Staying in shape

Introduction

Bacteria display a large variety in shape, size (typically between 0.5 and 5 µm) and colonial morphology. Most bacteria, however, come in one of the three general shapes as shown in figure 1 (adapted from 1). These are: ball-shaped (coccus), cylindrical (rod or bacillus) and curved/spiral-shaped (spirillum or spirochete). Cocci have been observed as spheres as well as bean-shaped or pointed variants. Likewise, rods also exhibit considerable structural variation, which comprise coccobacilli - filamentous, short and plump-like cells. Important spiral-shaped bacteria include curved, rod-like cells known as vibrios and cylindrical cells equipped with flagella for locomotion called spirilla (1). These different cellular shapes are dictated by the bacterial cell wall, which, according to live cell images, is a porous mesh-like hydrogel (8). In 1884, a staining procedure was developed by Christian Gramm that is based on the structural difference between bacterial cell walls and allows bacteria to be classified into two main groups. These are Gram-positive bacteria that retain the staining compound and Gram-negative ones that do not (2). It is now well established that the bacterial cell wall is a complex and multilayered structure that in most bacteria comprises a peptidoglycan (murein) layer. This is an elastic and mesh-like biopolymer that surrounds the cytoplasmic membrane (3). Functionally, the peptidoglycan layer maintains cell shape and protects the cell from lysis due to its internal osmotic pressure. Moreover, it also serves as attachment site for proteins and other polymers (4). The cell wall of Gram-positive bacteria is characterized by multiple thick peptidoglycan layers, while Gram-negatives contain a thin peptidoglycan cell wall (2). Chemically, peptidoglycan comprises repeating units of the disaccharide N-acetylglucosamine-N-acetylmuramic acid, which are cross-linked through their pentapeptide side chains. This results in a porous sacculus made up of roughly parallel peptidoglycan strands. The general peptidoglycan structure is conserved amongst all bacteria, although variations are known (3). The biosynthesis of peptidoglycan comprises three overall steps, namely: (i) production of precursors in the cytoplasm (ii) assembly of the precursors into lipid II, the basic peptidoglycan intermediate, at the cytosolic face of the cytoplasmic membrane, and (iii) crosslinking and maturation of glycan strands (3,4). Peptidoglycan synthesis is initiated in the cytoplasm with the production of the nucleotide-activated precursor molecules UDP-N-acetylglucosamine and UDP-N-acetylmuramyl pentapeptide. Subsequently, these precursors are assembled onto undecaprenyl, a lipid derivative of 55 C atoms, that functions as carrier. This generates lipid II, the lipid-anchored disaccharide-pentapeptide intermediate, which is flipped to the outer face of the cytoplasmic membrane. Finally, lipid II is polymerized and the new glycan strands are incorporated into the sacculus (3,4). These latter stages require a transglycosylation reaction for polymerization and transpeptidation to form peptide bond crosslinks between the glycan strands. It was thought for a long time that transglycosylation was catalyzed by bifunctional class A penicillin-binding proteins and monofunctional transglycosylases. However, accumulating genetic and biochemical evidence revealed another protein complex made up of RodA and PBP2 that facilitates peptidoglycan production in the absence of aPBPs, thereby serving as a peptidoglycan synthase (5,6). The detailed structure of this novel peptidoglycan synthase was determined recently, providing profound molecular mechanistic insight into bacterial cell wall biosynthesis (7). Here, I will discuss this structure as well as its current functional understanding.

