Application of serological typing to the investigation of outbreaks of Clostridium perfringens food poisoning, 1970-1978

Clostridium perfringens and Clostridium difficile in cooked beef sold in Côte d'Ivoire and their antimicrobial susceptibility

The aim of this study was to evaluate the prevalence of Clostridium difficile and Clostridium perfringens in cooked beef sold in the streets in Côte d'Ivoire and their antimicrobial susceptibility. A total of 395 kidney and flesh samples of cooked beef were collected from vendors at Abidjan and subjected to C. difficile and C. perfringens isolation and identification by using biochemical tests, API 20A system and PCR detection. Subsequently, the antimicrobial susceptibility test was performed for confirmed isolates. Our results showed the prevalence of 12.4% for C. difficile (11.04% in kidney and 13.45% in flesh) and 5.06% for C. perfringens (2.32% in kidney and 7.17% in flesh). Metronidazole and vancomycin remained the most potent antimicrobial agents against C. difficile while metronidazole and penicillin G were the most potent agents against C. perfringens. The resistance rates to tetracycline, doxycycline, chloramphenicol and erythromycin against C. difficile and C. perfringens isolates ranged from 2.05% to 8.16% and from 20% to 50%, respectively. Among all antimicrobial agents tested against C. difficile, percentages of resistance to quinolones ciprofloxacin, norfloxacin and nalidixic acid as well as to gentamicin and cefotaxime were the highest. Eight resistant phenotypes were defined for C. difficile isolates and eleven resistant phenotypes for C. perfringens isolates. Clindamycin/gentamicin/cefotaxime/ciprofloxacin/norfloxacin/nalidixic acid resistance was the most common phenotype for C. difficile (55.10% of isolates) while norfloxacin/nalidixic acid resistance was the most common phenotype for C. perfringens (20% of isolates).

Typing of Clostridium perfringens strains by use of Random Amplified Polymorphic DNA (RAPD) system in comparison with zymotyping

The definition of strain clonality postulates that strains showed identical phenotypic and genetic traits are likely to descend from a common ancestor even if they were isolated from different sources and locations. Regarding this definition, non-epidemiologically linked strains might be clonal strains. To overcome this ambiguity, the discriminatory capability of RAPD typing was assessed firstly on eight Clostridium perfringens strains proven to be chromosomally different with one being the mutant of another one. Thirteen primers were tested but only two were able to differentiate seven of the eight strains. With none of the used primers it was possible to differentiate the parental strain and its mutant harboring an insertion of 180 kb. The four most discriminant primers were retained to determine the RAPD fingerprints of a further 20 previously zymotyped strains from which seventeen were unrelated. To compare the two typing systems, the zymotype of the eight chromosomally different strains was determined. Thus, the discriminatory index was calculated on the basis of 25 unrelated C. perfringens strains. This was 0.97 with RAPD typing and 0.99 with zymotyping. From these results we conclude that the RAPD typing which is less fastidious than zymotyping can be used as an epidemiological marker for C. perfringens.

Multiple-locus variable-number tandem repeat analysis for strain typing of Clostridium perfringens

Clostridium perfringens is ubiquitous in the environment and causes diseases in man and animals, with syndromes ranging from enteritis, enterotoxemia, and sudden death to food poisoning and gas gangrene. Understanding the epidemiology of these infections and of the evolution of virulence in C. perfringens necessitate an efficient, time and cost effective strain typing method. Multiple-locus variable-number tandem repeat analysis (MLVA) has been applied to typing of other pathogens and we describe here the development of a MLVA scheme for C. perfringens. We characterized five variable tandem repeat (VNTR) loci, four of which are contained within protein encoding genes and screened 112 C. perfringens isolates to evaluate typability, reproducibility, and discriminatory power of the scheme. All the isolates were assigned a MLVA genotype and the technique has excellent reproducibility, with a numerical index of discrimination for the five VNTR loci of 0.995. Thus MLVA is an efficient tool for C. perfringens strain typing, and being PCR based makes it rapid, easy, and cost effective. In addition, it can be employed in epidemiological, ecological, and evolutionary investigations of the organism.

