The Staphylococcal Cell Wall

ABSTRACT

Dating back to the 1960s, initial studies on the staphylococcal cell wall were driven by the need to clarify the mode of action of the first antibiotics and the resistance mechanisms developed by the bacteria. During the following decades, the elucidation of the biosynthetic path and primary composition of staphylococcal cell walls was propelled by advances in microbial cell biology, specifically, the introduction of high-resolution analytical techniques and molecular genetic approaches. The field of staphylococcal cell wall gradually gained its own significance as the complexity of its chemical structure and involvement in numerous cellular processes became evident, namely its versatile role in host interactions, coordination of cell division and environmental stress signaling.

This chapter includes an updated description of the anatomy of staphylococcal cell walls, paying particular attention to information from the last decade, under four headings: high-resolution analysis of the Staphylococcus aureus peptidoglycan; variations in peptidoglycan composition; genetic determinants and enzymes in cell wall synthesis; and complex functions of cell walls. The latest contributions to a more precise picture of the staphylococcal cell envelope were possible due to recently developed state-of-the-art microscopy and spectroscopy techniques and to a wide combination of -omics approaches, that are allowing to obtain a more integrative view of this highly dynamic structure.

HISTORICAL OVERVIEW

Ever since the recognition of the cell wall in the late 1940s to early 1950s as a unique anatomical component of all eubacterial cells, Staphylococcus aureus has often served as the Gram-positive model in wall-related studies. One of the first demonstrations that bacterial cell walls can be isolated as physical entities with the size and shape of the whole bacterium was with S. aureus. It was in penicillin-treated S. aureus that the UDP-linked amino sugar-containing wall precursor peptides were discovered, providing the first insights into the unique building blocks of cell wall biosynthesis. The history of interest in the staphylococcal cell wall also reflects the history of success and failure of the antibiotic era. The clue that eventually led to the discovery of penicillin (and later to the autolytic enzymes) was provided by the lysis of staphylococcal colonies in the vicinity of a mold contaminant on an agar plate in Fleming’s laboratory. Elucidation of the mode of action of several important antibiotics in the 1960s and 1970s has been intimately linked to studies of the biosynthesis of staphylococcal cell walls. This included studies of the mode of action of penicillin and other β-lactam antibiotics as specific inhibitors (acylating agents) directed against the active site of penicillin binding proteins (PBPs)—transpeptidases with or without an additional transglycosylase function—which catalyze terminal stages in the assembly of the bacterial cell wall. It was mainly from studies of S. aureus and from parallel studies of Escherichia coli that by the early 1980s a coherent picture emerged of the biosynthetic pathway that leads to the formation of the lipid-linked disaccharide pentapeptide, which with some structural variations, is the universal building block of cell wall peptidoglycan in both Gram-positive and Gram-negative bacteria. Reviews and references summarizing various aspects of studies of the cell walls of staphylococci up to the late 1980s are available (1, 2).

Changing Image of Cell Walls

During the early antibiotic era, biochemists and microbiologists working with cell walls viewed these as more or less inert exoskeletons essential to withstand the turgor pressure of cytoplasm. However, these views are changing rapidly with the recognition of the complexity of chemical structure and biosynthetic pathways and the large number of genetic determinants involved with the synthesis of the cell wall. Cell walls are also intimately involved with host-related functions of staphylococci (see “Complex Functions of Cell Walls” below). The new image of cell walls emerging is that of a dynamic and very live structure. The assembly and replication of cell walls in dividing cells pose some of the most challenging questions of microbial cell biology, surpassing in complexity the questions of chromosome replication. How is the flow of precursor molecules and their polymerization on the outer surface of the plasma membrane controlled? What controls the unique species-specific chemical composition of peptidoglycan? What principles govern the organization of wall polymers into supermolecular sacculi which have the same size and shape as the particular bacterium? How and by what mechanism does the nascent innermost layer of this envelope mature while moving outward toward the cell surface? How and why and through what signals and by what catalysts are the outermost layers of cell wall shed into the medium during wall turnover? And how is the spatial and temporal accuracy of wall synthesis coordinated with cell division? Although large and successful efforts have been made in the past decade to provide answers to these questions, the scope of these key issues remains the roots of priority research lines.

ABOUT THIS REVIEW

In addition to the reemergence of interest in cell walls in the context of modern microbial cell biology, some approaches have been making great impact on discoveries in this field: in addition to high-resolution analytical techniques (high-performance liquid chromatography [HPLC] and mass spectrometry) and the application of molecular genetic approaches, more recently, cryo-electron microscopy, confocal fluorescence microscopy, superresolution microscopy, and nuclear magnetic resonance spectroscopy are contributing to a clearer picture of the ultrastructure of the Gram-positive cell envelope. Comprehensive coverage of all the data in a period of such rapid expansion of a field would be difficult. The purpose of an updated review may be better served by putting the interested reader “on track” of some of the most recent findings and emerging trends by quoting the relevant literature. The structure of the article will be as follows. After a brief reminder of the anatomy of staphylococcal cell walls, new information will be reviewed under four headings: –“High-Resolution Analysis of the Staphylococcal Peptidoglycan,” “Variation in Peptidoglycan Composition,” “Genetic Determinants and Enzymes in Cell Wall Synthesis,” and “Complex Functions of Cell Walls.”

ANATOMY OF THE S. AUREUS CELL WALL

Most clinical isolates of S. aureus express on their outermost surface one of the 13 chemically different capsular polysaccharides that have been identified so far (3, 4). The chemical subunit structures of these important and often antiphagocytic carbohydrate polymers have been elucidated, and rapid advances are being made in the identification of genetic determinants and the organization of capsular loci (5). Underneath these somewhat varied surface layers is the staphylococcal cell wall. In electron-microscopic thin sections followed by heavy metal staining, the cell wall appears as a triple layer: a diffusely staining middle layer sandwiched between two electron-dense lines (Fig. 1). The electromicrograph in Fig. 1 also shows the effect of inhibited cell wall turnover (Fig. 1B) resulting in the accumulation of large amounts of unstructured cell wall material on the cell surface of a vancomycin-resistant clinical isolate of S. aureus (6). The photograph of the isogenic vancomycin-susceptible parental cell recovered from the same clinical source is shown on the left side of the figure. Cell wall thickening is often the phenotype of S. aureus with vancomycin-intermediate S. aureus (VISA)-type vancomycin resistance (7).

FIGURE 1.

