NanJ Is the Major Sialidase for Clostridium perfringens Type F Food Poisoning Strain 01E809

ABSTRACT

Clostridium perfringens type F strains cause food poisoning (FP) when they sporulate and produce C. perfringens enterotoxin (CPE) in the intestines. Most type F FP strains carry a chromosomal cpe gene (c-cpe strains). C. perfringens produces up to three different sialidases, named NanH, NanI, and NanJ, but some c-cpe FP strains carry only nanJ and nanH genes. This study surveyed a collection of such strains and showed that they produce sialidase activity when cultured in Todd-Hewitt broth (TH) (vegetative cultures) or modified Duncan-Strong (MDS) medium (sporulating cultures). Sialidase null mutants were constructed in 01E809, a type F c-cpe FP strain carrying the nanJ and nanH genes. Characterization of those mutants identified NanJ as the major sialidase of 01E809 and showed that, in vegetative and sporulating cultures, nanH expression affects nanJ expression and vice versa; those regulatory effects may involve media-dependent changes in transcription of the codY or ccpA genes but not nanR. Additional characterization of these mutants demonstrated the following: (i) NanJ contributions to growth and vegetative cell survival are media dependent, with this sialidase increasing 01E809 growth in MDS but not TH; (ii) NanJ enhances 24-h vegetative cell viability in both TH and MDS cultures; and (iii) NanJ is important for 01E809 sporulation and, together with NanH, CPE production in MDS cultures. Lastly, NanJ was shown to increase CPE-induced cytotoxicity and CH-1 pore formation in Caco-2 cells. Collectively, these results suggest that NanJ may have a contributory role in FP caused by type F c-cpe strains that carry the nanH and nanJ genes.

INTRODUCTION

Clostridium perfringens can produce more than 20 different toxins, many of which are clearly associated with disease in animals or humans (1–3). However, considerable diversity in production of specific toxins exists among different C. perfringens isolates. Based upon this variability, carriage of genes encoding six of these toxins is used to classify C. perfringens isolates into seven toxin types (A to G) (4).

C. perfringens type F strains, which by definition must carry the genes encoding alpha toxin and C. perfringens enterotoxin (CPE), are an important cause of human food poisoning (FP) (4, 5). Animal model studies and epidemiologic data strongly suggest that CPE, a 35 kDa pore-forming toxin, is essential for FP caused by type F strains (6). In the United States, ~1 million cases of C. perfringens type F FP occur per year, resulting in more than $310 million in annual economic losses (7, 8). This FP can be fatal in the elderly, debilitated, or even in healthy younger people if they have severe constipation at the time of infection with a type F strain (9–11). For example, the 01E strains characterized in the current study originated from a 2001 Oklahoma type F strain FP outbreak that involved several cases of fatal necrotizing colitis occurring in people with psychiatric drug-induced constipation at the time of their infection with a type F strain (11). This constipation was thought to interfere with intestinal flushing from CPE-induced diarrhea during FP, thereby increasing contact time between colonic tissue and CPE. Animal model studies suggest that this prolonged contact facilitates CPE absorption into the bloodstream, causing a lethal hyperpotassemia that likely involves CPE-induced cardiac arrest (12, 13).

The cpe gene can be carried by type F strains on either their chromosome (c-cpe) or on large plasmids (p-cpe) (14–17). Type F strains carrying a plasmid-borne cpe gene cause nonfoodborne gastrointestinal (GI) diseases, such as antibiotic-associated diarrhea, as well as a minority of FP cases (16, 18, 19). However, most FP cases are caused by c-cpe type F strains, which are genetically distinct from other C. perfringens strains, including type F p-cpe strains (17, 20, 21). To cite two such differences, type F c-cpe strains lack the pfoA gene encoding perfringolysin O that is present in most other C. perfringens strains, while type F c-cpe strains typically produce spores with more heat, chemical, and cold resistance than the spores made by other C. perfringens strains, including type F p-cpe strains (17, 22, 23).

Besides CPE, sialidases are also becoming implicated as contributors to type F FP or type F nonfoodborne GI diseases (24). For example, the secreted sialidase NanI (75 kDa) significantly increases the adherence of type F nonfoodborne GI disease strain F4969 to enterocyte-like Caco-2 cells in vitro and in vivo colonization of mouse intestine by F4969 (25, 26). NanI sialidase also contributes to the in vitro and intestinal growth of F4969, especially in the presence of mucus (27). This sialidase also enhances CPE-induced cytotoxicity for Caco-2 cells and increases CPE-induced lethal enterotoxemia in mice (28, 29).

However, most type F c-cpe FP strains do not produce NanI sialidase (30). Instead, they make either the NanH (43 kDa) sialidase alone or both the NanH and NanJ (120 kDa) sialidases (30). A previous study (30) showed that the sole sialidase produced by FP strain SM101 is NanH, with this sialidase production occurring only during sporulation. Once produced by SM101, NanH remains cytoplasmic until it is released extracellularly when the mother cell lyses to release its mature spore. Once released, NanH enhances CPE cytotoxicity for Caco-2 cells (30).

Some type F FP strains, such as 01E809, carry both the nanH and nanJ sialidase genes but not the nanI gene (30). Production of NanJ sialidases, which are secreted (24), by type F c-cpe strains carrying the nanH and nanJ genes had received limited research attention until our recent study suggested that NanJ may be the major sialidase of 01E809 (31). The current study sought to rigorously confirm that suggestion and also evaluate whether NanJ supports C. perfringens growth and sporulation. Last, the current study tested whether NanJ can enhance CPE-induced cytotoxicity.

RESULTS

Comparison of sialidase activity and CPE production among C. perfringens c-cpe type F FP strains carrying the genes encoding both NanJ and NanH sialidases.

Our previous study conducted PCR and Southern blot analyses to survey the presence of each C. perfringens sialidase gene (nanJ, nanI, and nanH) in a large collection of c-cpe type F FP strains (30). That study then focused on the six identified type F FP isolates that only carry the nanH gene by showing (i) that those isolates produce NanH when cultured in modified Duncan-Strong (MDS) sporulation medium but not in Todd-Hewitt (TH) vegetative culture broth and (ii) that those isolates release NanH into MDS culture supernatants when the mother cell lyses to release its mature spore.

Additionally, that previous survey identified (but did not study further) other c-cpe type F FP strains carrying genes encoding both NanJ and NanH (30). Those strains originated from two separate FP outbreaks in the United States, i.e., the “01E” strains came from a 2001 FP outbreak in Oklahoma (11) while the “C” strains were from a 1980s FP outbreak in Vermont (32). Therefore, the current study first surveyed sialidase activity in TH or MDS cultures of those type F c-cpe FP strains carrying the nanH and nanJ genes but not the nanI gene (Table 1). This survey detected significant sialidase activity in both 24-h TH and MDS culture supernatants (Fig. 1A) for all 10 surveyed type F FP isolates carrying the nanH and nanJ genes. By comparison and consistent with previous reports, no significant sialidase activity was detected in 24-h TH culture supernatants of c-cpe type F FP strain SM101, which carries the nanH gene but not the nanI or nanJ genes (30). Notably, 7 of the 10 surveyed FP strains carrying both the nanH and nanJ genes had significantly higher sialidase activity in their 24-h MDS culture supernatants than their 24-h TH culture supernatants (Fig. 1A).

