Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) Hybridization of 2,659 Clostridium perfringens isolates with gene probes for seven toxins (α, β, ε, ι, θ, µ, and enterotoxin) and for sialidase Georges Daube, DVM, DVSc; Patricia Simon; Bernard Limbourg, DVM; Christophe Manteca, DVM; Jacques Mainil, DVM, DVSc; Albert Kaeckenbeeck, DVM From the Department of Bacteriology, Faculty of Veterinary Medicine, University of Liege, Sart-Tilman, Bat B43, Liege, B-4000 (Daube, Simon, Manteca, Mainil, Kaeckenbeeck), and Fédération de Lutte contre les Maladies Animales, 604 Chaussée de Marche, Er-pent, B-5101 (Limbourg), Belgium. Supported by Institut pour l'encouragement de la Recherche Scientifique dans l'Industrie et l'Agriculture grant No. 5313, 5452, and 5551 and by a grant from the Department of Agriculture of the Région Wallonne. This report represents a portion of a thesis submitted by the senior author to the university as partial fulfillment of the requirement for the Docteur en Sciences Vétérinaires degree. The authors thank L. A. Devriese for providing Clostridium perfringens isolates and M.-P. Marot, K. Renier, N. Collin, and V. Pirson for technical assistance. Objective — To genetically characterize Clostridium perfringens isolates for association of pathologic type with various diseases. Design — Prospective study. Sample Population — 2,659 C perfringens isolates from various nonhuman animals species, human beings, and foods. Procedure — Colony hybridization with DNA probes for 7 toxin (α, β, ε, ι (subunits a and b), θ, µ, and enterotoxin) genes and 1 sialidase gene were performed to group the isolates by pathologic type. Results — Enterotoxin-negative type-A isolates were the most common (2,575/2,659), were isolated from all sources, and were separated into 5 pathologic types. In cattle and horses with enterotoxemia, essentially only these pathologic types were identified. The enterotoxin-negative isolates of types C or D each had a single pathologic type. Type-C isolates were isolated only from swine with necrotic enteritis and type-D isolates from small ruminants with enterotoxemia, except that 1 type-D isolate was also found from a healthy fish. Type-B or E isolates were not found. Among the 47 enterotoxin-positive isolates, 5 isolates from sheep or deer were type D and the other 42 were type A. These 42 isolates were grouped into 3 pathologic types: 1 type was isolated from samples of almost all origins, but the other 2 types were found in only 5 fish, 4 human beings, and 1 dog. Conclusions and Clinical Relevance — Genetic characterization of these isolates allowed identification of 11 different pathologic types. This approach may be useful in molecular diagnosis and prophylaxis of clostridial disease. (Am J Vet Res 1996;57:496-501) Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) In recent years, rapid advances in the genetics of Clostridium perfringens virulence factors1 have allowed performance of extensive surveys of the strains isolated in various diseases. 2 The 4 major lethal toxins (α, β, ε, and ι) and enterotoxin are the most important factors in inducing C perfringens-related disease,3 but the other 7 minor toxins (δ, θ, Κ, λ, µ, v, and sialidase) also may participate in specific disease syndromes.4-6 However, except for θ-toxin,7 direct evidence for pathogenesis has not been reported. Type-A C perfringens produces αtoxin; causes gas gangrene in human beings, colitis in horses, and necrotic enteritis in birds; and is the more common type isolated from soil and sewage. The enterotoxigenic type-A isolates induce food poisoning in human beings. Type-B isolates, producing α-, β-, and ε-toxins, and type-C isolates, producing α- and β-toxins, are responsible for necrotic enteritis in young animals of most species and in human beings, in which enterotoxigenic type-C isolates are involved.8 Type-D isolates, producing α- and ε-toxins, cause enterotoxemia, especially in small ruminants, but this type also has been described in other species. Type-E isolates, producing α- and ι-toxins, have rarely been isolated and have been involved in necrotic enteritis in cattle. The purpose of the study reported here was to examine the pathologic types of clinical isolates of C perfringens via colony hybridization with 9 probes specific for the major toxins. Materials and Methods Gene probe — Six DNA probes were derived from multicopy recombinant plasmids carrying toxin- or sialidaseencoding genes (Table l).2 Plasmid DNA was isolated by ultracentrifugation in cesium chloride gradients as described.9 The other 3 gene probes were obtained by polymerase chain reaction. 