Structurally, peptidoglycan is made up of long glycan strands cross-linked by short peptide fragments. Figure 2 shows that these glycan strands are made up of repeating N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) residues, which are linked through a β-1→4 bond (9). The MurNAc moiety is equipped with a short (stem) pentapeptide whose composition is species-specific and comprises regular L-amino acids as well as unusual D-amino acids. The latter are synthesized by racemases from the respective L-enantiomer and modify the strength and flexibility of peptidoglycan (9). The amino acid composition of the E. coli and B. subtilis stem peptide is presented in the left panel of figure 2 (adapted from 4), revealing that it constitutes L-Ala-D-Glu-mDAP (meso-diaminopimelic acid)–D-Ala. The terminal Ala is released during the polymerization reaction. Although the type and extent of peptide crosslinking between the glycan strands is species-specific, crosslinking typically occurs between the carboxyl group of D-Ala at position 4 and the amino group of mDAP at position 3 (figure 2). A schematic representation of peptidoglycan biosynthesis is presented in the right panel of figure 2, showing that this process involves overall three main steps. The biosynthesis is initiated with production of the nucleotide-activated precursors (UDP-GlcNAc and UDP—MurNAc) in the cytoplasm. UDP-GlcNAc is obtained from fructose-6-phosphate by the sequential action of the enzymes GlmS, GlmM and GlmU (4). Of these, GlmU is a bifunctional enzyme, displaying acetyltransferase and uridyltransferase activity, that catalyzes the final step in the production of UDP-GlcNAc. This is a two-step conversion, requiring acetyl-CoA and UTP (11). Subsequent condensation with phosphoenolpyruvate (PEP) catalyzed by MurA and reduction by MurB generates UDP—MurNAc. Next, the stem peptide is appended onto UDP—MurNAc through sequential addition of L-Ala, D-Glu, mDap and D-Ala-D-Ala catalyzed by the ATP-dependent ligases MurC, MurC, MurD, MurE and MurF, respectively (4). Following this, phospho-MurNAc-pentapeptide is loaded onto undecaprenyl pyrophosphate, the membrane-embedded lipid carrier, by MraY (Phospho-N-acetylmuramoyl-pentapeptide-transferase). This yields lipid I (undecaprenyl pyrophosphate MurNAc-pentapeptide), the first lipid-anchored peptidoglycan intermediate. After this, MurG catalyzes the transfer of GlcNAc onto lipid I, generating lipid II (undecaprenyl pyrophosphate MurNAc-(pentapeptide)GlcNAc), the second lipid-anchored peptidoglycan intermediate (4).  The final biosynthetic reaction of peptidoglycan that occurs in the cytosol comprises the translocation of lipid II from the inner to the outer leaflet of the cytoplasmic membrane. This step is facilitated by a lipid II flippase, an enzyme of which the identity is debated. However, accumulating biochemical and genetic evidence strongly suggest that two cytoplasmic membrane proteins, MurJ and Amj, function as lipid II flippase (12). Following translocation of lipid II across the cytoplasmic membrane, the GlcNAc-MurNAc disaccharide is incorporated into the sacculus through polymerization of the glycan chains and crosslinking of the stem peptides. The polymerization reaction is catalyzed by glycosyltransferases (GTase), whereas crosslinking requires transpeptidases (TPase). These enzymes are collectively known as peptidoglycan synthases of which three distinct types have been described, namely: bifunctional GTase-TPase, monofunctional TPases and monofunctional GTases (3). Penicillin-binding proteins (PBPs), named after their ability to bind β-lactam antibiotics, represent an important class of membrane-embedded peptidoglycan synthases. Class A PBPs are bifunctional, while class B PBPs display only TPase activity (3). Bacteria possess multiple PBPs E. coli, for example, contains twelve PBPs that include three class A proteins (PBP1a, PBP1b and PBP1c) and two class B PBPs (PBP2 and PBP3). PBP1a and PBP1b are the most important class A PBPs and either one is required for cell viability (13). The bifunctional activity of these proteins has been confirmed through in vitro studies with their lipid II substrate (3). The other seven proteins are involved in cell separation, maturation of peptidoglycan or its recycling. PBP2 is required for cell elongation and PBP3 is a major component of the divisome, a protein complex that facilitates cell division (13).  Following incorporation into the existing sacculus, newly synthesized peptidoglycan strands are modified by different processing enzymes. For example, peptides are trimmed by carboxypeptidases or removed from glycan chains by amidases, crosslinks are cleaved by endopeptidases and glycan chains are cleaved by transglycosylases (3). Peptidoglycan is a crucial structural component of the cell wall that is essential for bacterial growth and survival. Many of the enzymes discussed above represent important targets for some of the most potent antibiotics (14). Methicillin, for example, prevents crosslinking of peptidoglycan strands through inhibiting the TPase activity of class A PBPs. Likewise, oxacillin and flucloxacillin also prevent crosslinking of glycan strands.

