Introduction

As per the UNICEF data, the global under-five mortality rate had dropped from 93 deaths per 1000 live births in 1990 to 41 in 2016 and India has also recorded 62% reduction in under-five mortality rate since 1990 (Ahmad et al. 2000), even though the proportional mortality for diarrheal diseases still remains high. According to a new Global Burden of Disease published in The Lancet Infectious Diseases journal in September 2017, India and Nigeria are responsible for 42% of child deaths on account of diarrheal diseases. It is well known that the diarrhoeal diseases are chiefly caused by foodborne pathogens such as bacteria, viruses, parasites or chemical substances which are usually infectious or toxic in nature, enter the body through contaminated food or water. These pathogens can cause severe diarrhea or debilitating infections including meningitis. Salmonella, Campylobacter, Listeria monocytogenes and Enterohaemorrhagic Escherichia coli are among the most common foodborne pathogens that affect millions of people annually, sometimes with severe and fatal outcomes (Bhaskar 2017). On the other hand, staphylococcal food poisoning is a gastrointestinal illness caused by eating foods contaminated with toxins of Staphylococcus aureus, a commensal bacterium found on the skin and in the nose of about 25–50% of healthy people and animals (Ouidri 2018).

According to FAO, about 2.5 billion people eat street food every day (FAO 2012). These foods are not only cherished for their unique flavors, convenience and the role they play in the cultural and social heritage of societies. They have also become important and essential for maintaining the nutritional status of the populations (FAO 1997). There are several kinds of vegetarian and non-vegetarian foods (animal origin foods consist of Milk, Kebab, Tikka, Paneer, Kala jamun, Omelet, Lassi, Burfi etc.) available on the roadsides of the Indian cities. On the other hand, the infrastructure facilities of these vendors are relatively limited for potable water, toilets, refrigeration, washing and waste disposal. Hand and dishwashing are usually done in plastic containers due to lack of running water and sometimes without soap (WHO 1996). Due to inadequate infrastructure during preparation and processing, the raw materials (meat, milk, eggs or fishes) are usually contaminated with feces (either of human or animal) and/or with unsafe water (during cleaning and processing), may increase the health risk potential. Thus, a majority of countries reported contamination of these foods with Escherichia coli, Klebsiella pneumonia, Staphylococcus aureus (three agents of greatest concern listed by WHO), Salmonella spp., Listeria monocytogenes, Pseudomonas spp. and Proteus spp. Nevertheless, many of the Indian is likely to have street food frequently because of delicacy, poverty, obligations, and idleness, regardless of the safety of health and life. Therefore, the microbial safety of street foods is questionable where foods and surrounding environment thought to be a source of not only virulent pathogens but also antimicrobial resistant pathogens (Iroha et al. 2011; Chrun et al. 2017). The traceability of food origin is also one of the important factors for food safety (El Sheikha and Xu 2017; El Sheikha 2017). Moreover, the consumption of antibiotics in India is expected to double by 2030 which may increase the burden of antimicrobial resistance (AMR) pathogens. However, improved water quality could significantly reduce the burden of childhood diarrheal diseases, which in turn would reduce antibiotic consumption (Nandi et al. 2017).

Resistant bacteria enter the food during various manufacturing and processing operations and are a potential threat to consumers. The crude infectious disease mortality rate in India is 416.75 per 100,000 persons (Laxminarayan et al. 2017) and there was a steep increase in MRSA from 29 to 47% during 2009–2014 (CDDEP 2015). Emerging reports from India documented the superbugs in food-producing animals (Pruthvishree et al. 2017; Nirupama et al. 2018). According to the Indian Network for Surveillance of Antimicrobial Resistance (INSAR), there is a widespread existence of superbugs throughout the country including a startling 41% of MRSA. The mortality due to S. aureus bacteremia remains approximately 20–40% despite the availability of effective antimicrobials (Mylotte et al. 1987). Several mechanisms play a key role in the development of resistance, for e.g. methicillin resistance in S. aureus is mediated through an altered protein called low-affinity penicillin-binding protein (PBP2a) which is encoded by mecA gene, present in chromosomal mobile genetic element called Staphylococcal cassette chromosome mec (SCCmec) (Matsuhashi et al. 1986; Katayama et al. 2000). Since the patient’s life is in danger, it is necessary to diagnose, treat and manage the MRSA promptly. Detection of methicillin resistance in India is based on cefoxitin and oxacillin disc diffusion methods with limited reports on MIC determination and detection of mecA gene by polymerase chain reaction (PCR) (Bhave et al. 2016; Sharma et al. 2017). In India the street vended foods are not regularly monitored, rather the report on AMR pathogens from street foods are very scanty. Keeping in view, we screened the street vended foods and their environmental samples for virulent MRSA, a pathogen of public health concern.

