Multidrug resistance, biofilm formation and detection of blaCTX-M and blaVIM genes in E. coli and Salmonella isolates from chutney served at the street-food stalls of Bharatpur, Nepal

Antimicrobial resistance (AMR) amid the bacteria found in ready-to-eat foods is a grave concern today warranting an immediate intervention. The current study was undertaken to explore the status of AMR in E. coli and Salmonella species in ready-to-eat Chutney samples (n = 150) served at the street food stalls in Bharatpur, Nepal, with a major focus on detecting extended-spectrum β-lactamase (ESBL) and metallo β-lactamase (MBL) genes along with biofilm formation. Average viable counts, coliform counts, and Salmonella Shigella counts were 1.33 × 106±141481.4, 1.83 × 105±91303.6, and 1.24 × 105±63933.19 respectively. Out of 150 samples, 41 (27.33%) harbored E. coli, of which 7 were E. coli O157:H7; whereas Salmonella spp. were found in 31 (20.67%) samples. Bacterial contamination of Chutney by E. coli and Salmonella and ESBL-production were both found significantly affected by different sources of water used, personal hygiene and literacy rate of the vendors as well as by the type of cleaning materials used to wash knives and chopping boards (P < 0.05). Antibiotic susceptibility testing revealed that imipenem was the most effective drug against both types of bacterial isolates. Additionally, 14 (45.16%) Salmonella isolates and 27 (65.85%) E. coli isolates were found to be multi-drug resistant (MDR). Total ESBL (blaCTX-M) producers reported were 4 (12.90%) Salmonella spp. and 9 (21.95%) E. coli. Only 1 (3.23%) Salmonella spp. and 2 (4.88%) E. coli isolates were blaVIM gene carriers. Dissemination of knowledge of personal hygiene amongst the street vendors and consumer awareness regarding ready-to-eat foods are crucial factors that can be suggested to curtail the emergence and transmission of food-borne pathogens.


Introduction
Foodborne illness has been a common problem worldwide most probably due to changes in marketable food production, such as minimal processing, and changing consumer demands for ready-to-eat (RTE) meals [1][2][3]. Foodborne disease, particularly diarrhea, is a major public health concern caused by eating microbially contaminated foods [4]. Because of the rise in food-related illnesses, food resistant outcomes from antibiotic susceptibility investigations suggest the notion of a possible MBL or ESBL production necessitating their confirmation phenotypically [14,24,25]. Biofilm-producing bacteria, according to infectious disease experts at the Centers for Disease Control and Prevention (CDC), could be responsible for 65% of all bacterial infections, compared to their free-floating counterparts [26]. Most importantly, production of biofilm, ESBL and MBL confer E. coli [14,27] as well as Salmonella [28][29][30][31] with increased resistance towards an array of antibiotics; thereby, contributing to failures in the treatment of several infections.
In Bharatpur, most of the of street food handlers are unaware of good hygienic practices because they lack knowledge about hazardous consequences of contaminated foods and some of them are deliberately ignoring this issue and take it merely as a way to earn money by ignoring the health of the consumers. So far, there are very negligible efforts launched by Bharatpur Metropolitan City and other stakeholders in order to raise awareness on possible health issues caused by consuming unhygienic street foods. Considering these facts, the present study was carried out to investigate the incidence of two important food-pathogens: Salmonella and E. coli in street-vended Chutney with an emphasis on antimicrobial resistance, ESBL, MBL and biofilm production in the bacteria and exploring some pertinent factors that might have contributed to their emergence.

Study site, design and sample size
A descriptive cross-sectional study was carried out from September 2019 to March 2020 in Bharatpur Metropolitan City of Chitwan District, Bagmati Province, Nepal. Chutney samples (n = 150) were collected from 3 major crowded areas of Bharatpur city: Narayani kinar, hospital area and school/college area (50 samples from each area) ( Fig. 1) using random sampling methods without repetition. The sample size was determined in accordance with the prevalence rate based on the previous study [21].

Sample collection and transportation
Chutney samples were collected in sterile zip-lock plastic bags from the street food stalls in Bharatpur city. The retrieved samples were placed in an icebox and transported aseptically to Birendra Multiple Campus's Microbiology laboratory within 1 h for further investigation.

Enumeration of bacteria from chutney sample
After transporting to the laboratory, Chutney (1 g) was immediately mixed with 9 mL of distilled water and serially diluted up to 10 − 5 . Spread plate technique was performed using Plate count agar, Eosin Methylene Blue (EMB) agar and Xylose Lysine Deoxycholate (XLD) agar. The plates were aerobically incubated for 24 h at 37 • C and CFU/mL was determined [29].

