* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Were Gram-positive rods the first bacteria?
Survey
Document related concepts
Signal transduction wikipedia , lookup
Cytoplasmic streaming wikipedia , lookup
Tissue engineering wikipedia , lookup
Cell membrane wikipedia , lookup
Extracellular matrix wikipedia , lookup
Endomembrane system wikipedia , lookup
Cell encapsulation wikipedia , lookup
Cellular differentiation wikipedia , lookup
Programmed cell death wikipedia , lookup
Cell culture wikipedia , lookup
Organ-on-a-chip wikipedia , lookup
Cell growth wikipedia , lookup
Transcript
166 Opinion TRENDS in Microbiology Vol.11 No.4 April 2003 Were Gram-positive rods the first bacteria? Arthur L. Koch Biology Department, Indiana University, Jordan Hall 142, 1001 East Third Street, Bloomington, IN 47405– 6801, USA At some point in the evolution of life, the domain Bacteria arose from prokaryotic progenitors. The cell that gave rise to the first bacterium has been given the name (among several other names) ‘last universal ancestor (LUA)’. This cell had an extensive, well-developed suite of biochemical strategies that increased its ability to grow. The first bacterium is thought to have acquired a covering, called a sacculus or exoskeleton, that made it stress-resistant. This protected it from rupturing as a result of turgor pressure stress arising from the success of its metabolic abilities. So what were the properties of this cell’s wall? Was it Gram-positive or Gram-negative? And was it a coccus or a rod? Bacteria evolved as a separate domain some time before Archaea or Eukarya [1]. They developed at a critical time that defined the ‘last universal ancestor (LUA)’, and arose after the basic groups of physiological processes had evolved to a functional point. These major processes must have included DNA replication, ribosome and protein synthesis, central intermediate metabolism, cell division, DNA repair and responses to fluctuation in the environment. Such properties are possessed by almost all modern cells [2] and each would have improved their ability to grow compared with the first living cell on this planet. The differences among the bacterial, archaeal and eukaryotic cells are slight compared with their similarities [2]. One important difference between the immediate precursor of the three cell domains and the first bacterium was the latter’s development of a functional and stressresistant wall [2,3], which allowed it to support a higher internal osmotic pressure and remain intact in a low osmotic pressure environment. The issue here is the nature of this cell wall. Was the wall thick or thin? Was it a coccus or a rod-shaped bacillus (Figs 1 and 2)? In the Kandler and Schleifer designation system [4,5], what category would it be? That is, what was the chemical structure of the original bacterium’s cell wall? This classification system uses three symbols. Thus, for Escherichia coli, it is A1g. The A indicates that the third amino-acid residue is meso-diaminopimelyl; the 1 indicates that there is no intervening peptide that bridges between the two peptides at the tail-to-tail linkage; and the g indicates that the third amino acid is meso and not LL. Corresponding author: Arthur L. Koch ([email protected]). The first cells: Gram-positive or Gram-negative? Arguments for Gram-positive rod-shaped cells Arguments from three groups suggest that the first cell to separate from the monophyletic prokaryotic predecessor of the bacteria was a Gram-positive, rod-shaped organism. Seifert and Fox [6] noted that rod-shaped structures cluster at the base of the bacterial phylogenetic tree. They compared the morphology of the cells in the various branches of the Woese’s 16S rRNA tree (see Ref. [1] and Olsen et al. [7]). Seifert and Fox stated: ‘It [also] seems likely that the last common ancestor of the domain Bacteria was rod-shaped’. Tamames et al. [8] drew conclusions from the analysis of the dcw (division and cell wall) clusters present in many bacteria. There are 15 genes in the dcw cluster of Escherichia coli and around the same number in other species. Tamames et al. compared the coding order of the genes (i.e. the sequence of genes on the chromosome) in a variety of species. They found that there was a pattern, and that the dcw cluster was more compact and conserved in bacilli, implying that rod-shaped bacteria came first. In the bacilli, in contrast to the cocci that they studied, there was a more constant order on the chromosome. These bacilli