino, L., 299, 305 Carrino, J . J . , 45. 64, 66 Carson, J . , 171, 241 Carter, K.I.,144, 243 Carter, L., 111, 118, 132 Casadaban, M. J . , 191, 242 Caugant, D. A , , 210, 233 Celis, J. E., 44, 64 Ceri, H., 226. 246 Chainey, A . M., 27. 28, 39, 47, 63,64 Chakrabarty, A . M., 146,157, 166,167, 184, 198, 199,206,221,222,224,225,232,233, 234,237, 240, 242, 243, 245 Chalutz, E., 279, 303, 305 Chamberlin, M. J., 119, 132 Chambers, I.. 72. 89, 105

Chambliss, G. H., 126, 133 Chang, M. T., 42, 66 Chapman, J. L., 299,305 Chapman, J. W., 114, 115, 133 Chatterjee, A , , 220, 233 Chatterjee, A . K.,220, 233 Cheah, K. -C., 192, 197, 233. 239 Chen, C. - S., 88, 105 Chen, J. F., 212, 233 Chen, V. J., 295, 299,306 Chen, W., 97, 105 Chen, Y., 95, 104 Chet, I., 279, 303 Chibazakura, T., 126, 132 Chikuma, M., 98, 107 Ching, W. -M. 72, 88, 89, 105, 108, 109 Chitnis, C., 167, 233 Choi, E. S., 85, 108 Choi, I. S., 73, 106 Chou, T. W., 281, 284,303 Choy, Y. -M. 157, 233 Christoffersen, R. E. 285, 295, 300, 305 Chu, L., 146, 166, 167. 184, 198, 199, 221, 222, 224, 233. 237, 240 Chuck, J. A., 253, 254, 270 Chun, W., 220, 233 Chye, M. L., 295, 305 Clarke, B., 142, 239 Clarke, B. R., 147, 149, 161, 190, 192, 210, 223,233, 241 246 Clarke, C. A,, 198, 233 Clausnizer, B., 47, 68 Clayton, L., 16, 21, 68 Cleary, J. M., 158. 170, 171, 205, 235, 240 Clementz, T., 163, 233 Cleveland, D. W., 43. 69 Clover, R. H., 196, 233 Cohen, G., 299,306 Cohen, G. N., 97, 105 Coira, J. A , , 159, 244 Cole, S. P., 116, 117, 133 Cole, S. T., 221, 240 Coleman, M., 220, 233 Coleman, M. J., 220, 242 Coley, J., 140, 233 Collins, H. H., 197, 220, 242, 243 Collins, L. V.. 161, 176, 193, 208, 233 Collins, 0. R., 4, 5, 11. 64,68 Colvin, J . R.,250, 271 Cone, J. E., 72, 74, 105 Connelly, C. J., 226, 236 Conner, R. K.,258, 271 Conner, R. L., 258, 270 Conque, J. J. R., 295, 299, 303 Cook, D. J., 6, 26, 33, 63, 64 Cooney, C. A., 52.64

AUTHOR INDEX

Coote, J. G., 112, 133 Coplin, D. L., 138, 190, 199, 204, 220, 232, 233, 234, 239, 241, 244

Coren, J. S., 8, 64 Corn, P. G., 226, 236 Costello, G . P., 187, 232 Costerton, J. W., 184, 185, 245, 246 Coucheron, D., 227, 233 Coucheron, D . H., 167, 244 Couso, R. O., 158, 171, 237 Cowan, N. J., 21, 69 Cowie, D. B., 97. 105 Cox, J., 71, 76, 78, 97, 98, 105 Cox, R. A., 8, IS, 17, 18, 19, 65, 67, 69 Crawford, K.,54, 64 Creeger, E. S., 187, 212, 233, 237 Creusot. F., 10, 64 Criddle, R. S., 97, 99, 107 Cripps, R. E., 171, 245 Crisel, R. M., 153, 233 Crosa, J. H.. 191, 192, 245 Crouch, R. J., 97, 109 Crowley, W. F., 199, 246 Culbertson, C . W., 100, 107 Cunningham, D . B., 6 , s . 16, 17, 18, 22, 25, 42, 49, 50, 62. 64, 67 Curtis, R., 279, 280, 304 Curvall, M., 147, 233 Cutter, G. A,, 100, 105 Cynkin, M. A , , 172, 241 Dafoe, L., 192, 193, 1%. 240 Dahnke, T., 222, 235 Dalbey, R. E., 183, 233 Dammin, G. J.. 197, 242 Danbara, H., 196, 237 Danchin, A,, 225, 233 Daniel, D. C., 52, 53, 64 Daniel, J. W., 41, 65 Daniels, L., 276, 086 Daniels, M. J., 223, 225, 233, 241, 244 Danilov, L. L., 172, 233 Dank, N., 88, 108 Dankert, M.,154, 156, 158, If@, 169, 171, 233, 237, 242, 246

Dankert, M. A , , 158, 232 Danner, D . B., 208, 243 Darvill, A . G . , 145, 231 Darzins, A. , 167, 221, 224, 234, 245 Dasgupta, T., 142, 238 Davidson, I. W., 167, 233 Davie, E., 213, 232, 234 Davies, B. H., 262, 273 Davies, J., 59, 65 Davis, D. P., 213, 239 Davis, F. M., 258, 271

313

Davis, J. B., 276, 303 Davis, J . N., 72, 73, 89, 94, 95, 107, 108 Dawes, E. A,, 263, 270 Dawid, I. B., 8, 66 Dawson, P., 16, 21, 68 Day, D. F., 212, 233 Day, J . , 122, 125, 126, 133 Dayhoff, M. O., 295, 305 Dazzo. F. B., 158, 235 De Crecy-Legard, V., 225, 233 De Kruyff, B., 258, 271 de Mendoza, D . , 120, 133 de Rosa, M., 250, 261, 270, 271 de Rosa, S . , 261, 270 Deb, S., 187, 237 Dee, J . , 4 , s . 6, 13, 15,26,27,28,30,31,33, 34, 35. 36, 37, 38, 39, 41, 47, 53, 61, 63, 64, 67, 68

Dee, L., 34, 63 delannino, N. I., 158, 232 Deich, R. A , , 208, 243 Deikman, J., 295, 303 Delahodde, A., 10, 64 DelCastillo, L., 26, 37, 64, 67 deLeij, L., 185, 233 DeLey, J., 263, 272 Dell, A , , 147, 237 Delmer, D. P., 167, 231 DeLuca, M., 61, 64 Demel, R. A , , 258, 271 DeMoss, J . A., 71, 76, 78, 97, 98, 105 DeMoss, R. D., 263, 271 Dengler, T., 142, 143, 157, 233, 237 Denny, T. P., 205, 223, 232, 233 dePaus, P., 228, 242 Deretic, V., 221.222,224,225,226,233,234, 238.240 DerVatranian, D. V., 84, 85, 86, 96, 106 DeVault, J. D., 146, 166, 167, 184, 198, IW, 221, 222, 224, 225, 232, 234, 237, 240 Devi, S. J . N., 152, 234 DeVos, G. F., 205, 234 deVroom, E., 228, 242 deWet, J . R.. 61, 64 Dhillon, N., 130, 133 Diaz, J., 187, 231 Dietrichs, D., 73, 74, 76, 89, 105 Diez, B., 299, 303 DiFabio, J. L., 149, 234 Diggins, M. A., 15, 18, 19, 29, 64, 65 Diggins-Gilicinski, M., 18.20.28, 29,32,33, 6468

Diguiiseppi, J., 283, 303 Dijkwel, P. A., 51, 52, 65 Dikshit, R., 221, 234 Dilworth, G. L., 72, 73, 87, 105

314

AUTHOR INDEX

Dilworth, M. J.. 277, 305 DiMarco, A. A., 78, 105 Dimitriev, B. A,, 151, 210, 239 Ding, M. -J., 192, 234 Ditta. G., 146, 198, 234 Ditta, G. S., 168, 180, 237, 243 Djordjevic, M. A., 229, 235 Djordjevic, S. R., 146, 231 Dodds. K. L.. 212, 234 Dodyk, F., 164, 231 Doelling, V. W., 140, 238 Doherty, D., 205, 223, 234 Doherty. D. H., 170, 171, 205. 236,245 Doi, R. H., 122, 132 Dolph. P. J., 190. 204, 220, 234, 244 Donahue-Rolfe, A , , 181. 234 Dong, J. G.. 288, 292, 295, 301 303 Dong, 0.. 224, 225, 234 Doolittle, R. F., 300. 301, 304 Doran, J. W..100. 105 Dorman, D. E.. 153, 233 Dotzlaf, J. E., 295, 299, 305 Douglas, C., 146, 198, 234 Douglas, D. J., 198, 234 Dove, W. F., 6, 8, 13, 15, 16, 17, 18. 19,20, 21, 22, 25, 27, 28. 29, 30, 32, 33, 34, 35, 37. 39, 41, 42, 43, 45, SO, 59, 60,61, 62, 64, 65, 66, 67, 68 Dow, J. M., 223, 244 Downie, J. A , , 178, 236 Drake, C. R.,199, 234 Drapeau. G. R.,217, 218, 219, 235 Drotar, A., 102, 105 Druzhinina, R. N.. 154. 156, 160, 169, 243 Dubnau, D., 119, 129, 132 Dubnau, E., 118, 119,124, 127,128,132, I33 Ducommun, B., 14,23,31,42,56,57,64,69 Dumanski, A. J.. 143, 232 Dunne, W. M.,Jr. 167, 234 Dunphy, W. G., 56, 57, 65 Durica, D. S., 49, 50, 62, 67 Diirre, P., 72, 73, 74, 87, 96, 105 Dutton. G. G . S., 147, 157, 233, 234, 241 Dyer, W. J., 253, 270 Dylan, T., 146. 168, 198, 234. 237 Easson, D. D., 226, 234 Ebright, R. H., 224, 225, 234 Edwards, E. S., 71. 76, 78, 97, 98, 105 Edwards, U., 172, 178, 180, 184, 202. 208, 235

Egan, W., 140, 152, 153, 234,244, 245 Eggerer, H.. 259, 271 Ehrenreich, A , , 92, 95, 96, 105 Eibl, H.,259, 270 Eichinger, G . H., 167. 168,198,228,229,246

Eidsness. M. K., 85, 106 Ekiel, I., 261, 271 Elias, P. M., 196, 233 Ellison, E. L.. 9. 67 Ely, S., 226, 234 Emerick, A. W., 167, 168, 198,228,229,246 Enenkel, B., 190, 233 Engberg, J., 8, 10, 65, 66 Engel, A , , 92, 106 Engelhardt, H., 92. 106 Enoch, H. G . , 76, 77, 78, 80, 81,83, 106 Enoki, A,, 279, 306 Enos, A . P., 58, 67 Entian, K. -D., 269, 270, 272 Epstein, E. M., 8, 64 Erbing, C., 153, 232 Ericson, J. U., 95, 105 Errington, J., 119, 132 Esaki, N., 98, 102, 106, 108 Evans, H. E., 41, 65 Evans, H. J . , 72,73, 87, 105 Evans, 1. J . , 178, 236 Evans, J. A . , 262, 271 Evans, L. R.,167, 239 Evans, T. E., 41, 52, 63,65 Eykholt, R. L., 52, 64 Ezaki, T., 220, 236 Fagelson, J. E., 128, 129, 132 Fahrenholz, F., 80, 105 Falkow, S., 218, 223, 226, 231, 240, 244 Fall, L. R., 102, 105 Fall, R., 102, 105 Farewell, A . , 213, 232, 234 Farwell, A . P.,96, 108 Fauque. G., 84, 85, 86, 96, 106, 108 Feagin, J. E., 12, 65 Fear, A. L., 167, 168. 198, 228, 229, 246 Feavers, 1. M., 119, 133 Feingold, D. S., 221. 242 Feinstein, N., 39, 41, 65 Fennessey, P., 154, 246 Ferguson, D. J. P.,226, 232 Ferguson. K. A., 258, 271 Fernandez-Maculet, J. C., 287, 288, 303 Ferrari, E., 119, 132 Ferrari, F. A,, 114, 115, 123, 132 Ferris. K. E.. 1 4 , 220, 243 Ferris, P.J., 9, 52, 65 Ferry, J. G., 76, 77, 78, 104, 107, I08 Fialho, A., 206, 243 Fiddler. A,, 228, 242 Fiedler, W., 186, 236 Filipova, T. M., 252, 271 Fimmel, A . L., 98. 106 Finan, T. M., 20s. 234

AUTHOR INDEX

Finke, A., 164, 165, 166, 173, 202, 234 Finn, C. W., 182, 183, 199, 201, 243 Fischer, R. L., 295, 303 Fisher, D. F., 277, 279, 305 Fisher, R. F., 198, 234 Fjaervik, E., 167, 244 Flanagan, R.,50, 51, 67 Flemming, H.-C., 161, 162, 234 Flesch, G., 252, 257, 262, 268, 269, 271 Flohk, L., 72, 73, 89, 106, 108 Flynn, J. L., 222, 234, 241 Fogher, C., 130, 133 Fokkens, R., 146, 168, 246 Folch, J., 253, 271 Foley, M., 182, 232 Foltz, C. M., 72, 108 Forchhammer, K., 72,76, 88,89, YO, 92,93, 94, 95, 96, 97, 99, 102, 105, 106, 107 Forkmann, G., 295, 303 Formal, S. B., 197, 209, 235, 238. 242, 243 Forney, J., 8, 65 Forster, C., 93, 106 Forster, J., 250, 271 Fortnagel, P., 192, 193, 197, 244 Foster, B. D., 213, 239 Foster, K. E., 17, I Y , 20, 21, 31, 39, 43, 63, 65, 67, 68

Foulds, J., 184, 235, 244 Fournet, B., 147, 149, 238 Fournier. J. -M., 221, 240 Foxon, J. L., 4, 28, 39, 41, 47, 61, 64 Frampton, J., 72, 89. 105 Frank, P., 97, 106 Franzen, L. -E., 145, 231 Fraser, B. A., 140, 235 Frederick, R. D., 220, 244 Freebairn, H. T., 277, 303 Freeley, T. C., 209, 235 Frehel, C., 186, 239 Freidman, D. J., 293, 304 Frerman, F. E., 139, 156, 160, 169, 244 Freudenberg, W., 74, 106, 109 Fridovich, I., 283, 303 Friebolin, H., 157, 233 Frolik, C. A., 295, 299, 306 Frosch, M., 164,172,173,178,180. 184,199, 202, 208, 229, 235, 245 Fry, J . , 41, 65 Fujii, T., 276, 277, 279, 280, 281, 282, 283, 284, 286, 287, 288. 293, 304,305 Fujisaki, S., 262, 264, 269, 271 Fujita, R. A, 45, 69 Fujiwara, T., 223, 237 Fukuda, H., 276,277,279,280,281,282,283, 284, 286, 287, 288, 291, 293, 304,305 Fuller, M. T., 18, 62, 67

