Research Note: Persistent Salmonella problems in slaughterhouses related to clones linked to poultry companies a-Cuellar ,z S. Gonzalez-Bodi,*,x L. Lorenzo-Rebenaque ,* and S. Vega * C. Marin ,*,y,1 M. Cerd * Facultad de Veterinaria, Instituto de Ciencias Biomedicas, Universidad Cardenal Herrera-CEU, CEU Universities, Alfara del Patriarca, Spain; yIRTA, Programa de Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Campus de la Universitat Autonoma de Barcelona (UAB), Bellaterra, 08193, Catalonia, Spain; zUnitat mixta d’Investigaci o IRTA-UAB en Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Campus de la Universitat Autonoma de Barcelona (UAB), Bellaterra, 08193, Catalonia, Spain; and xCentro de Biotecnologia y Genomica de Plantas, Universidad Politecnica de Madrid (UPM) - Instituto Nacional de Investigaci on y Tecnología Agraria y Alimentaria (INIA), Madrid, Spain ABSTRACT Salmonellosis remains one of the main slaughterhouse and that 46.3% of the different samples foodborne zoonoses in Europe, with poultry products as from carcasses were contaminated with Salmonella. The the main source of human infections. The slaughter- two serovars isolated at the different steps in the slaugh- house has been identified as a potential source for Salmo- terhouse were Enteritidis (98.2%) and Kentucky (1.8%). nella contamination of poultry meat. Despite the Pulsed-field gel electrophoresis analysis revealed a low mandatory programme of the EU, there are companies genetic diversity, with all S. Enteritidis isolates showing with persistent Salmonella that are unable to remove a nearly identical pulsotype (similarity >85%) and S. the bacteria from their processing environment, Kentucky strains showed the same XbaI PFGE profile compromising the entire production line. In this context, (95.0% genetic similarity). The results of this study an intensive sampling study was conducted to investi- showed a high genetic relationship among isolates recov- gate a slaughterhouse with persistent Salmonella prob- ered from carcasses and environmental samples in the lems, establishing the genetic relationship among slaughterhouse from both Salmonella-positive and Sal- Salmonella strains isolated during the slaughter process. monella-free flocks. Salmonella strains re-circulated A total of 36 broiler flocks were sampled during process- across to poultry flocks and re-entered the slaughter- ing at the slaughterhouse. Salmonella was identified house to survive on the processing line. Thus, it is neces- based on ISO 6579-1:2017 (Annex D), serotyped by sary to implement molecular diagnosis methods in time Kauffman-White-Le-Minor technique, and the genetic at the field level to determine the Salmonella epidemiol- relationship was assessed with ERIC-PCR followed by ogy of the flock, to make rapid decisions for the control PFGE. The outcomes showed that 69.4% of the batches of Salmonella and prevent entry into the slaughterhouse sampled carried Salmonella upon arrival at the environment. Key words: Salmonella, slaughterhouse, poultry, PFGE, zoonosis 2022 Poultry Science 101:101968 https://doi.org/10.1016/j.psj.2022.101968 INTRODUCTION infection. Poultry and poultry meat can become contam- inated with Salmonella throughout the entire poultry Salmonellosis remains one of the main foodborne zoo- production chain (breeder farms, fattening farms, trans- noses in Europe, with 87,923 cases in 2019 confirmed by portation, slaughterhouse, and retail; Ramirez- a laboratory test or based on the clinical symptoms and Hernandez et al., 2021). Thus, to ensure effective con- epidemiological link (EFSA and ECDC, 2021), and trol, insights into the occurrence of Salmonella and fac- poultry products are the main source of human tors affecting their prevalence are essential. Previous studies have identified the slaughterhouse as Ó 2022 The Authors. Published by Elsevier Inc. on behalf of Poultry a potential source for Salmonella contamination of poul- Science Association Inc. This is an open access article under the CC try meat, representing a critical stage for controlling its BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ dissemination in the food chain (Zeng et al., 2021). Dur- 4.0/). ing slaughter, carcasses may become contaminated by Received February 16, 2022. Accepted May 16, 2022. bacteria found in the intestinal content of the animals, 1 Corresponding author: clara.marin@uchceu.es either from the same flock or in previously slaughtered 1 2 RESEARCH NOTE flocks (Zeng et al., 2021). In recent years, strict biosecur- commercial conditions. The flock definition used was a ity and management protocols have been established at group of chickens reared in a broiler house during the farm level, which have made it possible to reduce and same time period and under the same rearing conditions. control the prevalence of the bacteria to a target estab- lished by National Salmonella Control Programmes (NSCP) set up by the European Commission. However, despite the important economic investment made by the Sample Collection governments and the poultry sector, there are companies From each flock, samples were collected at different with persistent Salmonella strains that cannot be times: when animals arrived at the slaughterhouse, removed from the processing environment, compromis- before slaughtering the flock, and while processing car- ing the entire production line. Therefore, when Salmo- casses (Figure 1). nella enters into the slaughterhouse from the farm, the From each flock studied, 2 feces samples (200−300 g bacterium is able to persist in the slaughterhouse envi- each, n = 72) were collected directly from truck crates ronment despite cleaning and disinfection procedures during unloading at the slaughterhouse (EC, 2005). The available in the agri-food chain, remaining viable for up flock was declared infected if at least one of the samples to 15 mo (Barron and Forsythe, 2007). Due to this situa- was tested as positive (n = 36). Before processing the tion, the rapid detection of the bacteria, characteriza- carcasses, environmental samples were taken. For this tion of the strain and identification of the contamination purpose, one 10 cm2 sample was collected by rubbing routes at the slaughterhouse are essential for field level the entire area with sterile gauze pads wet with a disin- control and to safeguard the food supply effectively fectant neutraliser (AES Laboratories, Bruz Cedex, (Bell et al., 2016). In this perspective, the use of techni- France) from the steel surface that came into contact ques with discriminatory power to explore the possible with carcasses in 6 selected stages: exsanguination, relationships among the circulating strains is of particu- scalding, defeathering, evisceration, grading, and air- lar interest for source attribution studies. To this end, chilling (Figure 1). One sample of the foodstuff contact an intensive sampling study was conducted to investi- surfaces of clean finished product transporting crates gate a slaughterhouse with persistent Salmonella prob- was also collected. Finally, one carcass was taken from lems, establishing the genetic relationship among each of the 3 selected processing line stages: after the Salmonella strains isolated during the slaughter process. exsanguination, defeathering, and air-chilling stages. Each carcass was collected under aseptic conditions and MATERIALS AND METHODS placed into an individual sterile bag (Seward, Worthing, UK). All the collected samples were transported in a All the procedures used in this study were performed cool box at ≤4°C to the laboratory for microbial analyses in accordance with Directive 2010/63/EU EEC for ani- and were analysed within 2 h after sampling. mal experiments. Laboratory carcass sampling was carried out by col- lecting neck skin and internal cavity surface samples. Study Design The skin samples (25 g) were placed into a sterile bag and diluted at 1:10 vol/vol in Buffered Peptone Water Over a year (2015), a poultry slaughterhouse with per- 2.5% (BPW, Scharlau, Barcelona, Spain). The mix was sistent Salmonella was intensively sampled. Overall, 36 homogenized by stomaching at 230 rpm for 120 s (Stom- broiler chicken flocks were sampled