A Metabolite Dehydrogenase Pathway Represses Sporulation of Clostridioides difficile

Abstract

Clostridioides difficile is a major gastrointestinal pathogen that is transmitted as a dormant spore. As an intestinal pathogen, C. difficile must contend with variable environmental conditions, including fluctuations in pH and nutrient availability. Nutrition and pH both influence growth and spore formation, but how pH and nutrition jointly influence sporulation are not known.

Objectives:

In this study, we investigated the dual impact of pH and pH-dependent metabolism on C. difficile sporulation.

Methods:

We examined the impacts of pH and the metabolite acetoin on C. difficile growth, gene expression, and sporulation.

Results:

We found that expression of the predicted acetoin dehydrogenase operon, CD0035-CD0039, was pH-dependent and repressed by acetoin and pyruvate. Regulation of the C. difficile CD0035-CD0039 locus is distinct from characterized orthologous systems and appears to involve a co-transcribed DeoR-family regulator, rather than a sigma⁵⁴-dependent activator. In addition, an CD0036 null mutant produced significantly more spores and initiated sporulation earlier than the parent strain. However, unlike other Firmicutes, growth and culture density of C. difficile was not increased by acetoin availability or disruption of the dehydrogenase pathway.

Conclusions:

Together, these results indicate that acetoin, pH, and the CD0036-CD0039 dehydrogenase pathway play important roles in nutritional repression of sporulation in C. difficile. However, the data do not support the involvement of the CD0036-CD0039 pathway in acetoin metabolism and acetoin is not a significant stationary phase energy source for C. difficile.

INTRODUCTION

Clostridioides difficile is an anaerobic, toxin-producing gastrointestinal pathogen that is the leading cause of antibiotic associated diarrhea (1). As a strict anaerobe, the ability of C. difficile to form spores is critical for both its survival outside of the host and the transmission of the pathogen to new hosts (2). While the formation of C. difficile spores appears morphologically similar to other endosporulating bacteria, the environmental cues and mechanisms that lead to sporulation of C. difficile differ from other spore-forming species and are not well defined (3–6).

Two conditional factors that greatly influence C. difficile spore formation are nutrient availability and the pH of the surrounding environment (3, 7). Prior work demonstrated an acute effect of pH on the sporulation of C. difficile; however, the mechanism by which pH affects sporulation is not understood (7). We hypothesized that pH influences the initiation of sporulation in C. difficile by changing the availability and metabolism of important nutrients. One important nutrient that is metabolized in response to pH in many bacteria is the intermediate metabolite, acetoin (3-hydroxy-2-butanone) (8). During active growth, bacteria decrease the environmental pH due to the production and export of acidic products of glycolysis (9). This drop in pH stimulates the synthesis of acetoin from pyruvate, reducing the acidification of the medium (10).

As preferred nutrients are depleted during stationary phase growth, many Firmicutes catabolize acetoin (9, 11, 12). Acetoin metabolism is performed by Acetoin Dehydrogenases (AoDH), which are enzyme complexes comprised of subunits E1α (AcoA), E1β (AcoB), E2 (AcoC), and E3 (AcoL) (8, 13, 14). In the model sporulating bacterium, Bacillus subtilis, the metabolism of acetoin during post-exponential growth supports spore formation (15, 16). Although acetoin synthesis and degradation has been examined in clostridial species as it relates to industrial fermentation, food science, and solvent production, the relationship between pH, acetoin metabolism, and sporulation is not clear (12, 13, 17–19)