315

Fulton, c., 14, 25, 65. 66 Funderud, S., 48, 51, 65 Furihata, K., 262, 272 Furuya, M., 14, 23, 24, 25, 26, 68, 69 Fuse, G., 279, 306 Fyfe, J. A. M., 222, 235 Gaffney, D., 213,235 Gaillard, J., 88, 107 Gallagher, M. P. 119, 120, 133, 178, 236 Gambacorta, A., 250, 261. 270, 271 Gambino, J., 21, 69 Gamian, A,, 141, 235 Gander, J. E., 171, 241 Cane, R., 276, 304 Gangarosa, E. J., 209, 235 Ganther, H. E . , 98, 106 Garcia, G. E., 74, 89. 95, 106 Gardiol, A. E., 158, 235 Gardlik, S., 83, 86, 106 Garegg, P. J., 145, 235 Cares, M., 56, 57, 64 Garland, P. B., 182, 232 Garon, C. F., 199, 243 Gartner, P., 266, 269, 272 Gaur, N. K., 128, 129, 132 Gautier, J., 56, 65 Gay, D. A., 43, 69 Geckeler, K.,259, 271 Geider, K., 220, 232 Geier, G., 168, 235 Gelder, K., 168, 235 Gelfand, D. H., 167, 168, 198,228. 229, 246 Gelpi, E., 250, 271 Gely, C., 14, 23, 31, 69 Gemski, P., 197, 235 Gervais, F. G., 217, 218, 219, 235 Ghalambor, M. A,, 157, 246 Gibbs, M., 263, 271 Gibson, B. W., 219, 241 Giegt, R. , 94, 104 Giesbrecht, P., 141, 238 Gilbert, J. M., 172, 241 Gill, D. R., 178, 236 Gill, J. F., 221, 222, 233, 234, 242 Gingold, E. C., 37, 65 Gibsburg, V., 227, 238 Giordano, G., 80,108 Glaser, L., 227, 240 Glaser, P., 225, 233 Glauner, B., 187, 237 Glazebrook, J., 205, 223, 224, 234,235, 241

Gleason, F. K., 74, 106 Glenn, A. R., 98.99, 107 Glover, L. A., 7, 60,66, 68

316

A I I ' ~ 1 I O R INDEX

Glyn, M. C., 18, 25. 29. 65, 68 Goguel, V., 10. 64 Gojobori, T., X9. 109 Gold, H., 279, 2x3. 304 Gold, L., 62, 65 Goldberg, J. B., 222, 235, 241 Goldberger. R. F., 93. 107 Goldemann, G . , 161, 162, 163, 235, 237 Goldfarb, P., 72, X9, 105 Golding. B. T., 279, 282, 303 Goldman, M. A., 50, 65 Goldman, R . C., 141, 143, 213, 235 Golecki, J. R . , 177. 181, 182, 183, 238, 240 Gollakota, K. G . , 276, 304 Gonzalez-y-Merchand, J. A , , 8, 65 Goto, M., 277, 281, 286, 287, 288, 293, 304, 305

Gotschlich, E. C., 140, 184. 235, 241 Gottesman, S.. 199, 203, 205, 206, 216, 217, 218, 219, 220, 223, 232, 235, 241, 244 Gottschalk, G . , 73, 74, 80, 86, 106 Gough. C. L., 223, 244 Gough, D. P., 262, 271 Gough, S., 93, 108 Govan, J. R. W . . 222, 234, 235, 240 Grabert, E.. 184, 235 Grafe, U., 262, 271 Graham, J. A , , 142. 243 Graham, L. L., 136, 137, 142, 182, 203,207, 232, 235, 238. 246 Grahame, D. A . 77, 78, X I , 83, 84, 104 Grant, W. D . , 37, 65. 69 Grasdalen, H., 167, 243 Graveline, J. F., 140, 238 Gray, A., 34, 63 Gray, J. X.,229, 235 Gray, L . , 178, 236 Green, L. L., IS, 18, 19, 25, 43, 45, 65 Grcril, A., 295. 305 Gricrson, D . , 292, 205. 301, 304 Gritz, L., 59, 65 Grobner, P.. 41, 44, 46, 55, 65, 66 Gross, M., 168, 235 Grossman, A , , 72, 89, 106 Grossman, A. D., 119, 120. 132, 133 Guerry. P., 197, 235 Guihard, G . , 187, 235 Gull, K., 13, 14, 15. 16, 17, 18, 19. 20, 21, 22.23. 25.27, 2n, 29, 3 1 , 39,42, 43. 63, 64,65, 67, 68, 69 Gumpfert, J., 262, 271 Gunderson, G . G . , 16, 65 Giinzler, W. A., 72, 89, 106 Gustafsson, C. E. D., 95. 105 Gutberlet, T., 141, 142, 237 Gutmann, M., 80, 105

Gutnick, D., 216, 217, 244 Gutteridge, J. M . C., 283, 304 Guttes, E.. 54, 65 Guttes, S., 48, 54, 65, 67 Guy, M., 287. 304 Hackenberg, H., 76, 77, 80, 107 Hackett, J., 161, 176, 193, 196,208,233, 235 Hagervall. T. G., 95, 105 Hahm, D. H., 277, 304 Hahn, J., 119, 132 Haigh, W. G., 250, 271 Haishima, Y., 141. 235 Hakamori, S., 149, 205, 239 Holboth, S., 84, 85, 04, 95, 106 Hale, T. L., 197, 235, 242 Halliwell, B., 283, 304 Hallmann, D., 2.59, 270 Hamada, Y.,26,66 Hamaoka, T., 196, 237 Hamelin, M., 37, 65 Hamilton, A. J . , 292, 295, 304 Hamlin, J. L., 51, 52, 65 Hancock, R. E. W.. 142, 184, 238, 245 Hiinel, F., 262, 271 Hansen, A , , 172, 173, IN,202, 208, 245 Hansen, T. A,, 87, 88, 109 Hanstein, C., 130, 133 Hanus, F. J., 72, 73, 87, 105 Hara, H., 264, 269, 271 Harada, T., 159, 236 Haraguchi, G . E. 191, 192, 196, 210, 213, 231. 234, 235

Harding. N. E., 158, 170, 171, 205, 235. 240 Hardman, N., 7, 8, 44, 60,65, 66, 68 Hardy, K. R., 177, 182, IW,201, 232 Harel, A , , 39, 41, 65 Harford. S., 299, 305 Harney, J . W., 95. 96. 104 Harper, S. H. T., 279, 305 Harris, R., 182, 235 Harrison, P. R . , 72, 89, 105 Hart, C. A , , 220, 221, 231 Hartmannis, M. G. N . , 97, 106 Hartner, T., 253, 254, 255, 256, 270, 272 Hartwell, L. H . , 37, 48, 58, 65 Hasegawa, M., 10, 65 Hasegawa, T., 164, IYI, 192, 204, 221, 238, 245

Hashimoto, Y.. 220, 236 Hasimoto, T.. 295, 300, 305 Haskin, M. A , , 154, 231 Hassid, W. Z., 166, 239 Hassler, R. A , , 170, 171, 205, 236, 245 Hatano, S.. 13. 64, 65 Hatch. C. L., 46, 67

AUTHOR INDEX

Hatch, L., 162, 163, 240 Hatfield, D., 73, 106 Hatfield, D. L., 73, 94, 106, 107 Haug. A., 166, 236, 238 Haugli, F., 11, 12, 16, 48, 51, 65, 67, 69 Haugli, F. B., 10, 13, 64,66 Haugli, K., 11, 12, 67 Havercroft, J. C., 13, 14, 23, 65, 68 Hayaishi, O., 291, 304 Hayakawa, Y.,262, 272 He, S. H., 85, 106 Heasley, F. A , , 162, 236 Heath, B., 39, 65 Heath, E. C., 139, 156, 157, 160, 169, 244, 246 Heidcr, J., 71, 72, 76, 78, 80, 89, 9 I . 93, 94, 95,96,97,99, 102, 104. 105, 106, 108, 109 Helinski, D. R., 61, 64, 180, 243 Hellerqvist, C. G . ,141, 150, 151, 212, 236, 239 Helmann, J. D., 119, 132 Helsinki, D.R., 146, 168, 198, 234, 237 Hemming, F. W., 262, 271 Henderson, E. R.,8, 65 Henderson, N. M., 171, 205, 245 Hendrickson, W., 225, 234 Hendrickson, W. A , , 97, 106, 109 Henning, U., 259, 260, 271 Heppel, L. A., 97, 104 Hermans, M., 253, 254, 255, 256, 259, 271 Hermondson, M. A , , 178, 236 Herrmann, D., 252, 272 Herzog, M., 10, 66 Hess, J . F., 218, 232 Heubner, V. D., 46, 65 Heuzenroeder, M. W., 184, 191, 102, 210, 231. 232, 236, 239, 240 Heyns, K., 145, 236 Hibler. N. S., 222, 224, 234 Higa, H . H., 170, 236 Higashi, B., 146, 231 Higashi, Y., 154, 236 Higerd, T. B., 128, 132 Higgins, C. F., 119, 120, 133, 178, 182, 224, 232. 236 Higgs, S. A , , 112, 132 High, N., 182, 200, 201, 242 Hignett, R. C., 144, 243 Higuchi, T., 154, 156, 169, 245 Hildebrandt, A., 6,41,49, 55, 56, 57.65.66 Hiles, I. D., 178, 236 Himawan, J., 205, 229, 239 Hinton, J. C. D . , 224, 236 Hirota, Y., 269, 271 Hirth, K. P., 8, 69 Hisamatsu. M., 146, 159, 168, 231, 236, 238

317

Hitchcock, P. J., 141, 236 Hitchin, E., 220, 233 Hobohm, U., 41. 55, 56, 57, 65 Hobot, J . A , , 136, 182, 236 Hoch, J . A., 113, 114, 115, 116, 117, 118, 119, 120,122, 123, 125. 126, 127, 128, 129. 130, 132, 133 Hodge, R., 177,178, 182, 199,200,201,232, 242 Hodgson, K. O., 97, 106 Hoffman, J., 150, 151, 236 Hoffman, J. L., 97, 106 Hoffmann, N. E., 287, 304. 305, 306 Hohman, S., 295, 305 Hoiby, N., 144, 149, 238 Hoiseth, S.K., 225, 226, 236 Holbein, B. E., 159, 240 Holcenberg, J. S., 87, 106 Holdsworth, M. J., 295, 301. 304 Holland, I. B., 178, 236 Hollibaugh, J. T., 100, 107 Hollingsworth, R. I., 196, 246 Holme, T., 145, 150, 151, 235, 236 Holmgren, A., 74, 98, 105, 106 Holmlund, C. E.. 258, 272 Holst, 0.. 141, 236, 252, 253, 272 Holt, C. E., 5, 27, 28, 29, 52, 67, 68, 69 Holt, S. C., 224, 234 Holtje, J . -V., 185, 186, 236. 245 Homonylo, M. K., 143,203,204,207,237,238 Honort, N., 221, 240 Hopklns, I., 178, 202. 226, 238 Horbach, S., 253, 257, 271 Horecker, B. L., 154, 156, 169, 245. 246 Horii, T., 204, 221, 245 Horiuchi, K., 264, 269, 271 Hormann, K . , 74, 106 Horton, J. R., 97, 106 Host, 0.. 141, 235 Hotte, B., 206, 211, 237, 238 Hottiger, T., 279, 284, 287, 304 Houng, H.-S.H, 220, 237 Howard, A , , 41, 65 Howard, S. P., 187, 232 Howden, L., 193, 241 Hoyle, H. D., 21, 65 Huala, E., 218, 231 Huang, J., 46, 65 Huber, R. E., 97, 99, 107 Hudman, J. F., 98, 99, 107 Huetterman, A., 27, 67 Hughes, C., 212, 231 Hughes, D. W., 144, 231 Hugirat, Y.,228, 242 Hull, R. A., 191, 192, 196, 210, 213, 231, 234, 235

318

A U T I I O R 1NDP.X

Hull, S. I.. 191, 192, 196, 210,213, 231, 234, 235

Hulton, C. S . J . , 224, 236 Hung. L., 10. 66 Hunt, F., 213, 235 Hunt, J . M., 276, 304 Hurst, L. D., 12, 65 Hussey, H., 172, 231 Hutchins, G., 34, 63 Hutzler, M . , 245 Huynh, B. H., X4, 85. 86, 96, 106 Hyde, S. C., 178, 236 Hyodo, H., 277. 281, 286, 288, 304 Ichihara. A,, 287, 304 Ielpi, 146, 158, 168, 170, 171, 180, 198, 234, 237, 240, 243

lida, K.-L.., 164, 191, 192. 238 lida, Y.,10, 65 linuma, H., 262. 272 Ikemura, T.. 89, 109 Ikeuchi, T., 114, 115, 132 Ilag, L., 279, 280. 304 Imhoff, D . , 87, 107 Imperiale, M. J . , 293. 304 Ina, K . , 163. 241 Incardona, J . P., 21, 66 Ince, J. E., 277. 281, 282, 284. 304 Ingledew, W. J., 102, 104 Inglis. R. J.. 55, 56, 63 Ingolia, T. D., 295, 298, 303, 304,305,306 Innerhofer, A , , 76, 77, 80, 107 Inoue, Y.,304 Isenberg, G., 7, 67 Ishaq, M.,263, 270 Ishibashi, F., 89, 109 Ishida, Y.,277, 281, 286, 288, 304 Ishidate, K., 187, 237 Ishiguro, E. E., 212, 237 Ishihara, K., 304 Isshiki. K . , 262, 271, 272 Ito, E., 144, 146 Ito, H., 164, 191, 192, 204, 221, 231, 238 Ito, J., 128, 132 luchi, S . , 223, 237 Iwabuchi. M., 10, 65 Jackson, G . D. F., 209, 241 Jacobitz, S . , 73, 88, 107 Jacobson, B., 73, 107 Jacq, C., 10. 64 Jahn, K., 234 Jalouzot, R., 50. 67 Jalouzot, T., 49. 65 Janas, T., 244 Janas, T., 244

Jankola, M., 161, 239 Jann, B., 138, 140, 141, 142, 143, 147, 151, 152, 153, 157, 160, 162, 163, 164, 165, 166. 173, 177, 182. 183, 185. 191. 192, 193, 197, 198. 199,202,203.210,233,234,235,237, 238, 239. 241. 242, 243, 244.245

Jann, K., 138, 140, 141. 142, 143, 147, 151. 152, 153, 157, 160, 162, 163. 164,165. 166. 173, 176, 177. 178. 181, 182, 183. 185, 191, 192, 193, 197, 198, l99,200,201,202, 203. 207, 2 10, 232, 233, 234, 235, 237, 238, 239. 241. 242, 243, 244,245 Jansson, P. -E., 142, 147, 237 Jayalakshmi, B., 276, 303 Jayaraman, J . , 262, 272 Jayaratne, P., 203, 219, 237 Jennings, H. J., 152, 232 Jensen, S. E., 299, 304,306 Jerzmanowski, A.. 45, 46, 65, 66 Jiang. X.-M.. 172, 173, 176, 192, 193. IYS, 206, 208, 21 I , 237 Jockusch, B. M.. 37, 69 Johansen, S., 8, 10, I I , 12.66, 67 Johansen, T., 10, 11, 12, 66, 67 John, P.. 287, 306 Johnson, A . W. B., 205, 229, 231, 232, 238 Johnson. E. M., 52, 53, 58.64, 68, 220, 237. 243

Johnson, J . G., 156, 160, 237 Johnson, K. A , , 22, 63 Johnson, L., 17, 19, 63 Johnson, R. T., 54, 68 Johnson. W. C., 127, 133 Joiner, K., 143, 213, 235 Jonas, R., 130, 133 Jonas, R. H., 126, 133 Jones, A. D., 255, 256, 270 Jones, E. P., 1&1 I , 66 Jones, J . B., 76, 78, 107 Jones, K. B., 56, 63 Jones, M. A , , 205, 231 Jones, R. T., 191, 237 Jones. R. W., 80, 107 Jonsson, Y. H., 95, 105 Jung, M. K., 33, 69 Junio, L. M.,197, 242 Jurgens, U . J., 252. 272 Jurgens. U . W., 258, 271 Kadam, S . K., 175, 208, 239 Kaim, W., 83, 84. 87, 101, 107 Kaiser, I. I . , 97, 109 Kaletta, C., 269. 270, 272 Kallio, P., 130, 133 Kallio, P. T., 128. 129, 132 Kalman, E. T.. 97, 109