Control of Clostridium perfringens germination and outgrowth by buffered sodium citrate during chilling of roast beef and injected pork

Inhibition of the germination and outgrowth of Clostridium perfringens by buffered sodium citrate (Ional) and buffered sodium citrate supplemented with sodium diacetate (Ional Plus) during the abusive chilling of roast beef and injected pork was evaluated. Beef top rounds or pork loins were injected with a brine containing NaCl, potato starch, and potassium tetrapyrophosphate to yield final in-product concentrations of 0.85, 0.25, and 0.20%, respectively. Products were ground and mixed with Ional or Ional Plus at 0, 0.5, 1.0, and 2.0%. Each product was mixed with a three-strain C. perfringens spore cocktail to obtain final spore concentrations of ca. 2.5 log10 spores per g. Chilling of roast beef from 54.4 to 7.2 degrees C resulted in C. perfringens population increases of 1.51 and 5.27 log10 CFU/g for 18- and 21-h exponential chill rates, respectively, while chilling of injected pork resulted in increases of 3.70 and 4.41 log10 CFU/g. The incorporation of Ional into the roast beef formulation resulted in C. perfringens population reductions of 0.98, 1.87, and 2.47 log10 CFU/g with 0.5, 1.0, and 2.0% Ional, respectively, over 18 h of chilling, while > or = 1.0% Ional Plus was required to achieve similar reductions (reductions of 0.91 and 2.07 log10 CFU/g were obtained with 1.0 and 2.0% Ional Plus, respectively). An Ional or Ional Plus concentration of > or = 1.0% was required to reduce C. perfringens populations in roast beef or injected pork chilled from 54.4 to 7.2 degrees C in 21 h. Cooling times for roast beef or injected pork products after heat processing can be extended to 21 h through the incorporation of > or = 1.0% Ional or Ional Plus into the formulation to reduce the potential risk of C. perfringens germination and outgrowth.

Amplified fragment length polymorphism (AFLP) analysis of Clostridium novyi, C. perfringens and Bacillus cereus isolated from injecting drug users during 2000

As part of the follow-up investigations associated with an outbreak of severe illness and death among illegal injecting drug users during 2000, 43 cultures of Clostridium novyi type A, 40 C. perfringens type A and 6 isolates of Bacillus cereus were characterised by amplified fragment length polymorphism (AFLP) analysis. Among the 43 C. novyi isolates, 23 different AFLP profiles were detected. The same AFLP profile was detected in isolates from 18 drug users investigated during 2000 from Scotland, England, the Republic of Ireland and Norway and a wound from a patient in 2000 who was not identified as a drug user. Unique AFLP profiles were obtained from four drug users from England and the Republic of Ireland, 10 historical isolates from culture collections, an isolate from food (1989) and three isolates from wounds (1995, 1991, 1988). The 40 C. perfringens isolates were from 13 drug users, the contents of one syringe and two samples of heroin. Sixteen AFLP types of C. perfringens were distinguished and there was little evidence for commonality among the isolates. The AFLP types of C. perfringens from heroin differed and were unique. Six isolates of B. cereus were from four drug users and two samples of heroin. Four different AFLP patterns were distinguished. Three AFLP types were isolated from four drug users. B. cereus isolates from an aspirate and a heroin sample collected from the same drug user were identical, and were also indistinguishable from an isolate from a groin infection in a second drug user. The AFLP type of the isolate from a second and unrelated heroin sample was unique. The AFLP results showed no or very limited evidence for commonality between the different isolates of B. cereus and C. perfringens. In marked contrast, the C. novyi isolates from the majority of the drug users during 2000 were homogeneous, suggesting a common source or clonal selection of a C. novyi type, or both, which either had an adaptive advantage in spore germination, survival or growth following the drug preparation and the injection procedure, or produced a more severe clinical presentation.

Amplified fragment length polymorphism (AFLP) analysis of Clostridium perfringens for epidemiological typing

Thirty-five Clostridium perfringens isolates from patients and foods implicated in seven outbreaks of suspected Cl. perfringens food poisoning together with five unrelated incidents were analysed by serotyping and amplified fragment length polymorphism (AFLP). Despite minor band differences, AFLP was found to be highly reproducible and 16 different profiles (each unique to the 12 incidents) were recognised. The results from both serotyping and AFLP analysis identified exactly the same groups of related cultures. It is concluded that AFLP can provide a rapid, sensitive and reproducible method for the typing of Cl. perfringens for outbreak investigation.

