FOOD ASSOCIATED LISTERIOSIS, YERSINIOSIS, AND BOTULISM; NEW DNA-BASED METHODS FOR STUDYING THE INFECTION ROUTES, CONTROLLING THE SPREAD AND ASSESSING THE RISKS IN FOOD PRODUCTION

Project leader: Hannu Korkeala, University of Helsinki, Department of Food and Environmental Hygiene, Faculty of Veterinary Medicine, FIN-00014 University of Helsinki, Finland, tel. +358-9-19149700, e-mail: Hannu.Korkeala@helsinki.fi
 
 

PUBLICATIONS

TIIVISTELMÄ SUOMEKSI

Reasearchers:
University of Helsinki:
Kristiina Asplund, tel. +358-9-19149701, e-mail: Kristiina.Asplund@helsinki.fi
Tiina Autio, tel. +358-9-19149766, e-mail: Tiautio@helsinki.fi
Johanna Björkroth, tel. +358-9-19149705, e-mail: Johanna.Bjorkroth@helsinki.fi
Maria Fredriksson-Ahomaa, tel.+358-9-19149701, e-mail: Fredrika@helsinki.fi
Sanna Hellström, tel. +358-9-19149701, e-mail: Sanna.Hellström @helsinki.fi
Sebas-tian Hielm, tel. +358-9-19149715, e-mail: Sebastian.Hielm@helsinki.fi
Eija Hyytiä, tel. +358-9-19149701, e-mail: Eija.Hyytiä@helsinki.fi
Katri Jalava, tel. +358-9-19149712, e-mail: Katri.Jalava@helsinki.fi
Tiina Korte, tel. +358-9-19149701, e-mail: Tiina.Korte@helsinki.fi
Miia Lindström, tel. +358-9-19149702, e-mail: Miia.Lindstrom@helsinki.fi
Janne Lundén, tel.+358-9-19149766, e-mail: Janne.Lunden@helsinki.fi
Annukka Markkula, tel. +358-9-19149703, e-mail: Annukka.Markkula@helsinki.fi
Mari Nevas, tel.+358-9-19149702, e-mail:  Mari.Nevas@helsinki.fi
Maria Mietti-nen, tel. +358-9-19149701, e-mail: Maria.Miettinen@helsinki.fi
Leena Oksa, tel. +358-9-19149701, e-mail: Leena.Oksa@helsinki.fi.
National Public Health Institute:
Anja Siitonen, tel. +358-9-47448245, e-mail: Anja.Siitonen@ktl.fi
Saija Hallanvuo, tel. +358-9-47448747, e-mail: Saija.Hallanvuo@ktl.fi
Susanna Lukinmaa, tel. +358-2-47448747, e-mail: Susanna.Lukinmaa@ktl.fi
Teemu Rinttilä, tel. +358-9-47448747, e-mail: Teemu. Rinttilä@ktl.fi.
University of Turku:
Mikael Skurnik, tel. +358-2-3337441, e-mail: Mskurnik@utu.fi
Saija Kilju-nen, tel. +358-2-3337444, e-mail: Saija.Kiljunen@utu.fi
Maria Pajunen, tel. +358-2-3337444, e-mail:  Maria.Pajunen@utu.fi
Lotta Söderholm, tel. +358-2-3337444, e-mail: Lotta.Söderholm@utu.fi

Financing SYTTY organisation: Tekes
Funding from SYTTY / Total funding of project (€): 276669 / 950934
Person-months of work funded by SYTTY / Total person-months of work: 56 / 198,5

KEY WORDS: food, epidemiology, listeriosis, yersiniosis, botulism
 

EXTENDED ABSTRACT

1 Introduction

Approximately 400 000 to 500 000 individuals become sick due to food poisonings in Finland annually. The proportion  of the population which is more prone to these poisonings because e.g. of aging or immunosuppressive therapy is increasing in Finland. The use of new production and packaging methods, together with long shelf-life periods demanded by retail supply, have generated a new problem in the form of certain pathogenic psychrotrophic bacteria.