AUTHOR INDEX

Kanegasaki, S . , 160, 161, 162, 163, 169, 170, 172. 176, 235. 237. 246 Kang, K. S . , 205, 235 Kannangara. C. G., 93, 108 Kannenberg, E., 253,255,256,259.271,272 Kao, C. C., 196, 205, 237 Kao, I.., 245 Kapfer, C., 197, 235, 242 Kapulnik, E., 279, 303 Karageorgos, L. E., 208, 209, 244 Karmazyn Campelli. C., 131, 133 Karst, F., 268, 272 Kashyap. L . , 146, 198, 234 Kastowsky. M., 141, 142. 237 Kat, B., 213, 239 Kates. M., 262, 271 Kato, J . , 146, 166, 167, 184, 192, 197, 198, 199, 221, 222. 224, 225, 234,237, 240, 246 Kato, N.. 163. 164. 191, 192, 204, 221, 231, 238, 241, 245

Katsuki, H., 262. 264, 269, 271 Katzir, N., 54, 67 Kawaguchi, M., 26, 66 Kawahara, K.. 196, 237 Kawamura, F., 113, 122, 123, 124, 126, 132, 133

Kawamura, T., 144, 243, 246 Kawano, S . , 5 , 10, 11, 12, 24, 66,67, 68 Kaya, S., 144, 246 Kazantsev, Y. Y.. 252, 271 Kazarinoff, M. N . , 58, 66 Kearny. J . J . , 77, 107 Keck, W . . 187, 232 Keenleyside, W. J . , 203, 219, 237 Keister, D. L., 168. 232 Kellenberger, E., 136, 182,186, 188,236,237 Keller. J . M..170, 198, 237, 242 Keller. M., 205, 231 Kelley. W. S., 154, 156, 233 Kellncr, R . , 266. 269, 272 Kende. H . , 287, 304,305 Kcnne, L., 138, 144. 145, 14% 153, 166,232, 237, 239

Kennedy. E. P., 146. 237, 240 Kenney, T. J., 130, 133 Kenny. C. P . . 152, 232 Kent, J . L.. 161. 175, 238 Kessel. I.. 259, 271 Kessler, A. C., 192, 210, 238 Kesslcr, D . , 16, 17. 18, 22, 25, 42, 64 Khawaja, S., 16, 65 Kido, N . , 163, 164, 191, 192, 204, 221, 231, 238, 241

Kieber. J.. 196, 233 Kieffer, J. D., 95, 96, 104 Kiessleing. G., 145. 236

319

Kilesso, V . A , , 154, 156, 160. 169, 243 Kim, S. -M. A , , 72, 89, 106 Kim, Y., 46, 65 Kim, Y. S., 73, 107 Kimbara, K., 222, 224, 234, 237 Kimble, M.. 21, 66 Kingma, J., 185. 233 Kirby, A. L., 262, 271 Korouac-Brunet, J., 5, 66 Kirsch, D. G., 100, 107 Kitajima, H., 284, 304 Kitano, K., 222. 224, 237 Kjosbakken, J., 167, 244 Klein, A , , 84, 85, 94, 95, 106, 107 Klein, S., 196, 246 Kleinig, H., 262, 271 Kleinkauf, H., 299, 306 Klena, J. D., 175, 176, 191, 208, 238, 243 Klimmek, O., 78, 80, 107 Kloser, A. W., 219, 241 Knights, J . M . , 152, 231 Knoche, K., 10, 66 Knowles, C. J . , 277, 281, 282, 284, 304 Kobayashi, N., 98, 107 Kobayashi, T . , 26, 66 Kobayashi, Y., 123, 124, 126, 133 Kochanski, R. S., 14, 66 Kochtov, N. K., 151, 210, 239 Kock, M., 295. 304 Koeltzow, D. E., 197, 235 Koeltzow, E. E., 191, 237 Kofoid, E. C., 114, 132 Kogan, L. M., 252, 271 Kohama, K., 26, 38, 66,69 Ktihrle, J . , 73, 96, 104 Koizumi, K., 146, 168, 238 Kojro, E., 80, 105 Kol, O., 147, 149, 238 Kolyva, S . , 220, 238 Komatsu,T., 164,191,192,204,221,23l,238 Kondo, M., 254, 269, 273 Kondo, N., 279, 280, 306 Konig, K., 86, 107 Konishi, K., 26, 66 Konyecsni, W. M., 221, 222, 224, 226, 233, 234. 238. 240

Kopecko, D. J., 146, 197,220,227,238,241, 243

Kdplin, R . , 1W. 206, 211, 233, 238 Kopmann, H. J . , 162, 165, 238 Korhonen, T. K . , 182, 201, 242 Kornfeld, R. H . , 227, 238 Koronakis, V . , 212, 231 Korsch, M. J., 209, 241 Kortner, C., 80. 105, 107 Koshland, D., 183, 238

320

AUTHOR INDEX

Kovacevic, S., 295, 299, 304 Kozlova, 1. V., 252, 271 Kraft, A . A,, 97, 99, 109 Krajewski-Bertrand, M. - A , , 259, 271 Krallmman-Wenzel, U., 191, 192, 210, 235 Kramer, G . F.,76, 88. 89, 95, 107 Krammer, G., 21. 66 Krausse, B., 172.178, I80,184,202,208,235 Krems, B . , 78, 80, 107 Krepinsky, J. J., 144, 231 Kretschmer, P. J., 199, 243 Krezel, A . M., 45, 65 Kroeplir-Rueff, L., 260, 271 Kroger, A,, 76, 77, 78, 80, 105, 107 Kroll, J. S . . 178,202,207,208,225,226,229, 234,238,246 Kroll, S. J., 226, 232 Kronke, K. D.,177, 181, 182, 183, 185, 238 Kropinski, A. M., 144, 149, 198, 212, 216, 232,238 Kropinski, A. M. B., 144, 231 Kriiger, B.. 73. 88, 107 Kubbies, M., 6.49. 66 Kudoh, J.. 114, 115, 132 Kueng, V.,45, 64 Kuhn, H. -M,, 140, 142, 143, 144, 175, 207, 238,239 Kull, F. J., 98, 99, 107 Kulpa, C. F., 185, 238 Kumar, S., 98, 105, 106 Kundig, F. D.. 164, 238 Kunkel, B., 205, 234 KUO, J . S. -C.. 140, 238 Kurahashi, K., 114, 115, 132, 169, 242 Kurata, H., 94, 107 Kuroiwa, T., 5 , 10, 11, 12, 24, 66, 67. 68 Kusecek. B., 184, 231 Kushwaha, S. C.,262, 271 Kusser, W., 212. 237 Kusuzaki, K., 163, 241 Kutsuki, H., 279, 283, 304 Kuzio, J., 198. 238 Kwak, M. Y.,277, 304 Kyraikopoulos, A . , 73, 96, 104 Kyte, L., 300, 301, 304

Lam, M. Y. C., 142, 144, 149, 238 Landrey, J. R. 258, 270, 271 Langan, T.A., 55, 56, 63 Langworthy, T. A., 250, 251, 252, 271 Larm, O., 150, 236 Laroche, A,, 8, 26, 37. 38, 50, 51, 61, 63, 67,68 Larsen, B., 166, 167, 236, 238,243 Larsen, P. R., 73, 95, 96, 104 Latchford, J. W., 229, 232,238 Lauterback, F.,80, 105 Lawson, C. J., 167. 170, 233, 243 Lazdunski, C.,187, 232 Le Minor, L., 196. 239,241 Leaveslay, D.I . , 192, 239 LeCoq, D., 114, 115, 132,133 Ledizet, M.,20, 66 Leduc, M., 186,239 Lee, B. J., 73, 94, 106, 107 Lee, C. C., 149, 180.204,205,223,224,229, 233,239. 246 Lee, K. J., 255, 272 Lee, S. J., 172, 173, 176, 192, 193, 195, 203, 206, 208, 21 I , 237,239,244 Lees, M.,253, 271 Lefebvre. P. A., 25, 66 LeGall, J., 84, 85, 86, 96, 106, 107, 108 Leigh, J. A., 138, 149, 180, 199, 204, 205, 220, 223, 224, 229, 233, 234, 239,244 Leinfelder, W., 72,78, 89,90.91,92,93,94, 95, 96, 97,98, 105, 106, 107, 109 Leitch, R. A,, 147, 241 Leive, L., 141, 143, 185, 213, 235, 236, 238 Lejeune, P., 225, 233 Leloir, L. F.,158, 159, 244 LeMaster, D. M., 97,106 Lemieux, G., 8, 26, 37, 38, 50, 51, 59, 61, 63.64,65,66,67,68 Lenaers, G., 10.66 Leonard, J . L., 96, I 0 8 Leonhardt, U., 76, 80, 107 Leontein, K., 147, 237 Leroux, B., 223, 239 Leskiw, B. W., 299, 304 Lespinat, P. A., 84, 85, 86, 96, 106, 108 Lester, R. L. 76, 77, 78, 80. 81, 83, 106 L'Hemault, S. W., 62, 66 Letellier, L., 187, 235 L'Vov, V. L., 151, 210, 239 Leu, T.-Z., 51, 52, 65 Laakso, D., 142, 239 Levengood, S. K.. 187, 239 Laakso, D. H.,203, 204, 207, 238 Levery, S. B.,149, 204,205, 223, 224, 239, Labischinski, H., 141. 238 246 Ladeaux, J . R., 119, I33 Levin, B. R., 210, 233 Laffler. T. G., 6, 42, 45, 64. 66 Levinthal, M..169. 246 Lai, E. Y., 25, 66 Levy, G. N.,162, 244 Laishley, E. J., 98, 99, 105 Lew, H. C., 207, 239 Lam, J . S . , 142, 143, 144, 149, 231,237,238 Lewandoski, M., 124, 132

AUTHOR INDEX

Lewin, B., 56, 66 Lewis, M., 46,67 Lewis, M. S., 143, 199, 235, 246 Lewis, S. . A , 21, 69 Li, J., 78, 80, 104, 146, 220, 243 Li, N . , 220, 236 Liback, L. B., 197, 242 Licht, A., 97, 106 Lieberman, M., 279, 303, 304,305 Lifely, M. R., 152, 239 Lightfoot, J., 142, 238 Lin, E. C. C., 223, 237 Lindberg, A. A., 141,150,151,195.2l2,236, 239,240,245

Lindberg, B., 138, 144, 145, 147, 148, 150, 151, 153, 166, 232, 233, 235, 236, 237, 239 Lindblow-Kull, C . , 98, 99, 107 Lindon, J. C., 152, 231, 239 Linker, A , , 167, 239 Linqvist, L., 195, 240 Lipman, D. J., 295, 304 Liras. P., 295, 299, 303 Little, M., 16, 21, 66, 68 Liu, D. E., 208, 21 I , 239 Liu, H. W.. 287, 305 Liu, S.-M., 77, 80. 109 Liu, T.-Y., 140, 164, 165, 170, 235, 245 Liu, Y.-N., 223, 244 Ljungdahl, L. G., 76, 77, 80,104, 107, 109 Llopiz, P.,252, 271 Logan, K. A., 44, 69 Lohka, M., 56, 65 Loidl, P., 41, 44, 46,54, 55, 57, 65, 66 Long, E. O., 8, 66 Long, R. A , , 213, 239 Long, S., 205, 223, 229, 239, 241 Long, S. R . , 198, 234 LGnngren, J., 147, 150, 151, 233, 236, 237 Lopez, J . , 186, 239 Lory, s., 222, 240 Losick, R., 120, 127, 130, 132, 133 Loughlin, R. E., 98, 106

Lowery, R., 10, 66 Loynds. B., 178,202,207,225,226, 229,238 Liideritz, 0.. 161, 239 Luduena, R. F., 21, 66 Luengo, J. M., 165, 242 Lugowski, C., 145, 239 Lugtenberg, B., 137, 239 Lugtenberg, B. J. J., 205, 233 Lui, H., 192. 197, 246 Lustre, V. M., 119, 132 Lutz, M., 88, 107 Lycett, G. W., 292, 304 Lynch, J . M., 250, 270, 279, 305 Lynen, F., 259, 260, 271

32 1

Ma, C., 51, 52, 65 MacAlister, T. J., 187, 237 MacDonald, L. A., 142, 143, 144, 149, 237, 238

Mackie, K. L., 157, 234 MacLachlan, P. R., 175, 191, 208, 239, 243 MacLachlan, R. P., 203, 219, 237 MacLean, D. B., 144,231 MacLean, L. L., 143,147,149, 190,232,239, 246 Macpherson, D. F., 192, 197, 239 Macy, J. M., 100, 107 Maest, A. S., 100, 107 Maharaj, R., 146, 166, 167, 184, 198, 199, 221, 240

Mahendran, R., 10, 11, 12, 66 Maiorino, M., 73, 108 Majerczak, D. R., 190, 204. 233, 234 Makela, P. H., 138, 141, 149, 161, 163, 169, 170,175,184,190, 191, 193, 197, 198,206, 207, 228,236,237,239, 240, 241, 243,244 Makini, F. W., 205, 233 Maleszewski, M., 45, 46,66 Maller, J., 56, 65 Mallory, F. B., 258, 270, 271 Mandel, S. J., 95, 104 Mandelstam, J., 112, 132 Mandrand, M.-A,, 80, 108 Mandrand-Berthelot, M.-A., 90, 91, 92, 95, 107 Mann, J., 250, 271 Manning, P. A., 191, 192, 197, 198, 208,2W, 210,212,213, 232, 233, 236,239, 240,244 Mansfield, J . W., 220, 233 Mansouri, S., 277, 279, 305 Marais, D. J . D . , 276, 305 Marceau-Day, M. L., 212, 233 Marcum, J. M., 22, 63 Margulis, L., 2, 66 Marino, P. A , , 173, 181, 239 Markert, C. L., 56, 67 Markovitz, A , , 146, 149, 166, 216, 240 Marolda, C. L., 192, 193, 196, 197,208,210, 21I , 240, 245 Marquez, L. M., 119, 132 Martel, R., 38, 66 Martin, A , , 152, 232 Martin, D. W., 221, 222, 224, 225, 234, 240 Martin del Rio, R., 72, 74, 105 Martin, J. F., 295, 299, 303 Martin, W., 295, 305 Martinetti, G., 180, 233 Marumo, K., 195, 240 Marzocca, M. P.,158, 170, 171, 240 Maskell, D. J., 190, 240 Mason, J., 213, 234

322

AlJ'I'HOR INDEX

Masson, L., 159, 240 Masson, S.. 5, 66 Masubo, Y . , 144, 243 Masui, Y.. 56, 67 Mathan, V., 196. 235 Matsuda, J., 295, 300, 305 Matsuda, Z., 223. 237 Matsuhashi, M., 154, 231 Matthews, H. R., 41, 44, 45, 46, 55, 56, 57. 63, 65, 67, 69

Mattingley, S. J . , 226, 246 Mattoo, A. K., 279, 303, 305 Maurizi, M. R.,216, 217, 244 May, G. S.. 21, 69 May, H. D., 78, 104. 108 May, T. B., 146, 166, 167, 184, 198, 199,206, 221, 233. 240, 242, 243

Mayberry, W. R.,250, 251, 252, 271 Mayer, F., 74, 106 Mayer, H.,138, 140, 142, 143, 144, 161, 162, 163, 164,175, 191.206,207.238,239,240, 242,243