Molecular subtyping of Clostridium perfringens by pulsed-field gel electrophoresis to facilitate food-borne-disease outbreak investigations

Clostridium perfringens is a common cause of food-borne illness. The illness is characterized by profuse diarrhea and acute abdominal pain. Since the illness is usually self-limiting, many cases are undiagnosed and/or not reported. Investigations are often pursued after an outbreak involving large numbers of people in institutions, at restaurants, or at catered meals. Serotyping has been used in the past to assist epidemiologic investigations of C. perfringens outbreaks. However, serotyping reagents are not widely available, and many isolates are often untypeable with existing reagents. We developed a pulsed-field gel electrophoresis (PFGE) method for molecular subtyping of C. perfringens isolates to aid in epidemiologic investigations of food-borne outbreaks. Six restriction endonucleases (SmaI, ApaI, FspI, MluI, KspI, and XbaI) were evaluated with a select panel of C. perfringens strains. SmaI was chosen for further studies because it produced 11 to 13 well-distributed bands of 40 to approximately 1,100 kb which provided good discrimination between isolates. Seventeen distinct patterns were obtained with 62 isolates from seven outbreak investigations or control strains. In general, multiple isolates from a single individual had indistinguishable PFGE patterns. Epidemiologically unrelated isolates (outbreak or control strains) had unique patterns; isolates from different individuals within an outbreak had similar, if not identical, patterns. PFGE identifies clonal relationships of isolates which will assist epidemiologic investigations of food-borne-disease outbreaks caused by C. perfringens.

Sandwich enzyme-linked immunosorbent assay by using monoclonal antibody for detection of Clostridium perfringens enterotoxin

Sandwich enzyme-linked immunosorbent assay (ELISA) was developed for the quantitative estimation of Clostridium perfringens enterotoxin (CPE) with monoclonal and polyclonal antibodies as capturing and detecting antibodies, respectively. The dose-dependent relationship between absorbance at 405 nm and concentration of purified CPE was obtained over the range of 0.64-400 ng/ml. The sandwich ELISA was fond to detect crude CPE in culture and CPE in 10% fecal extracts. This method is convenient, rapid and sensitive for specific detection of CPE.

Flow cytometric analysis for enterotoxin exposed on Clostridium perfringens spores

Flow cytometric method (FCM) with fluorescent-labeled anti-CPE antibody was applied to develop a rapid, specific, and convenient method to detect enterotoxin (CPE) exposed on the surface of spores of Clostridium perfringens. The results obtained indicate that FCM can specifically detect CPE exposed on C. perfringens spores for a short time. Thus, FCM is found to be a rapid, specific, and convenient assay method for detection of CPE exposed on C. perfringens spores, suggesting that it will be hopefully useful to diagnose food poisoning caused by C. perfringens.

Strain differentiation of Clostridium perfringens by bacteriocin typing, plasmid profiling and ribotyping

Bacteriocin typing, plasmid profiling and ribotyping were used to type 34 food and patient Clostridium perfringens isolates from 10 food poisoning cases, respectively, outbreaks. In nine cases/outbreaks bacteriocin patterns showed identical main groups. Subgroups differed within all cases/outbreaks. Plasmid profiles were identical for all isolates within each of three outbreaks. In eight food poisoning cases and outbreaks, all the ribotypes of each food and stool isolate were found to be identical. All three typing methods give valuable results for the characterization of C. perfringens beyond the species level. Bacteriocin typing represents a suitable addition to plasmid typing, particularly since the results do not show any correlation between losses of plasmids and changes in bacteriocin sensitivity patterns. Ribotyping was found to be a suitable tool to determine the genetic relationship of C. perfringens isolates in the context of food-borne poisoning.

