Prevalence and antimicrobial susceptibility of non-typhoidal Salmonella (NTS) from salad vegetables at farms and retail markets in Terengganu, Malaysia

Haslinda, W.H., *Tang, J.Y.H., Tuan Zainazor, T.C., Mohd Khairi Hilman, A.L., Wan Norezah, W.M., Irdawaty, T. and Noor Hafizatulakmal, H. Faculty of Bioresources and Food Industry, Universiti Sultan Zainal Abidin, 22200 Besut, Terengganu, Malaysia Food Safety and Quality Laboratory, Terengganu State Health Department, Kg. Bukit Tunggal, 21200 Kuala Nerus, Terengganu, Malaysia Faculty of Fisheries and Food Science, Universiti Malaysia Terengganu, 21030 Kuala Nerus, Terengganu, Malaysia


Introduction
The government of Malaysia through the Ministry of Agriculture and Food Industry has highlighted the importance of salad vegetables (locally known as ulam) consumption among Malaysian due to their nutritional value (Roshila-Murni, 2020) such as vitamins, minerals, dietary fibre and polyphenolic compounds (Sebastian et al., 2019). Salad vegetables such as ulam raja (cosmos), selom (water parsley), pegaga (pennywort), timun (cucumber), kacang botor (winged bean) and bayam brazil (sissoo spinach) are commonly available at retail markets all over the country. The recommended daily intake of vegetables for Malaysians is at least three servings which are equal to 240 g (Malaysian Dietary Despite the health benefits, salad vegetables have also been associated with contamination of enteric pathogens such as Salmonella spp., Listeria monocytogenes, thermophilic Campylobater spp. and enterohaemorrhagic Escherichia coli (Khalid et al., 2015;Tang et al., 2016;Koukkidis and Freestone, 2018;Mohd Noor et al., 2020). Pathogens contaminate vegetables via direct or indirect contact with humans, livestock, wild animals, manure and irrigation water (Mir et al., 2018). Thus, the microbial safety of salad vegetables is of concern as these foods are consumed raw. Salad vegetables have been recognized as one of the most common sources of foodborne illness (Crowe et al., 2015). Several reports indicated that foodborne outbreaks associated with fresh vegetables were mostly caused by non-typhoidal Salmonella (NTS) . According to the European Food Safety Authority (EFSA), ready-to-eat leafy vegetables were ranked in the first place in terms of outbreak occurrence risk due to the presence of Salmonella (McLauchlin, 2014).
NTS is a major foodborne pathogen causing diarrhoeal illness (Chang et al., 2020). Each year, NTS causes approximately 93.8 million cases of acute gastroenteritis and 155,000 human deaths worldwide (Chang et al., 2020). Gastroenteritis caused by NTS is commonly mild, self-limiting and does not require treatment, however, antibiotic therapy is necessary for severe and invasive infections such as bacteremia, septic arthritis, and meningitis mostly in children and the immunocompromised (Chang et al., 2020;Yang et al., 2020).
During the last few years, the prevalence of antibiotic-resistant and multidrug-resistant (MDR) Salmonella isolated from fresh vegetables was increased in Malaysia (Abatcha et al., 2018;Thung et al., 2020) as well as in other countries worldwide (Abatcha et al., 2020;Yang et al., 2020). A review by Abatcha et al. (2020) involving reports from several countries found relatively high resistances among Salmonella from raw vegetables towards erythromycin (82.3%), furazolidone (62.9%), streptomycin (49.6%), cefoperazone (48.6%), kanamycin (48.5%), cephalothin (44.7%) and amoxicillin-clavulanic acid (37.8%). Of particular concern is the probability of Salmonella strains to show resistance toward medically important classes of antibiotics such as extended-spectrum cephalosporins, fluoroquinolones, macrolides, aminoglycosides and penicillins which might compromise the treatment of complicated infections (Scott et al., 2019). Therefore, this study was conducted to determine the prevalence of Salmonella in salad vegetables from farms and retail outlets in Terengganu, Malaysia and their antibiotic susceptibility patterns. The findings in this study were expected to describe the role of salad vegetables as a vehicle of Salmonella transmission with their MDR characteristic in the food chain.