Identification of the isolates
The suspected colonies of E. coli on EMB agar and Salmonella on XLD agar were further subcultured on Nutrient agar (NA). Bacterial isolates were identified by their cultural, morphological and biochemical tests [32]. E. coli isolates were streaked on MacConkey Sorbitol agar and incubated at 37 • C for 24 h for the preliminary identification of E. coli O157: H7, with colorless colonies suspected to be E. coli O157: H7. The slide agglutination test, which uses anti-O157 and flagellar H7 serum, was used to confirm E. coli O157: H7 (Defico, USA).

Phenotypic assessment of ESBL and MBL producers
The ceftazidime (30 μg) and cefotaxime (30 μg) discs were used to screen ESBL producing isolates. According to the Clinical and Laboratory Standards Institute, a prospective ESBL producer has a zone of inhibition (ZOI) of ≤22 mm for ceftazidime and ≤27 mm for cefotaxime. A combined disc test was used to confirm phenotypic confirmation of potential ESBL producers, as suggested by CLSI [33]. The validation of ESBL-producing isolates was accomplished using ceftazidime and cefotaxime alone and in combination with clavulanic acid (CA) (10 μg). An increment in ZOI of 5 mm for either antimicrobial drug tested in conjunction with CA as compared to its zone when tested alone was considered as ESBL production [21]. Imipenem-resistant isolates were chosen for further detection of MBL production using the disc potentiation method using imipenem (10 μg) and meropenem (10 μg) with and without EDTA (1 μg) [24].
Metallo beta-lactamase production was defined as a difference of ≥7 mm between the zones of inhibition of any of the carbapenem discs with or without chelating agents (EDTA) [34].

Detection of biofilm by microtiter plate method
Detection of biofilm formation was conducted by microtiter plate method as suggested by Stepanović et al. [35] and Kuinkel et al. [36] with slight adjustments. Briefly, a loopful of bacteria was inoculated in 10 mL of tryptic soya broth (TSB) supplemented with 1% glucose. After proper incubation at 37 • C for 24 h, the culture solution was diluted 100 times with freshly prepared TSB and 200 μL of diluted TSB was distributed to each well of a 96-welled microtiter plate. The plate was incubated at 37 • C for 24 h, the suspension was removed gently from the wells, washed with phosphate buffer saline (4 times), and fixation of biofilm attached to the walls of the plate was done with 2% sodium acetate. To visualize the biofilm, staining was done with 0.1% crystal violet while the extra stain was removed with de-ionized water followed by the drying of plates. To remove the dye from the cells, all wells were filled with 200 μL of 95% ethanol. The optical density (OD) of stained adherent biofilms was then read using a microtiter reader at 630 nm, and the results were categorized as weak, moderate, or strong biofilm producers. The optical density cut-off value (ODc) was calculated by arithmetically averaging the OD of the sterile TSB broth wells and adding a standard deviation of +2. Positive samples had an optical density greater than the cut-off value, while the negative samples had a lower optical density than the cut-off value. The following criteria were used to determine the capacity of biofilm production: no biofilm production (OD ≤ ODc), weak bioflim production (ODc < OD ≤ 2ODc), moderate biofilm production (2ODc < OD ≤ 4ODc), and strong biofilm production (OD > 4ODc) [35,36].

DNA extraction and amplification of bla CTX-M and bla VIM gene by PCR
All of the phenotypically confirmed ESBL and MBL producers were inoculated into 5 mL of Luria-Bertani broth (Hi-media, India) and incubated at 37 • C for 24 h. The plasmid DNA was extracted using the alkaline lysis technique [37]. After that, the DNA samples The amplified PCR products were separated on a 1.5% agarose gel in 1× TAE buffer (0.04 Tris-acetate, 0.001 M EDTA, pH 8.0), dyed with ethidium bromide, and observed with a gel-doc system. The amplicon size for bla CTX-M and bla VIM were 560 bp [38] and 390 bp [39] respectively ( Table 1).

Preservation of the bla CTX-M and bla VIM genes positive isolates
Bacterial isolates were preserved using the glycerol stock method with some modifications [40]. Overnight culture of bacteria in LB broth (0.5 mL) was added into 50% glycerol stock (0.5 mL) in Eppendorf tube with proper labeling. Culture-inoculated Eppendorf tubes were preserved in the refrigerator at − 20 • C. Since subsequent freezing and thawing reduces the shelf life of the culture, while recovering, the frozen culture was scraped off with a sterile inoculating loop and directly transferred to culture media without letting the glycerol stock unthaw completely.