TABLE 1.

Isolates used in this study

Isolate	Toxin gene	Sialidase gene	Source and year
C-1841	cpe+, chromosomal	nanJ, nanH	Vermont outbreak, 1980s
C-1849	cpe+, chromosomal	nanJ, nanH
C-1851	cpe+, chromosomal	nanJ, nanH
C-1869	cpe+, chromosomal	nanJ, nanH
C-1881	cpe+, chromosomal	nanJ, nanH
C-1887	cpe+, chromosomal	nanJ, nanH
01E802	cpe+, chromosomal	nanJ, nanH	Food poisoning, Oklahoma, 1999
01E803	cpe+, chromosomal	nanJ, nanH
01E809	cpe+, chromosomal	nanJ, nanH
01E810	cpe+, chromosomal	nanJ, nanH
SM101	cpe+, chromosomal	nanH	Transformable derivative (21) of a 1950s food poisoning strain from Europe

FIG 1.

Survey of supernatant sialidase activity and CPE production by chromosomal cpe type F FP strains carrying both nanH and nanJ sialidase genes. (A) Sialidase activity was measured in supernatants from 24-h TH or MDS cultures of type F FP c-cpe strains that carry both the nanJ and nanH genes. SM101, whose only sialidase gene is nanH, served as a comparison control. Data were corrected for background by subtracting the absorbance at 595 nm values for supernatants of 24-h TH or MDS culture mixed with 0.05 M Tris-HCl (pH 7.2) buffer without the addition of sialidase substrate. Results shown are the averages of three repetitions; the error bars indicate the standard deviation (SD). *, P < 0.05 relative to 24-h TH culture supernatant sialidase activity. (B) Western blot analyses of CPE presence in the same supernatant samples used in panel A. Left panel samples were collected from TH culture, and right panel samples were collected from MDS culture. The right upper panel shows greater CPE production by type F FP strains carrying both the nanJ and nanH strains versus SM101. Because lanes for strains carrying nanJ and nanH exhibited film-saturated CPE production in this blot to allow comparison with CPE production by SM101, the CPE Western blot was repeated (right lower figure) using a 10× dilution of those same MDS culture supernatants to allow comparison of differences (if any) in CPE production among the nanJ- and nanH-carrying type F FP strains. The results shown are representative of three repetitions. The size of proteins in kilodaltons (kDa) are shown on the left.

With respect to CPE production, those FP strains carrying the nanH and nanJ genes all produced more CPE than SM101 after a 24-h TH culture (Fig. 1B). Compared to SM101, the FP strains carrying nanH and nanJ genes also produced more CPE after a 24-h MDS culture (Fig. 1B).

Construction and characterization of a 01E809-derived nanJ single null mutant, an nanH single null mutant, and an nanJ nanH double null mutant.

To evaluate definitively the contributions of NanJ and NanH to the sialidase activity of a type F c-cpe FP strain carrying both the nanH and nanJ (but not the nanI) genes, single nanJ or nanH null mutants and an nanJ nanH double null mutant were constructed in strain 01E809 using Clostridium-modified TargeTron technology (33). For this purpose, the previously prepared pJIR750nanJi and pJIR750nanHi vectors, which carry introns targeted to the nanJ or nanH genes, respectively, were electroporated into type F c-cpe FP wild-type (wt) strain 01E809 (30, 34).

PCR confirmed the presence of the desired intron insertions in each mutant. Using DNA isolated from wild-type 01E809, PCR with internal nanJ primers amplified a 437-bp product, while internal nanH primers amplified a 314-bp product. However, due to insertion of an ~900-bp intron, PCR with the same primers amplified an ~1.3-kb product from the nanJ mutant (named 01E809nanJKO) and an ~1.2-kb product from the nanH mutant (named 01E809nanHKO) (Table 2 and Fig. 2A). Similar PCR analyses also demonstrated the introduction of introns into both the nanH and nanJ genes in 01E809, creating an nanJ nanH double mutant named 01E809DKO (Fig. 2A).

TABLE 2.

Primers used in this study

Primer	Sequence	Size(s)	Use(s)
nanH(A)F	TCTGCACGCAGTACTGATTTT	314 bp (WT); 1,214 bp (KO)	nanH null mutant screening, RT-PCR
nanH(A)R	CCTAGCCATCCAATTGTATTACTTG
nanHcomFnew	CCCGggatccTCTTTAACAAATTCAAGAGCTAAGG (BamHI)	2,280 bp	nanH complementing plasmid
nanHcomRnew	CGTActgcagCAAATTTTGTGGCATTCTATTAATT(PstI)
JKOF	GAAACAACTAAAAAAGGAACGGTTT	437 bp (WT); 1,337 bp (KO)	nanJ null mutant screening, nanJ sequencing, RT-PCR
JKOR	CCATTTTGAATAAGGGTTTTATCAT
cpeF	GGAGATGGTTGGATATTAGG	233 bp (WT); 1,133 bp (KO)	cpe null mutant screening
cpeR	GGACCAGCAGTTGTAGATA
nanJcompFnew	CGGCggatccAACCTATCTTTGTGCAGTACTAC (BamHI)	4,432 bp	nanJ complementing plasmid, nanJ sequence
nanJcomRnew	GCAGgtcgacTCCAACTGCTCCTACAATTAATG (SalI)
qnanRF	AGGAGTAGGAGAAGCTACCATAA	145 bp	nanR RT-qPCR
qnanRR	ACTCTCATCCCTATGAACATGC
qccpAF	ACACATAGAACAAGAACTGTAGGT	87 bp	ccpA RT-qPCR
qccpAR	TACGTCCTCAGCACCTCTAA
qcodYF	GTGCTACAATAGTTGGGATGGA	94 bp	codY RT-qRCR
qcodYR	CCAATAGCTAATTGAACCACTGC
nanJseq1F	CAAGACTCAGCTTCAAGTGC		nanJ sequencing
nanJseq1R	GCAGTTTCAATAATAGCTTG		nanJ sequencing
nanJseq2F	AAAGTTTAATTGGTTTAAGTGATGG		nanJ sequencing
nanJseq2R	GTAGCAATTCTTGTATATGAACCA		nanJ sequencing
nanJseq3F	TGGAAGTTAGGAGAATCTCC		nanJ sequencing
nanJseq4F	ATCATAGGAGAGAGAGAGATAAC		nanJ sequencing
nanJseq4R	CTCCAACTCCTATTAAAGCTTTTGT		nanJ sequencing
nanJseq5F	ATAGGAGTTGGAGATGGAAA		nanJ sequencing

FIG 2.