In this study, we sought to determine if acetoin availability and utilization contributes to C. difficile spore formation. Herein, we investigated the predicted acetoin metabolic gene cluster (annotated as acoRABCL, CD0035-CD0039) and assessed the impact of this pathway on C. difficile gene expression, growth, and sporulation. We observed that expression of the CD0035-CD0039 operon was pH-dependent and subject to feedback regulation by acetoin. Further, we found that a mutant in this metabolic pathway (ΔCD0036) produced significantly more spores than the wild-type strain, but surprisingly, acetoin metabolism had no apparent effect on growth in the wild-type or mutant strain. These observations suggest that acetoin plays an important role in nutritional repression of sporulation in C. difficile, but does not function as an energy storage vehicle that can support post-exponential growth and cell proliferation.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Bacterial strains and plasmids used in this study are listed in Table 1. An anaerobic chamber (Coy Laboratory Products) was used to cultivate C. difficile at an atmosphere of 10% H₂, 5% CO₂, and 85% N₂ at 37°C, as previously described (20, 21). Strains were routinely grown fresh from −70°C stocks on brain heart infusion agar supplemented with yeast extract (22) (BHIS; Becton Dickinson Company) broth or agar plates or TY medium (20) in the presence of 1–5 μg/mL thiamphenicol (Sigma-Aldrich) or anhydrotetracycline (ATc) when needed. Escherichia coli was grown at 37°C in LB medium (23) with 100 μg/mL ampicillin and/or 20 μg/mL chloramphenicol (Sigma-Aldrich) as indicated. Following conjugation with C. difficile, E. coli was counterselected using 100 μg/mL kanamycin (24).

Table 1.

Bacterial Strains and plasmids

Plasmid or Strain		Relevant genotype or features	Source, construction or reference
Strains
E. coli
	HB101	F− mcrB mrr hsdS20(rB− mB−) recA13 leuB6 ara-14 proA2 lacY1 galK2 xyl-5 mtl-1 rpsL20	B. Dupuy
B. subtilis
	BS49	Tn916
	MC1849	BS49 Tn916:: CD0035–0039
C. difficile
	630Δerm	ErmS derivative of strain 630	Nigel Minton, (63)
	MC310	630Δerm spo0A::erm	(30)
	MC448	630Δerm pMC358	(36)
	MC492	630Δerm pMC369	This study
	MC1719	630Δerm ΔCD0036	This study
	MC1759	630Δerm ΔCD0036 pMC358	This study
	MC1760	630Δerm ΔCD0036 pMC369	This study
	MC1803	630Δerm ΔCD0036 pMC211	This study
	MC1827	630Δerm pMC1050	This study
	MC1828	630Δerm ΔCD0036 pMC1050	This study
	MC1850	MC1719 Tn916:: CD0035–0039	This study
Plasmids
	pRK24	Tra+, Mob+; bla, tet	(64)
	pSMB47	Tn916 integrational vector; CmR, ErmR	(65)
	pMSR	pseudo-suicide plasmid used for allelic exchange	(26)
	pMC123	E. coli-C. difficile shuttle vector; bla, catP	(66)
	pMC358	pMC123::phoZ	(36)
	pMC369	pMC358, PspoIIGA::phoZ	This study
	pMC911	pMSR with regions flanking acoA	This study
	pMC1050	pMC358, PCD0035::phoZ	This study
	pMC1058	pSMB47 PCD0035-CD0039	This study

Strain and plasmid construction

The oligonucleotide primers used in this study are listed in Table 2. Primer design and the template for PCR reactions were based on C. difficile strain 630 (GenBank accession NC_009089.1). All plasmids were sequenced prior to use (GenScript). Details of plasmid construction are provided in the supplemental Figure S7. Plasmids were conjugated into C. difficile from E. coli or B. subtilis as previously described (24, 25). The CD0036 deletion mutant strain was created after introducing using the pseudo-suicide allelic exchange vector pMC911 and screening by PCR for loss of acoA (26). Complementation of the CD0036 mutant was achieved by integration of Tn196 carrying the full CD0035-CD0039 operon and promoter sequence from pMC1058, which was introduced into the B. subtilis strain BS49 by transformation (MC1849), and conjugated to MC1719, as previously described (25, 27).

Table 2.

