c-di-GMP Inhibits Early Sporulation in Clostridioides difficile

ABSTRACT The formation of dormant spores is essential for the anaerobic pathogen Clostridioides difficile to survive outside the host gastrointestinal tract. The regulatory pathways and environmental signals that initiate C. difficile spore formation within the host are not well understood. One second-messenger signaling molecule, cyclic diguanylate (c-di-GMP), modulates several physiological processes important for C. difficile pathogenesis and colonization, but the impact of c-di-GMP on sporulation is unknown. In this study, we investigated the contribution of c-di-GMP to C. difficile sporulation. The overexpression of a gene encoding a diguanylate cyclase, dccA, decreased the sporulation frequency and early sporulation gene transcription in both the epidemic R20291 and historical 630Δerm strains. The expression of a dccA allele encoding a catalytically inactive DccA that is unable to synthesize c-di-GMP no longer inhibited sporulation, indicating that the accumulation of intracellular c-di-GMP reduces C. difficile sporulation. A null mutation in dccA slightly increased sporulation in R20291 and slightly decreased sporulation in 630Δerm, suggesting that DccA contributes to the intracellular pool of c-di-GMP in a strain-dependent manner. However, these data were highly variable, underscoring the complex regulation involved in modulating intracellular c-di-GMP concentrations. Finally, the overexpression of dccA in known sporulation mutants revealed that c-di-GMP is likely signaling through an unidentified regulatory pathway to control early sporulation events in C. difficile. c-di-GMP-dependent regulation of C. difficile sporulation may represent an unexplored avenue of potential environmental and intracellular signaling that contributes to the complex regulation of sporulation initiation. IMPORTANCE Many bacterial organisms utilize the small signaling molecule cyclic diguanylate (c-di-GMP) to regulate important physiological processes, including motility, toxin production, biofilm formation, and colonization. c-di-GMP inhibits motility and toxin production and promotes biofilm formation and colonization in the anaerobic, gastrointestinal pathogen Clostridioides difficile. However, the impact of c-di-GMP on C. difficile spore formation, a critical step in this pathogen’s life cycle, is unknown. Here, we demonstrate that c-di-GMP negatively impacts sporulation in two clinically relevant C. difficile strains, the epidemic strain R20291 and the historical strain 630Δerm. The pathway through which c-di-GMP controls sporulation was investigated, and our results suggest that c-di-GMP is likely signaling through an unidentified regulatory pathway to control C. difficile sporulation. This work implicates c-di-GMP metabolism as a mechanism to integrate environmental and intracellular cues through c-di-GMP levels to influence C. difficile sporulation.

the nisin-inducible cpr promoter (13,33,34) in two different C. difficile backgrounds. We included R20291, which is a clinically prevalent epidemic strain, and 630Derm, which is a spontaneous erythromycin-sensitive derivative of 630, a clinical isolate that has served as a long-term laboratory model strain (35,36). Of note, the amino acid sequences of DccA from R20291 and 630Derm are 100% identical. To assess sporulation frequency, we performed ethanol-resistant sporulation assays in these strains after 24 h of growth (H 24 ) on 70:30 sporulation agar.
When a plasmid copy of dccA (pDccA) was overexpressed from the nisin-inducible promoter in the R20291 background, the sporulation frequency significantly decreased from 33.7% in the absence of nisin to 19% in the presence of 0.5 mg/ml nisin (Fig. 1A), suggesting that the overexpression of dccA and, presumably, high intracellular levels of c-di-GMP inhibit C. difficile sporulation. To assess whether the decrease in the sporulation frequency was due to the production of c-di-GMP by DccA, we overexpressed a dccA allele encoding a mutated GGDEF domain (AADEF) that is unable to synthesize cdi-GMP (pDccA mut ) (13). Here, even at the highest expression level of dccA (0.5 mg/ml nisin), the sporulation efficiency remained unaffected, indicating that the DccA-dependent reduction in the sporulation frequency is due to DccA's diguanylate cyclase activity. We also visualized sporulation in these strains using phase-contrast microscopy, in which spores appear phase bright and vegetative cells are phase-dark rods. representative phase-contrast micrographs (B) of R20291 pMC211 (RT526), R20291 pDccA (RT527), and R20291 pDccA mut (RT539) grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol and 0 to 0.5 mg/ml nisin, as indicated, at H 24 . (C and D) Ethanol-resistant sporulation assays (C) and representative phase-contrast micrographs (D) of 630Derm pMC211 (RT762), 630Derm pDccA (RT763), and 630Derm pDccA mut (RT764) grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol and 0 to 0.5 mg/ml nisin, as indicated, at H 24 . The means and standard errors of the means from at least three biological replicates are shown. n.s., not significant; *, P , 0.05; **, P , 0.01; ***, P , 0.001; ****, P , 0.0001 (by one-way analysis of variance [ANOVA] with Dunnett's posttest comparing values to that of uninduced WT pDccA). c-di-GMP Negatively Impacts C. difficile Sporulation When dccA was overexpressed in R20291, not only were fewer spores visible, but this strain also formed long chains ( Fig. 1B; discussed below). The reduced sporulation frequency and cell morphology phenotypes are dependent on increased c-di-GMP concentrations as these effects were not seen in the R20291 strain overexpressing pDccA mut (data not shown).
