Three Orphan Histidine Kinases Inhibit Clostridioides difficile Sporulation

ABSTRACT

The ability of the anaerobic gastrointestinal pathogen Clostridioides difficile to survive outside the host relies on the formation of dormant endospores. Spore formation is contingent on the activation of a conserved transcription factor, Spo0A, by phosphorylation. Multiple kinases and phosphatases regulate Spo0A activity in other spore-forming organisms; however, these factors are not well conserved in C. difficile. Previously, we discovered that deletion of a predicted histidine kinase, CD1492, increases sporulation, indicating that CD1492 inhibits C. difficile spore formation. In this study, we investigate the functions of additional predicted orphan histidine kinases CD2492, CD1579, and CD1949, which are hypothesized to regulate Spo0A phosphorylation. Disruption of CD2492 also increased sporulation frequency, similarly to the CD1492 mutant and in contrast to a previous study. A CD1492 CD2492 mutant phenocopied the sporulation and gene expression patterns of the single mutants, suggesting that these proteins function in the same genetic pathway to repress sporulation. Deletion of CD1579 variably increased sporulation frequency; however, knockdown of CD1949 expression did not influence sporulation. We provide evidence that CD1492, CD2492, and CD1579 function as phosphatases, as mutation of the conserved histidine residue for phosphate transfer abolished CD2492 function, and expression of the CD1492 or CD2492 histidine site-directed mutants or the wild-type CD1579 allele in a parent strain resulted in a dominant-negative hypersporulation phenotype. Altogether, at least three predicted histidine kinases, CD1492, CD2492, and CD1579 (herein, PtpA, PtpB and PtpC), repress C. difficile sporulation initiation by regulating activity of Spo0A.

IMPORTANCE The formation of inactive spores is critical for the long-term survival of the gastrointestinal pathogen Clostridioides difficile. The onset of sporulation is controlled by the master regulator of sporulation, Spo0A, which is activated by phosphorylation. Multiple kinases and phosphatases control Spo0A phosphorylation; however, this regulatory pathway is not defined in C. difficile. We show that two predicted histidine kinase proteins, CD1492 (PtpA) and CD2492 (PtpB), function in the same regulatory pathway to repress sporulation by preventing Spo0A phosphorylation. We show that another predicted histidine kinase protein, CD1579 (PtpC), also represses sporulation and present evidence that a fourth predicted histidine kinase protein, CD1949, does not impact sporulation. These results support the idea that C. difficile inhibits sporulation initiation through multiple phosphatases.

INTRODUCTION

Clostridioides difficile undergoes a significant differentiation process to develop dormant endospores, which enable this anaerobic pathogen to survive outside the mammalian gastrointestinal tract for a prolonged period of time. The environmental cues and regulatory pathways that govern the initiation of sporulation all converge on Spo0A, the master regulator of sporulation (1–3). Spo0A is a conserved transcriptional regulator present in all endospore-forming bacteria and is essential to this process (4). Spo0A activity is controlled by the phosphorylation of an aspartate residue, allowing Spo0A to directly bind to specific target sequences in the promoters under Spo0A regulation (3, 5, 6). Thus, active Spo0A∼P drives transcription of sporulation-specific genes whose products are required for entry into the sporulation pathway (7, 8).

In other sporeformers, the opposing activities of numerous orphan histidine kinases and phosphatases contribute to Spo0A phosphorylation, presumably in response to environmental stimuli or nutritional cues. In the well-studied soil bacterium Bacillus subtilis, Spo0A is phosphorylated via an expanded two-component signal transduction system (TCS) known as a phosphorelay (9). The B. subtilis phosphorelay is comprised of multiple proteins: one of several sensor histidine kinases transmits a phosphoryl group to the response regulator Spo0F, and the phosphotransferase Spo0B transfers the phosphoryl group from Spo0F∼P directly to Spo0A. Phosphatases directly dephosphorylate Spo0A or a phosphotransfer protein in the phosphorelay or inhibit activation and/or autophosphorylation of the sensor histidine kinases. However, many of the key regulatory proteins that control Spo0A activation in B. subtilis are absent from the C. difficile genome (10–12), supporting the hypothesis that C. difficile controls the initiation of sporulation differently than the Bacillus sp.

The C. difficile genome does not encode orthologs of the Bacillus species’ intermediate phosphorelay proteins, suggesting that either unique orphan histidine kinases directly phosphorylate Spo0A or that other proteins transfer the phosphoryl group from the kinases to Spo0A (10). Reinforcing the former hypothesis, orphan histidine kinases promote spore formation and have been shown to directly phosphorylate Spo0A in clostridia, including in Clostridium acetobutylicum and Clostridium perfringens (13, 14). In C. difficile, however, only the putative sporulation-associated histidine kinase CD1492 has been studied in depth to ascertain its role in C. difficile sporulation (15). A CD1492 mutant exhibited a hypersporulation phenotype, had decreased TcdA production, and was significantly less virulent in the hamster model of C. difficile infection (15). Two other annotated sporulation-associated histidine kinases, the membrane-bound CD2492 and soluble CD1579, were briefly characterized in a previous study (16). This study showed that a CD2492 mutant has a decreased sporulation frequency via microscopy after extended growth in rich medium and provided evidence that CD1579 directly transferred a phosphoryl group to Spo0A in vitro. However, the conclusions of this study with regard to the function of either protein were limited.

Finally, RstA, a multifunctional protein in C. difficile, positively influences early sporulation events through an unknown mechanism (17, 18). The rstA mutant exhibits the opposite pattern of gene expression and reverse sporulation, toxin, and motility phenotypes from the CD1492 mutant (15). In addition, rstA transcript levels are decreased in the CD1492 mutant, and CD1492 transcript levels are increased in the rstA mutant (15). Altogether, these data suggest that the activities of CD1492 and RstA are linked in the same regulatory pathway.

Here, we further probed the function of CD2492 and CD1579 in C. difficile spore formation, as well as asked whether an additional conserved histidine kinase, CD1949, influences sporulation. Our results revealed that a null CD2492 single mutant and a combined CD1492 CD2492 mutant exhibited the same high sporulation frequency and increased sporulation-specific gene expression as the CD1492 mutant, indicating that these proteins function in the same regulatory pathway. A CD1579 mutant also exhibited a high, but variable, sporulation phenotype. We demonstrate that mutating the conserved histidine residues required for phosphoryl group transfer in each of these putative histidine kinases impacts C. difficile spore formation in various ways, providing evidence that phosphoryl group transfer is important for CD1492, CD2492, and CD1579 function. Finally, we show that CD1949 does not influence C. difficile sporulation. Because the functions of CD1492, CD2492, and CD1579 influence Spo0A phosphorylation, but their phenotypes do not support their primary activities as Spo0A kinases, we propose to name the corresponding loci phosphotransfer protein A (ptpA; CD1492), phosphotransfer protein B (ptpB; CD2492), and phosphotransfer protein C (ptpC; CD1579).

RESULTS

The PtpA (CD1492), PtpB (CD2492), and PtpC (CD1579) orphan histidine kinases inhibit C. difficile spore formation.