The anatomy of cell walls in normal (A) and vancomycin-resistant (B) S. aureus. Reproduced with permission from reference 32.

Earlier studies established basic compositional/structural features of the S. aureus cell wall. It is composed of a highly cross-linked A3α-type peptidoglycan with pentaglycine oligopeptide units connecting the ε-amino group of the lysine component of one muropeptide to the penultimate d-alanine of another. This peptidoglycan, together with ribitol-type teichoic acid chains [which are attached to the 6-hydroxyl groups of some of the N-acetylmuramic acid residues of the glycan chain through linkage units, (glycerol phosphate)3-N-acetylmannosaminyl beta(1→4)N-acetylglucosamine] (8, 9), surrounds the S. aureus cell in the form of a multilayered envelope (Fig. 2). The peptidoglycan of S. aureus is typically 20 to 30 nm thick and is characterized by short glycan strands (5 to 25 disaccharide units) (10) and a high degree of cross-linking (∼90%) (11), with striking specific structural characteristics that contrast not only with the peptidoglycan of Gram-negative bacteria but also with that of other Gram-positive model organisms (Fig. 3) (12). This review article only presents information on the peptidoglycan component and does not address teichoic acids.

FIGURE 2.

Three-dimensional structure of staphylococcal peptidoglycan. Straight lines of large globes represent sugar moieties of the peptidoglycan. Each globe in these lines symbolizes an amino sugar, N-acetylglucosamine (black globe), or N-acetylmuramic acid (white globe). Stem peptides, branching from N-acetylmuramic acid, are characterized by small dark globes with a white center. The connecting interpeptide bridges (pentaglycines) between the stempeptides are shown as small black globes. Schematic drawing by Peter Giesbrecht, Thomas Kersten, Heiner Maidhof, and Jorg Wecke; Robert-Koch Institute, Berlin, Germany. Reproduced with permission.

FIGURE 3.

Variation of peptidoglycan global structural parameters between S. aureus (blue), B. subtilis (red), and E. coli (black) (164). (A) Chain length, (B) peptidoglycan thickness, (C) percentage of peptidoglycan cross-linking, and (D) relaxation time-constants, with shorter values indicating a higher peptidoglycan flexibility. Hence, the peptidoglycan of S. aureus is the most rigid of the three bacterial species. Reproduced with permission from reference 12.

HIGH-RESOLUTION ANALYSIS OF THE S. AUREUS PEPTIDOGLYCAN

Progress in the high-resolution chemistry of the S. aureus cell wall came from the introduction of HPLC and mass spectrometric methods (13) for the analysis of the primary structure of peptidoglycan. First, studies with gel permeation HPLC established the presence of muropeptide oligomers in lengths extending to nanometers and beyond (14). This was followed by the adaptation of the reverse-phase HPLC system in combination with mass spectrometry for the analysis of strains of S. aureus (11, 15). Analysis of the peptidoglycan of methicillin-resistant S. aureus (MRSA) provided the first high-resolution view of the complexity of cell wall structures in Gram-positive bacteria.

Enzymatic hydrolysis (with the M1 muramidase) of the staphylococcal peptidoglycan was followed by reduction and separation of the muropeptide components on a reverse-phase HPLC column. This method resolved the hydrolysate to more than 21 distinct UV-absorbing peaks plus a “hump” of unresolved material of longer retention times which made up more than 25% of the muropeptides (Fig. 4A).

FIGURE 4.

(A) Analysis of the peptidoglycan of methicillin-susceptible S. aureus strain 27s and its isogenic derivatives ZOX3 (PBP2 TPase point mutant), 27sΔpbpD (PBP4 deletion mutant), and ZOX3 ΔpbpD double mutant. Numbers above the HPLC peaks identify the structure of the respective muropeptides. Peaks 5, 11, 15, 16, and 17 represent the monomeric muropeptide and its di-, tri-, tetra-, and pentameric derivatives, respectively. The poorly resolved part (“hump”) of the HPLC profile eluting after peak 18 contains highly cross-linked oligomeric components. (B) Model for the cooperative functioning of PBP2 and PBP4 for the cross-linking of the S. aureus peptidoglycan. Muropeptides are identified by the respective peak numbers in the above HPLC chromatograms. Monomers are indicated by solid circles, and the bead diagram symbolizes the number of monomeric units cross-linked. The primary TPase activity of PBP2 is suggested to involve the cross-linking of muropeptides of lower oligomerization degree (up to pentamers), and PBP4 is suggested to catalyze the cross-linking of higher multimers. Reproduced with permission from reference 55.

Analysis of muropeptide components obtained after digestion with muramidase, or with the combination of muramidase plus lysostaphin (a bacteriolytic endopeptidase that attacks the pentaglycine bridge), revealed that the major monomeric building block is a disaccharide pentapeptide carrying d-isoglutamine in position 2, an intact d-alanyl-d-alanine carboxyterminal, most frequently with a pentaglycine substituent attached to the ε-amino group of the lysine residue (muropeptide 5) or, occasionally, without it (muropeptide 1). These monomeric units make up about 6% of all muropeptides. Another 20% consist of dimers in which the pentaglycine substituent of one muropeptide unit is cross-linked to the penultimate d-alanine of a neighboring one (muropeptide 11). About 40% consist of higher oligomers containing 3 (muropeptide 15) to 9 muropeptide units generated by the same cross-linking principle. Still higher oligomers account for an additional 15 to 25% of the muropeptide units (the hump of unresolved components eluting with retention times longer than 110 min.).

A variation in the preparative technique also allows determination of the structure of the glycan chains. In this case, peptidoglycan is first hydrolyzed with lysostaphin to disrupt the pentaglycine bridges connecting neighboring muropeptides, followed by removal of stempeptides by the pneumococcal amidase (N-acetylmuramyl-l-alanine amidase), which hydrolyzes the covalent bond connecting the l-alanine residue to the acetyl muramic acid residue of the glycan chain. After removal of the stempeptides, the size distribution and composition of the glycan chains is determined by reverse-phase HPLC. Application of this method has resolved the glycan strands of S. aureus to a family of major peaks which represent oligosaccharides composed of repeating dissacharide units (N-acetylglucosamine beta-1,4-N-acetylmuramic acid) with different degrees of polymerization and terminating with N-acetylmuramic acid residues at the reducing ends. The method allowed separation of strands up to 23 to 26 disaccharide units. Minor satellite peaks were also present throughout the HPLC elution profile, most likely representing products of N-acetylglucosaminidase activity (10).