Sample collection
A total of 666 samples were collected from farms (270), wet markets (216) and supermarkets (180)  Descriptions of the samples were summarized in Table 1. Sample of approximately 250 g was aseptically transferred into sterile bags (3M TM plain sample bag) and transported to the laboratory in an icebox to maintain a temperature of 1-4°C.

Salmonella isolation and identification
Isolation of Salmonella from samples was carried out according to the International Standard protocol ISO 6579:2002(E) (ISO, 2002). For the first enrichment, 25 g of sample was weighed in a sterile stomacher bag and mixed thoroughly with 225 g of buffered peptone water (BPW) (Merck, 107228, Germany) prior to incubation at 37±1°C for 18±2 hr. Then, 0.1 mL and 1 mL of the preenriched cultures were incubated in 10 mL of Rappaport-Vassiliadis medium with soy (RVS) (Oxoid, CM0669, UK) at 41.5±1°C for 24±3 hrs and 10 mL of Muller-Kauffmann Tetrathionate novobiocin (MKTTn) broth (Merck, 105878, Germany)

Results
All farms involved in this study were operated traditionally in a controlled environment with proper management. Their farming areas were well separated from other nearby activities with appropriate fencing. The farmers used treated animal waste manure for fertilization and untreated water for irrigation. Their preference for untreated water is due to cost-saving and non-chlorinated properties as the chlorine toxicity to plants is a concern for them. The overall characteristics of each farm were summarized in Table 2.
The isolation of Salmonella in this study was presented in Table 3. Overall, Salmonella was found in 8 (2.96%), 44 (20.37%) and 6 (3.33%) samples collected from farms, wet markets and supermarkets, respectively. The wet market showed the highest incidence of Salmonella compared to the supermarket and farm, all varieties of salad vegetables from wet markets were contaminated at different percentages with the most contaminated were selom and pegaga (36.11%). Among samples taken at the farm, Salmonella was mostly found in irrigation water (11.11% of water samples tested). In soil samples, the percentage of Salmonella detection was 2.22%. Interestingly, Salmonella was not found in all manure samples tested. The distribution of Salmonella serovars isolated in this study was shown in Table 4. Of 58 isolates, 23 nontyphoidal Salmonella (NTS) serovars were identified, the most predominant serovar was S. enterica ser. Weltevreden (20.69%), followed by S. enterica ser. Albany (10.34%), S. enterica ser. Hvittingfoss (8.62%) and S. enterica ser. Aberdeen (6.90%).

Discussion
Finding on the prevalence of Salmonella in this study was in line with the previous study which reported a relatively high incidence of Salmonella in salad vegetables from the wet market including 21.5% in the northern and middle region of Malaysia (Abatcha et al., 2018), 12.9% in Mekong Delta, Vietnam (Nguyen et al., 2021) and 17.59% in Hue City, Vietnam (Chau et al., 2014). Cross-contamination through improper handling and poor hygienic practices might play a significant role as a source of Salmonella contamination to salad vegetables in the wet market, in addition, with the open display under ambient temperature, growth and multiplication of pathogenic bacteria including Salmonella will be promoted. In Malaysia, Nidaullah et al. (2017) reported consistent contamination of Salmonella (86.18% detection) in the environmental samples collected along the wet market chicken processing line, suggesting that poultry from the wet market can be an important vehicle for the transmission of Salmonella. Since selling locations for poultry and vegetables are adjacent in the wet market, Salmonella might be transferred to the salad vegetables through unhygienic and poor sanitation practices among retailers and consumers. Rodents that inhabit the wet markets could also be considered as the reservoirs and transmitters of Salmonella to salad vegetables. A study by Ribas et al. (2016) reported 49.10% prevalence of Salmonella in rats from wet markets in Thailand, they have belonged to three serovars: S. enterica ser. Typhimurium (30%), S. enterica ser. Weltevreden (12.7%) and S. enterica ser. 4,[5],12:i:-(6.4%). This finding indicates that rodents could act as a potential transmitter of Salmonella due to the contact they have with food products stored and sold in wet markets.