2 The fragment of the subunit b gene of the ι-toxin was obtained with the primers GATGAGAGCTGTGTTGAACTT and TAGAAGTTGAAGGATCCTTTG. The DNA template was extracted from the strain NCIB 10748, according to the method previously described.2,10 Clostridium perfringens isolates and DNA hybridization — Isolates of C perfringens (n = 2,659) from the collection of our laboratory were studied. The isolates had been collected between 1989 and 1993, and most of them were isolated from human beings or nonhuman animals having digestive tract disorders (Table 2). Most bovine, equine, or small ruminant isolates were from animals that had died of enterotoxemia. Samples from human beings, dogs, cats, and swine had been obtained from diagnostic laboratories; most of them had been submitted because the human being or animal had diarrhea, although clostridial disease had not been suspected. Eight isolates from 8 cases of necrotic enteritis in pigs also were included.a Isolates from fish and fowl were obtained from healthy animals, except for 6 chickens with diarrhea. The 45 food isolates were cultured from samples of foodstuffs of animal origin, especially meat, fish, or shellfish, without any association with an episode of food poisoning. One to 17 isolates (mean = 3) were purified from each of the 769 samples. Isolates of C perfringens were identified on blood agar platesb after incubation for 18 hours in anaerobic conditions (80% N2, 10% CO2, 10% H2). On the basis of colony appearances, hemolysis profile, and gram-stained smears, colonies considered to be C perfringens were sub-cultured for 24 hours anaerobically on agarc for lecithinase (α-toxin) production and sulfite reduction. Identification of the sulfite-reducing colonies was confirmed by disk method, using disks d to detect urease and indol production and glucose, lactose, sucrose, trehalose, and mannose fermentation. The toxin type of 133 isolates was obtained by use of classical mouse lethality assays.11,e The DNA colony hybridizations were performed as previously described. 2,f Nineteen C perfringens strains were used as controls in hybridization.2 Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) Table 1—Derivation of the 9 DNA probes used to classify isolates of Clostridium perfringens Probe Origin Endonuclease* Size (base pairs) Reference No. α β pTOX5† PCR fragment pGDS7† PCR fragment PCR fragment pLWL10† pRT1B† pNAG-S3† Unnamed† BamH1 and HindIII NA 950 1,025 17 2,21 HindIII NA 505 297 24 2,49 NA 331 49 Kpnl and HindIII Accl HindIII Kpnl and HindIII 504 736 1,300 1,400 50 51 52 53 ε ιa ιb E θ µ s *Used to derive the gene probe fragment from the plasmid; †plasmid. PCR = polymerase chain reaction; NA = not applicable; E = enterotoxin; and S = sialidase. Table 2—Origin of isolates of C perfringens Species No. of isolates No. of samples Cattle Human beings Sheep Dogs Cats Fish Swine Goats Fowl Deer Horses Other† Food‡ Total 1,443 224 219 200 171 77 63 55 42 29 25 66 45 2,659 390 46 63 35 22 16 48 20 40 8 8 41 32 769 Associated disease* (No. of samples) Enterotoxemia (360) Diarrhea (37) Enterotoxemia (63) Diarrhea (14) Diarrhea (4) None Diarrhea (15), necrotic enteritis (10) Enterotoxemia (19) Diarrhea (6) Enterotoxemia (8) Enterotoxemia (8) NA NA … *Other samples were obtained from animals without signs of clostridial disease. †Other animal species or unknown origin. ‡Of animal origin for human consumption. Table 3—Distribution of probe genes among isolates of C perfringens Origin Cattle (n = 1,443) Human beings (n = 224) Sheep (n = 219) Dogs (n = 200) Cats (n = 171) Fish (n = 77) Swine (n = 63) Goats (n = 55) Fowl (n = 42) Deer (n = 29) Horses (n = 25) Other* (n = 66) Foodt (n = 45) Total (n = 2,659) Hybridizing probe (%) α β ε ιa ιb E θ µ S 100 100 100 100 100 100 100 100 100 100 100 100 100 100 0 0 0 0 0 0 6.4 0 0 0 0 0 0 0.2 0 0 9.1 0 0 1.3 0 20 0 6.9 0 3 0 1.4 0 0 0 0 1.2 0 0 0 0 0 0 0 0 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.4 5.8 2.3 2.5 1.25 13 1.6 3.6 0 6.9 4 0 0 1.8 91.1 71.9 95 66.5 84.2 62.3 84.1 100 90.5 100 96 92.4 91.1 86.9 92.9 94.2 96.8 88 86.5 92.2 100 100 100 100 100 98.5 93.3 93.2 100 100 100 100 100 100 100 100 100 100 100 100 100 100 *Other animal species or unknown origin. tOf animal origin for human consumption. Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) Table 4—Comparison of mouse lethality assay and DNA-DNA hybridization results in 133 isolates of C perfringens Probe α β ε ιa ιb θ µ E S Total Nonlethal 14 0 0 0 0 14 13 0 14 14 Type A 96 0 0 1 0 88 93 8 96 96 Lethal TypeC 4 4 0 0 0 4 4 0 4 4 TypeD 19 0 19 0 0 19 19 2 19 19 Values are No. of isolates that hybridized with each probe. Results All 2,659 isolates hybridized with α-toxin and sialidase probes (Table 3). Moreover, those probes hybridized with lethal and nonlethal isolates (Table 4). Only 4 isolates from swine with necrotic enteritis hybridized with the β-toxin probe; those isolates were lethal for mice. Only isolates from sheep, goats, and deer hybridized with the ε-toxin gene probe; all of those tested isolates were lethal for mice. Two isolates from a healthy cat hybridized with the ιa-toxin (ADP-ribosyltransferase) probe, but not with the ιb-toxin probe. One of those isolates was classified as toxin type A by use of the lethality test. Enterotoxin probe hybridized with isolates from most origins, but such hybridization was more common in isolates from human beings or from fish than in those from other species. None of the 45 human food isolates hybridized with the enterotoxin gene probe. The θ-toxin gene was detected in 2,311 of the 2,659 isolates. Isolates that did not hybridize with this probe were more commonly from carnivorous (cats and dogs), aquatic (fish), or omnivorous (human beings and swine) species than from herbivorous animals (ruminants and horses). The same tendency was observed for the µ-toxin gene, but was less marked. The 2,659 isolates could be classified into 11 different pathologic types (Table 5). The most common pathologic type (2,093/2,659 isolates) was characterized by hybridization with only the α-, θ-, and µ-toxins and sialidase probes. This pathologic type corresponded to the common nonenterotoxigenic type-A strains. This toxin type also included 4 other pathologic types, 3 with different θ- and µ-toxin results and 1 that hybridized with the ιatoxin gene probe. The nonenterotoxigenic type-C and -D isolates hybridized with the θ- and µ-toxin probes. The enterotoxigenic isolates were divided into 4 pathologic types: toxin type-A or -D isolates hybridizing with the θ- and µ-toxin probes; type-A isolates hybridizing with the µ-toxin probe, but not with the θ-toxin probe; and type-A isolates not hybridizing with θ- and µ-toxin probes. In cattle, even those with enterotoxemia, only pathologic types corresponding to toxin type A were isolated (Table 6). Enterotoxigenic isolates were obtained from only 4 of the 390 studied cattle. In small ruminants and deer with enterotoxemia, the pathologic types corresponding to toxin type D were common (24/90 isolates). The enterotoxigenic type-D isolates were only from sheep and deer. In swine, dogs, cats, horses, and fowl, except for 4 type-C isolates from swine with necrotic enteritis, all isolates were toxin type A. The isolates of enterotoxigenic type A that did not hybridize with the θ-toxin probe were obtained only from human beings, fish, and 1 dog, whereas isolates of that type that hybridized with the θ-toxin probe were obtained from nearly all species. Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) Table 5—Distribution of pathologic types among the 2,659 isolates of C perfringens Hybridizing probes (%) Origin Cattle (n = 1,443) Human beings (n = 224) Sheep (n = 219) Dogs (n = 200) Cats (n = 171) Fish (n = 77) Swine (n = 63) Goats (n = 55) Fowl (n = 42) Deer (n = 29) Horses (n = 25) Other* (n = 66) Foodt (n = 45) Total (n = 2,659) α,S 1.1 2.2 0.5 4.5 0 2.6 0 0 0 0 0 0 0 1.2 α,θ,S 6.1 3.1 2.7 7.5 13.5 5.2 0 0 0 0 0 1.5 6.7 5.5 α,µ,S 7.8 23.7 4.6 28.5 15.8 24.7 15.9 0 9.5 0 4 7.6 8.9 11.4 α,θ,µ,S 84.6 65.2 82.6 57 68.4 53.2 76.2 76.4 90.5 89.7 92 87.9 84.4 78.7 α,ιa,θ,µ,S 0 0 0 0 1.2 0 0 0 0 0 0 0 0 0.1 α,E,S 0 0.4 0 