in 18 different visits acher 400 circulator, Seward Ltd., Worthing, UK). Sam- (the first- and last-processed flocks per visit). The proc- ples from the internal cavity surface of each carcass were essing line constituted part of an integrated broiler also collected with sterile wet gauze pads (AES laborato- chicken company and operated under standard ries, Bruz Cedex, France). To this end, each carcass was Figure 1. Samples collected at different times of the working day at the slaughterhouse. Abbreviation: ES, environmental samples. RESEARCH NOTE 3 aseptically opened and internally rubbed with sterile wet to the data to determine the presence and diversity of gauze. the bacteria in live chickens upon arrival at the slaugh- terhouse, in carcasses in the different poultry slaughter process stages, and the impact of each processing stage Salmonella Isolation and Serotyping on the microbiological contamination of chicken broiler For Salmonella detection, faeces (25g) and environ- carcasses. A flock was declared infected if at least one of mental wet gauze pads, neck skin homogenates, and the samples collected was positive. The association with internal surface cavity were analyzed (ISO 6579:2017). the status of live broiler chicken flocks and Salmonella- According to ISO, samples were first pre-enriched in contaminated carcass samples (neck skin and internal 1:10 vol/vol BPW 2.5% and incubated at 37 § 1°C for cavity surface) from the exsanguination, defeathering 18 § 2 h. Pre-enriched samples were transferred to Semi- and air-chilling stages was also studied. A carcass sam- -Solid Rappaport Vassiliadis agar plates (MSRV, Difco, ple was stated as being contaminated if at least one of Valencia, Spain) and incubated at 41.5 § 1°C for 24 to the samples (neck skin and/or internal cavity surface) 48 h. Suspicious plates were transferred to 2 different from each stage tested positive. In all the tests, an error agar plates, ASAP (AES Laboratories, Bruz Cedex, was designated as having a binomial distribution and France) and Xylose Lysine Deoxycholate agar (XLD, the probit link function was used. The binomial data for Liofilchem, Valencia, Spain), and were incubated at 37 each sample were assigned 1 if they presented prevalence § 1°C for 24 § 3 h. After the incubation period, 5 Salmo- of Salmonella, or 0 otherwise. A P value of less than 0.05 nella suspect colonies were selected and transferred to a was considered to indicate a statistically significant dif- nutrient agar plate (Scharlab, Barcelona, Spain) incu- ference. Data are presented as least squares means § bated at 37 § 1°C for 24 § 3 h. Finally, an API-20E bio- standard deviation. The analyses of Salmonella results chemical test (API-20, bioMerieux, Madrid, Spain) was were carried out by a commercially available software performed to confirm Salmonella spp. program (SPSS 16.0 software package; SPSS Inc., Chi- Salmonella isolates were serotyped by Kauffman- cago, IL, 2002). White-Le-Minor technique at the Centro de Calidad Avícola y Alimentaci on Animal of the Valencian Com- munity (Castellon, Spain). RESULTS AND DISCUSSION Since the National Salmonella Control Plans (NSCP) Molecular Typing of Salmonella Isolates were set out in poultry production, a significant reduc- tion of human cases of infection has been achieved. How- Two different subtyping methods were used for geno- ever, we have reached a prevalence of Salmonella that is typing Salmonella isolates. Clonality among the Salmo- difficult to eradicate from poultry production systems, nella isolates was initially assessed by enterobacterial with the presence of "persistent cases" in poultry facili- repetitive intergenic consensus (ERIC)-PCR, as previ- ties that allow Salmonella to remain the leading cause of ously described by More et al. (2017). Representative food outbreaks in Europe (EFSA and ECDC, 2021). isolates from the different Salmonella ERIC-PCR pat- During this study, a total of 504 samples were col- terns identified per sample were further analyzed by lected from eight points of the slaughterhouse. Samples pulsed-field gel electrophoresis (PFGE). were collected from feces (n = 36), environmental surfa- PFGE was performed according to the PulseNet ces (n = 252), neck skin (n = 108), and internal cavity standardized protocol ‘‘Standard Operating Procedure surfaces (n = 108). for PulseNet PFGE of Escherichia coli O157:H7, From all samples collected at the slaughterhouse, Escherichia coli non-O157 (STEC), Salmonella 45.0% (227/504) were positive for Salmonella. From serotypes, Shigella sonnei and Shigella flexneri” environmental surfaces, 40.4% (102/252) of the samples (www. pulsenetinternational.org). Isolates were ana- were positive for Salmonella, with statistically signifi- lyzed using the Xbal restriction enzyme (Roche Applied cant differences between the different stages, with Science, Indianapolis, IN). defeathering and grading (69.0 § 7.7 and 56.0 § 8.6%, ERIC-PCR and PFGE band patterns were analysed respectively) being the most contaminated stages in the using Fingerprinting II software, v3.0 (Bio-Rad, Hercu- slaughterhouse (P = 0.00). These stages were followed les, CA). Similarity matrices were calculated with the by evisceration, crate-transporting (47.0 § 8.3%, both) Dice coefficient and a cluster analysis was performed by and exsanguination (44.0 § 8.3%), and finally air-chill- the unweighted-pair group method with arithmetic ing and scalding (14.0 § 5.8 and 6.0 § 3.8%, respec- mean (UPGMA). The isolates with a minimum level of tively; P < 0.05). similarity of 90% were considered genetically similar or According to the different batches sampled, 69.4% identical and were assigned the same pulsotype. (25/36) arrived at the slaughterhouse shedding Salmo- nella in faeces, and 46.3% (100/216) of the different sam- Statistical Analysis ples from carcasses (neck skin and internal cavity surfaces) were also contaminated with Salmonella. The A generalized linear model, which assumed a binomial frequency of Salmonella-positive samples collected from distribution for the colonising of Salmonella, was fitted the chicken neck skin in the exsanguination, 4 RESEARCH NOTE defeathering and air-chilling stages was 67.0 § 7.9, 81.0 survive the cleaning and disinfection protocols, leading § 6.6, and 31.0 § 7.7%, respectively, being the percent- to cross-contamination among different slaughter stages. age of the Salmonella-positive samples statistically sig- In this study, exsanguination, evisceration, and crate- nificantly lower at the air-chilling stage (P < 0.05). For transporting were particularly relevant stages in the the internal cavity surface, the frequency of Salmonella- bacteria epidemiology. Exsanguination is one of the first positive samples in the exsanguination, defeathering, steps in the slaughter process, where birds can carry sig- and air-chilling stages was 3.0 § 2.7, 44.0 § 8.3, and nificant loads of microorganisms from the feathers, skin, 53.0 § 8.3%, respectively, being the percentage of the and feces from the farm environment and from their Salmonella-positive samples statistically significantly transport from the farm to the slaughterhouse. More- lower at the exsanguination stage (P < 0.05). For each over, the evisceration step is particularly relevant, where stages, there were statistically significant differences in a rupture of the intestinal tract could contaminate the both types of chicken carcass samples (neck skin and equipment. Finally, similarly to other studies, the crate- internal cavity surfaces; P < 0.05). transporting stage was considered a critical stage due to According to the status of the broiler flocks upon their the possibility of Salmonella adhesion to wet surfaces arrival at the slaughterhouse: when the Salmonella sta- and formation of biofilms. Bacteria within biofilms are tus of live flocks were positive (n = 25), 48 § 7.1% (12/ up to 1,000 times more tolerant to disinfectants, 25), 56 § 7.0% (14/25), and 55 § 7.3% (13/25) of the mechanical removal, and other stresses (Dieltjens et al., chicken carcass samples (neck skin and internal cavity 2020), facilitating bacterial survival. surfaces) were Salmonella-positive in the exsanguina- To