Quality control for test
Sterility and performance tests were performed on each batch of media and reagents. The control strains of E. coli ATCC 25922 and Salmonella Typhimurium ATCC 14028 were used to ensure quality control during the antibiotic susceptibility test. Media, antibiotics, and reagents were prepared, stored, and used according to the manufacturer's instructions for quality control. The antibiotic discs were kept in the refrigerator. For every PCR reaction, positive and negative controls were employed.

Data management and analysis
The SPSS software for windows (version 26) and R-programming statistical analysis tool were used to evaluate all the obtained data (version February 1, 5033, https://cran.r-project.org/). Chi-square test (χ 2 ) and logistic regression were used to scrutinize the data, and ggplot2 (grammar of graphics, version 3.3.2) package was used to construct the plots. The degree of significant associations Table 1 Nucleotide sequence of the primer for the detection of bla CTX-M and bla VIM gene.

Gene
Primers between the variables were presented as P < 0.001*** (highly significant), P < 0.01** (moderately significant) and P < 0.05* (significant). Likewise, Odds ratio (OR) and 95% confidence interval (CI) with both upper and lower limits were also calculated. Arc Geographic Information System (ArcGIS, version 10.2, https://www.arcgis.com/index.html) was used to locate the sampling points from the different sample collection areas.

Antibiotic susceptibility pattern (AST) of the isolates
Imipenem was the most efficient antibiotic as 87.80% E. coli and 90.32% Salmonella spp. isolates were sensitive to it. Cotrimoxazole, ciprofloxacin and gentamicin were the drugs which were effective at a fair rate against both the isolates. Amoxicillin and ampicillin were the least effective drugs against both isolates as Salmonella and E. coli isolates completely resisted these drugs ( Fig. 4A and B). Among 41 E. coli isolates, 27 (65.85%) isolates and 14 (45.16%) Salmonella spp. among 31 isolates were MDR.

Distribution of ESBL-and MBL-producing E. coli and salmonella isolates
The ESBL-screening test was positive in 37 out of the 72 isolates. The combined disk-diffusion test revealed that 21 isolates of E. coli and 8 isolates of Salmonella spp. were ESBL producers. Nine (21.95%) E. coli and 4 (12.90%) Salmonella isolates documenting a total of

Association between the presence of bla CTX-M and bla VIM genes with multidrug resistance, ESBL and MBL production
Amongst 41 E. coli isolates, the rate of MDR, ESBL and MBL production was observed higher (77.78%, 100.00% and 22.22% respectively) in the isolates which bore bla CTX-M gene as compared to the isolates which lacked the bla CTX-M gene (62.5%, 37.5%, 62.5% respectively). All Salmonella isolates carrying the bla CTX-M gene were MDR and ESBL producers, while 25% of them were found to produce MBL. Likewise, all bacterial isolates harbouring bla VIM gene were reported to be MBL producers. In contrast, significant association was observed between bla CTX-M possession, MDR status and ESBL production in case of Salmonella spp.; whereas only between bla CTX-M possession and ESBL production in case of E. coli (P < 0.05) ( Table 5).