Biochemical and structural features of a novel peptidoglycan synthase complex

The classical view of peptidoglycan biosynthesis as described above with class A PBPs as the principal peptidoglycan synthases was challenged in 2003 by the finding that mutant cells of B. subtilis lacking all class A PBPs were able to grow and produced peptidoglycan (5). This, therefore, points towards another peptidoglycan synthase that is complementary to class A PBPs. Recently, genetic and biochemical experiments were reported that showed that in B. subtilis the cytoplasmic membrane protein RodA functions as peptidoglycan polymerase, displaying GTase activity (6), suggesting that probably most bacteria employ two distinct classes of cell wall synthases. The cell cycle of rod-shaped bacteria, such as E. coli and B. subtilis, is regulated by two protein complexes, the elongasome and divisome. These facilitate cell elongation and cell division, respectively. The E. coli elongasome is made up seven proteins that include the peptidoglycan synthases RodA, PBP2 and PBP1A (4). B. subtilis RodA is a protein of about 43 kDa with 10 predicted transmembrane domains (TMDs). Moreover, RodA belongs to the conserved SEDS protein family (shape, elongation, division and sporulation) made up transmembrane enzymes with an essential but poorly understood roles in cell wall biosynthesis. Although it was demonstrated that RodA represents a peptidoglycan polymerase, this activity alone is not sufficient to build a cell wall. This, in addition to GTase activity, also requires TPase activity for peptide crosslinking that is catalyzed by class A and B PBPs. With regards to this it should be noted that biochemical evidence has been presented indicating that in E. coli cells RodA forms a complex with PBP2, a class B PBPs (15,16,            17). Thus, a complex between a SEDS protein and a class B PBP exhibits both polymerization and crosslinking activity similar to class A PBPs. The crystal structure of Thermus thermophilus RodA at a resolution of 2.9 Å was established recently (18). This is shown in figure 3 in surface (left panel with hydrophobic residues in red and hydrophilic ones in white) and ribbon representation (center panel), revealing that the overall structure comprises 10 TMDs that are connected by loops. RodA. The cytoplasmic loops between the TMDs are short, while the extracellular loops are large and contain many residues that are functionally important (shown as yellow sticks in the center panel). Interestingly, a long hydrophobic groove (indicated with asterisks in the left panel) was observed between TMD3 (in green) and TMD2 (in orange). This groove is bordered by several highly conserved residues (shown as yellow sticks in the right panel) and may, therefore, comprise the binding-site for lipid-anchored substrates. These residues include Glu108 and Lys111 that form an absolutely conserved salt bridge and are essential for RodA activity. Likewise, Asp255, Asp152 and Trp96 are crucial for proper RodA functioning. Thus, the presence of functionally important residues near the central cavity confirms that this region has a key role in peptidoglycan polymerization, making it catalytically essential. Furthermore, it was proposed that TMD8 and 9 contain the binding site between class B PBPs and RodA.

Recent insights from genetic and biochemical studies strengthened the view that SEDS proteins and class B PBPs function as a complex inside bacterial cells (15,16,17). This is in line with the proposed interaction between RodA and PBP2. To study this interaction in more detail, the crystal structure of the RodA-PBP2 complex was elucidated (7). To this end, a catalytically inactive variant of T. thermophilus RodA and wild-type PBP2 were coexpressed in E. coli cells and purified as a stochiometric complex. Subsequently, the crystal structure was solved at 3.3 Å resolution, which is represented in figure 4 in both surface (left panel) and ribbon representation (right panel). The catalytic residues of RodA (Asp255) and PBP2 (Ser308) are shown as orange spheres. The overall structure of the complex is remarkably compact and clearly reveals the two extracytoplasmic domains of PBP2 – the pedestal and transpeptidase domain (right panel).

The pedestal domain is located on top of the extracytoplasmic loops of RodA and is made up of two subdomains, namely: the anchor and head subdomains. The anchor is composed of a small β-sheet and is connected to hinge region that is attached to the N-terminal TMD. The anchor is C-terminally fused to the head domain that is made up of a small β-sheet with four α-helices. The C-terminal transpeptidase domain is located above the membrane plane and is connected to the head. This catalytic domain comprises a core β-sheet that is surrounded by α-helices. Within the dimeric complex, two main interaction surfaces between RodA and PBP2 are observed, designated interface 1 and interface 2 (left panel). The first interface is located in the lipid bilayer where the TMD of PBP2 interacts with TMD8 and 9 of RodA as noted previously. This interface is mainly hydrophobic. The second interface is above the membrane plane on the extracytoplasmic loops of RodA. Despite this second interaction site, it seems that the first interface provides most of the energy for RodA and PBP2 binding. A structural comparison between the structure of RodA alone (PDB 6BAR (18)) and the one in the complex (PDB 6PL6 (7)) is shown in figure 5 and reveals that, overall, both structures are highly similar. However, when bound to PBP2, TMD7 (in grey) of RodA is moved away from the core. This exposes a cavity accessible from the membrane (indicated by an asterisks) that is not present in the structure of RodA alone. Interestingly, this cavity is large enough to accommodate a lipid II molecule and it is, therefore, conceivable that it represents a substrate entry or exit site. 