Mayer, R., 167. 168, 198, 228. 229, 242, 246 Mayne, R. G . , 287, 305 McBain, W., 72, 89, I05 McCallum, K. L., 142, 220. 221, 239 McCammon, S. L.. 204. 239 McCloskey, M. A., 165, 176, 239, 244 McConnell, K. P.,97, 106 McConnell, M., 212, 213, 216, 239 McCoy, D. W., 140, 238 McCready. R. G . . 102, 107 McCurrach. K., 7, M), 66 McCurrach, K. J., 7, 68 McDonald, I. J . , 212, 234 McElhaney, R. N.. 259, 271 McGarvey, D. J., 287, 295, 300, 305 McGrath, B. C., 173, 175,176, 181,213,239 McGroarty, E. J . , 143, 144, 149, 212, 238, 239, 241, 242

McGuire, E. J., 152. 239 McGuire, S. L., 58. 67 McKean, M. L., 259. 260, 271 McKeon. T. A,. 276. 287. 305 McLean, R. J . C., 143. 232 McNeil. M., 145. 231 McQuade, B. A., 209, 305 Mdlui, K. E . , 163, 232 Meade, J . H., 167, 168, 198, 228, 229, 246 Medgley, M., 263, 270 Meier, U., 206, 240 Meier-Dieter, U., 140. 142, 143, 144, 161, 162, 163, 164, 175, 206, 238, 240 Meinhold, H., 73, 96, 104 Mekalanos, J . J., 218, 240 Meland, S., 11. 12, 67

Meldgaard, M., 295, 305 Melo, A., 227, 240 Meno, Y.,142, 231 Menon, N. K., 84, 85, 107, 108 Menzel, J., 185, 240 Mercer, A., 1x4, 231 Merker, R. I . , 244 Merrifield, E. H., 157, 234 Messen, A., 295, 305 Metzenberg, R. L., 5, 34, 67 Mevarech, M., 299, 304,306 Meyer, M., 73, 74, 76, 89, 105 Meyer, O., 73, 87, 88,107 Meyer, T. F., 164, 180, 202, 229, 235 Michaeli, D., 228, 242 Michel, G. P. F., 269, 271 Michel, T. A,, 100, 107 Middleditch, B. S . , 250, 270 Miller, D. L., 10, I I, 12, 270 Miller, E. V., 277, 279, 305 Miller, J., 196, 223, 231, 232 Miller, J. F., 218, 240 Miller, J. R., 295, 299, 304,305 Miller, K. J . , 146, 240 Miller, L. E., 209. 242 Miller, L. G., 100, 107 Miller, R. J . , 276, 304 Mimmack, M. L., 178, 236 Minale, L., 250, 261, 270 Minden, J., 30, 68 Mir, L., 13, 14, 23, 26. 31, 33, 63, 67, 69 Mirel, D. B . , 119. 132 Mischke, S . , 73. 106 Mishalanie, E. A , , 102, 105 Misra, T. K.. 146, 166, 167, 184, 198, 199, 221, 222, 224, 233, 234, 237, 240 Mitani, H., 146. 231 Mittermayer, C., 48, 63 Mizuno, T., 218, 240 Mizushima, S . . 218, 240 Mizutani, T., 94, 107 Mohberg, J., 6, 39, 67 Mohberg, J. M., 41, 65 Mohr, C. D., 221, 222, 224. 225, 234, 240 Moir, A , , 119, 133 Moisand, A . , 13, 14, 20, 23, 31, 33, 41, 43, 63. 68, 69

Monack, D., 223, 244 Monteiro, M. J . , 8, 15, 19, 67, 69 Montreuil. J., 147, 149, 238 Moos, M., IW, 246 Moran, C., Jr., 118, 132 Moran, C. P., Jr, 119, 130, 133 Moreau, R. A., 255, 256, 270 Moreno, C., 152, 239 Morgan, J. E., 46,69

AUTHOR INDEX

Mori, K., 12, 66 Mori, M., 204, 221, 231,245 Morino, Y.,277, 281, 288, 293, 304,305 Morishima, H.,262, 272 Morona, R., 192, 197, 208, 209, 239,244 Moms, E. R . , 170, 243 Moms, N. R.,21, 24, 58, 67,69 Morris, S. R., 38, 68 Morrison, D. C., 141. 236 Morschel, E., 84, 107 Moslein, E. M., 259, 271 Mottonen, J . M., 115, 133 Moulis, J.-M., 88, 1M Mountford, R., 178, 182, 200, 201, 242 Moura, I . , 84, 85, 86, 96, 106,108 Moura, J. J . G., 84, 85, 86, 96, 106, 108 Moxon, E.R.,178, 190, 202,207,208,225, 226, 229, 232,234,236, 238,240, 246 Mueller, R. D., 44, 46,67,69 Mugridge, A., 152, 231 Miihlradt, P. F.,185, 240 Mukhopadhyay, M. J . , 60.64 Mulford, C. A,, 175, 240 Miiller, A , , 180, 184, 235 Miiller, D., 180, 184, 235 Miiller, L., 172, 241 Miiller, M., 181, 183, 242 Miiller, S., 92, 106 Munoz, L. E., 122, 132 Munro, S. M., 144,243 Murakami, H.,286, 305 Murayama, S. Y.,196, 237 Murphy, D. B., 22, 63 Muscarella, D. E.,9, 67 Muth, E., 84, 107 Mysliwiee, T. H., 119, 132 Nader, W. F., 7, 27, 67 Nagahama, K., 277, 279, 281, 287, 288, 293, 304,305 Naganawa, H., 262, 271,272 Naide, Y.,161, 193, 240 Nainudel-Epszteyn, S., 39, 41, 65 Nair, G., 118, 119, 124, 125, 127, 132, 133 Nakae, T.,169, 227, 228, 240 Nakagawa, T., 98, 107 Nakamura, M., 196, 237 Nakatani, Y.,259, 270 Nakayama, M., 98, 107 Nakayama, T., 122, 132 Nakhla, N. A,, 144,231 Nanba, T.,10,11, 66 Nanninga, N., 136, 182, 235 Nano, F. E.,163, 232 Nassif, X., 221, 240, 245 Nath, K.,142, 243

323

Naumann, D., 141, 238 Neal, B., 172, 173, 176, 192, 193, 195, 206, 208, 211, 237 Neal, B. L., 192, 240 Nelson, D., 212, 240 Nemethy, E. K., 276, 305 Nes, W. R.,259, 260, 271 Nester, E. 146, 198, 234 Nester, E. W., 180, 198, 223, 233, 234,239 Neumann, H.,93, 107 Neumann, S., 266, 272 Neumeyer, B. A., 138, 162, 191, 242 Neunlist, S., 252, 253, 268, 271, 272 Neuss, B.,253, 254,255, 256, 257, 269, 259. 271,272 Newlon, C. S., 52, 67 Newport, J., 56, 57, 65 Nicholson, W. L.,131, 132 Nickerson, W. J., 279, 305 Nicolaus, B., 261, 271 Niehaus, K., 190,205, 231,233 Nielsen, H.,8, 10, 65,66 Nielsen, P. E., 275, 305 Nikaido, H., 137, 156, 161, 169, 183, 185, 193, 207, 227, 228, 239,240, 243,246 Nikaido, K., 156, 169, 240 Nikolaev, A. J . , 164, 165, 166, 173, 202, 234 Nimmich, W., 147, 233 Ninfa, A. J . , 113, 114, 133,218, 244 Nishimura, H . , 275, 305 Nishimura, O., 140.235 Nishimura, Y.,264, 269, 271 Nishino, T., 262, 264, 269, 271,272 Nishiyama, K.,277, 305 Nixon, T. B.,218, 242 Nock, S., 88, 108 Noel, K. D., 196, 232, 233 Noodleman, L.,88, 107 Noon, K. F.,220, 237 Norbury, C., 56, 65 Norval, M., 156, 160, 244 Nozaki, M., 291, 304 Nozawa, Y., 258, 273 Nunes-Edwards, P., 162, 231 Nurse, P., 37, 54, 56, 57, 64,67 Nygaard, O.,47, 68 Nygaard, 0. F., 48, 67 Nyman, K., 193, 199,241 O’Brochta, D., 180, 243 Oakley, B. R.,24, 33, 67,68 Obolnikova, E. A,, 252, 271 Ochs, D., 266, 269, 270, 272 Oetliker, M . , 16, 20, 21, 64 Ogasawara, N., 119, 132

324

AlITI{OR INDEX

Ogawa, T.. 276,277.279.280. 281,282, 283, 284.286,2n7.28~,2~3,304,305 Ohana, P., 228, 242 Ohba, Y.,277, 305 Ohman, D. E., 167. 222, 233, 234, 235, 241, 246 Ohta, M., 163, 164, 191, 192, 204, 221, 231, 238, 241, 245 Ohta, T., 24, 67 Okabe. S . , 295, 300, 305 Okada. Y.. 146, 168, 238 Okamura, N., 192. 197, 246 Olive, L. S., 4, 67 Olmsted, J. B., 22, 63,67 Olson, D., 288, 292, 295, 301, 303 Ornata, T., 304 Onn, T., 145, 235 Oppenheirn, A., 54, 67 Oppenheim, J . , 129, 132 Ordal, G. W., 119, 132 Orekh, G . T.. 252, 271 Oremland, R. S . , IOO, 102, 107, 108, 109, 276, 305 Orihara, K.,262, 272 Oro, J . , 250, 271 Oroszlan, S., 94, 106 Orr, E. C., 78, 108 Orskov,F., 153, 184. lY9,203,210,233,24l, 243,245 Orskov, I . , 143, 153. 184, 199,203,210,233, 235, 241, 243 Ortiz, A,, 165. 242 Osborn, M. J., 154, 156, 160, 161, 169, 171, 172, 173,175, 176, 181, 183, 185,213,238, 239, 240, 241, 245. 246 Osborne, D. J., 270, 306 Osbourn, A . .E. 223, 241 Osmani, A . H., 58. 67 Osmani, S. A . , 58, 67 Osterberg, L., 46, 65 Ott, G . , 91, 93, 106, I08 (itting, F., 72. 89, 106 Otvos, J. W . , 275, 305 Ou. J . T., 220, 227, 237, 241 Oulmouden, A,, 268, 272 Ourisson. G., 250, 251. 252, 253, 254, 255, 258, 259, 266, 270, 272 Oustrin, M. L.. 14, 23, 31, 33, 37.63, 64,69 Owen-Hughes, T., 224, 236 Owens, J.. IS. 68 Owens, L. D., 73, 106 Oxley, D., 147. 241 Paakanen, J . , 184, 241 Pachlatko, J. P., 262. 270 Pachter, J. S., 43. 69

Pagh, K. I . , 14, 23. 26, 67 Pakula, A. A , , 218, 232 Pallotta, D., 5,8,26,37,38, 50, 51,59,60,61, 62, 63, 64,65, 66, 67, 68 Palva. E. T., 141, 241 Pandian, S., 261, 262, 272 Pang, H., 144, 231 Papen, H., 72, 73, 87, 105 Parisi, E., 171. 241 Parker, C. T., 191, 219, 241, 243 Parker, H. M.,119, 132 Parkinson, J . S., 114, 132 Parolis, H., 147, 241 Parolis, L. A. S., 147, 241 Patel, P. S., 78, 104, 107 Patel, T. R., 144. 231 Patil, D. S.. 85, I06 Patino, C., 299, 305 Paul, E. C., 16, 20, 21,64 Paul, E. C. A.. 8, 14, 16, 17, 18, 19, 20. 22, 25, 31, 39, 42, 43. 64,67, 68 Pavelka, M.. 178, 201, 229, 241, 243 Pavitt, G. D., 224, 236 Payne, J. I., 102, 107 Pauani, C., 166,202, 234 Pearce, R., 220, 233 Pearce, S. R.,119, 120. 133, 178, 236 Pearson, N. B.. 191. 243 Pearson, W. R., 295. 304 Pecher, A., 78. 107 Pecht, I . , 97, 106 Peck, H. D., Jr, 84.85.86, 96, 106, IM.108 Pederson, S. S., 144, 149, 238 Peiseler, B., 252, 258, 268, 272 Peiser, G. D., 287, 305 Pele, S., 41, 65 Pelkonen, S., 177. 182, 241 Penalva, M. A., 299, 305 Peoples, 0. P., 226, 234 Perasso, R., 10, 63 Perea, J., 10, 64 Perego,M., 116,117, 119, 120,122,123, 127, 128, 130, 132, 133 Perry, D., 209,305 Perry, M. B., 143, 147, 149, 190. 212, 232, 234,239,241,246 Pesis, K., 46. 67 Petersen, L. A , , 268, 272 Petersom, P. A , , 295, 305 Peterson, A. A., 143, 144. 241 Petroni, E. A., 158, 170, 171, 240 Phan, K. A., 181, 239 Phoenix, P., 217, 218, 219, 235 Pierron, G., 6, 18, 26, 38,42,49, 50, 51. 67, 52, 53, 61, 62, 63, 66, 67 Pigeon, R. P., 243

AUTHOR INDEX

Piggot, P. J., 112, 114, 115, 119. 133 Pillat, M.,162, 237 Pilotti, A,, 150, 236 Pindar, D. F., 166, 241 Pinsent, J., 71, 78, 107 Piperno, G., 20, 66 Pirt, S. J.. 250, 270 Planques, V., 14, 20, 23, 31, 41, 43, 68, 69 Plosila, M., 193, 241 Poetsch, B., 16, 69 Pwtter, K.,220, 232, 241 Pogson, C. I . , 14, 22, 68 Pohl, A,, 184, 231 Politino, M., 88, 92, 95, 96, 107, 108 Pollock, T. J., 205, 244 Poole, K.,212, 216, 241 Popoff, M. Y.,196, 220, US,241 Popova, A . N., 154, 156, 160, 169, 243 Poralla, K.,250,251,253,254,255,256,259, 266,269, 270, 271, 272 Poth, H., 130, 133 Poulter, C. D., 268, 272 Poulter, R. T. M.,5, 30, 67 Poxton, I. R., 212, 240 Pradel, E., 175, 176, 208, 213, 216, 238, 241 Pradel, E. L., 191, 243 Prebble, J. N., 262, 271 Predich, M., 119, 124, 125, 127, 133 Prehm, P., 147, 241 Prescott, A. R., 4, 21, 67 Prescott, D. M., 42, 68 Presser, T. S., 100, 107 Price, J., 34, 63 Price, N. M., 73, 106 Price, V. A , , 119, 133 Prickril, B., 84, 85, 108 Prickril, 9. C., 85, 106 Primrose, S. B., 277, 279, 282, 303, 305 Prior, C. P., 58, 68 Przybyla, A. E., 84,85,96,96, 106, 107,108 Pua, E., 295, 305 Piihler, A., 190,205,206,211,231,233,237, 238

Puvion, E., 50, 51, 67 Qu, L.-H., 10, 63 Quandt, J., 205, 231 Queener, S. W., 295, 299, 303, 305, 306 Quigley, N. B., 195, 209, 245 Quinlan-Walshe, C., 218, 232 Raederstorff, D., 258, 272 Raetz, C. R. H., 138,141,163,191,232,233, 241

Raff, E. C., 21, 22, 65, 66, 68

325

Rajagopalan, K. V., 83, 86, 87, 106, 107 Rajagopalan, M., 73, 107 Raman, T. S., 261, 262, 272 Ramasarma, T., 262, 272 Ramon, D., 299,305 Ramsay-Sharer, L., 138, 162, 191, 242 Ramsden, M.,299, 305 Rao, P. N., 54, 56, 63,68 Rasmussen, N. S., 183, 185, 241 Rath-Arnold, 1. R., 206, 237 Raub, T. J., 4, 68 Ray, J., 295, 301, 304 Reamer, D. C., 102, 108 Reaves, C. B., 197, 235 Redmond, J . W., 146,205,209,231,233,241 Reed, J. W., 205, 224, 229, 239, 241 Reeny, M. A,, 78, 108 Reeves, P., 193,195, 196,208,209,235,241, 245, 246 Reeves, P. R., 172, 173, 176, 192, 193, 195, 196,197,203,206,208, 209,210,211,213, 231,232,237,238,239,240.244,245