encompassed both Gram-positive and Gramnegative forms. Gupta et al. [9– 11] analyzed the completed, published sequences of many genomes, both bacterial and archaeal, and concluded that Gram-positive bacteria arose first, and that Gram-negative bacteria arose from Gram-positive bacteria through a sequence of several other groups. Gupta’s phylogenetic tree [11] corroborates the standard 16S rRNA tree [12]. However, the Woese group has presented convincing evidence from the 16S rRNA Rod-shaped Bacillus Coccus Gram-positive Gram-negative TRENDS in Microbiology Fig. 1. Four possibilities for the wall of the first bacterium. These four types represent a majority of organisms. There are other shapes (curved, spiral and tapered) but these are probably less likely than the initial bacterial form. http://timi.trends.com 0966-842X/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0966-842X(03)00063-5 Opinion TRENDS in Microbiology 167 arose only when there were plants, animals and fungi to be parasitized and resisted. As a consequence, their evolution would have been a later event. Gram-positive Older wall Cytoplasmic membrane Gram-negative Penta-muropeptides Nonamuropeptide Up peptide To other tessera Down peptide TRENDS in Microbiology Fig. 2. Wall-side wall growth mechanisms. In Gram-positive organisms, the wall is formed by the continuous laying down of layers of murein just outside of the cytoplasmic membrane. After they are hydrolysed the layers become highly stretched and peripheral. This mode of the addition of underlying layers that move to the outside of the cell wall is called the inside-to-outside mechanism. In Gram-negative organisms, the wall grows by the extrusion of penta-muropeptide through the cytoplasmic membrane and their incorporation into a functioning wall structure under tension. This mode is called the nana-muropeptide stress mechanism. sequences to show that Archaea and Eukarya separated from a prokaryotic precursor and are not derivatives of the Bacteria [1,12], as Gupta believes [9]. Although the origin of Archaea or Eukarya is not directly relevant to the present discussion, this conflict is pertinent. Gupta et al. [9– 11] looked for corresponding regions in the available sequenced genomes and for differences characteristic to all members of a taxonomic group and that, additionally, were unique to some, but not other, taxonomic groups. Such differences of ‘significance’ they called ‘indels’ (insertions/deletions); consequently, these indels correlate with a taxonomic group. When these persisted in all members of a taxonomic group, Gupta concluded that the group originated from a founder cell generated from another phylogenetic group that, by chance, happened to have this particular indel. The indel was therefore present in all members of the group, and then was passed on when a founder from this group led to a newer group. Thus, the indel was common to all members of a group no matter what other lines of diversification occurred within it later. The important conclusion drawn was that major bacterial taxa arose linearly from each other [9,11] and that cells with one membrane, such as Gram-positive organisms, called ‘monoderms’, are the precursors of the cells with both a cytoplasmic membrane and an outer membrane. These are Gram-negative organisms, such as E. coli. All of these are called ‘diderms’ by Gupta’s group. A supporting argument for this order of evolution is that the Gram-negative cells are structurally better able to function as pathogens and to resist antibiotics. This would be in accordance with the idea that the Gram-negative cell http://timi.trends.com Vol.11 No.4 April 2003 Arguments for Gram-negative cells Woese [1] argued that Bacteria arose some time earlier than the split that led to the separation of Archaea and Eukarya from the ‘progenote’ or LUA. The derived phylogeny of Bacteria, which is currently well accepted, does not group the Gram-positive organisms together and does not group the rod-shaped cells separately from the cocci. A major proponent for the idea that Gram-negative cells arose first is Cavalier-Smith, who has presented extensive discussions of the origin and evolution of life [13,14]. He argued first that life started on the outside of a bilayered lipid vesicle that had been produced abiotically [15]. If life started on the outside of the lipid membrane, the problem of transport across a lipid bilayer of hydrophilic material is avoided but the theory requires that biomolecules of