assess the genetic relationship among isolates tion, defeathering and air-chilling stages, respectively. recovered from the different steps at the slaughter- However, when the Salmonella status of live flocks was house, the isolates were initially typed by ERIC-PCR. negative (n = 11), 5 § 4.4% (1/11), 77 § 8.9% (8/11), The result of this screening technique showed 89 differ- and 80 § 6.8% (9/11) of the collected chicken carcass ent patterns. One strain from each of the 89 different samples were Salmonella-positive in the exsanguination, ERIC-PCR patterns was selected for further analysis defeathering and air-chilling stages, respectively. The by PGFE. Strains of serovar Enteritidis were grouped chicken carcasses sampled from the exsanguination in 2 PFGE clusters (≥90% similarity each) which were stage showed significant differences according to live closely related (≥ 85.0% genetic similarity). Moreover, flock status (positive or negative) and when compared S. Kentucky strains showed two different pulsotypes to the different stages studied (P = 0.00). Nevertheless, among the 4 S. Kentucky isolates, showing the S. Ken- no significant differences were found between the Salmo- tucky strains from the environmental samples the nella-contaminated chicken carcass samples from either same XbaI PFGE profile (95.0% genetic similarity; the defeathering or the air-chilling stage, nor among Figure 2). stages when considering life flock status before slaughter Salmonella infection of chickens during rearing has (P ≥ 0.05). been studied by several authors in different countries, During the study, 227 Salmonella strains from 2 differ- including under Spanish conditions (Marin and Lai- ent serovars were isolated: 98.2% belonged to serovar nez, 2009). Nevertheless, in many cases, the results Enteritidis (223/227) and 1.8% to serovar Kentucky (4/ obtained were difficult to interpret, so it was not always 227). Salmonella enterica subsp. enterica serovar Enteri- possible to find a solution to the problem. Currently, the tidis was found in all different sample types collected at genotypic characterisation of isolates makes it possible all stages. Meanwhile, Salmonella enterica subsp. enter- to identify the different genotypes found within a partic- ica serovar Kentucky was found in faeces and neck skin ular serovar present in the slaughterhouse. In this study, samples (exsanguination and defeathering) from one the same Salmonella serovar were isolated in 98% of the flock. Moreover, S. Kentucky was recovered from the samples. Considering that the same genotypes could be environment in the exsanguination stage, after process- isolated from different farms, owned by the same inte- ing the S. Kentucky-contaminated flock. grated company, it strongly suggests a cross-contamina- The status of the flock when it arrives at the slaugh- tion throughout the broiler supply chain (Ha et al., terhouse is a challenge to avoid cross-contamination 2018). during processing, and an important factor for safe fin- This study demonstrated that in an integrated com- ished product quality. Notably, Salmonella Enteritidis is pany with a high presence of persistent cases of Salmo- the most frequent serovar isolated in this study. Since nella, there was a high genetic relationship among the the 1980s, this serovar has been the most frequently strains isolated from the incoming flocks, the slaughter involved in human salmonellosis in Europe, representing environment and their carcasses (Zeng et al., 2021). 