Intron-based mutagenesis to create an 01E809 nanJ single null mutant, an nanH single null mutant, or an nanJ nanH double null mutant and characterization of those mutants by PCR, an intron-specific Southern blot, and RT-PCR. (A) PCR analysis using nanJ internal primers (top) or nanH internal primers (bottom) for wild-type, null mutant, and complementing strains. The expected size of nanJ or nanH PCR products amplified from the wild-type parent or the null mutant strains are listed in Table 2. Size of DNA (bp) is shown at left. (B) Southern blot analysis of an intron-specific probe using DNA from wild-type parent or its null mutants. DNA from both wild-type parents and null mutants was digested with BsrGI overnight, electrophoresed on a 1% agarose gel prior, blotted, and hybridized with an intron-specific probe. Size of DNA fragments, in kilobases, is shown on the left. (C) RT-PCR analysis using nanJ (top) or nanH (bottom) primers for wild-type, null mutant, and complementing strains. The results shown are representative of three repetitions. Size of DNA fragments, in base pairs, is shown on the left.

To confirm that only a single intron had inserted into 01E809nanJKO or 01E809nanHKO and that two introns had inserted into 01E809DKO, Southern blotting was performed using an intron-specific probe. Results demonstrated (Fig. 2B) the presence of a single intron insertion in the 01E809 nanJ or nanH single null mutants and two intron insertions in 01E809DKO.

NanH or NanJ complementing strains of 01E809nanHKO or 01E809nanJKO were prepared by transforming those single sialidase null mutants with a shuttle plasmid carrying either the wild-type nanH or nanJ gene, creating the complementing strains 01E809nanHcomp and 01E809nanJcomp. It is not technically possible at present to complement simultaneously two inactivated genes in C. perfringens, so 01E809DKO could not be double complemented. Figure 2A PCR results confirmed the presence of the expected wild-type sialidase gene in the two complementing strains since PCR using DNA from 01E809nanHcomp or 01E809nanJcomp supported amplification of both the wild-type nanH and nanJ genes.

To evaluate sialidase gene expression in the sialidase mutants and complementing strains, reverse transcriptase PCR (RT-PCR) for nanH and nanJ genes was performed in both TH and MDS. The results showed that in both media, wild-type nanJ mRNA was not expressed by the nanJ null mutant strain but was expressed by the nanH null mutant. A large ~1.5-kb RNA band was detectable in the MDS cultures of this mutant and likely represents the presence of some nanJ mRNA disrupted by the 900-bp intron. The gene nanH was not expressed by the nanH null mutant but was expressed by the nanJ null mutant strain. Furthermore, RT-PCR also showed that the complementing strains recover wild-type nanJ or nanH mRNA expression, respectively, in the nanJ null mutant or nanH null mutant (Fig. 2C).

Comparison of growth, vegetative cell viability, spore formation, and CPE production levels by wild-type 01E809, its isogenic sialidase null mutants, and complementing strains cultured in TH broth.

The current study first determined that the 01E809 parent strain grew similarly to its isogenic sialidase null mutants and complementing strain derivatives when cultured in TH at 37°C (Fig. 3A). When 24-h TH cultures were plated onto brain heart infusion (BHI) agar plates to determine vegetative cell viability, the nanJ single mutant, nanH single mutant, and nanJ nanH double null mutant cultures each contained significantly fewer viable vegetative cells per milliliter compared to similar cultures of wild-type 01E809 or the nanJ or nanH complementing strains (Fig. 3B). These results indicated that both the NanH and NanJ sialidases contribute to 24-h vegetative cell viability in TH cultures.

FIG 3.

Comparison of postinoculation changes in culture OD₆₀₀, vegetative cell viability, heat resistant spore formation, and CPE production levels for 01E809 and its isogenic derivatives when cultured in TH medium. (A) postinoculation changes in the OD₆₀₀ of TH cultures from 0 h up to 8 h at 37°C. (B) Viable vegetative cells (top) and formation of heat-resistant spores (bottom) for wild-type 01E809, its null mutant, and complementing strains. The bacteria were cultured in TH for 24 h at 37°C, and 10-fold serial dilutions (in phosphate-buffered saline [PBS] buffer) were then plated onto BHI agar plates. After overnight anaerobic growth at 37°C, colony counting of viable vegetative cells was performed. To determine the presence of heat-resistant spores, an aliquot of each culture was then heat-shocked for 20 min at 70°C before plating. Shown are the mean values from three independent experiments. The error bars indicate the SD. *, P < 0.05 relative to wild type. (C) Western blot analyses of CPE levels in supernatants of panel B samples. The results shown are representative of three repetitions. The size of proteins in kilodaltons (kDa) are shown on the right.

Heat-resistant spore count results demonstrated that 24-h TH cultures of 01E809 contained only very low numbers (~100/mL) of spores, confirming that these are predominantly vegetative cultures. However, the nanH-single null mutant and the nanH nanJ double mutant both produced significantly more spores than their wild-type parent, and complementation of the nanH mutant reversed this phenotype (Fig. 3B).

When the same TH cultures were subjected to CPE Western blot analysis, the results (Fig. 3C) showed that CPE production levels were significantly greater for the nanH null mutant or the nanH nanJ double null mutant compared to those of the wild-type or NanH complementing strains. Since CPE production requires sporulation, these Western blot results were consistent with the spore formation results and indicated that NanH represses CPE production, as well as sporulation, in TH cultures.

Comparison of sialidase activity for TH cultures of wild-type 01E809, its isogenic sialidase null mutants, and complementing strains.

A previous study determined that when 01E809 (which encodes both nanH and nanJ) is cultured in TH medium, its sialidase activity was detectable at three time points (4, 8, or 24 h) using either whole (sonicated) cultures or culture supernatants (31). Therefore, the current study also compared the sialidase activities in both culture supernatants (which contain only extracellular sialidase activity) and sonicated whole cultures (which contain both intracellular and extracellular sialidase activity) for 01E809, its sialidase null mutants, and their complementing strains when cultured at 37°C in TH broth at those same three TH culture time points (4, 8, or 24 h).

The results (Fig. 4A) confirmed our previous report (31) that wild-type 01E809 produces detectable extracellular and intracellular sialidase activity in both whole TH cultures and TH culture supernatant at all three tested time points. However, even whole TH culture sialidase activity significantly decreased for the nanH or nanJ single null mutants compared to that of the wild-type parent at those same time points. Furthermore, complementation of those sialidase mutants restored wild-type sialidase activity. Notably, for the nanJ null mutant, sialidase activity was nearly undetectable in both TH culture supernatants and sonicated whole TH cultures at all three time points (Fig. 4A). In contrast, the nanH mutant retained partial sialidase activity relative to the parent strain when cultured in TH broth. These results demonstrate that NanJ is the predominant sialidase for 01E809 in TH cultures.

FIG 4.

Comparison of sialidase activity in TH culture supernatants and sonicated whole cultures and analyses of sialidase gene expression in those cultures by RT-qPCR. (A) Sialidase activity detected for wild-type 01E809, its single sialidase null mutants, and the complementing strains cultured in TH at 37°C, with samples collected at 4, 8, and 24 h. Orange bars show sonicated whole culture sialidase enzyme activity. Blue bars show culture supernatant sialidase activity. (B) nanH (top) or nanJ (bottom) RT-qPCR for samples collected from 4-h TH cultures incubated at 37°C. (C) nanR, ccpA, and codY RT-qPCR for samples collected from 4-h TH cultures incubated at 37°C. Shown are the mean values from three independent experiments. The error bars indicate the SD. *, P < 0.05 relative to wild type.