Discussion
In developing countries like Nepal, street foods are probably one of the prominent sources of food-borne illnesses, yet they are preferred and consumed by a large chunk of people. The present study investigated the bacterial quality of Chutney served at different street-vended shops with a major focus on Salmonella and E. coli.
A total of 150 Chutneys sold at the street food stalls in different areas of Bharatpur city were subjected to bacterial investigation. Chutney samples had average CFU counts of 1.33 × 10 6 ±141481.4, 1.83 × 10 5 ±91303.6, and 1.24 × 10 5 ±63933.19 for total viable growth, total coliform growth and total Salmonella Shigella growth respectively. These counts are higher than those reported in a study conducted by Eromo in Ethiopia, where the count was 1.7 × 10 5 in Chutney (Awaze) samples [41]. Another Ethiopian study reported that the CFU in street food samples was 6.64 × 10 4 [42]. In contrary, the mean CFU count of this study is lower than the study done in Gondar where it was 5.1 × 10 7 [43]. There is a variation in the rate of bacterial contamination of Chutney in different regions owing to many factors such as the environmental conditions which determine the bacterial growth and also the different types of ingredients (raw vegetables) and water used at different places to prepare Chutney [5]. Out of 150 samples, 27.33% samples were contaminated with E. coli. This rate is lower than some similar studies done in India and Bangladesh, which recorded 55.00% and 46.00% samples contaminated with E. coli respectively [44,45]. The incidence of Salmonella spp. reported in this study was 20.67%. A similar study in Deharadun city recorded 30.00% of Chutney contaminated with Salmonella spp. [44]. Though the rate of contamination is different, the nature of the contamination is similar as most of the studies in street foods have recorded a higher prevalence of E. coli than Salmonella isolates [21,44,45]. In the present study, E. coli O157: H7 have been recovered in 4.67% of Chutney samples. A study done in green salads served at hotels and restaurants of the same city, Bharatpur, showed 1.70% prevalence of E. coli O157: H7 [21]. E. coli O157: H7 contamination is more likely in vegetables grown in soil fertilized with animal manure [46]. Most of the Chutney samples from street food stalls located near school/college were found contaminated by E. coli (36.00%) and Salmonella spp. (24.00%) as compared to the food stalls at Narayani kinar and Hospital area, though it was not found statistically significant (P > 0.05). This can be ascribed to the fact that school/college areas are more crowded, which increases the chance of bacterial contamination leading to the food borne illnesses. The literacy rate of the chefs/vendors is an important factor that defines the extent to which the foods are contaminated. Chutney prepared by illiterate vendors had a higher incidence of E. coli (33.33%) (P < 0.001) and Salmonella spp. (31.25%) (P < 0.05). Vendors who are illiterate are unaware of health and hygiene issues, and they may cross-contaminate the Chutney during the preparation process. Vendors' cleanliness habits may have a direct impact on the dissemination of E. coli and Salmonella. Chutney prepared by non-gloved vendors accounted for a higher growth rate of E. coli (32.23%) (P < 0.05) and Salmonella spp. (23.97%). This finding echoes with another study which documented that non-gloved vendors served highly contaminated ready-to-eat foods compared to gloved vendors [18]. Foodborne pathogens such as Salmonella and E. coli O157: H7 are easily transmitted by contaminated hands [47]. Non-gloved vendors/persons may carry dirt and dust particles in their hands, which may contain a larger number of germs that can contaminate the Chutney while preparing. The cleaning trend of chopping boards and knives affects the frequency of Salmonella in Chutney [48]. A higher growth rate of E. coli (60.53%) and Salmonella spp. (39.47%) was    documented in the Chutney sold by the vendors who use water only to wash chopping board and knives (P < 0.05). The finding is in accordance with a previous study conducted with green salads served at hotels and restaurants in Bharatpur city [21]. Disinfection methods can lessen bacterial burdens on non-living surfaces, resulting in reduced coliform counts [49]. Chutney prepared by using underground water (open) accounted for higher occurrence of E. coli (64.29%) and Salmonella spp. (57.14%) in the present study. Municipal water is supplied only after treating with disinfectant which kills the bacteria present in it, and underground water (closed) has very little chance of being contaminated with microbes; whereas there is always a greater chance of contamination of open sources of water. Various factors that might contribute to the contamination of open water sources (wells) are-proximity of latrine to the wells, haphazard use of wells and waste management pits near the water sources [50]. Consumers' awareness on food safety is very important in order to curtail the incidences of foodborne illnesses and infections.
On performing antimicrobial susceptibility testing, imipenem and co-trimoxazole were found to be the most effective antibiotics against both the bacterial pathogens. A study performed in the same city revealed that co-trimoxazole was the most effective drug with   93.02% and 97.00% sensitivity against E. coli and Salmonella spp. detected from green salads respectively [21]. But another study done in street-vended Panipuri of the same city noted a lower sensitivity rates (46.67% and 42.85%) of co-trimoxazole against these two bacteria [18]. A similar study done on street foods in Ethiopia reported gentamycin (93.33%) and chloramphenicol (86.67%) sensitivity rates higher than our study [42]. Another Ethiopian study reported 88.90% resistance to chloramphenicol by Salmonella spp. which is higher than our study (48.39%) [41]. Amoxicillin and ampicillin were found to be the most ineffective drugs as all the isolates of E. coli and Salmonella spp. resisted these drugs. Some other studies on street foods also reported ampicillin as the least effective drug [41,42]. However, a study performed by Khadka et al. showed only 20.00% isolates of E. coli and 17.85% of Salmonella spp. resisted this drug [18]. The low efficacy of these medications could be owing to simple hydrolysis of the β-lactam ring by Salmonella and E. coli, as well as the widespread use of β-lactam antibiotics due to their low cost, leading to the development of resistance in most bacteria [51,52]. In our study, 65.85% isolates of E. coli and 45.16% Salmonella isolates were MDR. The rate of MDR isolates in this study is lower than some other studies done in vegetables and fruits [53,54]. In contrast, a study performed on street-vended Panipuri in the same city has reported only 36.70% E. coli and 33.33% Salmonella spp. as MDR. The problem of antibiotic resistance, which can be ascribed to irrational use of antibiotic, is increasing enormously day by day and is threatening the medical fraternity. The prevalence of ESBL-producing (bla CTX-M gene) E. coli and Salmonella spp. was 21.95% and 12.90% respectively in our study. A study done on retail foods in China reported 9.4% ESBL-producing Enterobacteriaceae [20]. Meanwhile, a study conducted by Raphael et al. [55] observed an ESBL incidence rate of only 2.3% among bacterial isolates from spinach sold at street vended shop; whereas another study performed in Bharatpur city recorded only 2.3% Salmonella isolates and 1.85% E. coli isolates as ESBL producers [21]. Little is known about the transfer of ESBL-producing bacteria in food, but the endogenous fecal flora of animal origin can certainly spread across the food chain and contaminate water [19]. Also, horizontal transfer of ESBL producers from humans to foods may occur due to the sheer negligence of food handlers [56]. This study recorded 4.88% and 3.23% isolates of E. coli and Salmonella spp. carrying bla VIM gene respectively. Iseppi et al. reported only 1.28% of isolates carrying bla VIM gene in ready-to-eat foods [25]. Following the emergence of ESBL-producing bacteria in the 2000s, usage of carbapenem antibiotics increased drastically resulting in the increased number of beta-lactamases-producing bacteria that can even hydrolyze carbapenems [57].
Of the 41 isolates of E. coli, 19 (46.34%) were biofilm producer among which 17 isolates were MDR, whilst 11 (35.48%) isolates were biofilm producers in case of Salmonella spp., of which 6 were MDR. Few years back, Behzadi et al. also reported that production of biofilm resulted in resistance to multiple classes of antibiotics [58] and not only in Enterobacteriaceae family, there has been report of similar trends in Pseudomonas spp. and Acinetobacter spp. as well [59]. Present study showed that the prevalence of MDR, ESBL and MBL production is pretty higher among the E. coli isolates possessing bla CTX-M gene as compared to the isolates which lacked bla CTX-M gene (77.78%, 100.00%, 22.22% vs. 62.50%, 37.50%, 62.50%) for MDR, ESBL and MBL respectively. All bacterial isolates harbouring bla VIM gene were reported to be MBL producers in the present study. Higher antimicrobial resistances are observed in E. coli and Salmonella harbouring bla CTX-M and bla VIM genes and are associated with the spread of clonal lineages as suggested by the previous findings [14,[60][61][62].

