Abstract
Products from three broiler abattoirs were sampled for Listeria species to evaluate the changes in the prevalence and contamination rates at two stages of processing. Sampling was performed at the evisceration stage and at the end of processing after packaging and refrigerating at 4°C for 24 h. A total of 212 samples were collected; 52 were from abattoir A, and 80 samples each were collected from abattoirs B and C. Among all samples, 99 (46.7%) tested positive for Listeria, including L. monocytogenes 19 (8.9%), L. innocua 69 (32.5%), L. grayi 10 (4.7%), and L. welshimeri 1 (0.5%). The L. monocytogenes contamination rate varied from 5% to 11.5% in the 3 abattoirs. L. innocua was the most common species identified and was found in 8.8% of the samples from abattoir A and 33.7% of the samples from both abattoirs B and C. Twenty-six of the L. monocytogenes isolates obtained from positive samples were subjected to serotyping by multiplex polymerase chain reaction and characterization by the pulsed-field gel electrophoresis (PFGE) method using two cutting enzymes, ApaI and AscI. Three molecular serogroups were identified: IIa, IIb, and IVb. Serogroup IIa was common to all abattoirs, and serogroups IIb and IVb were found only in abattoir C. The 10 different obtained PFGE profiles were grouped into 7 clusters; some of these clusters were common to the 3 abattoirs, and others were specific to the abattoirs in which they were identified. This study revealed a high prevalence of Listeria spp., particularly L. monocytogenes, in raw broilers. This high incidence presents a risk to consumers due to the potential occurrence of cross-contamination with other foods in domestic refrigerators and the ability of these microorganisms to survive in undercooked products.
Introduction
L
The involvement of L. monocytogenes as an infectious agent in both sporadic and outbreak listeriosis cases has been linked to the consumption of different food products with vegetal and animal origins, such as milk (Pini and Gilbert, 1988), meat products, vegetables, and fruits (Farber and Peterkin, 1991). While poultry products have not been directly implicated as sources of outbreaks, ready to-eat and undercooked chicken have been linked to sporadic cases of human listeriosis (Kerr et al., 1988; Ryser, 2007).
The prevalence of L. monocytogenes in broilers at abattoirs in many countries is well documented (Chasseignaux et al., 2001; Miettinen et al., 2001; Berrang et al., 2002; Vitas et al., 2004). The incidence of L. monocytogenes contamination is relatively high (Ryser, 2007). However, in Algeria, the presence of this pathogen in poultry products has not been investigated.
Characterizations of L. monocytogenes isolates can be based on serotyping, but this technique is not sufficiently discriminatory. Three serovars (4b, 1/2a, and 1/2b) are associated with the vast majority of listeriosis outbreaks and sporadic cases; however, most outbreak strains belong to serovar 4b (Chen et al., 2007).
Multiplex polymerase chain reaction (PCR) assays have been proposed by Doumith et al. (2004) as an alternative method for clustering the majority of serotypes into five groups (serogroups). Pulsed-field gel electrophoresis (PFGE) is the “gold standard” technique for molecular subtyping of L. monocytogenes due to its high discriminatory power (Graves and Swaminathan, 2001).
Raw broilers are cooked before consumption, but there is a risk of cross-contamination with other food, especially ready-to-eat items, in consumer kitchens (Almeida et al., 2005a). L. monocytogenes is able to grow and form biofilms on various food-processing surfaces, which permits its survival (Blackman and Frank, 1996). The objectives of this study were to estimate the prevalence of L. monocytogenes in poultry from three broiler abattoirs, to evaluate the changes in the contamination rate across two stages of processing, and to molecularly characterize the L. monocytogenes isolates with multiplex-PCR and PFGE techniques.
Materials and Methods
Sample collection and L. monocytogenes isolation and enumeration
Sampling was performed at three industrialized broiler abattoirs (A, B, and C) located in the Algiers region. The processing line speeds of both abattoirs A and B were 2000 birds/h, and 3500 birds/h for abattoir C.
The sites were sampled 1 time at 4 weeks from March to June in 2012, which resulted in 636 carcasses. Groups of 3 carcasses were pooled together, and a total of 212 samples were obtained and screened for Listeria; 52 were collected from abattoir A, and 80 samples each were collected from abattoirs B and C.
