Abstract
Meat products may be vehicles of bacterial pathogens to humans, and Salmonella spp., Campylobacter spp., and Shiga toxin–producing Escherichia coli (STEC) are the most relevant. The aim of this study was to generate data on prevalence of these three pathogens in 552 samples of meat products (hot dogs, pork sausages, raw ground beef, and raw chicken legs) sold at retail in the city of Sao Paulo, Brazil. Salmonella spp. was detected in 5.8% (32/552) of samples, comprising pork sausages 62.5% (20/32) and chicken legs 37.5% (12/32). The counts of Salmonella spp. were low, ranging from < 0.3 to 9.3 × 10 most probable number per gram and the most frequent serovars were Salmonella Typhimurium (28.1%), Salmonella I 4,[5],12:i:- (15.6%), Salmonella Enteritidis (12.5%), Salmonella Derby, and Salmonella Brandenburg (9.4%). Campylobacter spp. was detected in 33 samples (6.0%), comprising chicken legs (82%) and ground beef (18%). All samples were negative for STEC. These results suggest that meat products when subjected to inadequate cooking and/or cross-contamination with other products ready for consumption can lead to occurrence of outbreaks, highlighting the risks associated with them.
Introduction
D
In Brazil, there is no report on foodborne outbreaks caused by STEC or Campylobacter spp., but Salmonella spp. was responsible for the majority of outbreaks with known etiology occurring between 2000 and 2016 (MS, 2016). A study conducted in southern Brazil on Salmonella in foods between 2007 and 2012 has indicated that meat products were the most common vehicles (Capalonga et al., 2014). Owing to subnotification, the real occurrence of foodborne diseases in Brazil is not known. Most of the available data are from the southeast and southern states, where active sanitary and epidemiological surveillance services have been implemented. Furthermore, Brazilian official laboratories usually investigate only classical foodborne microorganisms (i.e., Salmonella spp., Bacillus cereus, Staphylococcus aureus, and Clostridium perfringens), but not emergent etiological agents (Ritter and Tondo, 2014).
Fragility of epidemiological data on the prevalence and characteristics of foodborne pathogens and scarcity of quantitative data on occurrence of these pathogens in the food chain are the main gaps in proper risk assessments in many countries, and Brazil is not different. This study was undertaken to generate data on Salmonella spp., Campylobacter spp., and STEC in meat products sold in retail stores, to subsidize government and industry with more information for risk assessments and for the establishment of proper measures to protect consumer health.
Materials and Methods
Samples
The study was conducted with 552 samples of refrigerated meat products (138 hot dogs, 138 raw pork sausages, 138 raw ground beef, 138 raw chicken legs) collected from 138 supermarkets located in 77 districts of the Sao Paulo city, Brazil. The supermarket in each region selected for sampling was determined by random draw from a list provided by the Sao Paulo Municipality. The samples were collected directly from the vending shelves by technicians of the Sanitary Surveillance Service of Sao Paulo city, transported to the laboratory in containers with recyclable ice and kept under refrigeration until analyzed (maximum 18 h).
Analytical methods
Detection and enumeration of Salmonella spp
The detection of Salmonella spp. was determined according to ISO 6579:2002, using a homogenate prepared with a 25 g portion of each sample and 225 mL of buffered peptone water (BPW) in a Stomacher 400 (Seward, UK) for 1 min. Enumeration was done using the three-tube most probable number (MPN) technique (Swanson et al., 2001) followed by ISO 6579:2002 for confirmation of Salmonella. The homogenate (25 g) and MPN tubes were incubated at 37 ± 1°C for 18 ± 2 h. After the pre-enrichment step, 1 and 0.1 mL from each BPW culture were transferred into Muller–Kauffmann tetrathionate–novobiocin broth and Rappaport–Vassiliadis soy peptone broth, respectively, and incubated for 24 ± 3 h at 37 ± 1°C and 41.5 ± 1°C, respectively. Aliquots of each enrichment broth were plated onto xylose–lysine–deoxycholate agar and Salmonella–Shigella agar, and incubated at 37 ± 1°C for 24 ± 3 h. Typical colonies were subjected to presumptive identification in IAL agar (Instituto Adolfo Lutz medium) at 37 ± 1°C for 24 ± 3 h (Pessoa and Silva, 1974) and those colonies presenting typical biochemical responses of Salmonella were plated onto nutrient agar, incubated at 37 ± 1°C for 24 ± 3 h, and confirmed by serotyping using polyvalent sera. Complete serotyping was performed according to Popoff and Le Minor (2005). Results of detection are expressed as presence or absence of Salmonella spp. in 25 g and of enumeration as MPN/g. Hoskins Table was used for calculation of MPN/g. All culture media and supplements were from Oxoid Ltd. (Oxoid, UK), except otherwise stated.