Overall occurrence of Salmonella in salad vegetables from a supermarket in this study was low, comparable to other studies in Malaysia (3.0%) and Turkey (3.38%) (Buyukunal et al., 2015;Saw et al., 2020). A supermarket is an organized grocery store selling a variety of goods. The goods were systematically arranged in different zones under specific conditions  depending on the type of goods to avoid crosscontamination. The environment in a supermarket is commonly clean, tidy and cool with proper maintenance, these conditions are required to maintain the quality of goods, especially food items (Saw et al., 2020). All these circumstances might contribute to the low incidence of Salmonella in salad vegetables sold in a supermarket.
In this study, only two (1.48%) Salmonella were detected from vegetable samples collected from the farm, this finding could be explained by the probability of low survivability of Salmonella on the vegetables during cultivation due to direct exposure to sunlight. According to Strawn, Gröhn, Warchocki et al. (2013), pathogens that present in soil are likely to contaminate vegetables, however, the presence of pathogens will decrease over time after cultivation, due to environmental conditions exposure such as ultraviolet light that reduces pathogen loads. Furthermore, the use of pesticides might be one of the causes of the low detection of Salmonella in farm vegetables, the pesticides have been shown to inhibit growth and kill soil microorganisms (Filimon et al., 2015). This finding was further evidenced by a study by Ottesen et al. (2015) which observed lower counts of Salmonella on tomato fruits and leaves that were regularly sprayed with pesticides as compared to controls that were not exposed to pesticides.
The prevalence of Salmonella from irrigation water in vegetable farms was also demonstrated by several other studies such as 12.5% in New York, USA, (Strawn, Gröhn, Warchocki et al., 2013), 7.7% in California, USA (Benjamin et al., 2013) and 3.4% in Kano, Nigeria (Abakpa et al., 2015). Through observation during the sampling process, all vegetable farmers involved in this study were found using irrigation pumps to water their crops from wells and streams (Table 3), sprinkler irrigation is the most common form of irrigation practised in the farms. Due to the usage of untreated water which is found to have Salmonella, vegetables might be contaminated through irrigation. Adetunde et al. (2015) revealed a strong positive correlation between the microbial counts on vegetables and in irrigation water, which makes it one of the main sources of vegetable contamination. Furthermore, the sprinkler method of irrigation used also contributed to the higher transfer and retention of bacteria on the edible portion of vegetables (Gupta and Madramootoo, 2017). Contamination of pathogens into streams and wells is commonly attributable to household or small industries wastewater (Keraita et al., 2014) and wild or domestic animals (Adetunde et al., 2015).
The detection of Salmonella in soil samples collected from farms was in agreement with other studies in Kano, Nigeria (2.5%) and New York, USA (2.0%) (Strawn, Fortes, Bihn et al., 2013;Abakpa et al., 2015). Salmonella may enter the soil environment from various sources such as through contaminated water, manure, livestock and wildlife (Alegbeleye et al., 2018), their survivability in the soil is influenced by various factors such as temperature, moisture, soil type, presence of plants, pH, nutrient availability, exposure to sun (ultraviolet) light and protozoan predation (Jacobsen and Bech, 2012). With regard to this study, the most probable source of Salmonella contamination into the soil is from untreated water used for irrigation, based on the prevalence of Salmonella in water samples analyzed. This finding could be related to a study by Leifert et al. (2008) which demonstrated the exposure of water through precipitation and irrigation as one of the most critical factors influencing microbial transport and survival in soil.