Sampling was performed at the processing lines, and carcasses were taken from two points on the processing lines. These points were after evisceration and at the end of the processing line after packaging and refrigeration at 4°C for 24 h. From each abattoir, 50% of carcasses were taken from the first point of sampling; the other 50% were taken from the second point of sampling.
The samples were packed in sterile bags and transported to the laboratory in ice containers. Twenty-five-gram samples of neck skin from each sample were used for Listeria screening according to the ISO 11290-1 method. L. monocytogenes enumeration was performed according to the ISO 11290-2 method.
Multiplex-PCR serogrouping
Multiplex serogrouping was performed on PCR template lysates obtained from bacterial colonies according to the method of Doumith et al. (2004). The five primers for the target fragments from the following genes: lmo 0737, ORF 2819, ORF 2110, lmo 1118, and prs (MWG-Biotech, Muenchenstein, Switzerland). PCR was performed in a thermocycler, and the PCR products were resolved on 2% agarose gels stained with ethidium bromide (Eurobio, France) and analyzed with a Bio-Rad Gel Doc 2000™ imaging system (Bio-Rad Laboratories, Segrate, Milan, Italy).
Molecular characterization by PFGE
L. monocytogenes isolates were characterized by DNA macrorestriction using PFGE according to the PulseNet standardized protocol (Graves and Swaminthan, 2001).
Bacterial cultures were embedded in graded agarose (Seakem Gold Agarose; Lonza, Rockland, ME), lysed, and washed. The DNA was digested in situ in separate reactions with the AscI (New England Biolabs, MA) and ApaI (Fermentas, Burlington, Canada) restriction enzymes.
The generated DNA fragment were resolved on 1% Seakem Gold Agarose, the gels were stained with ethidium bromide (Eurobio, France), and the image patterns were visualized and photographed with a Bio-Rad Gel Doc 2000™ imaging system.
The PFGE patterns were compared using the Bionumerics software package (Applied Maths, Belgium), and cluster analyses of the individual or combined pulsotypes were performed using the unweighted-pair group method using arithmetic averages and the Dice coefficient to analyze the similarities of the banding pulsotypes. The PFGE patterns were considered to be indistinguishable when their similarity exceeded 98% (Barrett et al., 2006).
The PFGE pulsotypes of the isolated strains were compared with the PFGE pulsotypes of the PFGE database of the Institut Pasteur, WHO Collaborating Centre, and French National Reference Centre for Listeria using the Bionumerics software package.
Results
Ninety-nine (46.7%) of the 212 tested samples were positive for Listeria spp. including 19 (8.9%) L. monocytogenes-positive samples. L. innocua was the most commonly isolated species from the 69 (32.5%) samples.
L. grayi was isolated in 10 (4.7%) samples, while L. welshimeri was found in only 1 (0.5%) sample. The enumeration revealed that, in all positive samples, the contamination rates did not exceed 100 CFU/g.
The distribution of Listeria across the slaughter line revealed a high prevalence of Listeria spp. at the end of processing after packaging and refrigeration at the 3 abattoirs: 77% (20/26), 72.5% (29/40), and 57.5% (23/40) in units A, B, and C respectively. However, at the evisceration stage, the prevalences were 7.7% (2/26), 22.5% (9/40), and 40% (16/40) in units A, B, and C, respectively. L. monocytogenes was isolated only at the end of the slaughter process in both units A and B in 6 (23%) and 4 (10%) samples, respectively.
In unit C, L. monocytogenes was isolated in both the evisceration stage from 1 (2.5%) sample and at the end of processing from 8 (20%) samples. L innocua was isolated in both of the stages at each of the three abattoirs (Table 1).
Altogether, 26 L. monocytogenes isolates were subjected to further characterization and represented 1 to 2 isolates from each positive sample. The multiplex PCR method revealed that 22 (84.6%) of the 26 collected isolates belonged to serogroup IIa (1/2a, 3a), 3 (11.5%) to serogroup IIb (1/2b, 3b, or 3c), and 1 isolate (3.8%) to serogroup IVb (4b, 4d, or 4e). Serogroup IIa was the most common in all of the abattoirs (100%) in both abattoirs A and B and 69% in abattoir C. Serogroups IIb (23%) and IVb (7.7%) were found only in abattoir C.
