SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Aims: In order to study the transmission of Listeria monocytogenes in a poultry and a pork meat plant, we analysed the contamination by this pathogen over several months.

Methods and Results: Five hundred and two isolates of L. monocytogenes were collected and characterized by genotyping and serotyping. Thirty-seven genotypes were obtained by ApaI-restriction analysis-pulsed field gel electrophoresis (REA-PFGE) and 35 by SmaI-REA-PFGE and resulted in 50 combined genotypes. The tracing of the contamination in both plants showed that some clones were able to survive for several months. However, some other clones were found only during processing operations, were not detectable after cleaning and seemed to enter continuously into the plant.

Conclusions: Some L. monocytogenes strains may persist for a long period in the plant environment. Different genotypes can be associated with poultry as well as pork meat.

Significance and Impact of the Study:Listeria monocytogenes contamination can be due to contaminated raw materials, bacterial spread and also ineffective cleaning procedures.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Listeria monocytogenes has emerged as one of the major human foodborne pathogens. Listeriosis can occur as sporadic disease or as outbreaks. In France, it has been associated with three major outbreaks in humans: in 1992 with pork tongue in jelly, in 1993 with pork ‘rillettes’ and in 1995 with a soft cheese ‘Brie de Meaux’ (Jacquet et al. 1995).

As outbreaks are linked with a food product, it is interesting to learn about the epidemiology of L. monocytogenes in food industry plants. Epidemiological studies are able to indicate the potential sources of contamination, trace the contamination in the plant and enhance knowledge about the environmental conditions where L. monocytogenes can survive and develop in a plant. Studies have shown that the first major amplification source of food product contamination with L. monocytogenes might be cross contamination, which occurs in the environment of the processing plants (Graham and Colins 1991; Wendtland and Bergann 1994). This phenomenon can be enhanced by the psychrotrophic nature of L. monocytogenes (Helke and Wong 1994) and its ability to adhere to various surfaces (Blackman and Frank 1996). The colonization of working surfaces represents a potential source of contamination of food products (Cox et al. 1989; Sammarco et al. 1997; Wong 1998). In a dairy plant, Unnerstad et al. (1996) showed contamination lasting several years with one particular clone. Destro et al. (1996) were able to trace contamination by L. monocytogenes in a shrimp-processing plant. Therefore, these studies evaluated the genomic diversity of the different clones of L. monocytogenes which are present in particular environments such as food industries.

The characterization of L. monocytogenes isolates can be achieved using phenotypic (serotyping and phage typing) and genotyping methods (DNA restriction analysis (REA) with pulsed field gel electrophoresis (PFGE), ribotyping, etc.). Serotyping is the first step in a typing scheme but is not sufficiently discriminatory. However, three serovars are associated with the majority of sporadic cases of listeriosis (4b, 1/2a and 1/2b) and serotype 4b is linked to almost all recent outbreaks of listeriosis (Farber and Peterkin 1991; Rocourt and Bille 1997). Therefore, serotyping and phage typing are useful in indicating a possible source for an outbreak in preliminary investigations. The REA-PFGE is highly discriminatory and reproducible (Rocourt and Bille 1997). Using different enzymes, this technique has been useful in proving that, in France, isolates collected in 1988 in a small outbreak were identical to isolates responsible for a major outbreak in 1975–77 (Buchrieser et al. 1992). It was also useful in appraising the polymorphism of isolates collected from cheeses and dairy products (Unnerstad et al. 1996; Margolles and de los Reyes-Gavilan 1998). This REA-PFGE method is often used in epidemiological studies in combination with serotyping (Buchrieser et al. 1992; Brosch et al. 1994; Jacquet et al. 1995; Destro et al. 1996; Boerlin et al. 1997).

In order to find the origin of L. monocytogenes on food products, we monitored two plants, a poultry meat plant (named A) for 1 year and a pork meat plant (named B) for 4 months for L. monocytogenes. We used serotyping and two genotypic methods (REA-PFGE with ApaI and SmaI) to characterize the isolates and to study the dissemination of L. monocytogenes types.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Detection and collection of Listeria monocytogenes isolates

Two types of sample were taken: swabs from the processing plants (232 in plant A and 116 in plant B) and meat products (raw and processed; 81 in plant A and 39 in plant B) (Table 1). In both plants, the finished products were uncooked and had to be well cooked before eating. The swabbing procedure was performed in three different workrooms of plants A and B: reception of raw materials, meat processing and product processing. The sampling surfaces were either in direct contact or without any contact with the raw meat or products and were assessed in two classes: environment (floors, walls and drain) and equipment (working tables, boxes, transport belts, machines, knives, cutter and churn). The swabbing was performed either during processing or after the cleaning operations. The meat products were divided into two classes: raw materials (pork and poultry raw meat, rind and water supply) and products (pork and poultry products at the end of the production and pork products at the end of the shelf-life at 4°C). All workrooms were studied at different times, for 1 year for plant A and 4 months for plant B.