Since both the nanJ and nanH null mutants produced less sialidase than 01E809 in TH cultures, reverse transcriptase quantitative PCR (RT-qPCR) was performed to determine whether, in this culture condition, nanJ expression affects nanH expression or nanH expression affects nanJ expression. As shown in Fig. 4B, the nanJ mutant exhibited increased nanH expression. Since NanH contributes less than NanJ to sialidase activity in TH cultures (Fig. 4A), the small increase in nanH expression levels observed for the nanJ mutant would have limited effect on the total sialidase activity of this mutant. As expected, complementation of the nanJ mutant restored nanH expression to wild-type levels (Fig. 4B). Interestingly, inactivation of nanH expression partially decreased nanJ expression compared to that of the wild-type parent (Fig. 4B), which likely contributes to the partial reduction in sialidase activity of the nanH mutant in Fig. 4A. Importantly, complementation to restore nanH expression also returned nanJ expression to wild-type levels.

RT-qPCR experiments were performed to begin exploring why the nanH null mutant has reduced nanJ expression and the nanJ null mutant has increased nanH expression in TH cultures. Since NanR is a transcriptional regulator that can control expression of some C. perfringens sialidase genes (31, 35, 36), nanR RT-qPCR was performed for 01E809 and its derivatives when cultured in TH. The results showed that nanR expression does not significantly change between the wild-type parent versus the nanJ or nanH mutants in TH cultures (Fig. 4C), i.e., the observed nanJ expression effects on nanH expression or nanH expression effects on nanJ expression do not require changes in nanR expression.

CodY or CcpA are two other candidate transcriptional regulators that might be involved in nanH expression effects on nanJ expression. Supporting their potential involvement in the observed 01E809 nanH expression effects on nanJ expression, or vice versa, production of NanI sialidase was shown to affect codY and ccpA expression in C. perfringens type D strain CN3718 (37).

Therefore, the current study performed ccpA and codY RT-qPCR for 01E809 and its derivatives when cultured in TH medium. The results showed that, as for nanR expression, expression of the ccpA gene does not significantly change between the wild-type parent and nanJ or nanH mutants (Fig. 4C). However, codY expression did change significantly between wild-type 01E809 and these mutants. For the nanJ mutant, codY expression was lower than the wild-type parent. In contrast, codY expression level is higher for the nanH mutant than the wild-type strain (Fig. 4C). Importantly, the complementing strains expressed similar levels of codY as wild-type 01E809 (Fig. 4C).

Comparison of growth, vegetative cell viability, spore formation, and CPE production levels by wild-type 01E809, its isogenic sialidase null mutants, and complementing strains when cultured in MDS sporulation medium.

It was next determined that the nanJ null mutant and double sialidase null mutant grow slower than wild-type 01E809, the nanH null mutant, or the complementing strains when cultured in MDS at 37°C (Fig. 5A). This result indicated that NanJ contributes to growth of 01E809 in this culture condition.

FIG 5.

Comparison of postinoculation changes in culture OD₆₀₀, vegetative cell viability, formation of heat-resistant spores, and CPE production levels for 01E809 and its derivatives when cultured in MDS medium. (A) Postinoculation changes in the OD₆₀₀ of MDS cultures from 0 h up to 8 h at 37°C. (B) Viable vegetative cells (top) and heat-resistant spore (bottom) levels for 01E809, its null mutants, and complementing strains. After culture in MDS for 24 h at 37°C, samples were serially diluted 10-fold in PBS buffer and plated onto BHI agar plates. Those plates were then grown anaerobically overnight at 37°C for colony counting of viable vegetative cells. To determine the presence of heat-resistant spores in these cultures, an aliquot of each culture was heat-shocked for 20 min at 70°C before plating. Shown are the mean values from three independent experiments. The error bars indicate the SD. *, P < 0.05 relative to wild type. (C) Western blot analyses of CPE levels in supernatants of panel B samples. Because the film was saturated in the upper panel to allow comparison of CPE production by 01E809 versus all of its derivatives, including 01E809DKO, which produces low amounts of CPE, this blot was also repeated using a 10× dilution of these same MDS samples to allow better comparison of CPE production between strains other than 01E809DKO (bottom). The results shown are representative of three repetitions. The size of proteins in kilodalton (kDa) are shown on the left.

When the 24-h MDS cultures were plated onto BHI agar, results demonstrated that the cultures of the nanJ mutant, nanH mutant, or nanJ nanH double null mutant each contained a modest but significant decrease in viable vegetative cells compared to those of cultures of the wild-type parent or the nanJ or nanH complementing strains (Fig. 5B). This result indicated that both NanH and NanJ contribute to the 24-h vegetative cell viability and survival of 01E809 in MDS medium. Counting heat-resistant spores in these cultures revealed that the nanJ null mutant and, particularly, the nanHnanJ double mutant produced significantly fewer spores compared to those of the wild-type parent or complementing strains (Fig. 5B). These results strongly suggested a role for NanJ, especially in the presence of NanH, as a contributor to 01E809 sporulation under these culture conditions.

When samples of the same 24-h MDS cultures were subjected to CPE Western blot analysis, the results (Fig. 5C) showed that CPE production levels were significantly less for the nanH nanJ double null mutant compared to those of the other strains. Those results strongly suggested a combined role for NanJ and NanH as contributors to CPE production by 01E809.

Comparison of sialidase activity for wild-type 01E809, its isogenic sialidase null mutants, and the complementing strains when cultured in MDS medium.

A previous study determined that 01E809, which encodes both nanH and nanJ, produces detectable levels of sialidase activity in whole (sonicated) cultures or culture supernatants when grown at 37°C in MDS medium for 8 or 24 h (31). Therefore, the current study compared sialidase activities in culture supernatants (which contain only extracellular sialidase activity) and sonicated whole cultures (which contain both intracellular and extracellular sialidase activity) of 01E809 versus its sialidase null mutants and their complementing strains when these bacteria are cultured in MDS at 37°C for 4, 8, or 24 h.

The results (Fig. 6A) confirmed our previous report (31) that wild-type strain 01E809 produces significant sialidase activity in both whole cultures and culture supernatants when cultured at 37°C in MDS for 8 and 24 h but not after culture in MDS at 37°C for only 4 h. Additionally, after similar culture in MDS for 8 or 24 h, the nanJ null mutant had significantly less sialidase activity in both whole cultures and culture supernatants than the wild-type parent, nanH null mutant, and complementing strains (Fig. 6A). This result demonstrated that NanJ is also the major sialidase for this strain in this culture condition.

FIG 6.