When dccA was overexpressed in the 630Derm background, we observed an ;4fold decrease in the sporulation efficiency (29.7% in the absence of nisin to 8.5% with 0.5 mg/ml nisin) (Fig. 1C). This reduction in sporulation frequency in the 630Derm pDccA strain was dose dependent, with the lowest sporulation frequency occurring at the highest expression level of dccA (0.5 mg/ml nisin). Again, the overexpression of the dccA mut allele resulted in wild-type (WT) levels of sporulation, indicating that the ability of DccA to repress sporulation relies on a functional GGDEF domain and its diguanylate cyclase activity. Similar to R20291, the overexpression of dccA in the 630Derm background resulted in fewer spores and a change in cell morphology when observed by phase-contrast microscopy, and these phenotypes were also dependent on a functional DccA GGDEF domain ( Fig. 1D and data not shown).
As noted above, the overexpression of dccA resulted in dramatic cell morphology changes in R20291 and 630Derm that are dependent on increased intracellular concentrations of c-di-GMP ( Fig. 1B and D). This change in cell morphology with dccA overexpression was expected and is likely due to the c-di-GMP-dependent increase in the expression of cmrRST, which encodes an atypical signal transduction system that regulates C. difficile colony morphology and motility (37,38). CmrT promotes bundled cell chaining. To determine whether CmrT is responsible for the formation of long chains at high c-di-GMP concentrations, we overexpressed dccA in the R20291 cmrT background. As predicted, the elongated cells observed at high c-di-GMP concentrations are no longer present in the cmrT mutant ( Fig. 2A and B), indicating that c-di-GMP promotes cell chaining through the activation of cmrT expression. CmrT is not involved in the c-di-GMP-dependent inhibition of sporulation, as the sporulation frequency of the cmrT mutant is decreased ;2-fold, similar to the R20291 parent strain, when dccA is overexpressed (Fig. 2C).
after 12 h of growth on 70:30 sporulation agar, which marks early stationary phase and the onset of sporulation (39). As expected, dccA and dccA mut transcript levels were increased ;4to 6-fold in both R20291 and 630Derm in the presence of 0.5 mg/ml nisin ( Fig. 3A and B), showing that dccA and dccA mut expression levels are consistent in both backgrounds. To better understand the effect of c-di-GMP on early sporulation events in C. difficile, we measured the steady-state transcript levels of an early sporulation-specific gene, sigE (spoIIG), using qRT-PCR. The transcription of sigE is dependent on active, phosphorylated Spo0A (40). The relative expression levels of sigE in both R20291 pDccA and 630Derm pDccA grown on 0.5 mg/ml nisin were decreased ;2-fold compared to the levels in their respective parent strains ( Fig. 3C and D). The decreased sigE transcript levels in R20291 pDccA were dependent upon a functional GGDEF domain FIG 3 Overexpression of dccA decreases Spo0A-dependent gene expression. (A and C) qRT-PCR analyses of dccA (A) and sigE (C) transcript levels in R20291 pMC211 (RT526), R20291 pDccA (RT527), and R20291 pDccA mut (RT539) grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol and 0 to 0.5 mg/ml nisin, as indicated, at H 12 . (B and D) qRT-PCR analyses of dccA (B) and sigE (D) transcript levels in 630Derm pMC211 (RT762) and 630Derm pDccA (RT763) grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol in the absence or presence of 0.5 mg/ml nisin at H 12 . The means and standard errors of the means from at least three biological replicates are shown. *, P , 0.05 (by one-way ANOVA with Dunnett's posttest comparing values to that of the uninduced WT vector). (E) Anti-Spo0A Western blot analysis of 630Derm pMC211 (RT762), 630Derm pDccA (RT763), 630Derm pDccA mut (RT764), R20291 pMC211 (RT526), R20291 pDccA (RT527), and R20291 pDccA mut (RT539) grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol and 0.5 mg/ml nisin at H 12 .
c-di-GMP Negatively Impacts C. difficile Sporulation as sigE transcript levels were unchanged when dccA mut was overexpressed in R20291 (Fig. 3C).