Reglero, A., 165, 242 Reid, W. W., 250, 270 Reinhardt, D., 292, 295, 306 Reinhardt, G., 262, 271 Reinhold, V. N., 146, 240 Remillard, S. P., 25, 66 Renoux, J. M., 252, 253, 272 Rensing, L., 41, 55, 56, 57, 65 Reuber, T. L., 205, 223, 241 Reumkens, J., 73, 108 Rhee, J. S., 277, 304 Richards, J. B., 262, 271 Richards, J. C., 143, 147, 149, 190,232,241, 246 Rick,P. D., 138, 143,161, 162,163, 164,183, 185, 191, 206, 231, 235, 240, 241, 242 Rieth, M.,73, 74, 76, 89, 105, 109 Rietschel, E. T., 141, 236. 238 Riley, L. W., 197, 242 Rivera, M.,149, 212, 239, 242 Robbins, J . B., 140, 152, 164, 165, 170, 234, 235,245

Robbins, P. W., 154, 156, 160, 169, 198,227, 232, 242, 246 Robbins, R. W., 154, 156, 233 Roberts,I., 166, 177, 178, 182, 183, 185,200, 201, 202, 234, 238, 242 Roberts 1. S., 164, 172, 178, 199, 200, 201, 202, 208, 220, 233, 234, 242, 243, 245 Roberts, L. S., 177, 178, 182, 199, 201, 178, 182, 199, 201, 232 Roberts. M., 182, 201, 242 Rodgriguez-Tebar, A., 187, 231 Rodriguez, M.-L., 153, 242

326

AUTHOR INDEX

Rodriguez-Aparicio, L. B., 165, 242 Rodwell, V. W . , 268, 270 Rogers, P. L., 255, 272 Rohmer, M.,250, 251, 252, 253, 254, 255, 257.258,262.264,266,268,269,270,271, 272

Rohr, T. E., 162, 164, 242, 244 Rolfe, B. G., 146, 229, 231. 235 Romana, L., 173,176,193,195,196,21I , 245 Romana, 172, 173, 176, 192, 193, 195, 196, l97,203,206,208,210,211,213,231,232, 237, 238, 239, 244 Romanowska, E.. 141,145,197.235.239.243 Roncero, C., 191, 242 Ronson, C. W., 218, 242 Roobol, A,. 13, 14, 19, 20, 22, 31, 39, 43, 67, 68 Roseman, S., 164, 231, 238 Rosen, I. G., 205, 235 Rosen, S. M.,156. 246 Rosenbaum, J. L., 25, 62, 66 Ross, 1. K., 28, 68 Ross, J., 43, 63 Ross, P., 167, 228. 242. 244 ROSS, P. Y.,167,228, 242,244 Rossen, L., 229, 232 Rossi, T., 262, 270 Rotering, H.. 186, 236 Roth, L. E., 196, 243 Rothfield, J., 154, 156, 169, 245 Rothfield, L. I., 187,212,213, 232,233,234, 237

Rothmel, R., 206, 243 Rothme1.R. K., 146,166, 167,184, 198,199, 221, 240 Rothnie, H. M.,7, 66,68 Roverie, A,, 73, 108 Rowley, D., 192, 239 Roxlau, A,, 205, 231 Roychoudhary, S., 146, 166, 167, 184, 198, 199, 221, 222, 240, 242 Royo. C., 223, 231 Rozhinova, S. S., 154, 156, 160, 169, 243 Rubin. F. A , , 220, 227, 241 Rubin, L. G., 208. 246 Rubin, R. A , , 146, 198, 220, 234,243 Rudner, D. Z., 119. 133 Rudolph, K.. 168, 235 Rugman, P. A , , 182, 232 Ruhnau, B., 295. 303 Ruoff, B. M.,9, 67 Rusch, H. P., 48, 54. 63, 67, 68 Russell, S. A,. 72, 73, 87. 105 Ruth, D. C., 58, 66 Ryan-Graniero, J.. 171, 205, 245 Ryser, U., 52, 69

Sa’Correia, I., 167, 245 Sachsenmaier, W., 47, 54. 55, 57, 66, 68 Sack, R. B., 209, 242 Saedler. H., 295, 305 Saengchjan, S., 261. 262, 272 Safran, M.,96, 108 Sagers, R. D., 77, 107 Saghari, H., 209. 235 Sahasrabuddhe, C. G., 56, 63 Sahm, H., 253, 254, 255. 256, 257. 259. 264, 268, 270, 271, 272 Saier, M. H., Jr, 181, 183, 242 Saiki, T., 77, 80, 109 Saito, H., 113, 122, 123, 124, 126, 132, 133 Sakai, F., 277, 305 Sakai, K., 221, 222, 242 Sakamura, S.. 287. 304 Sakasaki, R.,209, 242 Sakurai, H.,98, 107 Salles, I., 14, 23, 31, 69 Sallcs-Passador, I . , 14, 20, 41, 43, 68 Salmond, G. P. C., 178. 236 Saltmarsh-Andrew, M., 154, 156, 169, 245, 246 Samson, S. M.,295, 2Y9, 305 Samuelson, K., 150, 151, 236 Sanchez, F., 2 0 , 305 Sanderson, K. E., 175, 191, 208, 212, 239, 242,243

Sansonetti, P. J., 197, 221, 240. 242, 245 Santiago, F., 172, 173, 176, 192, 193, I%, 206, 208, 21 I, 237 Santini, D.-L., 80, I 0 8 Sarathchandra, S. U., 102, 108 Sasaki, T., 169, 209, 242 Sasse, R., 18, 29, 68 Sato, M.,277, 305 Satola, S., 130, 133 Satola, S. W., 130, 133 Sauer, H . W . , 6 , 7,49,50, 51, 56.62.63,66, 67, 68 Sauer, R. T., 183, 238 Saunders, J. R., 220, 221, 231 Saunders, K., 299, 305 Sauter, M.,78, 105 Savard. L.. 38, 68 Savelli, B . , 116, 123, 132 Sawa. T.. 262, 271 Sawada, S., 144, 243, 246 Sawers, G., 71,72,76,78,80, 89, YO, 92,94, 9.5, 96, 97, 99, 102, 105, 106, 107, 108 Schade. W..262, 271 Schaechter, M.,181, 234 Schatz, P. J., 183, 243 Schauer, N. L., 76, 77, 78, 108 Schedl, T.. 1.5, 18, 19, 37, 39, 42, 43, 64, 68

AUTHOR INDEX

Scheetz, M. E., 11 295, 299, 303 Schendel, P. F.,78, 108 Scherer, P. A . , 77, 81, 108 Schiller, N . H., 167, 243 Schlichtman, D.. 221, 242 Schlindwein, C., 80. 108 Schmidt, G . . 161, 191. 203, 243 Schmidt, M. A , , 140, 152, 243, 245 Schmoll. T..212, 231 Schnable, P. S.,295, 305 Schnaitrnan. C. A , , 175, 176, 191, 208. 213, 216, 219. 238. 241, 243 Schneerson, R.,140, 152. 153, 199,234,243, 245 Schneider, H.,250, 271 Schnell, J . , 74, 109 Schoelz. J . E., 213, 239 Schoenhals, G . , 142, 143, 203, 207. 239,246 Scholz. T.D., 92, 95, 96, 108 Schon, A . , 88, 91, 93, 108 Schoolnik, G . K . , 197, 242 Schrader, H., 74, 75, 76. 108 Schrager, H., 76, 77, 80, 107 Schrcider, I . , 78, 80, 107 Schroeder, M. M..15, 19, 65 Schuch, W., 295, 301, 304 Schuckelt, R . , 73, 108 Schulenberg-Schnell, H . , 253, 254, 255, 270 Schwartz. K., 72, 108 Schwartz, K. V . , 2, 66 Schwartz, R . M., 295, 305 Schwederski. B., 83, 84, 87, 101, 107 Scott, J.. 12, 68 Scott. R. A , , 85. 106 Seckler, B., 266, 272 Segel. I . H., 99, 107 Seid, R . C., 197, 235 Sela, M.,93, 107 Selander, R. K.,146, 210, 220, 233,243 Sen, K..183, 243 Sequeira, L., 196, 205, 237 Seraili, A . , 224, 236 Seraneeprakarn, V . , 98,102, 108 Setlow, B., 131, 132 Setlow, P., 131, 132, 133 Seto, H..262, 272 Shands, J. W., Jr. 142, 243 Sharma. B. V. S . , 262, 272 Sharrna, F., 184, 241 Sharma, S., 199,241 Shashkov, A. S.. 151, 210, 239 Shaw, D. H.,144, 231 Sheehy, T. W.,209, 243 Sherman, M. M.,268, 272 Shibaev, V. N.,138, 154, 156, 160. 162, 169, 172. 233, 237, 243

327

Shiffman, D., 299, 306 Shifrine, M., 258. 270 Shigeri, Y.,264, 272 Shiloach, J . . 152, 234 Shin-ya, K., 262, 272 Shinabarger, D., 146,166, 167, 184, 198,199, 206, 221, 240,243 Shinnick, T., 8, 37, 50, 51, 67 Shinnick, T.M.,27, 28, 29, 69 Shiomi, K., 262, 272 Shipley, G. L.,27. 28, 56. 63.67,68 Shipston, N.,277, 306 Shrift, A . , 98, 99, 100, 105. 107, 108 Shuber, A . P., 78, 108 Siewart, G . , 160, 243 Signer, E. R., 196, 205, 233,234, 246 Silflow, C.D., 22, 68 Silhavy, T. J., 183, 243 Silliker, M. E.,11, 68 Silver, R. P.. 178, 182, 183, 184, 1 9 , 201. 225,226,228, 229. 231,mi, 241,243,245 Silverman, M., 227, 231 Silverstone, A., 292, 295, 301, 303 Sim, G. K.,295, 305 Simmons, D. A. R., 197, 243 Simon, H . , 266, 272 Simon, M. I., 218, 232 Simon, R., 206, 21 1, 237,238 Simonin, P.,252, 258, 271, 272 Simpson, P. A . , 14, 65 Singh, M.. 172, 241 Singh, S. K . , 221, 242 Singhofer-Wowra, M., 16, 21, 66,68 Sinskey, A . J.. 226, 234 Sipe, J. D.,258, 272 Sirevag, R.,300, 305 Sismeiro, O.,225, 233 Sister, H., 279, 303 Skatrud, P. L.,295, 299, 303,305,306 Skjak-Braek, G.,166, 167. 243 Skotnicki, M. L.,255, 272 Skriabina, S. V.,252, 271 Skumik, M., 192, 193, 212, 213, 231 Skurray, R., 213, 235 Slettengren, K . , 147. 237 Sliedregt, L. A . J. M., 228, zQ2 Sllisz, M. L.,295, 299, 305 Sliwkowski, M. X.,74. 89. 108 Smets, P., 147, 149, 238 Smit, J . , 185, 243 Smith, A. N., 178, 201, 243 Smith, A. R. W . , 144,243 Smith, D.W. E., 94, 106 Smith, H. 0..208, 243 Smith, I., 118, 119, 124, 125, 127, 128, 129, 132, 133

328

AUI'HOR INDEX

Smith, 1. C. P., 152, 232, 261. 271 Smith, I. H., 170, 243 Smith, N . H., 146, 220, 243 Smith, P. F., 250, 251, 252, 271 Snellings, N . J., 220, 243 SO, J.-S., 196, 243 Soda, K., 98, 102, 106, 108 Soderstream, T., 153, 245 Sokol, P. A , , 226. 246 Sdll, D., 91, 93, 108 Solnica-Krezel, L., 18.20, 27, 28, 29, 30, 32, 33, 34. 35, 36, 41, 63, 64, 68 Sonati, A , , 200, 235 Sonenshein, G. E.. 52. 67 Sorbo, B., 260, 271 Spanu, P., 292, 295, 306 Spence, J . , 114, 115, I32 Spences. M., 279, 306 Speth, V., 185, 240 Spiegelman, G.. 130, 133 Spiegelman, G. B.. 122, 125, 127, 133 Spizizen, J., 128, 132 Spottswood, M. R . , 10, 11, 12, 66 Sprintz. H., 200, 243 Sprinzl, M., 88, 91, 93, 106, 108 Sprott, G. D., 261, 271 Squires, R. M.. 276. 303 St Geme. J. W., 111., 226, 244 Stacey, G., 196, 243 Stader, J.. 183, 243 Stadtman, T. C., 72, 73, 74. 76, 77, 78, 81, 83, 87, 88, 89, 91. 92, 94, 95, 96, 97, 98, 102, 104, 105, 106. 107, 108, 109 Stampf, P., 252, 272 Staneloni, R. J., 158, 159, 198, 234,244 Stanfield, S., 146, 198. 234 Stanfield, S. W., 168, 180, 237, 243 Stanley, G. H. S., 253, 271 Stanzel, M., 93, I08 Stark, J. W., 162, 244 Starman, R., 161, 162, 163, 164, 206, 240 Steed, M. M., 213. 239 Steenbergen, S. M., 164, 165, 172, 173, IN, 200, 202, 208, 229, 244, 245 Steffens, G. J., 72, 89, 106 Stein, M. A , , 219, 241 Steinberg, I. Z., 93, 107 Steinberg, N. A., 100, 108 Stevens, B. J. H.,223, 241 Stevenson, G . . 203, 206, 244 Stewart, V., 78, 80, 93, 104 Stibitz, S . , 223, 231 Stirm, S., 157, 233 Stock, A. M., 113, 114, 115, 133, 218, 244 Stock, J. 8 . . 113, 114, 115, 133,218, 244 Stocker, B. A . D., 138, 149, 161, 169, 170,

175, 190, 191, 193, 197, 198, 206,212. 228, 237. 239, 240, 242, 246 Storey, D. G., 226, 246 Stout, V., IYY, 205. 216. 217, 218, 223, 235, 244 Stragier, P., 115, 116, 123, 130,131,132, 133 Strassburger, W., 73, 108 Strauch, M.,130, 133 Strauch, M.A , , 120. 125, 126, 127, 128, 129. 132, 133 Strecker, G . , 147, 149, 238 Strittmatter, W., 141, 236 Stroeher, U. W., 208, 209, 244 Strominger, J. L., 154, 160, 231, 236, 243 Stuart, K., 12,65 Sturm, S. B., 192, 193, 197, 244 Subramani, S., 61, 64 Sugiyama, T . , 204, 221, 231 Sullivan, K. F., 15. 21, 68 Sullivan, M. A , , 95, 108 Sullivan, W.,30, 68 Sun, D., 131, 132, I33 Sunde, R. A., 72, 108 Sutcliffe, J., 184, 244 Sutherland, 1. W.,138, 145, 146, 156, 160, 166,167,168, 170, 171,227,233,235,244, 245 Sutter, B., 264, 268, 272 Sutton, A,, 184, 231 Suyama, Y., 11, 68 Suzue, G., 262, 272 Suzuki, S., 196, 237 Suzuki, T., 10, 66 Svanborg, C., 192, 234 Svanborg-Eden, C., 210. 233 Svenson, S. B., 147, 237 Svensson, S., 150, 151, 236 Sweeley, C. C., 154, 236 Swings, J., 263, 272 Swissa, M.,167, 244 Symes, K. C., 170, 243 Szabo, M. J., 190, 240 Szarek, W. A., 144, 231 Szulmajster, J., 115, 132

Tagaki, T., 26, 66 Tahara, Y.,254, 258, 269, 273 Takade, A., 142, 231 Takahashi, H., 123, 124, 126, 132, 133 Takahashi, M., 279, 281, 282,283,287, 304, 305 Takaiwa, F., 10, 65 Takano, H., I I , 12,66, 68 Takeshita, M., 169. 244 Takeuchi, T.. 262, 271, 272 Takikawa, Y.,277, 281, 286, 288, 304