Structure of the capsular polysaccharide of Clostridium perfringens Hobbs 10 determined by NMR spectroscopy

The complete primary structure of the type-specific capsular polysaccharide of Clostridium perfringens Hobbs 10 was determined. The polysaccharide was isolated from C. perfringens Hobbs 10 by cold-water extraction of whole, heavily encapsulated cells. The polysaccharide was purified, by ethanol precipitation, deproteination, selective precipitation with hexadecyltrimethylammonium bromide, ion-exchange chromatography and gel-filtration chromatography. The polysaccharide was comprised of D-glucose, D-galactose, N-acetylgalactosamine, and iduronic acid, in molar ratios of 2:2:1:1. Sequence and linkage assignments of the glycosyl residues were obtained by NMR spectroscopy, specifically by the combination of two-dimensional homonuclear DQF-COSY, TQF-COSY and TOCSY, heteronuclear ¿1H, 13C¿ single-quantum coherence (HSQC) and heteronuclear multiple-bond correlation (HMBC) experiments. The capsular polysaccharide of C. perfringens Hobbs 10 is a polymer composed of a hexasaccharide repeating unit with the following structure: [formula: see text] This structure is novel among bacterial cell-surface polysaccharides, and it is only the second of many serotypically distinct capsular polysaccharides of C. perfringens to be described.

Structure of the capsular polysaccharide of Clostridium perfringens Hobbs 5 as determined by NMR spectroscopy

The complete primary structure of the capsular polysaccharide of Clostridium perfringens Hobbs 5, an anaerobic bacterium implicated in food poisoning, was determined. The polysaccharide was isolated from C. perfringens Hobbs 5 cells, after deproteination, by ethanol precipitation and by ion-exchange chromatography. The polysaccharide was comprised of glucose, galactose, mannose, N-acetylglucosamine, N-acetylgalactosamine, and glucuronic acid, in equimolar ratios. Sequence and linkage assignments of the glycosyl residues were obtained by NMR spectroscopy, specifically by the combination of two-dimensional homonuclear TOCSY and NOESY experiments and heteronuclear (1H, 13C) multiple-quantum coherence (HMQC, HMQC-COSY, HMQC-TOCSY and HMBC) experiments. Thus, the envelope polysaccharide of C. perfringens Hobbs 5 was found to be a polymer composed of a hexasaccharide repeating unit with the following structure: [formula: see text] This structure is novel among bacterial cell-surface polysaccharides, and it is the first of many serotypically distinct capsular polysaccharides of C. perfringens to be described.

Nosocomial diarrhoea in the elderly due to enterotoxigenic Clostridium perfringens

To diagnose sporadic diarrhoea due to Clostridium perfringens infection, faecal specimens from elderly patients were examined directly for C.perfringens enterotoxin using reverse passive latex agglutination assay, and then cultured for this organism. C.perfringens isolates from those samples were grouped by slide agglutination and by pulsed-field gel electrophoresis (PFGE). Fifty of the 60 isolates agglutinated with newly raised antiserum WX2 and 38 shared the same genomic PFGE pattern. Characteristics of the epidemics and experimental data suggest that the diarrhoea was caused by a nosocomial spread of C.perfringens, and not by a food-borne outbreak.

Plasmid profiling for strain differentiation and characterization of Clostridium perfringens isolates

Plasmid profiling was used for the characterization of 13 Clostridium perfringens collection strains as well as 85 clinical and food isolates. Methodological details and limitations of the plasmid isolation procedure are outlined and discussed. For 19 clinical isolates obtained during seven group outbreaks, a close connection between at least some strains from each individual outbreak was established for six of them; only results from one outbreak were completely inconclusive, due to missing plasmids in one of the two isolates. The presence of multiple plasmid types within seven out of 12 given food samples, from which at least two C. perfringens isolates were obtained, indicates the importance of multiple isolates for meaningful typing results in epidemiological investigations. By including results from a previous report from this laboratory, baseline data on plasmid profiles for a total of 133 isolates are provided. The results of this study revealed that 36% of the food isolates unrelated to disease outbreaks carried no plasmids, as compared with 19-25% among disease-related isolates. A high prevalence (24.8%) of a 8.9 +/- 0.5 MDa plasmid was found among the 133 isolates, which contributed to one of four occurrences of identical plasmid profiles among strains that were initially considered unrelated. Two of these identical plasmid profiles were found among strains from the same culture collection, indicating the possibility of a common ancestor strain or cross-contamination.