Oligonucleotides

Primer	Sequence (5’→3’)	Use/locus tag/reference
oMC44	CTAGCTGCTCCTATGTCTCACATC	rpoC qPCR (66)
oMC45	CCAGTCTCTCCTGGATCAACTA	rpoC qPCR (66)
oMC344	GCGAATTCGGAGACATAACTCCAGAACAAG	Forward primer for PspoIIGA
oMC345	GCGGATCCGTATCCCCCCTTAAAACCTTT	Reverse primer for PspoIIGA
oMC2210	TAGAAATACGGTGTTTTTTGTTACCCTAAGTTTAAACGTTTAGCGAGGGCTTTTGGA	Forward primer for 5’ cassette CD0036, Gibson assembly into pMSR
oMC2215	GATTTTGGTCATGAGATTATCAAAAAGGAGTTTAAACCTGCTTTCAACAATCCCHAGCA	Reverse primer for 3’ cassette CD0036, Gibson assembly into pMSR
oMC2216	AGAAAGATAGATAAAAGHAACAGGAGGGAGAATGGTAAAGGAGGAAAATAAAAATGAGT	Forward primer for 3’ cassette CD0036, Gibson assembly into pMSR
oMC2217	ACTCATTTTTATTTTCCTCCTTTACCATTCTCCCTCCTGTTAACTTTTATCTATCTTTCT	Reverse primer for 5’ cassette CD0036, Gibson assembly into pMSR
oMC2329	CTCTATTAGGAGGGATTGGAGA	screening 5’ crossover CD0036
oMC2332	CATCHCTGGCACAAAACC	screening 3’ crossover CD0036
oMC2358	GCGGCTGTAATGGCTTTAGAA	CD0036 qPCR
oMC2359	ACCCATAAGTTCAGCCATCATTT	CD0036 qPCR / check CD0035–0036 co-transcription
oMC2705	GCCGGATCCTTGCTTTCCTCCAAAATCAA	Reverse primer for PCD0035
oMC2731	GGCGAATTCGCTTGATAAAGAATATAGTACTATACTTAATG	Forward primer for PCD0035
oMC2742	ATCAAAGAGTATTAGGAGTGAAACTGG	check CD0035–0036 co-transcription
oMC2743	GAAGATGCAGTAGAGTTTGCTCA	check CD0036–0037 co-transcription
oMC2744	GCTCCTGCTATAGCTGCTT	check CD0036–0037 co-transcription
oMC2745	GGAGAAATATCAGCCCTGATAAC	check CD0037–0038 co-transcription
oMC2746	ACACATCTACTAAATGAATCCTTCCAT	check CD0037–0038 co-transcription
oMC2747	CTTAGGGGTAGGTGCTACAC	check CD0038–0039 co-transcription
oMC2748	AGTACCTTCTTCCATGGCTACT	check CD0038–0039 co-transcription
oMC2749	GGAAGCTTTGGCAGATGTAAATG	check CD0039–0040 co-transcription
oMC2750	ACCTACTTGAAATCGCTTCATCTG	check CD0039–0040 co-transcription
oMC2772	GCCGGATCCGCTTGATAAAGAATATAGTACTATACTTAATG	CD0035-CD0039 complementation
oMC2773	GACGCATGCGGGGTATGGCTAAGTAAGTG	CD0035-CD0039 complementation

Growth and sporulation in 70:30 medium

Sporulation assays were performed as previously described using a slightly modified 70:30 medium without the addition of Tris base and adjusted for pH, as shown (7, 28, 29). For sporulation in 70:30 broth, C. difficile was first cultured in BHIS with 0.1% taurocholate until mid-log phase, diluted in 70:30 to an OD₆₀₀ of 0.3, and used to inoculate a 50 ml 70:30 culture (start OD₆₀₀ of 0.03), adjusted to pH 7.2 or 6.2, respectively. Two hours after reaching an OD₆₀₀ of 1 (time point T2), cultures were serially diluted and plated onto BHIS agar with 0.1% taurocholate for enumeration of total viable cells. After 24 h, culture samples were subjected to ethanol-resistance assays and spores enumerated as previously described (28).

For sporulation on 70:30 agar plates, log-phase BHIS cultures were diluted in BHIS to an OD₆₀₀ of 0.5 and 250 μl of culture applied on 70:30 plates adjusted to diverse pHs, as indicated (28). After 24 h of growth, cells were scraped from plates, suspended in BHIS to an OD₆₀₀ of 1, and evaluated for total cells and spores as previously described (7, 28). A spo0A mutant (MC310) was used as a negative control to ensure vegetative cell killing in ethanol resistance assays and phase-contrast microscopy was performed to confirm sporulation frequencies. The results represent three independent experiments and are presented as means with standard deviation of the means. Statistical significance was determined using a two-tailed Student’s t-test comparing the mutant to the parent strain or ANOVA, as indicated in the respective figure legends.

Quantitative reverse transcription PCR analysis (qRT-PCR)

Cultures were grown on 70:30 agar and harvested from the plates twelve hours after inoculation (H12) into 6 ml ice-cold water:ethanol:acetone (3:1.5:1.5), and stored at −70°C.

