The small acid-soluble proteins of Clostridioides difficile are important for UV resistance and serve as a check point for sporulation

Clostridioides difficile is a nosocomial pathogen which causes severe diarrhea and colonic inflammation. C. difficile causes disease in susceptible patients when endospores germinate into the toxin-producing vegetative form. The action of these toxins results in diarrhea and the spread of spores into the hospital and healthcare environments. Thus, the destruction of spores is imperative to prevent disease transmission between patients. However, spores are resilient and survive extreme temperatures, chemical exposure, and UV treatment. This makes their elimination from the environment difficult and perpetuates their spread between patients. In the model spore-forming organism, Bacillus subtilis, the small acid-soluble proteins (SASPs) contribute to these resistances. The SASPs are a family of small proteins found in all endospore-forming organisms, C. difficile included. Although these proteins have high sequence similarity between organisms, the role(s) of the proteins differ. Here, we investigated the role of the main α/β SASPs, SspA and SspB, and two annotated putative SASPs, CDR20291_1130 and CDR20291_3080, in protecting C. difficile spores from environmental insults. We found that SspA is necessary for conferring spore UV resistance, SspB minorly contributes, and the annotated putative SASPs do not contribute to UV resistance. In addition, the SASPs minorly contribute to the resistance of nitrous acid. Surprisingly, the combined deletion of sspA and sspB prevented spore formation. Overall, our data indicate that UV resistance of C. difficile spores is dependent on SspA and that SspA and SspB regulate/serve as a checkpoint for spore formation, a previously unreported function of SASPs.

Introduction to the main α/β type SASPs. All 4 contain the conserved "EIA" sequence that is cleaved by the germination protease (GPR) upon germination (Fig 1) [47][48][49]. Even with strong sequence similarity/identity, SASPs have varying roles between spore-forming organisms. To investigate the roles of C. difficile SspA, SspB, CDR20291_1130, and CDR20291_3080 in UV resistance, outgrowth, and chemical resistance, single gene deletions and pairwise deletions were generated using CRISPR-Cas9 genome editing [50]. Total amounts of spores produced and the sporulation frequency by the single deletion strains, C. difficile ΔsspA, C. difficile ΔCDR20291_1130, and C. difficile ΔCDR20291_3080 were indistinguishable from the wildtype C. difficile R20291 parental strain using heat or ethanol based sporulation assays (S1 and S2 Figs, respectively). Surprisingly, the C. difficile ΔsspB strain did not make spores and will be discussed later. These results indicate that the single deletions of C. difficile sspA, C. difficile CDR20291_1130 and C. difficile CDR20291_3080 do not impact C. difficile sporulation.

sspA is required for spore UV resistance
Because of the strong phenotype of B. subtilis SASPs in UV resistance, we hypothesized that C. difficile SspA, and/or the two annotated SASPs (CDR20291_1130 and CDR20291_3080) may function similarly. The viability of SASP mutants after a 10 minute exposure to 302 nm UV light was tested. Viable spores were quantified by colony formation on media supplemented with taurocholate (a C. difficile spore germinant) and then compared to spores derived from the wildtype parental strain [51]. Exposure of the C. difficile ΔsspA strain to UV light resulted in a~1,000x decrease in spore survival. Spores derived from the C. difficile ΔCDR20291_1130 and C. difficile ΔCDR20291_3080 mutants had a statistically significant difference in UV resistance compared to spores derived from the wildtype strain, but this difference is likely not biologically relevant (Fig 2A). These results indicate that C. difficile SspA is the most important SASP for UV resistance while the genes annotated as putative SASPs are not involved in UV resistance. In B. subtilis, SASP binding to DNA helps to protect from UV by encouraging the formation of spore photoproducts (SP) instead of cyclobutane thymine dimers [17,52]. The spore photoproduct lyase, SPL, repairs the SP [43][44][45]53]. To understand the impact of SPL on UV resistance in C. difficile, we engineered a deletion in the C. difficile spl gene. Spores derived from C. difficile Δspl strain have approximately 10x reduced survival compared to spores derived from the wildtype C. difficile R20291 strain (Fig 2A). Unsurprisingly, the C. difficile ΔsspA Δspl double mutant remains at the level of sspA mutant alone (0.1% of wildtype), suggesting that SPL function depends on SspA.
To determine if the putative SASPs have redundant roles in UV resistance, we generated deletions of all pairwise combinations of C. difficile CDR20291_1130, C. difficile CDR20291_3080, and C. difficile sspA. After 10 minutes of UV exposure, spores derived from  C. difficile sspA is necessary for UV protection. Spores were exposed to 302 nm UV light for 10 minutes. After the C. difficile ΔCDR20291_1130 ΔCDR20291_3080 were as resistant as wildtype spores, indicating that these annotated SASPs are not compensating for each other during UV exposure ( Fig 2B). Spores derived from the C. difficile ΔsspA ΔCDR20291_1130 mutant had no further reduction in survival compared to the C. difficile ΔsspA mutant alone. However, spores derived from the C. difficile ΔsspA ΔCDR20291_3080 double mutant and the ΔsspA ΔCDR20291_1130 ΔCDR20291_3080 triple mutant were not as sensitive to UV light as the C. difficile ΔsspA or the C. difficile ΔsspA ΔCDR20291_1130 strains (Fig 2B).
The sspA promoter is necessary for complementation of C. difficile ΔsspA UV resistance To understand the extent of protection that C. difficile SspA provides against UV damage, we quantified the viability of spores exposed to UV over time. Spores derived from the C. difficile ΔsspA strain showed a 10% loss of viability after 2.5 minutes of UV exposure. This loss increased to approximately 3 log 10 after 15 minutes of exposure. Expression of sspA from a plasmid under the control of its native promoter restored viability to the C. difficile ΔsspA strain (Fig 3A).
To further understand the role of C. difficile SspA in UV resistance, we tested the impact of different sspA expression constructs on spore survival. The sspA complement consisting of the sspA gene under its native promoter region resulted in restoration of spore viability. A 6x-histidine tag inserted on the C-terminus of the sspA gene also resulted in restoration of spore viability upon UV exposure ( Fig 3B). In B. subtilis, SASP genes can cross-complement a SASP mutant [35]. Therefore, plasmids were constructed that consisted of C. difficile sspB, CDR20291_1130, or the CDR20291_3080 genes, driven by their native promoter regions, introduced into the C. difficile ΔsspA mutant strain. After 10 minutes of exposure to UV light, spores derived with these plasmid constructs revealed that C. difficile sspB, CDR20291_1130, or CDR20291_3080 were unable to restore the UV resistance to spores derived from the ΔsspA mutant strain (Fig 3B). To determine if this is an issue with differences in expression, the promoter regions were changed and these genes were again tested for their ability to complement the C. difficile ΔsspA strain. When the sspA gene was placed under control of the sspB promoter region, complementation no longer occurred. This further supported our hypothesis that the SASPs have different expression levels. Swapping the promoter regions of C. difficile sspB, CDR20291_1130, or CDR20291_3080 complementation plasmids for the C. difficile sspA promoter region resulted in a restoration to approximately 5% of wildtype levels ( Fig 3C). However, as negative controls, the sspA promoter region alone, or the sspA promoter driving the gene encoding mCherry, could not complement the UV phenotype. These results suggest that despite what is observed for cross-complementation in other organisms, C. difficile sspB, CDR20291_1130, or CDR20291_3080 cannot fully complement the C. difficile ΔsspA strain phenotype, and to provide any complementation, they must be expressed from the sspA promoter.