To further assess how c-di-GMP impacts early sporulation events, we measured the transcript levels of additional early-sporulation-specific genes, including spo0A, murG (a Spo0A-dependent gene), sigF (encodes the early sporulation-specific sigma factor that is expressed in the forespore compartment), gpr (a SigF-dependent gene), and spoIID (a SigE-dependent gene). The transcript levels of early sporulation genes were minimally affected in the R20291 background; however, with the exception of spo0A, there was a significant decrease in early sporulation gene expression when dccA was overexpressed in the 630Derm background (Table 1). These results may reflect the stronger impact on sporulation that dccA overexpression has on 630Derm than on R20291. Mirroring spo0A transcript levels in R20291 and 630Derm, Spo0A protein levels were unchanged regardless of dccA overexpression (Fig. 3E). The lack of changes in spo0A transcript and protein levels is expected given that Spo0A activity is controlled by posttranslational phosphorylation. Furthermore, Spo0A autoregulation has not been observed in other C. difficile sporulation mutants, including the hypersporulating opp app and oligosporogenous rstA mutants (33,39), despite affecting Spo0A activity and Spo0A-dependent gene expression. These data indicate that the overexpression of dccA reduces early sporulation gene expression and confirm that the increased production of c-di-GMP is responsible for reduced sporulation in C. difficile.
Overexpression of dccA increases c-di-GMP-dependent gene expression in a dose-dependent manner on 70:30 sporulation agar. We previously utilized high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) to measure the intracellular concentration of c-di-GMP when dccA is overexpressed from the nisin-inducible promoter (13). To assess the relative increase in intracellular c-di-GMP with dccA induction during growth on 70:30 sporulation agar, we employed the previously described reporter controlled by the regulatory region of the pilA1 locus: the pilA1 promoter (P pilA ) and the 59 untranslated region (UTR) containing a c-di-GMP riboswitch (41). c-di-GMP directly and positively regulates pilA transcription via the c-di-GMP-responsive riboswitch located in the pilA 59 UTR (16). This P pilA1 -UTR-gusA reporter was introduced into 630Derm with chromosomal nisin-inducible dccA (41). As a control, we also utilized a P pilA1 -UTR A70G -gusA reporter where a single nucleotide substitution renders the riboswitch unresponsive to c-di-GMP (16,41). We measured P pilA1 -UTR-gusA reporter activity after growth on 70:30 sporulation agar in the presence of no nisin or 0.1 mg/ml or 0.5 mg/ml nisin at H 12 and H 15 . These time points represent early stationary phase, at the point where the sporulation regulatory cascade is initiated and sporulation-specific gene expression is active. Without induction, basal expression from the chromosomally encoded P cpr -dccA construct resulted in a slight increase (;1.2-fold) in b-glucuronidase activity compared to that of the parent strain at both time points (Fig. 4A). We observed ;1.4-fold and ;1.6-fold increases in b-glucuronidase activity in the presence of 0.1 mg/ml and 0.5 mg/ml nisin, respectively, compared to the parental control. As expected, the c-di-GMP-blind P pilA1 -UTR A70G -gusA reporter exhibited significantly reduced and constitutive activity under all tested conditions (Fig. 4A).
Because dccA is overexpressed from the chromosome in these reporter assays and not from a plasmid as we did in our previous experiments, we assessed the impact on sporulation frequency when dccA expression was induced from the chromosome. The sporulation frequency of 630Derm::P cprA -dccA grown on 70:30 sporulation agar plates supplemented with 0.5 mg/ml nisin was ;1.44-fold decreased compared to the 630Derm parent (Fig. 4B). As anticipated, the decrease in the sporulation efficiency was muted compared to dccA overexpression on the plasmid, likely due to differences in copy numbers. However, increasing the intracellular c-di-GMP concentration through dccA overexpression impacts both c-di-GMP-specific gene regulation and sporulation frequency. Altogether, these data confirm that c-di-GMP levels are modestly induced to physiologically relevant concentrations that affect c-di-GMP-dependent physiological processes under these conditions.