Comparison of sialidase activity in MDS culture supernatants and sonicated whole cultures of 01E809 and its derivatives, as well as RT-qPCR analysis of sialidase or regulatory gene expression in these cultures. (A) Sialidase activity measured in the wild-type parent, its null mutant, and the complementing strains using samples collected at 4, 8, and 24 h from MDS cultures incubated at 37°C. Orange bars show sonicated whole culture sialidase activity. Blue bars show culture supernatant sialidase activity. (B) nanH (top) or nanJ (bottom) RT-qPCR for samples collected at 4 h in MDS cultures incubated at 37°C. (C) nanR, ccpA, and codY RT-qPCR for samples collected from 4-h MDS cultures incubated at 37°C. Shown are the mean values from three independent experiments. The error bars indicate the SD. *, P < 0.05 relative to wild type.

RT-qPCR was performed to detect whether nanJ expression affects nanH expression or nanH expression affects nanJ expression when 01E809 or its derivatives are cultured in MDS medium. The results, shown in Fig. 6B, indicated that nanH expression modestly decreased in the nanJ null mutant under these culture conditions, possibly contributing to the virtual absence of sialidase activity for this mutant in MDS culture. In contrast, nanJ expression greatly increased for the nanH mutant, likely contributing to the observation that the sialidase activity of this nanH null mutant strain was comparable to that of the wild type under these culture conditions. Complementation of either the nanH or nanJ mutants restored expression of the sialidase genes to near wild-type levels.

RT-qPCR was also performed to compare nanR, ccpA, and codY expression by 01E809 and its derivatives when cultured in MDS medium. The results showed that expression of nanR or codY genes does not significantly change between the wild-type parent versus the nanJ or nanH mutants in this culture medium (Fig. 6C). However, ccpA expression did change significantly between wild-type 01E809 and its nanJ mutant or between this parent and its nanH null mutant. Furthermore, the complementing strains expressed similar ccpA levels as the wild-type strain. Therefore, in this culture condition, ccpA expression level is higher for both the nanJ and nanH null mutants than wild-type 01E809 (Fig. 6C).

Construction of a 01E809 cpe null mutant and a cpe nanJ nanH triple null mutant.

Sialidase activity can affect CPE production levels in MDS (Fig. 5). Therefore, to study whether NanJ affects CPE cytotoxic activity in the presence of 01E809 extracellular sporulating culture factors such as extracellular proteins, two additional 01E809 null mutants were prepared so that these mutants could be supplemented later with identical amounts of purified CPE. For this purpose, an intron was inserted into the cpe gene of 01E809 or the cpe gene of the 01E809 nanH nanJ double null mutant strain, creating (respectively) a cpe null mutant and a cpe nanJ nanH triple mutant. PCR results (Fig. 7A) confirmed that the cpe gene had been disrupted by an intron insertion in both mutants, while Southern blot analyses (Fig. 7B) demonstrated the expected presence of a single intron insertion in the cpe null mutant and three intron insertions in the cpe nanJ nanH triple mutant. A NanJ complementing strain was also prepared for the triple null mutant, with PCR demonstrating the presence of a wild-type nanJ gene in the nanJ-complemented triple null mutant strain (Fig. 7A).

FIG 7.

Characterization of an intron-based mutagenesis-created cpe null mutant, an nanJ nanH cpe triple null mutant, and an nanJ complementing strain of that triple mutant by PCR and an intron-specific Southern blot. (A) PCR analysis using cpe internal primers (top), nanJ internal primers (middle), or nanH internal primers (bottom) for wild-type 01E809 and its derivative null mutant strains. The expected size of cpe, nanJ, or nanH PCR products amplified from the wild-type parent or the null mutant strains are listed in Table 2. Size of the DNA is shown on the left. (B) Southern blot analysis of an intron-specific probe using DNA from wild-type parent or its null mutants. DNA from both wild-type parents and null mutants was digested with BsrGI overnight, electrophoresed on a 1% agarose gel prior, blotted, and hybridized with an intron-specific probe. Size of DNA fragments, in kilobase, is shown at left.

Presence of NanJ in MDS culture supernatants enhances CPE-induced cytotoxicity, binding, and CH-1 complex formation in Caco-2 cells.

The supernatants of overnight MDS cultures were first assessed for their sialidase activity. Results (Fig. 8A) showed that the 24-h MDS culture supernatants of the cpe null mutants contained the same amount of sialidase activity as produced by wild-type 01E809. Under the same culture condition, supernatant from the cpe nanJ nanH triple null mutant had no detectable sialidase activity, as expected. The NanJ complementing strain produced significantly more sialidase activity than the triple null mutant but a little less activity than the wild-type parent strain.

FIG 8.

Supernatants containing NanJ increase CPE-induced cytotoxicity, CPE binding, and CH-1 complex formation in Caco-2 cells. (A) comparison of supernatant sialidase activity for 01E809, its cpe null mutant, its triple cpe nanH nanJ null mutant, and a nanJ complementing strain of the triple null mutant. These supernatants were prepared from 24-h MDS cultures, before or after HBSS buffer exchange to remove toxic compounds in MDS. Shown are the mean values from three independent experiments. The error bars indicate the SD. There were no significant differences (P > 0.05) in sample sialidase activity before or after buffer exchange. *, P < 0.05 relative to wild type. (B) CPE Western blot for the sample of panel A, and CPE (0.5 μg/mL) was added to panel A sample. The results shown are representative of three repetitions. The size of proteins in kDa are shown at left. (C) Supernatants (1 mL) from panel B were applied to Caco-2 cells in 12-well cell culture plates for 1 h at 37°C. Supernatants were then collected for cytotoxicity detection using the Roche cytotoxicity detection kit (LDH). Shown are the mean values from three independent experiments. The error bars indicate the SD. *, P < 0.05 relative to 01E809cpeKO with 0.5 μg/mL of CPE. Caco-2 cells were then gently removed from plates and resuspended in RIPA buffer (Alfa Aesar) containing Benzonase (Millipore Sigma) for CPE Western blotting (D). The results shown are representative of three repetitions. The size of proteins in kilodaltons are shown on the left.

In order to evaluate whether NanJ can affect CPE-induced cytotoxicity in 24-h MDS culture supernatants, there was a complication to overcome. Previous studies showed that MDS itself contains small molecules that increase the background spectrophotometric reading for the lactate dehydrogenase (LDH) release cytotoxicity detection kit (30). To address these problems, we employed the same approach used previously to assess whether NanH can enhance CPE activity in 24-h MDS culture supernatant backgrounds (30). Specifically, the toxic material in MDS was removed from culture supernatants of all three cpe-negative mutants by buffer-exchange to Hanks balanced salt solution (HBSS) using Thermo Scientific Zeba spin desalting columns. Each buffer-exchanged sample was then supplemented with the same defined amount of CPE (0.5 μg/mL) that causes a nonsaturating amount of cytotoxicity when measured with the LDH cytotoxicity detection kit.

To assess whether buffer-exchange chromatography had affected sialidase activity, sialidase activity was compared in 24-h MDS culture supernatants before and after buffer exchange. This analysis detected no change in sialidase activity after buffer exchange (Fig. 8A). As expected, CPE Western blot analysis confirmed that the cpe mutant, the triple mutant, and the nanJ complementing strain of the triple mutant all failed to make CPE under this culture condition and that the same amount (0.5 μg/mL) of CPE supplementation could be added to the buffer-exchanged MDS supernatants of the cpe mutant, the triple mutant, or the nanJ complementing strain of the triple mutant (Fig. 8B).