C. difficile SASPs have redundant functions during outgrowth
C. difficile CDR20291_1130 and CDR20291_3080 do not contribute to spore UV resistance, but it is possible that their main role is to protect against other harsh environmental conditions the average from three independent experiments and the standard error from the mean. Statistical analysis was performed by one way ANOVA with Dunnett's multiple comparisons test in comparison to wildtype. � p < 0.01, �� p < 0.0001. https://doi.org/10.1371/journal.ppat.1009516.g002 Complementation of the C. difficile ΔsspA strain. A) Spores derived from wildtype C. difficile R20291 pEV •, the C. difficile ΔsspA pEV deletion strain □, and the C. difficile ΔsspA psspA strain ■ were exposed to UV for varying times and the viability assessed by normalizing to T 0 . B) Spores from the C. difficile ΔsspA mutant strain containing genes under control of the sspA promoter were exposed to 302 nm UV light for 10 minutes. After exposure, the spores were serially diluted and plated onto rich medium supplemented with germinants. The CFUs were enumerated and PLOS PATHOGENS C. difficile SspA and SspB contribute to sporulation or to serve as amino acid reservoirs (like γ-type SASPs of B. subtilis) during outgrowth of a vegetative cell from the germinated spore [23]. To determine if these annotated SASPs contribute to outgrowth, the OD 600 of spores derived from wildtype and mutant strains were analyzed over 12 hours in complex medium supplemented with germinants [51,54]. No difference in the outgrowth of vegetative cells was observed for the mutant strains, compared to the parental strain.
It is possible that the complex medium masked the hypothesized phenotype due to the sheer abundance of nutrients in the medium eliminating the need for the amino acids derived from the SASPs. Instead, we tested if outgrowth of a spore in minimal medium (CDMM) would be influenced in these mutant strains. Unfortunately, using a minimal medium resulted in an extreme delay in outgrowth that was not possible to practically measure. Therefore, spore outgrowth was analyzed in half-strength complex medium. In addition, to eliminate the possibility of extra resources being packaged into the spore when grown on a complex medium, and thus reducing the need for SASPs during outgrowth, spores were generated on minimal medium. Again, despite spore production on minimal medium and using halfstrength complex medium during the assay, there were no differences between the outgrowth of wildtype spores or spores derived from the C. difficile ΔsspA or the C. difficile ΔCDR20291_1130 ΔCDR20291_3080 double mutant strain (Fig 4A).
Due to the possibility that the SASPs can compensate for the deletion of one, a triple mutant was generated of sspA, CDR20291_1130, and CDR20291_3080. These spores were also generated on minimal medium and outgrowth analyzed in half-strength complex media. Spores derived from the C. difficile ΔsspA ΔCDR20291_1130 ΔCDR20291_3080 triple mutant had an approximate 2-hour delay in outgrowth compared to the wildtype strain ( Fig 4B). To eliminate the possibility that the difference in outgrowth is due to a germination defect in this triple mutant, we monitored germination by OD 600 in buffer supplemented with germinants [51,54]. During the very early events of endospore germination, the dormant, phase-bright, spore transitions to a phase-dark, germinated spore. Mutant spores germinated similarly to wildtype, suggesting no defect in germination ( Fig 4C). These results suggest that C. difficile SspA, CDR20291_1130, and CDR20291_3080 could be used as a nutrient/amino acid source during outgrowth of a germinated spore.

The C. difficile SASPs do not contribute to chemical resistance
To further characterize the role of these proteins in the spore, spores were exposed to various chemicals. Spores derived from the single deletion of C. difficile sspA, CDR20291_1130, and CDR20291_3080 and their complements, when necessary, were exposed to chemicals for 1 minute, 5 minutes, 10 minutes, and 30 minutes and spore viability was assessed by plating onto rich medium supplemented with germinant. Colony forming units were compared to T 0 and then this ratio compared to the ratio of wildtype survival. Spores exposed to 3% H 2 0 2 , 75% EtOH, 0.25% glutaraldehyde, 1 M HCl, 0.05% hypochlorite, and 2.5% formaldehyde did not exhibit reduced viability in comparison to spores derived from the wildtype strain (S3A- S3F  Fig). The mutant strains did have a slight reduction in viability after 5 or 10 minutes of exposure to 250 mM nitrous acid (S3G Fig). then normalized to T 0 CFUs. The ratio of the complemented C. difficile ΔsspA mutant strains were normalized to the ratio of the wildtype strain. C) Spores from the C. difficile ΔsspA mutant strain containing genes under the control of the sspA or sspB promoter regions were exposed to UV for 10 minutes. pEV indicates an empty vector. All data represents the average of three independent experiments and the standard error from the mean. Statistical analysis was performed by ANOVA with Dunnett's multiple comparison. A) two way, in comparison to wildtype, B) and C) one way, in comparison to C. difficile ΔsspA pEV. � p < 0.05, �� p < 0.0001. https://doi.org/10.1371/journal.ppat.1009516.g003