Table 1.   Distribution of samples and Listeria monocytogenes isolates obtained from contaminated samples in plants A and B Thumbnail image of

Meat products (10 g) were analysed using the VIDAS (bioMérieux, Marcy l’Etoile, France) method for L. monocytogenes detection according to the manufacturer’s recommendations (Pradel et al. 1995).

Swabbing was performed using a tissue swab moistened with 5 ml tryptone salt (AES Laboratoires, Combourg, France) containing 100 g l–1 Isobio® (Laboratoire LCB, Lugny, France). They were then treated according to the modified VIDAS method of detection for L. monocytogenes (Chasseignaux et al. 1999), i.e. the 48-h Palcam plates obtained after incubation of samples in half Fraser broth were soaked with 5 ml tryptone salt and the resulting bacterial suspension used for the VIDAS assay.

For each positive sample, isolates were then collected. On each plate, typical Listeria colonies were subcultured onto tryptone soy agar supplemented with yeast extract (TSAYe; (Oxoid, Unipath, Dardilly, France). Colonies with a bluish tinge on TSAYe agar were subcultured onto tryptone soy broth (AES Laboratoires) and identified using a microtiter plate method by testing catalase, haemolysis, nitrate reductase and fermentation of xylose, mannitol and rhamnose. An average of six isolates per sample was collected (Table 1); nevertheless, the collecting of isolates depended on the level of L. monocytogenes contamination, at low levels only one isolate could be recovered and, at high levels, up to 21 isolates were collected.

Preparation of plugs for pulsed field gel electrophoresis

Bacteria were grown on TSAYe at 30°C for 24 h. Cells were recovered by flooding the plates with 2·5 ml TN buffer (0·01 mol l–1 Tris HCl, pH 7·6 and 1 mol l–1 NaCl) and centrifuging (10 min, 5000 g). Bacterial cells were washed twice with TN buffer and then resuspended in 1 ml TN buffer. Washed cells were adjusted to an O.D. of 4 (at 600 nm) in TN buffer and mixed with an equal volume of 1% agarose (Eurobio, Les Ulis, France) solution in TN buffer. The mixture was distributed into insert moulds (100 μl in each mould; Bio-Rad, Ivry sur Seine, France) and then cooled for 10 min at 4°C. The plugs were incubated for at least 5 h at 37°C in a solution of lysozyme (0·01 mol l–1 Tris HCl, pH 8, 0·001 mol l–1 EDTA, 1 mol l–1 NaCl, 0·5 g l–1 N-lauryol-sarcosine and 10 g l–1 lysozyme) and then transferred in a proteinase K solution (0·5 mol l–1 EDTA, pH 9, 10 g l–1 N-lauryol-sarcosine and 1 g l–1 proteinase K) and incubated at 50°C for at least 40 h. They were then washed twice with TE buffer (0·01 mol l–1 Tris HCl and 0·001 mol l–1 EDTA) for 30 min at room temperature. The proteinase K was then inactivated with a solution of 0·002 mol l–1 Pefabloc® (Boehringer Mannheim, Meylan, France) in TE buffer for 2 h at 37°C (Benjamin and Datta 1995). Finally, the plugs were washed twice with TE buffer for 30 min at room temperature. The plugs were divided into four slices and stored in TE buffer at 4°C.

DNA restriction enzyme analysis-pulsed field gel electrophoresis

One quarter of a plug was used for endonuclease digestions in separate reactions with 60 units ApaI (Roche Diagnostics, Meylon, France) or 40 units SmaI under the conditions recommended by the manufacturer, in a final volume of 100 μl and for at least 5 h of incubation at 25°C. The plugs were then cast in 1·2% agarose prepared with 0·5 × TBE buffer (0·045 mol l–1 Tris, 0·045 mol l–1 boric acid and 0·001 mol l–1 EDTA; Eurobio).

The resulting DNA fragments were separated by PFGE using a CHEF DR III system (Bio-Rad). The electrophoresis for ApaI was performed at 14°C, 7·6 V cm–1, a first pulsed time ramp from 15 to 35 s for 7 h and then a second from 2 to 20 s for 13 h and for SmaI at 14°C, 7·6 V cm–1, with one ramp from 2 to 20 s for 20 h.

The resulting DNA patterns were visualized after staining in ethidium bromide solution (0·5 mg l–1) under short wave u.v. light. The images were captured by a video system (gel DOC 1000 system; Bio-Rad). The estimation of the size of fragments and the comparison of patterns belonging to different isolates were performed using Molecular Analyst Software Fingerprinting (Bio-Rad). Similarities between the profiles, based on band positions, were derived by the Dice correlation coefficient with a maximum position tolerance of 1·2% (Struelens 1996). Dendrograms were constructed to reflect the similarities between isolates in the matrix. Isolates were clustered by the Unweighted Pair Group Method using Arithmetic average (Struelens 1996).