Stability of Muropeptide Composition

Analyses of peptidoglycan prepared from a large number of S. aureus isolates of different clonal types, either susceptible or resistant to methicillin and grown without antibiotic challenge, all showed virtually identical HPLC muropeptide patterns (A. Ornelas-Soares, H. de Lencastre, and A. Tomasz, unpublished observations). Selective Tn551 inactivation of the mecA gene did not cause any detectable change in muropeptide composition (15). These data suggest that similar to other bacteria, the muropeptide composition of S. aureus cell walls is specific for the species.

Ultrastructural Analysis of the S. aureus Peptidoglycan

Knowledge of the primary structure of peptidoglycan and other cell wall components was more recently complemented by structural data provided by high-resolution approaches. Cryo-electron microscopy has allowed a change of the previous simple concept of a thick and undifferentiated cell wall. The substitution of chemical approaches of fixation by rapid freezing as a physical method with simultaneous chemical fixation allowed observation of the cell envelope organization, namely, the identification of a low-density inner wall zone limited by the membrane and a high-density outer wall zone. The low-density region, proposed to be a periplasmic space, lacks cross-linked polymeric structures and harbors soluble components, while the high-density region consists of an intricate net of peptidoglycan and teichoic acids. Furthermore, the cell wall structure is differentiated at the septum, with two high-density zones corresponding to nascent cross walls separated by a wide low-density mid-zone, suggested to result from the activity of autolysins, separating daughter cells during cell division (16, 17).

Another step forward on the definition of the structure of the peptidoglycan of S. aureus was possible due to solid-state nuclear magnetic resonance (NMR) spectroscopy of fully hydrated cell wall material. First, this method showed that the peptidoglycan structure is not altered by the removal of teichoic acids, validating the established HPLC-based peptidoglycan analysis (12). Also, it corroborated the longtime notion that glycan strands are more rigid than peptide stems and that the peptidoglycan of S. aureus is more rigid than that of E. coli or even Bacillus subtilis. This decrease in dynamics is directly related to the higher degree of cross-linking of S. aureus peptidoglycan, which compensates for the shorter glycan strands and the presence of interpeptide bridges (12), that should enhance peptidoglycan flexibility. Furthermore, rotational-echo double-resonance NMR, allowed measurement of bridge-glycan distances and determined that the pentaglycyl bridge is in close proximity (5 Å) of the neighboring glycan chain, suggesting a three-dimensional model in which the peptide stems are in a plane perpendicular to the glycan chain and parallel to each other. The study was widened to include fem cell wall mutants affected in the glycine interpeptide bridge and VISA strain Mu50 (18). More recently, the same approach was used to study the effects of glycopeptides on the biosynthesis of peptidoglycan and teichoic acids (19).

Fluorescence microscopy and, more recently, superresolution microscopy have driven important breakthroughs on cell morphology and division patterns. In fact, S. aureus cells were revealed not to be spherical throughout the cell cycle and to include cell elongation steps that depend on peptidoglycan synthesis and remodeling, allowing for new models of S. aureus cell division to be proposed (20–23).

VARIATION IN PEPTIDOGLYCAN COMPOSITION

Effect of Antibiotics

Growth of the highly methicillin-resistant S. aureus strain COL in a wide range of subinhibitory concentrations of methicillin caused striking changes in muropeptide composition: the representation of trimeric plus higher oligomeric components was drastically reduced (from a combined representation of close to 50% to less than 10%), while the proportion of monomeric and dimeric components increased (from about 15% to about 50%). It was suggested that the abruptness of this change in peptidoglycan composition may represent the switching over from the normal wall biosynthetic system (the four native PBPs) to another one (PBP2A, encoded by the methicillin resistance gene mecA) capable of functioning in the presence of high concentrations of methicillin (11). The anomalous composition of the peptidoglycan produced under these conditions would then reflect the limited capacity of PBP2A for cross-linking more than single monomeric muropeptides. Nevertheless, similar albeit much less abrupt compositional changes were also observed when methicillin-susceptible staphylococci or an isogenic derivative of strain COL (with inactivated mecA) were grown in sub-MIC concentrations of methicillin.

Effect of Vancomycin and Teicoplanin Resistance

Laboratory mutants with increased MIC values for glycopeptide antibiotics have been isolated (24–26). The cell wall composition was determined in a series of isogenic laboratory mutants of the highly methicillin-resistant S. aureus strain COL (27). The peptidoglycan of the mutants showed distorted muropeptide composition which paralleled the increasing vancomycin MIC values, and it was suggested that some of the unusual properties of these mutants and the mechanism of resistance are related to the enrichment of the cell walls in muropeptide monomers. Muropeptides terminating in the d-alanyl-d-alanine residues are known to form the binding sites for glycopeptide antibiotics. The mechanism of resistance may then be related to trapping the glycopeptides in the mature layer of the peptidoglycan enriched for muropeptide monomers, thus preventing the antibiotic molecules from reaching sites of wall biosynthesis at the plasma membrane (26, 28). Another structural change, increased glycan chain length, was also detected in a vancomycin-resistant mutant (29). Recently, the effects of glycopeptides were analyzed by rotational-echo double-resonance NMR (19). Vancomycin subinhibitory concentrations resulted in a strong reduction of d-alanine incorporation into wall teichoic acids but no changes in the peptidoglycan stem composition, suggesting that, upon glycopeptide challenge, the cell drives the available d-alanine to peptidoglycan biosynthesis at the expense of the wall teichoic acids.

VISA Isolates

Clinical isolates of S. aureus with reduced susceptibility to vancomycin have been repeatedly described in the literature, and in at least some of these so-called VISA isolates, alterations in peptidoglycan composition have been observed, including increased monomeric and decreased oligomeric components, excess of cell wall material, aberrant separation of daughter cells, and altered autolysis (6, 30–32). The genomic analysis of clinical VISA isolates, together with laboratory mutants, has contributed to the understanding of the genetic and biochemical mechanisms of intermediate-level vancomycin resistance in S. aureus. Isogenic groups of susceptible and resistant isolates recovered from patients with a history of vancomycin therapy or laboratory strains submitted to vancomycin selection have been studied using high-throughput whole-genome sequencing to identify the altered genes (33–36). The mutated genes were commonly identified in VISA isolates, with the most frequent mutations being in the vraRS (37), graRS (38, 39), and walKR (35, 40) two-component regulatory systems (detailed below). The most relevant mutated genes seem to be somehow related to cell wall metabolism and autolysis (35, 37, 38), but mutations in the RNA polymerase gene rpoB were also described (41). Nevertheless, the genetic alterations identified in the in vivo VISA isolates were not the same as the ones found in laboratory-derived VISA strains (42). For comprehensive reviews, see references 43 and 44.