Detection of Campylobacter spp
Twenty-five grams of each sample was homogenized with 225 mL of Bolton Selective Enrichment Broth containing Laked Horse Blood and Bolton Broth Selective Supplement (cefoperazone, vancomycin, trimethoprim, and cycloheximide) in a Stomacher 400 (Seward, UK) for 1 min (ISO 10272-1:2006, International Organization for Standardization, 2016). All culture media and supplements were from Oxoid Ltd., except otherwise stated. The mixture was incubated at 37 ± 1°C for 4 h, followed by incubation at 41.5 ± 1°C for 44 ± 4 h under microaerophilic conditions (5% O2, 10% CO2, and 85% N2) (BBL, US). The enrichment broth was plated onto charcoal cefoperazone deoxycholate modified agar and Skirrow agar, and incubated at 41.5 ± 1°C for 44 ± 4 h in microaerophilic atmosphere. Three to five presumptive Campylobacter colonies from each selective agar plate were transferred to Columbia agar plates and incubated at 41.5 ± 1°C for 24–48 h under microaerophilic conditions. Suspect colonies were subjected to Gram staining, catalase, oxidase, growth at 25°C and 41.5°C, resistance to nalidixic acid and cephalothin, and hydrolysis of hippurate and acetate. Results are expressed as absence or presence of Campylobacter spp. in 25 g.
Detection of STEC
Detection of STEC followed the methodology described by Meng et al. (2001). Portions (25 g) of the meat product samples were homogenized with 225 mL of modified E. coli (EC) medium (Difco, US) supplemented with 20 mg/L novobiocin (Inlab, Brazil) in a Stomacher 400 (Seward, UK) for 1 min, and incubated overnight at 35 ± 1°C. The enriched cultures were plated onto Sorbitol MacConkey agar (SMAC) and SMAC agar containing cefixime and tellurite (CT-SMAC), and incubated at 35 ± 1°C for 18–24 h. Both SMAC and CT-SMAC (Difco, US) were used to investigate sorbitol positive and negative E. coli O157:H7 and other STEC strains. Five to 10 sorbitol positive and negative colonies were transferred to IAL agar and incubated at 37 ± 1°C for 24 ± 3 h for identification of E. coli, according to Pessoa and Silva (1974). Suspect colonies were streaked onto tryptone soya agar, incubated at 35 ± 1°C for 18–24 h, and tested for lactose fermentation, utilization of citrate, and Voges Proskauer and methyl red tests.
Strains confirmed as E. coli were subjected to multiplex polymerase chain reaction (PCR) for detection of stx1 , stx2 , eaeA, and ehxA genes (Paton and Paton, 1998). The strains negative for stx1 and/or stx2 genes but positive for the eaeA gene were also tested for bfp gene (Gunzburg et al., 1995). The following strains were used as controls in the PCR tests: E. coli O157:H7 IAL 1848 positive for stx1 , stx2 , eaeA, and ehxA genes; E. coli O127:H6 strain E2348/69 positive for bfp gene; and E. coli ATCC 25922 negative control. All strains were grown in brain heart infusion broth at 35 ± 1°C for 18–24 h and aliquots (1 mL) were centrifuged (14,000 × g) for 2 min in a Mikro 120 centrifuge (Hettich, Germany). The DNA was extracted from cells using the Wizard® Genomic DNA Purification System (Promega, US), following the manufacturer's instructions. In these tests, the amplification cycle comprised an initial warming to 95°C for 5 min followed by 35 cycles of 45 s at 95°C, 45 s at 56°C, and 1 min at 72°C. The final extension was done at 72°C for 5 min.
The E. coli strains positive for at least one of the genes tested (stx1 , stx2 , eaeA, and ehxA) were serotyped by standard agglutination tests (Ewing, 1986) using O (O1-O181) and H (H1-H56) antisera produced at Adolfo Lutz Institute, Sao Paulo, Brazil.
Enumeration of thermotolerant coliforms as hygiene indicators
Enumeration of thermotolerant coliforms was done according to the classical MPN technique using sodium lauryl sulfate tryptose and EC broths and incubation at 35 ± 0.5°C for 48 ± 3 h and 45.5 ± 0.2°C for 48 ± 2 h, respectively (Kornacki and Johnson, 2001). The MPN of thermotolerant coliforms per gram of product (MPN/g) was calculated using the Hoskins Table.
Results
Salmonella spp. and Campylobacter spp. were detected in 32 (5.8%) and 33 (6.0%) of the 552 meat product samples, respectively. None of the samples was positive for STEC.
Among the 32 Salmonella-positive samples, 20 (62.5%) were pork sausages and 12 (37.5%) were chicken legs (Tables 1 and 2). None of the hot dog or ground beef samples was positive. The populations of Salmonella spp. were low, ranging from < 0.3 to 9.3 × 10 MPN/g. Three pork sausage samples and one chicken sample were positive only in the enumeration method.