PFGE was performed to determine the genetic relatedness among the 26 isolates. All of the isolates generated distinct electrophoretic profiles (Fig. 1).

Dendrograms resulted from the combination of profiles obtained by cutting enzymes ApaI and AscI, showing similarities among isolated Listeria monocytogenes from the broiler abattoirs. A total of seven (I–VII) clusters of pulsed-field gel electrophoresis (PFGE) profiles resulted from this combination. The polymerase chain reaction (PCR) serogroups were IIa, IIb, and IVb.
Typing based on the AscI enzyme revealed 7 different clusters; each of these clusters contained 1–10 distinct profiles with an overall similarity of 60%, whereas restriction with the ApaI enzyme revealed 7 different clusters that contained 1–10 distinct profiles with a similarity level of 58%.
The combination of the typing data from the 2 enzymes revealed 7 clusters (I–VII) with a similarity level of 58%. The largest group was cluster IV, which represented 38.4% of the isolates, and the least frequent clusters (I, II, III) contained 1 electrophoretic profile representing each of the serogroups and accounted for 3.8% of isolates.
The clusters I, III, IV, V, and VI contained 1, 1, 10, 8, and 2 isolates, respectively, that belonged to serogroup IIa, while the VII cluster contained 3 isolates that belonged to serogroup IIb. Cluster II contained only one isolate that belonged to serogroup IVb. The profiles belonging to the same cluster belonged to the same serogroup. Each abattoir exhibited a dominant profile; cluster V (75%) was dominant in unit A, clusters V and VI (33.3%) were dominant in unit B, and cluster IV (61.6%) was dominant in unit C. Some of the clusters were specific to these abattoirs: cluster I was specific to unit A; cluster II was specific to unit C; cluster III was specific to unit B. The other clusters were common to the three units (IV and V; Table 2).
The isolates collected from the same samples belonged to the same serogroups and had the same PFGE profiles (e.g., isolates number 42 and 54/serogroup IIb/cluster VII), while the other isolates each had 1 particular PFGE profile and belonged to different serogroups (e.g., isolates number 39 and 25/serogroups IIa and IVb/clusters II and IV).
Discussion
Listeria spp. and L. monocytogenes are typically detected in broiler carcasses during processing (Ojeniyi et al., 1996; Ryser, 2007). In our survey, the prevalences of Listeria spp. (46.7%) and L. monocytogenes (8.9%) were estimated on pooled samples derived from 3 carcasses. They were in agreement with those reported in previous studies by Whyte et al. (2004) and Almeida et al. (2005b): 6% and 9.7% respectively; however, these prevalences were fairly low compared to those reported by Genigeorgis et al. (1989) and Sjőman (2010): 36.7% and 62%, respectively.
L. monocytogenes contamination rates recorded in our study were all <100 colony-forming units (CFU)/g. This value is considered to be safe (FAO/WHO, 2004); however, in some cases, including the following, consumer food practices can make this level dangerous: (1) when poultry meat is stored for several days in domestic refrigerators at temperatures that range from 0°C to 4°C, the number of Listeria can increase given their psychrotrophic nature (Van Nierop et al., 2005) and thus reach and exceed the 100 CFU/g dose, recognized as unsafe by the European regulation (EC) No 2073/2005 (2005); (2) the meat can also contaminate other foods that are in the refrigerator, particularly the ready-to-eat foods via cross-contamination (Almeida et al., 2005a).
Evaluation of the changes in contamination prevalence was performed across two stages of processing in the three abattoirs: after evisceration, since it is a stage where Listeria contamination is high as observed by Whyte et al. (2004), and at end of processing to evaluate the impact of handling on the evolution of contamination. This assessment revealed that the prevalences of Listeria spp. and L. monocytogenes increased dramatically at end of processing after packaging and refrigeration compared to those observed at evisceration.
These results suggest that Listeria contamination of broiler carcasses at abattoirs occurs and increases during and after evisceration stage; this supposition agrees with data from Skovgaard and Morgen (1988), Genigeorgis et al. (1989), Berrang et al. (2000), Whyte et al. (2004), Almeida et al. (2005b), and Lopez et al. (2008).
Almeida et al. (2005b) reported L. monocytogenes contamination only at the end of processing, and we observed the same result in both units A and B. In contrast in unit C, L. monocytogenes was found at both the evisceration stage and at the end of processing, which agrees with the results reported by Lopez et al. (2008). We suggest that this contamination was caused by a highly contaminated environment at the evisceration stage in this unit.
