Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Molecular methods for the analysis of Clostridium perfringens relevant to food hygiene

Molecular methods for the analysis of Clostridium perfringens relevant to food hygiene Clostridium perfringens continues to be a common cause of food-borne disease [1,2]. It produces an enterotoxin (CPE) which is released upon lysis of the vegetative cell during sporulation in the intestinal tract. Catering premises with insufficient cooling and reheating devices often seem to be the cause of outbreaks of C. perfringens food poisoning. Typing of C. perfringens is of great importance for investigating sources of food poisoning cases and for studying the epidemiology of this microorganism. This report describes the examination of 155 C. perfringens isolates by molecular methods. Isolates were taken from 10 food poisoning outbreaks and cases (n = 34, food and fecal isolates) and from meat and fish pastes (n = 121). Isolates were characterized by plasmid profiling, ribotyping, and/or macrorestriction analysis by pulsed-field gel electrophoresis (PFGE). Results show that all three methods are suitable for classifying C. perfringens isolates below the species level. Ribopatterns and PFGE patterns can be interpreted more easily than plasmid profiling results and can be recommended for contamination studies and epidemiologic investigation of food poisonings associated with C. perfringens. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Clostridium perfringens ; Food poisoning ; Minced meat ; Plasmid pro¢ling ; Ribotyping ; Pulsed-¢eld gel electrophoresis 1. Introduction creasing extent in routine settings. For Clostridium perfringens, there are plasmid isolation [7], ribotyp- A variety of methods have been described for the ing of patients' and hospital environment isolates [8] strain di¡erentiation of bacteria, including biochem- and macrorestriction enzyme analysis by pulsed-¢eld ical characterization, bacteriocin typing, antibiotic gel electrophoresis (PFGE) [4,9]. Further DNA- resistance pro¢les, protein gel electrophoresis ¢nger- based methods are polymerase chain reaction printing, SDS-PAGE, or HPLC pro¢ling of bacterial (PCR)-related methods such as random ampli¢ed fatty acids [3^6]. Furthermore, several DNA-based polymorphic DNA (RAPD). Bacterial strain typing ¢ngerprinting methods are being applied to an in- methods are mainly used for epidemiological inves- tigations of the sources and spread of the causative agent in food-borne disease outbreaks but also sup- ply valuable information for contamination studies * Corresponding author ; E-mail : [email protected] in the food industry [10^12]. 0928-8244 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 928 -82 44( 99) 000 43- 7 Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 282 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 This article describes and evaluates the use of plas- [12]. Ensuing procedures were carried out as de- mid pro¢ling, ribotyping and PFGE for strain di¡er- scribed previously [23]. The DNA probe was pre- entiation of C. perfringens from food poisonings and pared from Escherichia coli 16S and 23S rRNA meat. Strains tested include patient stool and food (Boehringer, Mannheim, Germany) according to isolates from 10 food-borne outbreaks and cases. the manufacturer's instructions. Patterns were com- Furthermore, as meat is described to be the main pared visually. All 34 C. perfringens isolates from vehicle of infection in C. perfringens-associated food poisonings were analyzed three times. The 111 food poisoning, 121 meat and food isolates were minced meat and 10 food isolates were ribotyped analyzed. once, some isolates twice. 2.4. PFGE 2. Materials and methods The 34 isolates from food poisonings and the 10 2.1. Strains food isolates were analyzed once as described [24^ 26]. A total of 34 C. perfringens isolates associated with 10 food-borne outbreaks and cases (Table 1, designated 1^10) were analyzed by plasmid isolation, 3. Results ribotyping and PFGE. Isolates from food and stool were available. The strains were identi¢ed with the 3.1. Food poisoning isolates RAPID ID 32A identi¢cation kit (BioMerieux, Nu « rtingen, Germany) and additional con¢rmation A summary of the results for plasmid typing, ri- procedures as described previously [13,14]. botyping and PFGE of all 34 C. perfringens isolates As C. perfringens was found in 20% of minced is shown in Table 1. meat samples examined microbiologically according to the EU Directive 88/657/ECC [15], 111 C. perfrin- 3.2. Plasmid pro¢ling gens isolates were analyzed by ribotyping in order to collect basic information for a contamination study. For outbreaks 1, 2, and 5 (Table 1) all C. perfrin- Isolates were obtained by the pour-plate technique gens isolates, from foods as well as from fecal sam- with sul¢te-cycloserine-azide medium [16] and con- ples, showed identical plasmid pro¢les. One out of ¢rmed by acid phosphatase and reverse-CAMP test- two available fecal isolates as well as the food iso- ing as described [17]. Ten C. perfringens food isolates lates from outbreak 4 showed identical plasmid pro- from doner kebab, shrimp paste, crab paste, and ¢sh ¢les. For outbreak 3, one out of three fecal isolates sauce `tai pla' were analyzed by ribotyping and