NanR Regulates Sporulation and Enterotoxin Production by Clostridium perfringens Type F Strain F4969

Clostridium perfringens type F strains, which produce C. perfringens enterotoxin (CPE), are a major cause of gastrointestinal infections, including the second most prevalent bacterial foodborne illness and 5 to 10% cases of antibiotic-associated diarrhea. Virulence of type F strains is primarily ascribable to CPE, which is synthesized only during sporulation.

ABSTRACT

Clostridium perfringens type F strains, which produce C. perfringens enterotoxin (CPE), are a major cause of gastrointestinal infections, including the second most prevalent bacterial foodborne illness and 5 to 10% cases of antibiotic-associated diarrhea. Virulence of type F strains is primarily ascribable to CPE, which is synthesized only during sporulation. Many type F strains also produce NanI sialidase and carry a nan operon that likely facilitates uptake and metabolism of sialic acid liberated from glycoconjugates by NanI. During vegetative growth of type F strain F4969, NanR can regulate expression of nanI. Given their importance for type F disease, the current study investigated whether NanR can also influence sporulation and CPE production when F4969 or isogenic derivatives are cultured in modified Duncan-Strong sporulation (MDS) medium. An isogenic F4969 nanR null mutant displayed much less sporulation and CPE production but more NanI production than wild-type F4969, indicating that NanR positively regulates sporulation and CPE production but represses NanI production in MDS. Results for the nanR mutant also demonstrated that NanR regulates expression of the nan operon. A nanI nanR double null mutant mirrored the outcome of the nanR null mutant strain but with a stronger inhibition of sporulation and CPE production, even after overnight incubation. Coupled with results using a nanI null mutant, which had no impairment of sporulation or CPE production, NanR appears to carefully modulate the availability of NanI, nan operon-encoded proteins and sialic acid to provide sufficient nutrients to sustain sporulation and CPE production when F4969 is cultured in MDS medium.

INTRODUCTION

Clostridium perfringens is an important enteric and histotoxic pathogen of both humans and livestock (1), largely due to its ability to produce at least 20 recognized toxins (1–5). However, toxin production patterns differ significantly among isolates, which is used for classification purposes. The historic C. perfringens toxin typing-based classification system was recently expanded to assign strains of this bacterium to 7 different types (A to G) depending upon their production of six typing toxins (6). In this revised classification scheme, type F strains produce C. perfringens enterotoxin (CPE) and alpha toxin but not beta, epsilon, iota, or NetB toxin.

Type F strains are important enteric pathogens of humans, causing C. perfringens type F food poisoning, which was known as C. perfringens type A food poisoning prior to the classification system revision. This food poisoning is the third most common foodborne disease in the United States, where 1 million people/year are sickened (7–9). Type F strains are also responsible for ∼5 to 10% of nonfoodborne human gastrointestinal (GI) diseases, such as antibiotic-associated diarrhea and sporadic diarrhea (7, 10). CPE production is required for the enteric virulence of all type F strains, whether causing food poisoning or nonfoodborne human GI disease (11).

The sporulating ability of C. perfringens is also important for virulence (7, 12). The environmentally resistant spores made by this bacterium aid in the transmission of several C. perfringens diseases, including those caused by type F strains (7, 12). In addition, CPE production is strictly sporulation associated. Specifically, the production of CPE is controlled by the master sporulation regulator Spo0A and three sporulation-associated alternative sigma factors, named SigE, SigF, and SigK (13–15).

Sialidase production is emerging as yet another potential contributor to C. perfringens enteric virulence (16). C. perfringens can produce up to three sialidases, named NanH, NanI, and NanJ (16). When all three sialidases are produced, NanI is typically responsible for most of the extracellular sialidase activity of that strain (16–19). Most type F strains causing food poisoning do not make NanI, while the type F strains causing nonfoodborne human GI diseases usually produce this sialidase (18). These correlations between NanI production and specific GI diseases are interesting because type F food poisoning involves acute diarrhea, while CPE-associated nonfoodborne human GI disease caused by type F strains is often chronic, lasting up to several weeks, and more severe than the food poisoning (7).

Recent in vitro studies provided evidence of a potential role for NanI in nonfoodborne human GI diseases caused by type F strains (18, 20, 21). Specifically, using isogenic nanI mutants and complementing strains, it was shown that NanI is important for the adherence of F4969, a type F nonfoodborne human GI disease strain, to human enterocyte-like Caco-2 cells (18). Those same strains were also used in another study to demonstrate that NanI can support the vegetative growth and survival of F4969 using intestinally relevant, sialyated nutrient sources (20). A third study showed that NanI can increase CPE binding to, and cytotoxic effects on, Caco-2 cells, thereby suggesting that this sialidase may potentiate enterotoxigenicity (21).

NanI production is environmentally regulated in C. perfringens vegetative cultures (22, 23). For strain F4969, the presence of a 1.6 mM sialic acid concentration in a culture medium was shown to increase nanI transcription, while a 16 mM level of sialic acid or glucose in the culture medium decreased nanI transcription (22). NanR, a member of the RpiR family of transcriptional regulators, plays an important role in regulating transcription of the nanI gene and, possibly, the nan operon that encodes proteins needed for sialic acid uptake and metabolism, as well as nanR itself (22, 23).

