FEMS Microbiology Reviews 15 (1994) 175-183 ,~c? 1994 Federation of European Microbiological Societies 0168-6445/94/$15.

0(1 Published by Elsevier

175

F E M S R E 0/)429

Microbes in food processing technology

H. H o f s t r a , J . M . B . M . van d e r V o s s e n and J. van der Plas *
TNO, bzstitute ]br Nuttftion and Food Research, Dil'ision for Agrotechnology and Microbiolo~q3', PO Box 350. 3700 AJ Zeist. the Netherlands

Abstract: There is an increasing understanding that the microbial quality of a certain food is the result of a chain of events. It is clear that the microbial safety of food can only be guaranteed when the overall processing, including the production of raw materials, distribution and handling by the consumer are taken into consideration. Therefore, the microbiological quality assurance of foods is not only a matter of control, but also of a careful design of the total process chain. Food industry has now generally adapted quality assurance systems and is implementing the Hazard Analysis Critical Control Point (IIACCP) concept. Rapid microbiological monitoring systems should be used in these cases. There is a need for rapid and simple microbiological tests which can be adapted to the technology and logistics of specific production processes. Traditional microbiological methods generally do not meet these high requirements. This paper discusses the tests, based on molecular biological principles, to detect and identify microbes in food-processing chains. Tests based on D N A technology are discussed, including in vitro D N A amplification like the polymerase chain reaction (PCR) method and identifications based on RFLP, R A P D and D N A fingerprinting analysis. PCR-based methodology can be used for the rapid detection of microbes in food manufacturing environments. In addition, D N A fingerprinting methods are suitable for investigating sources and routes of microbial contamination in the fl)od cycle. Key words: Microorganisms: Food processing; HACCP: Detection: D N A amplification: Strain typing

Introduction

The microbiological quality, including safety, of food is influenced by various circumstances during its production and processing. Microbiological quality can only be guaranteed when the total production chain is taken into consideration. Quality Assurance systems (QA) can be used to improve quality and reduce costs; Hazard Analysis Critical Control Point (HACCP) programs can be used to better assure food safety. QA and HACCP, implemented in concert, facilitate improvements in product safety and production efficiency [1]. HACCP may virtually eliminate the need for extensive end product testing

* Corresponding author. Tel.: ( + 31-3404) 44709; Fax: ( + 31341)4) 54186; e-mail: vanderplas(ct voeding.tno.nl.

by identifying the safety risks at critical points and subsequent monitoring these control points in the production process. Through HACCP, i.e. continuous (real time) monitoring starting as far upstream in the processing system as possible, microbiological, chemical and physical risks are under control by anticipation and prevention. Traditional microbiological methods, like selective enrichment culturing and plating, generally lack the required speed and specificity. Modern microbiological methods, based on immunological and molecular genetic principles, are better suited for use in identification of CCPs and rapid assessment of microbiological quality. In this paper we will evaluate some of the molecular biological techniques or combinations of these techniques that can be used in microbiological quality assurance systems in the food industry.

SSDI 0 1 6 8 - ( ~ 4 4 5 ( 9 4 ) 0 0 0 3 1 - S

176

Rapid microbial detections for monitoring

A number of rapid analytical techniques in food microbiology are now available [2]. In certain areas of the food industry, automated instrumental microbiological analyses such as impedimetry and turbidimetry have had a significant impact as a hygiene parameter for total microbial and enterobacterial counts in routine food analysis [3]. The ATP bioluminescence assay has also found a wide application as a hygiene parameter, as it provides almost instantaneous information about total bacterial contamination of a product or a surface [4]. A major disadvantage of these methods is that they basically give information about colony forming units only. In contrast, modern techniques in microbial detection like immunological assays or nucleic acid hybridisation tests can provide information on specific bacterial species or virulence types and also about microbial metabolites (toxins etc.). One of the major concerns among microbiologists with respect to the implementation of these new techniques is their sensitivity, particularly when applied to complex samples like foods. Indeed, the sensitivity of most techniques is not high enough to allow direct detection of specific bacteria. Another drawback for implementation in routine analysis is the fact that the cost per test is relatively high. In addition, the methodological variations, especially in the treatment preceding the actual bioanalysis, are also responsible for the fact that biotechnological tests are only sporadically used as a routine analysis in microbiological monitoring systems. After a short description of biotechnological techniques that have potential to be sensitive, fast and specific, we will discuss some applications.