Conclusion
The results of the study hint that a poor quality of Chutney is being served to the consumers eating street-vended foods. Use of untreated water from open wells, unawareness about personal hygiene and improper cleaning of the utensils all have contributed to Chutney contamination by Salmonella and E. coli. The emergence of ESBL, MBL and biofilm-producing isolates in ready-to-eat Chutney is worrisome as it might lead to serious health consequences. Concerned authorities should act seriously and quickly to prevent outbreaks of food-borne illnesses and the spread of AMR in the metropolis.

Strengths and limitations
To the best of our knowledge, this is the first attempt exploring both phenotypic as well as the genotypic studies of ESBL, MBL and biofilm-producing E. coli and Salmonella from the Chutney samples sold in street-vended shops in Nepal. The results of the study might be helpful to the stakeholders to formulate policies and conduct programs related to food safety and consumer awareness. On the downside, the current work is confined to investigating only two bacterial pathogens (E. coli and Salmonella) and also to the molecular identification of only two genes. Furthermore, the potential sources of contamination of the Chutney were not identified. Future studies are recommended to address these limitations.

Contributions
SA, SK, SS and RSR conceived and designed the experiment. NP, SG, PB, PS, DT, RD, RSR carried out the laboratory works. SK, RSR and SS wrote the manuscript. SK and SA analyzed the data. SA, SK, AG, SS and KRR supervised the work. The final manuscript was reviewed and approved by all contributors.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
Not needed.

Consent of publication
Not applicable.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.