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation

C. difficile spore formation is influenced by SASPs
When generating the C. difficile ΔsspB mutant strain, we found that spores derived from this strain were phase-gray and were not released from the mother cell, compared to the wildtype strain (Fig 5A and 5B). Surprisingly, whole genome resequencing of the C. difficile ΔsspB strain revealed an additional, single nucleotide mutation in sspA. This mutation resulted in an to germinants taurocholic acid and glycine in a buffered solution. The OD 600 was normalized to T 0 . R20291 pEV • and triple mutant spores of C. difficile ΔsspA ΔCDR20291_1130 ΔCDR20291_3080 pEV ▼. pEV indicates an empty vector. All data points represent the average from three independent experiments and the standard error from the mean. Statistical analysis by two way ANOVA with Dunnett's multiple comparisons test. A) C. difficile ΔsspA, p < 0.001 at 11 minutes. B) C. difficile ΔsspA ΔCDR20291_1130 ΔCDR20291_3080 p < 0.05 at 6.5, 6, 8, 9 and 10.5 hours. p < 0.01 at 7.5 hours.
https://doi.org/10.1371/journal.ppat.1009516.g004 were fixed in 4% formaldehyde and 2% glutaraldehyde before imaging on a Leice DM6B microscope. I) DPA content was determined by boiling spore solutions and analyzing approximately 2.5 x 10 7 spores by terbium florescence. J) Western blot of spores exposed to 10 mM taurocholic acid in a rich medium for 1 hour and blotted against SleC. Arrow indicates processing of pro-SleC (above) to the active SleC form. K) Spores were exposed to 302 nm UV light for 10 minutes before serial dilution and plating on rich medium supplemented with germination. All values were normalized to T 0 and then those ratios were normalized to the ratio of wildtype survival. pEV indicates an empty vector. All data points represent the average from three independent experiments and the standard error from the mean. I) Statistical analysis by one way ANOVA with Dunnetts multiple comparisons test. K) Statistical analysis by unpaired t test. � p < 0.05, �� p < 0.01, ��� p < 0.001. https://doi.org/10.1371/journal.ppat.1009516.g005

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation sspA G52V allele, in addition to the sspB deletion (referred to as ΔsspB sspAG52V ). This phenotype could be complemented by expression of the sspB gene under the control of the sspB promoter region (Fig 5C) or the sspA gene under control of the sspA promoter region (Fig 5D). However, only a σ G -controlled promoter was able to complement the phenotype, other σ-factor controlled promoters could not complement (Fig 5E-5H). The rate of sporulation of the C. difficile ΔsspB mutant was also 5 log 10 less than the wildtype strain by heat sporulation assay and reduced by more than 4 log 10 with an ethanol based assay (S4 and S2 Figs). This was complemented by expression of sspB from a plasmid (S4 and S2 Figs).
Due to the phenotype of this strain, normal spore purification processes were unsuccessful. To encourage release of the immature spores from the mother cells, cultures from sporulating cells were incubated with lysozyme before resuming normal spore purification steps. This encouraged the spores to release from the mother cells and the recovery of a limited number of phase-gray spores. The phase-bright phenotype of wildtype spores is partially attributed to the dipicolinic acid (DPA) content of the spore core. Approximately 2.5 x 10 7 spores of the C. difficile wildtype, ΔsspB sspAG52V , and ΔsspB sspAG52V psspB strains were boiled and the DPA levels determined by Tb 3+ fluorescence [55]. The DPA content of the C. difficile ΔsspB sspAG52V strain was significantly lower than wildtype content and this was able to be complemented by expression of the sspB gene from its native promoter region ( Fig 5I).
Because germination assays rely heavily on the release of DPA during germination (phasebright to phase-dark transition), we were unable to use these assays to determine if this deletion and SNP combination altered germination capabilities [55,56]. Instead, the spores were exposed to germinants and the processing of spore proteins during germination were analyzed by immunoblot. In response to germinants, spores derived from the C. difficile ΔsspB sspAG52V strain processed proSleC to its active form, indicating that they are capable of receiving the germinant signals ( Fig 5J) [57][58][59].
We also tested the ability of spores derived from the C. difficile ΔsspB sspAG52V psspB strain to resist UV damage. The C. difficile ΔsspB sspAG52V alone was not assessed due to the difficulty in producing and purifying spores. In the assay, the C. difficile ΔsspB sspAG52V psspB strain was complemented to wildtype levels, revealing that the sspA G52V allele is not impaired in its ability to protect against UV damage but SspA may regulate C. difficile sporulation with SspB ( Fig 5K).

C. difficile sspA and sspB are required for spore formation
Due to the G52V mutation in sspA, we created a clean C. difficile ΔsspA ΔsspB strain to determine if the double mutant had the same phenotype as the C. difficile ΔsspB sspAG52V strain. Similar to above, the C. difficile ΔsspA ΔsspB strain produced phase-gray, immature spores ( Fig  6A). The rate of sporulation was also approximately 1,000x less than wildtype and could be complemented with expression of both sspA and sspB from a plasmid (S4 and S2 Figs). Moreover, analysis of the DPA content also showed little DPA in the double mutant strain, in comparison to wildtype ( Fig 6B). Finally, to evaluate whether these spores can germinate, we monitored SleC activation. These double mutant spores also processed proSleC to the active SleC form, showing that they are still able to germinate ( Fig 6C) [57][58][59]. Again, due to the difficulty in purifying this strain and despite density gradient purification, spore preparations still contained debris making it difficult to quantify the spores by microscopy. Therefore, we were forced to use approximate spore counts. The Coomassie-stained SDS PAGE gel shows that more protein is present in the double mutant sample than in wildtype or the complemented strains, even though the SleC bands detected by western indicate similar concentrations (S5  formaldehyde and 2% glutaraldehyde before imaging on a Leica DM6B microscope. B) DPA content was determined by boiling spore solutions and analyzing approximately 2.5 x 10 7 spores by terbium florescence. C) Western blot of spores exposed to 10 mM taurocholic acid (TA) in a rich medium for 1 hour and blotted against SleC. Arrow indicates processing of pro-SleC (above) to the active SleC form. pEV indicates an empty vector. D) Sporulation frequency was calculated by dividing the number of spores by total number of cells (spores + vegetative cells). Spore totals were determined by treating cultures with a final concentration of 28.5% ethanol and plating on BHIS supplemented with TA. Untreated cultures were plated onto BHIS without germinant to determine the number of vegetative cells. C. difficile Δspo0A was set to the limit of detection to determine sporulation frequency. Data represents the average of three independent experiments and the standard error from the mean. Statistical analysis was performed by one way ANOVA with Dunnett's multiple comparison test. � p < 0.01, �� p < 0.001. https://doi.org/10.1371/journal.ppat.1009516.g006

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation mutant preparation because of the difficulty in purifying or the spore counts could have been underestimated.
To understand how the sspA G52V allele affects sporulation, a plasmid containing the sspA promoter region and the sspA G52V allele was introduced into the C. difficile ΔsspA, C. difficile ΔsspB, or the C. difficile ΔsspA ΔsspB double mutant strains. As expected, expression of the sspA G52V allele in the C. difficile ΔsspA mutant strain led to a nearly 4 log 10 reduction in spore formation. Surprisingly, expression of sspA G52V in the C. difficile ΔsspB strain (which encodes a wildtype chromosomal sspA allele) caused a >10x reduction in sporulation. Finally, sporulation frequency in the C. difficile ΔsspA ΔsspB double deletion, which already has a sporulation defect, was further decreased upon expression of the sspA G52V allele (Fig 6D). These data further confirm that the sspA significantly contributes to the sporulation phenotype.