RNA isolation, DNase-I treatment (Ambion) and cDNA synthesis were performed as previously described (30, 31). Quantitative reverse-transcription PCR (qRT-PCR) was performed with three technical replicates using 50 ng cDNA and the SensiFAST SYBR & Fluorescein Kit (Bioline) on a Roche Lightcycler 96 instrument. To confirm the absence of contaminating genomic DNA, cDNA synthesis reactions also included no reverse transcriptase for all samples. Results were calculated using the comparative cycle threshold method (32), normalizing expression to the internal control transcript, rpoC. For expression by pH condition, a one-way ANOVA and Dunnett’s test was performed for statistical comparison to the standard pH condition.

Growth in minimal media

Growth curves were performed using a complete defined minimal media (CDMM) without the addition of D-glucose as previously described, (33) with slight modifications. Selenite (1 μM) (34) and zinc chloride (75 μM) (35), were added to the medium and the pH adjusted to 7.4 prior to filter sterilization. The CDMM base medium was supplemented with 30 mM acetoin (Sigma-Aldrich) or pyruvate (Fisher), as noted. Growth curves in minimal medium were carried out as follows: log-phase cultures were grown to an OD₆₀₀ of 0.5 in BHIS medium, then diluted 10-fold into CDMM. Diluted cultures were used to inoculate CDMM medium broth at a 10-fold dilution for the growth assays to start at an OD₆₀₀ of ~0.01. Statistical significance was determined using a two-tailed Student’s t-test comparing the mutant to the parent strain.

Alkaline phosphatase (AP) activity assays

C. difficile strains containing the transcriptional reporter fusions listed in Table 1 were grown in 70:30 broth pH 7.2, with or without addition of 30 mM acetoin or pyruvate, as indicated. Cells were harvested at the indicated time points. AP assays were performed as previously described (36), without the use of chloroform for cell lysis. Technical duplicates for each strain and condition were averaged, and the results provided as the means and standard deviation for three biological replicates. A two-tailed Student’s t-test was used to compare the activity in the CD0036 mutant to the parent strain or a two-way ANOVA and Dunnett’s test was performed for statistical comparison of multiple strains and conditions, as appropriate.

Detection of Toxins by ELISA

The C. difficile toxins TcdA and TcdB were quantified from the supernatants of cultures grown in TY broth (pH 7.4) for 24 h, according to the manufacturer instructions tgcBIOMICS (catalog no. TGC-E001–1). Technical duplicates were averaged and normalized per ml of cell culture. The results are presented as the means and standard deviations from three independent experiments. Statistical significance was determined using a two-tailed Student’s t-test comparing the mutant to the parent strain.

Acetoin quantitation assay

A modified Voges-Proskauer test was used as a quantitative colorimetric assay to measure acetoin concentration in C. difficile supernatants from cultures grown in the absence or presence of 500 μM, 100 μM, or 50 μM acetoin (37, 38). Briefly, active C. difficile cultures were diluted into duplicate BHIS cultures to an OD₆₀₀ = 0.03. At OD₆₀₀ = 0.5, the indicated acetoin concentration was added to one duplicate culture, and supernatants were harvested at this point (T0), after six hours (T6), or after 24 hours of growth, and frozen at −20°C overnight prior to assay. A BHIS medium control with and without acetoin was included to control for acetoin stability. C. difficile supernatants were diluted 1:3 in distilled H₂O. An acetoin standard curve was created in 0.25X BHIS. To a 96-well plate, 81 μl of the diluted culture supernatant or acetoin standard was added, followed by 56.7 μl of 0.5% creatine, 81 μl of 5% α-naphthol, and 81 μl of 40% KOH, mixing well after the addition of each reagent. After a 15 min incubation at RT, A₅₆₀ was measured using a BioTek Synergy H1 plate reader.