Based on those previous results (22, 23), the current model postulates that in an environment with limiting or no amount of sialic acid, NanR functions as a repressor by binding to the promoters of nanI and, likely, the nan operon. However, in C. perfringens vegetative cultures grown in the presence of moderate sialic acid levels, this repressor becomes inactivated, possibly due to its binding of ManNAC-Ap generated during sialic acid metabolism (23). Inactivation of NanR then leads to increased transcription of both the nanI gene and (likely) the nan operon, which allows C. perfringens to generate more sialic acid from environmental sialyconjugates using NanI and then transport and metabolize the resultant free sialic acid via the Nan utilization pathway encoded by the nan operon (22).

NanI production increases levels of sporulation and CPE production when F4969 is grown in glucose-free buffer with semipurified mucin, which is a poor medium that limits growth (20). However, in richer sporulation media, like Duncan-Strong or modified Duncan-Strong (MDS) sporulation medium, NanI production was shown to decrease levels of sporulation and CPE production by F4969 (18). While it is clear that NanR can regulate NanI production (22), and possibly sialidase utilization, under vegetative conditions, it has not yet been assessed whether NanR affects sporulation and CPE production. Consequently, the current study examined whether NanR also impacts sporulation and CPE production when F4969 is grown in MDS sporulation medium. The results obtained indicate that the nanR gene is expressed during growth in MDS and NanR positively regulates sporulation and CPE production by F4969.

RESULTS

The growth of C. perfringens type F human nonfoodborne human GI disease strain F4969 in MDS sporulation medium is repressed by NanR production.

Previous reports demonstrated that F4969, its isogenic nanR null mutant (F4969::nanR), and the reversed mutant (F4969nanRrev) all grew similarly in Todd-Hewitt (TH) medium and Dulbecco modified Eagle medium (DMEM), both of which support only vegetative growth (22). However, sporulation and CPE production, which is strictly sporulation dependent, also play an important role in type F enteric disease (7, 12), so the current study first compared the growth of wild-type F4969, the isogenic nanR null mutant, and the reversed-mutant strain in MDS sporulation medium. The results (Fig. 1A) showed that F4969::nanR and wild-type F4969 grew similarly during the first 3 h of growth at 43°C (the optimum temperature for C. perfringens growth and for F4969 sporulation in MDS medium [24]). However, by 5 h (when the culture was entering stationary phase) the mutant had grown significantly more than F4969. While growth of the F4969nanRrev reversed mutant yielded values for optical density at 600 nm (OD₆₀₀) intermediate between those of the wild-type and mutant strains during the initial 5 h of culture, this growth was not significantly different from that of either of those strains.

FIG 1.

NanR effects on growth of F4969 in MDS medium. (A) Growth curve for wild-type F4969 and isogenic nanR null mutant or reversed-mutant strains cultured at 43°C in MDS medium for 5 h, at which time the entire tube was transferred to 30°C for continued culture up to 24 h. At 2-h intervals between 1 h to 9 h, as well as at 24 h, the cultures were measured for their OD₆₀₀. All experiments were repeated three times, and mean values are shown. The error bars indicate SDs. *, P < 0.05 in comparison with null mutant strain as determined by ordinary one-way ANOVA, with post hoc analysis by Dunnett's multiple-comparison test. (B) (Top) RT-PCR analyses of nanR gene expression in MDS cultures of wild type, null mutant, and reversed-mutant strains grown for 2 h at 30°C after a 5-h culture at 43°C. (Bottom) Transcriptional analysis of the housekeeping 16S rRNA gene as a control for isolated RNA quality. The picture shows a representative result of three repetitions of the RT-PCR experiments.

All MDS cultures were then shifted to 30°C to enhance Ltr-mediated splicing of the intron from intron-disrupted mRNA (25), thus increasing nanR expression in the F4969nanRrev reversed mutant. Under these conditions, extended culture of wild-type F4969 up to 24 h continued to yield significantly lower OD₆₀₀ values than for the nanR mutant. In addition, within 2 h of this temperature shift, OD₆₀₀ values for F4969nanRrev also became significantly lower than those for F4969::nanR, indicating that the higher growth for this nanR mutant was not due to a spontaneous secondary mutation.

Results of a nanR reverse transcription-PCR (RT-PCR) then confirmed that the nanR gene is expressed by both the wild-type and reversed-mutant strains growing in these MDS cultures (Fig. 1B). Expression of nanR by the reversed mutant was lower than for the wild-type strain, as expected since Ltr-mediated splicing is not totally efficient (25). As also expected, nanR gene expression was absent in the nanR null mutant strain under the same culture conditions. As controls, Fig. 1B also shows that all three strains expressed 16S RNA, confirming the quality of all RNA preparations. However, purified RNA samples did not support amplification of 16S RNA without the addition of reverse transcriptase (data not shown), confirming that the RNA preparations were not contaminated with DNA.

NanR positively regulates both sporulation and CPE production in MDS cultures of F4969.

Previous studies showed that vegetative cultures of C. perfringens type F strain F4969 can use sialic acid for growth (20, 22). This process involves NanR regulating the expression of the nanI sialidase gene and, possibly, the nan operon (22, 23). The current study sought to determine whether NanR is also involved in regulating F4969 sporulation and CPE production, which is strongly sporulation associated.