Nucleic acid hybridisation technology

Several types of DNA hybridisation-based methods can be distinguished. Firstly, the classical hybridisation with a labelled DNA probe. In this technique, bacteria are applied to a solid phase, usually a nitrocellulose or nylon filter, After lysis of the bacteria the DNA is fixed to the

filter and labelled DNA probes are added and allowed to react with their counterstrand, if present [5,6]. In this type of reaction, the DNA present in the sample is not multiplied. Its sensitivity is therefore dependent on the initial amount of DNA. This is also the case when hybridisation with a labelled probe is performed in solution, followed by capture of the target DNA or RNAprobe hybrid by an immobilised second probe [7]. For such hybridisation tests, extensive (selective) pre-enrichment procedures are needed to obtain the minimum number of cells required (approximately 10 5 cells/ml). This can be done along the lines of the classical microbiological methodology. In case of Salmonella, this allows the detection of the bacterium within 2-3 days, including sample pretreatment. For this purose Salmonella probes derived from single copy DNA or from genes with a low copy number can be used. A more recent development is the primer-induced in vitro DNA amplification, e.g. the polymerase chain reaction (PCR). In the PCR type reactions, small but specific DNA primers are added to the DNA sample. If they meet their complementary strand, they will hybridise. A subsequent polymerase activity doubles the initial amount of specific DNA. This cycle can be repeated, in principle, for an unlimited number of times [8]. Alternative in vitro DNA amplifications have also been developed. Among these are Qbeta replication [10] and the newly introduced self-sustained sequence replication [11,12]. In principle, all these methods could be applied for obtaining highly specific single-step detections of low numbers of microorganisms. Several specific DNA probes and PCR primers suitable for use in food microbiological analyses are now available for implementation. Universal Salmonella probes and PCR primers, Campylobacter species-specific probes and PCR primers and several probes for Escherichia coli virulence types and Shigella species have been constructed [9]. DNA hybridisation techniques and PCR have the potential to play an important part in future microbiological identification and monitoring systems in the food sector as well as in clinical and environmental settings. DNA hybridisation on ill-

177

ters with labelled DNA probes will predominantly be used as identification or confirmation of pure cultures. DNA amplification is most suitable for the rapid detection ( p r e s e n c e / a b s e n c e ) of microorganisms in clinical, food or environmental samples and for definitive identification by direct sequencing of e.g. PCR-amplified rRNA genes. Another group of DNA analyses are those resulting in specific DNA patterns. These methods, as e.g. the Restriction Fragment Length Polymorphism (RFLP) analyses, the Random Amplification of Polymorphic D N A (RAPD), the DNA Amplified Fingerprinting (DAF) and other fingerprinting-like methods can be used for high resolution typing of microorganisms and they may gradually replace traditional serotyping systems. This paper will evaluate the suitability of some of these new approaches in the next paragraphs. Therefore, the following three areas can be distinguished for application of DNA probe technology: (1) Identification of microorganisms; (2) Rapid detection of microorganisms; (3) High resolution typing of microorganisms.

Microbial identification Microbial identifications can be performed on pure cultures by the classical DNA-hybridisation technique. Specific sets of labelled D N A probes may replace the biochemical and serological identification systems generally used today. D N A probes, to be used for identification purposes may be derived from several sources. The most common way to obtain specific probes is the selection of random D N A fragments from a gene bank [6,13]. In this case, fragments are selected on the basis of the specificity that is desired for their application. We selected random D N A probes specific for the different Campylobacter species using this method. DNA probes specific for C. jejuni, C. coli and the thermophilic Campylobacter species have thus been selected by cross-hybridisation of C. jejuni and C. coli D N A fragment libraries with labelled Campylobacter chromosomal DNAs [14,15]. In dotblot hybridisa-