RESULTS

Expression of the CD0035-CD0039 locus is pH-dependent

In previous work, we observed that sporulation of C. difficile substantially increased as the pH of the growth medium was raised, which corresponded with considerable changes in gene expression (7). In an effort to determine the mechanisms involved in pH-dependent sporulation, we identified transcripts that were differentially expressed under low and high relative pH conditions. Of these, the predicted acetoin dehydrogenase genes demonstrated significantly decreased expression that strongly correlated with increases in the pH of the medium (Figure 1; Figure S1). During growth on sporulation agar, we observed a ~5-fold decrease in CD0036 transcription as the pH increased from 5.5 to 6.5, ~8-fold decrease at pH 7.5 (relative to pH 5.5), and a greater than 10-fold decrease in transcription at pH 8.5. Hence, CD0036 expression is greatest at low pH and decreases with corresponding increases in the pH of the medium. Based on these data and prior work showing an effect of acetoin on sporulation in B. subtilis (16), we hypothesized that utilization of acetoin in C. difficile may contribute to the observed pH-dependent sporulation phenotypes.

Figure 1. Expression of the CD0035-CD0039 locus is pH-dependent.

qRT-PCR analysis of CD0036 expression in C. difficile strain 630Δerm grown 12 h on 70:30 agar at pH 5.5, 6.5, 7.5, and 8.5, normalized to pH 5.5. The means and standard deviation of the mean for three independent replicates are shown. Expression levels were analyzed by one-way ANOVA and Dunnett’s multiple comparisons test compared to pH 5.5. Asterisks indicate P values: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001

Defining the CD0035-CD0039 metabolism locus

To identify effects of acetoin metabolism on C. difficile, we first examined the predicted acetoin utilization gene cluster for its transcriptional composition. The CD0035-CD0039 gene cluster was annotated to include a predicted regulator, acoA, acoB, acoC, and acoL (Figure S1; CD0036-CD0039). The tandem arrangement of acoABCL in the C. difficile genome is consistent with the organization of the acetoin locus of B. subtilis. However, the predicted regulatory ORF, CD0035, is present upstream of acoA and is predicted to encode a DeoR-family transcriptional regulator, whereas B. subtilis encodes a sigma 54-dependent accessory activator that is transcriptionally separate and downstream of acoABCL (Figure S1). To define the transcriptional units within this cluster, we examined the entire CD0035-CD0039 locus for transcript linkage. Using cDNA generated from logarithmic cultures, we performed PCR to amplify across adjacent open reading frames (Figure S2). Products were amplified across all ORFs from CD0035 to CD0039, but not between acoL and the downstream ORF, CD0040.

Disruption of CD0036 results in greater sporulation of C. difficile.

In the model spore-former B. subtilis, acetoin reduces spore formation and sporulation is further decreased when the acetoin pathway is disrupted (16). To determine if acetoin utilization impacts spore formation or pH-dependent phenotypes in C. difficile, we generated a deletion mutant in the annotated acoA gene, CD0036, which is predicted to be required for acetoin pathway function (16) (Figure S3). The resulting mutant (MC1719) was examined for the ability to produce ethanol-resistant spores on sporulation agar under a range of pH conditions. Relative to the parent strain, the ΔCD0036 mutant demonstrated considerably greater spore formation, which was statistically significant at both pH 6.5 (6.9-fold) and pH 7.5 (2.8-fold) (Figure 2). The hypersporulation of the CD0036 mutant was fully rescued by chromosomal complementation with the CD0035-CD0039 operon (Figure S4). Further, the addition of acetoin to the medium did not consistently alter sporulation in the wild-type or CD0036 mutant strain. These data suggest that the metabolism via this locus either reduces the ability of the bacteria to engage the sporulation pathway or reduces mature spore production.

Figure 2. The CD0035-CD0039 operon reduces spore formation across a range of pH conditions.

Ethanol-resistant spore formation of wild-type (630Δerm) and the CD0036 mutant (MC1719) grown on 70:30 sporulation agar at pH 5.5, 6.5, or 7.5 for 24 h. Data represent the percentage of spores relative to total viable cells at 24 h, shown with the standard deviation for three independent replicates. A Student’s two-tailed t test was used to compare the 630Δerm and CD0036 mutant outcomes for each pH condition. **** indicates P value ≤ 0.0001; ** indicates P value ≤ 0.01

We further investigated the impact of the CD0035-CD0039 locus on the initiation of sporulation. To do so, we fused the promoter of the early sporulation locus, spoIIG, to the alkaline phosphatase (AP) reporter, phoZ, and examined activity over time during growth in sporulation broth cultures. The spoIIG operon encodes the early mother cell sigma factor, SigE, which is transcribed upon phospho-activation of the master sporulation regulator, Spo0A (39–41). During logarithmic phase growth, the CD0036 mutant and parent strains displayed similarly low levels of expression from the spoIIG promoter, signifying negligible initiation of sporulation within the population (Figure 3). But, by the onset of stationary phase (T₀), the CD0036 mutant demonstrated 9-fold greater reporter activity than the parent strain, indicative of robust sporulation-specific gene expression that precedes the initiation timing observed for the wild-type strain. These results suggest that it is the stationary phase activity of the CD0035-CD0039 dehydrogenase which suppresses initiation of the sporulation pathway, rather than the production or presence of acetoin.