crucial importance remain attached to the outer surface. Later, when life had developed adequately, this phospholipid vesicle engulfed the living portion. When fusion was complete, this formed the first cell that was surrounded by two bilayers, and was thus a Gram-negative organism. The space between these layers corresponds to the periplasmic space. Cavalier-Smith’s suggestion [15] is a modification of one proposed by Blobel [16], who originally suggests that the cell started ‘inside out’. Blobel presumed that life arose on a solid rock surface. Cavalier-Smith proposes that this location solves the energy problem because of the presence of polyphosphates in the rocks, and that only later did invagination take place, he assumes to convert this ‘obcell’ (obverse cell) to the usual arrangement [15]. Some descendents of this cell later degenerated to become the single-membraned, thick-cell-walled Gram-positive cells. Cell morphology: cocci or rod-shaped Until the cell developed mechanisms to establish its shape, the default appears to be that the phospholipid- or lipidbilayer-enclosed cell formed an ever-increasing sphere as it grew. Thus, unless the cell had already developed a cytoskeleton or a strong wall (that is, unless it had an exoor endo-skeleton), there was no mechanism to make it divide to produce new cells or cells of a constant mean size or particular shape. If the cell was not attached to any object, it must have grown larger and larger without limit. This certainly is not a working growth strategy. Attachment to surfaces could have helped such a cell to divide, although this would have happened irregularly and formed daughters of irregular shape. Only one alternative to this theory has been suggested in the literature as a possible mechanism for division during the time between the first cell and the LUA [17]. This proposal is that the lipid constituents of the bilayer were generated within the cell, and that they became inserted into the inner layer of the bilayer and this uneven growth caused the bilayer to invaginate. Eventually, this would lead to cell division. This strategy might have worked but would have been irregular and would not have been very effective [17]. It was possibly up to 800 million years later, when a 168 Opinion TRENDS in Microbiology prokaryote completed the evolution of a stretch-resistant ‘fabric’, that an effective alternative arose, which I argue was the creation of the domain Bacteria. Similarly, the development of pseudomurein could have led to the creation of the first Archaea. Of course, later, the cytoskeleton arose together with contractile proteins used by Eukarya. It is not enough in the development of bacteria to develop a means of forming an enclosing strong murein sacculus. Several other mechanisms must also have arisen. A mechanism that prevents wall growth in the established poles of cells is a key requirement. A mechanism that causes a pole to be metabolically inert provides a way to maintain the size and shape of cells in succeeding generations. For rod-shaped cells, inert poles provide support for the elongation of cylindrical growth [18]. However, this can only function to the degree that the poles are metabolically inert and rigid [19]. Thus, the inertness of the poles can also be the basis for the maintenance of the diameter of cylinder-shaped cells in balanced growth. Cell mechanisms must also function to foster wall growth in an amount consistent with the ongoing rate of protoplasmic synthesis. That is, cell biochemical growth must drive the enlargement of the cell wall. These aspects are, together, probably the answer to the question: why do bacteria not grow larger and larger and rounder and rounder? Experimental evidence for the poles of bacteria started with Cole’s and Hahn’s studies [20] of Streptococcus pyogenes, and Doyle’s early unpublished studies of Bacillus subtilis (published in [21]).These studies found that the established poles of both this Gram-positive coccus and rod were metabolically inert. There is now definitive evidence for the inertness of the poles of B. subtilis [22– 24] and of E. coli ([25]; A.L. Koch and M.A. De Pedro, unpublished). If these findings apply to other bacteria and it is found that the poles of bacteria, other than mycoplasma, are rigid, metabolically inert and cannot stretch further, then this rigidity is probably the defining feature of the domain of bacteria. The pole metabolisms of B. subtilis and E. coli only have been studied in sufficient