Materials and methods

A cross-sectional study was carried out from September 2015 to May 2016 in which a total of 430 samples comprising foods of animal origin and associated environmental swabs were collected randomly from different vendors located in Delhi (n = 114 from New Delhi, Old Delhi, Hazrat Nizamuddin and Anand Vihar) and Bareilly (n = 316 from Delapeer, Air Force station, Sahdana, Rajendra Nagar and Izatnagar). For each location, the sample size calculation was carried out using Epitools software (http://epitools.ausvet.com.au/content.php?page=home) with 95% CI, 80% power with Staphylococcus aureus prevalence between 5 and 20% (Chon et al. 2017). In Bareilly region the number of samples were collected marginally higher than the required samples size because of the easy convincing of the street vendors.

For convenience and as per the method of cooking/processing, the samples were categorized as environmental [Hand Swab (HS), Table Swab (TS), Cloth Swab (CS) and Plate Swab (PS)], raw foods (chicken, egg, milk, paneer and fish), ready to eat foods (lassi, rasmalai, burfi, pedha, curd, rasgulla, salad, chutney and masala) and cooked foods (chicken gravy, omelette, cooked fish, boiled egg and boiled milk). Based on availability of the type of samples and willingness of the vendors, 92 environmental swabs, 193 raw, 107 ready to eat and 38 cooked food samples were collected. Swab samples were collected by rubbing on l0 × 10 cm area of the table, plate, cloth and entire surface area of hand (palm) by moistened swabs. The collected swabs were kept in a separate screw-capped tube or test tube containing 10 mL sterile maintenance medium (0.9% NSS + 0.1% peptone) (Vaidya et al. 2007). Approximately 50–100 g each food sample was collected separately in sterile polythene bags. All the samples were transported under cold chain and immediately processed for isolation of S. aureus as per FSSAI (2012) followed by identification by Gram staining and biochemical characterization by catalase and coagulase test (BAM 2001).

Confirmation of S. aureus isolates by PCR

Genomic DNA of Staphylococcus isolates was extracted by snap chill method (Swetha et al. 2015) and analyzed for the presence of nuc gene by species-specific primers (Brakstad et al. 1992). PCR was optimized using different concentrations of reagents and cyclic conditions that affect the sensitivity and specificity. The optimization was carried out with different annealing temperature (as per Tm and TA value of the primer); primer concentration (5–20 pmol); template volume (2–10 µL) and Taq DNA polymerase (1–3 U) (Thermo Scientific, USA). After optimization, the PCR was carried out in 0.2 mL tube containing reaction mixture (25 μL) comprised of 2.5 μL of 10 × Taq buffer with 25 mM MgCl2, 2.5 μL of 2 mM of each dNTP, 10 pmol of each primer (forward and reverse), 1 U of Taq DNA polymerase, 2 μL of DNA template and nuclease-free water to make volume up to 25 μL. The details of primer used, cycling conditions and product size were listed in Table 1. PCR products were analyzed by resolving on 1.5% agarose gel containing ethidium bromide by electrophoresis (Sambrook and Russell 2001) and photographed using a gel doc system (UVP, UK).

Table 1 Primer sequence used for amplifying corresponding genes in PCR assay
Full size table

Isolation and identification of methicillin-resistant S. aureus (MRSA)

The confirmed S. aureus isolates were streaked on MeReSa agar (Hi-Media, India) and incubated at 35–37 °C for 18–48 h. Light pink colored colonies suspected of MRSA were selected for further confirmation and characterized for virulence and AMR genes.

