Comparison of the Levels of Heat Resistance of Wild-Type, cpe Knockout, and cpe Plasmid-Cured Clostridium perfringens Type A Strains

Abstract

An enterotoxin (cpe) plasmid was cured from a Clostridium perfringens non-food-borne gastrointestinal disease (NFBGID) isolate, and the heat resistance levels of wild-type, cpe knockout, and cpe plasmid-cured strains were compared. Our results demonstrated that (i) wild-type cpe has no influence in mediating high-level heat resistance in C. perfringens and (ii) the cpe plasmid does not confer heat sensitivity on NFBGID isolates.

Clostridium perfringens is a gram-positive, spore-forming, anaerobic bacterium that produces at least 15 different toxins (7, 10). A commonly used classification scheme (11) assigns C. perfringens isolates to one of five types (A through E) based upon their ability to produce alpha-, beta-, epsilon-, and iota-toxin. Although enterotoxin (CPE)-positive C. perfringens type A isolates represent <5% of the global C. perfringens population (10), these bacteria are very important human gastrointestinal (GI) pathogens, causing C. perfringens type A food poisoning (FP) and non-food-borne GI diseases (NFBGID) (10). CPE gene (cpe) knockout studies (13) have demonstrated that CPE expression is necessary for the pathogenesis of both C. perfringens type A FP and NFBGID isolates. In FP isolates, cpe is located on the chromosome, while NFBGID isolates carry cpe on large (∼75-kb) plasmids (3, 4, 8, 15). Although the cpe-carrying plasmid has been shown to be conjugative (2), these plasmids have not yet been fully characterized.

Recent studies (14) suggested that the specific association between chromosomal cpe isolates and FP is attributable, at least in part, to the chromosome-carried-cpe isolates being considerably more heat resistant than plasmid-carried-cpe isolates, which favors their survival in incompletely cooked or inadequately warmed foods. The basis for the differences in heat resistance between FP and NFBGID isolates remains unknown. However, at least two possible explanations for this phenomenon can be envisioned. First, since a previous study (1) revealed differences in heat resistance between CPE-producing and non-CPE-producing FP isolates, it is possible that CPE expression is involved in mediating high-level heat resistance in C. perfringens. In contrast, CPE-producing NFBGID isolates exhibit less heat resistance than do FP isolates because these isolates carry a large, 75-kb cpe plasmid (3, 6) that might carry a gene(s) conferring heat sensitivity. To evaluate these two hypotheses, in this study we cured the cpe plasmid from an NFBGID isolate and compared the heat resistance levels of wild-type, cpe knockout, and cpe plasmid-cured C. perfringens type A strains.

The cpe plasmid pMRS4969 was cured from C. perfringens cpe knockout mutant MRS4969 (13), which was constructed by replacing a 0.4-kb cpe internal fragment with a chloramphenicol (Cm) resistance gene from a cpe plasmid in F4969 (an NFBGID isolate). Strain MRS4969 was grown overnight at 50°C in TGY broth (9) without Cm. After five-times-repeated overnight growth at 50°C, the culture was spread onto brain heart infusion agar (BHIA) plates, which were then incubated overnight at 37°C under anaerobic conditions. Single colonies were patched onto a fresh BHIA plate or BHIA containing Cm (20 μg/ml). After 24 h of anaerobic incubation at 37°C, colonies sensitive to Cm (Cm^s) were selected. The curing of entire pMRS4969 from a Cm^s clone (MRS4970) was confirmed by Southern hybridization and restriction fragment length polymorphism (RFLP) analyses (13, 15). Using a cpe-specific probe, no hybridization signal was observed with HpaI-digested DNA of MRS4970. However, as expected from previous observations (2, 13), two hybridizing bands were observed with HpaI-digested DNA of MRS4969 (Fig. 1A). Further pulsed-field gel electrophoresis and Southern blot analyses (2, 3, 14, 15) showed that the cpe probe hybridized to a single plasmid DNA species of MRS4969 (Fig. 1B), while no hybridizing band was detected in total DNA of MRS4970. Finally, to demonstrate the curing of entire pMRS4969, RFLP analyses were performed. Plasmid DNA, isolated from MRS4969 and MRS4970 as described previously (12), was digested with XbaI and analyzed by agarose gel electrophoresis. These analyses demonstrated that at least 10 XbaI fragments (ranging from 3 to ∼>12 kb) of MRS4969 plasmid DNA were absent in XbaI-digested plasmid DNA of MRS4970 (Fig. 1C). The calculated size (∼75 kb) of the missing fragments matched the size of the entire cpe plasmid present in MRS4969 (Fig. 1B) and F4969 (the parental strain of MRS4969) (6). The cpe probe hybridized to an ∼>12-kb XbaI fragment of MRS4969 plasmid DNA, whereas no cpe-specific hybridization signal was obtained with XbaI-digested plasmid DNA of MRS4970 (Fig. 1D). The common DNA bands found in XbaI-digested plasmid DNA of both MRS4969 and MRS4970 (Fig. 1C) might have come from a non-cpe-bearing plasmid(s) present in these isolates, because NFBGID isolates can carry more than one plasmid (6). Collectively, these results indicated that the entire cpe plasmid had been cured from MRS4970.