Deletion of dccA results in variable sporulation frequencies in R20291 and 630Derm. Sporulation may be impacted by one or a subset of c-di-GMP metabolic enzymes. C. difficile 630 encodes 37 proteins containing GGDEF and/or EAL domains, and R20291 encodes 31 (10,11,35). The overexpression of dccA bypasses the endogenous control of c-di-GMP. To better understand the c-di-GMP regulatory mechanism, we next asked whether a null mutation in dccA alone affects the C. difficile sporulation frequency. To test this hypothesis, we employed the pseudo-suicide vectors pMSR and pMSR0 (tailored for the 630 and R20291 backgrounds, respectively), which take advantage of allelic exchange and an inducible C. difficile toxin-antitoxin system to create a markerless gene deletion (42).
The R20291 DdccA mutant produced slightly more spores than R20291 (42.3% in R20291 DdccA versus 31.7% in the isogenic parent) (Fig. 5A). However, the sporulation c-di-GMP Negatively Impacts C. difficile Sporulation frequencies were highly variable, and the differences between strains were not statistically significant. The R20291 DdccA mutant was complemented by integrating the dccA gene into the chromosome using the conjugative transposon Tn916. dccA is the second gene in a two-gene operon in R20291 and 630Derm, and the promoter of the upstream gene, CD1421, was used to drive dccA transcription in the complementation construct. The sporulation frequency of R20291 DdccA Tn916::P CD1421 -dccA was reduced (34.7%) compared to that of the dccA mutant, but again, the data were variable and not statistically significant.
The 630Derm DdccA mutant exhibited a slightly reduced sporulation frequency compared to the 630Derm parent (19.1% in the dccA mutant versus 25.6% in the parent strain) (Fig. 5B), an opposite trend compared to that of the R20291 DdccA mutant. The complemented strain 630Derm DdccA Tn916::P CD1421 -dccA showed an increased sporulation frequency (34.4%). But, as in the R20291 background, the sporulation frequencies were variable and did not achieve statistical significance. Altogether, these data suggest that c-di-GMP synthesis by DccA does not greatly and/or consistently contribute to sporulation initiation under the conditions tested or that other c-di-GMP metabolic enzymes are redundant with or compensate for the loss of DccA.
c-di-GMP likely does not signal through known C. difficile sporulation factors. To attempt to identify the regulatory pathway(s) through which c-di-GMP influences sporulation in C. difficile, we overexpressed dccA in several well-studied 630Derm sporulation mutants and performed ethanol-resistant sporulation assays after 24 h of growth on 70:30 sporulation agar. We included the parent 630Derm strain overexpressing dccA in the absence and presence of 0.5 mg/ml nisin as a reference in these experiments (Fig. 6A).
First, we assessed the effect of dccA overexpression on 630Derm codY and 630Derm opp app mutants. CodY is a global transcriptional regulator that binds to target DNA at high intracellular concentrations of the effectors GTP and branched-chain amino acids (BCAAs) (43). The loss of GTP and BCAA binding changes the conformation of CodY, the differential expression of metabolic pathways, and other physiological processes, including toxin production and sporulation (29,31,44,45). The opp and app operons each encode oligopeptide permeases that import small, heterogeneous peptides into the cell (46,47). The inactivation of these permeases in C. difficile results in significantly increased sporulation, suggesting that limited nutrient uptake triggers sporulation (33). The regulatory pathways and molecular mechanisms by which CodY, Opp, and App affect sporulation are not fully understood, although null mutations in these loci affect sporulation timing and result in increased sporulation frequencies (31,33). The sporulation frequencies of the codY mutant and the opp app double mutant decreased when dccA was overexpressed using 0.5 mg/ml nisin (2-fold and 3.7-fold, respectively, compared to each strain's vector controls grown in 0.5 mg/ml nisin) ( Fig. 6B and C).
These results indicate that c-di-GMP does not require CodY or the Opp and App oligopeptide permeases to inhibit C. difficile sporulation.