Psychrotrophic bacteria, such as nonproteolytic Clostridium botulinum, Listeria monocytogenes and Yersinia enterocolitica, capable of growing at chill temperatures, are considerable risk. These bacteria may enrich in food during cold storage if storage times are long enough. Thirty to 50 cases of listeriosis are reported in Finland annually. These have commonly been serious, life-threatening infections. There are approximately 600 to 700 annual cases of yersiniosis in Finland, in some of these cases reactive arthritis occurs as a complication. The first botulism outbreak caused by a product industrially manufactured in Finland was recorded 1999. In addition, Yersinia pseudotuberculosis emerged as a new outbreak-causing food-borne pathogen. Association of these risks with certain food types together with lacking means to control the risks have generated considerable difficulties for certain sectors in food industry.

The Finnish fishing waters and fish are commonly contaminated with C. botulinum spores. We have shown the spore contamination in fish products; of samples purchased from retail stores, 8% of hot-smoked rainbow trout and whitefish were positive, 3% of cold-smoked rainbow trout and 2% of raw-pickled, respectively. Botulism cases in Germany, caused by Canadian-origin whitefish warm-smoked in Finland, showed the hazard true also for Finland.  The causative strain was ascertained as type E in the Botulinum laboratory of the Department of Food and Environmental Hygiene. Vacuum-packaged, hot-smoked fish products can clearly be considered as a group of industrially produced food associated with a risk of C. botulinum.

L. monocytogenes survives in soil and at different food processing environments over long periods of time which may lead to food contamination during different production stages. This bacterium also tolerates low pH and high concentrations of sodium chloride. We have isolated L. monocytogenes from various vacuum-packaged meat products and especially from vacuum-packaged fish products. Prevalence of  L. monocytogenes in raw-pickled rainbow trout products was 29% and 17% in cold-smoked rainbow trout products, all products originating from retail sources

Pigs are the main reservoir of pathogenic Y. enterocolitica and insufficiently-heated pork is considered the main source of yersiniosis. Using a PCR-based method the prevalence of Y. enterocolitica in pig carcasses was shown to be 13 to 50% and 25% in pork-containing minced meat. The real link from food to yersiniosis is unclear because virulent strains have been isolated from food very seldom. This may, however,  be due to inefficient isolation methods.  The reservoirs for Y. pseudotuberculosis are not well known in Finland.

2 Methods

Together with the conventional methods, PCR methods and reporter bacteriophage (under development) are used for diagnostics. PFGE, ribotyping, rfb-gene typing, rep-PCR and RAPD have been developed and are used for characterisation of the isolates of C.botulinum, L. monocytogenes and Y. enterocolitica.

3 Results and discussion

Clostridium botulinum
Several techniques for molecular subtyping of C. botulinum strains have been developed, and protocols for the following methods have been published: ribotyping, RAPD- and Rep-PCR. Of these, ribotyping and Rep-PCR were able to distinguish between the species' proteolytic and non-proteolytic groups, whilst they were not that useful in distinguishing minute differences between strains belonging to the same group. RAPD-PCR and pulsed-field gel electrophoresis are able to distinguish between strains of the same group. RAPD-PCR has the advantage that it can type all strains, even those that are untypeable by PFGE on the account of DNase degradation. PFGE, on the other hand, is the method of choice for establishing a C. botulinum fingerprint database, for which purpose the chaotic nature of the RAPD-reaction is much too unreproducible. A multiplex PCR –method that detects different C. botulinum types has been developed. The method enables the simultaneous detection of all human pathogenic C. botulinum types, A, B, E, and F. The ability of commercial test systems for the classification of anaerobic bacteria was also tested, and it was concluded that they are not trustworthy enough for the identification of C. botulinum. We are also taking our work on isolation of C. botulinum strains further, and are now in the process of evaluating a new immunology-based method.

We looked at the growth and toxin production of C. botulinum type E in fish products and concluded that even a very low number of organisms is capable of establishing itself and to produce toxin in vacuum-packaged fresh and cold-smoked fish. In raw pickled fish we were not able to observe any growth or toxin production. We tested the ability of commercial food prediction models to foresee growth and toxin production in these fish products, and came to the conclusion that they were not dependable enough for this purpose.