The majority of our L. monocytogenes isolates belonged to serogroup IIa (1/2a, 2a), which contrasts with the results of a previous Algerian study (Belouni, 1990; Hamdi et al., 2007). In the latter study, the authors identified only serovar 4b isolates in food products that were tested in Algeria. However, the identifications of sergroups IIa, IIb, and IVb in broilers are in agreement with many studies (Chasseignaux et al., 2001; Gilberth et al., 2005; Chiarini et al., 2009; Nucera et al., 2010; Shi et al., 2010).
The predominant occurrence of serogroup IIa in the three abattoirs indicates its strong ecological fitness. The variety of serogroups found in unit C might indicate that the sources of contamination were numerous and variable (Chiarini et al., 2009).
Several studies have reported that >95% of human listeriosis cases are caused by the 4b, 1/2a, and 1/2b serotypes (Wiedmann et al., 2011). These important serotypes were dominant in the serogroups that we identified, and these finding suggest that when contamination conditions are combined, poultry might pose a health risk to consumers.
The study of the combined macrorestriction profiles generated by the ApaI and AscI enzymes provided a genetic fingerprint of all of the tested L. monocytogenes strains. Unfortunately, we were unable to compare our isolates to human clinical isolates to confirm whether these strains were implicated in human listeriosis cases in our country. There is no database containing information about genotypes of isolates from humans.
We compared the strains isolated in this study with those listed in the database of the Institut Pasteur WHO Collaborating Center and French National Reference Center for Listeria and found that the strains have molecular patterns that were observed in clinical databases.
The analysis of the genetic profiles revealed that each abattoir exhibited a dominant PFGE profile as previously reported by Miettinen et al. (2001) and Lopez et al. (2008).
Our study demonstrated that each abattoir possessed its own specific profile that might have been indicative of a persistent strain. Fugget et al. (2007) reported that all unit-specific strains were persistent strains.
Some strains obtained from the same samples exhibited different PFGE profiles and were of different serogroups. In a previous study, Nucera et al. (2010) noted that such differences can be higher than 80% and suggested that 3 colonies per sample are suitable for determining isolate variability within the same sample, particularly if the study aims to characterize L. monocytogenes populations in food.
The geographical distributions of PFGE profiles have revealed that some profiles are ubiquitous and found in many poultry farm locations that supply abattoirs, which corroborates the results of Buchrieser et al. (1991) and Chasseignaux et al. (2001). The ubiquity of genetic profiles might be attributable to the movement of poultry through trade channels.
The level of contamination of carcasses at abattoirs could be decreased by the introduction of more rigorous hygiene measures, by setting up cleaning and disinfection plans for the environment and for all contact surfaces, or through the implementation of a risk management system.
Conclusions
Results of the present study indicate a high prevalence of Listeria spp. (46.7%) and L. monocytogenes (8.9%) in broiler carcasses in abattoirs. The Listeria contamination of broiler carcasses at abattoirs likely occurs and increases during and after the evisceration stage, based on the high prevalences that were recorded at the end of processing after packaging and refrigeration. The molecular characterization provided a genetic fingerprint of all of the isolates of L. monocytogenes and identified groups of serotypes that had similar patterns seen in the majority of human listeriosis cases in the world. The serotype 4b cluster was associated with the soft cheese (Switzerland, 1983–1987; California, 1985) and pork tongue (France, 1992) outbreaks; serotype 1/2a was associated with the turkey deli meat (United States, 2000) outbreak (Chen and Knabel, 2007); and serotype 1/2b was associated with the multistate listeriosis cantaloupe (USA, 2011) outbreak (CDC, 2012).
The high prevalence of Listeria spp., particularly L. monocytogenes, in raw broilers is a problem that should be of concern to the authorities and food producers. Sanitation programs aimed at preventing or at least reducing Listeria contamination should be adopted.
Footnotes
Acknowledgments
We are grateful to the Laboratoire des Listeria, Centre National de Références des Listeria, WHO collaborating Center for Foodborne listeriosis, Institut Pasteur for allowing us to perform the serogrouping and PFGE typing of the strains examined in this work.
Disclosure Statement
No competing financial interests exist.