The C. difficile 630 genome contains genes that encode five orphan histidine kinases, CD1492, CD2492, CD1579, CD1949, and CD1352, which contain conserved catalytic domains (HisKA; PFAM00512) that share similarity to Bacillus sp. sporulation-associated kinases (16, 19) (see Fig. S1 in the supplemental material). The CD1352 kinase (CprK) governs a lantibiotic-responsive transporter with no sporulation phenotype and thus was not included in this study (20). A previous study found that disruption of CD2492 resulted in decreased sporulation frequency, while in vitro studies suggested that CD1579 directly phosphorylated Spo0A (16). Our previous work implicated CD1492 as an inhibitor of sporulation, as the sporulation frequency of a CD1492 mutant was significantly greater than that of the parent strain (15). To further investigate the impact of these four orphan histidine kinases on C. difficile sporulation, we recreated the previously published CD2492 mutant. We retargeted the group II intron from pCE240 utilizing the same CD2492 targeting site used by Underwood et al. to create a CD2492 mutant (referred to as the ptpB mutant). In addition, we created a CD1492 CD2492 double mutant (referred to as the ptpA ptpB mutant) by introducing the CD2492-targeted group II intron into the CD1492 background (see Materials and Methods for details and the PCR confirmation in Fig. S2 in the supplemental material; the original CD1492 mutant is referred to as the ptpA mutant herein).

We assessed the sporulation phenotypes of the ptpA, ptpB, and ptpA ptpB mutants by enumerating ethanol-resistant spores and vegetative cells after 24 h of growth on 70:30 sporulation agar. Under these conditions, the 630 Δerm parent sporulated at a frequency of ∼15.5%. As previously observed, the ptpA mutant exhibited a high sporulation frequency of 84.4% (15) (Fig. 1A). The ptpB mutant and the ptpA ptpB double mutant exhibited the same high sporulation frequencies as the ptpA mutant (83.6% and 83.7%, respectively) (Fig. 1A), indicating that the individual genes do not have an additive impact on sporulation. These results suggest that PtpA and PtpB function in the same regulatory pathway to inhibit spore formation. These sporulation phenotypes are also apparent by phase-contrast microscopy, as more phase-bright spores were visible in the ptpA, ptpB, and ptpA ptpB mutants compared to the parent strain (Fig. 1B).

FIG 1.

PtpA (CD1492), PtpB (CD2492), and PtpC (CD1579) inhibit C. difficile spore formation. (A) Ethanol-resistant spore formation and (B) representative phase-contrast micrographs of 630 Δerm, ptpA (MC674), ptpB (MC788), ptpA ptpB (MC802), rstA (MC1118), and ptpC (MC1646) grown on 70:30 sporulation agar at H₂₄ (defined as 24-h growth on plates). Sporulation frequency is calculated as the number of ethanol-resistant spores divided by the total number of spores and vegetative cells enumerated. The white scale bar represents 1 μm. *, P ≤ 0.01 by a one-way ANOVA followed by Dunnett’s multiple-comparison test.

Based on previous results, we hypothesized that the activity of the multifunctional regulator RstA (17) may be linked to PtpA (CD1492) activity, as the gene expression profiles and sporulation phenotypes of rstA and ptpA mutants are opposite (15). Because of both this inverse correlation and the observation that the ptpB and ptpA ptpB mutants phenocopy the ptpA mutant, we included the rstA mutant in this study as a comparator. As previously observed, the rstA mutant exhibited a significantly lower sporulation frequency compared to the parent (Fig. 1A and B).

To investigate the impact of PtpC (CD1579) on C. difficile sporulation, we created a clean deletion using allele-coupled exchange with a toxin-antitoxin system as a counterselectable marker to select for plasmid excision (21). Sporulation frequency in the CD1579 mutant was variable, but ∼3.5-fold greater than the 630 Δerm parent at 54.8% (Fig. 1A and B). This result was somewhat surprising given that PtpC was previously shown to directly phosphorylate Spo0A in vitro (16); however, it is common for HisKA family proteins to possess both kinase and phosphatase activities (22). These data suggest that PtpC also inhibits C. difficile sporulation, but may not be in the primary regulatory pathway controlling Spo0A dephosphorylation under the conditions tested.

Notably, the sporulation phenotype we observed in the ptpB mutant is the opposite of previously published results (16). When Underwood et al. created the original ptpB mutant, the sporulation frequencies were calculated after 72 h of growth in brain heart infusion (BHI) broth by directly counting carbol fuchsin and malachite green-stained bright-field micrographs. No additional experiments were performed to further probe the sporulation phenotype in the ptpB mutant, nor were complementation studies performed (16). We asked whether the ptpA, ptpB, and ptpA ptpB mutants exhibit an alternative sporulation phenotype under different growth conditions. We replicated the sporulation assays performed in BHI medium; however, to quantitate sporulation efficiency, we used the standard ethanol-resistance sporulation assays to enumerate spores, and we assessed sporulation by phase-contrast microscopy. The ptpA, ptpB, and ptpA ptpB mutants all hypersporulated in BHI medium (Fig. 2A), similar to the sporulation phenotypes observed on 70:30 sporulation agar. Due to the significant amount of cell lysis observed in the phase-contrast micrographs (Fig. 2B), vegetative cells could not be accurately enumerated at this time point from BHI cultures. Thus, the sporulation frequency was counted as spores per mL of culture. These data suggest that the original ptpB (low-sporulation) phenotype observed by Underwood et al. was inconsistent with our data due to the significant cell lysis present after 72 h in BHI. Altogether, our data demonstrate that PtpA and PtpB inhibit C. difficile sporulation.

FIG 2.

The sporulation frequencies of the ptpA (CD1492), ptpB (CD2492), and ptpA ptpB (CD1492 CD2492) mutants are increased in BHI medium. (A) Ethanol-resistant spore formation and (B) representative phase-contrast micrographs of 630 Δerm, ptpA (MC674), ptpB (MC788), and ptpA ptpB (MC802) grown in BHI medium at H₇₂. The means and standard errors of the means for three biological replicates are shown. The white scale bar represents 1 μm. No statistical significance observed via a one-way ANOVA followed by Dunnett’s multiple-comparison test.

CRISPRi knockdown of CD1949 expression does not affect C. difficile spore formation.

We next asked whether the orphan histidine kinase CD1949 contributes to C. difficile sporulation. After numerous unsuccessful attempts to create a CD1949 null mutant, we utilized the CRISPR interference (CRISPRi) tool, recently adapted for C. difficile, to directly repress CD1949 transcription (23). Here, the addition of xylose to the medium induced expression of the dCas9 gene, which encodes a nuclease-deactivated version of caspase-9. dCas9 is then guided to the target transcript by a gene-specific single guide RNA (sgRNA) and subsequently blocks gene transcription. We constructed two different CD1949-specific sgRNAs and expressed these in the 630 Δerm background. A previously published scrambled sgRNA (sgRNA-neg) was included as a control (23). No difference in sporulation frequencies was observed between strains containing the sgRNA-CD1949 targets compared to the sgRNA-neg-containing strain grown on sporulation agar, with or without xylose (Fig. 3A). To ensure that CD1949 was directly targeted by our sgRNA constructs, we measured CD1949 transcripts using quantitative reverse transcription-PCR (qRT-PCR). CD1949 transcripts were decreased by ∼10-fold or ∼20-fold using sgRNA-C1949-1 or sgRNA-CD1949-3, respectively (Fig. 3B). These data suggest that CD1949 does not play a role in controlling C. difficile sporulation.

FIG 3.

CRISPRi knockdown of CD1949 gene expression does not affect sporulation frequency. (A) Ethanol-resistant spore formation at H₂₄ and (B) qRT-PCR analysis of CD1949 transcript levels at H₁₂ in 630 Δerm strains expressing either a scrambled single guide RNA (sgRNA-neg) or sgRNAs targeting CD1949 (sgRNA-CD1949-1 and -3) grown on 70:30 agar supplemented with thiamphenicol (2 μg/mL) ± 1% xylose. Sporulation frequency is calculated as the number of ethanol-resistant spores divided by the total number of spores and vegetative cells enumerated. The means and standard errors of the means for three biological replicates are shown. *, P ≤ 0.001 by a one-way ANOVA followed by Dunnett’s multiple-comparison test.