In a survey of Finnish trout farms we found C. botulinum type E in both the environment (68 % of farm sediments positive) and in the produced fish (15 % of intestinal samples and 5 % of skin samples positive). During the survey, a total of 28 genotypes were identified from 42 C. botulinum type E strains isolated in this survey. The genetic diversity of C. botulinum type E seems very broad, in another study we were able to divide 92 isolates into 62 different genotypes by PFGE. This diversity can successfully be applied to epidemiological investigations of botulism outbreaks, as we are now able to establish a definitive connection between C. botulinum strains isolated from disease-causing foods and strains isolated from botulism patients. Many such studies are now under way, including investigations of Canadian infant and food borne botulism epidemics as well as the first ever food borne botulism outbreak in Finland, in which a Finnish 54-year old man became severely ill with type E botulism after having eaten spoiled whitefish roe.

C. botulinum type B spores frequently contaminate meat products and cause potential health hazards in vacuum packaged meat products. The elimination of the health hazard requires sufficient heating which in turn may alter the sensory quality of the product in a non desired manner. A heating process for meat products have been developed to eliminate C. botulinum type B spores without changing the sensory quality of the product.

The prevalence of C. botulinum in honey has been studied and the results show that  about 5 – 10 % of both Finnish honey samples and honey samples from other countries are contaminated with C. botulinum spores. This study has resulted in recommendations concerning the consumption of honey of infants in order to avoid infant botulism.

Listeria monocytogenes
Human isolates were collected from all Finnish clinical microbiological laboratories from 1990 through 2000, and all isolates (310 isolates) have been characterized with serotyping and pulsed field gel electrophoresis typing (PFGE). PFGE divided the isolates into 90 genotypes which demonstrates that the clinical isolates have a large biodiversity which gives a good ground for epidemiological studies. The typing of all new clinical isolates have continued timely in order to recognize possible epidemics as soon as possible. A considerable amount of food related L. monocytogenes strains have also been isolated and typed which have provided information about risk products. An epidemic, with butter as a source, was successfully identified and solved by sero- and PFGE typing of clinical and food isolates. An electronic fingerprint library has been created in order to facilitate the comparison of the PFGE patterns. The applicability of the automated ribotyping procedure as a first stage method in epidemiological clustering of L. monocytogenes strains was started to evaluate among a large set of isolates representing various sero- and PFGE types.

The PCR –based detection of L. monocytogenes in food samples have been improved. The most difficult step in PCR –based detection of L. monocytogenes in food samples is the elimination of inhibiting factors. The new PCR –protocol has shown to be successful in eliminating inhibiting factors.

L. monocytogenes in poultry processing, fish and fish products was studied, resulting in better understanding of the epidemiology. L. monocytogenes contamination patterns in pig slaughterhouses have been studied. L. monocytogenes was detected from the environment, carcasses and pluck sets. The highest prevalence was detected in tongue and tonsil samples, 14% and 12%, respectively. The results show that the handling of the tonsils and the tongue are important in the contamination of the carcass and altered slaughter procedures should be evaluated.

L. monocytogenes is able to cause persistent contamination of slaughterhouses and food processing plants leading to contamination of products. The persistent contamination is difficult to eradicate and it is probable that some L. monocytogenes strains have properties that enables them to withstand the eradication procedures in the plants. The reasons leading to persistent contamination have been studied and it seems that enhanced adherence to food contact surfaces and adaptation to disinfectants affect the persistence of L. monocytogenes

Yersinia enterocolitica
The efficiency of PFGE in typing of the pathogenic Y. enterocolitica bioserotype 4/O:3 isolates was studied. The discriminatory index increased from 74 % to 93 % and the number of different genotypes from 15 to 30 when isolates with the same PFGE pattern with NotI were further characterised with ApaI and XhoI, which shows that the PFGE is  an efficient technique for characterisation of bioserotype 4/O:3.

The PFGE typing method was used in characterising isolates from slaughterhouses and food samples. The results showed that the yadA-positive Y. enterocolitica contamination rate of edible pig offal in slaughterhouses was high. It was also shown that pig meat is a significant source of pathogenic Y. enterocolitica in retail level. The results of the project show that the slaughtering process of the pig affects significantly the spread of the bacteria, and by changing slaughtering techniques the spread of the bacteria could be prevented. The most important changes in the slaughtering technique should be done in the handling of the head and the offal. Cats and dogs have also been shown to be sources of pathogenic Y. enterocolitica. The pets are infected with raw meat, and humans, especially children, are at risk to be infected by faecal material.