detail to consider the question of metabolic turnover in the poles. The experimental finding is that the turnover at the poles is negligible [22,24]. However, the sidewalls of rod-shaped cells turn over with a half-life equal to their growth doubling time [24]. The inert nature of the poles is surprising from the biochemical point of view because attempts to find a significant chemical difference in the murein wall of the poles and of the sidewall regions of B. subtilis have failed [26]. Rigid poles allow the cells to maintain their size during growth because the poles provide a template for wall enlargement, and this determines the diameters of the next generation of poles. When initially linked into the wall of stress-bearing organisms, newly polymerized wall is not extended to its maximum size. Although the murein is elastic when inserted, it is unstretched. Of course, it will come to be stretched during growth. If there were no special controls on insertion and it occurred randomly over the cell surface, the cell would bulge like a soap bubble (as various bacteria do when treated with appropriate http://timi.trends.com Vol.11 No.4 April 2003 antibiotics [27]). The physical law, derived by LaPlace two centuries ago, would lead to cylindrical extension [28] if the poles were rigid. His law related the radius of a ‘bubble’ of arbitrary shape to the pressure difference across the wall, the amount of work needed to increase the surface by a unit amount and constraints upon its surface (see Ref. [34]). The implication of bacterial poles being metabolically inert is that the first bacterium, like all bacteria (with the exception of mycoplasma), formed new poles entirely by new synthesis and, subsequently, did not enlarge or alter them. This implies that the biochemistry and biophysics of bacterial cells must be such that the mature poles are blocked from further metabolism or turnover. This prevents the size of completed poles from being further modified by insertions or being turned over in future generations. It therefore appears likely that cocci only need to be able to form a septum centrally, and grow without changing their maximum diameter and the dimensions of a mature pole. They then only needed to allow or aid the septum to split to form two new daughter cells of the same diameter. This is the way that the coccus Enterococcus hirae and the rod Bacillus subtillis divide [28– 32]. Evidence from the electron microscope is that, at some critical stage of the cell cycle, a septum starts to grow inwards (from the site of the previous septa or in the middle of the cylindrical region). As it is formed, it starts to split from the outside in and the intervening split septal wall bulges outwards, forming two new poles. Consequently, the new pole has the same diameter as the old one and, in the next generation, these poles are the templates for the new nascent poles. This means that the diameter of poles in a culture in balanced growth is remarkably constant (^ 5%; Ref. [29]). The growth of cocci is clearly simpler than that of rodshaped organisms, which must form the cylindrical walls by a separate process in addition to septal formation and splitting. This would suggest that that the first bacterium was a coccus [33]. This is not in accordance with the logical extension of the ideas of Woese [1], or those of Seifert and Fox [6], Vicente’s group [8] and Gupta’s group [9]. Consequently, other roles for rod-shaped cells and their advantages will be considered later in this article to support their points of view. Wall growth mechanisms for the first bacterium The strategies for wall growth of Gram-positive and Gramnegative cells briefly presented here have been studied both theoretically and experimentally. The mechanism for Gram-positive cells is well-established [34]. Briefly, new layers of wall are added from the cytoplasmic membrane surface and autolysis removes the oldest wall. This is called the Gram-positive ‘inside-to-outside’ mechanism. By contrast, the mechanism for Gram-negative bacteria requires that new wall units be synthesized, inserted through the cytoplasmic membrane and covalently inserted into the stress-bearing wall. The most recent proposal depends on stress in the growing wall, altering the conformation of the new wall. This model is called the ‘nona-muropeptide stretch’ mechanism. It can be appreciated that both sidewall and pole formation for Opinion TRENDS in Microbiology Gram-positive organisms are quite simple