Next, we ascertained whether RstA or the phosphotransfer proteins CD1492 and CD2492 are part of the regulatory pathway that c-di-GMP employs to control sporulation. RstA is a multifunctional protein that serves as a transcriptional regulator and, through a separate domain, positively influences C. difficile sporulation initiation via an unknown mechanism (39). The overexpression of dccA in the rstA background resulted in an ;13-fold-decreased sporulation frequency compared to that of the rstA mutant containing the vector control (Fig. 6D), indicating that RstA is not necessary for c-di-GMP-dependent inhibition of sporulation. CD1492 and CD2492 are predicted histidine kinases that function as phosphotransfer proteins to repress sporulation and are c-di-GMP Negatively Impacts C. difficile Sporulation hypothesized to directly impact Spo0A phosphorylation (28,48). A CD1492 CD2492 double mutant exhibited significantly increased sporulation compared to the 630Derm parent ( Fig. 6A and E), and the overexpression of dccA in the CD1492 CD2492 background reduced the sporulation frequency ;3-fold (Fig. 5E), demonstrating that CD1492 and CD2492 are not a required part of the c-di-GMP signaling pathway.
We also asked whether c-di-GMP signals through SigD to control C. difficile sporulation. SigD is the flagellar alternative sigma factor that is required for both motility and efficient toxin production in C. difficile (14,39,49). Although a null sigD mutation does not affect sporulation under these conditions (comparing the 630Derm and 630Derm sigD vector control strains) (Fig. 6A and F) (39,49), we chose to investigate the sigD mutant because c-di-GMP directly represses sigD transcription, inhibiting C. difficile motility and toxin production (13,15,50). The sporulation frequency was reduced only 1.9-fold when dccA was overexpressed in the sigD background, and the difference was not statistically significant (Fig. 6F). Furthermore, in the 630E background, where the phase-variable switch that controls the transcription of the flagellar operon is locked off, resulting in low sigD expression levels (51)(52)(53), the overexpression of dccA also significantly decreased sporulation by ;20-fold (Fig. 6G). The DccA-dependent effect was not as dramatic in the 630Derm sigD background as in the other mutant backgrounds tested; however, these data altogether suggest that c-di-GMP does not primarily influence sporulation through SigD.
Finally, we tested whether c-di-GMP affects sporulation frequency through the catabolite control protein CcpA. As a transcriptional regulator, CcpA responds to glycolytic carbohydrate availability to regulate carbon and nitrogen metabolism as well as other physiological processes, including sporulation and toxin production (30,54). The sporulation frequency was again significantly decreased by ;20-fold when dccA was overexpressed in the 630E ccpA mutant (Fig. 6H), indicating that CcpA is not involved in c-di-GMP signaling to control sporulation initiation. Altogether, these data indicate that c-di-GMP does not significantly impact C. difficile sporulation through these known regulatory factors under the conditions tested.

DISCUSSION
In this work, we set out to determine the impact of the bacterial second messenger c-di-GMP on C. difficile sporulation. We found that the overexpression of dccA, encoding a DGC that synthesizes high intracellular levels of c-di-GMP when overexpressed (13), resulted in significant decreases in early sporulation gene expression and spore formation in the epidemic R20291 and historical 630Derm strains. The conserved catalytic GGDEF motif was required for DccA-dependent inhibition of sporulation, indicating that the diguanylate cyclase synthase activity of DccA is responsible for decreased sporulation. Consistent with this result, sporulation was inhibited in a dose-dependent manner, as higher transcript levels of dccA coincided with fewer detected transcripts of sigE, which encodes an early sporulation-specific sigma factor.
Because DccA overexpression drastically altered the sporulation frequency in R20291 and 630Derm, we had anticipated a stronger sporulation phenotype in the corresponding dccA mutants. However, there are many additional DGCs encoded in the C. difficile genome as well as many PDEs, and these likely contribute to the intracellular concentration of c-di-GMP also, as most are catalytically active (10). Given the sheer number of encoded DGCs and that their functions may inherently exhibit redundancy, it is, in retrospect, unsurprising that the deletion of a single GGDEF domain protein does not significantly impact sporulation. Any contribution of DccA to the intracellular c-di-GMP pool may be masked by the redundant functions of similar proteins. As such, we previously found that the overexpression of DccA resulted in modest increases in the transcript levels of several encoded PDEs, suggesting that C. difficile is able to somewhat compensate for high levels of intracellular c-di-GMP (55). It is also possible that changes in c-di-GMP levels upon the loss of DccA are compensated for by the upregulation or activation of other DGCs, the downregulation or inhibition of PDEs, or both. Importantly, previous studies have shown that distinct DGCs control different cdi-GMP-regulated phenotypes, suggesting that localized pools of c-di-GMP influence only a subset of phenotypes (56,57). It is plausible that the deletion of one or more of the other encoded DGCs in C. difficile could result in a stronger impact on the sporulation frequency. A recent report describes that the overexpression of a phosphodiesterase containing an EAL domain increases sporulation and that the deletion of that PDE decreases sporulation in C. difficile UK1, an epidemic strain that is nearly identical to R20291, corroborating our findings (58).
