Salmonellosis is a zoonotic disease caused by species of the genus Salmonella. They are Gram-negative microorganisms that include the species S. enterica and S. bongori, of which only the first one is of clinical relevance to humans and animals (1, 2). S. enterica includes over 2,610 serovars (3, 4). It is transmitted by direct contact with contaminated food from animal origin such as eggs and chicken meat (5). S. Enteritidis is considered the principal cause of human infection, and poultry products constitute the main source of the bacteria (6, 7, 8). The presence of Salmonella spp. in eggs is largely due to its ability to colonize the ovarian tissue and the surface of eggs during formation (9).

Salmonella spp. is prevalent in many developed countries including USA, Canada, UK, Norway, and Denmark (10, 11, 12),  where the bacterium is responsible for high morbidity and mortality ratios, particularly in children, aged people and immunocompromised patients (13). Nontyphoidal Salmonella might be responsible for about 1 million illnesses, 100,000 hospitalizations and 731 deaths in the United States annually (14). Salmonellosis in developing countries may also have a high prevalence; however, the disease is not properly notified-reported to the health-care centers, and the impact of the bacteria on public health is not accurately known. In Colombia, Salmonella was isolated in 31,7% (32 outbreaks) of foodborne diseases reported between January 2008 - August 2010 (15). The National Health Institute of Colombia reported 7,219 Salmonella isolates from clinical cases. S. Typhimurium (33,7%), S. Enteritidis (28,6%) and S. Typhi (9,2%) were the most prevalent serotypes (16). In the Tolima region, the epidemiological studies on the presence of Salmonella were very limited in different segments of the poultry industry. S. Enteritidisy S. Shannon were found as the main serotypes present in laying hen farms in this region (17), and subsequently a number of Salmonella serovarswere also identified in chicken carcasses sold at stores and supermarkets in Ibague city (18). This study was designed to estimate the prevalence of Salmonella in the surface of eggs marketed at stores of Ibague city for the period January – August, 2014, and to compare the conventional serotyping and genoserotypig using multiplex PCR-based typing.

MATERIAL AND METHODS

Study design

An observational cross-sectional study was conducted (between January - August, 2014) to establish the prevalence of Salmonella spp., in the surface of eggs marketed at stores and supermarkets in Ibague, Tolima. The sample size was calculated by the formula described by Thrusfield (2007) (19), with a 95% confidence level, 5% error, and an expected prevalence of 50%, given that no systematic studies were found on Salmonella in commercial eggs. The expected prevalence used in this study was based on the data from commercial egg laying hens, where the bacterium was more prevalent on eggshells than in feed or environmental samples (17).

The sampling included all the 13 communes  belonging to Ibague city. The number of samples per commune was proportional to the number of stores registered at the authority (Chamber of Commerce in Ibague). Each sample consisted on a pool of five eggs collected at each store, and a total of 341 samples was processed for Salmonella isolation. Thus, 1.705 eggs were sampled from the randomly selected stores or supermarkets across the city, where eggs were usually stored at room temperature (>25°C) or under cooling condition (<20°C). The eggs were packaged in sterile airtight plastic bags and transported to the Veterinary Diagnosis Laboratory for being processed within 12 hours. In addition, three S. Enteritidis isolates obtained from the egg surface in a pilot study (n=30) and six S. Enteritidis that were previously isolated from crushed eggshells (17) were also included in the study.

Isolation of Salmonella from egg surface

Salmonella spp. isolation followed the the international guidelines ISO 6579:2002/Amd1:2007. Briefly, eggs were immersed into a peptonated water solution and gentle rubbed on the shell´s surface for about five minutes to obtain a surface wash sample. A wash aliquot was inoculated into fresh peptone-buffered water for pre-enrichment and incubated at 37°C for 24 hr. Then, 1 ml of the pre-enrichment peptone water was sawn in 9 ml tetrathionate broth (Müller-Kauffmann) and incubated at 37°C. A second aliquot was inoculated in 9 ml of Rappaport Vassiliadis (MERCK®) broth and incubated at 42ºC for a selective enrichment. Later, bacterial colonies were sawn on SS (Salmonella-Shigella - MERCK®), XLD (Xylose Lysine Desoxycholate - MERCK®) and XLT4 (Xylose Lysine Tergitol 4 - MERCK®) agar. The compatible colonies were sawn in McConkey (MERCK®) and Rambach (MERCK®) agar and confirmed as Salmonella spp by an agglutination test with  Poli AI + Vi (Difco® 222641) antibodies. The positive colonies were also confirmed biochemically by using API® 20E gallery (Biomereux, France).