Use of plasmid profiling as a typing method for epidemiologically related Clostridium perfringens isolates from food poisoning cases and outbreaks

Plasmid profiling was used for the characterization of Clostridium perfringens isolates involved in disease outbreaks. The usefulness of this technique was demonstrated by the retrospective examination of food and patient isolates from 10 cases and outbreaks from 1984 to 1991. The origin of three outbreaks could be clearly confirmed due to identical plasmid profiles in all isolates. In one outbreak identical plasmid patterns were found between one food and one patient isolates, while one plasmid was missing in the second patient isolate. In an additional two cases a relationship between food and patient isolates is likely, if the possibility of the loss of one plasmid in one of the isolated strains is considered. In one outbreak two faecal isolates could be related to an isolate from one of the two foods implicated as outbreak source; isolates from the other food and a third faecal sample could not be linked to any other isolate. The results from three outbreaks were largely inconclusive because plasmids were not present either in all or in some of the isolates.

Multilocus enzyme typing of human and animal strains of Clostridium perfringens

Multilocus enzyme electrophoresis was developed to evaluate the genetic diversity of 71 human strains and 17 animal strains of Clostridium perfringens. Crude protein extracts, obtained by sonication of washed bacteria, were analyzed by polyacrylamide-agarose gel electrophoresis to characterize electrophoretic mobility variants of seven enzymes (esterase, glutamate dehydrogenase, glutamic-oxaloacetic transaminase, nucleoside phosphorylase, phosphoglucose isomerase, phosphoglucomutase, threonine dehydrogenase). Genetic diversity of the enzyme loci ranged from 0.340 to 0.813. Sixty-nine electrophoretic types were described among the 88 strains tested and the index of discrimination was 0.994. All strains were typable, and epidemiological relationships between isolates could be established. This method showed a fair correlation with esterase electrophoretic typing based on hydrolytic and electrophoretic polymorphism of esterases. This work demonstrates that multilocus enzyme polymorphism is a reliable and discriminant marker of genetic diversity of strains of C. perfringens.

Esterase electrophoretic polymorphism of human and animal strains of Clostridium perfringens

Esterase electrophoretic polymorphism in human and animal strains of Clostridium perfringens was studied by using polyacrylamide-agarose gel electrophoresis. Five types of esterases, designated E-I to E-V and defined by their hydrolytic specificities toward five synthetic substrates, were found in protein extracts of bacteria grown without glucose (glucose-containing media allowed only the expression of esterase E-I). Mobility variants of esterase E-I, which hydrolyzes alpha- and beta-naphthyl acetates and butyrates, were used as a basis for the distribution of strains into 11 zymogroups. When all five types of esterases and their electrophoretic variants were considered, 77 electrophoretic types (ETs) could be described for the 89 strains tested. Animal strains did not constitute a distinctive subpopulation, as revealed by their distribution in the zymogroups and by clustering analysis. Statistical analysis also emphasized the importance of esterase E-IV (which hydrolyzes only naphthyl acetates) and esterase E-V (which hydrolyzes only alpha-naphthyl acetate) in clustering by the relatedness of the ETs. ETs allowed the epidemiological characterization of stool isolates recovered from elderly inpatient residents and from adolescent chronic-care psychiatric patients. These results indicate that esterase electrophoretic typing may be a marker for epidemiological and ecological analyses.

Application of pyrolysis mass spectrometry to the investigation of outbreaks of food poisoning and non-gastrointestinal infection associated with Bacillus species and Clostridium perfringens

Eighteen isolates of Bacillus species and 15 of Clostridium perfringens, all of which had been associated with outbreaks of either food poisoning or non-gastrointestinal infection (NGI), were examined for relatedness by pyrolysis mass spectrometry (PyMS). The PyMS-analysis correctly clustered all the groups of epidemiologically related isolates of both genera, and distinguished all the single, epidemiologically unrelated isolates of the same species. PyMS is a simple, rapid and inexpensive technique which can provide useful and accurate inter-strain comparisons within both the Bacillus and Clostridium genera in complete accord with conventional serological typing results.

An outbreak of enterotoxaemia in cats

Eighteen local breed cats suffered from enterotoxaemia after scavenging on chicken-remnants. They suffered from vomitus and diarrhoea and collapsed dead after a few hours. An untypable strain of Clostridium perfringens was isolated in pure form and was suspected to be the cause of the outbreak.