Serotyping

Serotyping was performed using sera (Eurobio) according to the method described by Seeliger and Höhner (1979). For the O antigen, isolates were grown on brain heart infusion agar (BHI; 10 g l–1 agar) for 24 h at 37°C. A positive reaction consisted of an agglutination of the isolate after at least 1 min of mixing in one drop of serum. The polyvalent sera (OI/OII and OV/OVI) were tested first. In the case of a positive reaction with OI/OII, OI and OIV sera were tested. In the case of a positive reaction with OV/OVI, OVI, OVII, OVIII and OVIX serums were tested.

For the H antigen, isolates were grown in BHI half-agar (4·5 g l–1 agar) for 48 h at room temperature by sticking into the agar. Isolates at the far end of the growth were then grown on BHI broth for 24 h at room temperature. A volume (1·2 ml) of the culture was then mixed with 1·2 ml 9 g l–1 NaCl. In separate tubes, two drops of sera (A, AB, C, D and nothing) were mixed with 0·5 ml mixed culture. The five tubes were incubated at 50°C for 2–3 h. A positive result showed an agglutination at the bottom of the tube. The serotyping was then determined according to the table of Larpent (1995).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Status of Listeria monocytogenes contamination in both plants

Among the 468 samples, 65 of 313 in plant A and 33 of 155 in plant B were L. monocytogenes positive (Table 1). The environment and the equipment were not very contaminated (18·3 and 16·4% in plants A and B, respectively). Nevertheless, the finished products were more frequently contaminated (40 and 35·7% for plants A and B, respectively). There was an increase in contamination in plant A with regard to the level of contamination of the raw materials (19·5%) while the contamination of raw materials (36%) in plant B was similar to that determined in the finished products.

Analysis of the patterns obtained by ApaI-restriction analysis-pulsed field gel electrophoresis

Five hundred and two isolates were genotyped by ApaI-REA-PFGE. The results showed 37 different ApaI genotypes (named A1–A37) but three isolates were resistant to this restriction enzyme. Figure 1 represents the genetic relationships and the distribution of the 37 profiles obtained. Four major ApaI genotypes, A12, A20, A5 and A11, were predominant with 33·3, 11, 8 and 8% of the genotyped isolates, respectively.

image

Figure 1.  Schematic representation, genetic relationship and distribution of the 37 ApaI genotypes of the Listeria monocytogenes isolates. The matrix was subjected to cluster analysis by the unweighted pair group average linkage analysis clustering method using Dice’s index (maximum tolerance 1·2%). The serotypes of the isolates are also indicated. M, Lambda ladder marker in kb

Download figure to PowerPoint

A large diversity of the ApaI genotypes is obvious. However, some profiles showed great similarity (at least 0·85) and, therefore, nine clusters (named AA–AI) could be defined (Fig. 1). The main cluster AH (55·3% of the collection) was composed of seven ApaI genotypes (A11–A17) and included the major ApaI genotype A12 (33·3% of the collection); these profiles were quite similar and the main differences were due to the absence or presence of bands of 70–140 kb in length. Cluster AE contained three ApaI genotypes (A19, A20 and A28) which were very similar; it represented 11·4% of the genotyped isolates. Cluster AB also gathered two similar ApaI genotypes (A5 and A6) representing 8·4% of the isolates. Cluster AD contained three ApaI genotypes (A8, A18 and A26); the main differences were due to bands of 50–97 kb in length. Clusters AA, AC, AG and AI represented less than 5% of the collection.

Analysis of the patterns obtained by SmaI-restriction analysis-pulsed field gel electrophoresis

Following typing by ApaI-REA-PFGE, isolates were selected for the second genotyping method. At least one isolate per sample and per ApaI genotype was chosen for analysis by SmaI-REA-PFGE, thus, if isolates recovered from the same sample showed the same pattern by ApaI-REA-PFGE, only one was randomly selected. Therefore, 145 isolates were analysed, showing 35 different SmaI genotypes (named S1–S35). Figure 2 represents the genetic relationships and distribution of the 35 profiles obtained. Three major different SmaI genotypes were observed, S9, S4 and S14 with, respectively, 35·9, 8·3 and 5·5% of the isolates.

image

Figure 2.  Schematic representation, genetic relationship and distribution of the 35 SmaI genotypes of the Listeria monocytogenes isolates. The matrix was subjected to cluster analysis by the unweighted pair group average linkage analysis clustering method using Dice’s index (maximum tolerance 1·2%). The serotypes of the isolates are also indicated. M, Lambda ladder marker in kb

Download figure to PowerPoint

As for the ApaI genotypes, a great diversity was present and, as some SmaI genotypes showed great similarity (at least 0·85), eight clusters (named SA–SH) were defined (Fig. 2). The main cluster SF (42·6% of the genotyped isolates) was composed of five SmaI genotypes (S9–S13) with one major genotype S9 (35·9% of the isolates); the major differences were due to the absence or presence of three bands (120, 145·5 and 162 kb). Cluster SB contained only two closely related SmaI genotypes (S4 and S5) representing 9% of the isolates. Cluster SE (7·6% of the isolates) grouped three SmaI genotypes (S14, S25 and S26) with differences in lower DNA fragments (length below 100 kb). The other clusters represented less than 5% of the genotyped isolates.