tion experiments these probes were used for the identification of hundreds of previously isolated and characterised Campylobacter species. In all cases, a very good correlation with the few classical biochemical tests at hand was observed [15]. Using about the same procedure of selection by differential hybridisation, universal Salmonella probes were developed. A second source of DNA probes for microbial identifications are specific virulence genes [6,16]. Genes coding for microbial virulence factors allow a direct answer to the question whether certain pathogenic types of bacteria are present in a specimen, without the need to detect non-virulent types of the same or related microorganisms. For example, in the case of E. coli, this allows the detection of enteropathogenic, enterotoxigenic and entero-invasive E. coli as potential pathogens, as opposed to commensal E. coli, which could be detected as a hygiene parameter for faecal contamination [17]. A third source of DNA probes which is used very successfully by a number of research groups are the conserved and less conserved areas of common microbial genes. The most important group of genes in this respect are the ribosomal RNA coding genes, especially the 16S rRNA genes in numerous bacteria. The use of the variable regions of the 16S rRNA as a source of specific microbial probes has been documented extensively [6,18,19]. A less well-known source of probes, especially for the Gram-negative bacteria, is given by the genes coding for the outer membrane proteins. The enterobacterial outer membrane proteins cross the membrane several times and therefore exist of membrane-spanning conserved regions intermitted by exposed variable regions. The sophisticated use of selected nucleotide sequences from the variable regions created the opportunity to develop specific oligonucleotide probes for bacteria from the family Enterobacteriaceae [20-23]. Rapid detection In most practical situations, the number of cells of a particular pathogenic microorganism present in foods or other specimens is below the

17N level for direct detection by classical D N A - p r o b e hybridisation methods. Therefore, D N A probes as discussed in the previous paragraphs are better suited for the identification of the pure microbial cultures resulting from pre-enrichment and selective enrichment procedures. An alternative for microbiological enrichment cultures evolved in the specific in vitro enrichment (amplification) of genetic sequences ( D N A or RNA). Although a number of different in vitro amplification techniques exist (see previous section), only the Polymerase Chain Reaction (PCR) found a wider application so far, PCR assays have been developed that are specific for many bacteria relevant in food microbiology [24-31]. The PCR primers involved arc usually selected from the same sources of D N A as those discussed in the previous section on the origin of specific probes. Howcvcr, direct PCR detection of pathogens in fl)ods without extensive prctreatment of the samples appeared not to be feasible for several reasons. In the first place, the enzymatic D N A amplification reaction is strongly influenced by components from the food matrix in which the detections have to bc carried out [25,32,33]. In the second place, the sample size imposed by low contamination levels to be detected (e.g. 25 g for Sahnonella p r e s e n c e / a b s e n c e testing) usually cannot be accommodated by the small PCR reaction volume (5(I-100 /xl). In the third place, a significant contribution from the presence of dead cells (or even naked DNA) needs to bc eliminated, except in those cases where one is explicitly interested in previous contamination levels. In order to perform PCR analysis of food samples, creative combinations [3,34] of purification, cell concentration and culturing methods should be adapted to the requirements posed by each specific sample matrix, which varies widely from meat to milkpowder or vegetables. This tailoring of sample pretreatment before PCR is probably the most critical step hampering the implementation of PCR detection in the food sector, with the additional restriction that complicated methods for the isolation of microorganisms or their D N A as used in the clinical laboratory are not suitable for routine analysis of large numbers of samples in industrial food quality assurance. Simple and dependable protocols are currently being developed for the application of PCR detection of several pathogenic food microorganisms in a range of matrices: e.g. Listeria in milk [35], cheese [33,36] and meat products [37]; Salmonella and Campylobacter in poultry ([24,38], respectively). Regarding Salmonella PCR detection, our studies were directed towards the reduction of the detection time from the standard 4 - 6 days to less than 2 days. This Salmonella PCR was based on the best conserved region of a Salmonellaspecific chromosomal D N A fragment selected by differential hybridisation, as described above. This assay detected all 130 Sahnonella strains from all subclasses and no false-positives were fl)und among 49 non-Salmonella (manuscript in preparation). Our strategy for the implementation of this Salmonella-specific PCR in meat and meat products was to start with PCR from the standard resuscitation and enrichment culture {39]. PCR inhibiting components from meat and media were sufficiently removed by a simple centrifugation step. This already resulted in a reduction of 3 days when compared with the classical plate confirmation method fl)r Salmonella-positive samples. Excellent correlation was found between the classical Sahnonella p r e s e n c e / a b s e n c e test (ISO 6579) and the combination of standard enrichment with PCR. Subsequently, we further reduced the detection time to 24 h by modification of the enrichment procedure to a 20-h culturing step previous to PCR. This was possible because the selective enrichment for Sahnonella before PCR is not as critical as that before the classical plating on specific agars for colony identification purposes. The most important objectives of culturing before PCR is to dilute inhibiting components and nongrowing cells and to increase the concentration of viable cells. With PCR the presence of an excess of cells from the contaminating flora is much less important, as long as it does not reduce the growth of the target organism. From this point of view culturing before PCR deserves special attention.