0 0 0 0 0 0 0 0 0 0 0.04 α,µ,E,S 0 1.8 0 0.5 0 10.4 0 0 0 0 0 0 0 0.5 α,θ,µ,E,S 0.4 3.6 0.5 2 1.2 2.6 1.6 3.6 0 3.4 4 0 0 1.1 α,β,θ,µ,S 0 0 0 0 0 0 6.3 0 0 0 0 0 0 0.2 α,ε,θ,µ,S 0 0 7.3 0 0 1.3 0 20 0 3.4 0 3 0 1.2 α,ε,θ,µ,E,S 0 0 1.8 0 0 0 0 0 0 3.4 0 0 0 0.2 See Table 3 for key. Table 6—Main associations between pathologic types and origin of isolates of C perfringens Disease status Origin Enterotoxemia Cattle (n = 360) Sheep (n = 63) Goats (n = 19) Deer (n = 8) Horses (n = 8) Necrotic Swine (n = 8) enteritis None Human beings (n suspected = 46) Fish (n = 16) Dogs (n = 35) Cats (n = 22) Cattle (n = 30) Swine (n = 40) Fowl (n = 40) Food* (n = 32) α,S 2.5 1.6 0 0 0 0 Hybridizing probes {% of cases) Type A TypeC α,θ,S α,µ,S α,θ,µ,S α,ιa,µ,S α,E,S α,µ,E,S α,θ,µ,E,S α,β,θ,µ,S 15 15.3 94.4 0 0 0 1.1 0 7.9 11.1 85.7 0 0 0 1.6 0 0 0 52.6 0 0 0 5.3 0 0 0 100 0 0 0 12.5 0 0 12.5 100 0 0 0 12.5 0 0 0 60 0 0 0 0 40 TypeD α,ε,θ,µ,S α,ε,θ,µ,E,S 0 0 15.9 4.8 47.4 0 12.5 12.5 0 0 0 0 8.7 8.7 39.1 73.9 0 2.2 8.7 10.9 0 0 0 12.5 8.6 0 0 0 0 0 18.8 31.4 31.8 0 0 0 3.1 68.8 54.3 54.5 13.3 26.3 10 12.5 81.3 88.6 95.5 86.7 76.3 90 87.5 0 0 4.5 0 0 0 0 0 0 0 0 0 0 0 31.3 2.9 0 0 0 0 0 12.5 2.9 4.5 0 2.6 0 0 0 0 0 0 0 0 0 6.3 0 0 0 0 0 0 0 0 0 0 0 0 0 *Of animal origin for human consumption. Discussion Until recently, only phenotypic studies had been used to evaluate virulence factors in C perfringens.4-6 Since 1980, studies about virulence-based taxonomy of C perfringens have not been reported; however, the prevalence of the enterotoxin gene was determined in isolates from human food or feces in association with episode of food poisoning,12 from swine with diarrhea,13 and from various healthy animals.14 Genomic organization of virulence genes has been described in some reference strains of C perfringens.15,16 Pathologic typing of C perfringens isolated from animals with various clostridial diseases could help researchers in associating virulence factors with diseases or in detecting the reservoirs of some pathologic types. In this study, the major lethal toxins, except α-toxin, were always found by use of the mouse lethality test when the corresponding genes were detected by DNA-DNA hybridization. The C perfringens α-toxin possesses both phospholipase C and sphingomyelinase activities and is an important virulence factor.7,17,18 The α-toxin gene is located on the chromosome near the putative origin of replication. 15 All the isolates we evaluated hybridized Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) with the α-toxin gene probe, but had different in vitro α-toxin activity, as revealed by the lethality test or observation of hemolysis. Variations in gene expression or in toxin activity could explain variations in the virulence of type-A isolates in human gas gangrene and in animals with enterocolitis. 19,20 The β-toxin gene was recently cloned and sequenced,21 but the gene location has not yet been determined. Only a few isolates from swine with necrotic enteritis hybridized with the β-toxin gene probe. Other isolates, classified primarily as toxin type C, from similar swine did not hybridize with this probe and were classified as toxin type A with the mouse lethality test. Type-C strains have been responsible for cases of necrotic enteritis in young animals in most species, especially in young swine.22 However, necrotic enteritis was rarely described in animals in our study, and most studied swine were not neonates. Isolates that hybridized with the ε-toxin gene probe were almost exclusively from small ruminants. Although pulpy kidney disease is characteristic of sheep and goats, type-D isolates in other species have been described, especially in cattle.23 In this study, the 1,443 bovine isolates did not hybridize with the ε-toxin probe. The ε-toxin gene is located on an extrachromosomal element,16,24 and thus, the factors determining the host preference are probably located on the same genetic support. Our study revealed that type-D C perfringens was isolated only from sheep, goats, or deer with enterotoxemia and from 1 fish, but never from cattle with enterotoxemia. Type-E isolates were not found in this study. The virulence of the 2 feline isolates that hybridized only with the ιa-toxin gene probe is unknown. The ιb-toxin is involved in binding and internalization of the toxin in the cell. Further study has revealed that these strains produce a nonlethal ADP-ribosyltransferase, and that lethality was not recovered by complementation with C perfringens type E or with C spiroforme ιb, fraction.8 Our study suggested these isolates are rare, although type-E isolates have been found in other parts of the world. 