Deletion of ptpA (CD1492) and ptpB (CD2492) results in increased sporulation-specific gene expression and Spo0A activation.

To further characterize the sporulation phenotypes of the ptpA and ptpB mutants, we utilized qRT-PCR to measure transcript levels of sporulation-specific genes during the initiation of sporulation at 12 h of growth on sporulation agar (H₁₂). We examined expression of sigF, encoding the early sporulation forespore-specific sigma factor, sigE, which encodes the early mother cell-specific sigma factor, and spo0A. The ptpA, ptpB, and ptpA ptpB mutants all presented similarly increased sigF (∼2.1- to 2.4-fold) and sigE (∼1.8- to 2.1-fold) transcript levels (Fig. 4A). As previously observed, the rstA mutant had fewer sigF, sigE, and spo0A transcripts compared to the parent strain (17) (Fig. 4A). Although the ptpC mutant had a higher sporulation frequency than the 630 Δerm parent at H₂₄, sigF and spo0A transcript levels were marginally decreased at H₁₂; however, this effect was not statistically significant.

FIG 4.

Spo0A-dependent gene expression and Spo0A activation in histidine kinase protein mutants correlate with endpoint sporulation frequency. (A) qRT-PCR analyses of sigE, sigF, and spo0A transcripts, (B) anti-Spo0A Western blot (brightness adjusted) after Phos-tag gel separation of unphosphorylated and phosphorylated Spo0A (Spo0A∼P) species, and (C) quantitative analysis of the ratio of Spo0A∼P species versus total Spo0A protein in 630 Δerm, ptpA (CD1492; MC674), ptpB (CD2492; MC788), ptpA ptpB (CD1492 CD2492; MC802), rstA (MC1118), and ptpC (CD1579; MC1646) grown on 70:30 sporulation agar at H₁₂. The means and standard errors of the means for at least three biological replicates are shown. *, P ≤ 0.05 by a one-way ANOVA followed by Dunnett’s multiple-comparison test.

To determine whether Spo0A phosphorylation was affected during early sporulation in the ptpA, ptpB, and ptpC mutants, we employed Phos-tag SDS-PAGE. Here, total protein, harvested from cells after 12 h of growth on sporulation agar, was resolved by Phos-tag gel electrophoresis. The unphosphorylated (Spo0A) and phosphorylated (Spo0A∼P) forms were then detected by Western blotting with Spo0A antibody. As a control, an aliquot of 630 Δerm lysate was boiled to remove any heat-labile phosphate modifications. The protein representing the upper band is the Spo0A∼P species, as evidenced by the loss of this upper band after heating (Fig. 4B, lane 1 compared to lane 2). There was an increase in the ratio of Spo0A∼P to Spo0A in the ptpA, ptpB, ptpA ptpB, and ptpC mutants, confirming that a greater proportion of Spo0A protein was phosphorylated in these mutants, corresponding to the onset of sporulation (Fig. 4B and C). Likewise, a much lower ratio of Spo0A∼P to Spo0A was observed in the rstA mutant, also correlating with the decreased sporulation-specific gene expression and lower sporulation frequency observed in this mutant. Altogether, these data corroborate that PtpA, PtpB, PtpC, and RstA all affect Spo0A phosphorylation and, thus, early sporulation events in C. difficile.

PtpA (CD1492) and PtpB (CD2492) promote TcdA production.

Our previous work demonstrated that the ptpA mutant had an ∼2-fold decrease in tcdA transcript and TcdA protein levels compared to the 630 Δerm parent (15). We observed no change in tcdB transcript levels in the ptpA strain. To determine whether PtpB and PtpC impact toxin production, we measured tcdA, tcdB, and tcdR transcript levels in cells grown on 70:30 sporulation agar at H₁₂ using qRT-PCR. As we observed previously (15), the ptpA mutant exhibited an ∼2-fold decrease in tcdA transcript levels, but no significant change in tcdR or tcdB transcripts was observed (Fig. 5A). The ptpB and ptpA ptpB mutants mirrored the changes in toxin transcripts seen in the ptpA mutant, exhibiting an ∼2-fold decrease in tcdA transcript levels, with no effect on tcdR and tcdB transcript levels (Fig. 5A). Toxin transcript levels were not greatly impacted by the loss of ptpC, and as we previously observed, the absence of rstA resulted in significantly increased tcdR, tcdA, and tcdB transcripts, as RstA is a direct repressor of toxin gene transcription (17, 24).

FIG 5.

PtpA (CD1492) and PtpB (CD2492) promote TcdA production. (A) qRT-PCR analyses of tcdR, tcdA, and tcdB transcript levels at H₁₂ on 70:30 sporulation agar and (B) ELISA of TcdA and TcdB present in the supernatant at H₂₄ in TY medium in 630 Δerm, ptpA (MC674), ptpB (MC788), ptpA ptpB (MC802), rstA (MC1118), and ptpC (MC1646). The means and standard errors of the means from at least three biological replicates are shown. *, P ≤ 0.05 by a one-way ANOVA followed by Dunnett’s multiple-comparison test.

To further understand the impact that PtpA, PtpB, and PtpC exert on toxin production, we measured TcdA and TcdB present in the supernatants of the ptp mutants after 24 h of growth in tryptone-yeast extract (TY) medium. There was a slight decrease in total toxin production observed in the ptpA, ptpB, and ptpA ptpB mutants, but this effect was not statistically significant (Fig. 5B). Considering the qRT-PCR data, it is likely that the wild-type levels of tcdB transcription in the mutants offset the decrease in tcdA transcription. Since this enzyme-linked immunosorbent assay (ELISA) measures the presence of both toxins, the unchanged levels of TcdB in these mutants may mask the repression of TcdA production.

Similar to the variable increase in sporulation frequency in the ptpC mutant, we also observed variable concentrations of total TcdA and TcdB toxins present in the supernatant (Fig. 5B), suggesting that PtpC does not play a primary role in C. difficile toxin production. As expected, TcdA and TcdB toxins were significantly increased in the rstA mutant supernatant (17, 24) (Fig. 5B).

Although the decreases in tcdA transcripts and TcdA/TcdB toxin production in the ptp single and double mutants are not statistically significant in this study, we previously found that the decreased TcdA production in the ptpA mutant resulted in decreased virulence in the hamster model of infection (15). These data suggest that both PtpA and PtpB enhance C. difficile virulence by indirectly promoting tcdA transcription through an unknown mechanism. Further, these data comparing the single mutants to the double mutant provide additional support that PtpA and PtpB function in the same regulatory pathway to influence C. difficile physiological processes, as the toxin phenotypes in the single and double mutants are all identical.

Expression of rstA and the histidine kinases are impacted in sporulation mutants.

Our previous study suggested that the PtpA and RstA regulatory pathways are linked, as the ptpA and rstA mutants exhibit inverse gene expression and phenotype patterns (15). Because our data indicate that PtpA and PtpB function in the same regulatory pathway to control sporulation and toxin production, we asked whether the opposing activities of RstA and PtpA/PtpB impact the transcriptional patterns of each gene in their respective mutants. Interestingly, rstA transcript levels are significantly decreased (∼2- to 3-fold) in the series of ptp mutants (Table 1). In contrast, ptpA, ptpB, and ptpC transcript levels are significantly increased (∼1.5- to 2.5-fold) in the rstA mutant (Table 1). Finally, ptpA and ptpB transcript levels are decreased ∼2-fold in the ptpC mutant. These data further suggest that PtpA, PtpB, PtpC, and RstA have indirect effects on each other’s expression and that the activities of the proteins converge within a shared regulatory pathway.

TABLE 1.