showed a plasmid pro¢le identical to the isolate from PFGE (Table 2). one (heart goulash) of the three epidemiologically implicated foods. Considering that strains di¡ering 2.2. Plasmid analysis only by the presence/absence of a single plasmid might be clonally related, outbreaks 6 and 8 are Plasmid analysis was performed according to also characterized by the isolation of clonally related Mahony et al. [18,19] and Eisgruber et al. [7,20] strains from foods and fecal samples. and was carried out twice per isolate. For three outbreaks, the plasmid pro¢ling results do not allow a clear interpretation. All of the ¢ve 2.3. Ribotyping isolates from outbreak 7 lacked any detectable plas- mids. For outbreak 9, one of the four fecal isolates Ribotyping was performed according to the meth- appears to be clonally related to the food isolate as it od described by Grimont and Grimont [21]. DNA di¡ers only by the presence/absence of a single plas- was isolated using the guanidinium thiocyanate mid. The three other fecal isolates carried no detect- method of Pitcher et al. [22] with modi¢cations able plasmids. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 283 Fig. 1. Ribotype patterns of 15 C. perfringens isolates of minced meat. 3.3. Ribotyping 3.5. Meat and food isolates Altogether, 12 distinct ribotype patterns were Results of ribotyping and PFGE of 10 C. perfrin- found among the 34 food poisoning-associated C. gens isolates from doner kebab, shrimp paste, crab perfringens isolates. Those 12 patterns were clearly paste, and ¢sh sauce `tai pla' are shown in Table 2. reproducible in three di¡erent runs. Patterns di¡er- As far as the isolates were typeable by PFGE, cor- ing by one or more bands were considered di¡erent. responding results were obtained by both methods. In eight of 10 food poisonings (Table 1 : 1, 2, 4^8, Three isolates from doner kebab and crab paste 10) identical ribotype patterns for all isolates were could not be typed by PFGE. detected. Two unique ribotypes were found in out- Within 111 C. perfringens isolates from minced breaks 3 (954/85, fecal isolate) and 9 (174/90, poultry meat, 107 distinctly di¡erent ribopatterns were de- fricassee isolate). Outbreak 3 showed ¢ve identical tected. Fig. 1 shows the ribopatterns of 15 isolates. ribotypes among six isolates derived from foods Considering only the pattern of fragments smaller and feces. In outbreak 9 all four fecal isolates than 2.3 kb all 111 C. perfringens isolates could be showed identical ribotypes. However, the C. perfrin- divided into ¢ve groups. The pro¢les (A^E) are pre- gens strain isolated from the presumptive outbreak- sented in Fig. 2. 70.3% of all C. perfringens isolated related chicken fricassee (174/90) had a di¡erent and from minced meat belonged to pro¢le type A (see unique pattern. also Fig. 1, lanes 2, 4, 10), 12.6% to pro¢le B and 3.4. PFGE Our own examinations and those of other authors showed that PFGE could distinguish C. perfringens isolates with identical ribopatterns. But PFGE pat- terns interpreted according to the criteria of Tenover et al. [27], i.e. patterns within an outbreak that di¡er by one to three fragments only are considered to be subtypes, categorized the isolates as closely Fig. 2. Typical fragment patterns under 2322 bp of C. perfringens related. isolates from minced meat. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 284 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 Table 1 Plasmid, ribotype and PFGE patterns for 34 C. perfringens isolates from food poisoning cases Outbreak/case Isolate Plasmids (molecular weight in MDa) Ribotype [23] PFGE [4,9] 1 rabbit meat 721/84 7.1 1 1 feces 731/84 7.1 1 1 feces 732/84 7.1 1 1 2 rabbit meat 310/85 28.7 2 2 feces 313/85 28.7 2 2 feces 314/85 28.7 2 2 3 heart goulash 945/85 25.5, 22.5 3 3 cauli£ower salad 948/85 27.1 3 3 cauli£ower salad 949/85 27.1 3 3 feces 953/85 25.5, 22.5 3 3 feces 954/85 21.1 11 11 feces 955/85 22.5 3 3 4 pork 10/86 28.7, 8.4, 6.6 4 4 feces 18/86 s 33.0, 8.4, 6.6 4 4 feces 26/86 28.7, 8.4, 6.6 4 4 5 beef 349/86 30.7 5 5 feces 344/86 30.7 5 5 feces 345/86 30.7 5 5 6 pea mash 834/87 5.6 6 6 feces 835/87 29.3, 5.6 6 6 7 poultry meat 216/88 0 7 7 feces 227/88 0 7 7 feces 231/88 0 7 7 feces 234/88 0 7 7 8 chicken fricassee 1291/88 29.3, 8.7, 2.4 8 8 feces 1295/88 29.3, 8.7, 2.4, 2.3 8 8 9 poultry fricassee 174/90 27.9, 3.0 12 12 feces 175/90 27.9 9 9 feces 176/90 0 9 9 feces 192/90 0 9 9 feces 195/90 0 9 9 10 beef 344/91 0 10 10 feces 346/91 8.8 10 10 feces 347/91 0 10 10 9.9% to pro¢le C. Four C. perfringens strains showed 4.1. Plasmid pro¢ling di¡erent pro¢le types (D, E). The discriminatory power of ribotyping for these Plasmid pro¢ling has been used for strain di¡er- minced meat isolates was above 0.99, estimated ac- entiation in a variety of Gram-positive and Gram- cording to the recommendations of Hunter and Gas- negative bacteria [29]. The utility of plasmid pro¢ling ton [28]. for the di¡erentiation of isolates within a given spe- cies depends on multiple factors including (i) varia- bility of plasmid patterns within a species, (ii) fre- 4. Discussion quency of isolates without plasmids, (iii) stability of plasmids, and (iv) reproducibility of plasmid pat- This study compared the utility of plasmid pro¢l- terns. Our own investigations have shown that about ing, ribotyping and PFGE for the di¡erentiation of 30% of all C. perfringens isolates carry no detectable C. perfringens isolates from food poisonings, minced plasmids. Among plasmid-carrying isolates a signi¢- meat and food. cant diversity of plasmid patterns has been found. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 285 Table 2 the six isolates of outbreak 3 (Table 1) one unique Ribotype and PFGE patterns for 10 C. perfringens food isolates pattern was detected (954/85, feces isolate) by ribo- Origin Isolate Ribotype PFGE typing and PFGE. The other ¢ve isolates from foods Doner kebab 24019-1 A NT and stool were identical. In outbreak 9 the ribopat- 24019-2 B B tern and PFGE pattern of the food isolate (174/90) 24019-3 A NT di¡ered from those of the four feces isolates and Crab paste 22431-1 C C therefore can be judged to be genetically not related 22431-3 D NT to any of the patients' stool isolates. This conclusion Shrimp paste 42-1 E E 1 42-2 E E 1 could not be drawn from the plasmid pro¢ling re- 42-3 E E 2 sults. Fish sauce 264-1b F F Within the 10 C. perfringens isolates from doner 264-2 F F kebab, shrimp paste, crab paste, and ¢sh sauce cor- NT, not typeable. responding results were obtained by ribotyping and PFGE, as far as isolates were PFGE-typeable. The ribotyping results of 111 C. perfringens iso- This is consistent with results reported [19,30]. Mah- lates from minced meat underline the high discrim- ony et al. [19] reported that 92% of the clinical C. inatory power of the method. Furthermore, these perfringens isolates tested carried plasmids while only results clearly show that meat is contaminated by 77% of the isolates from other sources had detectable C. perfringens isolates with a great genetic variety. plasmids. Phillips Jones et al. [30] found detectable The use of multiple strain typing methods is nec- plasmids for 71% of their C. perfringens isolates from essary to achieve reliable information regarding the food poisonings. A lower frequency of plasmid car- connection between presumptive epidemiologically riage was found among isolates from other sources related isolates. Plasmid typing as outlined in this including pork, lamb and ground beef. Even the ab- communication is a relatively simple typing method sence of plasmids can give a hint on epidemiological with low initial setup costs and represents a useful relationship, e.g. within outbreak 7 (Table 1) no method for preliminary strain typing of C. perfrin- plasmids were isolated from all three clinical isolates gens isolates. Manual ribotyping and PFGE are and the associated food isolate. The probability of more expensive and laborious, but results can be such a ¢nding for four entirely unrelated isolates can interpreted easily. Both methods seem to be very be considered very small. Similarly, Mahony et al. promising for the epidemiological investigation of [31] reported two outbreaks for which C. perfringens food-borne diseases caused by C. perfringens and isolates without plasmids were obtained. In general, for contamination studies in the ¢eld of food hygiene the high diversity of plasmid patterns appears to and industry. make plasmid pro¢ling a sensitive method for the discrimination of C. perfringens isolates. Acknowledgements 4.2. Ribotyping and PFGE We thank Dr. habil. H.-P. Schau, National Clos- Ribotyping and PFGE both gave corresponding tridia Reference Center, Erfurt, Germany, for pro- identical results for eight cases of food poisoning viding the food poisoning strains. (Table 1). This indicates that the isolates within each of these food poisonings are genetically closely related. Since the food isolates tested were epidemio- References logically implicated as the source of the outbreak or case, it can be assumed that the C. perfringens strain [1] Brett, M.M. (1998) 1566 outbreaks of Clostridium perfringens found in food and stool caused the disease. food poisoning, 1970^1996. 4th World Congress on Food- In two cases di¡erent patterns were detectable borne Infections and Intoxications, 7^12 June, Berlin, p. 60. [2] FDA (1998) Bad Bug Book. Foodborne Pathogenic Micro- within the isolates of one food poisoning. Among Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 286 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 clostridia in food monitoring programs. Food Res. Int. 28, organisms and Natural Toxins Handbook. Clostridium per- 219^226. fringens. U.S. Food and Drug Administration, Center for [17] Schalch, B., Eisgruber, H., Geppert, P. and Stolle, A. (1996) Food Safety and Applied Nutrition. htttp ://www.cfsan.fda. Comparison of four routine procedures for the con¢rmation gov. of Clostridium perfringens from food. Arch. Lebensmittelhyg. [3] Mulligan, M.E., Halebian, S., Kwok, R.Y.Y., Cheng, W.C., 47, 27^30. Finegold, S.M., Anselmo, C.R., Gerding, D.N. and Peterson, [18] Mahony, D.E., Clark, G.A., Stringer, M.F., MacDonald, L.R. (1986) Bacterial agglutination and polyacrylamide gel M.C., Duchesne, D.R. and Marder, J.A. (1986) Rapid extrac- electrophoresis for typing Clostridium di¤cile. J. Infect. Dis. tion of plasmids from Clostridium perfringens. Appl. Environ. 153, 267^271. [4] Klein, G., Pack, A., Eisgruber, H. and Stolle, A. (1996) Mog- Microbiol. 