Vancomycin-Resistant S. aureus (VRSA)

In 2002, highly vancomycin-resistant strains of S. aureus were recovered from clinical specimens. These bacteria acquired the vancomycin resistance gene complex through the enterococcal transposon Tn1546. Expression of Tn1546 produced drastic changes in the cell wall composition: all pentapeptides were replaced by tetrapeptides, and the peptidoglycan contained at least 22 novel muropeptide species that frequently showed a deficit or complete absence of pentaglycine branches. The UDP-Mur-NAc pentapeptide, the major component of the cell wall precursor pool in vancomycin-sensitive cells, was replaced by the UPD-MurNAc depsipeptide in the resistant bacteria, and PBP2A, the product of the methicillin resistance gene mecA, was unable to utilize such precursors for wall synthesis. The MRSA strain COL carrying mecA and the vanA genes had extremely high MICs both for β-lactam antibiotics and for vancomycin (45). The major PBP responsible for the assembly of the abnormally structured peptidoglycan produced in the vancomycin-resistant cells was PBP2 (46).

Since 2002, the number of VRSA isolates in the United States has increased to 14 (47). Thirteen of these Tn1546-containing S. aureus isolates belong to the clonal complex 5 (CC5), an MRSA lineage that has been repeatedly isolated from clinical specimens (43, 44, 47), in contrast to VISA strains that have been associated with different genetic backgrounds, including CC5, CC8, CC30, and CC45 (48, 49). The fact that VRSA strains have not become more frequent is probably related to the specificities of the clinical scenario in which they emerge. Most VRSA isolates were retrieved from diabetic wounds of patients that carried both vancomycin-resistant enterococci and VRSA (45).

GENETIC DETERMINANTS AND ENZYMES IN CELL WALL SYNTHESIS

Penicillin Binding Proteins

Tentative roles for the four native staphylococcal PBPs in wall synthesis were originally assigned on the basis of morphological/biochemical effects of β-lactams that showed more or less selective binding to individual PBPs (50). More recent studies used genetic techniques and in vitro transpeptidation assays with lipid II and labeled substrates to clarify the functions of PBPs; the minimal protein composition of peptidoglycan synthesis machinery was addressed by generating a mutant strain in which seven of the nine genes encoding peptidoglycan synthesis enzymes (the five PBPs, four native and one acquired; the two MTGases, MGT and SgtA; and FmtA and FmtB, which have homology to transpeptidase [TPase] domains) were deleted. S. aureus was found to be able to grow and divide in vitro with only the PBP2 and PBP1 enzymes, without morphological changes, suggesting that the other enzymes have redundant functions. Transcriptomic analysis showed that, as could be expected, the vraSRF operon is upregulated and several murein hydrolases are inhibited in this mutant. However, cells with this minimal set of peptidoglycan synthetic enzymes were impaired in pathogenesis and antibiotic resistance, indicating that a complete synthetic pathway is important in natural conditions (51).

Allelic replacement approaches had already established the essential nature of PBP1 (52), and later the contribution of this PBP (a high-molecular-weight monofunctional transpeptidase) to cell division and separation was explored (53, 54). PBP1 localizes to the division septum and does not depend on other peptidoglycan synthetic enzymes for correct localization (51). A PBP1 depletion mutant rapidly loses viability but maintains a normal peptidoglycan composition (53). However, analysis of a PBP1 mutant with impaired TPase activity suggested that PBP1 plays a dual essential role, because its presence, independent of the TPase activity, is required for the formation of septa, while the TPase activity—being necessary for the mechanism of cell separation (54)—can act in a timely fashion at the end of the septal growth.

Of the four native PBPs, PBP2 is the only one composed of a TPase and a transglycosylase (TGase) domain, the former being essential for growth of the bacteria. Biochemical evidence for TPase activity was obtained through the analysis of the peptidoglycan muropeptide profile of a spontaneous ceftizoxime-resistant mutant with a point mutation in the TPase domain (55). The decreased level of highly cross-linked oligomers in the mutant suggests that the primary TPase function of PBP2 is in producing muropeptide dimers (Fig. 4A, peak 11), trimers, tetramers, and pentamers (peaks 15 through 17).

Studies of conditional mutants of pbpB demonstrated that this essential function of PBP2 may be replaced by PBP2A, the product of the acquired drug resistance gene mecA (56). The TGase domain of PBP2 was shown to cooperate functionally with PBP2A when the bacteria were growing in the presence of antibiotics (57). Critically located point mutations in pbpB were associated with methicillin resistance both in laboratory mutants and among some clinical isolates that did not carry the mecA gene but showed low-level β-lactam resistance (58). Recently, the substrate preferences of PBP2A, PBP2, and PBP4 were compared regarding the completeness of the glycine bridge. PBP2 was found to accept penta-, as well as tri- and monoglycine-associated muropeptides as substrates of transpeptidase activity, while PBP4 exhibits different enzymatic activities according to each of the three substrates, and PBP2A does not accept monoglycine muropeptides (59).

PBP3 is a class B high-molecular-weight protein with a C-terminal penicillin binding domain showing the conserved motifs characteristic of transpeptidases and an N-terminal nonbinding domain of unknown function. Unlike in the cases of pbpA and pbpB, inactivation of pbpC in either methicillin-susceptible or -resistant backgrounds allowed growth of the bacteria, and there was no change in the muropeptide composition. The only observable alteration of such mutants was a reduction in autolysis rates. However, when mutants were grown in the presence of sub-MIC concentrations of methicillin, the bacteria frequently showed aberrant morphologies (60). The possible function for this PBP in cell wall synthesis and cell division is not yet described.