Multiple typing techniques applied to a Clostridium perfringens food poisoning outbreak

Twenty-one stool specimens obtained from persons implicated in two food poisoning outbreaks at the same institution in Smith Falls, Ontario, were examined for Clostridium perfringens. Ninety-two colonies of Cl. perfringens (3-5 per stool specimen) were typed with antisera, bacteriocins and by plasmid analysis. They were also tested for the in vitro production of bacteriocin and enterotoxin. Sixteen of the 21 stool specimens were tested directly for enterotoxin. This was detected in 13, five of which were from individuals listed as 'asymptomatic' food handlers. The predominant strain isolated from 15 of the 21 stool samples produced bacteriocin and enterotoxin in vitro, contained no plasmids, and was of a common bacteriocin type and serotype.

An investigation into Clostridium perfringens enterotoxin-associated diarrhoea

An investigation was carried out to establish the incidence of diarrhoea associated with the presence of enterotoxigenic Clostridium perfringens. The results indicate a high risk group, namely elderly hospitalized patients, who should be investigated for this organism in a similar way to Clostridium difficile if symptoms occur. The significance of antibiotic association is suggested and cross-infection was shown to be a possibility. Detection of the enterotoxin was accompanied in all cases by the presence of high faecal counts of enterotoxigenic strains of C. perfringens.

Competitive erythroimmunoassay for detecting Clostridium perfringens type A enterotoxin in stool specimens

A competitive erythroimmunoassay (ERIA) is described for Clostridium perfringens enterotoxin (CPE) detection in stools. This technique uses sheep red blood cells sensitized by CPE and an anti-CPE-antibody-coated plate in which the results are read by eye. ERIA is simple, rapid, economic and more sensitive (2 ng/ml) than the enzyme-linked immunosorbent assay used for evaluation. ERIA is suitable for CPE detection in stool samples protected with phenylmethylsulphonylfluoride.

Analysis of plasmid profiling as a method for rapid differentiation of food-associated Clostridium perfringens strains

Plasmid analysis of over 120 strains of Clostridium perfringens, isolated during food-poisoning incidents and from animal carcasses and food constituents with no association with food poisoning, showed the potential of plasmid profiling as a means of differentiating epidemiologically related strains. On average 65% of freshly isolated strains contained one or more plasmids which could be used in the analysis. Comparison of profiles of strains from unrelated sources or unrelated strains from the same source showed a particularly wide variety of plasmid profiles. Thus the possibility that epidemiologically-unrelated strains might possess similar profiles appears to be very low in this organism. Analysis of serologically-related strains from the same source revealed similar plasmid profiles in all the plasmid-bearing strains examined. A high proportion (71%) of fresh and well-characterized food-poisoning strains possessed plasmids of 6.2 kb in size (compared with 19% of non-food-poisoning strains). The possible role of these plasmids is discussed, since the structural gene encoding the enterotoxin type A was not present on any of the plasmids in the food-poisoning strains tested.

Clostridium perfringens food poisoning: use of serotyping in an outbreak setting

An outbreak of Clostridium perfringens food poisoning occurred among attendees of a firehouse luncheon. The predominant symptoms of diarrhea (100%) and abdominal pain (81%) among case-patients, the mean incubation period (13.4 h), and the mean duration of illness (21.2 h) were all characteristic of C. perfringens enteritis. Roast beef, although not epidemiologically implicated, was the most likely vehicle of transmission. Fecal specimens from case-patients contained a median C. perfringens spore count of greater than 10(6) and yielded isolates that were heat sensitive and predominantly nonhemolytic, produced C. perfringens enterotoxin A, and, in the majority of specimens (four of five), were identical in serotype. Food samples were negative. This outbreak demonstrates that following enumeration of C. perfringens from a suitable number of fecal specimens from case-patients, serotyping of the isolates may be helpful in implicating C. perfringens as the cause of foodborne illness. This is especially true when implicated food items test negative or are no longer available for testing.