Combination of the two genotypic methods

In order to discriminate the L. monocytogenes isolates more precisely, the profiles obtained using the two genotyping methods were associated. The combination of both typing methods resulted in 50 different combined genotypes (named T1–T50) (Table 2). One ApaI profile can result in two or three different SmaI profiles, e.g. A13 is divided into S9 and S10 to give the combined genotypes T17 and T18 (Table 2). Correspondingly, one SmaI profile can result in different ApaI profiles, e.g. S9 is divided into A11, A12, A13 and A14 to give the combined genotypes T14, T16, T17 and T19 (Table 2).

Table 2.   Characteristics of the different combined genotypes (ApaI and SmaI) and associated serotypes Thumbnail image of
Table 3. Table 2 (Continued)Thumbnail image of

Comparing the clusters obtained for ApaI and SmaI, there were three notable occurrences. The first concerned isolates that were grouped in one cluster by ApaI as well as by SmaI: cluster AF matched with cluster SH and cluster AH with cluster SF. In the second, isolates that belonged to the same genotype by one method were divided in different genotypes by the other typing method: SmaI genotype S15 was distributed into cluster AC and ApaI genotype A32 into cluster SA. The last concerned isolates that were gathered in one cluster by one method but corresponded to different clusters and genotypes by the other method. Two examples can illustrate this occurrence. The first concerned cluster AC which is distributed by SmaI genotyping into clusters SD and SG. The second corresponded to cluster SC which included cluster AA and ApaI genotype A7.

Serotyping

Isolates were selected for serotyping as for typing by SmaI-REA-PFGE (at least one isolate per sample and per combined genotype; 133 isolates were serotyped). One major serotype was observed, 64% of the isolates being from 1/2a serotype. Serotypes 1/2b, 4a, 1/2c, 3b, 3a and 4ab represented, respectively, 14, 7·5, 4, 2, 1·5 and 1% of the chosen isolates. However, 6% of the isolates were not serotypable either because they did not react with the O or H factor antisera or with either of them.

Serotypes and profiles were then linked (Table 2). The analysis of the distribution of ApaI genotypes and serotypes resulted in three genomic divisions (Fig. 1). The first division contained clusters AA and AB and ApaI genotypes A4 and A7. These genotypes were either of serotype 1/2b or 3b. The second division gathered together clusters AC–AH and ApaI genotypes A19, A35 and A37, all of serotypes 1/2a, 3a, 1/2c or untypable. The last division grouped cluster AI and ApaI genotypes A29, A31, A32 and A34. These genotypes were either of serotype 4a, 4ab or untypable. The combination of SmaI genotypes and serotypes also resulted in three genomic divisions (Fig. 2). The first division gathered together cluster SA and SmaI genotypes S19, S20, S23, S24, S29, S32 and S35. These genotypes were either of serotype 4a, 4ab or untypable. The second division included clusters SB and SC and the SmaI genotype S18 and showed either serotype 1/2b or 3b. The last division grouped clusters SD–SH and SmaI genotypes S3, S15, S22, S30, S31 and S33. They were all of serotypes 1/2a, 3a, 1/2c or untypable.

When each combined genotype was considered, it was also observed that, generally, one combined profile corresponded to one serotype (Table 2). However, some profiles could have different serotypes. For example, the combined genotype T14 presented two serotypes, 1/2a and 3a. For a few profiles, isolates could either have a serotype or did not possess one. This was the case for the combined genotypes T28, T37 and T41. In the same way, three combined genotypes (T30, T34 and T44) were represented by only one isolate which did not show any serotype.

Analysis of the distribution of the combined types in the two different plants

Forty-one combined genotypes were observed in plant A, 37 during processing operations and eight after the cleaning and disinfecting procedures. Four of these latter genotypes (T40, T44, T49 and T50) were only detected after the cleaning and disinfecting procedures. Fourteen combined genotypes were found in plant B during processing, three of which were also recovered after the cleaning procedures. In contrast to plant A, no genotype was found only after the cleaning operations. Five combined genotypes (T14, T16, T17, T18 and T28) were isolated in both plants A and B during plant operation. Two of these were also isolated after the cleaning procedures in plant B. The categories of the different combined genotypes were defined according to their distribution in plants A and B (Table 2). The first category corresponded to the five combined genotypes present in both plants A and B, which were found in at least one workroom and on at least one sample. Categories II and III represented combined genotypes present only in plant B, largely spread in different workrooms (category II, genotypes T19 and T23) or only detected in one workroom and on one sample (category III, genotypes T3, T4, T21, T24, T25, T38 and T45). Categories IV, V and VI concerned those found in plant A in different workrooms, in one workroom but on many samples and restricted to only one sample, respectively.