AUTHOR INDEX

Tal. R., 167, 168, 198, 228, 229, 246 Tamamura, T . , 262, 271 Tamura, K.,209, 242 Tanaka, H., 74,76,98,102,107,108,279,306 Tanaka, K.,14, 39, 68 Tanaka, S.. 262, 272 Tanaka, T . , 259, 270 Tanase, S., 277. 281, 288, 293, 304,305 Tang, H.-C., 5, 64 Tang, J.-L., 223, 244 Tansey, L., 205, 244 Tappe, C. H., 266, 269, 271, 272 Tardif, M. C., 59, 64 Tarelli, E., 140, 233 Tatti, K. M., 119, 133 Tattrie, N. H., 250, 271 Tauber, A,, 283, 306 Tavernier, J. E., 102, 105 Taylor, R. F., 262, 273 Tazaki, M.. 277,279,281,282,283,284,286, 287, 288, 293, 304,305 Tecklenburg, M.,171, 205, 245 Teixeira, M., 84, 85, 86, %, 106, 108 Terakado, N., 196, 237 Tesche, N., 159, 164, 165, 244 Tessier, A., 8, 37, 38, 50, 51, 66, 67 Thauer, R. K.,77, 81, 108 Theines, C., 180, 233 Thomas, C. J . , 191, 192, 236 Thomas, K. C., 279, 306 Thompson, G. A., 258, 273 Thompson, R. H . , 261, 270 Thorne, K. J. I., 262. 273 Thorne, L., 205, 244 Thurow, H., 185, 232 Tiller, P. R., 147, 237 Tillmann, A. M., 262, 270 Timmis, K.,182, 200, 201, 242 Timmis, K. N., 177, 182, 192, 193, 197, 199, 201, 232, 244 Tippett, J.. 226, 234 Tobin, M. B., 295, 299, 304 Tokusbige, M., 264, 272 Tollon, Y.,47, 56, 57, 64,69 Tolmasky, M. E., 158, 159, 244 Tomibe, K.,144,243 Torgov, V. I . , 162, 237 Tonnay, P., 92, 95, 96, 105 Torres-Cabassa, A., 216,217, 218, 219,220, 235, 244 Toublan, B., 45, 49, 50, 65, 67, 69 Tovianen, P., 192, 193, 212, 213, 231 Trach, K.,113, 114, 115, 116, 117, 120, 122, 123, 125, 126, 130, 132, 133 Trach, K. A,, 113, 116, 117, 118, 132 Tribe, D. E., 255, 272

329

Trisler, P., 203, 206, 217, 218, 219, 235, 244 Troy, F. A., 138, 139, 156, 159, 160, 164, 165, 169,173, 176, 180, 181, 184, 185, 187, 227, 239, 242, 244, 245, 246 Truitt, C. L., 9, 65 Tsai, L., 88,92, 95,96, 105, 107, 108 Tsiolis, G. C., 192, 240 Tsuchiya, R., 89, 109 Tsui, F.-P., 152, 153, 244, 245 Tulley, R. E., 168, 232 Tullius, T . D., 97, 106 Turner, D. C., 74, 108 Turner, E. M.,279, 306 Turner, M. K.,299, 305 Turnock, G., 6, 37, 38, 68 Turowski, D. A., 196. 233 Tuve, T., 97, 108 Tyson, J. J., 47, 53, 68 Tyson, J. T . , 41, 42, 65, 66 Uchida, T., 169. 198, 209, 242 Ugalde, R., 168, 246 Uglade, R. A., 158, 159, 244 Umezawa, H., 262, 271, 272 Ursini, F., 73, 108 Ursinus-Wossner, A , , 185, 245 Urushizaki, S., 277, 305 Utamara, T., 146, 168, 238 Uyeda, T. Q. P., 14,23,24,25,26,38,68,69 Vaara, M., 137, 240 Vaarki, A., 170, 236 Vaidya, A. B., 119, 132 Vaillasante, A., 21, 69 Valla, S., 167, 244 Valvano, M.A., 191,192, 193,196,197,208, 210, 211, 240, 245, 246 Van Alphen, L., 137, 239 Van Alphen, L. F., 184, 245 van Boom, J. H., 228, 242 Van der Drift, C., 80,86, 108 Van der Marel, G. A., 228, 242 van Duin, J., 186,236, 245 Van Frank, R. M.,295, 299,305 Van Hoy, B., 122, 133 Van Kampen-de Troye, F., 184, 245 Van Veldhuizen, A., 146, 168, 246 Vanderslice, R. W., 171, 205, 245 Vandewel, D., 212, 237 Vann, W. F., 152, 153, 164, 165, 170, 182, 183, 199, 201, 234,243, 245, 246 Vasselon, T., 221, 240, 245 Vaughn, J. P., 51, 52, 65 VAzquez, D., 187,231 Veal, L. E., 295, 299, 305 Venis, M. A., 287, 306

330

AlJTlIOR INDEX

Veprek, B., 72, 89,92, 94, 95. 96, 07, 102. 105, 106, 107 Veres, Z., 88, 92. 95, 96, 107, 108 Verma, D. J . , 198, 245 Verma, N., 193, 195, 108, 209. 240, 245 Verma, N. K . , 208, 211, 239, 245 Ververidis, P . , 287, 306 Vijay, I. K.. 159, 164, 165, 187, 244. 245 Villiger, W., 136, 182, 235, 236 Vimr, E. R., 164, 165, 172, 173, 184, 185, 199,200,202.207.208,227,228,229,243, 244,245,246 Vining, L. C., 299, 304 Vogel, H. J., 226, 246 Vogels, G. D., 80, 86, I08 Vogt, V. M.,8, 9, 52, 53, 64. 65. 67, 69 von Dohren, H., 299, 306 Voordouw, G . , 8s. 108 Wacharotayankun, R.. 204, 221, 231, 245 Wada, K., 89, 109 Wagner, R.,76, 80, 81, 87, 109 Wakil, S . J . , 262, 273 Walden, P. D., 17. 18, 19, 69 Walderich, B., 185, 186, 236, 245 Walker, G. C., 205, 223, 224, 229. 234,235, 239, 241 Walsh, C. T., 287, 305 Walter, A , , 184, 244 Wang, D., 21, 69 Wang. G., 206, 21 1 , 238 Wang, H.. 295, 306 Wang, L., 173, 176, 193, 195, 196, 211, 245 Wang, S.-K., 167, 245 Wang, T. T., 287, 305 Ward, J. F., 223. 239 Ward, M. W. T., 279, 306 Ward, S . , 138, 162, 191, 242 Warn, R. M.,21, 67 Washington, O., 197, 238 Watanabe, H., 192, 197, 246 Watanabe, T., 279, 280, 306 Waterborg, J . H., 46,69 Watkinson, J. H., 102, 108 Watts, D. I., 6, 37, 68 Waxin, H., 220, 238 Weatherbee, J. A , , 21, 69 Webb, F. H., 95, 108 Webb, V., 125, 133 Webster, R. E., 186. 187, 188, 239. 245 Wei, Y.,46, 65. 67. 69 Weickert, M. J., 126, 133 Weigel, B. J . , 295, 298, 304,306 Wcinberger-Ohana, P., 228, 242 Weiner, I. M.,154, 156, 160. 169, 241, 245

Weiner, J. H., 100, 104 Weinhouse, D., 228, 242 Weinhouse, H., 167, 24l Weintaub, A., 195, 240 Weir, J., 118, 119, 127, 132, 133 Weisgerber, C., 162. 163, 164, 165, 172, 173. 178, 180, 184, 199,202,208,229,235,237, 245 Weiss, J., 245 Weiss, K. F., 97. Y9, 109 Weissborn, A . C., 146, 240 Welker, D. L. 8, 69 Welsh, J., 102, 193, 196, 240 Wendel, A , , 72, 89, 106 Weng, W. M.,205, 231 Werenskiold, A . K.. 16, 69 Werner, K. E., 140, 234 Werner, P. K., 181, 183, 242 Westhof, E., 94, 104 Westlake, D. W.. 298, 304 Wheals, A . E., 33, 34, 37, 65, 69 Whelan, J. K., 276. 304 White, D.. 143, 235 White, R. H., 261, 262, 264, 273 Whiteley, H. R.. 102, 109 Whitfield, C., 142, 143. 147, 149, 161, 171, 181, 184, 185,187, 190, 192,203,204,207, 210,219,220,221,233,237,238,239,245, 246 Wick, R. 6, 49, 66 Wickner, W., 183, 233 Widdel, F., 74, 87, 88, 109 Widmalm, G., 147, 237 Wieruszeski, J . M., 147, 149. 238 Wigender, J., 184, 235 Wijffelman, C. A , , 205. 233 Wijfyes, A. H. M.. 205, 233 Wikstrom, P. M.,95, 105 Wilcox, M.,14, 68 Wilhelm, F. X.,45, 49, 65, 69 Wilhelm, M. L., 45, 49, 65. 69 Wilkinson, R. G . , 161, 193, 212, 240, 246 Wilkinson, S. G., 147, 241 Willetts, N., 212, 213, 232, 235 Williams, A. E.. lY0, 240 Williams, D. S., 78, 104 Williams, H. H., 97, 108 Williams, J. M., 152, 231, 239 Williams, K. L., 8, 69 Williams, M. N. V.. 196, 246 Williams, P. H., 182, 201, 242 Wilmot, S. J., 143, 203, 204, 207, 237, 238 Wilson, D. B., 156, 160, 237 Wilson, L. G., 99,109 Winans, S. C., 223, 239 Wingfield, M. E,, 197, 235

AUTHOR INDEX

Winkler, E., 76, 77, 80, 107 Winkler, U. K., 184, 235 Winston, J. R., 277, 279, 305 Witholt, B., 185, 233 Wittwer, A,, 88, 105 Wittwer, A . J., 88, 89, 95, 96, 109 Wohlfarth-Botterman, K.-E., 13, 64 Wolf, B., 73, I08 Wolf-Ullisch, C., 162, 163, 237 Wolfe, R. S., 78, 105 Wolfe, S., 298, 304 Wolff, G., 259, 270 Wolski, S., 138, 162, 191, 242 Wong, H. C . . 167, 168, 198, 228, 229, 246 Wood, K. V., 61, 64 Woods, D. E., 226, 246 Woodson, W. R., 295, 306 Woolfolk, C. A., 102, 109 Wooton, J. C., 223, 244 Worland, P. J . , 73, 94, 106, 107 Worthington, R. D., 261, 270 Wozniak, D. J . , 222, 224, 246 Wrangsell, G., 147, 237 Wren, M.,37, 65 Wright, A , , 154, 156, 160, 169. 170, 172, 176, 198, 212, 216, 233, 237, 239, 242, 246 Wright, D. A,, 56, 63 Wright, L. F., 178, 201, 229, 241, 243 Wright, M., 13, 1 4 , 2 0 , 2 2 , 2 3 , 2 6 , 3 1 , 3 3 , 3 7 , 41, 42, 43, 47, 56, 57, 63, 64, 67, 68, 69 Wright, M. E., 227, 231 Wrona, T. J., 164, 165, 172, 173, 199, 202, 208, 229, 244 WU, J.-J., 119, 126, 130, 133 Wunder, D. E., 243 Wyk, P.,196, 209, 235, 246 Xavier, A. V., 84, 85, 108 Yamada, K., 94, 107 Yamada, Y., 254,258,269.273,295,299,304 Yamamoto, H., 220, 236 Yamamoto, I., 77, 80, 109 Yamamoto, T., 123, 133 Yamashita, K., 286, 304 Yamashita, S., 123, 124, 126, 133 Yamashita, T., 254, 269, 273 Yamazaki, S., 84, 109 Yang, S. F., 276,281,284,287,288,202,295, 301,303, 304,305, 306 Yang, W., 97, 109 Yang. Y. C., 10-11, 66 Yano, T., 10, 65 Yanofsky, M., 146, 198, 234 Yanofsky, M. F., 223, 239

33 1

Yao, Z., 192, 197, 246 Yasuda, H., 44, 46,67, 69 Yeadon, J.. 192, 239 Yeh, W. K., 295, 298, 305 Yen, T. J., 43, 69 Yokota, S., 144, 246 York, K., 130, 133 Yoshida, Y., 192, 197, 246 Yoshikawa, H., 119, 123, 124, 126. 132. 133 You, K. H., 73, 107 Young, P. A , , 97, 109 Youngman, P.,130, 133 Youngman, P. J., 5, 27, 28, 29, 69 Yu, H., 295, 305 Yu, S. H., 147, 246 Yuan Tze-Yuen, R., 156, 241 Yuasa, R., 169, 246 Yuhara, H., 258, 273 Yumoto, N.,264, 272 Yurewicz, E. C., 157. 246 Zahringer, U., 141, 191, 235, 236 Zakian, V. A , , 8, 69 Zalisz, R., 147, 149, 238 Zamojski, A . , 141, 236 Zamze, S., 202, 238 Zamze, S. E., 144, 243 Zanen, H. C., 184, 245 Zapata, G., 199, 246 Zehelein, E., 71, 78, 80, 90, 91, 92, 95, 102, 105, 107. 108

Zehr, J. A., 100, 109 Zehr, J. P., 102, 107 Zeleznick, L. D., 156, 246 Zellweger, A., 52, 69 Zevenhuizen, L. P.T. M., 146, 168, 246 Zhan, H., 149, 204, 205, 223, 224, 229, 239, 246 Zhao, J. Y., 56, 63 Zheng, Y., 33, 69 Zhou, D., 261, 262, 264, 273 Ziegler, S. F., 223, 239 Zielinski. N.A., 146,166, 167, 184,198,199, 221, 240 Zindel, U., 74, 109 Zinoni, F., 72, 78, 89, 90,92, 93, 94, 95.96, 97, 98, 102, 105. 107, 108, 109 Zlotkin, E., 39, 41, 65 Zoller, W. H., 102, 108 Zon, G., 140, 234 Zorreguieta, A., 168, 246 Zrike, J., 187, 237 Zuber, P.,127, 133 Zundel, M., 250, 252, 270, 273 Zwahlen, A., 208, 246

This Page Intentionally Left Blank

Subject Index

Numbers in bold refer to pages on which references are listed at the end of each chapter A

A1 mutants in Physarum polycephalum, 34-35 Abietinella abietina, 255 AbrB protein and sporulation in Bucillus subtilis, 127-128 ACC, see Aminocyclopropane Acetoacetate pathway and hopanoids, 260-262, 264 Acetohacter. 252, 258, 259 A . aceti, 251 A . pasteurianus, 251 A . xylinum. 167, 198, 214-215, 227, 250 Acetylation in histone modification, 45, 46 0-acetylhomoserine sulphydrylase and selenium metabolism, 98 Acinetobacter calcouceticus, 278, 282 Acremonum chrysogenum, 296-299 Actin and Physarum polycephalum. 13, 37-38.40-41 Adenosine phosphoselenate and selenium metabolism, 99 Adenosylhopanes and hopanoids, 249, 254 S-adenosylmethione and ethylene production, 281-282, 287 adhesion zones and cell-surface polysaccharide biosynthesis transport, 185-188 Aerobacter aernones and cell-surface polysactharidc biosynthesis, 139, 155, 1.56. 157. 160, 169. 177 Aeromonas hydrophila, 278, 282