compared with the process used by the thin-walled Gram-negative cell. What are the possibilities for how saccular growth occurred in the first place? Could either the mechanisms for Gram-positive or Gram-negative cells be those that functioned for the progenote LUA cell that had just perfected methods to form a cross-linked ‘fabric’ outside of its cytoplasmic membrane? Being able to form a polymer outside the cytoplasmic membrane is very complex [34] and there must have been many problems to overcome. However, in addition for bacterial growth, there must be mechanisms directing the insertion and leading to cell division. It is here at the LUA stage that an additional set of mechanisms had to be generated. Although the early mechanisms may not have been as sophisticated as those used in modern bacteria, at the start of the bacterial domain they had to be sufficient, simple and functional. On the basis of the Gram-positive and Gram-negative possibilities, it would appear plausible that something like a Gram-positive ‘inside-to-outside’ mechanism would be much simpler than the ‘nona-muropeptide stretch’ mechanism for Gram-negative cells. Moreover, it seems selfevident that a coccus instead of a rod-shaped organism is simpler and should have arisen earlier. So the first hunch could be that the original bacterium was a Gram-positive coccus. For division, this cell must have been capable of forming a septum or crosswall, possibly with a mechanism similar to that used for crosswall formation in B. subtilis or Enterococcus hirae. The septal crosswalls of Gram-positive cells are much thicker than a monolayer of murein and are around the same thickness as the Gram-positive sidewall of B. subtilis. Conclusions The first bacterial cell appears to have had a sacculus [2,3,34]. But other thoughts about the first member of the domain of Bacteria are more varied. It may have been a Gram-positive organism, that is, if Gram-positivity only denotes a thick peptidoglycan layer. I argued earlier in this article that the Gram-positive strategy for growth is much simpler than that of Gram-negative cells. Cavalier-Smith argued that a Gram-negative cell appeared first, although he considered mostly the membranes and not the murein wall and its role in bacterial growth [14]. It is probably the case that, during the time between the origin of the first cell and the first bacterium, cells had no murein layer. A wall was needed after organisms gradually became more successful because only then did their turgor pressure increase. The simplest mode for Gram-negative growth suggested so far is the nona-muropeptide stretch model, which is much simpler than earlier models [34]. If it functioned at a time when cells had only a partial functional sacculus and did not involve enzymes passing through the cytoplasmic membrane, proteins in holoenzyme clusters would not have been required, unlike in Höltje’s ‘three-for-one’ model [35]. Höltje’s model postulates that penta-muropeptides are secreted through the cytoplasmic membrane and are linked by glycan bonds to form oligopeptides of the same length as a template portion of the stress-bearing wall. Three of these chains are linked together by tail-to-tail bonds and the resultant ‘raft’ http://timi.trends.com Vol.11 No.4 April 2003 169 inserted around the template chain that is then removed from the wall. This enlarges the wall by doubling the area of the template strand. At present this model is the most well known model in the literature. However, with either model it is difficult to imagine from known biochemistry that the first cell was a Gram-negative cell, as CavalierSmith would postulate [13– 15]. The coccus is the simplest of possible cell shapes. If E. hirae were taken as the model for the first bacterium, such a cell would only need to form a thick septum and bisect it and then let the physical forces do the rest. The semiconservative process described above for E. hirae would not be too difficult to implement simply with a few or no extracytoplasmic enzymes. The reasons why a rod-shaped bacterial organism might have come first have been presented in this Opinion. This is the conclusion of three groups based on three quite different arguments. There are biophysical arguments, based on the kinetics of uptake issues [34], for the advantages of rod-shaped organisms over coccal-type cells. There are several different cases for diffusion up to the cell. In an environment with significant concentrations of resources, if every unit