Table 3 presents the distribution of the combined genotypes observed in the different workrooms for plants A and B. The combined genotypes of category I were generally spread over all the workrooms (environment of reception, equipment and raw meat in the processing area and transformed meat and product at the end of the production), except T18, which was not found in meat- and product-processing areas in plant A, and T28, which was detected in the meat-processing areas in both plants and in the product-processing area in plant A.

Table 3.   Distribution of the combined genotypes in the different workrooms of plants A and B in relation to the different samples Thumbnail image of

Contrasting situations appeared in plant A when the distribution of the 41 combined genotypes was analysed. A great diversity was observed in the environment of the reception and product-processing areas and also on the equipment of the meat-processing area. Moreover, the diversity diminished along the processing operation. From the three combined genotypes of category IV (genotypes present in different workrooms), genotype T5 was present in the three workrooms and on different samples, the environment in the reception area, the equipment in the meat-processing area and the environment and equipment in the product-processing area. For genotype T31 belonging to category V (genotypes restricted to one workroom), the contamination was limited to the product-processing area (environment, equipment and transformed meat). Some combined genotypes (T40, T44, T49 and T50) were found only after the cleaning and disinfecting procedures, in the reception area. Other genotypes (T22, T27, T34 and T37) were recovered both during processing and after the cleaning operations in the same workroom.

In plant B, diversity was observed on the equipment of the meat-processing area especially during plant operation. The two genotypes T14 and T16 were also detected in the different meat products and were largely spread over all the plant (environment of the reception area and equipment of the meat- and product-processing areas). Genotype T16 was also isolated, after the cleaning operations, on the equipment in the product-processing workroom. Three combined genotypes (T19, T23 and T24) were present in at least two workrooms. The two combined genotypes T19 and T23, belonging to category II, were detected in the meat-processing area during plant operation and also in the product-processing workroom after the cleaning procedures. Genotype T24 was first found in the environment of the reception area and also detected on the equipment of the meat-processing area. The other genotypes (category III) were generally found on the equipment of the meat-processing area.

The majority (74%) of samples in both plants A and B only presented one combined genotype. Nevertheless, 17·6% of the samples showed two genotypes and 6% contained three genotypes. However, some samples contained up to four different combined genotypes. Two samples presented four different genotypes: a floor in the reception workroom of plant A consisted of genotypes T5, T9, T42 and T48 and a working table in the meat-processing area of plant B consisted of genotypes T14, T16, T23 and T25.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Two different plants (A and B) were followed: one poultry-processing plant for 1 year and one pork-processing plant for 4 months. In both cases, plants were divided into different workrooms, studied separately and visited at least twice (during plant operation and after the cleaning and disinfecting operations). The 502 isolates of L. monocytogenes were characterized with the enzymes ApaI and SmaI into 50 combined genotypes. We have, therefore, observed great diversity in our collection. We found a higher number of PFGE profiles than in previous reports: 37 ApaI and 35 SmaI profiles instead of the 15 ApaI and 13 SmaI genotypes on 115 isolates tested by Destro et al. (1996), 17 ApaI genotypes on 42 isolates (Brosch et al. 1991) and 17 ApaI genotypes on 35 isolates (Buchrieser et al. 1991). The last two studies also combined the profiles obtained with three enzymes, resulting in only 24 different types whereas our combination of only two enzymes resulted in 50 different types. However, Buchrieser et al. (1991) obtained a greater diversity with fewer isolates than in the present study with regard to the ratio of genotypes vs studied isolates. Only one study (Brosch et al. 1994) showed a greater diversity, with 87 types for 176 tested isolates, but these isolates came from various collections and sources (human, food and animal) whereas our isolates were only collected from two plants in the raw meat industry.

These results also showed that three isolates were not cut by ApaI, despite several attempts, while they were digested by SmaI. Daniellson-Tham et al. (1993) and Brosch et al. (1994) observed the same phenomenon with the same enzyme. Considering the genotype obtained by SmaI for these three isolates, two different observations can be made: (i) for genotype T48, the SmaI genotype was already observed in two different combined types (T5 and T7) and (ii) for combined types T49 and T50, however, the SmaI genotypes (S32 and S33) were only observed for these two isolates. This shows that a resistance to ApaI is rare and probably due to a system of modification present in these isolates.