C. difficile sspB plays a minor role in UV resistance and does not play a role in outgrowth or chemical resistance
After discovering the SNP in the C. difficile ΔsspB strain, we generated a clean sspB deletion that produces phase-bright, released spores. The C. difficile ΔsspB strain had a sporulation frequency identical to wildtype (S4 and S2 Figs).
Since a clean sspB deletion produces spores, the UV resistance could be determined. Spores derived from the wildtype C. difficile R20291, C. difficile ΔsspB mutant, and C. difficile ΔsspB psspB strains were exposed to UV light for 10 minutes before plating on media with germinants to determine CFUs. Spores derived from the C. difficile ΔsspB strain had approximately 10% of the resistance observed for the wildtype strain. This phenotype was complemented by expression of sspB from its native promoter in trans ( Fig 7A). Therefore, C. difficile SspB plays a minor role in UV resistance. C. difficile ΔsspB has a minor UV defect and no role in outgrowth. A) Spores derived from C. difficile R20291 pEV, C. difficile ΔsspB with pEV, and C. difficile ΔsspB psspB were exposed to UV light for 10 minutes. The spores were serially diluted and plated onto medium containing germinants. Strains were normalized to CFUs at T 0 . The ratios were then normalized to the ratio of wildtype survival. B) Outgrowth of spores from C. difficile R20291 pEV • and C. difficile ΔsspB pEV^was monitored at OD 600 over 12 hours. Starting OD was 0.05 into half-strength complex medium supplemented with germinants. pEV indicates an empty vector. All data represents the average of three independent experiments and the standard error from the mean. A) Statistical analysis by one way ANOVA with Dunnett's multiple comparison test. � p < 0.001. B) Statistical analysis by two way ANOVA with Dunnett's multiple comparison test, p < 0.0001 at 10 hours. https://doi.org/10.1371/journal.ppat.1009516.g007

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation Next, we tested the outgrowth of spores derived from the C. difficile ΔsspB strain. Again, the spores were generated on a minimal medium and the outgrowth was tested in half-strength complex media supplemented with germinants. Similar to our observations with the ΔsspA strain, there was no difference in the outgrowth of spores from the C. difficile ΔsspB strain in comparison to wildtype spores ( Fig 7B).

Transcript levels from C. difficile mutant strains revealed SASP influence of σ G -controlled genes
To understand at what stage SspA and SspB exert their effects on C. difficile sporulation, we analyzed the abundance of transcript during sporulation for some select sporulation-related genes. RNA was extracted from the C. difficile R20291, C. difficile ΔsspA, C. difficile ΔsspB, C. difficile ΔsspB sspAG52V , C. difficile ΔsspA ΔsspB strains, and the C. difficile Δspo0A negative control strain at 18 hours, 24 hours, and 48 hours post plating. Quantitative reverse transcriptase PCR was performed and the fold change determined between wildtype and the indicated strains. For evaluation of sspA, sspB, CDR20291_3080, spoVT, and pdaA transcripts, cDNA was generated from a 48 hour time point. A 24 hour timepoint was used to determine sleC transcripts and the 18 hour timepoint for spoIVA. Technical triplicates of three biological replicates were performed for each strain and one replicate consistently presented as an outlier while the other two replicates had similar fold changes (S6 Fig). The transcript levels for sspA, sspB, CDR20291_3080, spoIVA, and sleC did not change in the mutant strains, when compared to the wildtype strain. However, the C. difficile ΔsspB sspAG52V strain had a trend for increased abundance of of spoVT and pdaA, both σ G -controlled genes, but this did not reach statistical significance (S6 Fig).