51, 521^523. [19] Mahony, D.E., Stringer, M.F., Borriello, S.P. and Mader, liche Zusammenha « nge zwischen Clostridium perfringens-Iso- J.A. (1987) Plasmid analysis as a means of strain di¡erenta- laten aus Lebensmitteln und Clostridium perfringens-Erkran- tion in Clostridium perfringens. J. Clin. Microbiol. 25, 1333^ kungen. 37. Arbeitstag. Arbeitsg. Lebensmittelhyg., DVG, Garmisch-Partenkirchen, pp. 248^255. [5] Harpold, D.J., Wasilauskas, B.L. and O'Connor, M.L. (1985) [20] Eisgruber, H., Wiedmann, M. and Stolle, A. (1996) Plasmid pro¢ling for strain di¡erentiation and characterization of Rapid identi¢cation of Clostridium species by high pressure Clostridium perfringens isolates. J. Vet. Med. B 43, 137^146. liquid chromatography. J. Clin. Microbiol. 22, 962^967. [21] Grimont, F. and Grimont, P.A.D. (1986) Ribosomal ribonu- [6] Krausse, R. and Ullmann, U. (1991) A modi¢ed procedure for cleic acid gene restriction patterns as potential taxonomic the identi¢cation of anaerobic bacteria by high performance liquid chromatography ^ quantitative analysis of short-chain tools. Ann. Inst. Pasteur/Microbiol. (Paris) 137, 165^175. [22] Pitcher, D.G., Saunders, N.A. and Owen, R.J. (1989) Rapid fatty acids. Zbl. Bakteriol. 276, 1^8. [7] Eisgruber, H., Wiedmann, M. and Stolle, A. (1995) Use of extraction of bacterial genomic DNA with guanidium thiocya- nate. Lett. Appl. Microbiol. 8, 151^156. plasmid pro¢ling as a typing method for epidemiologically [23] Schalch, B., Bjo « rkroth, J., Eisgruber, H., Korkeala, H. and related Clostridium perfringens isolates from food poisoning cases and outbreaks. Lett. Appl. Microbiol. 20, 290^294. Stolle, A. (1997) Ribotyping for strain characterization of Clostridium perfringens isolates from food poisoning cases [8] Forsblom, B., Palmu, A., Hirvonen, P. and Jousimiessomer, H. (1995) Ribotyping of Clostridum perfringens. Clin. Infect. and outbreaks. Appl. Environ. Microbiol. 63, 3992^3994. Dis. 20, (Suppl. 2) S323^S324. [24] Matushek, M.G., Bonten, M.J.M. and Hayden, M.K. (1996) Rapid preparation of bacerial DNA for pulsed-¢eld gel elec- [9] Ridell, J., Bjo « rkroth, J., Eisgruber, H., Schalch, B., Stolle, A. and Korkeala, H. (1998) Prevalence of the enterotoxin gene trophoresis. J. Clin. Microbiol. 34, 2598^2600. [25] Hielm, S., Bjo « rkroth, J., Hyytia « , E. and Korkeala, H. (1998) and clonality of Clostridium perfringens strains associated with food poisoning outbreaks. J. Food Prot. 61, 240^243. Genomic analysis of Clostridium botulinum group II by [10] Labbe, R.G. (1989) Clostridium perfringens. In : Foodborne pulsed-¢eld gel electrophoresis. App. Environ. Microbiol. 64, Bacterial Pathogens (Doyle, M.P., Ed.), pp. 191^234. Marcel 703^708. Dekker, New York. [26] Sperner, B., Schalch, B., Eisgruber, H. and Stolle, A. (1998) Short protocol for pulsed-¢eld gel electophoresis of a variety [11] Rood, J.I. and Cole, S.T. (1991) Molecular genetics and pathogenesis of Clostridium perfringens. Microbiol. Rev. 55, of Clostridia species. European Clostridia Conference, 4^7 621^648. October, Teistungen. [12] Bjorkroth, J. and Korkeala, H. (1997) Use of rRNA gene [27] Tenover, F.C., Arbeit, R.D., Goering, R.V., Mickelsen, P.A., restriction patterns to evaluate lactic acid bacterium contam- Murray, B.E., Persing, D.H. and Swaminathan, B. (1995) In- terpreting chromosomal DNA restriction patterns produced ination of vacuum-packaged sliced cooked whole-meat prod- uct in a meat processing plant. Appl. Environ. Microbiol. 63, by pulsed-¢eld gel electrophoresis : criteria for bacterial strain 448^453. typing. J. Clin. Microbiol. 33, 2233^2239. [13] Eisgruber, H. and Reuter, G. (1987) Einsatzmoglichkeiten ein- [28] Hunter, P.R. and Gaston, M.A. (1988) Numerical index of the facher und zeitsparender Verfahren zur orientierenden Identi- discriminatory ability of typing systems : an application of ¢zierung wichtiger Clostridien-Species. Arch. Lebensmittel- Simpson's Index of Diversity. J. Clin. Microbiol. 26, 2465^ hyg. 38, 141^146. 2466. [14] Eisgruber, H. (1992) Eignung des RAPID ID 32A0-Testsys- [29] Mayer, L.W. (1988) Use of plasmid pro¢les in epidemiologic tems zur schnellen Identi¢zierung lebensmittelhygienisch wich- surveillance of disease outbreaks and in tracing the transmis- tiger Clostridien-Spezies. Arch. Lebensmittelhyg. 43, 126^ sion of antibiotic resistance. Clin. Microbiol. Rev. 1, 228^243. 130. [30] Phillips Jones, M.K., Iwanejko, A. and Longden, M.S. (1989) [15] Schalch, B. and Eisgruber, H. (1995) Praktische Erfahrungen Analysis of plasmid pro¢ling as a method for rapid di¡eren- mit den mikrobiologischen Anforderungen der EG-Hack- tiation of food-associated Clostridium perfringens strains. £eischrichtlinie. 36. Arbeitstag. Arbeitsg. Lebensmittelhyg., J. Appl. Bacteriol. 67, 243^254. DVG, Garmisch-Partenkirchen, pp. 228^233 [31] Mahony, D.E., Ahmed, R. and Jackson, S.G. (1992) Multiple [16] Eisgruber, H. and Reuter, H. (1995) A selective medium for typing techniques applied to a Clostridium perfringens food the detection and enumeration of mesophilic sulphite-reducing poisoning outbreak. J. Appl. Bacteriol. 72, 309^314. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of the Endocrine Society Oxford University Press

Molecular methods for the analysis of Clostridium perfringens relevant to food hygiene

Download PDF
6 pages

Loading next page...