Serotyping of Salmonella spp. isolates

Salmonella spp. isolates were serotyped by using the White-Kauffman-Le Minor scheme (20) that identified the presence of specific somatic (O) and flagellar (H) antigens with the commercial antisera (Difco, Becton, Dickinson and Company Sparks, MD). Serotyping was performed based on the antigenic description established by Grimont and Weill (1) and the nomenclature described by Tindall et al. (21), and the Judicial Commission of the International Committee on Systematics of Prokaryotes (22). The procedure was carried out at the Colombian Agriculture Institute (ICA), following the methodology ISO 6579:2002/Amd1:2007.

Multiplex PCR-based typing of Salmonella spp.

The genomic DNA was extracted from each Salmonella isolate by the phenol chloroform isoamyl alcohol (25:24:1) method (23). All Salmonella samples isolated from the egg surface were subjected to 3 multiplex PCRs according to the previously described method (24, 25, 26). Three S. Enteritidis isolates from the egg surface in a pilot study (n=30) and six S. Enteritidis previously isolated from crushed eggshells (17) were also included in this analysis.  The procedure included a primary PCR reaction using 10 primers to identify Salmonella spp. serogroups based on the amplification of the genes involved in sugar biosynthesis pathways and transferases (rfB, wzX). Salmonella serovars were identified by two additional multiplex PCR reactions targeting fliC and fljB genes encoding the phase I (H1, 13 primers) and phase II (H2, 10 primers) flagellar antigens. The PCR reaction was performed in 25 µl total volume using the AccuPrime™ Taq DNA polymerase System (Invitrogen Life Technologies), containing 2.5 µl MgCl₂ (50 mM), 2.5 µl 10 × buffer, 0.8 µl of each primer (10 µM), 8 mM of dNTPs, 0.5 µl Taq DNA polymerase and 1 µl of DNA template. Amplification was carried out in a T-100 (Bio-Rad) thermal cycler with the following program: an initial denaturation step at 95°C for 5 minutes, followed by 35 denaturation cycles at 94°C for 1 min, annealing at 55°C for 30s and extension at 68°C for 30s, and a final extension step at 68 °C for 7 minutes. Salmonella TyphimuriumATCC 14028 and Salmonella EnteritidisATCC 13076 were used as positive controls and the negative control did not contain DNA template. PCR products were mixed with 2µl of 10× loading buffer and then resolved by electrophoresis on 2% agarose gel with 100 bp DNA ladder (Invitrogen®). The gel was stained with ethidium bromide and visualized under the UV light by using an ENDURO^TM GDS (Labnet International, Inc.) gel documentation system.

Epidemiologic survey 

To identify the potential risk factors for Salmonella spp. contamination, a questionnaire was administered to the shop owner/administers during an interview at the sampling time. The variables collected followed those previously described (27), and included the presence of pets in the store, cleaning conditions and frequency of cleaning and disinfection of the store (one vs. more than once per week), presence of sanitary unit or sink, presence of flies, presence of dirty (stool or dust) eggs, type of surface (smooth or rough), presence of long nails in the staff (long vs. short nails), storage temperature of eggs (20-25°C vs. >25 °C),  and staff personal clothing. The variables were optimized in the pilot study conducted by our research group at the University of Tolima.

A descriptive analysis of the epidemiologic variables evaluated at each store was conducted by using IBM SPSS Statistics® 20 version software and GraphPad Prism® 5.03 version software. The independence between the epidemiologic variable and the positivity to Salmonella spp. was determined by cross tabulation analysis. The association strength was calculated by the odds ratio. Prevalence was determined as the proportion of positive samples over the total samples, expressed as a percentage.

RESULTS

A total of 341 samples was cultured for Salmonella isolation, and 10 different isolates were obtained from egg surface. Thus a 2.93% prevalence of Salmonella in the surface of eggs marketed at Ibague, Colombia, was estimated for the period January - August 2014. Salmonella isolates were identified by conventional serotyping as S. Enteritidis (n=8), S. ParatyphiB (n=1) and S. bongori (n=1) (Table 1).

The multiplex PCR targeting the genes encoding the O-antigen showed two amplification products, one of 615 bp and the second of 230 bp (Fig. 1). All Salmonella Enteritidis (serogroup D1) amplified a single band of 615 bp, whereas S. Paratyphi B (serogroup B) amplified a band of 230 bp. The species S. bongori amplified 2 bands, one of 615 bp and one of 230 bp. The positive controls S. Enteritidis ATCC 13076 and S. typhimurium ATCC 14028 (serogroup B) also amplified a single band of 615 bp and 230 bp, respectively.

