Penicillin-binding proteins exhibit functional redundancy during asymmetric cell division in Clostridioides difficile

Abstract

Peptidoglycan synthesis is an essential driver of bacterial growth and division. The final steps of this crucial process involve the polymerization of glycan strands by shape, elongation, division, and sporulation (SEDS) family glycosyltransferases and the cross-linking of peptide cross-bridges by class B penicillin-binding proteins (bPBP). While many bacteria use distinct bPBPs to perform specialized roles during a given cellular process, some bPBPs can play redundant roles, particularly in the presence of certain cell wall stresses. Our understanding of these compensatory mechanisms, however, remains incomplete. Endospore-forming bacteria typically encode multiple bPBPs to drive morphological changes required for sporulation. The sporulation-specific bPBP, SpoVD, synthesizes the polar division septum and the cortex peptidoglycan layer during sporulation in the pathogen Clostridioides difficile. Although SpoVD catalytic activity is essential for cortex synthesis, we show that it is partially dispensable for asymmetric division. The dispensability of SpoVD's catalytic activity requires the presence of its SEDS partner, SpoVE, and another sporulation-induced bPBP, PBP3. While PBP3 localizes to the polar septum and interacts with components of the polar division machinery, the ability of PBP3 to promote division during sporulation occurs independent of its catalytic activity. Notably, this latter finding differs from previously reported modes of functional redundancy in bacteria, indicating that there are diverse mechanisms by which penicillin-binding proteins can be functionally redundant in bacteria.IMPORTANCEPeptidoglycan synthesis requires the transpeptidase activity of penicillin-binding proteins (PBPs), which have specialized functions during cell growth, division, and differentiation. However, many bacteria produce PBPs with overlapping functions, and this functional redundancy can lead to increased antibiotic resistance. While the major pathogen, Clostridioides difficile, requires the SpoVD PBP to form spores, we found that its transpeptidase activity is dispensable for asymmetric division, the first morphological stage of sporulation, because a sporulation-induced PBP, PBP3, partially substitutes for SpoVD's function during this stage. Since PBP3's ability to promote asymmetric division in this context does not depend on the its catalytic activity, unlike prior studies of PBP functional redundancy, our analyses highlight the diversity in mechanisms used to enable functional redundancy between PBPs.

Keywords: Clostrioides difficile; asymmetric division; functional redundancy; penicillin-binding proteins; sporulation.

Fig 1

SpoVD catalytic activity is partially dispensable for its function during asymmetric division. (a) Schematic of a SEDS-bPBP peptidoglycan synthase complex. The SEDS glycosyltransferase polymerizes nascent glycan strands from lipid-linked PG precursors in the cytoplasmic membrane, while the class B penicillin-binding protein (bPBP) cross-links the stem peptides between the growing strands. (b) Cytological profile of individual cells representing each of the five morphological stages of sporulation, as indicated. Representative phase-contrast and fluorescence micrographs show wild-type (WT) cells sampled from sporulation-inducing 70:30 plates after 18 h of growth. The nucleoid was stained using Hoechst, and the cell membrane was stained using FM4-64. Cells undergoing asymmetric division (AD) have a flat polar septum; cells undergoing engulfment (EI) have a curved polar septum; cells that have completed engulfment (EC) are indicated by bright-membrane staining around a fully engulfed forespore; phase-visible forespores (PFs) indicate forespores completing maturation visible as phase-dark or phase-bright forespores (yellow arrowheads) associated with the mother cell; mature free spores (FSs) are observable as independent phase-bright particles. (c and d) Quantification of the cytological profiling of cells sampled from sporulation-inducing plates after 20–22 h of growth. White circles indicate data from each replicate; bars indicate the average means; and error bars indicate standard deviation. More than 1,000 total cells and over 100 visibly sporulating cells were analyzed per sample from a minimum of three biological replicates, except for ∆spoVD/spoVD^S311A, for which two biological replicates were conducted. For representative micrographs, see Fig. S2. (c) Distribution of visibly sporulating cells among the indicated stages of sporulation. (d) Proportion of cells that complete and progress beyond asymmetric division, i.e., all visibly sporulating cells, as a percentage of the total cells profiled. Note that the data are normalized to WT (dotted line). ****P < 0.0001. Statistical significance was determined using one-way ANOVA and Tukey’s test. (e) Western blot analyses of SpoVD levels in the indicated strains 14 h after growth on sporulation-inducing plates. The anti-Spo0A antibody was used as a proxy for measuring sporulation induction.

