The bacterial cell wall takes centre stage

An unexpected function has been assigned to part of the molecular machinery that synthesizes the bacterial cell wall — a dramatic shift in our understanding that may have major implications for antibiotic development. See Article p.634

Shakespeare's comedies are full of mistaken identities — a woman is thought to be a man, lovers pursue egregiously misidentified objects of their affections, a duke pretends to be priest — and hilarity ensues. Although this works well in comedy, it's not so humorous in science. A paper on page 634 by Meeske et al.1 raises the curtain on an analogous plot twist as the authors shatter a deeply held assumption about how bacterial cell walls are built, a conclusion that is also supported by the work of Cho et al.2 in a paper published in Nature Microbiology.

Previously, only class A penicillin-binding proteins (aPBPs) were known to polymerize the molecule peptidoglycan3, the backbone of most bacterial cell walls. Now, Meeske et al. and Cho et al. report that this enzymatic step can be performed by an unrelated protein, RodA, which had been thought to have a completely different role. The discovery overturns long-held ideas about a fundamental biological process and exposes a different target towards which antibiotic development might be directed.

Our view of cell-wall synthesis had seemed fairly straightforward (Fig. 1a). Two carbohydrate molecules are linked to one another to form a disaccharide, which is then 'flipped' from inside the cell across the cell membrane to the cell exterior. There, the glycosyltransferase domain of an aPBP adds this disaccharide to a growing polysaccharide chain, and a transpeptidase enzyme binds this new chain to older ones by creating peptide crosslinks4. The result is a large, interconnected, net-like mesh known as a peptidoglycan that wraps around the cell and acts like an external skeleton. Breach this wall, and most bacteria will burst like overinflated balloons, which is one reason that cell-wall synthesis has been studied so intensively as a potential target for antibiotics.

Figure 1: Models of peptidoglycan synthesis in bacterial cell walls.

a, In the conventional model, a dissacharide (not shown) is flipped across the cell membrane from the cytoplasmic side to the cell exterior, and delivered to a class A penicillin-binding protein (aPBP), which contains a glycosyltransferase enzyme domain (GT, blue oval) and a transpeptidase enzyme domain (TP, red oval). The glycosyltransferase domain polymerizes the disaccharide substrates into long glycan chains (blue lines), and the transpeptidase domain links these chains through short peptides (red lines) to create the peptidoglycan cell wall. The RodA protein and a class B penicillin-binding protein (bPBP) have unknown roles in this model. b, In the model proposed by Meeske et al.1 and Cho et al.2, substrate is delivered to RodA (blue rectangle), the newly recognized glycosyltransferase that catalyses chain formation. This chain is crosslinked to existing strands by the transpeptidase domain of bPBP. In this scheme, the aPBPs also polymerize and crosslink glycan chains, but their relationship to the RodA or bPBP complex is undefined.

In the established view, only aPBPs have glycosyltransferase polymerizing activity, and for decades no other such enzymes were thought to be involved. The major upheaval precipitated by Meeske et al. and Cho et al. is caused by their finding that RodA can perform this glycosyltransferase step, perhaps even displacing aPBPs from a central role in the process (Fig. 1b).

Although this is something of a shock, we really should have been prepared. Whispers (or shouts) that something was amiss were heard as long as 13 years ago, with reports that two bacteria, Bacillus subtilis5 and Enterococcus faecalis6, can survive in the absence of all known aPBPs. This led to predictions that an unidentified protein — possibly RodA (ref. 5) — might have glycosyltransferase activity. But RodA represents an entirely different family of proteins (the SEDS family) and, in its absence, bacterial cells that are usually rod-like in shape7 round up and grow as spheres. What RodA actually does has engendered vigorous debate8. An old but recently resurrected proposal is that RodA and other SEDS proteins flip disaccharides across the membrane to deliver these substrates to the aPBPs8,9.

Thus, the idea that RodA and other SEDS proteins might function as peptidoglycan polymerases was not just unexpected, it was all but actively disregarded. That it has taken so long to confirm the identity of this new class of polymerase is a somewhat dubious tribute to our ability to ignore data that run counter to preconceptions.