The analysis of the genetic relationship of either ApaI or SmaI genotypes resulted in great diversity. However, some types showed great similarity and a genotype which was defined with one enzyme can match a cluster defined by the other enzyme. In contrast, one ApaI profile can result in different SmaI genotypes and, conversely, one SmaI profile can result in several ApaI genotypes. This could be due to a mutation only noticed by one of the two enzymes and the relationship between these isolates is, therefore, very important. These facts were observed in the two plants. It is interesting to note that genotypes T1 and T2 were found in plant A whereas the close genotype T3 was detected in plant B. These three genotypes possessed the same SmaI genotype (S7) and three related ApaI genotypes (A1, A2 and A3) with similarities greater than 85%. Genotypes T9, T10 and T11 also illustrated these observations; they corresponded to the combination of the ApaI genotype A8 and the SmaI genotypes S1, S2 and S16. If genotypes S1 and S2 were closely related, genotype S16 only showed 65% of similarity with S1 and S2. Therefore, genotypes T9 and T10 were closely related. Moreover, they were not found in the same plant (plant A, genotype T9 and plant B, genotype T10). In both cases, these observations supported the hypothesis of clonal origins for genotypes T1 and T2 and T9 and T10, respectively. Therefore, the use of two restriction enzymes showed greater discrimination than the use of a single enzyme.

The other observation regarding the genetic relationship was that isolates could be grouped in the same cluster by both methods. For example, cluster AH was the major cluster by the ApaI method and corresponded to the major cluster SF by SmaI. In both clusters, we observed one major genotype: (i) A12 representing 33·3% of the genotyped isolates in cluster AH and (ii) S9 representing 35·9% of the genotyped isolates in cluster SF. In cluster AH, we also observed two secondary genotypes, A11 and A13, representing, respectively, 8 and 5·8% of the genotyped isolates. The other genotypes were less important. Moreover, genotypes A11 and A13 were highly similar to genotype A12 (respectively 97 and 92%). In cluster SF, except for genotype S9, all the genotypes were minor. All these isolates were closely related. These results suggested that the secondary and minor genotypes derived from the major genotype by different mutations and may reveal a major clone of L. monocytogenes which could be characterized by the ApaI profile AH and the SmaI profile SF.

The serotyping results showed that different types were present: 1/2a, 1/2b, 1/2c, 3a, 3b, 4a and 4ab, the major serotype being 1/2a. Different studies have shown that only three serotypes (4b, 1/2a and 1/2b) were associated with the majority of sporadic cases of listeriosis but were also present in outbreaks. However, serotype 4b is predominant in most European outbreaks whereas serotypes 4b, 1/2a and 1/2b are equally responsible for epidemic cases in Canada and the USA (Farber and Peterkin 1991; Rocourt and Bille 1997). No serotype 4b isolates were found in this study. Indeed, L. monocytogenes serotype 4b have caused almost all recent outbreaks of foodborne listeriosis. Accordingly, these results indicate that the presence and predominance of both serotypes 1/2a and 1/2b in these pork and poultry industries might be a source for sporadic cases in France.

Moreover, Buchrieser et al. (1992) showed that there was no direct correlation between the restriction profiles using three different enzymes and the different serotypes. Our results confirmed this observation that six different combined profiles (T1, T5, T14, T28, T37 and T41) each displayed more than one serotype. Nevertheless, the majority of combined types presented only one serotype. Furthermore, the profiles which were obtained after cleavage of genomic DNA by ApaI as well as by SmaI resulted in different genomic divisions that were more discriminating: the first included serotypes 4a, 4ab and unserotypable isolates; the second included serotypes 1/2a, 1/2c, 3a and unserotypable isolates and the last included serotypes 1/2b and 3b. Our results confirm those of Brosch et al. (1994) who observed two major genomic divisions correlated with the serotype. The first genomic division included serotypes 1/2a, 3a, 1/2c and 3c and the second included serotypes 1/2b, 3b, 4b, 4d and 4e. These two genomic divisions were therefore linked with the flagellar antigen type.

Following this analysis of genotypes, the L. monocytogenes isolates obtained can be divided into four different classes. The first category (I) is composed of isolates collected during processing but also after the cleaning operations (e.g. genotype T16) after an interval of 9 months for plant A and 3 months for plant B. These isolates probably show an adaptation to survival in a food industry plant. This has previously been observed by Giovannacci et al. (1999) in pork slaughtering and cutting plants, Boerlin et al. (1997) in fish products and Unnerstad et al. (1996) in a dairy plant. The study of Giovannacci et al. (1999) pointed out the persistence of L. monocytogenes strains over a 1-year period in the environments of two pork plants. Boerlin et al. (1997) noted survival for several months whereas Unnerstad et al. (1996) showed survival over 7 years. Two hypotheses can be expressed: either the cleaning operations are not sufficiently efficient to eliminate these clones or these clones are adapted to animals and continuously enter the plants and find appropriate conditions to survive and eventually grow in the plants.

The second category (II and IV) is composed of isolates found only during plant processing operations in different workrooms, at different times. These isolates probably entered continuously into the plant but were eliminated by the cleaning and disinfecting procedures.

The third category (V) gathers together isolates found only during processing, in one workroom but on different samples. For example, genotype T31 was recovered in three sample types (environment, equipment and raw meat) during the product processing in plant A. These isolates were probably brought in by the raw meat and contaminated the equipment and environment of the workrooms.