PBP4 is a nonessential low-molecular-weight PBP. Bacteria with inactivated PBP4 grew normally but produced a peptidoglycan in which the highly oligomeric components were substantially reduced. An extensive deletion in the promoter region was identified in β-lactam-resistant laboratory isolates which showed increased peptidoglycan crosslinking (61). Inactivation of pbpD was demonstrated in a highly vancomycin-resistant lab mutant which produced peptidoglycan with greatly reduced cross-linking (28). These findings confirm earlier proposals that PBP4 can act as a secondary transpeptidase. It was recently demonstrated that PBP4 not only acts preferentially as a transpeptidase, accepting both nascent peptidoglycan and lipid II as substrates but also is the main agent responsible for the incorporation of exogeneous d-amino acids into peptidoglycan. It was proposed that this PBP may be responsible for adding cross-links to partially cross-linked peptidoglycan, primarily to pentamers and higher oligomeric muropeptides (55) and to repair peptidoglycan damage (62). Furthermore, regarding the cross-linking preferences, PBP4 is much more selective than PBP2 and shows a wider range of activities; PBP4 acts exclusively as a transpeptidase for pentaglycine muropeptides and shows 90% of hydrolytic activity with triglycine muropeptides and carboxypeptidase activity for monoglycyl muropeptides (59). Initially, PBP4 was thought not to be required for β-lactam resistance as tested in hospital-acquired MRSA strains. However, in the background of community-acquired strains it was deemed to be involved in the optimal expression of resistance (63). Using atomic-force microscopy to determine the elasticity of the cell wall, the analysis of PBP4 mutants showed that S. aureus cells suffer an overall enhancement in elasticity upon decrease of the cross-linking level and that this effect was stronger in the community-acquired MRSA background (64).

PBP2A—the protein product of the resistance gene mecA–was crystallized, allowing the identification of structural features responsible for the extremely low affinity of this protein for most β-lactam antibiotics (65). The basis of the PBP2A preference for a peptidoglycan substrate is an allosteric site (distal by 60 Å from the active site) that regulates the conformation of a structural loop guarding the active site. When the allosteric site is bound to a neighboring peptidoglycan strand, the active site becomes available to its ligand, peptidoglycan. Furthermore, the catalytic serine residue changes conformation to maintain the substrate preference toward the β-lactam molecule (66–68). Such structural insights are crucial for mechanistic studies aimed to understand the mode of action of the different β-lactam molecules and the emerging PBP2A mutations. Regarding the transpeptidase activity, it was earlier proposed that PBP2A would only cross-link stem peptides with a pentaglycine branch; however, PBP2A was recently found to also accept as a substrate triglycine-containing strands but not monoglycine ones (59).

Methicillin-Conditional Mutants

The central genetic determinant of MRSA strains is the heterologous gene mecA. Several lines of evidence suggest that mecA evolved from a housekeeping gene of Staphylococcus sciuri (69–72). The evolution toward a resistance determinant, promoted by antibiotic use in humans and animals, involved alterations in the promoter region, structural changes in one protein domain, and the adaptation of the genetic background of bacteria (73).

Inactivation of mecA results in complete loss of resistance. Nevertheless, a reduction in the resistance level from a modest decrease to a virtually complete loss of resistance can also be the consequence of inactivation of a surprisingly large number of domestic genes, over 30, which have been referred to as fem or auxiliary genes and are not directly involved in the expression of the mecA gene (74–77). The transposon mutants of methicillin-resistant S. aureus strain COL, which led to the identification of auxiliary genes, were originally isolated to clarify some complex genetic features of staphylococcal methicillin resistance (75, 78). Biochemical analysis indicated that in several of these mutants the inactivated genes were involved with staphylococcal cell wall synthesis. This finding suggested that a reduction in the methicillin resistance level may provide a relatively easily selectable phenotype for the identification of genes in cell wall synthesis and metabolism (75). As a working model, it was proposed that the methicillin MIC value in such “methicillin-conditional” mutants may primarily reflect the success of the mecA gene product (PBP2A), positioned close to the end of a long and complex metabolic pathway, to continue catalyzing peptidoglycan incorporation even when this metabolic pathway is disrupted by an inhibited auxiliary gene.

Tn551 MRSA mutants selected for reduced MIC (such as femXA/B, gluM, murE, llm, fmtA to C, etc.) or specifically constructed mutants (murF, murT-gatD, etc.) have led to insights into hitherto unknown or not fully characterized steps in staphylococcal cell wall precursor biosynthesis (74–76, 79–85). More recently, the list of auxiliary genes was widened by the large-scale identification, using antisense interference conditional mutants, of essential genes required for β-lactam optimal resistance (86). This study allowed the identification of essential cell division (ftsZ and ftsA) and teichoic acid biosynthesis genes (tarL), among others.

The Role of the Stringent Response in the β-Lactam Resistance Mechanism

Recently, the above-described model has been challenged by a sequence of studies that culminated in the identification of the stringent response as a key player in a mecA-dependent mechanism of homogeneous and high-level β-lactam resistance (87–89). The stringent response is a bacterial reaction to a nutrient shortage, with the aim to economize metabolic resources. Such a response involves the transcriptional regulation by RNA polymerase and the modulation of the activity of certain proteins and RNAs through the cellular level of guanosine pentaphosphate (p)ppGpp (90). A comparative study based on whole-genome sequencing of highly resistant subpopulations of early MRSA strains with different degrees of oxacillin resistance identified mutations in 27 genes of diverse functional categories. These mutations were shown to induce high-level β-lactam resistance by triggering the stringent stress response, causing increased production of PBP2A (91). Nevertheless, the mechanism of high-level resistance induced by the stringent response corroborates the previously established model of β-lactam resistance, which considers the essentiality of the transglycosylase activity of PBP2 (87). Moreover, auxiliary mutants were shown to regain the homogeneous resistance level of the parental strain COL when challenged with mupirocin, an isoleucyl-tRNA homologue that induces stringent stress (92).

FemAX/B

The femX (Fmhb), femA, and femB genes are part of an operon involved with the synthesis and attachment of the pentaglycine branches to the epsilon amino group of the lysine residue of the S. aureus muropeptides. The synthesis of pentaglycine bridges occurs in three sequential steps catalyzed by the FemABX nonribosomal peptidyl transferases, which cannot substitute for one another. The protein product of Fmhb catalyzes the addition of the first glycine residue to the muropeptides. This gene is essential for bacterial survival and growth (93). The protein product of femA is involved with the addition of the first and second, and the protein product of femB, with the addition of the third and fourth, glycine residues to the cross-bridge (83, 94). In an elegant series of studies, the entire sequence of pentaglycine branch synthesis and attachment has been determined and reproduced in vitro using bactoprenol-linked disaccharide pentapeptide as the acceptor and t-RNA-linked glycine as the source of the amino acid residues (95, 96). Structural NMR studies of femAB-deletion mutants with shortened pentaglycine interpeptides showed that the pentaglycine bridge is required to reduce the average distance between the nearest-neighbor bridge and glycan and that a poorly cross-linked cell wall may not be able to resist the cell’s internal turgor. To counteract the internal osmotic stress in the absence of femAB genes, transcriptomic studies suggested that the cell resorts to metabolic alterations, namely, of the arginine-deiminase pathway and nitrogen metabolism (18, 97). Nevertheless, viability in the absence of femA and femB suggests that one or more of the native PBPs can cross-link stem peptides with monoglycyl or triglycyl bridges.