 
/lp/oxford-university-press/molecular-methods-for-the-analysis-of-clostridium-perfringens-relevant-JITHxGbs4T

References (31)

Publisher
Oxford University Press
Copyright
© 1999 Federation of European Microbiological Societies.
eISSN
2472-1972
DOI
10.1111/j.1574-695X.1999.tb01295.x
pmid
10397312
Publisher site
See Article on Publisher Site

Abstract

Clostridium perfringens continues to be a common cause of food-borne disease [1,2]. It produces an enterotoxin (CPE) which is released upon lysis of the vegetative cell during sporulation in the intestinal tract. Catering premises with insufficient cooling and reheating devices often seem to be the cause of outbreaks of C. perfringens food poisoning. Typing of C. perfringens is of great importance for investigating sources of food poisoning cases and for studying the epidemiology of this microorganism. This report describes the examination of 155 C. perfringens isolates by molecular methods. Isolates were taken from 10 food poisoning outbreaks and cases (n = 34, food and fecal isolates) and from meat and fish pastes (n = 121). Isolates were characterized by plasmid profiling, ribotyping, and/or macrorestriction analysis by pulsed-field gel electrophoresis (PFGE). Results show that all three methods are suitable for classifying C. perfringens isolates below the species level. Ribopatterns and PFGE patterns can be interpreted more easily than plasmid profiling results and can be recommended for contamination studies and epidemiologic investigation of food poisonings associated with C. perfringens. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Clostridium perfringens ; Food poisoning ; Minced meat ; Plasmid pro¢ling ; Ribotyping ; Pulsed-¢eld gel electrophoresis 1. Introduction creasing extent in routine settings. For Clostridium perfringens, there are plasmid isolation [7], ribotyp- A variety of methods have been described for the ing of patients' and hospital environment isolates [8] strain di¡erentiation of bacteria, including biochem- and macrorestriction enzyme analysis by pulsed-¢eld ical characterization, bacteriocin typing, antibiotic gel electrophoresis (PFGE) [4,9]. Further DNA- resistance pro¢les, protein gel electrophoresis ¢nger- based methods are polymerase chain reaction printing, SDS-PAGE, or HPLC pro¢ling of bacterial (PCR)-related methods such as random ampli¢ed fatty acids [3^6]. Furthermore, several DNA-based polymorphic DNA (RAPD). Bacterial strain typing ¢ngerprinting methods are being applied to an in- methods are mainly used for epidemiological inves- tigations of the sources and spread of the causative agent in food-borne disease outbreaks but also sup- ply valuable information for contamination studies * Corresponding author ; E-mail : [email protected] in the food industry [10^12]. 0928-8244 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 928 -82 44( 99) 000 43- 7 Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 282 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 This article describes and evaluates the use of plas- [12]. Ensuing procedures were carried out as de- mid pro¢ling, ribotyping and PFGE for strain di¡er- scribed previously [23]. The DNA probe was pre- entiation of C. perfringens from food poisonings and pared from Escherichia coli 16S and 23S rRNA meat. Strains tested include patient stool and food (Boehringer, Mannheim, Germany) according to isolates from 10 food-borne outbreaks and cases. the manufacturer's instructions. Patterns were com- Furthermore, as meat is described to be the main pared visually. All 34 C. perfringens isolates from vehicle of infection in C. perfringens-associated food poisonings were analyzed three times. The 111 food poisoning, 121 meat and food isolates were minced meat and 10 food isolates were ribotyped analyzed. once, some isolates twice. 2.4. PFGE 2. Materials and methods The 34 isolates from food poisonings and the 10 2.1. Strains food isolates were analyzed once as described [24^ 26]. A total of 34 C. perfringens isolates associated with 10 food-borne outbreaks and cases (Table 1, designated 1^10) were analyzed by plasmid isolation, 3. Results ribotyping and PFGE. Isolates from food and stool were available. The strains were identi¢ed with the 3.1. Food poisoning isolates RAPID ID 32A identi¢cation kit (BioMerieux, Nu « rtingen, Germany) and additional con¢rmation A summary of the results for plasmid typing, ri- procedures as described previously [13,14]. botyping and PFGE of all 34 C. perfringens isolates As C. perfringens was found in 20% of minced is shown in Table 1. meat samples examined microbiologically according to the EU Directive 88/657/ECC [15], 111 C. perfrin- 3.2. Plasmid pro¢ling gens isolates were analyzed by ribotyping in order to collect basic information for a contamination study. For outbreaks 1, 2, and 5 (Table 1) all C. perfrin- Isolates were obtained by the pour-plate technique gens isolates, from foods as well as from fecal sam- with sul¢te-cycloserine-azide medium [16] and con- ples, showed identical plasmid pro¢les. One out of ¢rmed by acid phosphatase and reverse-CAMP test- two available fecal isolates as well as the food iso- ing as described [17]. Ten C. perfringens food isolates lates from outbreak 4 showed identical plasmid pro- from doner kebab, shrimp paste, crab paste, and ¢sh ¢les. For outbreak 3, one