The last category (III and VI) is composed of isolates found only once in one workroom. The majority of these isolates can be considered as transitory as they do not survive to the cleaning and disinfecting procedures. However, four genotypes, T40, T44, T49 and T50, were only detected after the cleaning operations but never during processing. These isolates, which were not eliminated by the cleaning and disinfecting procedures, were still present during processing but not recovered, probably because of their low levels.

Some profiles were also common to both plants, which are from two different geographical locations (combined profiles gathered in category I, Table 2). This was previously observed by Buchrieser et al. (1991) and Boerlin et al. (1997). We can, therefore, postulate that some clones are widely spread. Moreover, when these genotypes were compared with the genotypes obtained by Brosch et al. (1994), isolate 1570 (CDC/G-4484) was very similar to our major genotype A12. This isolate came from the collection of the Center for Disease Control and Prevention (CDC, Atlanta, GA, USA) and had been isolated from food. Its serotype was 1/2a as with the majority of our isolates showing ApaI genotype A12. When our ApaI genotypes were then compared with the genotypes obtained by Giovannacci et al. (1999), their major cluster was quite similar to cluster AH. Furthermore, plants A and B processed different kinds of meat, poultry meat in plant A and pork meat in plant B. This strengthened the idea that some genotypes might be widely spread in the food industry.

The results of the present study allow us to trace the contamination in the two plants, in a similar way to Destro et al. (1996). Some isolates were found during the plant processing operations and after the cleaning and disinfecting procedures. For plant A, contamination was present in 12 different points of the reception area, only in the environment and never on the equipment. A deep cleaning of this area might be necessary. However, two genotypes (T22 and T34) found after cleaning and disinfecting procedures and during plant operation were limited to the reception area. Four genotypes (T40, T44, T49 and T50) were found only after cleaning and disinfecting procedures and were not recovered during plant operation. These genotypes might not be predominant and might be masked by the major genotypes collected.

For plant B, the contamination of the entire plant with genotypes T14 and T16 was obvious; these genotypes were found in different places and on different processed meat products. The contamination with three different combined types (T16, T19 and T23) in the product-processing area after the cleaning and disinfecting procedures highlights the problem of the efficiency of these procedures.

In both plants, some samples were contaminated with up to four different types. However, a majority of samples (74%) contained only one genotype. These types were either closely related or very different. This has already been observed by Margolles and de los Reyes-Gavilan (1998) who found four samples contaminated with two different L. monocytogenes isolates (about 85% of similarity). It is, therefore, of interest to have more than one isolate per sample, when possible, to give better understanding of the spreading.

In conclusion, the molecular genotyping methods (i.e. digestions using a restriction enzyme followed by PFGE) helped us to trace contamination by L. monocytogenes in both plants. Some clones were recovered in plants for several months while others were eliminated during the cleaning and disinfecting procedures. Finally, some clones were found only during processing operations but in different workrooms and seemed to enter continuously into the plant. Different genotypes can be associated with poultry as well as pork meat. Therefore, these clones might be associated with meat production. Further studies are required to prove whether these clones are found in other meat plants and meat products.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