of surface area has the same concentration of uptake sites, the only relevant factor is the ratio of surface to volume. Smaller cells have higher ratios. For a fixed volume, thin rods or flat leaf-like structures have higher ratios. By contrast, spheres have the lowest ratio. Therefore, a narrow rod would have been an optimal shape. If different regions of the cell under different environmental conditions have a different number of absorbing sites, then the actual surface area of the cell is not the only significant quantity, and the number and kind of uptake systems per unit of surface area must also be taken into consideration. However, when nutrients are present in very low concentration, the factor of prime significance is the diffusion process from the bulk medium to the cell. Now the mathematics for different-shaped objects is quite different. But the effect of cell shape is actually minor in either low or high substrate concentration. Of course, even if the effect on growth rate is very slight over many generations, it might still be important. The general conclusion from these considerations is that a rod shape is better than a coccoid shape for a cell of fixed volume. Are there advantages of rod-shaped growth over coccaltype growth? Two potential advantages can be suggested. One is that the individual cells in a massed aggregate of cells might have a more effective exposure to the environment if the aggregate is an assortment of randomly oriented rod-shaped cells than if it is a compact mass of more spherical particles. This is a possibility that must be mathematically explored. The second, and more probable, suggestion is that a rod shape is useful when cell division is controlled by cell growth. At some point in the E. coli cell cycle, the initiation of DNA replication takes place and this, in turn, controls the timing of the subsequent cell division [36]. This strategy may have been selected to preclude the possibility that the cell division event would interfere and break a chromosome. Consequently, the rod-shaped mode of cell growth is safer for the genetic integrity of the cell. 170 Opinion TRENDS in Microbiology Contrast this with the behavior of cocci. In E. hirae, it appears that splitting of the septum is initiated when wall growth cannot occur fast enough for the cell’s needs for space for its cytoplasm [29– 31]. Because of the disparity of the ratio of volume growth rate to wall surface growth rate throughout the cell cycle, the trigger might be turgor pressure. Turgor pressure must increase inside the cell towards the end of septal closure and it is the pressure that triggers the wall splitting event. Such a system could err if the chromosome and the cell division cycle were not in proper temporal relationship, but would require an additional regulatory system. Relevant to this, Seifert and Fox [6] observed that, in several branches of the 16S rRNA evolutionary tree, the cells switched from rod-shape to cocci but then did not revert in further evolution. The first bacterium might have been a rod, to allow safe growth but much later in the development of the bacterial clades, genetic deletions occurred to eliminate the sidewall region. This could have been dangerous, unless new developments of various kinds occurred to control the chromosome cycle to be in accordance with the cell division cycle. This is an attractive possibility but more sequence data for more species will be needed to check the mode of cell growth. However, for now, this is the best guess for the apparent initial role of rods and subsequent emergence of cocci. This hypothesis would require early and sophisticated linkages between the morphological aspects of the cell cycle with the chromosome cycle. Vol.11 No.4 April 2003 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 References 1 Woese, C.R. (1987) Bacterial evolution. Clin. Microbiol. Rev. 51, 221 – 271 2 Koch, A.L. (1994) Development and diversification of the last universal ancestor. J. Theor. Biol. 168, 269 – 280 3 Koch, A.L. (1995) Origin of intracellular and intercellular pathogens. Q. Rev. Biol. 70, 423 – 437 4 Schleifer, K.H. and Kandler, O. (1972) Peptidoglycan types of bacterial cell walls and their taxonomic implications. Bacteriol. Rev. 36, 407 – 477 5 Tipper, D.J. and Wright, A. (1979) Structure and biosynthesis of bacterial cell walls. In The Bacteria (Sokatch, J.R. and Ornston, L.N., eds) pp. 291 – 426, Academic Press 6 Seifert, J.L. and Fox, G.E. (1998) Phylogenetic mapping of the bacteria morphology. Microbiology 144, 2803– 2808 7 Olsen, G.J. et al. (1994) The winds of (evolutionary) change: breathing new life into microbiology. J. Bacteriol. 