Analysis of RodA’s glycosyltransferase activity established that this was only detected when both RodA and PBP2 were present. Possibly, the pedestal domain of PBP2 is involved in the allosteric activation of RodA because it is associated with an extracellular loop of RodA that contains many functionally important residues (interface 2). Upon disruption of this interaction site by site-directed mutagenesis within PBP2, the glycosyltransferase activity of RodA was strongly reduced, while copurification was not affected. This, indeed, suggests that the pedestal domain of PBP2 is crucial for stimulating the polymerization activity of RodA. The pedestal domain is not only important for controlling the activity of RodA but, in fact, ensures a proper functioning of the elongasome – the cell elongation complex formed by RodA-PBP2 and its partners.  

RodA-PBP2-meditated peptidoglycan cell wall biosynthesis

The compact conformation of the RodA-PBP2 complex raises the question how the transpeptidase domain of PBP2 ensures crosslinking of glycan strands near the cell wall above the membrane plane. With regards to this question, it should be noted that data obtained by negative-stain microscopy reveal that the RodA-PBP2 complex is able to adopt a variety of conformations, including compact and extended conformations. In the latter state, PBP2 might be able to reach above the membrane plane with its active site near the peptidoglycan layer. Taken together, the structural data discussed above establish that the pedestal domain is a crucial switch that controls peptidoglycan polymerization and crosslinking activities. The model shown in figure 6 (adapted from 7) proposes that by adopting either a compact or extended conformation, PBP2 controls the polymerization of newly synthesized glycan strands through RodA near the membrane or crosslinking by placing its transpeptidase domain near the sacculus. 

Summarizing conclusion

Although bacteria display a large variety in shape and size, they all rely on peptidoglycan within their cell wall to maintain their proper cell shape and prevent the cell from rupture due to its internal pressure. Peptidoglycan represents a unique mesh-like biopolymer that surrounds the cytoplasmic membrane as a bag-shaped macromolecule (2,3,4). Structurally, peptidoglycan is made up of long glycan strands cross-linked by short peptide fragments. These glycan strands are made up of repeating N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) residues, which are linked through a β-1→4 bond. The MurNAc moiety is equipped with a short (stem) pentapeptide (3,4,9). The biosynthesis of peptidoglycan comprises three overall steps, namely: (i) production of precursors in the cytoplasm (ii) assembly of the precursors into lipid II, the basic peptidoglycan intermediate, at the cytosolic face of the cytoplasmic membrane, and (iii) crosslinking and maturation of glycan strands (3,4). Peptidoglycan synthesis is initiated in the cytoplasm with the production of the nucleotide-activated precursor molecules UDP-N-acetylglucosamine and UDP-N-acetylmuramyl pentapeptide. Subsequently, these precursors are assembled onto undecaprenyl, that functions as lipid carrier. This generates lipid II, the lipid-anchored disaccharide-pentapeptide intermediate, which is flipped to the outer face of the cytoplasmic membrane. Finally, lipid II is polymerized and the new glycan strands are incorporated into the sacculus (3,4). These latter stages require a transglycosylation reaction for polymerization of the disaccharide units and transpeptidation to form peptide bond crosslinks between the glycan strands. The enzymes that catalyze peptidoglycan biosynthesis are collectively known as peptidoglycan synthases of which three distinct types have been described, namely: bifunctional GTase-TPase, monofunctional TPases and monofunctional GTases (10). Penicillin-binding proteins (PBPs), represent an important class of membrane-embedded peptidoglycan synthases. Class A PBPs are bifunctional, while class B PBPs display only TPase activity (3). However, accumulating genetic and biochemical evidence indicates that bacteria employ a second type of peptidoglycan synthase that functions in parallel of class A PBPs (5). This novel cell wall synthase is made up of the SEDS protein RodA and PBP2, its cognate class B PBP (6,7,15). RodA facilitates the polymerization of newly synthesized glycan chains, while PBP2 drives the crosslinking of the glycan strands followed by their incorporation into the existing sacculus. Thus, SEDS-class B PBP systems comprise a newly identified cell wall synthase that are present in nearly all bacterial genomes unlike class A PBPs. PBPs represent important targets for some of the most potent antibiotics (14). Methicillin, for example, prevents crosslinking of peptidoglycan strands through inhibiting the TPase activity of class A PBPs. Likewise, oxacillin and flucloxacillin also prevent crosslinking of glycan strands. To combat the growing number of antibiotic resistant bacteria, novel antibiotic targets are highly desired. With regards to the latter, it should be mentioned that the RodA-PBP2 complex provides an attractive drug target and it can therefore be expected that the detailed structure of the RodA-PBP2 complex, that was assessed recently (7), will aid the design of novel antimicrobials.   