The authors acknowledge Europe (FEDER) for financial support of the molecular assays and the Ultra-propre Nutrition Industrie Recherche group for financial support of the investigations in plants, the French Ministère de l’Education Nationale, de la Recherche et de la Technologie and the Agence Nationale pour la Recherche et la Technologie (ANRT) for its grant. The authors also thank Y. Le Nôtre-Michel and S. Gorin for their technical help during the study.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  • 1
    Benjamin, M.M. & Datta, A.R. (1995) Modified pulsed field gel electrophoresis technique using Pefabloc(R)sc for analysing Listeria monocytogenes DNA. Biochemica 2 , 3031.
  • 2
    Blackman, I. & Frank, J. (1996) Growth of Listeria monocytogenes as a biofilm on various food-processing surfaces. Journal of Food Protection 59 , 827831.
  • 3
    Boerlin, P., Boerlin Petzold, F., Bannerman, E.J.B., Jemmi, T. (1997) Typing Listeria monocytogenes isolates from fish products and human listeriosis cases. Applied and Environmental Microbiology 63 , 13381343.
  • 4
    Brosch, R., Buchrieser, C., Rocourt, J. (1991) Subtyping of Listeria monocytogenes serovar 4b by use of low-frequency-cleavage restriction endonucleases and pulsed-field gel electrophoresis. Research Microbiology 142 , 667675.
  • 5
    Brosch, R., Chen, J., Luchansky, J. (1994) Pulsed-field fingerprinting of Listeriae: identification of genomic divisions for Listeria monocytogenes and their correlation with serovar. Applied Environmental Microbiology 60 , 25842592.
  • 6
    Buchrieser, C., Brosch, R., Catimel, B., Rocourt, J. (1992) Pulsed-field gel electrophoresis applied for comparing Listeria monocytogenes strains involved in outbreaks. Canadian Journal of Microbiology 39 , 395401.
  • 7
    Buchrieser, C., Brosch, R., Rocourt, J. (1991) Use of pulsed field electrophoresis to compare large DNA-restriction fragments of Listeria monocytogenes strains belonging to serogroups 1/2 and 3. International Journal of Food Microbiology 14 , 297304.
  • 8
    Chasseignaux, E., Michel, Y., Toquin, M.T., Ermel, G., Salvat, G., Colin, P. (1999) Comparison of two rapid methods for the detection of Listeria monocytogenes with a standard procedure in naturally contaminated environments of raw poultry processing. Journal of Rapid Methods in Automated Microbiology 7 , 147153.
  • 9
    Cox, L.J., Kleiss, T., Cordier, J.L., Cordella, C., Konkel, P., Pedrazzini, C., Beumer, R., Siebenga, A. (1989) Listeria spp. in food processing, non-food and domestic environments. Food Microbiology 6 , 4961.
  • 10
    Daniellsson-Tham, M.L., Bille, J., Brosch, R., Buchrieser, C., Persson, K., Rocourt, J., Schwarskopf, A., Tham, W., Ursing, J. (1993) Characterization of Listeria monocytogenes strains isolated from soft cheese. International Journal of Food Microbiology 18 , 161166.
  • 11
    Destro, M., Leitao, M., Farber, J. (1996) Use of molecular typing methods to trace the dissemination of Listeria monocytogenes in a shrimp processing plant. Applied Environmental Microbiology 62 , 705711.
  • 12
    Farber, J. & Peterkin, P. (1991) Listeria monocytogenes, a foodborne pathogen. Microbiological Review 55 , 476511.
  • 13
    Giovannacci, I., Ragimbeau, C., Queguiner, S., Salvat, G., Vendeuvre, J.L., Carlier, V., Ermel, G. (1999) Listeria monocytogenes in pork slaughtering and cutting plants: use of RAPD, PFGE and PCR-REA for tracing an molecular epidemiology. International Journal of Food Microbiology 53 , 127140.DOI: 10.1016/s0168-1605(99)00141-5
  • 14
    Graham, C.G.M. & Colins, J.K. (1991) Listeriosis: biology and implications for the food industry. Trends in Food Science Technology 1991 , 8993.
  • 15
    Helke, D.M. & Wong, A.C.L. (1994) Survival and growth of Listeria monocytogenes and Salmonella typhimurium on stainless steel and buna-N rubber. Journal of Food Protection 57 , 963968.
  • 16
    Jacquet, C., Catimel, B., Goulet, V., Lepoutre, V., Veit, P., Dehaumont, P., Rocourt, J. (1995) Typing of Listeria monocytogenes during epidemiological investigations of the French listeriosis outbreaks in 1992, 1993 and 1995. XII International Symposium on Problems of Listeriosis, Perth, Western Australia, 2–6 October 1995. Canning Bridge, Western Australia: Promaco Conventions Pty Ltd.
  • 17
    Larpent, J.P. (1995) Les Listeria. Paris:Lavoisier.
  • 18
    Margolles, A. & De Los Reyes-Gavilan, C. (1998) Characterization of plasmids from Listeria monocytogenes and Listeria innocua strains isolated from short-ripened cheeses. International Journal of Food Microbiology 39 , 231236.DOI: 10.1016/s0168-1605(97)00132-3
  • 19
    Pradel, J.M., Battail, N., Atrache, V. (1995) Evaluation of an automated immunoassay for the specific detection of L. monocytogenes. Food Micro′95, Melbourne, 8–10 February 1995. Waterloo, Australia: Australian Institute of Food Science and Technology Inc.
  • 20
    Rocourt, J. & Bille, J. (1997) Foodborne listeriosis. World Health Statistics Quarterly 50 , 6773.
  • 21
    Sammarco, M.L., Ripabelli, G., Ruberto, A., Iannito, G., Grasso, G.M. (1997) Prevalence of Salmonella, Listeriae and Yersiniae in the slaughterhouse environments. Food Microbiology 6 , 4961.
  • 22
    Seeliger, H.P.R. & Höhner, K. (1979) Serotyping of Listeria monocytogenes and related species. In Methods in Microbiology ed. Bergan, T. and Norris, J.R. pp. 31–49. New York: Academic Press.
  • 23
    Struelens, M.J. (1996) Consensus guidelines for appropriate use and evaluation of microbial epidemiologic typing systems. Clinical Microbiological Infection 2 , 211.
  • 24
    Unnerstad, H., Bannerman, E., Bille, J., Danielsson Tham, M., Waak, E., Tham, W. (1996) Prolonged contamination of a dairy with Listeria monocytogenes. Netherlands Milk and Dairy Journal 50 , 493499.
  • 25
    Wendtland, A. & Bergann, T. (1994) Listeria monocytogenes: occurence in a factory for slaughtering, carvening and meat processing. Fleischwirtschaft 74 , 13291331.
  • 26
    Wong, A.C.L. (1998) Biofilms in food processing environments. Journal of Dairy Science 81 , 27652770.