Phenotypic detection of MRSA and beta-lactamase resistance by disc diffusion assay

Antibiotic sensitivity of the confirmed S. aureus isolates was carried out by the Kirby–Bauer disc diffusion technique using Mueller–Hinton agar (CLSI 2013). Briefly, each culture was inoculated into sterile BHI broth (Hi-Media, India) and incubated at 37 °C for overnight. The turbidity of the inoculum was compared with 0.5McFarland standards. Pure broth culture of each isolate was spread on to the Mueller–Hinton agar (Hi-Media, India) supplemented with 2% NaCl and kept for drying. Antibiotic discs viz. oxacillin (1 μg), cefoxitin (30 pg), and penicillin G (10 units) (BD BBL Sensi-Disc, USA) were aseptically placed over the dried agar surface of plates and incubated at 35 °C for 24 h. Inhibition zone diameter of < 21 mm, < 12 mm and < 28 mm were considered as resistant to cefoxitin, oxacillin, and penicillin G, respectively (CLSI 2013).

Genotypic detection of MRSA, beta-lactamase resistance and virulence genes by PCR

The primers and PCR cycling conditions used for amplification of MRSA and beta-lactamase resistance genes (mecA and blaZ) and virulence (coa, clfA, spA, FnbA, and sea) genes were given in Table 1. All the primers were custom synthesized from M/s. Eurofins India Ltd. The procedure for PCR amplification and agar gel electrophoresis was carried out as described earlier.

Enterobacterial repetitive intergenic consensus (ERIC)-PCR

The details of ERIC-PCR primer and cycle conditions were given in Table 1. The amplicons were electrophoresed in 1.5% agarose gel for determining phylogenetic relationship among the isolates by dice coefficient using unweighted pair group method with arithmetic mean (UPGMA) method (UVP, UK). The discriminatory power (D) value was calculated by using an online calculator for discriminatory power (http://insilico.ehu.es/mini_tools/discriminatory_power/).

Statistical analysis

Statistical analysis was done with SPSS version 20.0 on Windows platform. The association between the location, type of samples, number of S. aureus and MRSA was tested by Chi square test with Yates correction/Fisher’s exact (two-tailed).

Results and discussion

Indians have the delicacy and practice of consuming street vended foods, where, poor quality raw foods, unhygienic and inadequate sanitary practices are rampant which may disseminate harmful bacteria. The National Policy for Urban Street Vendors/Hawkers in India stated that street vendors constitute approximately 2% of the population of a metropolis (Bhowmik 2005). It is well known that the foods of animal origin such as chicken meat and raw milk are prone to many bacterial contaminations including MRSA (Rodriguez-Lazaro et al. 2017). Therefore, the present study was aimed to determine the AMR and virulence genes in Staphylococcus aureus, for which a total of 430 street vended foods of animal origin and associated environmental samples were processed for isolation and identification of S. aureus. Of the 430 samples, 52 (52/430, 12.09%) samples were positive for S. aureus by cultural, morphological, biochemical and PCR assay. Relatively higher number of S. aureus was isolated in raw food samples (22.27%), followed by ready to eat foods (4.67%) and environmental samples (4.34%). Methicillin-resistant S. aureus has emerged as a pathogen of global importance due to public health significance and rising antibiotic resistance pattern (Ferreira et al. 2012). It can be detected by a number of methods like bacteriological isolation in differential chromogenic media, disc diffusion agar test with oxacillin or cefoxitin. In the present study, while analyzing 52 S. aureus isolates for methicillin-resistant pattern by disc diffusion assay, it was observed that 13 (25%), 19 (36.5%) and 43 (82.7%) isolates were resistant to cefoxitin, oxacillin, and penicillin G, respectively (Table 3). As per the CLSI (2013) guidelines, if any isolate is resistant to any one of the two antimicrobials namely cefoxitin and oxacillin, it may be considered as methicillin-resistant. Therefore, 42.30% (22/52) isolates were considered as methicillin-resistant. Thus, overall presence of methicillin resistance was observed as 5.11% (22/430) in street vended foods and its associated environment, of which majority were from milk and milk products (16/430, 3.72%). Of the 22 resistant S. aureus isolates, only nine isolates showed characteristic pink color colonies on MeReSa agar. However, to rule out the false positivity in the phenotypic test, it is equally important to confirm these isolates by a molecular method like PCR targeting mecA gene. On genotyping screening, the 9 isolates which showed characteristic growth on MeReSa agar harbored mecA gene. Thirteen out of 22 MRSA (detected by cefoxitin disc diffusion method) isolates did not carry mecA gene, indicated existence of some other mechanism of resistance, either by mecC gene or unknown. Adhikari et al. (2017) also noted that the mecA gene was absent in 7 of the MRSA isolates detected by the cefoxitin disc diffusion method. Further, the mecA gene was absent in methicillin- susceptible S. aureus (MSSA) isolates. Presence of Methicillin-sensitive and Methicillin-resistant S. aureus in foods of animal origin sold at the street was also reported by Lozano et al. (2016). Even in India, S. aureus was reported from street vended foods sold at Gangtok, Nainital and Tumkur cities (Kharel et al. 2016; Sudeep Kumar et al. 2017). However, the prevalence of MRSA in these foods has not been studied in detail. Recent reports suggest that human clinical samples have a higher prevalence of MRSA i.e. 53.74% and 76.75% from Moradabad and Jaipur, India (Kumar and Bhadauria 2016; Gupta and Sinha 2017).