25,26 The number of enterotoxin-positive isolates was less than that found in another study.14 The rate of enterotoxinpositive isolates was 1% in our study vs 22% in that study for cattle, 13% vs 14% for horses, and 0% vs 10% for poultry. However, sample origin, isolation procedures, and geographic origin were different in these studies. Indeed, in the other study, the isolates were collected in Switzerland from healthy animals at slaughterhouses, and the isolates were cultured on tryptose-sulphite-dextrose agar. In this study, the human isolates were from fecal samples submitted to hospital laboratories, and were not associated with episodes of collective food poisoning. The human carriage in this population was important, approximately 22%. This high prevalence could have been associated with antibiotic use, 27 but specific data were not available. The 45 human food isolates did not hybridize with the enterotoxin probe, and thus none could have caused food poisoning. Another study of C perfringens isolated from fecal or food samples in food-borne disease outbreaks revealed that the prevalence of enterotoxigenic isolates was 66 and 57%, respectively.12 Thus, human foodstuffs seem to be rarely contaminated with enterotoxigenic C perfringens. In our study, only fish, with a prevalence of 44%, were especially hazardous. Human carriers of enterotoxigenic C perfringens must also be considered as potential sources of food poisoning. The enterotoxigenic C perfringens isolates can carry an enterotoxin structural gene on a chromosomal fragment or on an extrachromosomal element.16,28 Most of the food poisoning-associated isolates carried the gene on the chromosome. The genetic support for the enterotoxin gene has not been determined in isolates of various origins. Genetic location could be associated with variation in gene expression and, thus, in virulence. All isolates hybridized with the sialidase gene probe in our study. The sialidase gene was located on the chromosome in an instable region.15 The unique sialidase-negative C perfringens described is a laboratory strain, JIR323, which has probably lost the sialidase gene.2 The lack of sialidase-negative wild isolates, as with the αtoxin gene, suggested that this factor is important for resistance against natural selection, but was not absolutely necessary in vitro. This sialidase, unlike other neuraminidases, seems to be a cytoplasmic protein and, therefore, probably has a minor role in pathogenesis.29 The θ- and µ-toxin genes are located on the same chromosomal fragment, with substantial variations but without position changes.16 In one study,30 θ-toxin-negative isolates, without expression of this gene, were found. Two regulation systems have been described.31-33 In this study, the lack of 1 or both genes in some isolates was detected, but the prevalence varied among species of animals. This trend was especially obvious for the θ-toxin gene, with > 90% positive isolates in herbivore Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) species and only 62, 67, and 72% in fish, dogs, and human beings, respectively. Characterization of these isolates allowed us to identify 11 different pathologic types or gene combinations. With pathologic types corresponding to toxin types B and E and to the sialidase-negative toxin type-A strain,2 14 pathologic types were described with these 9 probes. The nonenterotoxigenic type-A isolates of this study could be grouped into 5 pathologic types. The most common type corresponded to the toxin type A1, described by Niilo. 4 Strains positive for α- and µ-toxins and positive or negative for θ-toxin were previously described.2 The other 2 pathologic types, negative for µ-toxin, had not been reported. The nonenterotoxigenic type-C and -D isolates, each with 1 unique pathologic type, corresponded to the toxin types C3 or C4 and D, respectively. 