Microbial Quality, Safety, and Pathogen Detection by Using Quantitative PCR of Raw Salad Vegetables Sold in Dhanbad City, India

Fresh vegetables are an essential part of the healthy diet of people in all the economic classes in India. Several studies have suggested that consumption of fresh vegetables and fruits provides high concentrations of vitamins and minerals, reduces mortality from cardiovascular and cerebrovascular diseases, and has benefits toward cancer prevention (3, 28, 51). The microbial quality of vegetables can depend on many factors, including the use of contaminated irrigation or process water, the use of manure for fertilization, poor equipment sanitation, and poor worker hygiene (24, 26, 31). Irrigation water is considered a potential source of preharvest pathogen contamination of vegetables (50). Recently, disease caused by foodborne pathogens has become an important public health problem that is resulting in significant morbidity and mortality (17, 52). Many reports specified that raw vegetables may harbor various pathogenic bacteria such as Salmonella spp., Listeria monocytogenes, and Escherichia coli that are involved in large foodborne outbreaks worldwide, causing symptoms of gastroenteritis and even chronic infections (29, 33, 40). Vegetables that have come into direct contact with the soil may have surface contamination by L. monocytogenes, Salmonella spp., and E. coli. These bacteria may survive the washing and sanitizing steps by forming biofilms on the surface of the vegetables or via protection within the cuticle of the vegetable (41).

The documented worldwide presence of foodborne pathogens in raw vegetables has given rise to public health concerns. However, information on the microbiological quality of salad vegetables, especially with respect to incidence of pathogens, is limited in Dhanbad, India. The main objective of this study was to assess the microbiological quality of salad vegetables consumed in Dhanbad to guide future improvement of food safety measures.

Sample collection. In total, 480 samples of fresh salad vegetables (n = 60 of each) including cucumber, tomato, carrot, coriander, cabbage, beetroot, radish, and spinach were collected from five retail markets during their production seasons (between July 2013 and June 2014) in Dhanbad. Samples of the vegetables from retailers were purchased at different times and brought to the laboratory for analyses. Samples from the leafy vegetables spinach and coriander were obtained in small bunches, whereas carrots, tomatoes, cucumber, beetroot, and radish were evaluated intact. The samples were transported at 25°C to the laboratory for analysis in sterile bags and analyzed within 24 h. Before a sample was taken out of its bag, the surface of the packaging was carefully sterilized using 70% ethanol swabs to prevent cross-contamination. Samples that had been damaged or visibly compromised before analysis were discarded. All samples were washed aseptically with sterile distilled water in a laminar air flow chamber before sample preparation.

Sample preparation. Twenty-five grams of each sample (fresh weight) of raw vegetables was diluted in 225 ml of buffered peptone water and aseptically homogenized in a stomacher.

Microbiological analysis. Total plate count, total coliforms, and E. coli were determined using the standard methodologies. Four 10-fold dilutions were made for each sample. One milliliter of each step dilution was inoculated into duplicate plates and analyzed. The colonies formed on the plates were counted and expressed as log CFU per gram.

E. coli O157:H7 detection and identification were determined by the International Organization for Standardization (ISO) 16654 (21) reference method. Twenty-five grams of each sample was diluted in 225 ml of modified tryptic soy broth (Oxoid, Basingstoke, UK) with novobiocin (Himedia, Mumbai, India), and then the sample was homogenized for 2 min at 260 rpm with a Stomacher 400 (LabFriend, New Delhi, India). Next, the sample was incubated for 24 h at 41.5°C according to the ISO 16654 (21) reference method, as well as the remaining steps. After enrichment, the selective and differential isolation of E. coli O157:H7 was streaked onto tellurite cefixime sorbitol MacConkey agar (Oxoid).

Isolation and identification of Salmonella spp. were done according to ISO 6579 (22) and performed using the homogenate in buffered peptone water. Quantities of 1 ml of the buffered peptone water were inoculated into tetrathionate broth with novobiocin and Rappaport-Vassiliadis broth (Himedia). The enrichment broths were incubated for 24 ± 2 h at 37 ± 1°C for tetrathionate broth and 42°C for Rappaport-Vassiliadis broth. The positive cultures were streaked onto xylose lysine desoxycholate Salmonella agar (Himedia) at 37 ± 1°C for 24 h, and the confirmation was done using an API 20E kit (bioMeriéux, Inc., Marcy l'Etoile, France) (42, 45).