Figure 3. The CD0035-CD0039 operon delays sporulation.

Alkaline phosphatase (AP) activity of the early sporulation PspoIIG::phoZ reporter fusion in C. difficile strain 630Δerm (MC492) and the CD0036 mutant (MC1760), grown in 70:30 sporulation medium containing 1 μg/ml thiamphenicol. Samples were assayed from logarithmic growth (OD₆₀₀ 0.5), early stationary phase (OD₆₀₀ 1.0; T0), and two hours after the onset of stationary phase (T2). The promoterless phoZ reporter carried by the parent strain (MC448, 630Δerm) and CD0036 mutant (MC1759) were included as negative controls. The means and standard deviations for three independent replicates are shown. An unpaired Student’s t-test was used to compare the activity in the CD0036 mutant to the parental strain, per timepoint. ** indicates P value ≤ 0.01

Acetoin metabolism has limited effects on growth or toxin production in C. difficile.

A key trigger for sporulation is limitation of nutrients in the environment (30, 42–44). Similarly, the nutrient limiting conditions that favor the sporulation of C. difficile generally support toxin production. The concurrent production of toxin and spores occurs due to an overlap of regulatory factors that control both processes, including CcpA, CodY, RstA, and Spo0E (3, 31, 44–48). Considering the increased sporulation of the CD0036 mutant, we examined the impact of the locus on toxin production. To this end, wild-type and the CD0036 mutant strain were cultivated in TY medium and toxin TcdA and TcdB levels in the supernatant were assessed by ELISA after 24 h of growth (Figure S5). While there was a trend for the CD0036 mutant to generate more toxin, the differences did not reach statistical significance (P=0.051).

The production and degradation of acetoin can independently impact the environmental pH, cellular redox balance, and growth of bacteria (8). To determine if C. difficile uses acetoin as a carbon source to support cell growth and replication, we assessed growth of the wild-type and mutant strains in a minimal medium with and without acetoin (Figure 4). No significant differences in growth or pH were observed between the CD0036 mutant and parent strain, with or without the addition of acetoin. However, the addition of 30 mM acetoin did decrease the overall growth and culture density by ~20% for both strains (Figure 4ABC), suggesting that excess acetoin has deleterious effects on cell proliferation. Similar, but respectively lesser effects were observed with the addition of acetoin to the medium at 10 mM and 20 mM (not shown). These growth results are in contrast to those seen for B. subtilis, which demonstrates increased proliferation with the addition of acetoin to wild-type cultures and can use acetoin as a sole carbon source (11, 16). Moreover, disruption of the acetoin pathway in B. subtilis resulted in reduced growth for an acoA mutant, even in the absence of exogenous acetoin (11). The lack of demonstrable growth impact of acetoin catabolism in C. difficile suggests that acetoin may play a lesser role as a carbon storage molecule in this bacterium than in other Firmicutes (Bacillota).

Figure 4. Growth and pH are not impacted by the disruption of the CD0035-CD0039 pathway.

Strain 630Δerm (WT) and the CD0036 mutant (MC1719) were cultivated in A) CDMM or B) CDMM with 30 mM acetoin. C) Maximum cell density of WT or CD0036 mutant cultures at 12 h in CDMM with and without acetoin. D) pH measured from the same cultures. Graphs are plotted as the means +/− SD from three independent replicates. Differences in the mean values of WT and ΔCD0036 at each time point and between the same strain +/− acetoin were analyzed by two-way ANOVA with Tukey’s post-hoc test; no significant differences in growth or pH were observed between the strains in either condition. Differences in density for individual strains grown with and without acetoin (C) were assessed by Student’s two-tailed t test. **P ≤ 0.01

Expression of the CD0035-CD0039 operon is repressed by pyruvate and acetoin.