Clinical Y. enterocolitica isolates have been continuously collected and typed with PFGE for epidemiological purposes. The epidemiology of Finnish Y. enterocolitica O:3 isolates have also been studied further by genotyping over 120 strains of human and porcine origin with probe pAY100 (carrying a cloned fragment of O antigen gene cluster). The pAY100 sequences responsible for the genotyping potential were found also in other pathogenic serotypes in addition to serotype O:3, which gives an opportunity to genotype of group of serotypes (O:3, O:9, O:5,27, O:2 and O:1) responsible for over 90% of Y. enterocolitica infections worldwide. The pAY100 sequences responsible for the genotyping potential were absent in nonpathogenic Y. enterocolitica strains. The potential of this finding in differentiating between pathogenic and nonpathogenic Y. enterocolitica strains was studied further. A large set of Finnish clinical Y. enterocolitica strains was examined by comparing the presence of certain virulence associated phenotypic properties to the presence of pAY100 typing sequences. The true clinical significance of  Y. enterocolitica isolates collected from patients in different hospital laboratories in Finland are being evaluated with help of these studies. According to very preliminary results, some 30% of about 650 cases of yersiniosis officially reported in Finland annually may be of no clinical significance.

Construction of a reporter bacteriophage for rapid diagnostics of Y. enterocolitica O:3 has not been completed during the project. The characterisation of the wild type bacteriophage is completed and the hole genome has been sequenced. The host range of the bacteriophage T3 was changed to Y. enterocolitica O:3 –specific by swabbing the adhesin gene.

4 Conclusions

PCR-based detection and typing methods have been developed and improved for C. botulinum, L. monocytogenes and Y. enterocolitica.

PFGE typing, possessing a high discriminatory index, is a very efficient tool in typing C. botulinum, L. monocytogenes and Y. enterocolitica. It has been applied successfully in solving food borne outbreaks and studying contamination routes.

Genotyping Y. enterocolitica with probe pAY100 has proven to be a powerful typing method. This method enables also the development of a combined detection and subtyping method for pathogenic Y. enterocolitica strains.
Clinical isolates are continuously typed for epidemiological purposes. A national electronic fingerprint database has been created for L. monocytogenes. This enables rapid recognizion of epidemics and sources of epidemics.

The knowledge of the epidemiology of C. botulinum, L. monocytogenes and Y. enterocolitica has increased, and riskproducts have been detected.

L. monocytogenes is commonly found in raw meat. The contamination of carcasses could be affected by changes in slaughtering techniques.

Pig meat is a significant source of pathogenic Y. enterocolitica in retail level. Changes in slaughtering technique could prevent the spread of the bacteria to the meat. Cats and dogs are also sources of pathogenic Y. enterocolitica.

The bacteriophage T3 has been manipulated into Y. enterocolitica O:3 –spesific which will be used in the development of a rapid diagnostic tool for Y. enterocolitica O:3.

C. botulinum type E is capable of producing toxin in vacuum-packaged fresh and cold-smoked fish at 4º C even at a very low number of organisms. This has lead to increasing pressure to adjust storage temperature and time of these products.

A heating process for meat products has been developed to eliminate C. botulinum type B spores without changing the organoleptic quality of the product.

Honey was found to be contaminated with C. botulinum spores. These results have lead to recommendations of the consumption of honey of children under one year.

5 Co-operation

National collaborators: prof. Tuula Honkanen-Buzalski (L. monocytogenes) (National Veterinary and Food Research Institute),  Anna-Maija Sjöberg (L. monocytogenes) (Technical Research Centre of Finland), Raija Ahvenainen (C. botulinum) (Technical Research Centre of Finland).

International collaborators: John Austin (C. botulinum) (Health Protection Branch, Ottawa, Canada), prof. Wilhelm Holzap-fel (Taxonomy) (Insti-tu-te of Hygiene and Toxico-logy, Federal Re-search Centre for Nutriti-on, Karlsru-he, Germany), prof. Andreas Stolle (C. botulinum) (De-partment of Hygiene and Techno-logy, Univer-sity of Munich, Germany).
 

[ Projects | Main Page ]