Discussion
C. difficile infections occur in individuals with a disrupted microbiota, commonly due to the use of broad-spectrum antibiotics [1,4,5]. Vegetative C. difficile colonizes the lower GI tract and produces toxins that disrupt the epithelial lining integrity and cause colitis and diarrhea [4]. While the toxin producing vegetative cells are eliminated with antibiotic treatment, the spores can persist in the gut and become vegetative cells again, or they can shed into the environment [4,5]. This shedding allows for the transmission of disease by passing spores into an environment where they can then transfer to other individuals [4,8]. Spores can withstand environmental insults, such as UV light, and many chemicals, and persist long term in most environments [10,17,24].
The small acid-soluble proteins (SASPs) are well established to protect the spores from environmental insults (e.g., UV light or chemicals) [10,17,18,24]. Here we investigated the functions of C. difficile SASPs in response to UV light and chemicals. We found that C. difficile SspA and SspB have a role in UV resistance but none of the SASPs significantly contribute to chemical resistances. Surprisingly, we discovered that the deletion of C. difficile ΔsspA ΔsspB resulted in phase-gray, unreleased spores and we hypothesize that SspA and SspB are working together to influence sporulation, a SASP function not previously reported.
C. difficile SASPs are involved in protecting spores from UV exposure. This was not surprising since the SASPs in all other organisms have had a large role in UV resistance [28,34,37,60,61]. In C. difficile, SspA is the major contributor to UV resistance. UV protection is attributed to the binding of the SASPs to the DNA, which changes its conformation and encourages spore photoproduct mutations over cyclobutane pyrimidine dimers [18,24,38,40,43,45,52,62]. The loss of the spore photoproduct repair system, SPL, interestingly only resulted in a 10x loss of viability. This highlights the importance of SspA in protecting the genome from lethal UV irradiation. However, in a C. difficile ΔsspA Δspl double mutant, the viability remained the same as for a C. difficile ΔsspA single deletion, suggesting SspA is the major contributor to UV resistance. This also suggests that SspA primarily influences the change in DNA structure that encourages spore photoproduct formation upon UV exposure.
In B. subtilis, other SASPs than SspA and SspB are considered minor SASPs with many redundant roles [18,24,35]. Unexpectedly, whenever the double deletion C. difficile ΔsspA ΔCDR20291_3080 or the triple deletion C. difficile ΔsspA ΔCDR20291_1130 ΔCDR20291_3080 were tested, they resulted in 10% viability. We hypothesized that these deletions would result in either the same defect of 0.1% viability seen with C. difficile ΔsspA alone or that UV sensitivity would be increased upon more deletions. This increase in survival would suggest that an unknown factor may be compensating for the loss of these SASPs (there were no obvious candidates from the whole genome re-sequencing). We also found that the other 3 C. difficile SASPs, sspB, CDR20291_1130, and CDR20291_3080 are unable to complement a C. difficile ΔsspA deletion, but hypothesized that this could be due to differences in expression levels. Indeed, when the sspA promoter region was used to drive expression of the other SASPs, complementation occurred to approximately 5% of wildtype level. We hypothesize that these other SASPs can bind DNA but need higher expression to do so.
The transcript levels determined by RT-qPCR showed similar levels of sspA, sspB, CDR20291_3080, spoIVA (σ E -controlled), and sleC (σ K -controlled) between the mutant and wildtype strains. However the σ G -controlled genes, spoVT and pdaA, trended toward higher levels in the ΔsspB sspAG52V strain, though this difference was not statistically significant. But, this trend, and that expression of sspB is required to be driven by a σ G -dependent promoter (Fig 5), supports our hypothesis that SspA and SspB influence sporulation within the forespore compartment. Due to asynchronous sporulation of C. difficile cultures, we could not quantify the absolute abundance of the SASP transcripts-the variability between experiments was considerable.
Other analyses revealed that the SASPs SspA, CDR20291_1130, and CDR20291_3080 only minorly contribute to outgrowth. A triple mutation was needed before a~2 hour defect was observed. We predict that deletion of all three SASPs depletes the amino acid pool and results in a delay in protein production and outgrowth of the vegetative cell.
In other organisms, SASPs play large roles in protection from chemicals. Interestingly, C. difficile SASPs were not important for survival of the spores upon chemical exposure. Nitrous acid was the only chemical where the SASPs may play a minor role in protection. Of note, there were some reductions in spore survival of some mutants compared to wildtype across the other chemicals tested. However, in these cases complementation plasmids did not restore viability, leading to the conclusion that some other factor, besides the SASP deletion, was causing the reduced viability. For evaluating the effect of hypochlorite and glutaraldehyde, reproducibility was an issue. Every trial varied in response to these chemicals, making it difficult to draw a confident conclusion from these data. We conclude that C. difficile SASPs do not contribute to chemical resistance. Indeed, the C. difficile spore coat is established to contain proteins involved in some resistances [63,64].
In B. subtilis, the ΔsspA ΔsspB strain is commonly used to characterize the SASPs. It was very surprising that the C. difficile ΔsspB with the sspA G52V allele (ΔsspB sspAG52V ) and the C. difficile ΔsspA ΔsspB double mutant were unable to form phase-bright spores; instead, they formed phase-gray spores lacking DPA. The DPA phenotype could be caused by various

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation reasons such as DPA not being produced by the mother cell, the SpoVA proteins not being produced and therefore not packaging DPA into the spores, or the cortex/other surrounding spore layers not being produced and enclosing the DPA into the spore. Because DPA contributes to heat resistance and heat resistance is a classically tested feature of endospores and used in sporulation assays, this could have impacted the results of the sporulation assay by killing fully formed spores that do not package DPA and thus exacerbating the sporulation frequency phenotype [10,56,[65][66][67]. However, due to the mutant strains withstanding ethanol treatment in our chemical assays, we performed another sporulation assay using ethanol resistance. Indeed, using ethanol resistance as a metric for spore formation, we observed similar levels of a sporulation defect for the C. difficile ΔsspB sspAG52V and the C. difficile ΔsspA ΔsspB double mutant strains (compared to the heat based assay). Interestingly, the double mutant is fully complemented with only the sspA gene, however, the sspB gene only restores 10x the amount of sporulation from the mutant alone. This could suggest that sspB requires sspA in order to complete its role in sporulation.
Due to the phenotype, it was very difficult to purify spores from C. difficile ΔsspB sspAG52V and C. difficile ΔsspA ΔsspB. The spores were treated with lysozyme to digest the mother cell peptidoglycan and release the phase-gray spore. These spores are then less dense, further complicating the purification process. Because of these difficulties, the spore amounts used in DPA content assays and immunoblots are rough estimates based on microscope counts. The Coomassie stained gel shows that more protein was present in the C. difficile ΔsspA ΔsspB sample than in the wildtype and the complemented samples even though the immunoblots show similar levels of protein present. This discrepancy could be due to vegetative cells/debris still present after purification through a density gradient or it is possible we underestimated the number of spores in the microscope counts.
Furthermore, evaluation of UV viability on the C. difficile ΔsspB sspAG52V strain complemented with sspB shows that these spores have approximately 60% of wildtype viability. Thus, the sspA G52V allele does not impair the ability of SspA to protect against UV damage which indicates that DNA binding still occurs. Also, expression of the sspA G52V allele from a plasmid reduced sporulation frequency in strains that sporulated normally. Furthermore, whenever sspA G52V was expressed in strains with that encoded a wildtype sspA allele, the sporulation frequency was reduced leading to the conclusion that sspA G52V has a dominant-negative effect of the allele. However, this reduction could be due to the overexpression of the allele from a plasmid in comparison with the expression of a chromosomal copy. Due to the observed phenotypes, we hypothesize that SspA and SspB together regulate spore formation. Interestingly, the C. difficile ΔsspA ΔsspB phenotype is similar to the phenotype of a C. difficile ΔspoVT mutant [16,68]. In B. subtilis, SpoVT is involved in a feed forward loop to regulate sporulation. While its expression is through σ G , it works to enhance some, and repress other, σ G -controlled genes [69]. SpoVT regulates the SASPs, SspA and SspB, with a mutant having a 30% reduction in these proteins [70]. This mutant strain is still able to form phase-bright spores but they have a reduced sporulation frequency [70]. In C. difficile, the spoVT mutant has a different phenotype than that observed in B. subtilis. This protein is under σ G regulation and possibly also σ F [15,16,68]. The C. difficile ΔspoVT strain forms immature spores that are phase-dark [16]. Similarly to B. subtilis, SASP expression was reduced in C. difficile; C. difficile sspA had a 70 fold reduction and sspB a 12 fold reduction [16]. The authors suggested that this SASP reduction was probably not sufficient to explain the phenotype [16]. However, our data suggests that the C. difficile ΔspoVT phenotype may be due to reduced sspA and sspB levels. This further highlights the differences in sporulation and SASP function in C. difficile compared to the model B. subtilis and also furthers our hypothesis of SspA and SspB regulating spore formation. Furthermore, in the B. subtilis ΔsspA ΔsspB double mutant, expression levels of some genes were changed [71]. However, the expression levels suggest that the SASPs work to negatively regulate multiple forespore genes and even one mother cell gene [71]. They do note that the changes in transcription may not be due to regulation but to actual binding of the SASPs to the genome, which represses transcription.
B. subtilis sporulation program contains multiple checkpoints. Most of these checkpoints occur in the early stages of sporulation (before/during commitment). More recently, a checkpoint was found that functions later, during the assembly of the spore layers. CmpA is produced in sporulating cells and functions to prevent spores whose pathways have gone awry from completing the sporulation processs. During normal sporulation, CmpA is detected at early sporulation but not in later stages [72,73]. C. difficile does not encode many of the early checkpoints that occur before sporulation commitment because Spo0A activation does not occur through phosphorelay, unlike B. subtilis. C. difficile also does not encode cmpA, but it is likely that it could contain a protein that would function in a similar capacity. Our findings indicate that C. difficile SspA and SspB are necessary for spore maturation and could be a checkpoint for the sporulation. What spores could be generated from the C. difficile ΔsspA ΔsspB strains are without DPA. However, we do not know if cortex is synthesized or if coat layers are formed. Unlike with the CmpA pathway, the C. difficile ΔsspA ΔsspB strains remain stalled post-engulfment and seemingly do not proceed to lysis. Further investigations into the spore assembly process for C. difficile ΔsspA ΔsspB would reveal where SspA and SspB are vital for sporulation and potentially the pathway that they function in. The novel finding of immature spore formation in C. difficile ΔsspA ΔsspB double mutant suggests that C. difficile SASPs perform a previously unreported function in the sporulation process. This insight will open new doors for understanding the regulation of sporulation of C. difficile.