Despite the similarity of the C. difficile CD0035-CD0039 amino acid sequences to orthologs of other Firmicutes, the predicted AcoR regulator of C. difficile, CD0035, bears no significant resemblance to other characterized AcoR regulators (Figure S1). The previously characterized AcoR regulators are sigma 54-dependent accessory activators that are often transcribed independently of the acetoin metabolism genes (8, 49–51). Unlike other aco regulatory factors, CD0035 is a DeoR-family transcriptional regulator encoded within the same operon as the predicted aco metabolic genes. In addition, the CD0035-CD0039 locus of C. difficile does not appear to have sigma⁵⁴/SigL regulatory elements or dependence (52, 53). The aco genes of other Gram-positive bacteria are inducible by pH and acetoin, and are also subject to regulation by carbon catabolite repression through the global regulator, CcpA (16, 50, 54). Though the C. difficile CD0035-CD0039 locus is regulated by pH (Figure S1), no CcpA or glucose-dependent regulation has been observed for this operon, nor is a characteristic cre site present (43). Thus, regulation of C. difficile CD0035-CD0039 gene expression does not appear to be under the same regulatory influences as previously characterized acetoin metabolism systems.

To understand how this pathway and the associated sporulation phenotypes are controlled, we examined expression from the predicted CD0035-CD0039 promoter under different conditions. For this, we generated a transcriptional fusion of the predicted promoter region upstream of the CD0035-CD0039 operon to the phoZ reporter (Paco::phoZ) and expressed the resulting plasmid in the wild-type and ΔCD0036 strains. In contrast to other species, expression from the CD0035 promoter was repressed by acetoin in wild-type C. difficile, rather than induced (Figure 5). PCD0035::phoZ activity was also reduced in the CD0036 mutant relative to the parent strain, which further supports the contribution of acetoin to feedback regulation. The addition of pyruvate to the cultures also resulted in significant reduction in promoter activity, indicating transcriptional repression by this metabolite. However, the mechanism for this regulation is not clear.

Figure 5. Expression of the CD0035-CD0039 operon is impacted by metabolites and function of the dehydrogenase pathway.

Alkaline phosphatase (AP) activity of the PCD0035::phoZ reporter fusions in C. difficile strain 630Δerm (MC1827) and the CD0036 mutant (MC1828), grown in 70:30 medium containing 1 μg/ml thiamphenicol, with or without 30 mM acetoin or pyruvate. Samples were taken during logarithmic growth (OD₆₀₀ 0.5). The promoterless phoZ reporter carried by the parent strain (MC448, 630Δerm) and CD0036 mutant (MC1759) were included as negative controls. The means and standard deviations for three independent replicates are shown. A two-way ANOVA with Tukey’s post-hoc test was used to compare the activity of strains by growth condition. Asterisks indicate P values: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001.

Acetoin dehydrogenase systems (AoDHs) are large, lipoate-dependent enzyme complexes that share high sequence similarity to pyruvate dehydrogenases (PDH) [19–21]. To validate the functions of the enzymes encoded by the CD0035-CD0039 genes, we performed modified Voges-Proskauer tests to assess the utilization and production of acetoin by the wild-type and CD0036 mutant (37). As shown in Figure S6, no significant difference in acetoin concentration was observed between wild-type C. difficile (strain 630Δerm) and the CD0036 mutant grown in BHIS medium with 500 μM, 100 μM, or 50 μM acetoin. Additional tests were performed to assess acetoin production at multiple growth stages and with the addition of precursor substrates glucose and pyruvate; however, no acetoin was detectable by our assays under the conditions tested (not shown). These results suggest that the CD0035-CD0039 locus is not involved in acetoin metabolism, however, it cannot rule out acetoin production or metabolism under other growth conditions, including within the host.