out of three fecal isolates sauce `tai pla' were analyzed by ribotyping and showed a plasmid pro¢le identical to the isolate from PFGE (Table 2). one (heart goulash) of the three epidemiologically implicated foods. Considering that strains di¡ering 2.2. Plasmid analysis only by the presence/absence of a single plasmid might be clonally related, outbreaks 6 and 8 are Plasmid analysis was performed according to also characterized by the isolation of clonally related Mahony et al. [18,19] and Eisgruber et al. [7,20] strains from foods and fecal samples. and was carried out twice per isolate. For three outbreaks, the plasmid pro¢ling results do not allow a clear interpretation. All of the ¢ve 2.3. Ribotyping isolates from outbreak 7 lacked any detectable plas- mids. For outbreak 9, one of the four fecal isolates Ribotyping was performed according to the meth- appears to be clonally related to the food isolate as it od described by Grimont and Grimont [21]. DNA di¡ers only by the presence/absence of a single plas- was isolated using the guanidinium thiocyanate mid. The three other fecal isolates carried no detect- method of Pitcher et al. [22] with modi¢cations able plasmids. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 283 Fig. 1. Ribotype patterns of 15 C. perfringens isolates of minced meat. 3.3. Ribotyping 3.5. Meat and food isolates Altogether, 12 distinct ribotype patterns were Results of ribotyping and PFGE of 10 C. perfrin- found among the 34 food poisoning-associated C. gens isolates from doner kebab, shrimp paste, crab perfringens isolates. Those 12 patterns were clearly paste, and ¢sh sauce `tai pla' are shown in Table 2. reproducible in three di¡erent runs. Patterns di¡er- As far as the isolates were typeable by PFGE, cor- ing by one or more bands were considered di¡erent. responding results were obtained by both methods. In eight of 10 food poisonings (Table 1 : 1, 2, 4^8, Three isolates from doner kebab and crab paste 10) identical ribotype patterns for all isolates were could not be typed by PFGE. detected. Two unique ribotypes were found in out- Within 111 C. perfringens isolates from minced breaks 3 (954/85, fecal isolate) and 9 (174/90, poultry meat, 107 distinctly di¡erent ribopatterns were de- fricassee isolate). Outbreak 3 showed ¢ve identical tected. Fig. 1 shows the ribopatterns of 15 isolates. ribotypes among six isolates derived from foods Considering only the pattern of fragments smaller and feces. In outbreak 9 all four fecal isolates than 2.3 kb all 111 C. perfringens isolates could be showed identical ribotypes. However, the C. perfrin- divided into ¢ve groups. The pro¢les (A^E) are pre- gens strain isolated from the presumptive outbreak- sented in Fig. 2. 70.3% of all C. perfringens isolated related chicken fricassee (174/90) had a di¡erent and from minced meat belonged to pro¢le type A (see unique pattern. also Fig. 1, lanes 2, 4, 10), 12.6% to pro¢le B and 3.4. PFGE Our own examinations and those of other authors showed that PFGE could distinguish C. perfringens isolates with identical ribopatterns. But PFGE pat- terns interpreted according to the criteria of Tenover et al. [27], i.e. patterns within an outbreak that di¡er by one to three fragments only are considered to be subtypes, categorized the isolates as closely Fig. 2. Typical fragment patterns under 2322 bp of C. perfringens related. isolates from minced meat. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 284 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 Table 1 Plasmid, ribotype and PFGE patterns for 34 C. perfringens isolates from food poisoning cases Outbreak/case Isolate Plasmids (molecular weight in MDa) Ribotype [23] PFGE [4,9] 1 rabbit meat 721/84 7.1 1 1 feces 731/84 7.1 1 1 feces 732/84 7.1 1 1 2 rabbit meat 310/85 28.7 2 2 feces 313/85 28.7 2 2 feces 314/85 28.7 2 2 3 heart goulash 945/85 25.5, 22.5 3 3 cauli£ower salad 948/85 27.1 3 3 cauli£ower salad 949/85 27.1 3 3 feces 953/85 25.5, 22.5 3 3 feces 954/85 21.1 11 11 feces 955/85 22.5 3 3 4 pork 10/86 28.7, 8.4, 6.6 4 4 feces 18/86 s 33.0, 8.4, 6.6 4 4 feces 26/86 28.7, 8.4, 6.6 4 4 5 beef 349/86 30.7 5 5 feces 344/86 30.7 5 5 feces 345/86 30.7 5 5 6 pea mash 834/87 5.6 6 6 feces 835/87 29.3, 5.6 6 6 7 poultry meat 216/88 0 7 7 feces 227/88 0 7 7 feces 231/88 0 7 7 feces 234/88 0 7 7 8 chicken fricassee 1291/88 29.3, 8.7, 2.4 8 8 feces 1295/88 29.3, 8.7, 2.4, 2.3 8 8 9 poultry fricassee 174/90 27.9, 3.0 12 12 feces 175/90 27.9 9 9 feces 176/90 0 9 9 feces 192/90 0 9 9 feces 195/90 0 9 9 10 beef 344/91 0 10 10 feces 346/91 8.8 10 10 feces 347/91 0 10 10 9.9% to pro¢le C. Four C. perfringens strains showed 4.1. Plasmid pro¢ling di¡erent pro¢le types (D, E). The discriminatory power of ribotyping for these Plasmid pro¢ling has been used for strain di¡er- minced meat isolates was above 0.99, estimated ac- entiation in a variety of Gram-positive and Gram- cording to the recommendations of Hunter and Gas- negative bacteria [29]. The utility of plasmid pro¢ling ton [28]. for the di¡erentiation of isolates within a given spe- cies depends on multiple factors including (i) varia- bility of plasmid patterns within a species, (ii) fre- 4. Discussion quency of isolates without plasmids, (iii) stability of plasmids, and (iv) reproducibility of plasmid pat- This study compared the utility of plasmid pro¢l- terns. Our own investigations have shown that about ing, ribotyping and PFGE for the di¡erentiation of 30% of all C. perfringens isolates carry no detectable C. perfringens isolates from food poisonings, minced plasmids. Among plasmid-carrying isolates a signi¢- meat and food. cant diversity of plasmid patterns has been found. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 285 Table 2 the six isolates of outbreak 3 (Table 1) one unique Ribotype and PFGE patterns for 10 C. perfringens food isolates pattern was detected (954/85, feces isolate) by ribo- Origin Isolate Ribotype PFGE typing and PFGE. The other ¢ve isolates from foods Doner kebab 24019-1 A NT and stool were identical. In outbreak 9 the ribopat- 24019-2 B B tern and PFGE pattern of the food isolate (174/90) 24019-3 A NT di¡ered from those of the four feces isolates and Crab paste 22431-1 C C therefore can be judged to be genetically not related 22431-3 D NT to any of the patients' stool isolates. This conclusion Shrimp paste 42-1 E E 1 42-2 E E 1 could not be drawn from the plasmid pro¢ling re- 42-3 E E 2 sults. Fish sauce 264-1b F F Within the 10 C. perfringens isolates from doner 264-2 F F kebab, shrimp paste, crab paste, and ¢sh sauce cor- NT, not typeable. responding results were obtained by ribotyping and PFGE, as far as isolates were PFGE-typeable. The ribotyping results of 111 C. perfringens iso- This is consistent with results reported [19,30]. Mah- lates from minced meat underline the high discrim- ony et al. [19] reported that 92% of the clinical C. inatory power of the method. Furthermore, these perfringens isolates tested carried plasmids while only results clearly show that meat is contaminated by 77% of the isolates from other sources had detectable C. perfringens isolates with a great genetic variety. plasmids. Phillips Jones et al. [30] found detectable The use of multiple strain typing methods is nec- plasmids for 71% of their C. perfringens isolates from essary to achieve reliable information regarding the food poisonings. A lower frequency of plasmid car- connection between presumptive epidemiologically riage was found among isolates from other sources related isolates. Plasmid typing as outlined in this including pork, lamb and ground beef. Even the ab- communication is a relatively simple typing method sence of plasmids can give a hint on epidemiological with low initial setup costs and represents a useful relationship, e.g. within outbreak 7 (Table 1) no method for preliminary strain typing of C. perfrin- plasmids were isolated from all three clinical isolates gens isolates. Manual ribotyping and PFGE are and the associated food isolate. The probability of more expensive and laborious, but results can be such a ¢nding for four entirely unrelated isolates can interpreted easily. Both methods seem to be very be considered very small. Similarly, Mahony et al. promising for the epidemiological investigation of [31] reported two outbreaks for which C. perfringens food-borne diseases caused by C. perfringens and isolates without plasmids were obtained. In general, for contamination studies in the ¢eld of food hygiene the high diversity of plasmid patterns appears to and industry. make plasmid pro¢ling a sensitive method for the discrimination of C. perfringens isolates. Acknowledgements 4.2. Ribotyping and PFGE We thank Dr. habil. H.-P. Schau, National Clos- Ribotyping and PFGE both gave corresponding tridia Reference Center, Erfurt, Germany, for pro- identical results for eight cases of food poisoning viding the food poisoning strains. (Table 1). This indicates that the isolates within each of these food poisonings are genetically closely related. Since the food isolates tested were epidemio- References logically implicated as the source of the outbreak or case, it can be assumed that the C. perfringens strain [1] Brett, M.M. (1998) 1566 outbreaks of Clostridium perfringens found in food and stool caused the disease. food poisoning, 1970^1996. 4th World Congress on Food- In two cases di¡erent patterns were detectable borne Infections and Intoxications, 7^12 June, Berlin, p. 60. [2] FDA (1998) Bad Bug Book. Foodborne Pathogenic Micro- within the isolates of one food poisoning. Among Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99 286 B. Schalch et al. / FEMS Immunology and Medical Microbiology 24 (1999) 281^286 clostridia in food monitoring programs. Food Res. Int. 28, organisms and Natural Toxins Handbook. Clostridium per- 219^226. fringens. U.S. Food and Drug Administration, Center for [17] Schalch, B., Eisgruber, H., Geppert, P. and Stolle, A. (1996) Food Safety and Applied Nutrition. htttp ://www.cfsan.fda. Comparison of four routine procedures for the con¢rmation gov. of Clostridium perfringens from food. Arch. Lebensmittelhyg. [3] Mulligan, M.E., Halebian, S., Kwok, R.Y.Y., Cheng, W.C., 47, 27^30. Finegold, S.M., Anselmo, C.R., Gerding, D.N. and Peterson, [18] Mahony, D.E., Clark, G.A., Stringer, M.F., MacDonald, L.R. (1986) Bacterial agglutination and polyacrylamide gel M.C., Duchesne, D.R. and Marder, J.A. (1986) Rapid extrac- electrophoresis for typing Clostridium di¤cile. J. Infect. Dis. tion of plasmids from Clostridium perfringens. Appl. Environ. 153, 267^271. [4] Klein, G., Pack, A., Eisgruber, H. and Stolle, A. (1996) Mog- Microbiol. 