Bacterial Cell Enlargement Requires Control of Cell Wall Stiffness Mediated by Peptidoglycan Hydrolases

Abstract

Most bacterial cells are enclosed in a single macromolecule of the cell wall polymer, peptidoglycan, which is required for shape determination and maintenance of viability, while peptidoglycan biosynthesis is an important antibiotic target. It is hypothesized that cellular enlargement requires regional expansion of the cell wall through coordinated insertion and hydrolysis of peptidoglycan. Here, a group of (apparent glucosaminidase) peptidoglycan hydrolases are identified that are together required for cell enlargement and correct cellular morphology of Staphylococcus aureus, demonstrating the overall importance of this enzyme activity. These are Atl, SagA, ScaH, and SagB. The major advance here is the explanation of the observed morphological defects in terms of the mechanical and biochemical properties of peptidoglycan. It was shown that cells lacking groups of these hydrolases have increased surface stiffness and, in the absence of SagB, substantially increased glycan chain length. This indicates that, beyond their established roles (for example in cell separation), some hydrolases enable cellular enlargement by making peptidoglycan easier to stretch, providing the first direct evidence demonstrating that cellular enlargement occurs via modulation of the mechanical properties of peptidoglycan.

Importance: Understanding bacterial growth and division is a fundamental problem, and knowledge in this area underlies the treatment of many infectious diseases. Almost all bacteria are surrounded by a macromolecule of peptidoglycan that encloses the cell and maintains shape, and bacterial cells must increase the size of this molecule in order to enlarge themselves. This requires not only the insertion of new peptidoglycan monomers, a process targeted by antibiotics, including penicillin, but also breakage of existing bonds, a potentially hazardous activity for the cell. Using Staphylococcus aureus, we have identified a set of enzymes that are critical for cellular enlargement. We show that these enzymes are required for normal growth and define the mechanism through which cellular enlargement is accomplished, i.e., by breaking bonds in the peptidoglycan, which reduces the stiffness of the cell wall, enabling it to stretch and expand, a process that is likely to be fundamental to many bacteria.

FIG 1

Morphological dynamics during the cell cycle of S. aureus. (a) FM 1-43 labeling of living S. aureus SH1000 cells shows that the bacteria change shape rapidly immediately after division (see dashed boxes). (b) Images of S. aureus after labeling of the cell wall with Van-FL (~1.7-kDa) and WGA-AF350 (~38-kDa) fluorescent probe molecules. Arrowheads show cells where Van-FL is bound to regions from which WGA-AF350 has been excluded.

FIG 2

Role of glucosaminidases in population growth. (a) Physical map showing the domain structure of putative glucosaminidases of S. aureus. Percentage homology to Atl glucosaminidase domain and length of homologous amino acid sequence (aa) are indicated in brackets. Black, signal peptide; green, propeptide; grey, repeat domains R1, R2, and R3; blue, N-acetyl-muramyl-l-alanine amidase domain; red, N-acetyl-β-d-glucosaminidase domain; purple, cysteine, histidine-dependent amidohydrolase/peptidase (CHAP) domain. (b) Growth of SH4615 (P_spac-sagB atl sagA scaH) with or without IPTG.

FIG 3

Morphological defects in S. aureus cells lacking glucosaminidases. (a) Images of fixed, Van-FL-labeled cells showing altered morphology. Arrowheads indicate hemispherical cells. Cells of this shape are not found in wild-type populations unless attached to a sister cell. (b) Examples of hemispherical cells. In some cells, septa are visible in bacteria that have not completed the shape change to the mature spherical morphology. This shows that correct shape change is not taking place within the duration of the cell cycle. (c) Quantification of the proportion of hemispherical cells in each sample. P values are the result of Fisher’s exact tests comparing the wild type with each mutant.

FIG 4

Mechanical properties of the S. aureus cell wall. (a) AFM heights and effective spring constants (stiffness maps) of SH1000 and SH4608 (sagB) derived from force maps. In the height map, regions with a lighter color are higher than darker regions. In the stiffness map, regions with a lighter color are stiffer than darker regions. Scale bars, 200 nm; height scale, 500 nm; stiffness scale, 0.010 to 0.018 Nm⁻¹. (b) Stiffness of the cell wall of wild-type and glucosaminidase mutant strains, derived from AFM force maps.

FIG 5

Role of glucosaminidase activity in glycan chain length determination in S. aureus. Strains lacking sagB had substantially longer glycan strands than did SH1000. The presence or absence of other glucosaminidase-encoding genes (atl, sagA, and scaH) had minimal effect on strand length. Annotations show proportions of glycan strands within ranges of numbers of disaccharides (DS); grey traces show the proportions in SH1000 cells for comparison. To compensate for the fact that longer glycan strands incorporate more [¹⁴C]GlcNAc, radioactivity counts (cpm) were divided by the corresponding theoretical molecular weight (see Materials and Methods). The glycan chain abundance is plotted normalized relative to the maximal ratio between radioactivity counts and theoretical molecular weight (cpm/MW).