Transcript levels of rstA, ptpA, ptpB, and ptpC are reciprocally altered in the histidine kinase and rstA mutants

Transcriptsa	Fold change relative to the 630 Δerm parent strain (mean ± SEM)b
	ptpA	ptpB	ptpA ptpB	ptpC	rstA
rstA	0.42 ± 0.05	0.35 ± 0.03	0.38 ± 0.04	0.65 ± 0.11	ND
ptpA	ND	0.91 ± 0.08	ND	0.47 ± 0.06	1.45 ± 0.16
ptpB	0.83 ± 0.08	0.10 ± 0.02 c	0.08 ± 0.00 c	0.56 ± 0.05	1.50 ± 0.16
ptpC	0.70 ± 0.08	0.67 ± 0.06	0.78 ± 0.09	ND	2.54 ± 0.45

^aqRT-PCR analyses of rstA, ptpA, ptpB, and ptpC transcripts in 630 Δerm, ptpA (MC674), ptpB (MC788), ptpA ptpB (MC802), ptpC (MC1646), and rstA (MC1118) strains grown on 70:30 sporulation agar at H₁₂.

^bThe means and standard errors of the means (SEM) for at least four biological replicates are shown. Values in boldface indicate statistical significance (P < 0.05) compared to the 630 Δerm strain, calculated by one-way ANOVA followed by Dunnett’s multiple-comparison test. ND, transcripts not detected.

^cSome ptpB transcripts are expected to be detected in the ptpB mutants as these mutants were constructed using TargeTron disruption (see Materials and Methods).

PtpA (CD1492) and PtpB (CD2492) are both required for repression of C. difficile sporulation, but the conserved histidine residue is not required for PtpB (CD2492) function.

To ensure that the sporulation phenotypes exhibited by the ptpA, ptpB, ptpA ptpB, and ptpC mutants were due to disruption or loss of the targeted gene, we complemented these mutants by expressing each locus under the control of its native promoter on an exogenous plasmid (pMC123). Expression of ptpA or ptpB from their native promoters restored the ptpA and ptpB single mutants’ sporulation frequencies to below wild-type levels (Fig. 6A). However, expression of ptpA in the ptpB mutant or ptpB in the ptpA mutant did not complement sporulation, further supporting that PtpA and PtpB functions are not redundant (Fig. 6A). Complementation of the CD1492 CD2492 double mutant required the expression of both ptpA and ptpB; expression of a single Ptp protein in the double mutant was not enough to exert any impact on sporulation frequency (Fig. 6A). Altogether, these data indicate that PtpA and PtpB function together in a regulatory pathway to inhibit spore formation and that their functions are not interchangeable.

FIG 6.

PtpA (CD1492) and PtpB (CD2492) are both required for repression of C. difficile sporulation, but the conserved histidine residue is not required for PtpB (CD2492) function. Ethanol-resistant spore formation in (A) 630 Δerm pMC123 (MC324), ptpA pMC123 (MC964), ptpA pptpA (MC998), ptpA pptpA-H668A (MC1812), ptpA pptpB (MC965), ptpB pMC123 (MC966), ptpB pptpB (MC967), ptpB pptpB-H664A (MC1030), ptpB pptpA (MC999), ptpA ptpB pMC123 (MC968), ptpA ptpB pptpA (MC1000), ptpA ptpB pptpB (MC969), and ptpA ptpB pptpA-ptpB (MC1396) strains and (B) 630 Δerm pMC123 (MC324), ptpC pMC123 (MC1672), ptpC pptpC (MC1673), ptpC pptpC-H372A (MC1706), ptpC pptpC-T376R (MC2051), and ptpC pptpC-N493D (MC2052) strains grown on 70:30 sporulation agar supplemented with 2 μg/mL thiamphenicol at H₂₄. Sporulation frequency is calculated as the number of ethanol-resistant spores divided by the total number of spores and vegetative cells enumerated. The means and standard errors of the means for at least three biological replicates are shown. Note the difference in scales between panels A and B. *, P ≤ 0.05; **, P ≤ 0.001 by a one-way ANOVA followed by Dunnett’s multiple-comparison test.

The autophosphorylation and phosphotransferase activities of sensor histidine kinases rely on a conserved histidine residue located in the dimerization and histidine phosphotransfer domain (DHpt) (25, 26). These conserved histidine residues are present in PtpA, PtpB, and PtpC and were proposed to be critical for phosphotransfer to an aspartyl residue in Spo0A (16). Replacing the conserved histidine residue with alanine in a histidine kinase disables the autophosphorylation and phosphotransfer activity of the protein, resulting in a nonfunctional protein (27). Our previous work showed that overexpression of ptpA-H668A in the ptpA background did not reduce sporulation (15), suggesting that the histidine residue is critical for CD1492 function in sporulation. We were able to replicate these results by expressing ptpA-H668A from its native promoter, rather than the inducible promoter used previously (Fig. 6A). Surprisingly, the corresponding ptpB-H664A allele did complement the ptpB mutant (Fig. 6A), indicating that this histidine residue is not necessary for PtpB to repress sporulation.

To confirm that the ptpC mutation was responsible for the hypersporulation phenotype observed, the ptpC gene was expressed from its native promoter on an exogenous plasmid (pMC123). Although the sporulation phenotypes were variable in the ptpC strains harboring the control and the pptpC plasmids, expression of the ptpC complemented the increased sporulation phenotype of the ptpC strain harboring the vector control (Fig. 6B). Because of the variable phenotype, we performed whole-genome sequencing on the ptpC mutant and found no additional mutations besides the replacement of the ptpC allele with the ermB cassette (data not shown).

To further elucidate the function of PtpC, we performed a series of site-directed mutations. We targeted the conserved histidine residue H372 and a highly conserved threonine residue, T376, which is located in the E/DxxT/N motif immediately adjacent to the conserved histidine residue. The conserved T/N residue is critical for phosphatase activity in bifunctional HisKA proteins, but is not required for autophosphorylation or kinase activity (22, 28). We also mutagenized N493, a conserved asparagine that is required for autophosphorylation in bifunctional histidine kinases (26, 29). The ptpC-H372A, ptpC-T376R, and ptpC-N493D site-directed mutants complemented the increased sporulation frequency observed in the ptpC mutant (Fig. 6B), although due to the high variability in sporulation frequencies in these strains, none of these effects were statistically significant. These data suggest that the potential kinase and phosphatase activity of CD1579 does not play a primary role in regulating sporulation events under the conditions tested.

Expression of the ptp site-directed mutants results in a dominant-negative phenotype.

To further probe the function of the conserved histidine residues, we expressed the ptpA, ptpB, and ptpC wild-type alleles and histidine site-directed mutations from their native promoters in the 630 Δerm background. Comparable to our previous study (15), sporulation frequency decreased when ptpA was expressed from its native promoter, compared to the parent strain containing the empty vector (from 33.3% in 630 Δerm pMC123 to 14.4% in 630 Δerm pptpA) (Fig. 7A and B). We observed a similar effect when ptpB was expressed in 630 Δerm (to 16.0% in 630 Δerm pptpB) (Fig. 7A and B), indicating that PtpA and PtpB are able to reduce sporulation in an otherwise wild-type background.