72.4% of confirmed foodborne outbreaks in 2019 This fact could be explained by the survival of certain (EFSA and ECDC, 2021). clones linked to the companies, with a high resistance to The persistent cases of Salmonella can be explained by adverse conditions that allows them to recirculate con- reasons such as inaccurate cleaning and disinfection pro- tinuously throughout the production chain. In addition, cedures (Zeng et al., 2021), the development of biofilms it has been observed that regardless of whether the flocks (Marin and Lainez, 2009), or the high resistance of the arrive at the slaughterhouse excreting the bacteria or circulating strains (Bridier et al., 2019). Indeed, in this not, the slaughterhouse surfaces continued to be highly study, the same strain of S. Enteritidis was able to contaminated (Zeng et al., 2021). RESEARCH NOTE 5 Figure 2. Dendrogram of the XbaI PGFE profiles from a subset of Salmonella strains isolated from environmental and chicken samples (n = 89). Abbreviations: ES, environmental sample; FS, feces sample; ICS, Internal surface cavity; NSS, neck skin sample. Arrival: Truck crates. In Spain, 80% of the broiler production chain is con- transmission and maintenance of other zoonotic bacteria trolled by integrated companies which have not only such as Salmonella, Listeria or Campylobacter. This broiler farms but also feed plants, breeders, and slaugh- highlights the need for improvement of the biosecurity terhouses. This has been considered a risk for the measures and vaccination programmes to control 6 RESEARCH NOTE persistent Salmonella strains from breeder farms to REFERENCES broiler farms, to reduce infection pressure in the slaugh- Barron, J. C., and S. J. Forsythe. 2007. Dry stress and survival time of terhouse. In addition, it is necessary to implement Enterobacter sakazakii and other Enterobacteriaceae in dehy- molecular diagnosis methods in time at the field level, drated powdered infant formula. J. Food Prot. 70:2111–2117. enabling us to track the potential source of strains, to Bell, R. L., K. G. Jarvis, A. R. Ottesen, M. A. McFarland, and make rapid decisions to control the pathogen, limiting E. W. Brown. 2016. Recent and emerging innovations in Salmo- nella detection: a food and environmental perspective. Microb. the entry of the bacterium into the slaughterhouse. Biotechnol. 9:279. Bridier, A., P. Le Grandois, M.-H. Moreau, C. Prenom, A. Le Roux, C. Feurer, and C. Soumet. 2019. Impact of cleaning and disinfec- tion procedures on microbial ecology and Salmonella antimicrobial ACKNOWLEDGMENTS resistance in a pig slaughterhouse. Sci. Rep. 91:1–13 2019. Dieltjens, L., K. Appermans, M. Lissens, B. Lories, W. Kim, This study was funded by the CEU-Cardenal Herrera E. Van der Eycken, K. Foster, H. Steenackers, and D. Ermolat’ev. 2020. Inhibiting bacterial cooperation is an evolu- University as part of the “Programa de Consolidaci on de tionarily robust anti-biofilm strategy. Nat. Commun. 11:107. Indicadores de Investigaci on” (INDI21/35). Laura Lor- EFSA and ECDC. 2021. The European Union One Health 2019 zoo- enzo-Rebenaque was supported by a research grant from noses report. EFSA J. 19:6406. the Generalitat Valenciana- European Social Fund Ha, J. S., K. Won Seo, Y. Bin Kim, M. S. Kang, C.-S. Song, and Y. J. Lee. 2018. Prevalence and characterization of Salmonella in (ACIF/2020/376). The authors wish to thank the two integrated broiler operations in Korea. Ir. Vet. J. 71:1–9. “Improvement of Production System-related Food Marin, C., and M. Lainez. 2009. Salmonella detection in feces during Safety and End Products” research group (Veterinary broiler rearing and after live transport to the slaughterhouse. Faculty, University Cardenal Herrera-CEU) for the Poult. Sci. 88:1999–2005. More, E., T. Ayats, P. G. Ryan, P. R. Naicker, K. H. Keddy, technical support. The CERCA Programme from the D. Gaglio, M. Witteveen, and M. Cerd a-Cuellar. 2017. Seabirds Generalitat de Catalunya is also acknowledged. The (Laridae) as a source of Campylobacter spp., Salmonella spp. and English text version was revised by N. Macowan English antimicrobial resistance in South Africa. Environ. Microbiol. 19:4164–4176. Language Service. Ramirez-Hernandez, A., A. K. Carrascal-Camacho, A. Var on-García, M. M. Brashears, and M. X. Sanchez-Plata. 2021. Genotypic char- acterization of antimicrobial resistant Salmonella spp. strains from three poultry processing plants in Colombia. Foods 10:491 2021. DISCLOSURES Zeng, H., K. De Reu, S. Gabri€el, W. Mattheus, L. De Zutter, and G. Rasschaert. 2021. Salmonella prevalence and persistence in The authors declare no conflict of interest. industrialized poultry slaughterhouses. Poult. Sci. 100:100991.