51, 521^523. [19] Mahony, D.E., Stringer, M.F., Borriello, S.P. and Mader, liche Zusammenha « nge zwischen Clostridium perfringens-Iso- J.A. (1987) Plasmid analysis as a means of strain di¡erenta- laten aus Lebensmitteln und Clostridium perfringens-Erkran- tion in Clostridium perfringens. J. Clin. Microbiol. 25, 1333^ kungen. 37. Arbeitstag. Arbeitsg. Lebensmittelhyg., DVG, Garmisch-Partenkirchen, pp. 248^255. [5] Harpold, D.J., Wasilauskas, B.L. and O'Connor, M.L. (1985) [20] Eisgruber, H., Wiedmann, M. and Stolle, A. (1996) Plasmid pro¢ling for strain di¡erentiation and characterization of Rapid identi¢cation of Clostridium species by high pressure Clostridium perfringens isolates. J. Vet. Med. B 43, 137^146. liquid chromatography. J. Clin. Microbiol. 22, 962^967. [21] Grimont, F. and Grimont, P.A.D. (1986) Ribosomal ribonu- [6] Krausse, R. and Ullmann, U. (1991) A modi¢ed procedure for cleic acid gene restriction patterns as potential taxonomic the identi¢cation of anaerobic bacteria by high performance liquid chromatography ^ quantitative analysis of short-chain tools. Ann. Inst. Pasteur/Microbiol. (Paris) 137, 165^175. [22] Pitcher, D.G., Saunders, N.A. and Owen, R.J. (1989) Rapid fatty acids. Zbl. Bakteriol. 276, 1^8. [7] Eisgruber, H., Wiedmann, M. and Stolle, A. (1995) Use of extraction of bacterial genomic DNA with guanidium thiocya- nate. Lett. Appl. Microbiol. 8, 151^156. plasmid pro¢ling as a typing method for epidemiologically [23] Schalch, B., Bjo « rkroth, J., Eisgruber, H., Korkeala, H. and related Clostridium perfringens isolates from food poisoning cases and outbreaks. Lett. Appl. Microbiol. 20, 290^294. Stolle, A. (1997) Ribotyping for strain characterization of Clostridium perfringens isolates from food poisoning cases [8] Forsblom, B., Palmu, A., Hirvonen, P. and Jousimiessomer, H. (1995) Ribotyping of Clostridum perfringens. Clin. Infect. and outbreaks. Appl. Environ. Microbiol. 63, 3992^3994. Dis. 20, (Suppl. 2) S323^S324. [24] Matushek, M.G., Bonten, M.J.M. and Hayden, M.K. (1996) Rapid preparation of bacerial DNA for pulsed-¢eld gel elec- [9] Ridell, J., Bjo « rkroth, J., Eisgruber, H., Schalch, B., Stolle, A. and Korkeala, H. (1998) Prevalence of the enterotoxin gene trophoresis. J. Clin. Microbiol. 34, 2598^2600. [25] Hielm, S., Bjo « rkroth, J., Hyytia « , E. and Korkeala, H. (1998) and clonality of Clostridium perfringens strains associated with food poisoning outbreaks. J. Food Prot. 61, 240^243. Genomic analysis of Clostridium botulinum group II by [10] Labbe, R.G. (1989) Clostridium perfringens. In : Foodborne pulsed-¢eld gel electrophoresis. App. Environ. Microbiol. 64, Bacterial Pathogens (Doyle, M.P., Ed.), pp. 191^234. Marcel 703^708. Dekker, New York. [26] Sperner, B., Schalch, B., Eisgruber, H. and Stolle, A. (1998) Short protocol for pulsed-¢eld gel electophoresis of a variety [11] Rood, J.I. and Cole, S.T. (1991) Molecular genetics and pathogenesis of Clostridium perfringens. Microbiol. Rev. 55, of Clostridia species. European Clostridia Conference, 4^7 621^648. October, Teistungen. [12] Bjorkroth, J. and Korkeala, H. (1997) Use of rRNA gene [27] Tenover, F.C., Arbeit, R.D., Goering, R.V., Mickelsen, P.A., restriction patterns to evaluate lactic acid bacterium contam- Murray, B.E., Persing, D.H. and Swaminathan, B. (1995) In- terpreting chromosomal DNA restriction patterns produced ination of vacuum-packaged sliced cooked whole-meat prod- uct in a meat processing plant. Appl. Environ. Microbiol. 63, by pulsed-¢eld gel electrophoresis : criteria for bacterial strain 448^453. typing. J. Clin. Microbiol. 33, 2233^2239. [13] Eisgruber, H. and Reuter, G. (1987) Einsatzmoglichkeiten ein- [28] Hunter, P.R. and Gaston, M.A. (1988) Numerical index of the facher und zeitsparender Verfahren zur orientierenden Identi- discriminatory ability of typing systems : an application of ¢zierung wichtiger Clostridien-Species. Arch. Lebensmittel- Simpson's Index of Diversity. J. Clin. Microbiol. 26, 2465^ hyg. 38, 141^146. 2466. [14] Eisgruber, H. (1992) Eignung des RAPID ID 32A0-Testsys- [29] Mayer, L.W. (1988) Use of plasmid pro¢les in epidemiologic tems zur schnellen Identi¢zierung lebensmittelhygienisch wich- surveillance of disease outbreaks and in tracing the transmis- tiger Clostridien-Spezies. Arch. Lebensmittelhyg. 43, 126^ sion of antibiotic resistance. Clin. Microbiol. Rev. 1, 228^243. 130. [30] Phillips Jones, M.K., Iwanejko, A. and Longden, M.S. (1989) [15] Schalch, B. and Eisgruber, H. (1995) Praktische Erfahrungen Analysis of plasmid pro¢ling as a method for rapid di¡eren- mit den mikrobiologischen Anforderungen der EG-Hack- tiation of food-associated Clostridium perfringens strains. £eischrichtlinie. 36. Arbeitstag. Arbeitsg. Lebensmittelhyg., J. Appl. Bacteriol. 67, 243^254. DVG, Garmisch-Partenkirchen, pp. 228^233 [31] Mahony, D.E., Ahmed, R. and Jackson, S.G. (1992) Multiple [16] Eisgruber, H. and Reuter, H. (1995) A selective medium for typing techniques applied to a Clostridium perfringens food the detection and enumeration of mesophilic sulphite-reducing poisoning outbreak. J. Appl. Bacteriol. 72, 309^314. Downloaded from https://academic.oup.com/femspd/article-abstract/24/3/281/626679 by Ed 'DeepDyve' Gillespie user on 28 April 2018 FEMSIM 1032 8-6-99

Journal

Journal of the Endocrine Society – Oxford University Press

Published: Jul 1, 1999

There are no references for this article.