FIG 7.

ptpA (CD1492) and ptpB (CD2492) expression in the 630 Δerm background decreases sporulation frequency, while ptpA-H668A and ptpB-H664A expression results in a dominant-negative phenotype. (A) Ethanol-resistant spore formation and representative phase-contrast micrographs in panels B and C 630 Δerm pMC123 (MC324), 630 Δerm pptpA (MC2024), 630 Δerm pptpA-H668A (MC2025), 630 Δerm pptpB (MC2026), 630 Δerm pptpB-H664A (MC2027), 630 Δerm pptpC (MC2030), 630 Δerm pptpC-H372A (MC2031) grown on 70:30 sporulation agar supplemented with 2 μg/mL thiamphenicol at H₂₄. Experiments with ptpA and ptpB were performed at different times than with ptpC, and the 630 Δerm pMC123 (MC324) control strain is shown for each. Sporulation frequency is calculated as the number of ethanol-resistant spores divided by the total number of spores and vegetative cells enumerated. The means and standard errors of the means for three biological replicates are shown. The white scale bar represents 1 μm. *, P ≤ 0.05 by a one-way ANOVA followed by Dunnett’s multiple-comparison test comparing each strain to 630 Δerm pMC123 (MC324).

Since histidine kinases function as oligomers, we next hypothesized that expression of the nonfunctional ptpA-H668A allele in the parental background would result in nonfunctional hetero-oligomers. These hetero-oligomers would be unable to function as phosphatases, resulting in increased sporulation and a dominant-negative phenotype, similar to the effect observed when nonfunctional KinA variants are expressed in B. subtilis (30, 31). As predicted, sporulation frequency increased in 630 Δerm pptpA-H668A by ∼1.7-fold compared to the parent strain (Fig. 7A and B). Although the ptpB-H664A allele complemented the ptpB mutant, we also observed a dominant-negative phenotype when pptpB-H664A was expressed in 630 Δerm, as sporulation frequency increased ∼1.9-fold (Fig. 7A and B). In contrast to the complementation study (Fig. 6A), these data suggest that the conserved histidine residue of PtpB plays a role in C. difficile sporulation.

Finally, we examined the effect on sporulation when ptpC is expressed in the 630 erm background. Although not statistically significant because of variability, spore formation was increased by ∼2-fold when the wild-type ptpC allele was expressed in 630 Δerm (Fig. 7A and C). However, sporulation frequency was increased by only ∼1.3-fold when the ptpC-H372A allele was expressed, suggesting that the histidine residue impacts the ability for PtpC to influence C. difficile spore formation.

DISCUSSION

Although the morphological changes that produce a dormant spore are conserved between clostridia and the well-studied bacilli, the regulatory pathway and factors that control sporulation initiation in C. difficile are not well defined (10–12). In all spore-forming bacteria, the onset of sporulation is governed by the essential regulator of sporulation, Spo0A, which is activated by phosphorylation and inactivated by dephosphorylation (10, 32). The broadly studied Bacillus species use an expanded two-component system, known as a phosphorelay, to transfer phosphoryl groups from sporulation-associated sensor histidine kinases via two intermediate proteins to Spo0A, to trigger the onset of sporulation. An orthologous expanded sporulation regulatory pathway leading to Spo0A activation is not encoded in the C. difficile genome or other clostridia (7, 10, 12); however, several orphan sensor histidine kinases, including CD1492, CD2492, and CD1579, have been implicated in controlling sporulation initiation by influencing Spo0A phosphorylation (15, 16).

This investigation has expanded what was previously known about how CD1492 (PtpA), CD2492 (PtpB), and CD1579 (PtpC) impact C. difficile spore formation (15, 16). The data demonstrate that PtpA and PtpB function in the same regulatory pathway to control Spo0A activation, as the sporulation phenotypes and gene expression profiles of the single and double mutants are identical. However, neither protein can fulfill the role of the other, indicating that PtpA and PtpB functions are nonredundant. Further, the data suggest that these proteins function together, not necessarily stepwise, as neither protein is epistatic to the other. It is possible that PtpA and PtpB form hetero-oligomers and/or do not function as phosphatases without both proteins present. We hypothesize that PtpA and PtpB directly bind to and dephosphorylate Spo0A, although alternatively, PtpA and PtpB may interact with an intermediate factor or factors that directly phosphorylate Spo0A or serve as an endpoint in a serial dephosphorylation pathway. We attempted to assess potential direct protein-protein interactions by using tagged, full-length proteins in previously validated bacterial adenylate cyclase two hybrid (BACTH) and split luciferase assays (33, 34). However, these approaches have been unsuccessful thus far, likely because PtpA and PtpB are membrane proteins that are toxic when expressed in Escherichia coli, resulting in unstable constructs. We also unsuccessfully attempted to capture these interactions in vivo with coimmunoprecipitation assays using dually tagged full-length recombinant proteins, followed by Western blotting, and we were unsuccessful in probing the full-length Ptps’ phosphorylation statuses with Phos-tag gel. In future studies, we will employ alternative approaches, such as working with cytosolic PtpA and PtpB truncations (14, 26, 35) and performing coimmunoprecipitation studies followed by mass spectrometry analyses. Identification of the direct binding partners of the Ptp proteins is a high priority.

The role of PtpC in sporulation is less clear. A ptpC null mutant had increased, but variable, sporulation. In contradiction, overexpression of ptpC in 630 Δerm also increased sporulation frequency. However, expression of the ptpC-H372A allele had no impact on sporulation. Further, ptpC alleles containing site-directed mutations in conserved residues critical for kinase and phosphatase activity in bifunctional histidine kinases complemented the variably increased sporulation phenotype of the ptpC mutant. PtpC was previously shown to phosphorylate Spo0A in vitro, suggesting that PtpC positively controls C. difficile sporulation initiation (16). However, the contribution of the conserved PtpC histidine residue for in vitro phosphoryl group transfer to Spo0A was not tested. Considering both the previously published in vitro assays and our data presented here, it appears that PtpC has the ability to perform both kinase and phosphatase functions. However, the effects mediated by PtpC may be masked in vivo by other factors (i.e., PtpA and PtpB) that negatively regulate sporulation, PtpC may indirectly impact C. difficile sporulation through additional regulatory pathways, or PtpC may function more robustly under conditions not yet tested.

The environmental and/or intracellular signals that control the activities of PtpA, PtpB, and PtpC are unknown. These proteins may have kinase or phosphatase activity under differing conditions, similar to the well-studied EnvZ histidine kinase of Escherichia coli and Salmonella enterica (25, 26), the pH-sensing HK853 from Thermotoga maritima (36), and the quorum-sensing sensor kinases of Vibrio sp., LuxN and LuxQ (37, 38). LuxN requires a conserved aspartate residue, but not the conserved histidine residue, for phosphatase activity (38), and EnvZ retains phosphatase activity when several other residues are substituted for the histidine residue (39). Further, another C. difficile histidine kinase, CprK, has also exhibited potential phosphatase activity in the absence of its conserved histidine residue (20). Retention of phosphatase activity may explain why the PtpB-H664A mutant remained functional in complementation studies yet displayed a dominant-negative phenotype in the parent strain. We hypothesize that the histidine residue is not required for PtpB phosphatase activity, but is required for PtpA activity. Further, PtpA, PtpB, and PtpC all contain a conserved E/DxxT/N motif in which the E/D residue is critical for kinase activity and the T/N motif is necessary for phosphatase activity (22, 40). Along with our data, the presence of this conserved motif provides additional evidence that PtpA, PtpB, and PtpC may possess dual kinase and phosphatase activities. Elucidating the molecular mechanisms by which PtpA, PtpB, and PtpC control phosphate flux to Spo0A will be a focus of our future studies.

Regulation of ptpA, ptpB, and ptpC gene expression may also influence the timing of accumulation and activity of these proteins. The transition phase sigma factor SigH directly activates ptpB transcription (41), while the inactivation of sigB, a general stress response sigma factor, results in decreased ptpA expression and increased ptpC expression (42). Additionally, the catabolite control protein, CcpA, appears to indirectly repress ptpC expression in response to glucose (43). Altogether, the expansive list of global regulators that influence ptpA, ptpB, and ptpC gene expression underscores that the pathways that control Spo0A phosphorylation and dephosphorylation are under complex regulatory control. Understanding when the Ptp proteins are expressed and active may provide insight into what signals the onset of spore formation in C. difficile.