179

Further reduction of the detection time for Salmonella in meat products is now being evaluated using a combination of enrichment and immunomagnetic separation (IMS) [24,40,41]. This gives an additional purification from matrix components and contaminating flora, although it adds to the complexity of the procedure. With regard to the duration of the assay it should be noted that usually for samples taken at day 1 the results will not become available before day 2, unless the detection can be performed well within 8 h. An overnight incubation (24-h assay) is often a good way to take advantage of this procedural lag-time. Sample pretreatment should be as simple as possible to facilitate introduction of the developed techniques in routine laboratories. A fast screening method should be completely adapted to the specific requirements of the PCR and the particular food matrix. A similar approach as described for Salmonella is followed for other food-borne pathogens, such as Campylobacter. A multiprimer PCR for C. jejuni/C, coli gives direct identification of the species after enrichment from poultry. Where the classical hippurate hydrolysis is tedious and time-consuming [42]. The results of the PCRbased screening method showed a 98% correlation with the classical detection method.

High-resolution typing
Bacteria
High resolution typing of microorganisms is important in many areas of microbiology [43]. Essential for the HACCP concept is the need to identify and type microorganisms in order to design measures to avoid their growth and to trace the route of contamination [44]. Microbial ecology of potentially pathogenic microorganisms in food, in farm animals and in the environment is the area of our prime interest and Salmonella has been the first organism to test certain methodological approaches for strain monitoring [45]. Detailed strain identification, i.e. subdivision of the most common serovars and other serovars of particular importance, is essential for investiga-

tions of e.g. Salmonella outbreaks. Clonal associations of isolates in an outbreak may be shown by demonstrating common characteristics. Equally important is that strains can be discriminated from unrelated isolates in the same set of characteristics. Complete identity of genotypes is normally not considered as a prerequisite for strain identity, as bacteria may undergo genetic variation during epidemic spread [46]. The situation to be studied dictates both the stringency of the term clone and the typing methodology to be used. In general, serological tests in combination with phage typing are being used to elucidate the type and origin of Salmonella infection. A major drawback of these specialised techniques for typing in detail is that only specialised institutes can afford to use them routinely, as these techniques rely on the availability of highly specific antisera and phage lysates. Moreover, these sera and lysates are difficult to standardise with the consequence that international standardisation is also difficult. However, in order to compete with the golden standard of serotyping, new methods should have favourable features, like universal applicability, stability and polymorphism. For the investigation of local minor outbreaks, rapid comparative methods, which tend to require typing of all isolates in parallel, may be used. However, methods that are both definitive and highly reproducible must be applied to compare isolates over a considerable time span and at different geographical locations [43], DNA methods, like plasmid analysis, restriction enzyme analysis (REA), pulsed field gel electrophoresis (PFGE) and restriction fragment length polymorphism (RFLP) analysis can be used to study the genotypes of a variety of both Gram-positive and Gram-negative bacteria, without the need of specific modifications to the species to be studied [43]. Furthermore, DNA typing methods, unlike bio-, sero- and phagetyping, are not based on often complex expression of phenotypic traits, and seldom show untypable strains. The use of ribosomal RNA probes in a Southern blot analysis to determine restriction fragment length polymorphisms in the rRNAcoding DNA of bacterial strains provides a more

180

stable and conserved marker system than plasmid DNA. Ribosomal RNAs (rRNA) are ubiquitous and highly conserved molecules. Escherichia coli 16 + 23S r R N A was proposed as a universal probe to visualize rRNA gene restriction patterns [47]. This r R N A gene restriction pattern (ribotype) approach is increasingly used to build molecular identification or typing schemes. More recently, rapid PCR-typing methods evolved that are based on random amplification of polymorphic DNA (RAPD) [48,49], on amplification of restriction fragment polymorphisms [50,51], or on amplification with primers derived from conserved repetitive sequences: REP-PCR (REP stands for Repetitive Extragenic Palindromic element) and ERIC-PCR (ERIC stands for Enterobacterial Repetitive Intergenic Consensus sequence) [52,53] or eukaryotic repeated sequence motives [54,55]. We extended the ribotyping method by developing additional generally applicable RFLP probes from common (eu)bacterial genes. Protein encoding sequences from macromolecules that are widely distributed often show a high degree of functional constancy and are sufficiently conserved to span the full evolutionary spectrum. The only requirement for production of such probes by PCR [8] is the availability of nucleotide sequence information and a minute amount of the original DNA, in the form of a plasmid or chromosomal DNA or even intact cells. Based on the sequence, PCR primers are synthesized that allow amplification of the region of choice: e.g. the complete coding sequence of the gene of interest or if possible only the precise portion that is best conserved among the known amino acid sequences. In this way primers were selected from the genes encoding ATPase, RNA polymerase, elongation factor Ef-Tu and ribosomal protein S12, and RFLP probes were produced by PCR amplification (Van der Plas et al., in preparation). The universal probes hybridised with a wide range of bacteria and were successfully used to type larg e numbers of Salmonella [45], Lactobacillus, Campylobacter and Helicobacter isolates with high resolution [15]. The RFLP analyses with these D N A probes supplement the ribotyping data and add extra levels of discrimination because the probes map to different sites of the