176, 1 – 6 8 Tamames, J. et al. (2001) Bringing gene order into bacterial shape. Trends Genet. 17, 124 – 126 9 Gupta, R.S. (2002) Phylogeny of bacteria, are we now close to understanding it? ASM News 68, 284 – 291 10 Gupta, R.S. and Griffiths, E. (2002) Critical issues in bacterial phylogeny. Theor. Popul. Biol. 61, 423 – 434 11 Gupta, R.S. (2001) The branching order and phylogenetic placement of 27 28 29 30 31 32 33 34 35 36 species from completed bacterial genomes, based on conserved indels found in various proteins. Int. Microbiol. 4, 187– 202 Olsen, G.J. (2001) The history of life. Nat. Genet. 28, 197– 198 Cavalier-Smith, T. (2001) Obcells as proto-organisms: membrane heredity, lithophosphorylation, and the origins of the genetic code, the first cells, and photosynthesis. J. Mol. Evol. 53, 555 – 595 Cavalier-Smith, T. (2002) The neomuran origin of Archaebacteria, the negibacterial root of the universal tree and the bacterial megaclassification. Inter. J. System. Evol. Microbiol. 52, 7 – 76 Cavalier-Smith, T. (1987) The origin of cells: a symbiosis between genes, catalyst, and membranes. Cold Spring. Harbor Symp. Quant. Biol. 52, 805 – 824 Blobel, G. (1980) Intracellular membrane topogenesis. Proc. Natl. Acad. Sci. U. S. A. 77, 1496 – 1500 Koch, A.L. (1985) Primeval cells: possible energy-generating and celldivision mechanisms. J. Mol. Evol. 21, 270– 277 Koch, A.L. (1982) On the growth and form of Escherichia coli. J. Gen. Microbiol. 128, 2527 – 2540 Koch, A.L. (2002) Why are rod-shaped bacteria rod-shaped? Trends Microbiol. 10, 452– 455 Cole, R.M. and Hahn, J.J. (1962) Cell wall replication in Streptococcus pyogenes. Science 135, 722 Koch, A.L. et al. (1981) The coupling of wall growth and chromosome replication in Gram-positive rods. FEMS Microbiol. Lett. 12, 201 – 208 Merad, T. et al. (1989) Cell wall assembly in Bacillus subtilis: visualisation of old and new material by electron microscopic examination of samples selectively stained for teichoic acid and teichuronic acid. J. Gen. Microbiol. 135, 645– 655 Clarke-Sturman, A.J. et al. (1989) Cell wall assembly in Bacillus subtilis: partial conservation of polar wall material and the effect of growth conditions on the pattern of incorporation of new material at the polar caps. J. Gen. Microbiol. 135, 657 – 665 Koch, A.L. and Doyle, R.J. (1985) Inside-to-outside growth and the turnover of the Gram-positive rod. J. Theor. Biol. 117, 137– 157 De Pedro, M.A. et al. (1997) Murein segregation in Escherichia coli. J. Bacteriol. 179, 228– 234 Buchanan, C. (1979) Altered membrane proteins in a mini-cell mutant of Bacillus subtilis. J. Bacteriol. 139, 305– 307 Schwarz, U. et al. (1969) Autolytic enzymes and cell division of Escherichia coli. J. Mol. Biol. 41, 419 – 429 Koch, A.L. et al. (1981) Surface tension-like forces determine bacterial shapes: Streptococcus faecium. J. Gen. Microbiol. 123, 151 – 161 Higgins, M.L. and Shockman, G.D. (1970) Model for cell wall growth of Streptococcus faecalis. J. Bacteriol. 101, 643– 648 Koch, A.L. and Burdett, I.D.J. (1986) Biophysics of pole formation of Gram-positive rods. J. Gen. Microbiol. 132, 3451– 3457 Koch, A.L. et al. (1982) The role of surface stress in the morphology of microbes. J. Gen. Microbiol. 128, 927 – 945 Koch, A.L. and Higgins, M.L. (1984) Control of wall band splitting in Streptococcus faecalis ATCC 9790. J. Gen. Microbiol. 130, 735 – 745 Koch, A.L. (1996) What size should a bacterium be? A question of scale. Annu. Rev. Microbiol. 50, 317 – 334 Koch, A.L. (2001) Bacterial growth and form, 2nd edn, Kluwer Höltje, J-V. (1993) ‘Three for One’: A simple growth mechanism that guarantees a precise copy of the thin, rod-shaped sacculus of E. coli. In Bacterial Growth and Lysis, Metabolism and Structure of the Bacterial Sacculus (de Pedro, M.A., Höltje, J-V. and Löffëlhardt, W., eds) pp. 419– 426, Plenum Press Cooper, S. (1991) Bacterial Growth and Division, Academic Press News & Features on BioMedNet Start your day with BioMedNet’s own daily science news, features, research update articles and special reports. Every two weeks, enjoy BioMedNet Magazine, which contains free articles from Trends, Current Opinion, Cell and Current Biology. Plus, subscribe to Conference Reporter to get daily reports direct from major life science meetings. http://news.bmn.com http://timi.trends.com