Plasmid construction
The C. difficile sspA targeted CRISPR plasmid was constructed by amplifying 500 bp upstream and downstream from the sspA gene using primer pairs 5'sspA_MTL, 3'sspA_up and 5'sspA_down, 3'sspA_downMTL, respectively. The fragments were cloned into the NotI and XhoI site of pKM126 by Gibson assembly [75]. The gRNA was retargeted to sspA by inserting gBlock CRISPR_sspA_165 into the KpnI and MluI sites by Gibson assembly, generating the pHN05 plasmid. The spl, sspB, CDR20291_1130 and CDR20291_3080 targeting plasmids were similarly constructed. For spl plasmid construction, 5' spl_UP and 3' spl_UP amplified the upstream homology arms and the downstream homology by 5' spl_DN and 3' spl_DN. These were cloned into pKM197 at the NotI and XhoI sites. The gRNA was retargeted by cloning into the KpnI and MluI sites gBlock CRISPR_spl_647 resulting in plasmid pHN61. Primers 5' sspB UP and 3' sspB UP amplified the upstream region, while primers 5' sspB DN and 3' sspB DN
For C. difficile ΔsspA complementation plasmids, the genes and promoter regions were inserted by Gibson assembly into the NotI and XhoI sites of pJS116 [51] by Gibson assembly [75]. The sspA gene plus 500 bp upstream was amplified using primer pair 5'sspA_MTL and 3' sspA.pJS116 to create plasmid pHN11. The vector, pHN14, consisting of the sspB gene and 500 bp upstream was generated using primers 5' sspB UP and 3'sspBpJS116. The sspA portion of the sspA and sspB double mutant complementation vector, pHN30, was amplified with 5'sspApJS116 and 3' sspAsspB. The sspB portion of the double mutant complement was amplified with 5' sspAsspB and 3' sspBpJS116. For pHN84, the sspA complement with a 6x histidine tag on the C-terminus, the primers 5'sspA_MTL and 3' sspA.His_pJS116 were used. The primers 5' sspB UP with 3' PsspB_sspA and 5' PsspB_sspA with 3' sspA.pJS116 were used to generate a plasmid with 500 bp upstream of the sspB gene driving expression of the sspA gene for pHN91. The 500 bp upstream of sspA was amplified with 5'sspA_MTL and 3'sspB_sspA and used to drive expression of the sspB gene, amplified with 5'sspB_sspA and 3'sspBpJS116 for the pHN83 complementation vector. To generate the pHN56 vector of the CDR20291_1130 gene and 500 bp upstream, the primer pair 5' 1130comp and 3' 1130comp were used. For amplification of the CDR20291_3080 gene and 500 bp upstream, pHN57, the primers 5' 3080comp and 3' 3080comp were used. Generation of 500 bp upstream of sspA was amplified with primers 5'sspA_MTL and 3' sspA_CD1130 and used to drive expression of the CDR20291_1130 gene amplified with primers 5' CD1130_sspA and 3' 1130comp for plasmid pHN96. The plasmid pHN97 was generated with 500 bp upstream of sspA, with primers 5'sspA_MTL and 3' sspA_CD3080 and the CDR20291_3080 gene amplified with 5' CD3080_sspA and 3' 3080comp. For the negative control, pHN109, of mCherry driven by sspA promoter region, the sspA promoter region was amplified with primers 5'sspA_MTL and 3' PsspA.mCherry, while the mCherry gene was amplified from pRAN473 [77] with primers 5' PsspA.mCherry and 3' mCherry.PsspA. Another negative control of just the sspA promoter region, pHN102, was generated by using 5'sspA_MTL and 3' PsspA.pJS116 to amplify the sspA promoter region. To generate the σ E _sspB plasmid, the promoter region of bclA2 was amplified with 5' sigE. bclA2_pJS116 and 3' sigE.bclA2_sspB and the sspB gene with 5' sigE.bclA2_sspB and 3' sspB_pJS116, creating pHN80. For the σ K _sspB complement vector, the promoter region of sleC was amplified with 5' pJS116_sigK and 3' sigK_sspB and the sspB gene with 5' sigK_sspB and 3'sspB_pJS116, generating pHN49. The promoter region of gpr was amplified with 5' pJS116_sigF and 3' sigF_sspB and the sspB gene with 5' sigF_sspB and 3' sspB_pJS116 to generate the σ F _sspB plasmid, pHN47. To generate pHN118, the plasmid containing the sspA G52V allele and promoter, primers 5'sspA_MTL and 3' sspA.pJS116 were used to amplify from HNN04. All plasmid sequences were confirmed by DNA sequencing. Cloning was done in E. coli DH5α. A complete list of oligonucelotides used and the strains and plasmids generated can be found in S1 and S2 Tables, respectively.