For this purpose, F4969 or previously prepared (22) isogenic F4969::nanR and F4969nanRrev strains were first grown for 5 h in MDS sporulation medium at 43°C. When spores in those 5-h cultures were counted, F4969::nanR was found to have produced significantly fewer spores than the wild-type or reversed-mutant strains (Fig. 2A). After 5 h, all cultures were transferred to 30°C for overnight culture to improve splicing of the intron in the reversed mutant. After this extended overnight culture, the nanR null mutant was found to have produced more spores than it had after 5 h, although it still made significantly fewer spores than the wild-type and reversed-mutant strains (Fig. 2B). The level of sporulation did not increase further, even when the nanR mutant was grown in MDS medium for 3 days (data not shown).

FIG 2.

Comparison of heat-resistant spore formation and CPE production in MDS cultures of F4969, isogenic nanR null mutant, and reversed-mutant strains. The bacteria were cultured in MDS medium for 5 h at 43°C (A, left panel) and then transferred to 30°C for overnight culture (B, left portion). Shown are the average results from three repetitions (log 10 scale). Error bars depict SDs. For panels A and B, an asterisk indicates a P value of <0.05 compared to the null mutant strain by ordinary one-way ANOVA with post hoc analysis by Dunnett's test. Also shown are Western blot results for CPE production by wild-type, isogenic nanR null mutant, and reversed-mutant strains when cultured in MDS medium for 5 h at 43°C (A, middle portion) or for 5 h at 43°C before continued overnight culture at 30°C (B, middle portion). The right portion of panel A shows RT-PCR results for cpe and 16S RNA expression by wild-type F4969, the isogenic nanR null mutant, or the reversed mutant after culture in MDS medium for 5 h at 43°C, followed by a shift to 30°C for 2 h. Panel C shows detection of GUS activity levels in the wild-type (black bars) and nanR null mutant (red bars) strains transformed with a reporter plasmid where GUS production is driven from the promoter of the spoIIA operon, which encodes SigF. The bacteria were cultured in MDS medium 5 h at 43°C (5 h) or first cultured 5 h at 43°C and then transferred to 30°C for overnight culture (O/N). Shown are the average results from three repetitions. Error bars depict SDs. *, P < 0.05 compared to the null mutant strain by Student's t test.

Since CPE is produced only during sporulation, a CPE Western blot was performed to compare CPE production by F4969, the isogenic nanR null mutant strain, and the reversed mutant after growth in MDS medium for 5 h at 43°C or after shifting the 5-h culture to 30°C, followed by overnight incubation to increase NanR production by the reversed mutant. Results of this experiment (Fig. 2) indicated that in the 5-h cultures, there was strong CPE production by wild-type F4969 and weaker CPE production by F4969nanRrev. In contrast, there was no detectable CPE production by the F4969::nanR mutant in the 5-h cultures. After overnight culture, the F4969nanRrev reversed mutant made amounts of CPE similar to those produced by wild-type F4969. While CPE production by the nanR mutant became detectable after this long incubation, it remained less than observed with the wild-type or reversed-mutant strains.

The Western blots in Fig. 2 tested for the presence of CPE in MDS culture supernatants. Since CPE is released only when the mother cell lyses upon the completion of sporulation and the nanR mutant sporulated poorly in Fig. 2A, an RT-PCR analysis was performed to distinguish whether the smaller amounts of CPE detected in MDS supernatants of the nanR mutant were simply due to impaired CPE release or, instead, involved a reduction in cpe expression. As shown on the right side of Fig. 2A, this analysis demonstrated significantly lower cpe expression by the nanR mutant than by the wild-type or reversed mutant strain.

The results in Fig. 2 showed that NanR significantly enhances F4969 sporulation and CPE production, so a further experiment was performed to evaluate when NanR affects sporulation. This involved use of a previously prepared reporter plasmid in which the gusA gene is driven by the promoter controlling expression of the spoIIA operon that encodes SigF, a key early sporulation-specific sigma factor for C. perfringens (14). This reporter plasmid was transformed into wild-type F4969 or the isogenic nanR null mutant strain and GUS activities were measured after 5 h of culture in MDS medium at 43°C or after those cultures were shifted to 30°C and incubated overnight. GUS activity measured in cultures of the wild-type strain was significantly higher than for nanR null mutant cultures at both time points (Fig. 2C), indicating that NanR has an early effect on sporulation and, by extension, CPE production.

NanR represses the exosialidase activity of F4969 in MDS medium.

Our group recently reported that when cultured in vegetative growth medium (TH medium or DMEM), the isogenic nanR mutant had more exosialidase activity (sialidase activity in the culture supernatants) than wild-type F4969 or the reversed mutant (22). This pattern was observed in both young (4 h) and older (overnight) culture supernatants (22). Therefore, the current study measured sialidase activities in both 5-h and overnight MDS culture supernatants.