Those CPE-supplemented and buffer-exchanged supernatants were then used to evaluate whether NanJ can enhance CPE cytotoxicity for Caco-2 cells in a background containing extracellular C. perfringens sporulation-associated factors, such as proteins. For this CPE cytotoxicity experiment, HBSS buffer was used as a negative control and HBSS buffer containing 1% Triton-100 served as a positive control. Cytotoxicity results for this experiment showed (Fig. 8C) that, without the addition of CPE, less than 15% cytotoxicity was detected using buffer-exchanged MDS supernatants from any of the three tested strains. However, when buffer-exchanged supernatants from the cpe mutant were supplemented with 0.5 μg/mL of CPE, cytotoxicity increased to about 80%. In contrast, similar addition of 0.5 μg/mL CPE to buffer-exchanged supernatants of the triple null mutant strain caused only about 40% cytotoxicity for Caco-2 cells. When buffer-exchanged supernatants from the nanJ complementing strain triple mutant were supplemented with 0.5 μg/mL of CPE, cytotoxicity was restored back to 80%. These results demonstrated the ability of NanJ to enhance CPE-induced cytotoxicity.

After CPE binds to receptors, formation of the CH-1 CPE pore complex is required for cytotoxicity in Caco-2 cells (38). Therefore, the presence of that complex in Caco-2 cells was assessed by CPE Western blotting. As expected, no CH-1 complex was detected after treatment of Caco-2 cells with buffer-exchanged MDS supernatants from cultures of the cpe null mutant or triple mutant (Fig. 8D). However, after addition of equivalent concentrations (0.5 μg/mL) of CPE to each buffer-exchanged MDS supernatant, CPE Western blotting detected less CPE binding and CH-1 pore complex formation in Caco-2 cells treated with buffer-exchanged MDS supernatants of the triple null mutant compared to buffer-exchanged MDS supernatant from the cpe mutant or nanJ complementing strain. Furthermore, after addition of equivalent amounts of CPE, the nanJ complementing strain caused more CPE binding and CH-1 pore complex in Caco-2 cells compared to the triple mutant (Fig. 8D). These CPE Western blotting experiments indicated that the presence of NanJ enhances CPE binding and CH-1 formation in Caco-2 cells.

DISCUSSION

The current study surveyed a collection of C. perfringens type F c-cpe FP strains carrying both the nanJ and nanH genes that originated from two FP outbreaks in the United States, i.e., the “C” strains came from an FP outbreak in Vermont in the 1980s (32), while the “01E” strains came from an FP outbreak in Oklahoma in the 2001 (11). This survey showed that these FP strains carrying nanJ and nanH differed from SM101 (which only carries the nanH gene) by producing supernatant sialidase activity in both MDS sporulation medium and TH broth. Most of the surveyed type F c-cpe FP strains carrying nanJ and nanH, including 01E809, produced greater sialidase activity in MDS versus TH. This finding demonstrated the influence of the culture medium on production of sialidase activity for these FP strains. Similarly, the current study confirmed a previous finding (31) that culture medium influences the kinetics of sialidase activity production, i.e., 01E809 did not begin to produce sialidase activity until between 4 and 8 h of culture in MDS, in contrast to more rapid production of sialidase activity when this strain is cultured in TH.

Compiling the current and previous results (25, 30) suggests that C. perfringens strains carrying all three sialidase genes typically produce the most sialidase activity, strains carrying only the nanH and nanJ genes possess intermediate sialidase activity, and strains carrying only the nanH gene produce low sialidase activity. However, the production of sialidase activity is more complicated than mere sialidase gene carriage. As already mentioned, the sialidase activity of a strain is clearly influenced by the culture medium. For example, SM101 produces little sialidase activity in TH but detectable levels of sialidase activity in MDS (30). In contrast, type F p-cpe nonfoodborne GI disease strain F4969 has moderate sialidase activity in both TH and MDS. For F4969, which carries all three sialidase genes, NanI sialidase is considered the most important contributor to sialidase activity, whether this strain is cultured in TH or MDS (25).

NanJ appears to be only a minor contributor to the sialidase activity of F4969 (25). However, the contribution of NanJ to the sialidase activity of nanJ and nanH-carrying type F c-cpe FP strains like 01E809 had received limited attention until a recent study (31) suggested that this sialidase may be the major contributor to the sialidase activity of 01E809. That suggestion was based upon (i) 01E809 making more sialidase activity than SM101 (which lacks nanJ) in TH or MDS medium, as confirmed in the present study; (ii) the coupled observations indicating that a nanR null mutant of 01E809 has less sialidase activity than its wild-type parent in TH and RT-qPCR showing that NanR regulates nanJ expression, but not nanH expression, in TH cultures of 01E809; and (iii) the majority of sialidase activity in mid-log phase (~4 h) TH cultures of 01E809 being present in culture supernatants, which supports the importance of NanJ for sialidase activity since this sialidase is secreted while NanH is not.

The most definitive approach to assessing the relative contributions of different sialidases to the overall sialidase activity of a strain involves use of sialidase null mutants. Therefore, the current study constructed nanH or nanJ sialidase gene knockout mutants in strain 01E809. Experiments using those mutants and complementing strains indicated that both NanJ and NanH contribute some to sialidase activity in TH cultures of 01E809, but NanJ activity predominates. The negligible sialidase activity in TH cultures of the nanJ null mutant of 01E809 is consistent with a previous report indicating that SM101, which only makes NanH sialidase, produces little or no sialidase activity in TH (30). NanJ is also clearly responsible for nearly all sialidase activity in MDS cultures of 01E809, which differs from SM101, where NanH contributes significantly to sialidase activity in MDS cultures (30). These observations, coupled with earlier results for F4969, identify diversity in the relative contributions of different sialidases to the overall sialidase activity of various type F strains (25).

Interestingly, RT-qPCR comparisons of TH cultures of 01E809 versus its sialidase null mutants revealed that, in the absence of nanJ expression, 01E809 compensates by upregulating nanH expression. Despite that modestly increased nanH expression, the negligible sialidase activity measured in TH cultures of the nanJ mutant of 01E809 indicates that this upregulated nanH expression has little impact on total sialidase activity of the mutant in this culture condition. In MDS cultures of the nanJ mutant, the absence of nanJ expression slightly decreased nanH expression, which may contribute to the near absence of total sialidase activity for MDS cultures of this mutant. Conversely, the absence of nanH expression decreased nanJ expression in TH cultures of the nanH mutant but strongly increased nanJ expression in MDS cultures of this strain. The latter effect may help to explain why the nanH mutant retains full sialidase activity in MDS culture. Overall, these results identify an interesting, and previously unappreciated, interplay between nanH expression and nanJ expression that varies with culture medium.