Fig 2

Catalytically inactive SpoVD requires its SEDS partner, SpoVE, to facilitate asymmetric division. (a and b) Quantification of the cytological profiling of cells sampled from sporulation-inducing plates after 20–22 h of growth. White circles indicate data from each replicate; bars indicate average means; and error bars indicate standard deviation. More than 1,000 total cells and over 100 visibly sporulating cells were analyzed per sample from a minimum of three biological replicates. (a) Distribution of visibly sporulating cells among the indicated stages of sporulation. See Fig. 1 for the distribution of WT cells. (b) Proportion of cells that complete and progress beyond asymmetric division, i.e., all visibly sporulating cells, as a percentage of the total cells profiled. Note that the data are normalized to WT (dotted line) and that the spoVD^S311A data were derived from Fig. 1. ****P < 0.0001. Statistical significance was determined using one-way ANOVA and Tukey’s test. (c) Western blot analyses of SpoVD levels in the indicated strains 14 h after growth on sporulation-inducing plates. The anti-Spo0A antibody was used as a proxy for measuring sporulation induction.

Fig 3

PBP3 is a non-essential bPBP that is involved in spore formation. (a) Protein schematic comparing SpoVD and PBP3. Functional domains and catalytic sites were predicted using HMMER (http://hmmer.org/) and revised to reflect the identification of the pedestal domain which interacts with SEDS proteins (11). TM: transmembrane domain; PASTA: PBP and serine/threonine kinase-associated domain. The catalytic serine residues are shown in green. (b) Western blot showing the levels of SpoVD, PBP3, and Spo0A in cells sampled from sporulation-inducing plates after ~14 h of growth. SpoVD and PBP3 are not detected in the ∆spo0A strain, which cannot initiate sporulation. * indicates a non-specific band detected by the anti-PBP3 antibody. (c) Efficiency of heat-resistant spore formation (sporulation efficiency) of the pbp3 mutant and complemented strains relative to WT. Means with standard deviations are indicated. Cells were collected from sporulation-inducing 70:30 plates ~20 to 22 h after inoculation. Data are from a minimum of five biological replicates. **P < 0.01, ***p < 0.001, ****P < 0.0001. Statistical significance was determined using one-way ANOVA and Tukey’s test. (d) Representative phase-contrast micrographs of WT, pbp3 mutant, and complemented cells collected from sporulation-inducing 70:30 plates after ~20 h of growth. Examples of phase-bright spores are indicated by yellow arrowheads. The blue arrowhead highlights an elongated forespore in the ∆pbp3 mutant. Scale bar, 5 µm. (e) Representative transmission electron microscopy (TEM) images of WT and ∆pbp3 strains. The forespore of ∆pbp3 cells was frequently observed to form an elongated shape in the TEM samples. Scale bar, 500 nm.