4 The enterotoxigenic isolates could be classified into 4 pathologic types. The first type, ε-toxin positive, was isolated in 3 independent fatal cases of enterotoxemia in sheep and in 1 in deer. Enterotoxin could be important in the pathogenesis of pulpy kidney disease by assisting ε-toxin transfer in the bloodstream, as has been shown for enterotoxigenic type-C isolates.8 The relative importance of this type of isolate in small ruminants has not been determined. Among the other 3 enterotoxin-positive pathologic types, 1 corresponded to the toxin type A2 and was the main enterotoxigenic type isolated in our study. 4 This pathologic type was isolated from samples of almost all origins, unlike the other 2 pathologic types, which were found in only 5 fish, 4 human beings, and 1 dog. Because the same pathologic types also were observed in most of the reference food poisoning isolates, 2 these isolates are probably the most hazardous. Although we associated the pathologic type with the origin of the isolates and with disease states, a causal relationship could not be determined, because bacterial counts or phenotypic results (toxins in intestinal content) were not available. Indeed, for most of the intestinal diseases caused by Clostridia, clinical signs are induced by the exotoxins produced in situ in high concentrations by a large number of organisms. For example, for C perfringens food poisoning in human beings, ingestion of many enterotoxigenic bacteria is responsible for massive intestinal sporulation, with subsequent enterotoxin production.4,34 For pulpy kidney disease in small ruminants, C perfringens type D must be present in high numbers for a period sufficient to produce enough εtoxin to cause the disease.35 In necrotic enteritis caused by C perfringens type C in young animals or human beings, bacterial numbers also must rapidly increase to enable the β-toxin to induce lesions.36 In that disease, nutritional factors were shown to be essential for disease expression.3,37 Other factors, especially stress, antibiotic treatment, or intestinal motility modification, could induce intestinal multiplication of C botulinum, spiroforme, difficile, or perfringens and toxin production.3,38-41 Detection of pathogenic Clostridia (eg, C botulinum, difficile, tetani, spiroforme, or various toxin types of C perfringens) in animals without signs of clostridial disease has been demonstrated.3,37,42-48 For all these reasons, the study of intestinal clostridial diseases by DNA-DNA colony hybridization must be interpreted with bacterial counts in mind. In addition, detection of toxigenic isolates by gene probes does not allow determination of toxin gene expression or gene product activity in the intestine. Thus, epidemiologic surveys of nonclinical bacterial carriers, regulation of toxin gene expression, and activity of expressed products are needed to determine, in each species and in each disease, the best diagnostic tests. The colony hybridization method allowed ready characterization of many isolates of known origin. Use of ELISA or easy in vitro neutralizing tests may more completely address the role of C perfringens isolates in disease. For ethical and practical reasons, we do not think that the mouse lethality test is a method for further use. As in other bacterial species (eg, Escherichia coli), distinguishing between saprophytic or pathogenic isolates is important. Moreover, such evaluation could help to assess the role of minor determinants in C per-fringensassociated diseases. Our approach may allow optimization of molecular diagnosis and prophylaxis in clostridial disease.10 ------------------------------------------a Isolates obtained from MR Popoff, Institut Pasteur de Paris, Paris, France. Anaerobic blood agar base with 10% cattle blood, Gibco-BRL, Paisley, Scotland. c Shahidi-Ferguson perfringens agar, Oxoid, Hampshire, United Kingdom. d Rosco disks, International Medical, Brussels, Belgium. e Clostridium welchii type A, C, D, E antitoxins, Wellcome Diagnostics, Dartford, England. f Whatman 541 filters, VWR Scientific, San Francisco, Calif. g Popoff MR, Institut Pasteur de Paris, Paris, France: Personal communication, 1992. b Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) References 1. Rood JI, Cole ST. Molecular genetics and pathogenesis of Clostridium perfringens. Microbiol Rev 1991;55:621-648. 