L. monocytogenes was detected according to the ISO 11290-1 (19). Samples (25 g) were weighed into sterile stomacher bags, diluted, and homogenized with 225 ml of Fraser broth (Himedia). After homogenizing and preculturing at 37 ± 1°C for 48 ± 2 h, the positive broth was streaked onto Listeria PALCAM agar (Himedia) and incubated at 37 ± 1°C for 24 ± 2 h. Characteristic colonies were Gram stained, tested for motility, and tested for oxidase and catalase activity, followed by identification with the API Listeria system (bioMérieux, Inc.).

Detection of pathogens. Detection of Salmonella spp., L. monocytogenes, and E. coli O157:H7 was confirmed by real-time quantitative PCR after selective enrichment steps. The analyses were performed according to ISO 7579 (20). Samples were homogenized in buffered peptone water, incubated at 37°C for 24 h (preenrichment), and subsequently enriched in broth media, as described in “Microbiological analysis.” After these enrichment steps, 100-μl aliquots of the samples were diluted 1:10 (vol/vol) in sterile distilled water and kept at −20°C for quantitative PCR; 500-μl aliquots were mixed with 500 μl of 400 g liter⁻¹ sterile glycerol and kept at −20°C for use in pathogen detection from quantitative PCR–positive samples. For all tested raw vegetables, the detection limit was the same as using the standard ISO protocol. A method targeting the invasion gene invA (primers Strinva-JHO-2F/R and Strinva-JHO-2p TaqMan probe [FAM and TAMRA dual labeled]) previously reported by Hoorfar et al. (18) was used to detect Salmonella spp., whereas primers stx1-F and GLT were used for E. coli O157:H7 and L. monocytogenes, respectively. Uniplex reactions were performed using a PCR Core Reagents kit (Applied Biosystems–Roche Molecular Systems Inc., Branchburg, Germany). Reactions were run on an ABI PRISM 7900HT sequence detection system (Thermo Fisher). All samples were analyzed in triplicate. Both the negative (distilled water plus all PCR components) and positive (a DNA template plus all PCR components) controls were included in each run.

Legislation. The microbial quality of the food was evaluated according to the specifications of the Indian Food Safety and Standards Act 2006 of the Food Safety and Standards Authority of India (9). According the act, a microorganism of concern is not permitted in 25 g of food.

Quality control and quality assurance. Analytical grade chemicals and culture media were used for quantitative PCR and microbiological analyses. Double-distilled deionized water and Milli-Q water (Millipore, Billerica, MA) were used for preparation of all reagents and calibration standards. Calibrated glassware was used for experimental work. To avoid other microbial contamination, special care was taken to transfer the samples from the sampling site to the laboratory.

The results of total plate count, total coliforms, and pathogens are presented in Tables 1 through 3. The highest microbial counts were associated with spinach, beetroot, and tomato (7.1, 5.8, and 5.2 log CFU/g for total plate count; 5.7, 5.0, and 4.5 log CFU/g for total coliforms), and these counts exceeded the adaptation limit per Indian Food Safety and Standards Act 2006 of Food Safety and Standards Authority of India. A total plate count >5.0 log CFU/g was found in 224 (46.7%) of the samples (Fig. 1), indicating that 46.7% of the samples were unacceptable for consumption in India (10) and several others countries (2, 4, 6, 7, 9, 14). Numerous studies reported high aerobic counts on salad vegetables (5, 13, 25, 33, 41). This corroborates findings by Pingulkar et al. (34) for which the overall microflora count on beetroot and cabbage was 6.8 and 6.7 log CFU/g, respectively, both higher than the counts in the present study. In contrast to the present study, Tzschoppe et al. (47) noted the mesophilic counts of salad and sprouts were between 6.4 and 9.4 log CFU/g, with a median of 7.6 log CFU/g. Correspondingly, Garg et al. (13) reported high mesophilic mean counts for cabbage (6.1 log CFU/g), carrot (5.6 log CFU/g), and spinach (5.9 log CFU/g) samples. Similarly, in Singapore Seow et al. (42) reported that the mean value of mesophilic microorganisms on tomato was 4.2 log CFU/g, with a range of 2.4 to 5.5 log CFU/g. Our study produced similar results with those reported for fresh organic vegetables and ready-to-eat salad (32, 34). With an exception to spinach, beetroot, tomato, and cabbage, almost all other vegetables had a similar mean count between 4.3 and 4.8 log CFU/g, with a range of 2.3 to 8.0 log CFU/g. There are a few possible explanations to account for the relatively high mean counts on spinach (mean of 7.1 log CFU/g), one of which might be due to its large surface area. The open leaves may also be in contact with soil and irrigation water, making them more susceptible to bacterial contamination (1).