chromosome and have flanking regions with differing sequence stability. The resolution of the RFLP analysis can be adjusted at will by choosing the appropriate number and kind of restriction e n z y m e / p r o b e combinations.

Yeasts and fungi

The identification, classification, and typing of the eukaryotic microorganisms using classical methods based on morphological, physiological or biochemical characteristics have proven to be rather difficult and time-consuming. Recently, Meyer et al. [56,57] have shown the applicability of the oligonucleotide fingerprinting for the analysis of fungal genomes. These fingerprints showed a high degree of polymorphism demonstrating the applicability of this method to discriminate between fungal species and strains. Lieckfeldt et al. [58] suggested reclassification of the genus Trichoderma, based on DNA fingerprint results. During classification of other fungi, in particular Fusarium species that cause losses in the primary production in the agricultural sector, we also came to the conclusion that the traditional classification does not always accurately reflect genetic relationships. The observation by Lieckfeldt et al. [58] that yeasts analysed by D N A fingerprinting showed representative D N A patterns characteristic for each species investigated (e.g. Candida, Kluyveromyces, and Saccharomyces), is confirmed by RAPD analysis of DNA from yeast species in our laboratory (e.g. Sacharomyces cerevisiae, Zy-

gosaccharonyces baili, Z. rouxii, Candida valida, C. lipolytica, Rhodotorula rubra). Within the genus Saccharomyces, two subclasses could be distinguished using different oligonucleotides in RAPD. Minor bands occasionally observed in the RAPD patterns allowed discrimination on the strain level. In addition, we found that RFLP analysis with probes derived from the highly conserved eukaryotic 17S rRNA and ATPase genes is also applicable to differentiate between yeast species.

Conclusions
PCR-typing methods, RAPD and PCR fingerprinting with repetitive sequences appear to be

181

rapid and discriminative and should be very helpful in the comparison of limited numbers of microbial isolates in parallel, e.g. for the investigation of local minor outbreaks. For a wider application the reproducibility has to be enhanced as the method is too susceptible to minor variations in experimental conditions. D N A typing based on hybridisation like RFLP, though more laborious, is highly reproducible and better suited for comparing isolates over a considerable time span and at different geographical locations. The genomic variation as observed in D N A typing is not only extremely helpful in taxonomic and phylogenetic studies, but will also be of great help for tracing routes of contamination and designing measures to prevent food infection and spoilage.

References
1 Microbiology and food safety committee of the national food processors association (1992) HACCP and Total Quality Management - Winning concepts for the 90s: a review. J. Food Protect. 55, 459-462. 2 Huis in 't Veld, J., Hartog, B. and Hofstra, H. (1988) Changing perspectives in food microbiology: implementation of rapid microbiological analyses in food microbiology. Food Rev. Int. 4, 271-329. 3 Huis in 't Veld, J.H.J. and Hofstra, H. (1991) Biotechnology and the quality assurance of foods. Food Biotechnol. 5, 313-322. 4 Jay, J.M. (1978) Modern Food Microbiology. Van Nostrand Company, New York, NY. 5 Matthews, J.A. and Kricka, L.J. (1988) Analytical strategies for the use of DNA probes. Anal. Biochem. 169,
1-25.