FIG. 1.

Molecular characterization of a cpe plasmid-cured strain. (A) Southern blot analysis of HpaI-digested DNA prepared from parent strain MRS4969 and cpe plasmid-cured strain MRS4970. The Southern blot was probed with a 639-bp cpe-specific probe. The migration of the hybridizing bands derived from each strain is indicated by arrows. (B) Pulsed-field gel electrophoresis and Southern blot analysis of undigested DNA, prepared in agarose plugs, from MRS4969 and MRS4970. The blot was probed with a 639-bp cpe-specific probe. The pulsed-field gel was calibrated with bacteriophage lambda DNA markers (Bio-Rad), whose migration is shown at the left of the blot. (C) RFLP analyses of XbaI-digested plasmid DNA prepared from MRS4969 and MRS4970. The plasmid DNA was digested with XbaI and analyzed by subjecting each digested DNA sample to electrophoresis at 100 V in a 1% agarose gel, followed by ethidium bromide staining and visualization under UV illumination. The molecular sizes (in kilobases) of DNA markers are indicated at the right of the gel. The DNA bands that were present in XbaI-digested plasmid DNA of MRS4969 but were absent in XbaI-digested plasmid DNA of MRS4970 are indicated by lines at the left of the gel. The XbaI DNA fragment that was hybridized with the cpe probe is shown by an open arrow. (D) Southern blot of the gel shown in panel C. The blot was probed with a 639-bp cpe-specific probe. The migration of the hybridizing band derived from MRS4969 DNA is indicated by an arrow at the left of the blot.

Having obtained evidence for successful construction of a cpe plasmid-cured derivative of MRS4969, we then compared the sporulation capability of MRS4970 with that of its parent strain, MRS4969. The numbers of sporulating cells present per milliliter of Duncan-Strong (DS) (9) culture were found to be very similar for MRS4969 (1.1 × 10⁷ CFU/ml) and MRS4970 (2.3 × 10⁷ CFU/ml), suggesting that the cpe plasmid does not carry any sporulation-specific genes. To evaluate whether the cpe plasmid plays any role in regulation of CPE expression, we incorporated the recombinant plasmid pJRC200 (5) carrying wild-type cpe into MRS4970 by electroporation (5) and tested for CPE expression. CPE Western blot analyses (9, 14) demonstrated that MRS4970 carrying pJRC200 expresses CPE in a sporulation-associated pattern (data not shown) that mimics CPE expression by its parent cpe-positive strain, F4969, indicating that no cpe regulatory factor(s) is encoded by any gene located on the cpe plasmid. However, this experiment does not rule out the possibility that the cpe plasmid might affect the amount of CPE made by an isolate.

The heat sensitivities of C. perfringens strains were compared by determining the D values (i.e., the time that a culture must be held at a given temperature to obtain a 90% reduction in viable cell numbers) for each strain as previously described (14). To test the reliability of our technique to determine the D values, the heat sensitivities of vegetative cells and spores of SM101 (a transformable derivative of FP isolate NCTC8798 carrying chromosomal cpe [16]) were first compared with the heat sensitivities of vegetative cells and spores of F4969 (an NFBGID isolate carrying a cpe plasmid). As shown in Table 1, vegetative cells of SM101 were found to have an approximately 1.5-fold-higher D value at 55°C than vegetative cells of F4969. Furthermore, the spores of SM101 had approximately a 124-fold-higher D value at 100°C than the spores of F4969. These results are consistent with the results obtained in previous studies (14), indicating that our method is reliable for determining the D value at any specific temperature.