The results (Fig. 3) indicated that exosialidase activity is significantly higher when wild-type F4969 is cultured in MDS medium (this study) versus TH medium (18). Furthermore, MDS cultures of the nanR null mutant had significantly more exosialidase activity than MDS cultures of the wild-type or reversed-mutant strain after either 5 h or overnight culture (Fig. 3). Figure 3 results also strongly suggested that for the overnight cultures, most sialidase production by F4969 occurred during the initial 5 h of culture at 43°C.

FIG 3.

Comparison of the exosialidase activity in MDS cultures of F4969, isogenic nanR null mutant, and reversed-mutant strains. (A) Exosialidase activity in MDS cultures grown at 43°C for 5 h; (B) exosialidase activity in MDS cultures grown at 43°C for 5 h before continued overnight culture at 30°C. Shown are the average results from three repetitions. Error bars depict SDs. *, P < 0.05 compared to the null mutant strain by ordinary one-way ANOVA, with post hoc analysis by Dunnett's multiple-comparison test.

NanR represses nanA gene expression in F4969 MDS cultures.

On the basis of indirect results from gel shift experiments (23), it has been suggested that NanR can repress nan operon expression. Since it has not been directly evaluated if nanR regulates nan operon expression, and to determine whether this regulation occurs under MDS culture conditions, a quantitative RT-PCR (RT-PCR) was performed for the nanA gene, which is contained within the nan operon (23). This analysis (Fig. 4) detected much higher nanA expression levels in the nanR null mutant than in wild-type F4969 or the reversed mutant. These results demonstrate for the first time that NanR can regulate nanA expression and that this regulation occurs in MDS cultures.

FIG 4.

Quantitative RT-PCR analyses of nanA (as a representative gene in the nan operon) expression levels in F4969, an isogenic nanR null mutant strain, and a reversed-mutant strain grown in MDS medium. Transcript levels were determined using 10 ng of the RNA isolated from MDS cultures grown for 2 at 30°C after a 5-h culture at 43°C. Average C_T values were normalized to that of the 16S rRNA housekeeping gene, and fold differences were calculated using the comparative C_T method (2^−ΔΔCT). Values for each bar indicate the fold change versus wild type. Experiments shown in all panels were repeated three times, and mean values are shown. The error bars indicate standard deviations. *, P < 0.05 compared to the null mutant strain by ordinary one-way ANOVA, with post hoc analysis by Dunnett's multiple-comparison test.

NanI affects sporulation levels and CPE production in F4969 MDS cultures.

A previous study reported that at 37°C, MDS cultures of the F4969 nanI null mutant strain produces significantly more spores and CPE than does wild-type F4969 (18); it was also shown that this effect apparently involves increased expression of spo0A, which encodes the Spo0A master sporulation regulator (18). Therefore, the current study evaluated whether NanI also affects sporulation and CPE production when F4969 is grown in MDS medium under the conditions used in the experiments whose results are shown in Fig. 1 to 4. After 5 h of growth at 43°C, MDS cultures of wild-type F4969 and the isogenic nanI null mutant had similar OD₆₀₀s, but the OD₆₀₀ of the complementing strain culture was significantly higher. When culturing was continued overnight at 30°C, the nanI null mutant culture had an OD₆₀₀ that was significantly lower than those of the wild-type and complementing strains (Fig. 5).

FIG 5.

Comparison of the growth, exosialidase activity, heat-resistant spore formation and CPE production in MDS cultures of F4969, an isogenic nanI null mutant, and reversed mutant strains. (A) Detection of growth, exosialidase activity, spore formation, and CPE production by the indicated MDS cultures incubated for 5 h at 43°C; (B) detection of growth, exosialidase activity, spore formation, and CPE production by the indicated MDS cultures incubated for 5 h at 43°C before overnight incubation at 30°C. Experiments shown in all panels were repeated three times, and mean values are shown. The error bars indicate SDs. *, P < 0.05 compared to the null mutant strain by ordinary one-way ANOVA, with post hoc analysis by Dunnett's multiple-comparison test.

As expected since NanI is the major sialidase of F4969 (18), the exosialidase activity of the nanI null mutant strain was significantly decreased compared to those of the wild-type and complementing strains in either the 5-h or overnight MDS cultures (Fig. 5). As noted previously for F4969 grown in TH medium (18), the complementing strain produced more exosialidase activity at both time points in these MDS cultures.

While the nanI mutant produced less sialidase, both 5-h and overnight MDS cultures of this mutant made significantly more spores and CPE than matching cultures of the wild-type or complementing strains. The complementing strain produced significantly fewer spores and less CPE than wild-type F4969 at both time points.

NanR and NanI double null mutant preparation and characterization.

Since the experiments whose results are shown in Fig. 5 indicated that exosialidase activity affects sporulation and CPE production by F4969 grown in MDS medium, a double nanR nanI mutant was constructed to understand further the effects of the nanR mutation on sporulation and CPE production in MDS cultures. For this purpose, a nanI null mutation was introduced into the nanR null mutant strain to construct a double nanI nanR null mutant. PCR results confirmed this second mutation by showing that DNA from the double null mutant amplified both larger nanI and nanR PCR products (Fig. 6A). Furthermore, an intron-specific Southern blot demonstrated that this double null mutant strain carries two intron insertions, in contrast to the nanI or nanR single null mutant strains, which each had only one intron insertion (Fig. 6B). RT-PCR results clearly demonstrated that the double null mutant strain did not express either the nanI or nanR gene, with a reversed mutant recovering expression of both expression (Fig. 6C). RT-PCR of the 16S rRNA housekeeping gene confirmed the quality of all RNA samples.