In defence of the field, measuring this kind of glycosyltransferase activity in living cells was mostly out of reach until the advent of super-resolution microscopy about five years ago. Meeske and colleagues monitored peptidoglycan synthesis by visualizing the microscopic motion of Mbl, a protein that is part of the RodA machinery that elongates the rod-shaped cells of B. subtilis. Cho and colleagues used a similar approach by monitoring MreB, a protein in the RodA complex of Escherichia coli. In both species, these complexes constantly circle the cells, inserting new peptidoglycan as they go. However, Meeske and colleagues confirmed that adding a β-lactam antibiotic inhibited cell-wall synthesis and stopped this motion, indicating that peptidoglycan polymerization is the motor that drives the RodA machine.

Surprisingly, though, both groups found that the RodA apparatus continued to move when aPBPs were absent or inhibited; this indicated that another synthase enzyme must be operating. Furthermore, Cho and colleagues observed that motion stopped when RodA was removed or inactivated. Meeske et al. found that overexpression of RodA restored rapid growth to a mutant that lacked aPBPs, and a partially purified cell extract containing RodA exhibited glycosyltransferase activity. All of these results argue that RodA can polymerize peptidoglycan.

Why is this transformative? First, we now know that aPBPs are not essential for peptidoglycan synthesis, although they help a lot. For example, B. subtilis strains that lack aPBPs produce aberrant clumps of peptidoglycan and have abortive and oddly placed cell-wall invaginations, and such mutants grow much more slowly than normal5. Thus, aPBPs might still be required for robust growth and division in most bacterial cells. Second, as both Meeske et al. and Cho et al. infer, other members of the SEDS protein family might also have glycosyltransferase activity. For example, the SEDS protein FtsW, which is associated with cell division, might drive peptidoglycan polymerization during this stage of the bacterial cell cycle. In other circumstances, SEDS proteins might be co-opted for different purposes, and some bacteria might exist with no aPBPs at all. Third, because RodA is impervious to compounds that normally inhibit aPBPs, there might be undiscovered antibiotics that target the newly recognized glycosyltransferase abilities of SEDS proteins.

As always with new discoveries, there are questions and concerns. For example, if the SEDS proteins are glycosyltransferase enzymes, then why are there any aPBPs at all? Why has evolution produced (and combined) these two disparate agents? Might RodA flip its own peptidoglycan precursors across the membrane? And how does RodA interact with aPBPs in cells that require these latter enzymes?

Finally, and perversely, I worry that we might exchange one case of mistaken identity for another. Both groups propose that all SEDS proteins have this enzymatic activity, a conclusion drawn from the abilities of a single member, RodA, and the fact that two possible catalytic residues are evolutionarily conserved among the RodA and FtsW proteins of other organisms. Such inductive reasoning certainly opens other lines of inquiry, but the danger is that focusing attention too strongly in one direction might blind researchers to alternative possibilities, recreating our earlier error. In the end, Shakespeare resolves his characters' mistakes, unites the lovers and allows all to live (mostly) happily ever after. But, for microbiologists, this is only Act II, and the stage may yet be set for even more interesting drama.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Meeske, A. J. et al. Nature 537, 634–638 (2016).

  2. 2

    Cho, H. et al. Nature Microbiol. http://dx.doi.org/10.1038/nmicrobiol.2016.172 (2016).

  3. 3

    Egan, A. J. F., Biboy, J., van' t Veer, I., Breukink, E. & Vollmer, W. Phil. Trans. R. Soc. B 370, 20150031 (2015).

  4. 4

    Typas, A., Banzhaf, M., Gross, C. A. & Vollmer, W. Nature Rev. Microbiol. 10, 123–136 (2012)

  5. 5

    McPherson, D. C. & Popham, D. L. J. Bacteriol. 185, 1423–1431 (2003).

  6. 6

    Arbeloa, A. et al. J. Bacteriol. 186, 1221–1228 (2004).

  7. 7

    Bendezú, F. O. & de Boer, P. A. J. J. Bacteriol. 190, 1792–1811 (2008).

  8. 8

    Ruiz, N. Lipid Insights 8 (Suppl. 1), 21–31 (2015).

  9. 9

    Mohammadi, T. et al. EMBO J. 30, 1425–1432 (2011).

Download references

Author information



Corresponding author

Correspondence to Kevin D. Young.

Related links

Related links

Related links in Nature Research

Structural biology: Lipopolysaccharide rolls out the barrel

Microbiology: The dark side of antibiotics

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Young, K. The bacterial cell wall takes centre stage. Nature 537, 622–624 (2016). https://doi.org/10.1038/537622a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.