Secondary Modifications

Secondary modifications of the glycan strands of the staphylococcal peptidoglycan, namely, O-acetylation and N-deacetylation of N-acetylglucosamine, have been mostly associated with evasion of the human host. O-acetylation occurs at the C-6 position of muramic acid residues and is catalyzed by the integral membrane O-acetyltransferase OatA; this modification confers resistance to lysozyme, a muramidase present in most human biological fluids and secreted by cells of the immune system (98–100). The proposed basis of resistance is that the O-acetyl group sterically prevents the binding of lysozyme to peptidoglycan (101), and only pathogenic species of Staphylococcus have O-acetylated peptidoglycan (98). The stem peptide of S. aureus also suffers modifications to the primary structure, including the addition of the pentaglycine bridge (see previous section) and amidation of the second residue d-glutamate to d-iso-glutamine. The genetic determinants of the amidation reaction, the murT and gatD genes, were described in 2012, and the protein products form a bi-enzymatic complex, with MurT being a member of the mur ligase family and GatD a glutamine amino transferase that uses glutamine as the source of the amino group (79, 102). As for O-acetylation, peptidoglycan amidation is involved in lysozyme resistance, but in contrast, it is not restricted to pathogenic species but, rather, is widely distributed among Gram-positive bacteria. The MurT-GatD catalyzed reaction is essential for viability and for resistance to β-lactams, showing different levels of involvement in the resistance mechanism according to the genetic background of the strain (79, 102, 103).

The pentaglycine bridge is a modification of S. aureus peptidoglycan and the specific target of lysostaphin, a bacteriolytic enzyme secreted by Staphylococcus simulans. This species is resistant to the activity of lysostaphin since it contains serine/glycine cross-bridges, a modification catalyzed by Epr (lysostaphin endopeptidase resistance) of the FemAB protein family (104).

Insights into Regulation and Contribution of Transcriptomic Studies

Upregulation of the transcription of pbpB, the structural gene of PBP2, by exposure of S. aureus cells to oxacillin, vancomycin, and other cell wall inhibitors has been reported (105), suggesting transcriptional regulatory pathways linking the cell wall structure and metabolic functions. This first observation was corroborated and widened by a large number of transcriptomic studies using state of the art DNA microarrays to study cell wall exogenous stress conditions (37, 106–110) or cell wall mutations (37, 97, 107, 111–114). Utaida et al. were the first to describe a large set of genes, the so-called cell wall stimulon (CWS), the expression of which was activated in response to cell wall-targeting antibiotics of diverse classes, suggesting the sensing of cell wall damage (110, 115). The CWS included cell wall-associated genes such as pbpB, murZ, and sgtB, encoding a TGase, and fmtA, encoding an enzyme with TPase activity. The vancomycin resistance-associated (VraSR) two-component system, identified for being activated by vancomycin challenge, was found to induce all members of the CWS (37, 116). Upregulation of the CWS is suggested to provide resistance to VraSR-inducing agents, including β-lactams.

S. aureus contains 16 sets of 2-component regulatory systems that monitor cell wall damage and activate an appropriate physiological response; among these is not only VraSR, but also GraRS and WalKR, which interestingly were all found to be implicated in the development of VISA strains (40, 117, 118).

The VraSR system autoactivates its own expression and most CWS genes, which are suggested to contribute to the repair of the sensed cell wall damage (113, 119). VraS is an intramembrane sensing kinase that undergoes autophosphorylation upon challenge with a cell wall-targeting antibiotic and subsequently phosphorylates the VraR response regulator (120), which binds to its own promoter and activates or represses the genes of the regulon (121). The molecular signal detected is not yet defined but is suggested to be a by-product of cell wall damage since a diverse class of unrelated antibiotics was shown to induce the system. The Vra system effects methicillin resistance by a still unclear mechanism, independently of mecA gene (116).

The glycopeptide resistance-associated (GraSR) system confers resistance to cationic antimicrobial peptides by increasing the positive charge of the cell surface, contributing to electrostatic repulsion through the modulation of the d-alanylation of teichoic acids and of lysinylation of phosphatidyl glycerol (122, 123). A transcriptomic study showed that the GraSR system is also involved in pathogenesis, stress response, and cell wall signal transduction, and its regulon overlaps the WalKR regulon (112). This system is proposed to be a five-component system composed of GraS, an intramembrane sensing kinase, GraR, the response regulator, GraX, a regulatory cofactor, and the VraFG ABC transporter that senses the presence of cationic antimicrobial peptides and is involved in signaling transduction (112, 124).

The two-component system, WalKR, is essential for S. aureus viability and is primarily involved in the regulation of cell wall metabolism by positively modulating autolytic activity through the control of the expression of most of the cell wall hydrolase genes, including the direct binding of WalR to the atl promoter (125). LytM autolysin seems to play a particularly important role in this mechanism by reducing the degree of cross-linking in the peptidoglycan through its endopeptidase activity toward the pentaglycine bridge (126). Mutants for this system showed enhanced biofilm formation, decreased peptidoglycan biosynthesis and autolytic activity, thickened cell walls, and reduced vancomycin susceptibility (125, 126). A role in virulence was also described for this system, because WalKR was found to trigger the host inflammatory response through the release of peptidoglycan fragments as a consequence of increased autolysis (111). A mutation in WalK sensor kinase that was also found in a clinical VISA strain was responsible for deficient autophosphorylation and concomitant reduced phosphorylation of WalR, which binds less efficiently to the atl promoter, impairing the control of the autolytic system (118).