FIG 6.

(A) Preparation and characterization of a nanI and nanR double null mutant strain. The left lane shows a 1-kb molecular ruler (Thermo Fisher). The second and third lanes show the nanI PCR product amplified using DNA from wild-type strain F4969 or the nanI nanR double null mutant strain. The fifth and sixth lanes show the nanR PCR product amplified using DNA from wild-type strain F4969 or the nanI nanR double null mutant. Note that DNA from the double null mutant strain supported amplification of larger nanR and nanI products due to the insertion of an intron into the nanI and nanR genes of the double mutant. (B) Intron-specific Southern blot hybridization with DNA from wild-type F4969, single nanI and nanR mutants, or the double null mutant strain. DNA from each strain was digested with EcoRI overnight at 37°C and electrophoresed on a 1% agarose gel. The sizes of DNA fragments are shown to the left. Using DNA from wild-type F4969, no intron-specific band was detected. However, a single intron-specific band was detected for the nanI or nanR null mutant strains, while two intron-specific bands were detected for the double null mutant strain. (C) RT-PCR analysis for 16S RNA (top), nanI (middle), or nanR (bottom) transcription of wild-type F4969, the double null mutant (F4969DKO), and reversed double null mutant strain (F4969DKOrev). Wild-type F4969 DNA was used as a positive control. The leftmost, unlabeled lane contains a 1-kb molecular ruler (Thermo Fisher).

NanR affects F4969 sporulation and CPE production by additional mechanisms besides regulating NanI production.

Once the nanR and nanI double null mutant became available, it was grown in MDS medium for 5 h at 43°C or grown for 5 h at 43°C, followed by overnight culture at 30°C. OD₆₀₀ values of those cultures were then compared against OD₆₀₀ values of similarly grown cultures of wild-type F4969 or a double-reversed mutant (Fig. 7). After 5 h, the double null mutant strain MDS culture had a significantly higher OD₆₀₀ reading than the matching MDS culture of F4969 or the reversed double mutant, resembling the 5-h results for the nanR single mutant in Fig. 1. However, after overnight culture, the double null mutant strain showed a lower OD₆₀₀ reading than the other two strains. This overnight result resembles the phenotypic results for the nanI mutant grown overnight in Fig. 5. Sialidase measurements of culture supernatants detected, at both the 5-h and overnight time points, little exosialidase activity for the double mutant.

FIG 7.

Comparison of the growth, exosialidase activity, heat-resistant spore formation, and CPE production in MDS cultures of F4969, a nanI nanR double null mutant, and reversed double null mutant strains. (A) Detection of growth, exosialidase activity, spore formation, and CPE production by the indicated MDS cultures incubated for 5 h at 43°C; (B) detection of growth, sialidase activity, spore formation, and CPE production by the indicated MDS cultures incubated for 5 h at 43°C before overnight incubation at 30°C. Experiments shown in all panels were repeated three times, and mean values are shown. The error bars indicate SDs. *, P < 0.05 compared to the null mutant strain by ordinary one-way ANOVA, with post hoc analysis by Dunnett's multiple-comparison test.

In contrast to the nanI mutant results (Fig. 5), the double null mutant produced no detectable spores or CPE after either 5-h or overnight culture in MDS medium (Fig. 7).

DISCUSSION

NanR is a transcriptional regulator produced by most, if not all, C. perfringens strains (23). It binds to the promoters for both the nanI gene and the nan operon, which encodes the expression of several proteins involved in sialic acid uptake and metabolism, as well as NanR itself (23). Previous studies showed that under sialic acid-free conditions, NanR represses NanI sialidase production in vegetative cultures (22). The current study now formally demonstrates that NanR can also repress transcription of the nan operon, consistent with previous suggestions (23). This finding establishes that C. perfringens tightly coordinates production of both NanI sialidase, which is (for most strains) the major sialidase generating sialic acid from sialyconjugates (17–19), and the nan operon-encoded proteins that likely transport and metabolize sialic acid.

A second major contribution of this study is establishing a role for NanR during sporulation and CPE production. Despite its importance for disease transmission and enterotoxin production (12), the process of C. perfringens sporulation remains poorly understood. The current study demonstrates that nanR is expressed in sporulating cultures and that the NanR transcriptional regulator positively controls sporulation and CPE production. Studies with a reporter construct indicate that this regulation occurs prior to SigF expression, which is an early event during sporulation.

The results obtained during this study suggest the following model for NanR involvement in regulating C. perfringens sporulation and CPE production. In MDS medium, which contains low but detectable (66 μM) levels of free sialic acid (data not shown), wild-type F4969 produces NanR, NanI, and nan operon-encoded proteins like NanA. In these cultures, NanR modulates production of NanI and sialic acid uptake and metabolism proteins, resulting in steady generation and metabolism of small amounts of sialic acid over an extended time period. When produced, NanI not only removes terminal sialic acid from glycoconjugates but also exposes underlying sugars and amino acids for processing by other enzymes (e.g., glycosylases and proteases) to generate additional nutrients for subsequent metabolism (20). This generation of sustained low-level energy sources by NanI helps F4969 to initiate and complete sporulation.