Agarius bisporus, 278 Agrobacterium, 145-146, 168 A . radiobacter, 205 A . tumefaciens, 146, 179-180. 198, 223, 228, 262 Alcaligenes faecalis , 145-146 var. myxogenes. 159 alginate synthesis and cell-surface polysaccharide biosynthesis, 221-222, 224 Alicyclobacillus, see Bacillus allosteric activation of cellulose synthetase, 228 Alternaria solani, 278 I-aminocyclopropane-lsarboxylicacid pathway in ethylene production by higher plants, 281, 287-288, 302 amoeba amoebal mitosis, 29-31 amoebal phase, 3, 4 -flagellate transition in cytoskeleton development, 23-26 -plasmodium transition in cytoskeleton development, 26-33, 38 see also Physarum polycephalum apogamic strains in Physarum polycephalum, 3, 6, 26, 28-29, 30, 33-34 Arabidopsis thaliana, 16 Arthrobacter, 278 Ascochyta imperfecti, 278 Aspergillus A . candidus, A . clavatus, A . flavus, A . ustus and A . variecolor, 278

334

SIJI3Jb( 'I' INDEX

Aspergillus (cont.) A . nidulans, 17, 21, 58 azasqualene and hopanoids, 257-258, 266 Azotobacter A . chroococcum, 255 A . vinelandri and cell-surface polysaccharide hiosynthesis. 146, I 6 6 and hopanoids. 25 I , 255, 261, 262

B Bacillus R. acidocaldarius. 251, 255-256, 259, 266, 269

B. acidoterrestris, 25 I R. megaterium. 102 B. mycoides, 278, 282 B. suhtilis, 278 see also sporulation hacteriohopanepentol and hopanoids, 249, 254

bacteriohopanepolyols and hopanoids, 25S, 267

bacteriohopanetetrol and hopanoids, 248, 254. 256. 259

ether, 249, 254, 259 glycoside, 248, 254, 259 Beijerincka, 255 B. indica and R. mohilis. 251 biosynthesis and ethylene production, 281-288 and genetics o f hopanoids, 25Y-270 see also isopentcnyl pyrophosphate and selenium metabolism, 89-96 seleno-t RNAs, 95-96 see also selcnoproteins see also cell-surface polysaccharides Blastomyces dermatitidb, 278-279 Bordetella pertussis, 223 Botrytis spectabilis. 278 Rradyrhizobium japonicicm. 73. 87. 196 Brassica jucea, 294-295 Rrevihacterium linens, 278 C CAD-like proteinsand cxtracellular regulation of cell-surface polysaccharides, 224-225 Caldariella acidophila, 26 I capsular polysaccharides (CPS) and cellsurface biosynthesis, 137, 169-170 export, 172-173, 177-184 genetics, 190, 199-204, 207-208 process, 156, 159-161, 164-166 regulation, 221 structure and attachment, 139-144, 146, 148-149

carhon-monoxide dehydrogenases and selenium metabolism, 73 carotenoids and hopanoids, 258 catalysis in selenium metabolism versus sulphur, 96-97 cell membrane stability see hopanoids cell-surface polysaccharides in gram-negative bacteria. hiosynthesis and expression of, 135-246 structure and attachment, 13X-I53 multiple, expression of, 149 repeating unit structure, 144-148, 150-1 53 surface association. 13X-I44 seealso Export; Genetics o f polysaccharide hiosynthesis; Proccsses; Regulation o f cell-surface plysaccharides cellulose synthetase, allosteric activation o f , 228

Cephalosporium C. acremonium , 294298 C . gramineum, 12, 17 Chaetomium C . chlamaloides. 278 C. globosum, 284 chimeric mitosis in Physarum polycephalum, 29-30

Chinia ricbra, 262 Chlamydia trachomatis, I63 Chlamydomonas reinhardtii. 1 I , 16, 17, 62 chromatin and Physarum polycephalum, 44, 45, 46-47

Chromohacteriicm violaceum, 278 chromosomes chromosomal genes for 0-polysaccharides, 19&196

replication in Physarum polycephalum, 48-52

origins, 51-52 timing in individual genes. 49-51 see also genetics Citrohacter. 145-146, 278 C. freundii, 98. 214, 220, 227 Clostridium 76 C. acidiurici, 73, 77, X I , 86 C. barkeri, 87 C. cylindrosporum, 73, X I , 86 C . formicaceticum 80 C. histolyticum. 73 C. litorale, 73 C. pasteurianum. 17, 81, 99 C. purinolyticum, 73. 95 C. sporogenes, 73 C. sticklandii, 73, 14. 88, 98 C. thermacetrcum , 71, 80 Codium latum. 279, 280

.

~

335

SUBJECT INDEX

colanic acid, 145-146 competition during incorporation in selenium metabolism, 97 Coriolus hirsutus and C. versicolor. 278 Corynebucterium, 282 C. aquaticum, 278 CPS see capsular polysaccharides Cryptocvccus C . ulbidus. 278-279,281,282-284,289, 302 C. luurentii, 278-279 Cucumis melo, 289 cyclin and mitotic regulation in Physarum polycephahrm, 51-58 cycloartcnol and hopanoids, 258, 267 cytoplasmic membrane and cell-surface polysaccharide biosynthesis, 136-137 see ulso Export of polysaccharidcs; Transport of polysaccharides cytoskcletal organization of Physarum polycephalum, 13-39 in development see Development inferences, 38-39 microtuhule organization, 13-14 microtubule-associated proteins, 22-23 other genes differentially expressed, 37-38 see also Tubulins

D Daeduleu dickinsii, 278 dehydrogenases and selenium metabolism, 73, 8 ~ 7 see ulso FDH Dematium pulluluns, 278 Desulfohucterium niacini, 87 Desulfococcus multivoruns, 88 DesulfomicrobiumlDesulfovibrio D.hucirlutum. 84-85. 87 D . gigus, 85-86 D . sulexigens, 84 D. vulgaris, 84, 85-86 development of Physurum polycephalum cyctoskeleton, 23-33 amoeba-flagellate transition, 23-26 amoeha-plasmodium transition, 2 6 3 3 mutants, 3-3-37 Dianthus caryophyllus. 294-295 Dictyostelium discoideum, 8 Didymium iridis, 4, 8, 11 diffusion uptake of introduced molecules and Physurum polycephalum. 58 dimethylallyl pyrophosphate and hopanoids, 265 diploptene and hopanoids, 248,25&25 1,253, 266267 diplopterol and hopanoids, 248, 25(b-251, 253, 258, 259. 2 6 2 6 7

D N A , 103 rearrangements and IS elements in instability of polysaccharides, 225-227 transformation and Physarum polycephalum, 7, 47-52, 54-62 gene targeting, 6 1-62 mitochondrial, 10-13 nucleolar genome, 8-10 replication, 52-53 stable expression, 61 transient expression, 5 5 M 1 see also genetics Drosophilu melanoguster, 6-7, 16, 17. 20, 21, 41

E ECA see enterobacterial common antigen Emericellu niduluns. 296-298 Enterobucter uerogenes and E. clouceue, 278 cnterobacterial common antigen (ECA) and cell-surface polysaccharide hiosynthesis, 175 process, 161, 162-163 structure and attachment, 139, 142, 144-145, 148-149 enzymes activity in precursor formation and ccllsurface polysaccharide biosynthesis, 227-228 selenium-containing, 7 2 4 8 not containing selenocysteine residues,

86-88 see also prokaryotic selenoproteins EPS see extracellular polysaccharides

Erwiniu, 202 E. amylovoru, 168, 204, 214, 220 E . chrysanthemi, 144 E. herbicolu, 278 E. rhapontici, 278, 286 E . slewartii, 190, 199, 204, 214, 220 Escherichiu coli and ccll-surface polysaccharide biosynthesis export, 172, 174-175, 177-185, 187-188 genetics, 190-193, 194-197, 195)-204, 2 6 2 0 8 , 210-211 process, 157, 15%165 regulation, 212-221, 224-225 structure and attachment, 139-144, 146-148, 151-153 and ethylene production, 277-278, 281-282, 292-293, 302-302 and hopanoids, 261-262, 264, 268-269 and selenium metabolism, 71 enzymes, 77-79, 83 geochemistry. 100

336

SUHJECI' INDEX

Escherichia coli (cont.) and sulphur, 96, 97 transport and metabolism, 98-99 tRNAs and selenoprotein biosynthesis, 89.92, 93-95 ethanol concentration and hopanoids, 257-258 ether, bacteriohopanetetrol, 249, 254, 259 ethylene production by microorganisms, 275-306 biosynthetic pathways, 281-288 see also aminocyclopropane; methylthiobutyric acid; oxyglutarate reports of, 277-281 see also P. syringae under pseudomonas Eubacterium acidaminophilum, 73. 74-76 eukaryotes selenoproteins from, 73, 89 sterols in, 250, 258, 266-267 export of polysaccharides and cell-surface assembly, I7 1-188 biosynthetic complexes located at cytoplasmic membrane, 171-174 translocation from cytoplasmic membrane to surface, 182-188 see also transport of polysaccharides extracellular polysaccharides (EPS) hiosynthesis, 137, 212 export, 168, 170-171, 184 genetics, 190, 196, 198-206 process. 154, 156, 158 structure and attachment, 138439, 142, 144-146

extracellular regulation of cell-surface polysaccharidcs, 216-229 allosteric activation of cellulose synthetase, 228 CAD-like proteins, 224-225 enzyme activity in precursor formation, 227-228 histone-like proteins in synthesis of alginate in Pseudomonas aeruginosa, 224 IS elements and DNA rearrangements in instability of, 225-227 protein-protein interactions and translational regulation, 228-229 see also two-component systems F FI mutant in Physarum polycephalum. 35 farnesyl-pyrophosphate synthase and hopanoids, 244-265, 268-269 FDH (formate dehydrogenase H) and selcnium metabolism, 72-73, 76-84, 89-90. 93, 96-97 Fistulina hepatica, 278

flagellates and Physarum polycephalum. 4, 14, 34-35 transition to, 23-26 Flammulina veltipes, 278 Fomitopsis pinicola, 278 formate dehydrogenase H see FDH Frankia, 255, 256 Fusarium oxysporum , 284 f s p . tulipae. 287 G GI, G2 and G3 mutants in Physarum polycephalurn. 35 Gallus domesticus, 17 genetics gene expression and molecular cloning, 292-293 gene for UGA-decoding tRNA, 90-92 of hopanoids see biosynthesis and genetics of hopanoids of Physarum polycephalum gene targeting and DNA transformation, 61-62 see also genome organization of plysaccharide biosynthesis, 188-21 1 chromosomal genes for, 190-196 extracellular, 198-206 housekeeping, 188-190 molecular basis for antigenic variation, 207-2 1 1 multiple clusters, relationships between, 26207 of, 190-198

plasrnid-encoded genes for, 196-197 rfe-independent, transport of, 175-177 structure modification genes, 197-198 see also chromosomes; DNA; genome; mitosis; mutants; RNA genome organization of Physarum polycephalum, 6-13 mitochondrial, 10-13 nuclear chromosomal, 6-8 nucleolar DNA, 8-10 geochemistry of selenium, 100, 102-103 geohopanoids, 250 geranyl pyrophosphate and hopanoids, 265 Gloephyllum saepiarium and G . trabeum, 278 glucose metabolism and hopanoids. 262-264 L-glutamic acid and ethylene production, 281 gluthathione peroxidase isoenzymes and selenium metabolism, 73, 89 glycine reductase and selenium metabolism, 72, 73-76, 89 glycoside, bacteriohopanetetrol, 248, 254, 259 gram-negative bacteria see cell-surface polysaccharides

SUBJECT INDEX

group-Hike polysaccharides, 157, 216-221 gene clusters, 202-204 group-11 and group-TI-like polysaccharides, 152-153 gene clusters, 199-201 polymerized in E. coli, 164-166 transport of, 177-182

H Haemophilus influenzae and cell-surface polysaccharide biosynthesis export, 178-180 genetics, 190, 202, 207-208 regulation, 214-215, 225, 229 structure and attachment, 139-140, 148, I53 Halobacterium H. cutirubrum, 261, 262, 264 H . halobium, 261 Hansenula subpelliculosa, 278 heterophasic and heteroploidic fusions in mitotic regulation of Physarum polycephalum. 53-54 heterothallic strains in Physarum polycephalum. 3, 5-6, 28-29, 34 Hirshioporus abietinus, 278 histones histone-like proteins in synthesis of alginate in Pseudomonas aeruginosa, 224 and Physarum polycephalum, 4041, 58 periodic variation, 44-47 hopanoids in bacteria, biochemistry and physiology of, 247-273 detection and analysis, 253-254 distribution and physiological role, 254-259 structural diversity, 248-249, 250-252 see aho under biosynthesis Hordeum vulgare, 294-295 hpr gene and transition-state regulators and sporulation in Bacillus subtilk, 128 hydrogenases and prokaryotic selenoproteins, 84-86 see also dehydrogenases hydroxymethylglutaryl CoA and hopanoids, 260-261 Hymenochaete tabacina, 278 Hyoscyanus niger, 294-295 I lepex lacteus, 278 incorporation in selenium metabolism, competition during, 97 inheritance of Physarum polycephalum, 3,

-5-6

337

initiation of sporulation in Bacillus subtilis, 130-131 introduced molecules and Physarum polycephalum, 58-62 diffusion uptake, 58 DNA transformation, 59-62 macroinjection, 58-59 IPP see isopentenyl pyrophosphate IS elements in instability of polysaccharides, 225-227 isopentenyl pyrophosphate hopane pentacyclic skeleton from, 264synthesised, 259-264

K Kl mutant in Physarum polycephalum. 36 KDO see octulosonic acid kinases initiating sporulation in Bacillus suhrri 113-115 control of phosphate flow, 120-123 see also phosphorelay isolation of genes for, 1 16-1 18 kinase activity in Physarum poll 4041, 4547, 55-58 Kitasatosporia, 262 Klebsiella, 157, 190 K . aerogenes, 156, 160, 170, 215 K . pneumoniae, 278 and cell-surface polysaccharide biosynthesis genetics, 192-193, 202, 204, 210-211 process, 157, 161 regulation, 214, 220-221 structure and attachment, 146-148, 149 KMBA see methylthiobutyric acid L Ll mutant in Physarumpolycephalum, 35-36 Lactobacillus plantarum, 262, 264 Laetiporus sulphureus, 278 lanosterol and hopanoids, 258, 267 Lentinus lepideus, 278 Lenzites betulina, 278 life cycle of Physarum polycephalum, 3, 4-6 amoeba1 phase, 3, 4 plasmodia1 phase, 3, 4-5 sexual cycle and inheritance, 3, 5-6 lipids see hopanoids liposaccharides (LPS) and cell-surface polysaccharide biosynthesis, 137 export, 172, 174, 176, 181, 183, 185, 188 Lycopersicon esculentum, 294-295, 301

M M I mutant in Physarum polycephalum, 35

macroinjection and Physaruni pol.ycephalum, Sn-59 Malus sylvestris. 294-205, 301 MAP (microtubule-associated proteins) and Physarum polycephalum. 22-23 matA mutations linked to, 33-34 mutations unlinked to, 34 maturation promoting factor see MPF membrane see cell membrane; cytoplasmic membrane Merhanohacteriurnformicicum, 77. 83 Methanococcus M . formicicum. 78. 96 M . vannielii, 76. 78, 84 M . voltae, 84, 95

methione and ethylene production. 281-282, 2x7 L-methionine catabolism in ethylene production. 277, 281, 287 methyldiplopterol and hopanoids, 248, 259 Methylohacterium, 255 M . fujisawaense, 264 M. organophilum, 251, 262. 264. 268 Methylococcus M . capsulatuc. 251 M . luteus. 251 Methylocystis parvus, 25 1 Methylonionas merhanica, 251 Methylosinus trichosporiurn, 25 1

methylsclenides/methylselenoxidesand selenium metabolism, 102-103 2-keto- 1 -nieth ylthiobutyric acid pathway in ethylene production, 277, 279, 281, 282-284, 302 mevalonate and hopanoids, 20-262. 264 Micrococcus M . luteus, 262, 278 M . lysodeikticus. 162 Microcystis aeruginosa, 25 1 microtubule and Physarum polycephalum