6 Tenover, F.C. (1988) Diagnostic deoxyribonucleic acid probes for infectious diseases. Clin. Microbiol. Rev. 1, 82-10l. 7 King, W., Raposa, S., Warshaw, J., Johnson, A., Halbert, D. and Klinger, J.D. (1989) A new colorimetric nucleic acid hybridization assay for Listeria in foods. Int. J. Food Microbiol. 8, 225-232. 8 Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S., Higuchi, R., Horn, G.T., Mullis, K.B. and Erlich, H.A. (1988) Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487-491. 9 Hofstra, H. and Huis in 't Veld, J.H.J. (1990) Bioanalytical methods; The application of monoclonal antibodies and DNA hybridization in food microbiology. In: Processing and Quality of Foods. Vol. 2. Food Biotechnology (P. Zeuthen et al., Eds.), pp. 281-287. Elsevier, London. 10 Lizardi, P.M., Guerra, C.E., Lomeli, H., Tussie-Luna, I.

and Kramer, F.R. (1988) Exponential amplification of recombinant RNA hybridization probes. Bio/Technology 6, 1197-1202. 11 Gingeras, T.R., Kwoh, D.Y., Fahy, E., Whitfield, K., Richman, D.D. and Guatelli, J.C. (1990) Self-sustained replication (3-SR) reaction: an isothermal method for the in vitro amplification of nucleic acids. VIth Int. Congr. Rapid Methods Autom. Microbiol. lmmunol. 17. (Abstract) 12 Gingeras, T.R., Richman, D.D., Kwoh, D.Y. and Guatelli, J.C. (1990) Methodologies for invitro nucleic acid amplification and their applications. Vet. Microbiol. 24, 235-251. 13 Fitts, R., Diamond, M., Hamilton, C. and Neri, M. (1983) DNA-DNA hybridization assay for the detection of Salmonella spp. in foods. Appl. Environ. Microbiol. 43, 1146-1151. 14 Van der Plas, J. and Hofstra, H. (1991) Rapid detection and identification of Campylobacter species in food using DNA probes. In: Agricultural Biotechnology in Focus in the Netherlands (J.J. Dekkers H.C. Van der Plas and D.H. Vuijk, Eds.), pp. 140-145. PUDOC, Wageningen. 15 Van der Plas, J., Hofstra, H., Huis in 't Veld, J.H.J., Bijker, P., Biesheuvel, M., Glupczynski, Y., Goossens, H. and Butzler, J.P. (1993) DNA probe assays for detection, identification and typing of Campylobacter species and Helicobacter pylori. Microb. Ecol. Health Dis. 4, $60. 16 Rubin, F.A., Kopecko, D.J., Noon, K.F. and Baron, L.S. (1985) Development of a DNA probe to detect Salmonella typhi. J. Clin. Microbiol. 22, 600-605. 17 Hofstra, H. and Huis in 't Veld, J.H.J. (1988) Methods for the detection and isolation of Escherichia coli including pathogenic strains. J. Appl. Bacteriol. Syrup Suppl. 197S212S. 18 Giovannoni, S.J., DeLong, E.F., Olsen, G.J. and Pace, N.R. (1988) Phylogenetic group-specific oligodeoxynucleotide probes for identification of single microbial cells. J. Bacteriol. 170, 720-726. 19 Haun, G. and G6bel, U. (1987) Oligonucleotide probes for the genus-, species- and subspecies specific identification of the genus Proteus. FEMS Microbiol. Lett. 43, 187-193. 20 Spierings, G., Hofstra, H., Huis in 't Veld, J.H.J., Hoekstra, W. and Tommassen, J. (1989) Development of enterobacterium-specific oligonucleotide probes based on the surface-exposed regions of outer membrane proteins. Appl. Environ. Microbiol. 55, 3250-3252. 21 Spierings, G., Van Silfhout, A., Hofstra, H. and Tommassen, J. (1992) Identification of Klebsiella pneumoniae by DNA hybridization and fatty acid analysis. Int. J. Syst. Bacteriol. 42, 252-256. 22 Spierings, G., Elders, R., Van Lith, B., Hofstra, H. and Tommassen, J. (1992) Characterization of the Salmonella typhimurium phoE gene and development of Salmonellaspecific DNA probes. Gene 122, 45-52. 23 Spierings, G., Ockhuijsen, C., Hofstra, H. and Tommassen, J. (1992) Characterization of the Citrobacter freundii phoE gene and development of C. freundii-specific oligonucleotides. FEMS Microbiol. Lett. 99, 199-204. 24 Fluit, A.C., Widjojokusumo, A., Box, A.T., Torensma, R.