Fig 4

PBP3 partially compensates for the loss of SpoVD catalytic activity during asymmetric division. (a, b) Quantification of the cytological profiling of cells sampled from sporulation-inducing plates after 20–22 h of growth. White circles indicate data from each replicate, bars indicate average means, and error bars indicate standard deviation. More than 1,000 total cells and over 100 visibly sporulating cells were analyzed per sample. Data are from a minimum of three independent experiments. For representative micrographs, see Fig. S3. (a) Distribution of visibly sporulating cells among the indicated stages of sporulation. See Fig. 1b for the distribution of WT cells. (b) Proportion of cells that complete and progress beyond asymmetric division, i.e., all visibly sporulating cells, as a percentage of the total cells profiled. Note that the data are normalized to WT (dotted line) and that the spoVD^S311A data were derived from Fig. 1. ns, not significant. *P < 0.05, **P < 0.01, ****P < 0.0001. Statistical significance was determined using one-way ANOVA and Tukey’s test. (c) Western blot showing the levels of SpoVD, PBP3, and Spo0A in cells sampled from sporulation-inducing plates after ~14 h of growth. SpoVD and PBP3 are not detected in the ∆spo0A strain, which cannot initiate sporulation. *Indicates a non-specific band detected by the anti-PBP3 antibody.

Fig 5

PBP3 interacts with components of the polar divisome. (a) Bacterial two-hybrid analysis of interactions between PBP3 and other PG synthases or components of the polar divisome. The β-galactosidase activity was normalized to the negative control. N-terminal T18 or T24 fusion to PBP3 was paired with reciprocal N-terminal fusions to the indicated proteins. Data are from three technical replicates. (b) The schematic shows interactions detected in the bacterial two-hybrid analyses. Components of the predicted polar divisome are indicated. PBP1 may also be a part of the polar divisome based on co-immunoprecipitation analyses using SpoVD-FLAG₃ as bait. (c and d) Co-immunoprecipitations performed on cells sampled from sporulation-inducing plates after 12 h of growth. (c) PBP3-FLAG₃ was used as bait in the ∆pbp3/pbp3-FLAG₃ strain background; (d) SpoVD-FLAG₃ was used as bait in the ∆spoVD/spoVD-FLAG₃ (WT) and ∆spoVD∆pbp3/spoVD-FLAG₃ (∆pbp3) strain backgrounds. ∆pbp3/pbp3 and ∆spoVD/spoVD strains were used as negative controls (no tag). The presence of SpoVD, PBP3, and PBP1 in the pull-downs was probed using antibodies against the indicated proteins and Western blotting. The FLAG-tagged proteins were detected using an anti-FLAG antibody. SpoIID was used as a control protein because it is also a PG-associated transmembrane protein localized to the forespore membrane, but it is not predicted to be a part of the polar divisome.

Fig 6

PBP1 and PBP3 localize to the site of asymmetric division. C. difficile strains harboring either mScI3-pbp1 (a–d) or mScI3-pbp3 (e and f) expression constructs under an aTc-inducible promoter were grown on sporulating-inducing plates containing 10 ng/mL aTc for 12 h. Cells were then labeled with HADA and fixed for fluorescence microscopy. A representative cell undergoing asymmetric division is shown in the inset for each genotype. The mScI3 and HADA fluorescence along the medial axis of a cell was quantified for 20 individual cells from two independent experiments, and the fluorescent signal was normalized to the maximum fluorescence for each cell before generating aggregate curves. The mean normalized fluorescence ± standard deviation is graphed along the normalized cell distance. The level of induction by aTc on agar medium was variable, likely due to altered diffusion through the bacterial lawn on the agar surface, resulting in some bacteria showing a decreased mScI3 signal (insets in a–d). To facilitate comparison between images, the insets in panels a and d were adjusted to the same brightness/contrast, and those in panels e and f were adjusted to the same brightness/contrast. Yellow arrowheads point to the site of asymmetric division. Scale bars, 2 µm. Images are representative of at least two independent experiments.

Fig 7

Prevalence of bPBP and SEDS enzymes in Firmicutes. Heatmaps showing the distribution of class B penicillin-binding protein (bPBP) and SEDS protein numbers encoded in the genomes of sporulating (n = 328) and non-sporulating (n = 166) Firmicutes organisms. Sporulation ability was inferred by the presence of broadly conserved sporulation-specific genes spo0A and spoIIE in the genome. The data set comprises 494 diverse Firmicutes organisms as reported in reference .