The multiplex PCR for phase I flagellar antigen H1 showed amplification products of 500 and 333 bp that correspond to the flagellar antigens H: 9, 12 and H: b: 1, 2. All Salmonella Enteritidis showed 2  bands, one of 500 bp and a second one of 333 bp. S. ParatyphiB showed a band of 150 bp. The positive controls Salmonella Enteritidis ATCC 13076 and Salmonella Typhimurium ATCC 14028 showed the corresponding 500 bp, 333 bp and 250 bp, respectively (Fig. 2). The third multiplex PCR directed to detect phase II flagellar antigens showed amplification products of 400 bp corresponding to the H:b:1,2 flagellar antigens described by Herrera-León et al. (25) and Echeita et al. (24). All S. Enteritidis showed no clear amplification products, whereas S. Paratyphi B showed a single amplification product of 400 bp. The positive control S. Enteritidis ATCC 13076 showed no amplification products, whereas S. Typhimurium ATCC 14028 showed a single band of about 400 bp (Fig. 3). The species S. bongori showed two bands, one of 400 bp and a second one of 100 bp.

Salmonella spp. was more frequently isolated (42.85%) from eggs at stores having a frequency of cleaning and disinfection once a week. A reduced frequency of cleaning at the store was identified as a potential risk factor for the presence of Salmonella (OR: 19.84, p<0.05) on eggs. The lifestyle of staff may also contribute to the contamination of eggs by Salmonella. The store staff with long nails were identified as a risk factor (OR: 5.46, p<0.05) for the presence of Salmonella on marketed eggs. Finally, in the present study, eggs kept at temperatures less than 25°C were identified as a risk factors for being contaminated with Salmonella (OR: 10.05, p<0.05) than eggs stored at temperatures over than 25°C.

DISCUSSION

Salmonella prevalence in eggs and laying-hen farms may vary considerable between different countries and geographical regions (27). In Europe, Salmonella prevalence in laying-hen farms ranged from 0% to 79.5% in 2007, with an average of 30.8%. In UK, a prevalence of 18% Salmonella in laying hens was reported in 2010 (28), and the frequency of Salmonella in eggs at the market place in Europe may vary between 7 - 24% (29). In Argelia, a prevalence range of 7.14% - 41.17% Salmonella in laying hens was reported in 2012 (30). In Colombia, a prevalence of 45.71% Salmonella in laying hen farms was reported in a study that included Cundinamarca, Santander, Bolivar and San Andres regions in 2013 (31), and recently our group reported a prevalence of 33.33% Salmonella in egg-laying-farms in the Tolima region (17). In this study, Salmonella prevalence in egg surface was slightly lower than that determined in crushed eggshells (5.16%) from eggs (8/155) in laying-hen farms in the Tolima region (17). The difference might be due to the methodology used to isolate Salmonella from eggs. The authors used eggshells and therefore incubated them into peptone water, whereas, in this study, we wanted to explore the possibility to isolate the bacteria from the egg surface without the need of breaking the eggs and by using a surface rinse obtained by 5 minute-incubation and rubbing the egg surface in peptone water. The prevalence of Salmonella estimated in this study is within the range (0 – 22.6%) reported in eggs for human consumption in the European Union and the United Kingdom (32); and it was very similar to the prevalence (3%) of Salmonella enterica reported in eggs from laying hen farms in other regions of Colombia (7). The use of an egg surface rinse to establish the Salmonella load in the egg surface may constitute a practical procedure to screen large volumes of eggs from laying-hen farms during the process of egg disinfection.

This study identified S. Enteritidis (80%) and S. Paratyphi B (10%) as the main serovars of Salmonella enterica present on the surface of eggs marketed in stores and supermarkets of Ibague city (Table 1). These results are in partial agreement with the results of Rodríguez et al. 015) (17), who found S. Enteritidis (42.85%) and S. Shannon (57.15%) in laying hen farms in this region. The authors isolated most frequently Salmonella from crushed eggshells (57.15%) than from feed (28.57%) or environmental (14.29%) samples (17).  S. Enteritidis was the most prevalent serovar found on the egg surface (80%), and this is the predominant serovar associated to Salmonella outbreaks in several countries. In 2012, S. Enteritidis accounted for 179 outbreaks and 2,177 human cases of salmonellosis in the European Union, most of which were attributed to eggs and egg products (168 outbreaks, 22%) (33). Given the lack of epidemiologic data on Salmonella in the poultry industry of Tolima, the finding of S. Enteritidis in the surface of eggs commercialized in Ibague city may raise concern on the potential transmission to egg consumers and may indicate the need to assess its relationship with clinical cases of salmonellosis in this region.