Interestingly, two early C. difficile sporulation regulators that are orthologous to the B. subtilis SinR transcriptional repressor regulate C. difficile intracellular c-di-GMP levels. Null mutations in the two C. difficile SinR orthologs, known as SinRR9 in R20291 and CD2214-CD2215 in 630Derm, result in increased dccA transcription and c-di-GMP levels and an asporogenous phenotype (20,59). Furthermore, the deletion of CD2214-CD2215 resulted in the differential expression of additional DGC and PDE genes in C. difficile 630Derm (20). It will be intriguing to determine if the effect of SinRR9 on C. difficile sporulation occurs through alteration of c-di-GMP levels by regulating DGC/PDE gene expression or if SinRR9 affects sporulation initiation through multiple regulatory pathways.
Identification of the c-di-GMP effector(s) that mediates c-di-GMP-dependent sporulation regulation in C. difficile is of great interest. A variety of c-di-GMP receptors that directly bind to c-di-GMP have been characterized in bacteria. These encompass a number of protein-based receptors, including proteins containing degenerate GGDEF and/or EAL domains, and two distinct types of riboswitches that alter downstream gene expression in response to c-di-GMP binding (8). C. difficile encodes a single PilZ domain protein (6), a domain that often directly binds c-di-GMP (60,61), and a type IV pilus PilB ATPase similar to orthologs that have been shown to bind c-di-GMP (62), but there are no other known or predicted c-di-GMP protein receptors reported in C. difficile (8). C. difficile encodes at least 11 functional riboswitches that alter gene expression in response to c-di-GMP and contains 5 additional predicted riboswitches (15,55). None of these riboswitches appear to affect the expression of sporulation-related genes; however, the conditions used in this study do not support efficient sporulation initiation in C. difficile (55). Performing transcriptomics under conditions that favor sporulation may provide insights into which regulatory pathway(s) or factor(s) is required to mediate this c-di-GMP-dependent response.
The regulatory pathway that c-di-GMP utilizes to influence sporulation remains unknown. The decrease in spore formation mediated by dccA overexpression does not appear to signal through CodY, CcpA, the Opp or App oligopeptide permeases, RstA, or the CD1492 and CD2492 phosphotransfer proteins. Although the effect of DccAmediated inhibition of sporulation was slightly decreased in the sigD mutant, c-di-GMP is unlikely to signal solely through SigD under these conditions. Given that we know that c-di-GMP directly affects sigD transcription in C. difficile through the Cdi-1-3 riboswitch (13,15,50), it is attractive to hypothesize that SigD might have a role in sporulation. It may be possible that decreased levels of SigD are necessary for c-di-GMP to affect sporulation; in this case, testing the c-di-GMP-dependent effects on sporulation in a sigD mutant or in JIR8094, a phase-off, nonmotile strain with low SigD levels, may not directly answer this question. Thus far, there is no published evidence of a regulatory role for SigD in sporulation (39,49,55).
The impact of c-di-GMP on sporulation has been investigated in only a few other spore-forming bacteria. Interestingly, using an mCherry reporter fused to a c-di-GMPregulated riboswitch, high c-di-GMP levels were observed in sporulating B. subtilis cells, suggesting a correlation (63), but the impact of c-di-GMP on B. subtilis sporulation has remained relatively unexplored. Studies in Bacillus thuringiensis and Bacillus anthracis found no effect on the sporulation efficiency when any of the catalytically active GGDEF/EAL/HD-GYP-encoding genes were individually deleted (64,65). These studies in Bacillus sp. further underscore the differences between C. difficile and other endospore-forming bacteria in their early sporulation regulatory networks. Finally, direct c-di-GMP regulation of sporulation has been identified only in Streptomyces sp., where c-di-GMP inhibits spore formation directly by antagonizing the sporulation-specific sigma factor WhiG and binding directly to the transcriptional regulator BldD (66,67). The contribution of c-di-GMP to sporulation remains an understudied field.