Determined by the ISO 6579:2002 method (ISO, 2002).
Determined by the MPN method (Swanson at al., 2001).
MPN, most probable number.
Determined by the ISO 6579:2002 method (ISO, 2002).
Determined by the MPN method (Swanson et al., 2001).
The Salmonella spp. isolated from all positive samples belonged to 14 serovars: Salmonella Typhimurium (28.1%), Salmonella I 4,[5],12:i:- (12.5%), Salmonella Enteritidis (12.5%), Salmonella Derby (12.5%), Salmonella Brandenburg (9.4%), Salmonella enterica subsp. diarizonae 61: c: z35 (6.2%), Salmonella Infantis (6.2%), Salmonella Anatum (3.1%), Salmonella Newport (3.1%), Salmonella Ohio (3.1%), Salmonella Rissen (3.1%), Salmonella Schwarzengrund (3.1%), Salmonella enterica subsp. diarizonae 61:c:- (3.1%), and Salmonella I 4,[5],12:-:- (3.1%). Two pork sausage samples contained Salmonella spp. belonging to more than one serovar simultaneously (Table 2).
Among the 33 samples positive for Campylobacter spp., 27 (82%) were chicken legs and 6 (18%) were ground beef. None of the hot dog or pork sausage samples was positive. Chicken legs contained mainly Campylobacter coli (51.8%) and C. jejuni (33.3%) and three samples (11.1%) contained the two species simultaneously. In one sample, species of Campylobacter could not be determined. All isolates from the ground beef samples were characterized as C. jejuni.
Among the 171 E. coli isolates evaluated for stx1 , stx2 , eaeA, and ehxA genes, 3 (1.7%) were positive for at least one of them. One isolate, obtained from pork sausage, was positive for eaeA and ehxA genes, and two from chicken legs were positive only for the eaeA gene. These three isolates were then tested for the bfp gene and found negative, belonging to the atypical enteropathogenic E. coli (EPEC) group. The detected serotypes were ONT:H11 (pork sausage), OR:HNT (chicken leg), and O181:HNM (chicken leg).
Presence of thermotolerant coliforms was observed in 52.9% of the samples (chicken legs 94.9%, ground beef 59.4%, pork sausages 47.8%, and hot dogs 9.4%). The populations ranged from <3.0 to 4.6 × 103 MPN/g in the chicken legs and from <3.0 to ≥2.4 × 103 MPN/g in pork sausages, hot dogs, and ground beef samples. The populations were <10 MPN/g in 65.4% of the meat products (Table 3). Table 4 shows that the most samples positive for pathogens contained <102 MPN/g thermotolerant coliforms.
Discussion
Prevalence and quantification studies of pathogens in commercially acquired meat products provide estimates that reflect the consumer exposure level. The results of this study showed the presence of Salmonella spp., Campylobacter spp., and EPEC but not STEC in the evaluated meat products. Hot dogs were the only products that did not present any of the pathogens studied, certainly in consequence of the thermal processing.
Among the tested meat products, Salmonella spp. was detected in pork sausages and chicken legs only. A previous study with hot dogs (Martins et al., 2008; Aslam et al., 2012), raw ground beef, and other meats (Vipham et al., 2012) also failed in detecting Salmonella spp. Most studies describe low levels of Salmonella spp. in raw meat products (Mürmann et al., 2009; Ta et al., 2014; Wang et al., 2014b; D'Ostuni et al., 2016), as also observed in this study (<0.3–0.91 MPN/g in chicken legs and <0.3–9.3 × 10 MPN/g in pork sausages).
In three samples, Salmonella spp. was detected only by the enumeration method, probably because the effect of spices and other interfering food components was diminished by the serial dilution used in this method.
The prevalence of Salmonella spp. in chicken leg is comparable to other studies in Brazil (Medeiros et al., 2011) and other developing countries (Akbar and Anal, 2013; Naik et al., 2015). However, much higher rates (>35%) have been recently reported in China (Huang et al., 2016) and Vietnam (Ta et al., 2014).
The most common Salmonella serovar in the chicken legs was Salmonella Enteritidis (33.3%), in accordance with results of previous studies in poultry products in Brazil (Tavechio et al., 2002; Costa et al., 2013; Capalonga et al., 2014) and other countries (Wang et al., 2014b; Abd-Elghany et al., 2015). In the pork sausages, Typhimurium (45%) and Derby (20%) serovars prevailed, as in other studies with pigs and related products (Kich et al., 2011; Bonardi et al., 2013). It is noteworthy that the Salmonella I 4,[5],12:i:- serovar found in the pork sausage and chicken leg samples has been implicated in foodborne disease outbreaks in recent years (Guillier et al., 2013; EFSA and ECDC, 2014). Also, this serovar is one of the most frequently detected in clinical samples in Brazil (Fernandes et al., 2006).