Our data demonstrate that C. difficile utilizes at least two nonredundant pathways to regulate Spo0A activation. This appears in contrast to B. subtilis, which positively controls phosphate flux from the kinases to Spo0A through the intermediate proteins Spo0F and Spo0B. To modulate the phosphate flow to Spo0A, B. subtilis employs several aspartyl phosphatases, which directly dephosphorylate Spo0F or Spo0A (44–46), and kinase inhibitors, which prevent KinA autophosphorylation and/or phosphotransfer (47, 48). C. difficile also carries orthologous aspartyl-phosphatase and kinase inhibitor genes (11), potentially providing additional regulatory mechanisms to inhibit sporulation under specific conditions, even if the precise function or target is not conserved with B. subtilis. Although the clostridia are hypothesized to have a simplified Spo0A activation pathway, C. difficile and its relatives Clostridium acetobutylicum, Acetivibrio thermocellus, and C. perfringens employ multiple, potentially dual-function, histidine kinases to control Spo0A phosphorylation (13, 14, 49, 50). These results suggest that sporulation initiation is more tightly regulated in response to environmental and intracellular cues in clostridia than previously credited.

As PtpA, PtpB, and PtpC all inhibit sporulation, one of the biggest questions remaining is what factors are primarily responsible for Spo0A activation? The multifunctional regulator RstA positively influences Spo0A phosphorylation through an unknown molecular mechanism (17, 18). Although there is no evidence that RstA directly binds Spo0A or functions as a kinase, it remains possible that RstA phosphorylates Spo0A or an intermediate or blocks Spo0A dephosphorylation by steric hindrance. Spo0A phosphorylation may also be controlled directly by unidentified kinases. These potentially unknown kinases are difficult to predict based on knowledge from well-studied systems, as there is low conservation between clostridial or Bacillus spore formers. Identifying proteins that directly interact with known sporulation factors may uncover additional regulators that impact sporulation initiation, helping to unravel the regulatory pathways and molecular mechanisms that influence the ability for C. difficile transmission and survival.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains and plasmids used for this study are listed in Table 2. Clostridioides difficile strains were routinely cultured in a 37°C anaerobic chamber (Coy) with an atmosphere of 10% H₂, 5% CO₂, and 85% N₂, as previously described (51), either in BHI-supplemented (BHIS) or tryptone-yeast extract (TY) medium at pH 7.4. C. difficile cultures were supplemented with 2 to 10 μg/mL thiamphenicol if necessary for plasmid maintenance, and overnight cultures included 0.1% taurocholate to promote spore germination and 0.2% fructose to inhibit sporulation, as indicated (52, 53). Escherichia coli strains were grown at 37°C in LB with 100 μg/mL ampicillin and/or 20 μg/mL chloramphenicol, as indicated, and 50 to 100 μg/mL kanamycin was used to counterselect against E. coli HB101/pRK24 after conjugation with C. difficile (54).

TABLE 2.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant genotype or features	Source, construction, or reference
Strains
E. coli
HB101/pRK24	F− mcrB mrr hsdS20(rB− mB−) recA13 leuB6 ara-14 proA2 lacY1 galK2 xyl-5 mtl-1 rpsL20 pRK24	B. Dupuy
C. difficile
630 Δerm	Erms derivative of strain 630	Nigel Minton (60)
MC310	630 Δerm spo0A::erm	58
MC324	630 Δerm pMC123	58
MC674	630 Δerm ΔCD1492	15
MC788	630 Δerm CD2492::erm	This study
MC802	630 Δerm ΔCD1492 CD2492::erm	55
MC964	630 Δerm ΔCD1492 pMC123	This study
MC965	630 Δerm ΔCD1492 pMC658	This study
MC966	630 Δerm CD2492::erm pMC123	This study
MC967	630 Δerm CD2492::erm pMC658	This study
MC968	630 Δerm ΔCD1492 CD2492::erm pMC123	This study
MC969	630 Δerm ΔCD1492 CD2492::erm pMC658	This study
MC998	630 Δerm ΔCD1492 pMC673	This study
MC999	630 Δerm CD2492::erm pMC673	This study
MC1000	630 Δerm ΔCD1492 CD2492::erm pMC673	This study
MC1030	630 Δerm CD2492::erm pMC683	This study
MC1396	630 Δerm ΔCD1492 CD2492::erm pMC731	This study
MC1646	630 Δerm ΔCD1579::erm	This study
MC1672	630 Δerm ΔCD1579::erm pMC123	This study
MC1673	630 Δerm ΔCD1579::erm pMC707	This study
MC1706	630 Δerm ΔCD1579::erm pMC982	This study
MC1873	630 Δerm pMC1064	This study
MC1874	630 Δerm pMC1062	This study
MC1963	630 Δerm pMC1095	This study
MC2024	630 Δerm pMC673	This study
MC2025	630 Δerm pMC1000	This study
MC2026	630 Δerm pMC658	This study
MC2027	630 Δerm pMC683	This study
MC2030	630 Δerm pMC707	This study
MC2031	630 Δerm pMC982	This study
MC2051	630 Δerm ΔCD1579::erm pMC1128	This study
MC2052	630 Δerm ΔCD1579::erm pMC1129	This study
Plasmids
pRK24	Tra+ Mob+ bla tet	61
pCR2.1	bla kan	Invitrogen
pUC19	Cloning vector; bla	62
pCE240	C. difficile TargeTron construct based on pJIR750ai (group II intron, ermB::RAM ltrA); catP	C. Ellermeier
pJIR1457	ermB oriCP oriEC oriT	63
pMSR	Pseudo-suicide plasmid used for allele exchange in C. difficile 630; Ptet-CD2571.1 catP	21
pIA33	Pxyl::dCas9-opt Pgdh::sgRNA-rfp catP	23
pMC123	E. coli-C. difficile shuttle vector; bla catP	56
pMC330	pCR2.1 with group II intron targeted to CD2492	This study
pMC333	pCE240 with CD2492-targeted intron	This study
pMC336	pMC123 with CD2492-targeted intron; erm::RAM ltrA catP	This study
pMC658	pMC123 expressing CD2492 from its native promoter	This study
pMC673	pMC123 expressing CD1492 from its native promoter	This study
pMC681	pUC19 expressing CD2492-H664A from its native promoter	This study
pMC683	pMC123 expressing CD2492-H664A from its native promoter	This study
pMC707	pMC123 expressing CD1579 from its native promoter	This study
pMC731	pMC123 expressing CD1492 and CD2492 from their native promoters	This study
pMC982	pMC123 expressing CD1579-H372A from its native promoter	This study
pMC919	pMSR with homology regions flanking CD1579 and ermB	This study
pMC1000	pMC123 expressing CD1492-H668A from its native promoter (synthesized by GenScript)	This study
pMC1062	pIA33 with sgRNA-neg	This study; 23
pMC1064	pIA33 with sgRNA-CD1949-3	This study
pMC1095	pIA33 with sgRNA-CD1949-1	This study
pMC1128	pMC123 expressing CD1579-T376R from its native promoter	This study
pMC1129	pMC123 expressing CD1579-N493D from its native promoter	This study

Strain and plasmid construction.