In the present study, total coliform counts (Table 2) were detected in the majority of samples (96.7%), with the range 0.0 to 7.8 log CFU/g. All vegetables had mean coliform counts ranging from 2.1 to 5.7 log CFU/g, and 70.2% of samples recorded counts <5.0 log CFU/g. These results are similar with many investigations where most of the salad vegetables showed coliform counts <6 log CFU/g (5, 32, 41, 42, 49). The frequency (0.0 to 3.9 log CFU/g) of total coliforms in cabbage was very low, with 100% of cabbage samples containing <4.0 log CFU/g and the mean count noted as 2.1 log CFU/g. These findings are very similar to those reported by Viswanathan and Kaur (49) who found a total coliform mean count of 2.1 log CFU/g.

E. coli is frequently of fecal origin, and it can be used as indicator microorganisms of fecal contamination (39). In this present study, E. coli was detected in 11.5% of all samples; however, three samples (two spinach and one beetroot) contained E. coli O157:H7. None of the tomato, cucumber, carrot, coriander, radish, and cabbage analyzed was found positive for E. coli O157:H7 (Table 3). Farm animals are the major reservoir of E. coli O157:H7 (16), and application of E. coli O157:H7–contaminated manure and water to a production field may result in contamination of the vegetables in that field. A study conducted by Singh et al. (44) showed that water used by vegetable venders for washing and sprinkling purposes was the potential source of pathogenic microorganisms. However E. coli O157:H7 was not detected in studies conducted in Ireland (29), Norway (23), the United Kingdom (37), Hong Kong (8), and the United States (30, 48). Furthermore, in the present study E. coli was not detected in cabbage. Our results are consistent with a study by Viswanathan and Kaur (49), but they reported the presence of E. coli in salad vegetables. In contrast to the present study, García-Villanova et al. (11, 12) reported 86.1 and 75% of vegetables were contaminated with E. coli. The most likely reason for the presence of E. coli on the surface of vegetables are sprinkling of gray water, growing the vegetables on contaminated soil, and irrigation with contaminated water (10, 15, 37, 43, 46). In the present study, 3.75% of samples were contaminated with pathogenic microorganisms. L. monocytogenes (14 of 480 samples), Salmonella spp. (16 of 480 samples), and E. coli O157:H7 (3 of 480 samples) were detected from the surface of tomato, spinach, cucumber, carrot, beetroot, and radish. None of the cabbage and coriander samples were found to harbor L. monocytogenes, Salmonella spp., or E. coli O157:H7 (Table 3). In a previous study conducted in Dhanbad by Kumar (26), Salmonella spp. were detected in 7.8% of the analyzed samples. This study also reported 20% spinach contaminated with Salmonella spp., a value that is slightly higher than that of the present study. Similarly, Singh et al. (44) reported the prevalence of Salmonella spp. in vegetables in cities in northern India. The prevalence of L. monocytogenes in the salad vegetables analyzed was 2.9% (14 of 480) and was in accordance with previous results reported by Porto and Eiroa (36), who found 3.2% of samples positive for L. monocytogenes. Lin et al. (27) also reported low (1.6%) incidence of L. monocytogenes in salad samples. However, conflicting results were reported by Ponniah et al. (35) in a study in Malaysia, where L. monocytogenes was detected in 22.5% of minimally processed vegetables samples.

The present results are comparable to those of previous studies, indicating the need to adopt hygienic practices by food vendors, processors, and consumers to minimize the risks of transmission of pathogens. Water used for irrigation or other uses is also an issue of concern. However, the potential hazard of pathogenic bacteria present in vegetables should not be underestimated, particularly for those that are eaten raw. The most plausible causes of vegetables serving as vehicles of the pathogens might be via irrigating them with contaminated water, growing them on contaminated soil, or washing them with contaminated water. The data obtained in the study may be used by the government agencies for quantitative risk assessments.

The authors thank the Department of Environmental Science and Engineering, Indian School of Mines, Dhanbad, India, for providing necessary support to conduct this research.