Utilizing c-di-GMP to inhibit sporulation may be advantageous to C. difficile, as c-di-GMP can be rapidly degraded when environmental and intracellular conditions favor sporulation. The c-di-GMP metabolic activity of a protein is often controlled by the corresponding domains encoded within the same protein. Identifying the DGCs and PDEs that affect C. difficile sporulation and investigating the function of their associated domains may reveal the environmental and intracellular signals that promote or delay sporulation. Finally, the finding that c-di-GMP is a regulator of early sporulation events in C. difficile creates new research opportunities for discovering the potentially novel regulatory pathways, c-di-GMP effectors, and molecular mechanisms that control spore formation in this significant pathogen.

MATERIALS AND METHODS
Bacterial strains and growth conditions. The bacterial strains and plasmids used for this study are listed in Table 2. Clostridioides difficile was routinely cultured in brain heart infusion-supplemented (BHIS) medium in a 37°C vinyl anaerobic chamber (Coy) with an atmosphere of 5% CO 2 , 10% H 2 , and 85% N 2 (68). C. difficile cultures were supplemented with 2 to 10 mg/ml thiamphenicol as necessary for plasmid maintenance. Escherichia coli strains were grown aerobically at 37°C in LB with 100 mg/ml ampicillin and/or 10 to 20 mg/ml chloramphenicol, and counterselection against E. coli after conjugation with C. difficile was performed using 100 mg/ml kanamycin (13).
Strain and plasmid construction. C. difficile strains 630 (GenBank accession no. NC_009089.1) and R20291 (GenBank accession no. FN545816.1) were used as the basis for primer design and PCR amplification (oligonucleotides used in this study are listed in Table 3). The dccA mutants were constructed using the pseudo-suicide vectors pMSR and pMSR0 (42). Upstream and downstream homology regions were amplified from the 630Derm genome with primer pairs R2928/R2929 and R2930/R2931, respectively. The fragments were Gibson assembled (New England BioLabs [NEB]) into SalI/XhoI-digested pMSR to create pMSR::DdccA. Similar fragments were amplified from the R20291 chromosome using the same primers and assembled into pMSR0 to create pMSR0::DdccA. Chloramphenicol-resistant colonies were confirmed by PCR with plasmid-specific primer pair R838/R1832 (pMSR) or R2743/R2744 (pMSR0).
To create the dccA mutants, the pMSR::DdccA and pMSR0::DdccA plasmids were transformed into E. coli HB101 pRK24 for conjugation with C. difficile 630Derm and R20291, respectively. Subsequent steps were done essentially as previously described (42). Briefly, large thiamphenicol-and kanamycin-resistant colonies, which presumably had integrated the plasmid into the chromosome to allow for more rapid growth, were streaked again on BHIS medium with 10 mg/ml thiamphenicol and 100 mg/ml kanamycin to ensure purity. Next, large colonies were streaked onto BHIS medium with 100 ng/ml anhydrotetracycline (ATc) to induce the expression of the toxin gene and eliminate bacteria that still contained the vector. Colonies were screened for the 0.8-kb deletion of dccA using primer pair R2926/R2927 (Fig. 7A). Genomic DNA was isolated from potential mutants, and the dccA region was PCR amplified using primer pair R2926/R2927 and sequenced to confirm the integrity of the sequence.
The 630Derm and R20291 dccA mutants were complemented using a Bacillus subtilis donor strain, BS49, carrying the conjugative transposon Tn916 to transfer the dccA gene driven by its native promoter, which is encoded upstream of CD1421 (MC1959). To create the plasmid carrying the Tn916::P CD1421 -dccA construct (pMC1094), P CD1421 and dccA were spliced by overlapping PCR using primer pairs oMC2910/ 2911 and oMC2912/2913 and Gibson assembled into BamHI/SphI-digested pSMB47. Erythromycin-resistant colonies were confirmed by PCR with primer pair CD1420qF/CD1420qR (Fig. 7B).
To create the 630Derm DCD1492 CD2492::erm double mutant, the Targetron-based group II intron from pCE240 was retargeted using the targeting site reported previously by Underwood et al. to create pMC336 (69). Briefly, the CD2492-targeted intron was amplified using primers oMC317, oMC318, oMC319, and EBSu and TA cloned into pCR2.1 to create pMC330. A BsrGI/HindIII-digested fragment containing the CD2492-targeted intron was subcloned into pCE240 to create pMC333. Finally, an SphI/SfoIdigested fragment containing the CD2492-targeted intron was subcloned into pMC123 to create pMC336. The resulting construct, pMC336, was conjugated into the 630Derm CD1492 strain (MC674), and erythromycin-resistant mutants were screened for the 2-kb Targetron insertion within CD2492 using primer pair oMC309/338 (Fig. 7C). Notably, the targeting site was not located in the 254a site within the CD2492 coding region noted by Underwood et al. (69) but rather was located at 318s (data not shown).