182 and Verhoef, J. (1993) Rapid detection of Salmonellae in poultry with the magnetic immuno-polymerase chain reaction assay. Appl. Environ. Microbiol. 59, 1342 1346. Meyer, R., Li~thy, J. and Candrian, U. (1991) Direct detection by polymerase chain reaction (PCR) of Escherichia coli in water and soft cheese and identification of enterotoxigenic strains. Lett. Appl. Microbiol. 13, 268-271. Rossen, L., Holmstrom, K., Olsen, J.E. and Rasmussen, O.F. (1991) A rapid polymerase chain reaction (PCR)based assay for the identification of Listeria monocytogenes in food samples. Int. J. Food Microbiol. 14. 145 152. Oyofo, B.A., Thornton. S.A., Burr, D.H., Trust, T.J., Pavlovskis, O.R. and Guerry, P. (1992) Specific detection of Campylobacter jejuni and Campylobacter coli by using polymerase chain reaction. J. Clin. Microbiol. 3(/, 26132619. Islam, D. and Lindberg, A.A. (1992) Detection of Shigella dysenteriae type-1 and Shigella flexneri in feces by immunomagnetic isolation and polymerase chain reaction. J. Clin. Microbiol. 30, 2801-2806. Kwaga, J., Iversen, J.O. and Misra, V. (1992) Detection of pathogenic Yersinia enterocolitica by polymerase chain reaction and digoxigenin-labeled polynucleotide probes. J. Clin. Microbiol. 30, 2668-2673. Koch, W.H., Payne, W.L., Wentz, B.A. and Cebula, T.A. (1993) Rapid polymerase chain reaction method for detection of Vibrio cholerae in foods. Appl. Environ. Microbiol. 59, 556-56(I. Tsen, H.Y. and Chen, T.R. (1992) Use of the Polymerase Chain Reaction for specific detection of Type-A, Type-D and Type-E enterotoxigenic Staphylococcus aureus in foods. Appl. Microbiol. Biotechnol. 37, B85 690. Rossen, L., Norskov, P., tIolmstmm, K. and Rasmussen, O.F. (1992) Inhibition of PCR by components of food samples, microbial diagnostic assays and DNA-extraction solutions. Int. J. Food Microbiol. 17, 37-45. Wernars, K., Heuvelman, C.J., Chakraborty, T. and Notermans, S.H.W. (1991) Use of the polymerase chain reaction for direct detection of Listeria monoo, togenes in soft cheese. J. Appl. Bacteriol. 7{I, 121-126. Sharpe, A.N. (1991) On multiplexed separations in quick detection of micro-organisms in foods - Letter to the editor. Food Microbiol. 8, 167-170. Starbuck, M.A.B., Hill, P.J. and Stewart, G.S.A.B. (1992) Ultra sensitive detection of Listeria monoeytogenes in milk by the polymerase chain reaction (PCR). Lett. Appl. Microbiol. 15, 248-252. Fluit, A.C., Torensma, R., Visser, M.J.C., Aarsman, C.J.M., Poppelier, M.J.J.C., Keller, B.H.I., Klapwijk, P. and Verhoef, J. (1993) Detection of Listeria monocytogenes in cheese with the magnetic immuno-polymerase chain reaction. Appl. Environ. Microbiol. 59, 1289 1293. Niederhauser, C., Candrian, U., Hofelein, C., Jermini, M., Buhler, H.P. and Luthy, J. (1992) Use of polymerase chain reaction for detection of Listeria monoeytogenes in food. Appl. Environ. Microbiol. 58, 1564-1568. Giesendorf, B.A.J., Quint, W.G.V., Henkens, M.tt.C., Stegeman, H., Huf, F.A. and Niesters, H.G.M. (1992) Rapid and sensitive detection of Campylobacter spp. in chicken products by using the polymerase chain reaction. Appl. Environ. Microbiol. 58, 3804-3808. Ten Bosch, C., Van der Plas, J.. Havekes, M., Geurts. J.M., Van der Palen, C., Huis in 't Veld, J.lt.J. and Hofstra, H. (1992) PCR assay for Salmonella. Proceedings Salmonella and Salmonellosis, pp. 44-49. P l o u f r a g a n / Saint-Brieuc, France. Skjerve, E. and OIsvik, O. (1991) Immunomagnetic separation of Salmonella from foods. Int. J. Food Microbiol. 14, 11-17. Vermunt, A.E.M., Franken, A.A.J.M. and Beumer, R.R. (1992) Isolation of Salmonellas by immunomagnetic separation. J. Appl. Bacteriol. 72, 112-118. Morris, G.K., El Sherbeeny, M.R., Patton, C.M., Kodaka, H., Lombard, C.L., Edmonds, P., Hollis, D.G. and Brenner, P.J. (1985) Comparison of four hippurate hydrolysis methods for identification of thermophilic Campylobacter spp. J. Clin. Microbiol. 22, 714-718. Olsen, J.E. and Brown, D.J. (1991) Molecular methods for defining a clone. In: Salmonella colonization in poultry, (Duchatelle, R., Haesebrouck, F. and Mulder R.W.A.W., Eds.), Proceedings of a F L A I R No B / C O S T No 9(16 meeting in Ghent. pp. 49-56. Spelderholt, Beekbergen. Slade, P.J. (1992) Monitoring Listeria in the food production environment. 2. Identification Techniques. Food Res. Int. 25, 21)3-214. Van der Plas, J., ten Bosch, C., Brinks, G., Geurts, J.M., Huis in "t Veld, J.H.J. and Hofstra, H. (1992) Restriction F r a g m e n t Length Polymorphism (RFLP) typing of Salmonella strains. Proceedings Salmonella and Salmonellosis, pp. 61-67. Ploufragan/Saint-Brieuc, France. Orskov, F. and 10rskov, I. (1983) Summary of a workshop on the clone concept in epidemiology, taxonomy and ew> lution of the Enterobacteriaceae and other bacteria. J. Infect. Dis. 148, 346-357. Grimont, F. and Grimont, P.A.D. (1986) Ribosomal nucleic acid gene restriction patterns as potential taxonomic tools. Ann. Inst. Pasteur/Microbiol. 137B, 165-175. Welsh, J. and McClellan& M. (1990) Fingerprinting genomes using PCR with arbitrary primers. Nucleic Acids Res. 18, 7213-7218. Williams, J.G.K, Kubelik, A.R., Livak, K.J., Rafalski, J.A. and Tingey, S.V. (1990) D N A polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 18, 6531-6535. Akopyanz, N., Bukanov, N., Westblom, T.U. and Berg, D.E. (1992) PCR-based RFLP analysis of D N A sequence diversity in the gastric pathogen Helieobacter pylori. Nucleic Acids Res. 2(/, 6221-6225. Ralph, D., McClelland, M., Welsh, J., Baranton, G. and Perolat, P. (1993) Leptospira species categorized by arbitrarily primed polymerase chain reaction (PCR) and by mapped restriction polymorphisms in PeR-amplified r R N A genes. J. Bacteriol. 175, 973 981. Versalovic, J., Koeuth, T. and Lupski, J.R. (1991) Distribu-