This study implemented a series of 3 multiplex PCRs designed for typing commonly isolated Salmonella spp. from clinical samples that targeted the serogroups D (615 bp), C1 (483 bp), E (345 bp), B (230 bp) and C2 (154 bp) (24, 25, 26). This PCR-based method identified O:D1 and O:B as the serogroups of Salmonella present in the surface of eggs marketed in Ibague Tolima; and these results correlated with the results obtained by the conventional serotyping (Table 1). The multiplex PCR also identified correctly a number of S. Enteritidis previously isolated from crushed eggshells (17). Thus, the use of multiplex PCR for typing Salmonella spp. present in eggs might be advantageous by the limited number of Salmonella spp. subspecies that survived in this product and might be a useful tool to speed up the identification of Salmonella spp. in poultry products.

The second multiplex PCR for phase 1 flagellar antigens allowed the identification of serovar S. Enteritidis with two amplification products, one of 500 bp that corresponded to the G-complex and a second band of 333 bp specific for this serovar (Fig. 2). The serovar S. Paratyphi B showed the specific 150 bp band described by Herrera-Leon et al. (25). The isolate corresponding to the species S. bongori showed a single band of about 250 bp, similar to the single band amplified from the positive control S. Typhimurium ATCC 14028.

The third multiplex PCR for phase 2 flagellar antigens showed no bands in all S. Enteritidis, according to the findings of Echeita et al. (24). This multiplex also showed a correct identification of S. Paratyphi B with an amplification product of 400 bp and S. Enteritidis ATCC 13076 with no amplification products as described by others (24), whereas S. bongori showed 2 bands (400 bp and 100 bp) that could differentiate this species from S. enterica. However, the multiplex PCRs were developed for the identification of serogroups and serovars mostly present in the subspecies I, and they did not take into account subspecies of S. bongori. The results of PCR-typing indicated that this methodology could identify S. Enteritidis, S. Typhimurium and S. Paratyphi B isolated from eggs commercialized at stores in Ibague city, and the banding pattern of those serovars could differentiate them from the members of the species S. bongori, another Salmonella that could be recovered from the egg surface but lacking association to clinical diseases in animals and human, they are usually present in the environment (reptiles, amphibians and fish) (34).

The epidemiological survey in each store and supermarket revealed that a low frequency (once per week) of cleaning and disinfection of the stores (OR: 19.84, p <0.05), the staff with long nails (OR: 5.468, p <0.05) and the storage of eggs at temperatures less than 25°C (OR 10.05, p <0.05) may be potential risk factors for the presence of Salmonella spp. in eggs marketed in the city of Ibague, Tolima (35). The finding also reflected the importance of strengthening the cleaning and disinfection procedures along the chain of egg production, as it has been observed in the egg packing plants in Europe, where significant differences were found between the prevalence of Salmonella in plants with and without disinfection plans (36). It has also been observed that a high frequency (21.6%) of meat plant operators with long nails was positive to Salmonella in Nariño, Colombia (37).  Finally, at low temperatures, Salmonella may slow down its metabolism and could survive longer on the outer shell surface (38). Other possibility that could explain this finding is that Salmonella may be susceptible to desiccation and dryness of the egg surface at temperatures higher than 25°C (39). 

Concluding, this study estimated for the first time a prevalence of 2.93% Salmonella in the surface of eggs marketed at shops and supermarkets of Ibague city. S. Enteritidis and S. Paratyphi B were the main serotypes of Salmonella circulating in eggs, and the use of 3 independent multiplex PCRs for the typing of Salmonella showed a good correlation with serotyping and then constituted a useful tool to speed up the identification of this pathogen at the species level upon isolation.  The results may support the need to conduct studies with a larger coverage. In addition, they reveal the need to increase awareness of poultry producers, vendors and health authorities on the health risk imposed by Salmonella, and the need to implement appropriate diagnosis and epidemiological surveillance tools to detect and identify Salmonella species potentially contaminating food products in a short time.

ACKNOWLEDGEMENTS

REFERENCES