2. Daube G, Simon P, Kaeckenbeeck A, et al. Molecular typing of Clostridium perfringens. Ann Med Vet 1994;138:183-191. 3. Daube G. Clostridium perfringens et pathologies digestives. Ann Med Vet 1992;136:5-30. 4. Niilo L. Clostridium perfringens in animal disease: a review of current knowledge. Can Vet J 1980;21:141148. 5. McDonel JL. Clostridium perfringens toxins (type A, B, C, D, E). Pharmacol Ther 1980;10:617-635. 6. McDonel JL. In: Dorner F, Drews J, eds. Toxins of Clostridium perfringens types A, B, C, D and E. Oxford, England: Pergamon Press, 1986;477-517. 7. Stevens DL, Mitten J, Henry C. Effects of α- and θ-toxins from Clostridium perfringens on human polymorphonuclear leukocytes. J Infect Dis 1987;156:324-333. 8. Skjelkvale R, Duncan CL. Enterotoxin formation by different toxigenic types of Clostridium perfringens. Infect Immun 1975;11: 563-575. 9. Sambrook J, Fritsch EF, Maniatis T. Purification of closed circular DNA by equilibrium centrifugation in CsCl ethidium bromide gradients. In: Molecular cloning: a laboratory manual. New York: CSH Laboratory, 1989. 10. Daube G, China B, Mainil J, et al. Typing of Clostridium perfringens by in vitro amplification of toxin genes. J Appl Bacterial 1994;77:650-655. 11. Sterne M, Batty I. Pathogenic Clostridia. London: Butter-worths, 1975;22-26. 12. Van Damme-Jongsten M, Rodhouse J, Notermans S, et al. Synthetic DNA probes for detection of enterotoxigenic Clostridium perfringens strains isolated from food borne disease outbreaks. J Clin Microbiol 1990;28:131-133. 13. Van Damme-Jongsten M, Haagsma J, Notermans S. Testing of Clostridium perfringens type A strains isolated from piglets with diarrhea for the presence of the enterotoxin gene. Vet Rec 1990;126: 191-192. 14. Tschirdewahn B, Notermans S, Untermann F, et al. The presence of enterotoxigenic Clostridium perfringens strains in faeces of various animals. Int f Food Microbiol 1991;14:175-178. 15. Canard B, Cole ST. Genome organization of the anaerobic pathogen Clostridium perfringens. Proc Natl Acad Sci U S A 1989;86: 6676-6680. 16. Canard B, Saint-Joanis B, Cole ST. Genomic diversity and organization of virulence genes in the pathogenic anaerobe Clostridium perfringens. Mol Microbiol 1992;6:1421-1429. 17. Saint-Joanis B, Gamier T, Cole ST. Gene cloning shows the alpha-toxin of Clostridium perfringens to contain both sphingomyelinase and lecithinase activities. Mol Gen Genet 1989;219:453-460. 18. Williamson ED, Titball RW. A genetically engineered vaccine against the alpha-toxin of Clostridium perfringens protects mice against experimental gas gangrene. Vaccine 1993;11:1253-1258. 19. Lyristis M, Bryant AE, Rood JI, et al. Identification and molecular analysis of a locus that regulates extracellular toxin production in Clostridium perfringens. Mol Microbiol 1994;12:761-777. 20. Rood JI, Lyristis M. Regulation of extracellular toxin production in Clostridium perfringens. Trends Microbiol 1995;3:192-196. Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) 21. Hunter SEC, Brown JE, Titball RW, et al. Molecular genetic analysis of beta-toxin of Clostridium perfringens reveals sequence homology with alpha-toxin, gamma toxin, and leukocidin of Staphylococcus aureus. Infect Immun 1993;61:3958-3965. 22. Barnes DM, Moon HW. Enterotoxemia in pigs due to Clostridium perfringens type C.J Am Vet Med Assoc 1964;144:1391-1394. 23. Griner LA, Aichelman WW, Brown GD. Clostridium perfringens type D (epsilon) enterotoxemia in Brown Swiss dairy calves. J Am Vet Med Assoc 1956;129:375-376. 24. Daube G, Simon P, Kaeckenbeeck A. IS1151, an IS-like element of Clostridium perfringens. Nucleic Acids Res 1993;21:352. 25. Niilo L, Avery RJ. Bovine enterotoxaemia: 1. Clostridium perfringens types isolated from animal sources in Alberta and Saskatchewan. Can Vet ] 1963;4:31-37. 26. Itodo AE, Adesiyun AA, Umoh JU, et al. Toxin-types of Clostridium perfringens isolated from sheep, cattle and paddock soils in Nigeria. Vet Microbiol 1986;12:93-96. 27. Samuel SC, Hancock P, Leigh DA. An investigation into Clostridium perfringens enterotoxin-associated diarrhoea. J Hosp Infect 1991;18:219-230. 