Plasmid analysis as a means of strain differentiation in Clostridium perfringens

A total of 114 Clostridium perfringens isolates were serotyped and examined for plasmids. Fifty-two strains were from hospitalized patients with diarrhea or from hospital environments, and 62 epidemiologically unrelated isolates were obtained from food poisoning outbreaks. All strains were screened for bacteriocin production against a common indicator strain of C. perfringens. In the one significant hospital outbreak of C. perfringens diarrhea, three to five plasmid types were found in strains of the predominant serotype, but no similar correlation between serotype and plasmid type was found in random isolates from a variety of sources. All of the strains associated with the diarrhea outbreak produced bacteriocins, whereas 63% of the strains from various sources produced bacteriocins. The typing data suggest a promising differentiating capability for plasmid analysis in the epidemiological study of outbreaks of food poisoning, diarrhea, or infections caused by C. perfringens.

A double antibody sandwich enzyme-immunoassay for Clostridium perfringens type A enterotoxin detection in stool specimens

A double antibody sandwich enzyme-immunoassay has been developed for detection of Clostridium perfringens enterotoxin. Anti-enterotoxin immunoglobulin G-alkaline phosphatase conjugates were prepared using a rapid minicolumn procedure. The assay can achieve a sensitivity of greater than or equal to 1 ng/ml with purified enterotoxin. Sensitivity for detection of cases of C. perfringens enteritis in a C. perfringens outbreak (86 individuals tested) was between 85.7 and 98.0 per cent depending upon stringency of criteria for defining positive cases. Specificity of the assay was demonstrated by the lack of positive results in 53 individuals involved in a gastroenteritis outbreak of unknown etiology.

Faecal carriage of Clostridium perfringens

The numbers and serotypes of Clostridium perfringens present in the faeces of three groups of hospital patients and young healthy laboratory workers were examined in studies lasting between 10 and 13 weeks. In one hospital some long-stay geriatric patients carried relatively high numbers of C. perfringens (greater than 10(7)/g) most of the time and it was not unusual in any one week for the majority of these patients to carry the same serotype(s). However, the numbers of C. perfringens in the faeces of young long-stay patients in the same hospital were in the range of 10(3)-10(4)/g and carriage of common serotypes was not observed. These results were similar to the findings with the young laboratory workers. This investigation indicates that two of the laboratory criteria often used in the investigation of C. perfringens food poisoning, i.e. faecal counts of greater than or equal to 10(5) C. perfringens/g and patients carrying the same serological type need to be interpreted with caution with suspected outbreaks involving some groups of geriatric long-stay hospital patients.

Heat resistance, spore germination, and enterotoxigenicity of Clostridium perfringens

Heat resistance at 95 C, heat activation at 75 C, and germination response were determined for spores of 10 serotype strains of Clostridium perfringens type A, including five heat-resistant and five heat-sensitive strains. The D95-values ranged from 17.6 to 63.0 and from 1.3 to 2.8 for the heat-resistant and the heat-sensitive strains, respectively. The heat-activation values, the ratios between the heated and unheated viable counts of spore suspensions, ranged from 0.0035 to 0.65 and from 6.5 to 60.0 for the heat-sensitive and the heat-resistant strains, respectively. Spores of these strains were divided into two distinct germination types on the basis of their germination response; spores of the heat-resistant strains germinated in KC1 medium after heat activation (K-type), and spores of the heat-sensitive strains germinated in a mixture of L-alanine, inosine, and CaCl2 in the presence of CO2 without heat activation (A-type). The strains were tested for enterotoxigenicity by a reversed passive latex-agglutination (RPLA) test. All the heat-resistant strains were RPLA-positive, whereas the heat-sensitive strains were all RPLA-negative. A total of 37 strains of the organism isolated from food-poisoning outbreaks were tested for spore germination and enterotoxin formation. All of the 20 heat-resistant strains showed K-type spore germination and, except for three strains, were RPLA-positive, whereas all of the 17 heat-sensitive strains showed A-type spore germination and, except for only one strain, were RPLA-negative.

The assessment and application of a bacteriocin typing scheme for Clostridium perfringens

A collection of 50 bacteriocins was assembled and used to type 802 isolates of Clostridium perfringens from food poisoning outbreaks and a variety of other sources. It was found that strains of the same serotype within an outbreak showed similar patterns of susceptibility to bacteriocins, and the use of "one difference' rule is proposed for interpretation of the typing patterns of epidemiologically related strains. Isolates of different serotype or of the same serotype isolated from different sources produced many variations in bacteriocin susceptibility patterns. Two computer programs were developed to assist in the interpretation of bacteriocin typing patterns. Their use showed that related and unrelated strains formed different clusters and enabled a range of the 20 most discriminatory bacteriocins to be selected. Isolates of C. perfringens from a wide range of sources were screened for their ability to produce bacteriocins. A much greater proportion of the strains from food poisoning outbreaks was bacteriocinogenic than were isolates from human and animal infections, various foods and the environment. The relevance of these findings to the occurrence of C. perfringens food poisoning is discussed.