Results obtained using several isogenic mutants support this model. First, compared to wild-type F4969, the nanR mutant maintained high production levels of sialidase activity and increased expression of nan operon-encoded proteins, even in older cultures. Consequently, in the short term, the mutant quickly uses the free sialic acid present in MDS medium but also generates substantial amounts of additional sialic acid and, perhaps, other exposed nutrients from sialyconjugates present in MDS medium. This surge in energy sources results in the nanR mutant remaining as vegetative cells and growing to higher levels than wild-type F4969 by the end of 5 h. When sialic acid and sialyated glycoconjugates present in MDS medium start to become depleted, the mutant does not respond to this at least in part, because it cannot produce NanR to repress nan operon expression. Therefore, these cells remain committed to growth in a nutrient-poor environment and consequently die before sporulating.

When the nanI gene is inactivated, F4969 produces lower, but still appreciable, exosialidase activity. In this mutant, NanR is still available to repress expression of the nan operon, so the culture grows similarly to wild-type F4969 up to 5 h, modulating its use of free sialic acid and using its other sialidases to generate lower, yet substantial, amounts of sialic acid from sialyconjugates in MDS medium. This sialic acid generation is sufficient to allow the nanI mutant to sporulate and produce CPE similarly to the wild-type strain. However, upon extended culture, the OD₆₀₀ values of nanI mutant cultures become lower than those of the wild-type strain, likely due to incomplete processing of sialyconjugates in MDS medium because NanI is absent. Also, C. perfringens sialidases possess various properties, such as substrate specificity (26), so even though some NanH and NanJ are still produced by the nanI mutant, the absence of NanI may render some portions of sialyated molecules inaccessible for use by this mutant. This further reduces its generation of free sialic acid and, perhaps, other underlying sugars and amino acids, thereby contributing to increased cell death in overnight cultures.

For the double nanI nanR null mutant, the absence of NanR regulation of nan operon expression causes the mutant culture to quickly utilize the free sialic acid available in MDS medium. However, this mutant has little exosialidase activity, so it generates fewer free sialic acids and, perhaps, underlying sugars and amino acids. Consequently, the double mutant dies without sporulating.

The model proposed above, in which NanR finely regulates the generation and use of sialic acids to support sporulation and CPE production, predicts a role for this regulator in virulence, which should be tested in animal models once an in vivo sporulation model is developed. However, sialic acid regulation in vivo would seem to represent an attractive pathogenic strategy for type F strains during enteric disease. The intestines are rich in sialyated glycoconjugates (27), like mucins (in mucus) and glycoconjugates (on host cell surfaces), that represent a pool of potential nutrients for growth and, under the proper conditions, sporulation. Sialic acids could then be harvested from those sources using exosialidases, particularly NanI in the case of strains like F4969, followed by sialic acid transport and metabolism by the nan operon-encoded proteins. NanR would modulate those effects to ensure that sialic acids are not too quickly depleted before sporulation and CPE production are completed, which can take >8 h (7).

While the current results indicate that modulating sialic acid levels can impact C. perfringens sporulation, there are likely other signals involved in initiating sporulation. For example, sporulation is often induced in laboratory media by the addition of polysaccharides like starch or raffinose (28, 29); however, the mechanism behind this effect remains elusive. It has also been shown (30–32) that in vitro, some (but not all) C. perfringens strains respond to intestinally relevant factors like exposure to low pH (as would occur in the stomach after ingestion) and bile salts or phosphate (both present in the small intestines). Once the initiating factors in C. perfringens sporulation become better understood, it will be important to integrate their effects with the influence of NanR and sialic acids on sporulation reported in the current study to gain a fuller appreciation of the C. perfringens sporulation process.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and chemicals.

C. perfringens F4969 is a type F isolate from a case of nonfoodborne human GI disease in Europe during the 1990s (18). The isogenic F4969 nanI null mutant strain and nanI-complemented strain, as well as an F4969 nanR null mutant strain and nanR reverse-mutant strain, had been prepared previously (18, 22). A previously prepared plasmid named pJIR750nanRi (22), which carries a nanR-targeted group II intron in the sense orientation, was used in the current study to construct (i) an F4969 nanR nanI double null mutant strain from the F4969 nanI null mutant strain and (ii) an F4969 nanR nanI reversed-mutant strain from the F4969 nanR nanI double mutant.

Media used to culture C. perfringens in this study included cooked meat medium (CMM; Difco Laboratories), fluid thioglycolate (FTG) medium (Difco Laboratories), modified Duncan-Strong (MDS) medium (15 g/liter of proteose peptone, 4 g/liter of yeast extract, 1 g/liter of sodium thioglycolate, 10 g/liter of disodium phosphate, 4 g/liter of raffinose, and 19.2 g/liter of caffeine), TGY medium (30 g/liter of tryptic soy broth [Becton Dickinson], 20 g/liter of glucose [Fisher Scientific], 10 g/liter of yeast extract [Becton Dickinson], and 1 g/liter of sodium thioglycolate [Sigma-Aldrich]), and brain heart infusion (BHI) agar plates (Becton Dickinson).