Conjugations
The resulting plasmids were conjugated separately into C. difficile R20291 using B. subtilis BS49 as a conjugal donor, as described previously [50]. Briefly, 0.25 mL of C. difficile R20291 overnight culture was back diluted into 4.75 mL fresh BHIS and incubated for 4 hours. Meanwhile, 5 mL of BHIS supplemented with chloramphenicol and tetracycline was inoculated with 1 colony of B. subtilis BS49 containing the plasmid and was incubated at 37˚C for 3 hours. After incubation, the B. subtilis culture was passed into the anaerobic chamber and 100 μL culture was plated on TY agar medium, along with 100 μL of C. difficile R20291. This was incubated for approximately 24 hours. Growth was then suspended in 1.5 mL of BHIS and plated onto BHIS agar supplemented with thiamphenicol and kanamycin for selection. Resulting transconjugant colonies were streaked, twice, onto BHIS supplemented with thiamphenicol and kanamycin (TK) or BHIS supplemented with tetracycline (to screen for the conjugal transfer of the Tn916 transposon). Thiamphenicol resistant and tetracycline sensitive colonies were tested by PCR for plasmid components.

CRISPR induction
Overnight cultures of 5 mL TY medium supplemented with thiamphenicol were inoculated with one colony of a C. difficile R20291 strain containing a CRISPR mutagenesis plasmid with tet-driven cas9 (pGC05 and pHN05). After approximately 16 hours of growth, 0.25 mL of overnight culture was back diluted with 4.75 mL TY broth and supplemented with thiamphenicol and anhydrous tetracycline and incubated for 6 hours [50]. A loopful (approximately 10 μL) of culture was plated onto BHIS and the individual colonies were tested by PCR for the mutation. Once PCR confirmed, the plasmid was cured by passaging onto BHIS plates. For induction of xyl-driven cas9 plasmids (pHN61, pHN101, pHN32 and pHN34), colonies were passaged 3 times on TY agar supplemented with 1% xylose and thiamphenicol. Mutants were detected by PCR and the plasmid was cured by passaging in BHIS broth supplemented with 0.5% xylose. Pairwise mutants were generated by conjugating the appropriate plasmid into the necessary mutant strain and inducing as described above.

Spore purification
Cultures were plated onto 70:30 media or CDMM minimal media where indicated and incubated 5 days. Spores were purified as previously described [51,55,56,78]. Briefly, plates were scraped and contents suspended in dH 2 O overnight at 4˚C. The pellets were resuspended then centrifuged at maximum speed. The upper, fluffy-white, layer was removed and resuspended again in dH 2 O. This process was repeated approximately 5 times. The spores were then separated by density gradient in 50% sucrose solution at 3,500 xg for 20 minutes. The spore pellet was washed approximately 5 times in dH 2 O. The spores were stored at 4˚C until use.
To purify the ΔsspB sspAG52V and the ΔsspA ΔsspB double mutant spores, the pellets were scraped into dH 2 O and left overnight at 4˚C. The pellets were then resuspended with 1 μg of lysozyme and incubated for 4 hours at 37˚C. The suspension was centrifuged at max speed for 1 minute and the upper phase removed, and the pellet resuspended with dH 2 O. This process was repeated approximately 5 times before 5 mL of spores were layered onto a HistoDenz

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation gradient of 10 mL of 50% and 10 mL of 25%. This was centrifuged 30 minutes at 18,900 xg at 4˚C. The pellet was then washed approximately 5 times in dH 2 O and stored at 4˚C until use.

Germination assay and DPA content
Germination was monitored using a Spectramax M3 plate reader (Molecular Devices, Sunnyvale, CA). 5 μL of OD 600 = 100 spores were added to 95 μl germination buffer consisting of a final concentration 1x HEPES, 30 mM glycine, 10 mM TA and the OD 600 was monitored for 1 hour at 37˚C. To assay total DPA content, 1 x 10 8 spores in 20 μL, were boiled at 95˚C for 20 minutes. 5 μL (an equivalent of 2.5 x 10 7 spores) of the solution was added to 95 μL of 1X HEPES buffer with 250 μM TbCl 3 and analyzed by excitation at 275 nm and emission at 545 nm with a 420 nm cutoff [54,55,59,66,79].

Western blotting
For the ΔsspB SNP strain, approximately 1 x 10 9 spores were incubated for 1 hour in 50 μL of BHIS with or without 10 mM TA at 37˚C. The solutions were boiled for 20 minutes in 2x NuPAGE buffer at 95˚C. 10 μL (equivalent to approximately 1 x 10 8 spores) of each solution was separated on a 15% SDS PAGE gel. For the double mutant strain, ΔsspA ΔsspB, approximately 4-8 x 10 6 total spores were separated. The protein was transferred to polyvinylidene difluoride (PVDF) membranes at 0.75 amps for 1.5 hours for the ΔsspB sspAG52V strain and at 1 amp for 30 minutes for the ΔsspA ΔsspB strain. The membranes were blocked overnight in 5% milk in TBST. The following day, the membranes were washed 3 times, 15 minutes each, at room temperature. SleC anti-sera was added to 5% milk in TBST at a 1:20,000 dilution for 1 hour. The membranes were washed again, as above. The anti-rabbit secondary antibody was diluted to 1:20,000 in 5% milk in TBST and incubated for 1 hour. The membranes were again washed, as above. The membranes were incubated with Pierce ECL Western Blotting Substrate (ThermoScientific) for 1 minute and then x-ray film was exposed and developed.

Phase contrast imaging
Strains were plated onto 70:30 (70% SMC medium and 30% BHIS medium) sporulation media and incubated for 6 days. After the incubation period, the growth was harvested and suspended in fixative (4% formaldehyde, 2% glutaraldehyde in 1x PBS). The samples were imaged on a Leica DM6B microscope at the Texas A&M University Microscopy Imaging Center.

Heat resistance
Heat resistance assays were completed as previously described [67]. Briefly, three 70:30 plates were inoculated with fresh colonies of the indicated strains. These were grown for 48 hours before half of the plate was harvested into 600 μL of pre-reduced PBS solution and the pellets resuspended. 300 μL of suspension was removed from the chamber and heated for 30 minutes at 65˚C and mixed every 10 minutes by inversion. After heating, these samples were passed into the anaerobic chamber for enumeration. Both the heat-treated and the remaining 300 μL of untreated suspension was serially diluted and plated in technical triplicates of 4.5 μL spots on rich, BHIS, medium supplemented with 0.1% TA. After 22 hours of growth, the CFUs were enumerated. The ratios of treated to untreated were calculated and then the efficiency of sporulation determined in comparison to the wildtype strain. The experiment was performed in biological triplicate.