This study also began investigating how expression of one C. perfringens sialidase gene might impact expression of other sialidase genes. One plausible explanation could have involved sialidase-induced changes in nanR expression levels for the nanH or nanJ null mutants since NanR can regulate C. perfringens sialidase gene expression (31, 35). However, the current study determined that nanJ expression effects on nanH expression, or vice versa, do not involve changes in nanR expression by 01E809. In contrast, RT-qPCR did detect differences in codY or ccpA expression between the nanJ or nanH mutants versus wild-type 01E809 when cultured in TH or MDS, respectively. Those differences may be important for regulating sialidase gene expression since CodY and CcpA are sensitive to bacterial nutritional status, which could be influenced by the generation of sialic acids due to sialidase activity (37). Specifically, sialic acids are thought to be converted in the C. perfringens cytoplasm to molecules such as fructose biphosphate, which can regulate CcpA activity and generate energy intermediates, which could also be important since GTP can influence CodY activity (37, 39). Conceivably, NanJ and NanH may vary in how they remove sialic acids from sialyl conjugates or in the specific sialic acids they remove from sialyl conjugates (40). Those differences might then impact codY or ccpA expression, or CodY or CcpA activity, and lead to changes in expression of other sialidase genes. Further study of this regulatory network is warranted to obtain a better understanding of the interplay between the effects of individual sialidases on expression of other sialidase genes.

In addition to elucidating the relative contributions of NanH and NanJ to the overall sialidase activity of 01E809, the isogenic sialidase mutants also provided insights into NanH and NanJ contributions to growth, vegetative cell survival, sporulation, and CPE production by this strain. Neither NanH nor NanJ production was required for growth into stationary phase when 01E809 is cultured in TH broth, although those sialidases did increase 24-h viability of vegetative cells under this growth condition. This observation may indicate that these two sialidases are not necessary in fresh rich media having abundant readily utilizable nutrients but, as those are depleted during culture, 01E809 then relies upon sialidases to help generate sialic acids as new nutrients. A previous study showed that NanI also can contribute to F4969 growth and survival in TH medium (27).

Interestingly, nanH expression was found to significantly repress sporulation of 01E809 in TH cultures. The observed high level of sporulation for the nanH null mutant in TH broth is unusual since this medium generally supports vegetative growth, as evident by the wild-type parent producing only 100 spores/mL even in 24-h TH cultures. This suggests that sialidases, particularly NanH, generate nutrients to continue vegetative growth and delay sporulation in TH broth. Last, only the mutants unable to express nanH made detectable levels of CPE in TH cultures, which is consistent with the sporulation results and the strict association of CPE production with sporulation (5).

For MDS cultures, results of the current study indicated that NanJ does contribute to growth during log-phase, highlighting that sialidase growth contributions vary in different media. Since MDS induces sporulation, it is plausible that, in less favorable media (like MDS) for vegetative growth, NanJ is needed for optimal growth. With respect to 24-h viability/survival in MDS, both NanJ and NanH contributed. Relative to 01E809, the nanJ null mutant (but not the nanH null mutant) showed a 1-log decrease in sporulation levels when cultured in MDS. That decrease is relatively small, so no significant decrease in CPE production was detected for the nanJ single mutant. However, when cultured in MDS, the nanJ nanH double null mutant showed a larger (two log) drop in sporulation, which was sufficient to detect significantly less CPE production. This result suggests that NanH also makes a subtle but significant contribution to sporulation and CPE production in MDS cultures of 01E809.

Last, results from the current study demonstrated that NanJ can enhance the cytotoxicity of CPE. This involved increased formation of CH-1 pore complex. Since all of the bound CPE in treated Caco-2 cells was present in CH-1 complex, this NanJ promotion of CH-1 complex formation involves enhanced CPE binding. Previous studies have shown that NanI sialidase enhances the binding and cytotoxicity of CPE, C. perfringens epsilon toxin, and C. perfringens beta toxin in vitro and CPE in vivo (28, 34). Another study showed that NanH can also enhance CPE binding and cytotoxicity in vitro (30). With the current demonstration that NanJ promotes CPE activity, it is now established that all three sialidases can enhance the cytotoxicity of C. perfringens toxins by increasing toxin binding. Whether this binding promotion involves trimming away sialic acid residues on the host cell surface that sterically reduce toxin binding or by reducing sialic acid-induced charge inhibition of toxin binding remains to be determined.

In summary, this study adds NanJ to the list of C. perfringens sialidases that could act as accessory virulence factors by promoting growth, sporulation, toxin production, and toxin activity. Collectively, those effects could assist the development of type F gastrointestinal diseases.

MATERIALS AND METHODS

C. perfringens strains, media, and chemicals.

C. perfringens strains used in this study are listed in Table 1. C. perfringens stock cultures were maintained in cooked meat medium (CMM) (Difco Laboratories) at −20°C. For vegetative cultures, FTG (fluid thioglycolate) medium (Difco Laboratories), TH medium (Bacto Todd Hewitt Broth [Becton, Dickinson], with 0.1% sodium thioglycolate [Sigma-Aldrich]), or TGY medium (3% tryptic soy broth [Becton, Dickinson], 2% glucose [Fisher scientific], 1% yeast extract [Becton, Dickinson], and 0.1% sodium thioglycolate [Sigma-Aldrich]) were used. For sporulating cultures, C. perfringens strains were grown in MDS sporulation medium (Bacto proteose peptone, 15 g/L; yeast extract, 4 g/L; disodium phosphate, 10 g/L; sodium thioglycolate, 1 g/L; raffinose, 4 g/L; and caffeine, 19.2 g/L). For selection and plate counting of mutant and complementing strains, BHI agar plates (brain heart infusion) (Becton, Dickinson) with or without chloramphenicol (Cm) (from Fisher Scientific) were used. Anaerobic incubations were performed using Gas Pack jars.

To culture Escherichia coli DH5α transformants and isolate plasmids, Luria-Bertani (LB) broth (1% tryptone [Becton, Dickinson], 0.5% yeast extract [Becton, Dickinson], 1% NaCl [Fisher scientific]) was used.

Construction and characterization of 01E809 null mutants.

The nanJ, nanH, or cpe genes in 01E809 were inactivated by insertion of a targeted group II intron using the Clostridium-modified TargeTron system (33). For this purpose, previously prepared pJIR750nanJi, pJIR750nanHi, or pJIR750cpei plasmids were electroporated into wild-type strain 01E809 (30, 34). Mutants were then selected using BHI agar plates containing 15 mg/L Cm. Mutant screening primers and expected PCR product sizes are listed in Table 2. For this screening, PCRs included were as follows: 1 μL of each pair of primers (at a 0.5 μM final concentration), 1 μL of purified DNA template (100 ng), and 25 μL of 2× DreamTaq green PCR master mix (Fisher Scientific). These were mixed together and double-distilled water (ddH₂O) was then added to reach a total volume of 50 μL. Each reaction mixture was placed in a thermal cycler (Techne) and subjected to the following amplification conditions: 1 cycle of 95°C for 2 min, 35 cycles of 95°C for 30 s, 55°C for 40 s, and 72°C for 1 min 40 s, and a single extension of 72°C for 5 min. PCR products were then electrophoresed on a 1.5% agarose gel, which was stained with ethidium bromide. After curing the pJIR750nanJi, pJIR750nanHi, or pJIR750cpei plasmid, the resultant single sialidase gene null mutants were named 01E809nanJKO, 01E809nanHKO, or 01E809cpeKO. The double nanJ nanH null mutant was named 01E809DKO, and the triple nanJ nanH cpe null mutant was named 01E809TKO.