TABLE 1.

Heat resistance of vegetative cells and spores of C. perfringens^a

Strain	Genotypeb	D value (min) for:		Reference
		Vegc cells at 55°C	Spores at 100°C
SM101	WT	11.0 ± 0.7	62.0 ± 5.0	16
MRS101	cpe−	12.0 ± 2.0	61.0 ± 2.1	13
F4969	WT	8.0 ± 0.5	0.5 ± 0.1	3
MRS4969	cpe−	7.0 ± 1.0	0.6 ± .05	13
MRS4970	cpe plas−	7.0 ± 0.5	0.6 ± 0.1	This study

^aResults shown are based on at least three determinations for each experimental parameter for each strain.

^bWT, wild type; cpe⁻, cpe knockout derivative of SM101 or F4969; cpe plas⁻, cpe plasmid-cured derivative of MRS4969.

^cVegetative cells of each strain.

When the heat resistance levels of vegetative cells and spores of FP isolate SM101 were compared with the heat resistance levels of vegetative cells and spores of MRS101 (an isogenic cpe knockout mutant derivative of SM101), no significant differences in heat resistance were observed between the vegetative cells and spores of SM101 and MRS101 (Table 1). These results suggest that chromosomal wild-type cpe has no role in mediating high-level heat resistance in FP isolates. Although one study (1) previously compared the heat resistance levels of spores of CPE-producing FP isolates with those of non-CPE-producing FP isolates, our study is the first study to directly compare the heat resistance levels of vegetative cells and spores of C. perfringens FP isolates carrying wild-type versus knockout cpe.

When similar experiments were performed to compare the heat sensitivities of vegetative cells of MRS4970 with the heat sensitivities of vegetative cells of MRS4969, no significant differences of D values at 55°C were obtained between the vegetative cells of MRS4969 and those of MRS4970 (Table 1). These results suggest that the cpe plasmid has no role in mediating heat sensitivity in vegetative cells of NFBGID isolates. To compare the heat sensitivities of spores produced by MRS4970 and those produced by MRS4969, D values were determined for DS medium cultures of MRS4970 and MRS4969 strains heated at 100°C as previously described (14). The results shown in Table 1 clearly indicate that there are no significant differences in heat sensitivities between the spores of MRS4970 and those of MRS4969. Phase-contrast microscopy analysis confirmed that high levels of spores were present in heat-shocked DS medium cultures of both MRS4969 and MRS4970 (data not shown). Collectively, our results suggest that the cpe plasmid does not carry a gene(s) that might confer heat sensitivity on NFBGID isolates.

In conclusion, our present study initiated experiments to identify the possible basis for the differences in heat resistance between C. perfringens FP and NFBGID isolates. Our results demonstrated that (i) wild-type cpe has no role in mediating high-level heat resistance in C. perfringens and (ii) the cpe plasmid does not carry a gene(s) that might confer heat sensitivity on NFBGID isolates. However, our study cannot rule out the possibility that a non-cpe plasmid(s) present in NFBGID isolates might play a role in heat sensitivity. This received support from our observations that plasmid DNA prepared from cpe plasmid-cured strain MRS4970 produced multiple DNA fragments after XbaI digestion (Fig. 1C), suggesting that MRS4970 may carry a non-cpe cryptic plasmid. Note that these XbaI-DNA fragments were absent in XbaI-digested plasmid DNA prepared from FP isolate SM101 (data not shown). Further analyses of the non-cpe plasmid(s) present in MRS4970 should address the role of that plasmid, if any, in mediating heat sensitivity in NFBGID isolates. Finally, this report provides an invaluable tool, the cpe plasmid-cured strain, to probe plasmid-encoded virulence factors by comparing the phenotypes of isogenic cpe plasmid-cured and plasmid-carrying C. perfringens strains.