Two monofunctional glycosyl transferases were identified and characterized in S. aureus (127, 128), and genomic analysis suggested that besides PBP2, only MGT and SgtA contain TGase motifs. MGT enzyme was shown in vitro to catalyze glycan chain polymerization, using lipid II as a substrate, with an efficiency of 5,800 M⁻¹s⁻¹ (129). All three enzymes have the capacity to elongate glycan chains, and although the monofunctional glycosyl transferases are not essential, S. aureus cannot survive without the TGase activity of PBP2 or MGT, indicating that the activities/specificities of MGT and SgtA are different. Both MGT and SgtA were found to interact with PBP2, suggesting the existence of a cell-wall synthetic complex in S. aureus (127). Expression of the mgt gene was found to be upregulated in the presence of cell wall antibiotics (110). However, although MGT can compensate for the lack of PBP2 TGase activity in methicillin-susceptible S. aureus and MRSA strains, it does not replace PBP2 upon challenge with high concentrations of oxacillin (127).

Although never fully described, several pieces of evidence support the existence of a coordinated multienzyme complex responsible for cell wall assembly. One piece of evidence was provided by the analysis of the peptidoglycan of a PBP2 TPase mutant and of its corresponding double PBP4 deletion mutant (Fig. 4A). This study allowed the description of the cooperative functioning of PBP2 and PBP4 in which the TPase activity of PBP2 is suggested to produce the substrates for PBP4, dimers, trimers, and tetrameric muropeptides, which are subsequently transformed into higher oligomers by PBP4 activity (Fig. 4B). In the same study, PBP2A was proposed to directly provide “assistance” to this PBP2-PBP4 transpeptidation system through a still undescribed mechanism (55).

COMPLEX FUNCTIONS OF CELL WALLS

The peptidoglycan is a protective barrier against the host and also a scaffold for the attachment of surface proteins and polysaccharides, which are essential for cell division and pathogenesis. The carboxyterminal end of staphylococcal surface proteins (such as protein A) are covalently bound to the oligoglycine cross-bridge of peptidoglycan muropeptide units through a threonine residue in the protein (generated after cleaving of the wall sorting signal LPXTG), which is then linked to the amino terminal glycine residue of the muropeptide. Sortase-anchored surface proteins were shown to be released into the external environment together with peptidoglycan immunostimulatory fragments linked to their C-terminus by the action of murein hydrolases (130). The structure of these cross-bridges is under the control of the fem gene complex. Therefore, the already pleiomorphic femA/B mutant (showing defective wall structure, reduced methicillin resistance, and slow wall turnover) may also be affected in virulence.

The active involvement of the cell wall in complex functions is also being recognized in studies of the transport (release and uptake) of large molecules (131). Increased susceptibility of femA/B null mutants to some antibacterial agents other than β-lactams may be related to increased porosity of the peptidoglycan. Increased extractability of proteins (wall porosity?) was noted in a highly teicoplanin-resistant staphylococcal mutant with greatly decreased peptidoglycan cross-linking (132).

Murein Hydrolases in Cell Division and Antibiotic Resistance

Peptidoglycan is a dynamic structure, being constantly synthesized and hydrolyzed. The cell biological functions of staphylococcal murein hydrolases and their regulation have been intensively investigated, and these studies benefited greatly from zymography for profiling bacteriolytic enzymes and from genetic work. The major autolysin gene atl was found to encode a bifunctional protein that undergoes proteolytic processing at the cell surface into the independent N-acetylmuramyl-l-alanine amidase and endo-β-N-acetyl glucosaminidase domains (133, 134). The still “double-headed” atl gene product containing the fused amidase and glucosaminidase proteins was found to be localized on the plasma membrane as a circumferential ring at the future cell divisional site, and this localization precedes the appearance of centripetally growing cell wall (135) (Fig. 5). The molecular signature that directs the autolysin gene to its site of action at the cellular equator was shown to reside within the three repeat elements present in each domain (136, 137). The presence of wall teichoic acids at the septum was described to prevent the binding of Atl domains to the septal region through an avoidance strategy (138). Atl is primarily involved in the separation of daughter cells during cell division but also contributes to cell wall turnover and antibiotic-induced lysis (139, 140).

FIGURE 5.

Localization of atl gene products on the cell surface of S. aureus during the division cycle as determined by scanning electron microscopy. (a to d) The gold labeling patterns on cells at different stages of the cell cycle. Bar, 100 nm. Reproduced with permission from reference 135.

In zymograph gels, a Δatl mutant showed a single lytic band that was determined to correspond to the amidase Sle1 (141), also involved in cell separation. Besides the atl, sle1, and lytM genes, S. aureus is thought to contain 10 additional genes encoding peptidoglycan hydrolases, although their products are not biochemically characterized.

Alteration of autolytic properties has been consistently seen in staphylococcal mutants isolated for changed β-lactam resistance. Increases in autolytic rates (and cell wall turnover) in parallel with the increase in resistance level were demonstrated in cefotaxime-resistant mutants (142), while reduced rates were observed in femA, femB, and femC mutants. Also, the downregulation of PBP2 and PBP1 expression resulted in an associated reduction in the transcription of atl and sle1, indicating a coupling of cell wall synthesis to the autolytic enzymes at the transcriptional level (54, 143).

Cell Division

Recent methods allowed the identification of functional links between the peptidoglycan structure and synthesis pathway and the cell division process. Fluorescence microscopy localization studies demonstrated the essential role of FtsZ and of lipid II peptidoglycan precursor in the correct localization of PBP2 to the septum (144, 145). On the other hand, wall teichoic acids were shown to indirectly control PBP4 localization to the septum (146).

The process of cell division in S. aureus occurs equatorially and on consecutive orthogonal planes in three dimensions in accordance with the previous suggestion that peptidoglycan insertion only occurs at the septum (145, 147). It was possible to correlate the architecture and dynamics of peptidoglycan with the cell cycle using atomic force microscopy and fluorescence microscopy by observing a thick band of peptidoglycan, the so-called piecrust, at the division plane before the formation of the septum. After division, these structures remained as orthogonal ribs, marking the location of the previous division event, possibly to enable the cell to maintain the division plane localization sequentially orthogonal over the generations. Upon cell expansion after separation, the peptidoglycan mesh must undergo rearrangements since the area of the flat septal disc is much less than the area of half of the new cell (Fig. 6A) (21, 148). Some peptidoglycan hydrolases (Atl, SagA, ScaH, and SagB) with glucosaminidase activity have recently been shown to be required for cell enlargement, presumably by altering the mechanical properties of peptidoglycan that becomes more flexible for stretching (23).

FIGURE 6.