Ethanol resistance
Ethanol resistance assays were performed as previously described with modification [80]. A 70:30 plate was inoculated and incubated for 48 hours. After incubation, ¼ of the plate was scraped and resuspended in 1 mL of PBS. To enumerate vegetative cells, 500 μL of the culture was added to 500 μL of BHIS and serially diluted and plated on BHIS. To enumerate spores, 500 μL of the culture was added to 300 μL of 95% ethanol and 200 μL of dH 2 O and incubated 20 minutes. After incubation, the solution was serially diluted into PBS with 0.1% TA and plated onto BHIS TA. The sporulation frequency was generated by dividing the number of spores by the total cell count.

UV exposure
The 302 nm UV lamp (Fisher Scientific) was allowed to warm-up for 20 minutes before UV exposure. Spores were diluted in PBS to 1 x 10 9 spores/mL. A sample was collected at T 0 for initial, untreated, spore calculations. 1 mL of a spore solution was added to a miniature glass (UV penetrable) petri dish. These dishes were placed under the UV lamp and exposed with constant agitation for the indicated time. After exposure, the untreated and treated samples were serially diluted then introduced into the anaerobic chamber where they were plated onto BHIS medium supplemented with taurocholic acid and thiamphenicol, for plasmid maintenance. After 48 hours of incubation, the CFUs were enumerated. Treated spore counts were normalized to untreated and then this ratio was normalized to the ratio for wildtype spores.

Outgrowth
Outgrowth, post-germination, was performed in half-strength BHIS supplemented with 30 mM glycine and 10 mM taurocholic acid. Thiamphenicol was supplemented where necessary to maintain plasmids. Spores were added to an OD 600 of 0.05 and the OD 600 was recorded over time.
Chemical resistance 1 x 10 9 spores were suspended in PBS with the indicated chemical concentration for various exposure times. The exposed spores were immediately serially diluted into PBS tubes and plated onto rich media supplemented with taurocholate. For generation of nitrous acid, a one to one volume of 1 M sodium acetate, pH 4.5 and 1 M sodium nitrite were combined and incubated for at least 30 minutes before spore exposure. For formaldehyde exposure, the exposed solution was serially diluted into PBS with 400 mM glycine and incubated for at least 20 minutes to quench the formaldehyde before plating.

RNA extraction
Strains were incubated on 70:30 agar for 18, 24, or 48 hours. The cells were harvested and RNA extractions completed using the FastRNA Pro Blue Kit (MP Biologicals). Cells were lysed in a bead-beater with 3 rounds of 45 seconds on and 2 minutes off on ice. Contaminating DNA was treated with TURBO DNA-free kits (Invitrogen) with 3 rounds of DNase incubation. After DNase treatment, cDNA was generated using SuperScript III First-Strand Synthesis System (Thermo Scientific) with 100 ng of total RNA. Negative controls without the reverse transcriptase were also generated for the samples. 3 biological replicates were extracted and treated for each strain.

PLOS PATHOGENS
C. difficile SspA and SspB contribute to sporulation
Supporting information S1 Fig. Heat based sporulation efficiency is not affected by deletion of sspA, CDR20291_1130 or CDR20291_3080 alone. Strains were grown on sporulation medium for two days. Sporulating cultures were heat treated at 65˚C. Sporulation rate was determined by comparison of the CFU of heat treated culture to CFU of untreated culture and then the ratios were compared to wildtype. pEV indicates an empty vector. All data represents the average of three independent experiments. Statistical analysis by one way ANOVA with Dunnett's multiple comparison test. (TIF) S2 Fig. Ethanol-based sporulation assay. The strains were treated with 28.5% final concentration of ethanol. The sporulation frequency was determined by dividing the number of spores by the total number of cells (spores + vegetative cells). C. difficile Δspo0A was set to the limit of detection to determine sporulation frequency. All data represent the average of three independent experiments and the standard error from the mean. Statistical analysis by one way ANOVA comparison to wildtype with Dunnett's multiple comparison test. � p < 0.001. (TIF) S3 Fig. Chemical resistance is not impacted by loss of individual SASPs. 1 x 10 9 spores were exposed to chemicals for 1, 5, 10, or 30 minutes. After exposure, solutions were serially diluted and plated onto rich medium with germinants. The CFUs were enumerated and compared to unexposed samples and then this ratio was compared with that of the wildtype spores. A) 3% H202 B) 75% EtOH C) 0.25% Glutaraldehyde D) 1 M HCL E) 0.05% hypochlorite F) 2.5% Formaldehyde G) 250 mM Nitrous Acid. pEV indicates an empty vector. All data represents the average of three independent experiments. Statistical analysis by two way ANOVA with Dunnett's multiple comparison. A) C. difficile ΔsspA pEV and C. difficile ΔsspA psspA P<0.0001 at 30 minutes. C) P<0.001 for C. difficile ΔsspB psspB 5 minutes and C. difficile ΔsspA psspA at 10 minutes. P<0.05 for C. difficile ΔCDR20291_3080 pEV at 10 minutes. D) P<0.05 for C. difficile ΔsspA psspA at 5 minutes and C. difficile ΔsspB psspB at 30 minutes. P<0.01 for C. difficile ΔsspA psspA at 10 minutes. P<0.001 for C. difficile ΔsspA pEV and C. difficile ΔsspA psspA at 30 minutes. E) P<0.01 for C. difficile ΔsspB psspB at 1 minute. F) P<0.05 for C. difficile ΔCDR20291_3080 at 10 minutes. G) P<0.05 for C. difficile ΔCDR20291_1130 at 5 minutes, C. difficile ΔsspA pEV and C. difficile ΔCDR20291_1130 pEV at 10 minutes. P<0.01 for C. difficile ΔCDR20291_3080 at 5 minutes. P<0.001 for C. difficile ΔsspA pEV at 5 minutes. (TIF)

S4 Fig. Sporulation efficiency of mutant strains as determined by heat treatment.
Strains were grown on sporulation medium for two days. Sporulating cultures were heat treated at 65˚C. Sporulation rate was determined by taking the ratio of the CFU of heat treated culture to CFU of untreated culture and then the ratios were compared to wildtype. pEV indicates an empty vector. All data represents the average of three independent experiments. Statistical analysis by one way ANOVA with Dunnett's multiple comparison test. hours of incubation, spoVT transcript levels. G) RNA extracted after 48 hours of incubation, pdaA transcript levels. BLQ is below limit of quantification; the cycle threshold is below the limit of which we can accurately quantify. All data represents the average of three independent experiments and the standard error from the mean. Statistical analysis by one way ANOVA with Dunnett's multiple comparison test. (TIF) S1