Although Campylobacter spp. is one of the main causes of foodborne disease outbreaks in the United States and Europe (CDC, 2013; EFSA and ECDC, 2014), there is no information on cases of campylobacteriosis associated with foods in Brazil. However, the pathogen is often isolated from children with diarrhea (Moreno et al., 2010; Quetz et al., 2012).
In this study, Campylobacter spp. was isolated from 19.6% of chicken leg samples, which is similar to previous studies in Brazil: 10.8% (De Carvalho et al., 2014) and 19.6% (De Moura et al., 2013). In other countries, the prevalence seems to be higher, but a wide variation can be observed, from 0.29% to 90% (Medeiros et al., 2008; Hara-Kudo et al., 2013; Scheinberg et al., 2013; Wang et al., 2013; Ma et al., 2014; Sison et al., 2014; Huang et al., 2016). These variations can be attributed to many factors such as sampling, analytical methodology, chicken brands, animal handling, and hygienic practices (Scheinberg et al., 2013; Wang et al., 2013; Huang et al., 2016).
The prevalence of Campylobacter spp. in ground beef was 4.5%, similar to observations of Little et al. (2008) in the United States (4.9%) and Trokhymchuk et al. (2014) in Canada (4.5%), but much higher than the prevalence reported by Llarena et al. (2014) in Finland (0.6%). Campylobacter spp. was not detected in the hot dogs or pork sausages, which is in contrast with other studies in pork products that reported the presence of this pathogen in 2% of samples in the United States (Noormohamed and Fakhr, 2013) and in 15% in Ireland (Scanlon et al., 2013). To the authors' knowledge, there is no other study on prevalence of this pathogen in ground beef, hot dogs, and pork sausages in Brazil. The processing steps to which meat products are subjected as well as heat treatment, cooling, and freezing can reduce or even eliminate this heat-sensitive and fragile pathogen (Ghafir et al., 2007; Medeiros et al., 2008), which may explain the absence of Campylobacter spp. in hot dog and pork sausage samples.
The absence of STEC in the meat samples in this study is in accordance with Zhao et al. (2001) in the United States, Kumar et al. (2014) in India, and Wang et al. (2014a) in China. In Brazil, despite the low frequency of proven human infections caused by STEC (Lozer et al., 2013; Dias et al., 2016), this pathogen has been detected in Brazilian cattle (Rigobelo et al., 2008; Feng et al., 2014; Ferreira et al., 2014), raw ground beef (Bergamini et al., 2007), and raw beef (Feng et al., 2014), in contrast to reports in other countries (Beneduce et al., 2008; Lee et al., 2009; Brusa et al., 2013; Momtaz and Jamshidi, 2013) that detected STEC in beef, poultry, and pigs.
Despite the negative results for STEC in the tested meat products, atypical EPEC was detected in pork sausage and chicken leg samples, which is worrisome as some studies have associated atypical EPEC with diarrhea in different geographic areas in Brazil (Trabulsi et al., 2002; Vieira et al., 2016). In addition, it has been suggested that atypical EPEC isolates containing eaeA and ehxA genes can be STEC that have lost the bacteriophages carrying the stx genes (Bielaszewska et al., 2008).
As indication of the hygienic–sanitary conditions of food samples, the current Brazilian regulation requires testing for thermotolerant coliforms (Brasil, 2001) and the limits for pork sausages, hot dogs, and chicken legs are 5 × 103 MPN/g, 103, and 104, respectively. The Brazilian regulation does not contain a limit for thermotolerant coliforms for raw beef. Most samples (89.3%) had low populations of thermotolerant coliforms, usually <102 MPN/g (Table 3), indicating satisfactory hygienic conditions. Only four (0.7%) samples, three pork sausages and one hot dog, failed to meet the requirements set by the Brazilian regulation. Among these samples, the pork sausages were also positive for Salmonella spp.
Table 4 shows that Salmonella spp. and Campylobacter spp. were detected in samples with counts of thermotolerant coliforms <3 MPN/g, demonstrating that low counts of fecal indicators in meat products do not guarantee the absence of pathogens.
In conclusion, this study indicated that meat products might represent a risk to health. Data on prevalence of Salmonella spp. and Campylobacter spp. are relevant for estimating the risks of foodborne diseases associated with consumption of meat products in Sao Paulo, and for establishing science-based intervention strategies aimed at reducing these risks.
Footnotes
Acknowledgments
The authors thank Fundação de Amparo à Pesquisa do Estado de Sao Paulo (FAPESP) (Grant No. 2007/54650-5) for financial support and COVISA (Coordenação de Vigilância em Saúde, Secretaria Municipal de Saúde de São Paulo, Brasil) for providing the meat product samples.
Disclosure Statement
No competing financial interests exist.