To construct nanJ or nanH complementing strains, which were named 01E809nanJcomp or 01E809nanHcomp, respectively, we prepared pJIR750 C. perfringens-E. coli shuttle plasmids (41), carrying either a cloned nanJ or nanH gene. PCR primers used in this study for construction of these complementing vectors are also listed in Table 2. These complementing plasmids were then electroporated into the nanJ or nanH null mutants. Using the pJIR750nanJcomp vector and the primers listed in Table 2, the nanJ gene of 01E809 was sequenced by GENEWIZ. Sequencing primers are listed in Table 2.

Southern blot analyses using an intron-specific probe were performed to confirm that single null mutants had only had one intron insertion, the double null mutant had two intron insertions, and the triple null mutant had three intron insertion in their genomes. An intron-specific probe was prepared by PCR and a DIG labeling kit (Roche Applied Science), as previously described (42). DNA was prepared from wild-type 01E809 and their null mutant strains using the Epicentre DNA purification kit (purchase from Fisher Scientific). An aliquot of each purified DNA (3 μg) was digested overnight at 37°C with BsrGI and then electrophoresed on a 1% agarose gel. Digested DNA was alkali transferred to a nylon membrane (Roche), and the blot was then hybridized with a digoxigenin-labeled intron specific probe. DIG detection reagents with CSPD substrate (Roche) were used to detect hybridized probes, as specified in the manufacturer’s instructions.

C. perfringens growth curve analyses and quantitative counts of viable vegetative cells or heat-resistant spores in TH or MDS cultures.

As described previously (43), growth (optical density at 600 nm [OD₆₀₀]) was measured, and viable vegetative cells and heat-resistant spore formation were quantified in TH or MDS cultures of 01E809 wild-type or their derivatives. To enumerate heat-resistant spores, the cultures were heat-shocked at 70°C for 20 min before plating in order to kill vegetative cells and promote heat-resistant spore germination.

Measurement of sialidase enzyme activity.

To measure sialidase activity, a 0.2-mL aliquot from an FTG overnight culture was transferred into 10 mL of fresh TH or MDS medium. Those cultures were then incubated at 37°C for specified times (see Results). Aliquots from those TH or MDS cultures were then removed and sonicated using a Qsonica sonicator, as described previously (30). Culture supernatants or supernatants of sonicated culture were used to determine sialidase activity. For this purpose, an aliquot (60 μL) of each supernatant was added to a 40-μL of substrate (4 mM 5-bromo-4-chloro-3-indolyl-α-d-n-acetylneuraminic acid [Santa Cruz]). Each reaction was then incubated at 37°C for 1 h, and absorbance at 595 nm was measured using a Bio-Rad microplate reader.

C. perfringens mRNA isolation, RT-PCR, and RT-qPCR analysis.

Strain 01E809, its isogenic nanJKO or nanHKO null mutants, and complementing strains were grown in TH or MDS broth for 4 h at 37°C. From those pelleted cultures, RNA was extracted with saturated phenol and purified by TRIzol and chloroform (Life Technologies and Sigma), as described previously (43). PCR without reverse transcriptase (RT) assessed whether the isolated RNA samples were DNA-free before RT-PCR or RT-qPCR analysis. When DNA contamination was detected, DNase (Thermo Fisher) was used to remove the residual DNA contamination. Each purified RNA sample was quantified by measuring A₂₆₀ and used to prepare cDNA with a Maxima first-strand cDNA synthesis kit (Thermo Scientific) according to the manufacturer’s instructions. An aliquot (1 μL) of each cDNA was used for RT-PCR to check nanJ or nanH gene expression in wild type, nanJ or nanH null mutants, and their complementing strains. The primers used are listed in Table 2. Each cDNA was diluted 10 times to 5 ng/μL for use in RT-qPCRs. RT-qPCR was performed using Power SYBR green PCR master mix (Thermo Fisher Scientific) and a StepOnePlus RT-qPCR instrument (Applied Biosystems), as described previously. Following RT-qPCR, relative mRNA expression was normalized to constitutive expression levels of the housekeeping recA RNA and then calculated using the comparative threshold cycle (2^−ΔΔCT) method (43, 44). RT-qPCR was also used to compare nanJ, nanH, nanR, ccpA, or codY gene expression in the wild type, nanJ or nanH null mutants, and their complementing strains; primers used for those analyses are also listed in Table 2.

CPE Western blotting.

Supernatants of 24-h TH or MDS cultures of specified strains were subjected to 12% SDS-PAGE, followed by Western blotting with CPE polyclonal antiserum, as previously described (25).

Evaluation of CPE-induced cytotoxicity and CPE binding/CH-1 complex formation.

Caco-2 cells were cultured and treated with CPE-supplemented, buffer-exchanged MDS culture supernatant, as described previously (30). Briefly, 24-h cultured supernatants of 01E809cpeKO, 01E809TKO, and 01E809TKOnanJcomp MDS cultures were buffer exchanged to Hanks balanced salt solution with calcium and magnesium without phenol red (HBSS buffer from Corning) using Thermo Scientific Zeba spin desalting columns. A 0.5-μg aliquot of purified CPE was then added to 1 mL of each buffer-exchanged MDS supernatant, and that mixture was applied for 1 h at 37°C to confluent Caco-2 cells that had been grown in 12-well cell culture plates. After this treatment, each culture supernatant was used for cytotoxicity detection with a cytotoxicity detection kit (LDH- release) purchased from Roche. Caco-2 cells gently removed from plates were resuspended in radioimmunoprecipitation assay (RIPA) buffer (Alfa Aesar) containing Benzonase (Millipore Sigma). Samples in 5× concentrated SDS loading buffer were electrophoresed on 6% SDS-acrylamide gels for CPE Western blotting. This CPE Western blotting detected CPE CH-1 complex formation (30).

To ensure that CPE was not produced by the cpe mutant, the triple cpe nanH nanJ null mutant or a nanJ complementing strain of that triple mutant, a 0.2-mL aliquot of an FTG culture of those strains was inoculated into 10 mL of MDS or TH media for 24 h at 37°C. Culture supernatants were then removed and mixed with 5× SDS loading buffer. These samples were loaded onto 12% SDS-acrylamide gels for CPE Western blotting, performed as described.

Statistical analyses.

Statistical analyses were performed using GraphPad Prism 8. To compare more than 2 samples, one-way analysis of variance (ANOVA) was used with post hoc analysis by Dunnett’s multiple-comparison test. To compare two samples, Student's t test was used. Significant differences had P value less than 0.05.

Data availability.

The nanJ gene of 01E809 was sequenced and deposited in GenBank under accession number OQ095367.