Chloramphenicol (CM) used in this study was purchased from the Fisher Scientific Company.

Construction of nanI nanR double null mutant and reversed-mutant strains from a C. perfringens F4969 nanI null mutant strain.

Utilizing the Clostridium-modified Targetron mutagenesis system (33), an F4969 nanI nanR double null mutant (named F4969DKO) was constructed. This involved electroporation of plasmid pJIR750nanRi, carrying a nanR-targeted intron, into the previously constructed F4969::nanI strain (18, 22). When the targeted group II intron was inserted between nucleotides 546 and 547 of the nanR open reading frame (ORF) in the F4969 nanI null mutant, the nanR gene was inactivated (see Results). Growth on BHI agar plates containing 15 mg/liter of CM was used to select for transformants. A PCR screen (see Results) using PCR primers nanRKOF and nanRKOR confirmed the insertion of the intron into the nanR gene of the F4969DKO mutant. The pJIR750nanR plasmid was then cured from the mutant, as described previously (22).

To create a nanR and nanI double reversed mutant (named F4969DKOrev), the pJIR750nanRi plasmid was reelectroporated back into the F4969DKO mutant. Consistent with other reversed C. perfringens mutants, growth of the F4969DKOrev reversed mutant at 30°C increased Ltr-mediated splicing removal of both introns during transcription, partially reversing the phenotypic effects of mutations in both the nanR gene and the nanI gene (see Results). PCR and RT-PCR, described below, were used to confirm the identity of the F4969 double-reversed nanR nanI mutant.

Measurement of C. perfringens growth and sporulation.

For analysis of C. perfringens vegetative growth, 0.2-ml aliquots of overnight FTG cultures of the F4969, F4969::nanR, and F4969nanRrev strains were each inoculated into 10 ml of MDS. The cultures were incubated at 43°C; thereafter, at 1, 3, and 5 h, 1-ml samples were removed for measurement of optical density at 600 nm (OD₆₀₀). The cultures were then transferred to 30°C at 5 h to promote partial nanR expression in the reversed mutant, where every 2 h (from 1 h up to 9 h) a sample was removed for measurement of OD₆₀₀. A final OD₆₀₀ measurement of the cultures was taken 24 h later.

To measure spore formation, a 0.2-ml aliquot of an overnight FTG culture of the F4969, F4969::nanR, or F4969nanRrev was added to 10 ml of MDS. The cultures were then incubated at 43°C for 5 h before each culture was heated at 70°C for 20 min to kill vegetative cells and stimulate spore germination. Alternatively, the cultures were incubated at 43°C for 5 h and then transferred to 30°C for overnight incubation before being heated at 70°C for 20 min. These heat-shocked solutions were serially diluted by 10² to 10⁷ with sterile water and then plated on BHI agar plates. The colonies on each plate were counted after overnight incubation in an anaerobic jar at 37°C.

For assessment of the growth of F4969::nanI and F4969nanIcomp strains, as well as that of F4969DKO and F4969DKOrev, media and procedures identical to those described above were used.

Measurement of supernatant sialidase enzyme activity.

For measurement of supernatant sialidase enzyme activity, a 0.2-ml aliquot of an overnight FTG culture of F4969, F4969::nanR, or F4969nanRrev was transferred to 10 ml of MDS. The cultures were incubated at 43°C for 5 h; alternatively, similar cultures were incubated at 43°C for 5 h and transferred to 30°C for overnight incubation. After incubation, a 20-μl sample of supernatant was taken from each culture and added to 40 μl of 0.05 M Tris-HCl buffer (pH 7.2) in a microtiter plate. A 40-μl aliquot of substrate (4 mM 5-bromo-4-chloro-3-indolyl-α-d-N-acetylneuraminic acid [Sigma]) was then added, and the mixture was incubated at 37°C for 30 min. The absorbance at 595 nm for the mixture was measured using a Bio-Rad microplate reader.

For assessment of supernatant sialidase enzyme activity in cultures of F4969::nanI and F4969nanIcomp strains, as well as that of F4969DKO and F4969DKOrev strains, all media and procedures used were identical.

DNA isolation, PCR analysis, and intron Southern blotting.

C. perfringens DNA was isolated with a MasterPure Gram-positive DNA purification kit (Epicentre), as previously described (18). PCRs were performed using the following amplification program: cycle 1, 95°C for 5 min; cycles 2 through 35, 95°C for 30 s, 55°C for 40 s, and 68°C for 80 s; and a final extension for 5 min at 68°C. A 20-μl aliquot of each PCR sample was electrophoresed on a 1.0% agarose gel and visualized by staining with ethidium bromide. For wild-type DNA, the PCR should amplify a 500-bp product from the nanI gene, using nanIKOF and nanIKOR primers (18), or a 300-bp product from the nanR gene, using the primers nanRKOF and nanRKOR (22). For the null mutant strains, PCR with the same pair of the primers should amplify a 1,400-bp product from the intron-disrupted nanI gene and a 1,100-bp product from the intron-disrupted nanR gene.