(A) AFM height (H) and phase (P) images of purified sacculi from exponential phase S. aureus and key architectural details of the piecrust, division rings, and knobbles (21). (Top panel) Sacculus showing apparent incomplete septum parallel to the plane of the image, encircled by a piecrust feature associated with the initiation of septum formation; inset shows detail of piecrust features. (Middle panel) Sacculus showing ring texture associated with nascent peptidoglycan and knobble texture associated with older peptidoglycan; insets show details of rings and knobbles. (Bottom panel) Interpretive diagram drawn from the yellow rectangle in the middle panel. Reproduced with permission from reference 148. (B) SEM images of S. aureus strain COL showing an asymmetric scar (blue line) corresponding to the previous division site and a fissure located in the middle of the cell (red line), presumably corresponding to the next division site. The new cell wall (light brown), resulting from septal material from the mother cell, which has a smooth surface immediately after division (first panel), occupies less than half of the total surface. Scale bars, 600 nm. (C) Comparison of two models for S. aureus growth and division. (a) An earlier model which assumed that S. aureus cells remained approximately spherical over the cell cycle and that, on division, the cell wall material from the septum of the mother cell increased in cell surface area to constitute half (one hemisphere made of new cell wall) of the cell surface of each daughter cell. As a consequence, “scars” of previous divisions were proposed to encode epigenetic information that could be used to determine orthogonal placement of division septum. (b) In the other model proposed, S. aureus cells are approximately spherical at the beginning of the cell cycle and elongate as the cell cycle progresses. On division, there is no increase in the surface area of the previous septum, which becomes ∼33% of the surface area of each daughter cell. This asymmetry in the regions composed of new and old cell wall results in scars of previous divisions that do not divide cells in quadrants. Consequently, T-junctions of scars of two previous divisions are not located at the cell poles. Two consecutive divisions in orthogonal planes are depicted in panels a and b. Reproduced with permission from reference 20.

The model of functional association between peptidoglycan synthesis and cell division still needs further work in light of new observations obtained by super-resolution microscopy. In contrast to original beliefs, S. aureus cells were shown to elongate before dividing, and the septum forms in a slightly unsymmetrical plane, suggesting that the peptidoglycan “piecrust” does not mark quadrants of the cell (20) (Fig. 6B and C).

Interaction with the Host

Purified insoluble peptidoglycan of S. aureus was shown to induce arthritic symptoms in mice and interleukin-1b expression in macrophages, indicating that live bacterial cells are not necessary to elicit an inflammatory response (149, 150). The fact that this heterogeneous polymer is ubiquitous, essential, and unique to bacteria makes it the perfect target for the innate immune system through the action of lysozyme, a muramidase that is well distributed among the animal kingdom (151).

S. aureus peptidoglycan undergoes several secondary modifications (see above), such as O-acetylation of MurNAc and amidation of the second position of the stem peptide, contributing to lysozyme resistance, with a ΔoatA mutant showing impairment in the capacity to mediate septic arthritis (152). Other surface alterations that add to such resistance in S. aureus by modification of the net charge include the incorporation of d-alanine in teichoic acids and l-lysine in the membrane phosphatidylglycerol (153, 154).

Peptidoglycan fragments, released by cell wall turnover, have strong proinflammatory capacity, since they are recognized by host immune receptors, including nucleotide oligomerization domain protein-2 (NOD2), which detects l-lysine-containing muropeptides and the muramyl dipeptide as the minimal recognition structure (155). Peptidoglycan is also recognized by the extracellular Toll-like receptor-2 (TLR2) (156) and by peptidoglycan recognition proteins (PGRPs) that are present in leucocytes, liver, and epithelial cells and also in insects (157). Another clever host concealment strategy of S. aureus was described to rely on the activity of autolysins to trim the peptidoglycan fragments that stand out from the cell wall surface and become exposed to PGRP (158). Recognition of peptidoglycan by all these receptors, and perhaps others that remain to be identified, leads to the activation of multiple host defense responses (159).

Interaction with the External Environment and Recycling

Bacteria build and remodel their peptidoglycan scaffold not only through the primary biosynthetic pathway but also by interaction with the external environment. One form of interaction is peptidoglycan recycling, which although well described in Gram-negative bacteria for which turnover products are retained in the periplasm, has been a subject of debate for Gram-positive organisms. Recently, recycling of N-acetylmuramic acid was described for S. aureus, B. subtilis, and Streptomyces coelicolor, mediated by MurNAc-6P etherase, which converts per generation about 5 to 10% of MurNac-6P to N-acetylglucosamine-6-phosphate and lactate intracellularly. Deletion mutants showed no growth defects during exponential phase, because recycling was shown to occur mainly upon transition to stationary phase, suggesting that this process contributes to survival in the late phase of growth (160). Another form of interaction with the environment is the peptidoglycan chemical plasticity in challenging conditions. Noncanonical d-amino acids are known to be released to the environment by a wide range of bacterial species and to be exogenously incorporated into the mature peptidoglycan structure, through the action of surface associated enzymes, as a mechanism of peptidoglycan editing (161). In fact, the effect of exogenous glycine on the structure of staphylococcal peptidoglycan was described in early studies. Growth of S. aureus in the presence of high concentrations of glycine, d-serine, or other d-amino acids caused dramatic distortions of the peptidoglycan scaffold with reduction in cross-linkage and replacement of the muropeptide components by abnormal structures in which the carboxyterminal d-alanine residues were replaced by glycine or the other amino acids added to the medium (11, 162). One of the best-documented roles for such peptidoglycan editing capacities is to confer resistance to glycopeptides, but other functions related to adaptation to sudden environmental stresses, presence of the host, or an aging cell population have been suggested (161, 163). Most recently, incorporation of exogenous amino acids into S. aureus peptidoglycan was described to occur inside the infected live organism Caenorhabditis elegans, revealing a process of in vivo cell surface remodeling, eventually as a communication platform between pathogen and host (163).

SUMMARY

Through these studies, the staphylococcal cell wall is beginning to emerge not only as a critically important exoskeleton responsible for the structural integrity of the bacterium, but also as an organelle of multiple functions: it is intimately involved with interactions with the system of innate immunity of the host; it is a dynamic scaffold undergoing permanent remodeling upon external changes; it provides covalent attachment sites for a number of host-related surface proteins; and it occupies center stage in resistance to β-lactam and glycopeptide antibiotics. Some of these observations have also begun to throw light on how the control of degradative and synthetic activities and morphogenetic principles come together in the assembly and replication of cell walls.