Our results suggest that the application of treated manure to fields can significantly decrease the risk of Salmonella contamination as none of the manure samples tested in this study were contaminated with Salmonella. All manure samples taken for this study were in the form of processed pellets of chicken faeces. The manure management practices, with regards to the collection, transport, handling, treatment, disposal and utilization might affect the survival of foodborne pathogens. Treatment methods for manure such as thermal, chemical, biological and physical are commonly used, all these types of treatment may reduce pathogen loads in manure before releasing into the environment (Alegbeleye and Sant'Ana, 2020). Different farms handle manure using various types of systems and their combinations, each treatment is capable of reducing pathogen loads in manure, even though the efficiency of different methods may vary (Millner, 2014).
S. enterica ser. Weltevreden was also reported as the most found serovar in salad vegetables from other studies in Vietnam (Nguyen et al., 2021), Malaysia (Abatcha et al., 2018) and Indonesia (Kusumaningrum et al., 2012). This serovar has been identified as the common pathogen associated with human salmonellosis in the South and South-East Asian regions (Gunasena and De Silva, 2021). S. enterica ser. Weltevreden was recognized for the first time to cause a salmonellosis outbreak related to plant products in Scandinavia (Norway, Denmark and Finland), this outbreak occurred in 2007 involving 45 cases resulting from consumption of contaminated alfalfa sprouts (Gunasena and De Silva, 2021). The capability of S. Weltevreden to grow and survive in a plant is possibly due to the presence of additional carbohydrate metabolism clusters in their genomes as compared to other serovars, which enable eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER this serovar to metabolize carbohydrates better in plant tissue (Brankatschk et al., 2012). Brankatschk et al. (2012) also demonstrated the presence of genes for myoinositol utilization in S. enterica ser. Weltevreden isolates tested, the capability of this serovar to grow on stereoisomers of myo-inositol might also contribute to their survival on plants as this carbohydrate is commonly present in soil and the phosphorylated form of myoinositol is stored in plant tissue. The presence of these types of genes in S. enterica ser. Weltevreden might be considered as the reason for their capability to become prevalent in salad vegetable samples involved in this study.
A study on the prevalence of Salmonella in salad vegetables from Peninsular Malaysia by Abatcha et al. (2018) found a relatively comparable percentage of S. enterica ser. Albany (4.60%), S. enterica ser. Hvittingfoss (5.75%) and S. enterica ser. Aberdeen (4.60%) with this present study. These serovars are rarely associated with salmonellosis outbreaks, however, in 2016, S. enterica ser. Hvittingfoss has been reported to cause outbreaks associated with fresh produce (cantaloupes) in Australia (Flynn, 2016).
On the other hand, S. enterica ser. Newport is one of the most important NTS serovars that has been associated with several major outbreaks due to consumption of raw vegetables such as alfalfa sprouts, tomatoes, lettuce and cucumber (El-Dougdoug et al., 2019;Yang et al., 2020). Latest in 2020, a multistate outbreak of S. enterica ser. Newport infections linked to onions affecting 1127 people from 48 states in the USA has been reported (CDC, 2020). This serovar has previously been listed among the three highest ranks of Salmonella serovars associated with foodborne outbreaks in the USA since 1970 (CDC, 2013). Furthermore, the Public Health Agency of Canada has listed S. enterica ser. Newport in the ten most commonly reported Salmonella serotypes in Canada accounting for human illness in 2018 (Government of Canada, 2020). In Malaysia, there is no published report of a foodborne outbreak associated with S. enterica ser. Newport so far. However, S. enterica ser. Newport was present in salad vegetables tested in this study and also in a study by Abatcha et al. (2018) with 3.45% and 2.30% detection, respectively. This indicates the probability for the occurrence of salmonellosis associated with salad vegetables contaminated with S. Newport in Malaysia.
Multiple types of other serovars were also found among samples tested (Table 4), the diversity of these serovars might have been influenced by the different sampling locations that were exposed to different sources of contamination (different farms, transports, retail outlets or retailers).