Aliquots (3 μg each) of wild-type, single null mutant strain, or double null mutant strain DNA samples were digested overnight with EcoRI at 37°C according to the manufacturer's instructions (New England BioLabs). Intron-specific Southern blotting was performed as described previously (18).

RNA isolation, RT-PCR, and qRT-PCR.

For RNA isolation, a 0.2-ml aliquot from an overnight FTG culture of F4969, F4969::nanR, or F4969nanRrev was added to 10 ml of MDS medium. The inoculated cultures were incubated at 43°C for 5 h and then transferred to 30°C for 2 h. Total RNA was extracted from pelleted cells of those cultures using the saturated phenol (Fisher Scientific) method, as described previously (22). Purified RNA was quantified by determination of the absorbance at 260 nm and stored in a −80°C freezer.

RT-PCR analysis of gene expression was performed with the AccessQuick RT-PCR system (Promega) using 20-μl reaction volumes that contained 20 ng of purified RNA samples and primers (18). RT-PCR of 16S RNA gene expression served as a control to demonstrate the presence of equivalent RNA amounts in different samples (22). Primers for analyzing nanR expression were nanRKOF and nanRKOR (22); expression of the cpe gene was analyzed using primers cpeF (5′-GGAGATGGTTGGATATTAGG-3′) and cpeR (5′-GGACCAGCAGTTGTAGATA-3′), while expression of the nanI gene was evaluated using the nanIKOF and nanIKOR primers (18).

For qRT-PCR analyses, an aliquot (1 μg) of RNA was first synthesized to cDNA using a Thermo Scientific Maxima first-strand cDNA synthesis kit for qRT-PCR according to the manufacturer's instructions. Briefly a 20-μl aliquot of the reaction mixture was amplified at 25°C for 10 min, then at 50°C for 30 min, and finally at 85°C for 5 min. Before qRT-PCR, this cDNA was diluted 10 times to 5 ng/μl. All qRT-PCR primers were designed using the Integrated DNA Technologies (IDT) website. The qRT-PCR primers used to amplify the 16S RNA gene sequences were published previously (22). The primers used for nanA qRT-PCR were qnanA-F (5′-GAAATGATGTTACCAGCTACAGT-3′) and qnanA-R (5′-GCTTCTTTCATAGGTTGTCTACA-3′). Power SYBR green PCR master mix (Thermo Fisher Scientific) and a StepOnePlus qRT-PCR instrument (Applied Biosystems) were used to perform qRT-PCR. After qRT-PCR, the relative quantitation of mRNA expression was normalized to the level of constitutive expression of the housekeeping 16S RNA gene and calculated by the comparative threshold cycle (C_T; 2^−ΔΔCT) method.

Western blot analyses for CPE production.

A 0.2-ml aliquot was taken from an overnight FTG culture of F4969, F4969::nanR, or F4969nanRrev and inoculated into 10 ml of MDS, followed by growth at 43°C for 5 h. Another group of 0.2-ml aliquots from the same strains was inoculated into 10 ml of MDS, followed by growth at 43°C for 5 h before transfer to 30°C for overnight incubation. To perform the CPE Western blot assay, culture aliquots were removed from all strains in both groups and then adjusted to equivalent OD₆₀₀ values. A 1-ml aliquot of each adjusted culture aliquot was then centrifuged. Equal volumes of the resultant supernatants were mixed with 5× SDS loading buffer, and a CPE anti-rabbit polyclonal antibody was used to perform the actual blotting, as described previously (18).

The same media and procedures were also used to analyze the CPE Western blot assay for the F4969::nanI and F4969nanIcomp strains, as well as the F4969DKO and F4969DKOrev strains.

Measurement of GUS activity.

Plasmid pJIR750-sigFp-Gus, prepared previously, was utilized as a β-glucuronidase (GUS) reporter vector to monitor expression from the promoter of the spoIIA operon, which encodes the SigF alternate sigma factor (34). For this purpose, 0.2-ml aliquots of F4969 or F4969::nanR transformed with the pJIR750-sigFp-Gus vector were incubated overnight in FTG medium at 37°C, followed by inoculation at 0.2 ml into 10 ml of MDS medium and growth at 43°C for 5 h. The OD₆₀₀ of each 5-h culture was measured and cells were collected by centrifugation. These 5-h cell pellets were resuspended in 5 ml of phosphate-buffered saline (PBS) and sonicated using a Qsonica sonicator. Lysates were then collected for GUS activity detection. Another group of 0.2-ml aliquots from the same strains were inoculated into 10 ml of MDS medium at 43°C for 5 h before being transferred to 30°C for overnight incubation. The OD₆₀₀ of the overnight cultures were measured, the cultures were centrifuged and sonicated, and the lysates were retained to measure GUS activity.

Using those samples, GUS activity was measured by adding 50 μl of 6 mM 4-nitrophenyl-β-d-glucuronide (in PBS) to 100 μl of each lysate prepared as described above. Following a 30-min or 60-min incubation at 37°C, the absorbance at 405 nm was read. The GUS activities were calculated and are provided as Miller specific activity units.

Statistical analysis.

When comparing results for two groups, Student's unpaired t test was used to test for statistical significance; when comparing the results of more than two groups, ordinary one-way analysis of variance (ANOVA) with post hoc analysis by Dunnett's test was used.