-associated proteins, 22-23 organization. 13-14 see also MTOCs mitochondria1 genome of Physarum polycephalum, 10-13 mitotic cycle, 39-58 chromosome replication, 48-52 periodic variations, 42-48 plasmodial, 3 W 2 ribosomal DNA replication, 52-53 see also Regulation, mitotic molecular basis for antigenic variation in genetics of polysaccharidc biosynthesis, 207-21 1 cloning and expression of gene, 292-293

Moraxella nonliquefaciens. 152

MPF (maturation promoting factor) and mitotic regulation in Physarum polycephalum. 55-58 MTOCs (microtubule organizing centres), 13-14, 18, 20, 29-33, 38 Mucor hiemalis, 278 multiple cell-surface polysaccharides, expression of, 149 multiple clusters, relationships between in genetics o f polysaccharide biosyn thesis, 206-207 multiple tubulins and Physarum polycephalum. 20-22 M u musculus. 16, 17 mutants developmental in Physarum polycephalum. 33-37 and sporulation in Bacillus subtiIisseespo0 genes see also genetics Mycobacterium, 280 Mycoplasma mycoides, 258 myosin and Physarum polycephalum, 13, 38 Myrothecium roridum, 278 Myxococcus fulvus. 262, 264

N Naeglaeria gruberi, 25 Neisseria N. gonorrhoeae, 202 N. meningitidis and cell-surface

polysaccharide biosyn thesis export, 172, 178-180, 184 genetics, 202, 208 process, 159, lWl6.5, 170 regulation, 22Y structure and attachment, 140, 146, 148, 152 Neurospora, N N. crassa, 278

nicotinic acid and selenium-dependent enzymes, 73. 87 Nocardia lactamdurans, 294-298 non-reducing terminus, growth of 0polysaccharide at, 161-164 Nostoe. 25 1 npf’ mutants in Physarum polycephalum, 34-37 nuclear chromosomal genome of Physarum polycephalum, 6-4 nucleolar DNA genome of Physarum polycephalum, 8-10 0 3-deoxy-D-manno-octulosonic acid (KDO)

SUBJECT INDEX

and cell-surface polysaccharide biosynthesis, 183, 191, 202 process, 163. 165-166 structure and attachment, 140, 144. 148 0-polysaccharides see cell-surface polysaccharide biosynthesis outer-membrane proteins, role of, 18.3-185 oxygen and selenium metabolism, 102-103 tension and ethylene production, 277, 279 2-oxyglutarate pathway in ethylene production, 281, 284-287, 295. 302

P p34 and mitotic regulation in Physarum polycephalum, 55-58 Padina urborescens, 279, 280 Pasturella haemolytica, 146-1 47, 152 Penicillium P. chrysogenurn, 294-298 P. corylophilum, 278 P . cyclopium, 278, 284-285 P. digitatum, 27b28 1 , 284-290, 302 P. luteum. 278 P. patulum, 278 pentacyclic skeleton from isopcntenyl pyrophosphate, 264-268 pentacyclic triterpenoids of hopane series, 247, 250 periodic variations in mitotic cycle in Physarum polycephalum , 4 2 4 8 Persea americana, 294-295 Petunia hybrida, 294-295 Phaeolus schweinitzii, 278 Phanerochaete chrysosporium, 278 279 Pholiota adiposa, 278 phosphate flow and sporulation in Bacillus suhtilis, 120-123 see ulso phosphorelay level and Physarum polycephalum, 4 - 4 1 . 45 phospholipids and hopanoids, 256, 259 phosphorelay in sporulation of Bacillus suhtilis, 113-120 control of, 120-126 see also kinases; s p 0 0 genes phosphorylation in histone modification. 45, 46 Phy comyces P. blakesleeanus, 278 P. nitens. 278, 284 Phycoporus coccineus, 278 Physarum polycephalum, 1-69 genome organization, 6-13 introduced molecules, 58-62 ~

339

life cycle, 3, 4-6 see also cytoskcletal organization; mitotic cycle plasma protein P, 73 plasmid-encoded genes for 0pol ysaccharides, 196- 197 plasmodia plasmodia1 mitotic cycle in Physarum polycephalum. 29-31, 39-42 plasmodia1 phase of Physarum polycephalum, 3. 4-5 see also Physarum polycephalum polymerization reactions in cell-surface polysaccharide biosynthesis, 159-168 group41 capsular polysaccharides polymerized in E. coli, 164-166 non-reducing terminus, growth of 0pol ysaccharide at, I6 1-1 64 reducing terminus, growth of 0polysaccharide at, 10-161 undecaprenol-independent mechanisms, 166-168 polypeptides see tubulins polysaccharides see cell-surface pol ysaccharides Porphyria tenera, 280 processes of cell-surface polysaccharide biosynthesis, 154-168 polysaceharide-modification reactions, 168-171 undecaprenol-linked intermediates formed, 154-159 see also polymerization reactions profilin and Physarum polycephalum, 13, 38 prokaryotic selenoproteins, 73-86 formate dehydrogenases, 76-84 glycinc reductase, 72, 73-76, 89 hydrogenases, 84-86 see also FDH protaminc in histone modification, 46 proteins outer-membrane, role of, 183-185 protein-protein interactions and translational regulation in cell-surface polysaccharide biosyn thesis, 228-229 Proreus, 99 P. mirabilis, 143 P . vulgaris, 91 Pseudomonas, 100, 262 P . aeruginosa and cell-surface polysaccharide biosynthesis export, 184 genetics, 19bIW. 206 process, 166-167 regulation, 214, 221, 224-226

340

SUBJECT INDEX

Pseudomonas (cont.) structure and attachment, 142, 144, 146, 149 P. atlantica, 214, 227 P. caroxydopava, 88 P. puorescens, 218 P. indigofera, 278 P. mevalonii, 268 P. pisi, 282 P. pufida, 223 P. sesami, 261 P. solanacearum. 196, 205, 223. 217, 278, 286 P. syringae, 144 P. syringae and P. syringae p.v. phaseolicola and ethylene production, 277-279, 303 biosynthetic pathways, 281, 284-288 comparison with related enzymes, 295-302 mechanisms for, 288-292 molecular cloning and expression of gene, 292-293 P. syringae pv. glycinea, 278 P. syringae pv. mori, 278 Pueraria lobara, 277, 288 pyruvate and hopanoids 264

Rhizobium, 149 R. fredi, 168 R. leguminosarum, I%, 198,205,215,229 R. melilofi and cell-surface polysaccharide biosynthesis export, 179-180 genetics, 190, 196, 204-205 regulation, 214-215, 223, 229 structure and attachment, 145-146, 158 R. trifolii, 158, 278, 282 Rhodobacfercapsulatus, 255. 269 Rhodocyclus gelatinosus, 255 Rhodomicrobium vannielii, 251, 255 Rhodopseudomonas, 264 R. acidophila, 251, 255, 262, 268 R. palusfris, 251, 255, 261, 262, 268 R. viridis, 141 Rhodospirillum rubrum, 251, 255 ribosomal DNA replication in Physarum polycephalum, 52-53 RNA and Physarumpolycephalum, 7,10, 12,25, 37, 43 and selenium metabolism, 8 H 9 , 94, 95-96, 97, 102, 104 see also genetics; selenocysteyl-tRNA

R

S Saccharomyces cerevisiae comparison with Physarum polycephalum, 7, 10, 16,4n, 57 and ethylene production, 278, 279, 292 and hopanoids, 255, 268, 269 Salmonella, 99, 102, 141, 14.5-146, 149-151, 214 S. enferica and cell-surface polysaccharide biosynthesis, 215 export, 172-177, 181, 185, 187 genetics, 189-196, 198, 206-211 process, 154-155, 159-161, 169-170 structure and attachment, 144, 148-149 S. enferica serogroup kentucky, 156 S. enterica serogroup newport, 156 S. enterica serogroup seftenberg, 156 S. enterica serovar anatum, 156, 160, 169-170, 185, 187, 195,213 S. enferica serovar boecker. 149 S. enrerica serovar dublin, 196, 220 S. enferica serovar madelia, 149 S. enterica serovar minnesofa, 144 S. enferica serovar monfevideo, 162 S. enfericaserovarparatyphi, 194.21 1,220 S. enterica serovar typhi, 194. 220 S. enterica serovar fyphimurium, 99. 119 and cell-surface polysaccharide biosynthesis

reducing terminus, growth of 0polysaccharide at, 160-161 reductase, glycinc 73-76 regulation in Bacillus subrilis see transition-state regulators of cell-surface polysaccharide biosynthesis, 2 12-229 liposaccharides, 212-216 see also extracellular regulation of cellsurface polysaccharides mitotic, in Physarum polycephalum, 53-58 factors, 54-55 heterophasic fusions, 53-54 heteroploidic fusion, 53 MPF,p34 and cyclin, 55-58 transcriptional and cell-surface polysaccharide biosynthesis, 223-224 of genes for phosphorelay components, 123-126 relative transcription timing in chromosome replication in Physarum polycephalum, 5&51 repeating unit structure in cell-surface polysaccharides, 144-1 48, I 5 0 4 53 rfe-independent 0-polysaccharides, transport of, 175-177

SUBJECT INDEX

structure and attachment, 144, 154-155 export, 172, 175, 185, 187 genetics, 190-191, 194, 196, 206-207, 208-209 process, 161-162, 169 regulation, 212-213, 220, 227 S. typhimurium see S . enterica serovar typhimurium above Schizophyllum commune, 278 Schizosaccharomyces S . octosporrcg, 278, 279 S . pombe, 16, 17, 56 Sclerotinia laxa, 278 Scopulariopsis brevicaulis, 278 selgenes for UGA-decoding t RNA, 90-9-92.93 selenide, 100, 102-103 selenite and selenate, 99-100, 102 selenium metabolism in micro-organisms, 71-109 biosynthesis see biosynthesis and selenium metabolism geochemistry, 100, 102-103 selenium-containing enzymes see enzymes, selenium-containing selenium-containing tRNAs, 88-89 transport of compounds, 98-99 see also selenoproteins selenocysteine, 72. 97, 98 codons discrimation from stop codons, 93-94 incorporation, evolution of, 94-95 selenocysteyl-tRNA, 73, 91 from seryl-tRNA, 92-93 unique elongation factor for, 93 selenomethionine, 97, 98, 102 Selenomonas ruminantium, 99 selenoproteins, biosynthesis, 89-95 from eukaryotes, 73, 89 gene for UGA-decoding tRNA: selC, W92 from prokaryotes see prokaryotic selenoproteins see aLvo selenocysteine; selenocysteyltRNA; sulphur Serratia S . liquefaciens, 278 S . marcescens, 146-147, 278 seryl-tRNA, selenocystcyl-tRNA from, 91, 92-93 sexual cycle of Physarum polycephalum, 3. 5-6

Shigella. 141 S . boydii, 192, 197, 210-21 I S . dysenteriae, 190, 192-193, 197 S.flexneri, 192, 197 S . sonnei, 192, 197

34 1

side-chains and cell-surface polysaccharide biosynthesis, 168-171 sin gene and transition-state regulators and sporulation in Bacillus subtilis, 128-129 Spongiporus sinuosus, 278 sp00 genes and sporulation in Bacillus subtilis, 1 12-1 20 functions of, 115-116 and initiation of sporulation, 130-131 sensor kinases isolated, 116-1 18 see also phosphorelay ; transition-state regulators sporulation in Bacillus subtilis, 11 1-1 33 alternatives to, 129-130 initiation of, 130-131 phosphorelay, 113-120 control of, 120-126 transition-state regulators, 126.129 squalene biosythesis, 257, 264-267, 269 stable expression and DNA transformation and Physarum polycephalum, 61 Staphylococcus, 262 S . aureus, 278 S . carnosus, 264 star mitosis in Physarum polycephalum, 29-30 sterols in eukaryotes, 250, 258, 2 6 2 6 7 stop codons discrimation from seloncysteine codons, 93-94 Streptococcus S . faecium, 262 S. mutans, 262 Streptomyces, 255, 279-280 S . aeriouvifer, 262 S . cluvuligerus, 294-298 S . griseus, 262 S. hygroscopicus, 262 S . jumonjinesis, 296-298 S . lipmanni, 294-298 S . noursei, 262 S . tendae, 251 structural diversity of hopanoids, 248-249, 250-252 structure modification genes of 0polysaccharides, 197-198 Stylonichia lemmae, 16 succinate and ethylene production, 285 succinoglycan, 145-146 sulphur versus selenium, 96-97, 98, 99-101 in catalysis, 96-97 competition during incorporation, 97 surface association of bacterial polysaccharide biosynthesis, 138-144 Synechococcus, 255

342

SUNJECT INDtX

Synechocystis, 25 1. 255, 258 T tetrahymanol and hopanoids, 258. 2 6 2 6 7 Tetrahymena, 9 T. pyriformis, 258. 266-267 T . thermophila. 8, 62 T G A codon, 72, 89 Thamnidium elegans, 27X Thieluvia data, 278 Thiobacillus T . ferrooxiiiuns, 102, 278 T. novellus, 278. 281 thyroxine de-iodinase, 73 timing in chromosome replication in Physarum polycephalum, 49-5 1 transcriptional regulation and cell-surface polysaccharide biosynthesis, 223-224 of genes for phosphorelay components, 12.3-1 26 transient expression and DNA transformation and Physarum polycephalum, 59-61 transition-state regulators and sporulation in Racillicr sithti1i.s. 126-129 AbrB protein, 127-128 hpr gene. 128 sin gene, 128-129 translational regulation in cell-surface polysaccharide hiosynthesis, 22X-229 transport o f polysaccharides across cytoplasmic memhrane. 175-18 I to cell surface, 182-188 energies of, 1x1 group-11-like capsular polysaccharides, 177-182 rfe-independent 0-polysaccharides, 17.5-177 o f selenium-containing compounds, 98-99 transposons. 7 Trypanosoma. 8 T . hrucei, 12, 17 tryptophan synthase. 9X tubulins and Physarum polycephalum 13, 25, 35-36, 38, 4W-4-11 genes and polypeptides, 14-17 microtuhule-associated proteins, 22-23 multiple, 20-22 periodic variation, 42-44

utilization, 17-20 two-component systems and cell-surface polysaccharide biosynthesis group-I-li ke, 216-22 I and synthesis o f alginate, 221-222 and transcriptional regulators, 22.3414 Tyromyces palustris, 278 U

ubiquitination in histonc modification, 45. 46 U G A codon, 89.93-9.5 UGA-decoding tRNA, gene for. 90-02 undecaprenol -independent mechanisms in cell-surface polysaccharidcs, 166-16X -linked intermediates formed in cellsurface polysaccharides, 1 54-159

v

Veillonella ulcalescens, 102 viantigen, 145 Vihrio chobrue, 192, 2OX-209 Viola odorata. 255

W Wolinellu succinogenes, 77, 78. 80, 83, Oh

X xanthine dehydrogenases and seleniumdependent cnzymcs, 73, 8 6 4 7 Xunthomonas cumpestris, 278 and cell-surface polysaccharide hiosynthesis genetics, 205, 21 I process, 156, 170--171 regulation, 214, 223, 224-225, 229 Xenopus laevis, 56, 57. 58, 292

Y Yersinia Y . enterocoliticu. 192-193, 213 Y . pseudotuherculosis. 192, 210 Y . ruckeri, 144 I,

Zea mays, 294-295 zones of adhesion see adhesion Zooloea rumigera, 214. 226 Zymomonas mohilis and hopanoids 25 I 254-259, 262-264, 26x-269