C. difficile 630 (GenBank accession no. NC_009089.1) was used as the template for primer design, and C. difficile 630 was used as the template for PCR amplification and mutant construction. Oligonucleotides used in this study are listed in Table 3. The 630 Δerm CD2492 mutant (ptpB; MC788) was recreated by retargeting the group II intron from pCE240 using the targeting site published by Underwood et al. in 2009 (16). Notably, the targeting site was not located in the 254A site within the CD2492 coding region noted by Underwood et al., but rather at 318S. The CD1492 CD2492 double mutant (ptpA ptpB; MC802) was constructed similarly using the 630 Δerm ΔCD1492 (ptpA; MC674) background (55). All strains were confirmed by PCR analysis (Fig. S1A).

TABLE 3.

Oligonucleotides used in this study

Primer	Sequence (5′→3′)	Use/locus tag (reference)
oMC44	5′ CTAGCTGCTCCTATGTCTCACATC	Forward primer for rpoC qPCR (56)
oMC45	5′ CCAGTCTCTCCTGGATCAACTA	Reverse primer for rpoC qPCR (56)
oMC301	5′ CAAATAATGCAGTATTTAGTCATGTG	Forward primer for screening 3′ crossover ΔCD1579
oMC304	5′ CAGCCAACGGACTCTTCTC	Reverse primer for screening 5′ crossover ΔCD1579
oMC309	5′ GGAGAATACAGAGATTTGATTGATTC	Forward primer for PCR verification of CD2492::erm
oMC317	5′ AAAAGCTTTTGCAACCCACGTCGATCGTGAAGTGATCTTAATCGTGCGCCCAGATAGGGTG	CD2492 IBS; similar to IBS_CD1A (16)
EBS1d_CD1A; oMC318	5′ CAGATTGTACAAATGTGGTGATAACAGATAAGTCTTAATCTCTAACTTACCTTTCTTTGT	CD2492 EBS1 (16)
oMC319	5′ CGCAAGTTTCTAATTTCGATTATCACTCGATAGAGGAAAGTGTCT	CD2492 EBS2; similar to EBS2_CD1A (16)
oMC331	5′ CTCAAAGCGCAATAAATCTAGGAGC	Forward primer for spo0A
oMC332	5′ TTGAGTCTCTTGAACTGGTCTAGG	Reverse primer for spo0A
oMC338	5′ TCCCATTTGCCTTTATTTGAACTTGA	Reverse primer for PCR verification of CD2492::erm
oMC339	5′ GGGCAAATATACTTCCTCCTCCAT	Forward primer for sigE qPCR (58)
oMC340	5′ TGACTTTACACTTTCATCTGTTTCTAGC	Reverse primer for sigE qPCR (58)
oMC352	5′ GGAGTAGGTTTAGCTTTGTTATTAGGAACC	Forward primer for PCR verification of ΔrstA (24)
oMC355	5′ CTGTTGGAATATCTAGGCGATAAGC	Forward primer for rstA qPCR (17)
oMC356	5′ TGGTCCTCAGCCTTGTTTAATTC	Reverse primer for rstA qPCR (17)
oMC914	5′ GCGCGGCCGCCAGCCTTGTCATTTTTTAGATTG	Reverse primer for PCR verification of ΔCD1492 (15)
oMC937	5′ GCTTTATCAGAGGCTATGAATA	Forward primer for PCR verification of ΔCD1492
oMC956 (fliCqF)	5′ TACAAGTTGGAGCAAGTTATGGAAC	Forward primer for fliC qPCR (64)
oMC957 (fliCqR)	5′ GTTGTTATACCAGCTGAAGCCATTA	Reverse primer for fliC qPCR (64)
oMC1201	5′ CGTAGTGACTGGCCGAAA	Forward primer for CD1949 qPCR
oMC1202	5′ CCCATAAACTCTATTTCCACTAGAATC	Reverse primer for CD1949 qPCR
oMC1204	5′ TTCCACAACTTGCTGTTATTTCTC	Reverse primer for PCR verification of ΔrstA (17)
oMC1481	5′ GCATGGATCCTCTAGCAGAAAGAATTGCATGATT	Forward primer for CD2492
oMC1482	5′ TAGCGCATGCCCTTATGATAGCCTATTTCTTACAACTTA	Reverse primer for CD2492
oMC1537	5′ GACTCGGATCCTCAGAGGCTATGAATAGTAAAGAAG	Forward primer for CD1492
oMC1538	5′ GATGAGCATGCACGCATCAAATACAACTAAAGTAATAAA	Reverse primer for CD1492
oMC1603	5′ GCATGGATCCAAAGATGACTATTGATAAGTAAGAGA	Forward primer for CD1579
oMC1604	5′ TAGCGCATGCAAACTTATAAATCCGAGAACTCTAT	Reverse primer for CD1579
oMC1749	5′ CCAATATAATCATGCAATTCTTTCTGCTAGAGGATCCTCAGAGGCTATGAATAGTAAAGAAG	Forward primer for CD1492 to Gibson assemble into pMC658
oMC1750	5′ CAGTCACGACGTTGTAAAACGACGGCCAGTGAATTCAACGCATCAAATACAACTAAAGTAATAAA	Reverse primer for CD1492 to Gibson assemble into pMC658
oMC1997	5′ GTAGAAATACGGTGTTTTTTGTTACCCTAAGTTTAAACTGCGCCAGGTGCTATTTT	Forward primer for CD1579 (5′) homology region for Gibson assembly
oMC1998	5′ GGATTTTGGTCATGAGATTATCAAAAAGGAGTTTAAACGTAACTTCAGACCACAGCTCC	Reverse primer for CD1579 (3′) homology region for Gibson assembly
oMC2065	5′ CTGCGCCAGGTGCTATTTTTG	Forward primer for screening 5′ crossover ΔCD1579
oMC2066	5′ CATCCCTATATAAAGGGACGAGTC	Reverse primer for screening 3′ crossover ΔCD1579
oMC2139	5′ ATAATCTCATGACCAAAATCCCTTAACGATTCTAACCACTACCTTTCAATGTTATTTA	Reverse primer for overlapping PCR with CD1579 upstream flanking region and ermB
oMC2140	5′ TAAATAACATTGAAAGGTAGTGGTTAGAATCGTTAAGGGATTTTGGTCATGAGATTAT	Forward primer for overlapping PCR with CD1579 upstream flanking region and ermB
oMC2141	5′ TTTTTAAAATTTTATTTTTTATATTTAAACCTCCTTGGAAGCTGTCAGTAGTATACCT	Reverse primer for overlapping PCR with CD1579 downstream flanking region and ermB
oMC2142	5′ AGGTATACTACTGACAGCTTCCAAGGAGGTTTAAATATAAAAAATAAAATTTTAAAAA	Forward primer for overlapping PCR with CD1579 downstream flanking region and ermB
oMC2498	5′ CATTGAAAGGTAGTGGTTAGAATATGGATACCCATAATAAATATGTAAATTTT	Forward primer for CD1579-H372A site-directed mutagenesis
oMC2499	5′ AAAATTTACATATTTATTATGGGTATCCATATTCTAACCACTACCTTTCAATG	Reverse primer for CD1579-H372A site-directed mutagenesis
oMC2785	5′ AATTAAACTGTAAATGGCCATACTATTCAGAAACCAAATGGTTTTAGAGCTAGAAATAGC	Forward primer for sgRNA-CD1949-1 (targeting sequence underlined)
oMC2787	5′ AATTAAACTGTAAATGGCCAAGAAAATACCTATTACTGTCGTTTTAGAGCTAGAAATAGC	Forward primer for sgRNA-CD1949-3 (targeting sequence underlined)
oMC3062	5′ GAAGATGATATTAAATTTATAGGTCTTCTAAGTTCATGGGATAAATTTGCAAAAAATTCCATC	Forward primer for CD1579-T376R mutagenesis
oMC3063	5′ GATGGAATTTTTTGCAAATTTATCCCATGAACTTAGAAGACCTATAAATTTAATATCATCTTC	Reverse primer for CD1579-T376R mutagenesis
oMC3064	5′ TTTTCCATCTTTCTTATTGTACTTAATTCCATCAGAAAGTAAATTTAATATTATTCTTTCCAATTTTTCT	Forward primer for CD1579-N493D mutagenesis
oMC3065	5′ AGAAAAATTGGAAAGAATAATATTAAATTTACTTTCTGATGGAATTAAGTACAATAAGAAAGATGGAAAA	Reverse primer for CD1579-N493D mutagenesis
4084	5′ AACTTATAGGATCCGCGGCCGCTAGTCAGACATCATGCTGATCTAGA	Reverse primer for sgRNA amplification (23)
4238	5′ AATTAAACTGTAAATGGCCAAGACCGCTAAACTGAAAGTTGTTTTAGAGCTAGAAATAGC	Forward primer for sgRNA-neg amplification (targeting sequence underlined) (23)