A few reports were published concerning the antibiotic-resistant Salmonella associated with vegetables in Malaysia and other Southeast Asian countries (Abatcha et al., 2018;Nguyen et al., 2021). Among the isolates recovered in this study, resistance to ampicillin, tetracycline and trimethoprimsulfamethoxazole occurred most frequently, these results are consistent with the report from a study of resistance among Salmonella isolates from raw vegetables in China . It is noteworthy that the highest resistance is toward ampicillin as this antibiotic is categorized as the critically important antibiotic with high priority in treating serious bacterial infections in people (Scott et al., 2019;WHO, 2019). Resistance toward tetracycline, doxycycline, trimethoprimsulfamethoxazole and chloramphenicol could also be of concern since these antibiotics are categorized as highly important for human medicine by the WHO (2019). All the resistant NTS found in this study were isolated from wet markets and supermarkets, thus, the most probable source of antibiotic resistance is from direct contamination of resistant isolates from humans during harvesting and handling of salad vegetables due to insufficient hygiene measures. Similarly, Mesbah Zekar et al. (2017) demonstrated that raw eaten vegetables at the market constitute a reservoir of resistant and MDR bacteria that might be reflected from improper hygiene practices during handling.
Attention should be given to the presence of ertapenem resistant S. enterica ser. Albany since carbapenems are the last line of antibiotics used to treat human infections caused by MDR bacteria (Monte et al., 2019). However, this finding is relatively low as compared to the finding by Igbinosa et al. (2017), who reported the presence of 16 (22.22%) imipenem (carbapenem class of antibiotic) resistant Salmonella isolated from fresh vegetables. According to WHO (2019), there is still limited transmission of carbapenemresistant Salmonella from non-human sources, however, based on reports from several countries, the spread in the food chain is increasing. Carbapenem-resistant bacteria have caused severe outbreaks, the infections are difficult to treat and associated with high costs and mortality (Malchione et al., 2019). The increasing trend of carbapenem-resistance in bacteria necessitates a move to the usage of colistin as the last-resort antibiotic where this antibiotic is active against carbapenem-resistant bacteria (Malchione et al., 2019;Wakabayashi et al., 2020). The finding of colistin-resistant NTS in this study is comparable to a report by Verma et al. (2018) which found two (7.7%) isolates of colistin-resistant Salmonella from vegetable samples examined. The eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER emergence of colistin-resistant bacteria is a worldwide concern as this antibiotic is recognized as the last choice of antibiotic for the treatment of infections in hospitalized patients caused by carbapenem-resistant Gram-negative bacteria (Elbediwi et al., 2019). The presence of carbapenem and colistin-resistant NTS isolated from salad vegetables in this study, even in low percentage, indicate the emerging spread of resistance toward these last-resort antibiotics in the food chain.
The level of MDR NTS (18.97%) in this study was lower than that previously reported by Abatcha et al. (2018) in Malaysia (55.1%). Abatcha et al. (2018) also reported the presence of two (50%) MDR isolates of S. Albany from vegetables with high MAR index of 0.66. Several other studies also reported the high prevalence of MDR among S. enterica ser. Albany such as in the chicken food chain in Cambodia (96.1%), poultry processing plant and retail markets in South Korea (93.2%) and chicken meat in Korea (80%) (Shang et al., 2019;Vuthy et al., 2017;Sin et al., 2020). The MDR characteristics among most of S. enterica ser. Albany isolated in this study and also from others might indicate the specific S. enterica ser. Albany in the food chains have acquired resistance to various types of antimicrobial agents including the medically important antibiotics.

Conclusion
Findings in this study reveal the relatively high occurrence of NTS contamination of salad vegetables from wet markets in Terengganu, Malaysia, suggesting that salad vegetables could represent a potential source of salmonellosis. Furthermore, the resistance toward medically important antibiotics among NTS in this study highlight the potential risk to public health. More attention should be given to the presence of carbapenem and colistin-resistant isolates as these antibiotics are categorized as last-resort antibiotics in the treatment of severe foodborne illness cases. These data may contribute to providing useful information for the development of effective strategies to ensure the safety of salad vegetables sold at retail outlets in Malaysia.