25

39

26

40

27

41

42

28

29

43

30

44

31

45

32

46

33

47

34

48

35

49

36

50

51

37

38

52

183 tion of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 19, 6823-6831. 53 Giesendorf, B.A.J., Van Belkum, A., Koeken, A., Stegeman, t t , Henkens, M.H.C., Van der Plas, J., Goossens, H., Niesters, H.G.M. and Quint, W. (1993) Development of species specific DNA probes for Campylobacter jejuni, Campylobacter coli and Campylobacter lari by PCR fingerprinting. J. Clin. Microbiol. 31, 1541-1546. 54 Riley, D.E., Samadpour, M. and Krieger, J.N. (1991) Detection of variable DNA repeats in diverse eukaryotic microorganisms by a single set of polymerase chain reaction primers. J. Clin. Microbiol. 29, 2746-2751. 55 Niesters, H.G.M., Goessens, W.H.F., Meis, J.F.M.G. and Quint. W.G.V. (1993) Rapid, polymerase chain reactionbased identification assays for Candida species. J. Clin. Microbiol. 31,904-910. 56 Meyer, W., Koch, A., Niemann, C., Beyermann, B., Epplen, J.T. and B6rner, T. (1991) Differentiation of species and strains among filamentous fungi by DNA fingerprinting. Curt. Genet. 19, 239 242. 57 Meyer, W., Lickfeldt, E., Woestemeyer, J. and Boerner, T. (1992) DNA fingerprinting for differentiating aggressivity groups of the rape seed pathogen Leptosphaeria macuhms. Mycol. Res. 96, 651 657. 58 Lieckfeldt, E., Meyer, W., Kuhls, K. and B6rner, T. (1992) Characterization and differentiation of filamentous fungi and yeasts by DNA-fingerprinting and random amplified polymorphic DNA. Belg. J. Bot. 125, 226-233.