The CD1579 (ptpC) mutant was created using the pseudo-suicide allele-coupled exchange (ACE) vector as previously described (21), with some modifications. A pMSR-derived vector, pMC919, containing ∼740 bp of the upstream CD1579 homology arm and ∼500 bp of the downstream CD1579 homology arm, flanking an ermB cassette, was conjugated into 630 Δerm using 15 μg/mL thiamphenicol for plasmid selection and 100 μg/mL kanamycin for counterselection of E. coli. Faster-growing colonies were streaked onto BHIS supplemented with 10 μg/mL thiamphenicol and screened by PCR for upstream or downstream crossover events. Positive colonies were grown in 10 mL BHIS with 100 ng/mL anhydrotetracycline (ATc) and 5 μg/mL erythromycin to induce expression of the CD2517.1 toxin and cure the plasmid. After 24 h of growth, 2 μL of this culture was streaked on BHIS agar supplemented with 5 μg/mL erythromycin, and colonies were PCR verified for homologous recombination and thiamphenicol sensitivity (Fig. S1B).

The CD1492-H668A (pMC1000) and CD2492-H664A (pMC681) alleles were synthesized and cloned into pMC123 and pUC19, respectively, by GenScript (Piscataway, NJ). The Benchling CRISPR Guide RNA Design tool was used to create an sgRNA targeting CD1949 (23). The details of vector construction are given in the supplemental material (see Fig. S3).

Sporulation assays and phase-contrast microscopy.

C. difficile strains were grown overnight in BHIS supplemented with 0.1% taurocholate, to promote spore germination, and 0.2% fructose, to inhibit sporulation. In the morning, cells were diluted slightly into BHIS medium and grown until mid-exponential phase (defined as an optical density at 600 nm [OD₆₀₀] of approximately 0.5). Sporulation was examined on 70:30 agar plates or from BHI broth, independently. Aliquots of 0.25 mL were either spread as a lawn onto 70:30 agar supplemented with 2 μg/mL thiamphenicol (53) or diluted 1:10 into BHI broth. Ethanol-resistant sporulation assays were performed after 24 h of growth (H₂₄) on 70:30 agar or after 3 days of growth in BHI broth (H₇₂), as previously described (15). Cells were either collected from 70:30 agar and suspended in BHIS medium to an OD₆₀₀ of approximately 1.0 or taken directly from BHI broth. Vegetative cell counts were determined by immediately serially diluting and plating suspended cells onto BHIS. At the same time, ethanol-resistant spore numbers were ascertained by mixing a 0.5-mL aliquot of resuspended cells with 0.3 mL ethanol and 0.2 mL distilled water (dH₂O) to a final concentration of 28.5% ethanol. This mixture was vortexed and incubated for 15 min to eliminate all vegetative cells. Ethanol-treated cells were serially diluted in 1×phosphate-buffered saline (PBS) containing 0.1% taurocholate and plated onto BHIS with 0.1% taurocholate. CFU were enumerated after at least 36 h of growth, and the sporulation frequency was calculated as the total number of spores divided by the total number of spores and vegetative cells. An spo0A mutant (MC310) was used as a negative control for sporulation and vegetative cell death. The results represent the means and standard error of the means from at least three independent biological replicates. Statistical significance was determined using a one-way analysis of variance (ANOVA), followed by Dunnett’s multiple-comparison test (GraphPad Prism v8.3). Phase-contrast microscopy was performed at H₂₄ or H₇₂, using the resuspended cells, with a Ph3 oil immersion objective on a Nikon Eclipse Ci-L microscope, and at least two fields of view were captured with a DS-Fi2 camera from at least three independent experiments.

Quantitative reverse transcription-PCR analysis.

C. difficile cells were cultured on 70:30 agar as a lawn as described above. Cells were collected at H₁₂, suspended in 6 mL 1:1:2 ethanol-acetone-water solution, and stored at −80°C. RNA was isolated and subsequently DNase I treated (Ambion) as previously described (56–58). cDNA was synthesized (Bioline) using random hexamers (58). Quantitative real-time reverse transcription-PCR (qRT-PCR) analysis was performed in triplicate on 50 ng cDNA using the SensiFAST SYBR & Fluorescein kit (Bioline) and a Roche Lightcycler 96. The results were calculated by the comparative cycle threshold method (59), were normalized to the rpoC transcript, and represent the means and standard errors of the means from at least three independent biological replicates. Statistical significance was determined using a one-way ANOVA, followed by a Dunnett’s multiple-comparison test (GraphPad Prism v8.3).

Enzyme-linked immunosorbent assay.

Quantification of TcdA and TcdB toxin present in culture supernatants was performed on C. difficile cultures grown in TY medium (pH 7.4) at H₂₄, as previously described (18). Briefly, cultures were pelleted, and the supernatants, diluted with the provided dilution buffer, were assayed in technical duplicates using the tgcBIOMICS kit for simultaneous detection of C. difficile toxins A and B, according to the manufacturer’s instructions. The averaged results were normalized to the OD₆₀₀ of each respective culture at H₂₄, and the results are provided as the means and standard errors of the means from three independent biological replicates. Statistical significance was performed using a one-way ANOVA, followed by a Dunnett’s multiple-comparison test (GraphPad Prism v8.3).

Phos-tag gel electrophoresis and Western blotting.

C. difficile strains were grown on 70:30 sporulation agar and harvested at H₁₂. Lysates were prepared as previously described (15); however, 0.1% phosphatase inhibitor cocktail II (Sigma) was included in the lysis buffer to prevent global protein dephosphorylation. Total protein from lysates was quantitated using the Pierce Micro BCA (bicinchoninic acid) protein assay kit (Thermo Scientific). Prior to gel electrophoresis, lysates were prepared at 4°C, with the exception of an additional 630 Δerm aliquot that was briefly heated to 99°C before loading to remove any heat-labile phosphoryl groups. Approximately 3 μg of total protein was separated by electrophoresis on a precast 12.5% Super-Sep Phos-tag SDS-PAGE gel (Fujifilm Wako Chemicals, Inc., USA) at 90 V for 3.5 h at 4°C. Protein was transferred to 0.2-μm-pore nitrocellulose membrane in transfer buffer containing 10% methanol and 0.04% SDS. Western blot analysis was performed using mouse anti-Spo0A (53) as the primary antibody and goat anti-mouse conjugated with Alexa 488 (Invitrogen) as the secondary antibody. Imaging and densitometry were performed with a ChemiDoc and Image Lab 6.0 software (Bio-Rad), respectively, for three independent experiments. To calculate the ratio of phosphorylated Spo0A for each strain, the adjusted total band volume of phosphorylated Spo0A was divided by the sum of the phosphorylated and unphosphorylated Spo0A species adjusted total band volumes (Spo0A∼P/Spo0A plus Spo0A∼P).