Sporulation assays. C. difficile strains were grown overnight in BHIS medium supplemented with 0.1% taurocholate and 0.2% fructose to aid in spore germination and prevent spore formation, respectively (70,71). Thiamphenicol (5 mg/ml) was included for plasmid maintenance when necessary. When strains reached mid-exponential phase (optical density at 600 nm [OD 600 ] of ;0.5), 250-ml aliquots were applied to the surface of 70:30 agar containing 2 mg/ml thiamphenicol and 0 to 0.5 mg/ml nisin (71). After 24 h of growth, ethanol-resistant sporulation assays were performed as previously described (48,72). Briefly, cells were scraped from the plate surface and suspended in BHIS medium to an OD 600 of ;1. To enumerate Phase-contrast microscopy. Phase-contrast microscopy was performed at H 24 as described previously (33). Briefly, cells were concentrated by pelleting 0.5 ml of the cell suspension, and the concentrated cell suspension was applied to a 0.7% agarose pad on a slide. Cells were imaged with a 100Â Ph3  Quantitative reverse transcription-PCR analysis. RNA was isolated from C. difficile strains grown on 70:30 sporulation agar at H 12 and DNase I treated as previously described (13,29,33,34). cDNA was synthesized using random hexamers, and quantitative real-time-PCRs were performed in triplicate (Bioline) and monitored using a Roche LightCycler 96 system. The rpoC transcript (primer pair oMC44/ oMC45) was used as the reference gene (34). Controls with no reverse transcriptase were included for all templates and primer sets. The data were analyzed by the 2 2DDCT method (73), with normalization to rpoC and the indicated reference condition or strain. The results represent the means and standard errors of the means from at least three independent experiments.
Western blotting. The indicated strains were grown on 70:30 sporulation agar supplemented with 2 mg/ml thiamphenicol and 0.5 mg/ml nisin and harvested at H 12 . Total protein from the cell lysates was quantitated using the Pierce Micro bicinchoninic acid (BCA) protein assay kit (Thermo Scientific), 2.5 mg of total protein was separated by electrophoresis on a precast 4 to 20% TGX stain-free gradient gel (Bio-Rad), and total protein was imaged using a ChemiDoc system (Bio-Rad). Protein was transferred to a 0.45-mm nitrocellulose membrane, and Western blot analysis was performed with mouse anti-Spo0A (71) as the primary antibody and goat anti-mouse conjugated with Alexa Fluor 488 (Invitrogen) as the secondary antibody. Imaging and densitometry were performed with a ChemiDoc system and Image Lab software (Bio-Rad), respectively, for three independent experiments. b-Glucuronidase reporter assays. b-Glucuronidase assays were performed as previously detailed (41,74). Briefly, strains were grown on 70:30 sporulation agar as indicated above and harvested at H 12 and H 15 by scraping the plates and suspending the cells in BHIS medium to an OD 600 of ;0.5 to 0.7. Two 1-ml aliquots were pelleted and stored at 220°C overnight. The pellets were suspended in 0.8 ml Z buffer and 0.05 ml 0.01% SDS. The samples were vortexed, incubated at 37°C for 5 min, and then chilled on ice for 5 min. After a 1-min incubation at 37°C to warm the samples up to room temperature, the enzymatic reaction was started with the addition of 100 ml of 40 mg/ml p-nitrophenol-b-D-glucuronide and stopped with 0.4 ml 1 M Na 2 CO 3 . Cell debris was pelleted, and the A 420 and A 550 were measured using a BioTek Synergy H1 plate reader. Specific activity was normalized by the OD 600 . Three independent biological replicates were used to calculate the means and standard errors of the means.

ACKNOWLEDGMENTS
We are grateful to the members of the Tamayo and McBride labs for helpful suggestions and discussions during the course of this work. This research was supported by the U.S. National Institutes of Health through research grants AI107029 and AI143638 to R.T. and AI116933 and AI156052 to S.M.M. C.L.W. is supported by an Institutional Research and Academic Career Development Award (IRACDA) fellowship under K12-GM000678. The content of the manuscript is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institutes of Health. c-di-GMP Negatively Impacts C. difficile Sporulation