28. Cornillot E, Saint-Joanis B, Cole ST, et al. Variable location of the enterotoxin gene (cpe) in Clostridium perfringens. Mol Microbiol 1995;15:639-647. 29. Roggentin P, Rothe B, Schauer R, et al. Conserved sequences in bacterial and viral sialidases. Glycoconjugate 1989;6:349-353. 30. Higashi Y, Chazono M, Amano T, et al. Complementation of theta-toxinogenicity between mutants of two groups of Clostridium perfringens. Biken J 1973;16:1-9. 31. Shimizu T, Okabe A, Hayashi H, et al. An upstream regulatory sequence stimulates expression of the perfringolysin O gene of Clostridium perfringens. Infect Immun 1991;59:137-142. 32. Shimizu T, Ba-Thein W, Hayashi H, et al. The virR gene, a member of a class of two-component response regulators, regulates the production of perfringolysin O, collagenase, and hemagglutinin in Clostridium perfringens. ] Bacteriol 1994;176:1616-1623. 33. Skjelkvale R, Duncan CL. Enterotoxin formation by different toxigenic types of Clostridium perfringens. Infect Immun 1975;11: 563-575. 34. Genigeorgis C. Public health importance of Clostridium perfringens. J Am Vet Med Assoc 1975;167:821827. 35. Bullen JJ, Scarisbrick R. Enterotoxaemia of sheep: experimental reproduction of the disease. J Pathol Bacteriol 1957;73:495-509. 36. Smith HW, Jones JET. Observations on the alimentary tract and its bacterial flora in healthy and diseased pigs. J Pathol Bacteriol 1963;86:387-412. 37. Niilo L. Enterotoxemic Clostridium perfringens. In: Gyles CL, Thoen CO, eds. Pathogenesis of bacterial infections in animals. Ames, Iowa: Iowa State University, 1986. 38. McFarland LV, Surawicz CM, Stamm WE. Risk factors for Clostridium difficile carriage and C. diffîcileassociated diarrhea in a cohort of hospitalized patients. J Infect Dis 1990;162:678-684. 39. Peeters JE, Geeroms R, Wilkins TD, et al. Significance of Clostridium spiroforme in the enteritis-complex of commercial rabbits. Vet Microbiol 1986;12:25-31. Published in: American Journal of Veterinary Research (1996), vol.57, pp. 496-501 Status: Postprint (Author’s version) 40. Borriello SP, Carman RJ. Association of iota-like toxin and Clostridium spiroforme with both spontaneous and antibiotic-associated diarrhea and colitis in rabbits. J Clin Microbiol 1983;17:414-418. 41. Thompson JA, Glasgow LA, Olson C, et al. Infant botulism: classical spectrum and epidemiology. Pediatrics 1980;66:936-942. 42. Holmes HT, Sonn RJ, Patton NM. Isolation of Clostridium spiroforme from rabbits. Lab Anim Sci 1988;38:167-168. 43. Wilkins CA, Richter MB, Carstens J, et al. Occurrence of Clostridium tetani in soil and horses. S Afr Med J 1988;73:718-720. 44. Borriello SP, Honour T, Barclay F, et al. Household pets as a potential reservoir for Clostridium difficile infection. J Clin Pathol 1983;36:84-87. 45. Hentges DJ, Marsh WW, Carter MK, et al. Influence of infant diets on the ecology of the intestinal tract of human flora-associated mice. J Pediatr Gastroenterol 1983;14:146-152. 46. Brant PC, Franti CE, Riemann HP, et al. Prevalence of Clostridium perfringens type A in two groups of food handlers in Bello Horizonte, Minas Gerais, Brazil. Rev Latinoam Microbiol 1978;20: 131-133. 47. Smith LDS. Inhibition of Clostridium botulinum by strains of Clostridium perfringens isolated from soil. Appl Microb 1975;30: 319-323. 48. Uemura T, Genigeorgis C, Franti CE, et al. Antibody against Clostridium perfringens type A enterotoxin in human sera. Infect Immun 1974;9:470-471. 49. Perelle S, Gilbert M, Popoff MR, et al. Characterization of Clostridium perfringens iota-toxin genes and expression in Escherichia coli. Infect Immun 1993;63:5147-5156. 50. Van Damme-Jongsten M, Wernars K, Notermans S. Cloning and sequencing of the Clostridium perfringens enterotoxin gene. An-tonie Van Leeuwenhoek 1989;56:181-190. 51. Tweten RK. Nucleotide sequence of the gene for perfrin-golysin O, theta-toxin, from Clostridium perfringens: significant homology with the genes for streptolysin O and pneumolysin. Infect Immun 1988;56:3235-3240. 52. Canard B, Garnier T, Cole ST, et al. Molecular genetic analysis of the nagH gene encoding the hyaluronidase (mu-toxin) of Clostridium perfringens. Mol Gen Genet 1994;243:215-224. 53. Roggentin P, Rothe B, Schauer R, et al. Cloning and sequencing of a Clostridium perfringens sialidase gene. FEBS Lett 1988; 238:31-34.