DISCUSSION

Spore formation is a critical process for C. difficile and other anaerobes that allows the bacteria to survive in atmospheric oxygen and spread in the environment. Although sporulation is essential, it is also energy intensive and represents a state of dormancy that requires specific conditions to escape. As a result, all spore-forming bacteria have evolved regulatory mechanisms to prevent the initiation of sporulation when conditions can support growth and replication. The availability of nutrients and the pH of the environment are major factors that limit the entry into sporulation (3, 7). In this study, we examined the impact of the metabolite acetoin to determine if this nutrient is important to post-exponential growth and sporulation in C. difficile, as it is for other organisms. We found that while a predicted acetoin metabolism locus plays a significant role in the production of C. difficile spores and the initiation of sporulation, there are considerable differences in how this bacterium reacts to acetoin and the phenotypes associated with acetoin metabolism, compared with other species. Further, we found no evidence that the predicted acetoin metabolism pathway functions in acetoin or pyruvate utilization.

In most bacteria, including the model spore-former B. subtilis, acetoin acts as an important energy storage molecule that supports post-exponential growth and cell proliferation (16). Surprisingly, the addition of acetoin to the medium did not change the apparent growth rate or final cell density of C. difficile cultures and added acetoin had no significant effects on sporulation (Figure 3, S4). We were also unable to detect acetoin catabolism in rich medium (Figure S6) or minimal medium (data not shown). Loss of CD0036 increased spore formation, but did not have an apparent impact spore maturation or the synchronicity of sporulation. Also, despite the pH-dependent expression of the CD0035-CD0039 operon, disruption of CD0036 had no significant impact on the pH of the medium (Figure 1, 4). These findings suggest that acetoin is not a preferred carbon source for growth, but is sufficient to delay sporulation initiation (Figure 3). It is also possible that the predicted aco pathway is involved in metabolism of a different substrate that is not indicated by the homology of these proteins to acetoin and pyruvate dehydrogenases, while its expression is still regulated by these metabolites. Considering that acetoin can be reduced or oxidized to 2, 3-butanediol, it is also possible that acetoin utilization serves more as a mechanism for balance of redox for C. difficile than as a significant source of energy or a mechanism for reducing acidification of the environment (37). Redox balancing contributes to both toxin production and sporulation; however, the specific connections between these processes have not been determined (55, 56).

In addition to differences in acetoin growth and sporulation phenotypes, the regulation of the CD0035-CD0039 metabolic genes in C. difficile is unlike that of previously characterized dehydrogenase systems (8, 11, 13, 50, 51). In contrast to other bacteria, expression of the CD0035-CD0039 operon is not induced by acetoin in C. difficile, but rather CD0035-CD0039 gene expression is reduced in the presence of acetoin and the precursor metabolite, pyruvate (Figure 5). The most obvious cause of these differences in regulation is that the CD0035 regulator of C. difficile is a co-transcribed DeoR-family regulator, rather than the sigma⁵⁴-dependent activator that regulates transcription of the aco genes in other species (Figure S1) (8, 50, 57). DeoR-family regulators often bind directly to sugars, which act as effectors and control regulator function, while sigma⁵⁴-dependent genes often respond to nitrogen availability (58). It is likely that the C. difficile CD0035 binds to a metabolic effector(s), which would result in more direct regulation, rather than regulation by general nutrient depletion that occurs with AcoR regulators. Further, acetoin metabolism of many characterized species are subject to carbon catabolite repression (CCR), which restricts the utilization of lower quality metabolites when optimal nutrients are available (16, 59). Based on published information, the regulation of CD0035-CD0039 transcription in C. difficile does not appear to be dependent on the characterized metabolic-associated regulators CodY, CcpA, SigH, SigB, SigD, SigL, Spo0A, Rex, PrdR, or ClnR (31, 33, 39, 43, 45, 52, 53, 55, 60–62). However, further studies are needed to characterize CD0035 involvement and potential mechanisms of CD0035-CD0039 transcriptional regulation.

These results demonstrate both the conditional expression of the predicted acetoin metabolic genes and that the CD0035-CD0039 operon has a significant impact on C. difficile spore formation. The contrasts in acetoin utilization and regulation suggest that the differences in niches and nutritional requirements of C. difficile have influenced the utility of acetoin as an energy source. Further study is needed to determine the mechanism of CD0035 regulation in C. difficile and the role of acetoin metabolism in pathogenesis.

Highlights.

• We report on a pH-responsive expression of the locus CD0035-CD0039.

• The CD0035–0039 locus encodes a predicted acetoin dehydrogenase pathway

• The dehydrogenase pathway reduces C. difficile sporulation frequency

• Acetoin metabolism is unchanged in a dehydrogenase pathway mutant, suggesting an alternate metabolic function