Development and application of an enzyme linked immunosorbent assay for Clostridium perfringens type A enterotoxin

An enzyme linked immunosorbent assay (ELISA) has been developed to quantitate faecal Clostridium perfringens enterotoxin in the investigation of C perfringens food poisoning. The sandwich ELISA could be carried out in 24 h and was sensitive enough to detect as little as 5 ng/g of enterotoxin in faeces. Specificity of the assay was shown by comparing results with those obtained from other standard toxin assays, such as double gel diffusion and counterimmunoelectrophoresis, and by the assay of faecal material from control groups. By means of the ELISA method, 515 faecal samples from 50 separate outbreaks of C perfringens food poisoning were examined, together with 21 food samples from 12 of the outbreaks. A clear distinction was noted between faecal samples collected on the first two days of an outbreak, where 77% were enterotoxin positive, and those specimens collected later than the second day, when only 33% had detectable enterotoxin. The ELISA is recommended as a valuable tool in the investigation of C perfringens foodborne illness.

Analysis of the common polysaccharide antigens from the cell envelope of Clostridium perfringens type A

The major, common antigen of Clostridium perfringens type A, isolated and purified independently from three selected strains (Hobbs 5, Hobbs 9, and Hobbs 10), is composed of equimolar amounts of 2-acetamido-2-deoxy-D-mannose (Man-NAc) and 2-acetamido-2-deoxy-D-glucose (GlcNAc). The purified antigen gave a strong immunoprecipitin line by double immunodiffusion in gel. Smith degradation of the major, common antigen caused decomposition of all of the GlcNAc, without concomitant loss in ManNAc, or a perceptible change in serological activity. Therefore, the serological activity of the major, common antigen depended solely on the presence of ManNAc. Data obtained by the 13C-n.m.r.-spectral analysis of the Smith-degradation product revealed that it was a linear-backbone polysaccharide analogous to a Rhodotorula glutinis mannan, but composed of pairs of 2-acetamido-2-deoxymannopyranosyl residues alternately linked beta-(1 leads to 3) and beta-(1 leads to 4). The one-bond, carbon-hydrogen coupling-constant of 162 Hz for both anomeric centers was consistent with the proposed beta-linkages. A similar, 13C-n.m.r.-spectral analysis of the native, common antigen indicated that the GlcNAc residues were randomly connected to three of the four hydroxyl groups not already involved in linking the ManNAc backbone, the 4-hydroxyl group being the exception. A second, serologically inactive, polysaccharide composed of rhamnose, GalNAc, and galactose was identified, but not obtained in homogeneous state. The rhamnosyl residues were probably situated as nonreducing antennae, as they were quantitatively removed by Smith degradation without concomitant decomposition of the polymeric structure of the remaining residues.

Common polysaccharide antigens from the cell envelope of Clostridium perfringens type A

Soluble antigens were obtained by extracting five serotype strains of Clostridium perfringens type A with water at 100 degrees C. The type-specific polysaccharides were precipitated with ethanol, and the common antigens were recovered from the ethanol supernatants by concentration, dialysis, and lyophilization. Refluxing the water-extracted cell residues with 1% acetic acid followed by concentration, dialysis, and lyophilization gave additional common antigen fractions. A comprehensive, side-by-side comparison of the antigen fractions, the ethanol precipitate, the ethanol supernatant, and the acetic acid supernatant, revealed that common antigens were recovered in all three fractions, and that three distinct entities were responsible for the formation of the observed common immunoprecipitin lines; whereas many fractions possessed all three immunoprecipitin lines, others contained only one or two. The serological homology observed between the various antigen fractions was apparently a consequence of N-acetylglucosamine- and N-acetylmannosamine-containing polymers. The common antigens were presumably associated with the cell envelope and may be the type of markers sought previously by others for the serological identification of C. perfringens.