Resistance to penicillin G can be considered as beta-lactamase resistance in S. aureus (CLSI 2013). The detection of the blaZ gene by PCR is equally important to know the presence of beta-lactamase producing gene. Of the 82.69% (43/52) phenotypic beta-lactam resistant isolates, 69.23% (36/52) isolates harbored blaZ gene. Among 9 MRSA isolates, phenotypic and genotypic beta-lactam resistance was detected in 8 and 7 isolates, respectively. Beta-lactamase production was also observed by Adhikari et al. (2017) in 71.82% of the S. aureus isolates.

The nuc and coa genes are considered as important virulent determinants of S. aureus (Chesneau et al. 1993; Da Silva and Da Silva 2005). In this study, all the 52 S. aureus isolates harbored nuc gene (100%) and 50 isolates carried coa gene (96.15%). Similar findings have been reported by Salem-Bekhit et al. (2010) and Momtaz et al. (2013). The Staphylococcal protein A (spA), fibronectin binding proteins A and B (FnbA and FnbB), collagen-binding protein, and clumping factor (clf) A and B proteins are necessary for attachment of S. aureus to the host cell surface to initiate the colonization (Foster and Hook 1998). In the present study, 30.76% (16/52) isolates harbored clfA; 55.76% (29/52) spA; 50% (26/52) FnbA and 7.69% (4/52) entA (sea) genes. Similarly, Momtaz et al. (2013) reported 76.92% isolates with clumping factor A and 26.82% isolates with spA gene. In chicken and bovine mastitis samples, 4.5–25.5% isolates carried sea gene (Hwang et al. 2010; Mashouf et al. 2015). In contrast, Kumar et al. (2011) reported FnbA gene in all the isolates from mastitis milk samples while, Kalorey et al. (2007) could not detect any of the isolate carrying sea gene among 37 Staphylococcal isolates screened in Nagpur, India. Surprisingly, all the MRSA isolates recovered in the present study was from raw food samples and the majority of them were from milk and milk products (16/22, 72.72%) (Tables 2, 3). It was also observed that egg, meat, and their products were less contaminated by S. aureus as compared to milk and its products.

Table 2 Details of S. aureus isolates from different street vended food samples collected from Delhi and Bareilly
Full size table
Table 3 Antimicrobial resistance and virulence profile of S. aureus isolates from street vended foods of animal origin and associated environment
Full size table