References

1.      Talaro, K. P., & Talaro, A. (2002). Foundations in microbiology (4th ed.). Boston: McGraw-Hill.

2.      Silhavy TJ, Kahne D, Walker S. 2010. The bacterial cell envelope. Cold Spring Harb Perspect Biol. 2: a000414.

3.      Typas A, Banzhaf M, Gross CA, Vollmer W. 2011. From the regulation of peptidoglycan synthesis to bacterial growth and morphology. Nat Rev Microbiol. 10: 123-136.

4.      Pazos M, Peters K. 2019. Peptidoglycan. Subcell Biochem. 92: 127-168.

5.      McPherson DC, Popham DL. 2003. Peptidoglycan synthesis in the absence of class A penicillin-binding proteins in Bacillus subtilis. J Bacteriol. 185: 1423-1431.

6.      Meeske AJ, Riley EP, Robins WP, et al. 2016. SEDS proteins are a widespread family of bacterial cell wall polymerases. Nature. 537: 634-638.

7.      Sjodt M, Rohs PDA, Gilman MSA, et al. 2020. Structural coordination of polymerization and crosslinking by a SEDS-bPBP peptidoglycan synthase complex. Nat Microbiol. doi: 10.1038/s41564-020-0687-z.

8.      Pasquina-Lemonche L, Burns J, Turner R, et al. 2020. The architecture of the Gram-positive bacterial cell wall. Nature. doi: 10.1038/s41586-020-2236-6.

9.      Vollmer W, Blanot D, de Pedro MA. 2008. Peptidoglycan structure and architecture. FEMS Microbiol Rev. 32: 149-167.

10.   Vollmer W, Bertsche U. 2008. Murein (peptidoglycan) structure, architecture and biosynthesis in Escherichia coli. Biochim Biophys Acta. 1778: 1714-1734.

11.   Olsen LR, Vetting MW, Roderick SL. 2007. Structure of the E. coli bifunctional GlmU acetyltransferase active site with substrates and products. Protein Sci. 16: 1230-1235.

12.   Meeske AJ, Sham LT, Kimsey H. et al. 2015. MurJ and a novel lipid II flippase are required for cell wall biogenesis in Bacillus subtilis. Proc Natl Acad Sci U S A. 112: 6437-6442.

13.   Sauvage E, Kerff F, Terrak M. et al. 2008. The penicillin-binding proteins: structure and role in peptidoglycan biosynthesis. FEMS Microbiol Rev. 32: 234-258.

14.   Cochrane SA, Lohans CT. 2020. Breaking down the cell wall: Strategies for antibiotic discovery targeting bacterial transpeptidases. Eur J Med Chem. 194:112262.

15.   Cho H, Wivagg CN, Kapoor M et al. 2016. Bacterial cell wall biogenesis is mediated by SEDS and PBP polymerase families functioning semi-autonomously. Nat Microbiol. 1:16172.

16.   Rohs PDA, Buss J, Sim SI et al. 2018. A central role for PBP2 in the activation of peptidoglycan polymerization by the bacterial cell elongation machinery. PLoS Genet. 14(10):e1007726.

17.   Taguchi A, Welsh MA, Marmont LS et al. 2019. FtsW is a peptidoglycan polymerase that is functional only in complex with its cognate penicillin-binding protein. Nat Microbiol. 4:587-594.

18.   Sjodt M, Brock K, Dobihal G et al. 2018. Structure of the peptidoglycan polymerase RodA resolved by evolutionary coupling analysis. Nature. 556: 118-121.