The raw food items that showed high isolation rates from Delhi were raw milk (50%); raw chicken (21.42%); paneer (14.28%) and raw egg (5.88%) with the overall isolation rate of 19.35%. In Bareilly, items that revealed high isolation rates were raw milk (43.58%), channa (40.0%), rasmalai (40.0%), curd (33.33%), raw chicken (28.0%), burfi (20.0%), paneer (19.23%), hand swab (17.64%), masala (12.5%) and cloth swab (6.25%) with an overall of 12.65% (40/316). The food item wise analysis for presence of S. aureus and MRSA isolates in Delhi and Bareilly region showed that raw foods of Bareilly region carried significantly (p < 0.05) higher number of S. aureus isolates than environment, ready to eat and cooked foods. Between the two places there was no significant difference (p > 0.05) in the presence of S. aureus and MRSA in different food samples (Table 2). The isolation rate from milk vendors, meat vendors, and egg vendors were 17.7%, 13.6%, and 2.12%, respectively. The present study revealed the highest isolation rate of S. aureus in milk samples, and similar findings were reported by Kamal et al. (2013) and Kalorey et al. (2007). This might be due to well adaptation of S. aureus with the udder tissue and cause of mastitis. Further, this pathogen is sturdy in the environment and could contaminate the surfaces and hands of street vendors, especially the milk vendors. There are only a few studies on S. aureus isolation from street vended foods from India and the studies carried out in Silchar city, Assam by Sharma and Mazumdar (2014) revealed that 14.2% street vended food samples were positive for S. aureus while in Gangtok and Nainital, it was detected in 19.5% and 33% street vended food samples (Kharel et al. 2016). Various workers reported that between 2.42% and 69.74% restaurant/street food samples harbored MRSA (Rhee and Woo 2010; Rizek et al. 2011; Ranjbar et al. 2017), thus responsible for health implications to the consumer. In another study, enterotoxigenic S. aureus was detected in 91 (60%) samples of coriander sauce, 87 (58%) samples of coconut slices and 129 (86%) samples of ready-to-eat salads in New Delhi and Patiala City (Ghosh et al. 2007). In contrast to our study, Sharma and Mazumdar (2014) and Ghosh et al. (2007) reported a high isolation rate of S. aureus.

All the 52 isolates were characterized by ERIC PCR to determine the genetic diversity and phylogenetic relationship among the isolates. All the isolates were typeable by ERIC PCR. The Dendrogram analysis of 52 S. aureus isolates revealed 8 distinct types/clades with discriminatory power (D value) of 0.77. These isolates formed two main clusters (C and D) with the heterogeneity of 43.4%. The two main clusters further divided into sub-clusters and formed 8 clades (D3, D4, D5, D7, D8, C1, C3, and C4) (Fig. 1). A low D value was observed in this study compared to Ye et al. (2012) who reported that ERIC-PCR classified 35 S. aureus isolates into 28 ERIC types with a D value of 0.98. Only 12 isolates were grouped in D clade while remaining all in clade C in which 11 isolates (from one cloth swab, three paneer, one hand swab, one raw egg and five raw milk) formed a subgroup C1 and five on them were MRSA. It is surprising to note that, of the 12 isolates grouped in clade D, all the isolates were either sensitive or resistant to both cefoxitin and oxacillin except the 3 isolates (two from raw chicken and one from raw milk) which were either resistant to one. Remaining 10 phenotypically confirmed MRSA isolates were grouped in C2 and C3 clade. The similarities between the ERIC profiles among isolates from diverse sources as identified in clades C1, C2 and C3 indicate that ERIC PCR fingerprints were effective in differentiating the isolates from various sources. Further, 100% type ability of ERIC PCR reaffirms the fact that this technique is very reliable in genotyping of isolates and hence is a useful tool in food microbiology. The ERIC-PCR has been reported as an effective tool in typing S. aureus isolates from various sources including animals (Arslan and Mtulu 2016).

Fig. 1
figure 1

Phylogenetic analysis of S. aureus isolates isolated from street vended foods of animal origin and associated environment

Full size image

Conclusion

The present study revealed that raw foods, environmental samples and ready to eat food samples were a major source of S. aureus and MRSA which is a public health concern. Though all the environmental swabs and many ready-to-eat foods were contaminated, all cooked food samples were free from S. aureus and MRSA. The ERIC-PCR analysis showed relatedness between the isolates from different sources indicated that a common source of contamination and warrants proper hygiene measures. There is also an urgent need for continuous surveillance of street vended foods to tackle the threat of antibiotic resistance.