Genetic diversity and antibiotic resistance of Shigella spp. isolates from food products

Abstract The emergence of multidrug‐resistant Shigella is a significant threat to global public health. Limited studies have investigated the incidence, antimicrobial susceptibility, and genetic diversity of Shigella isolated from food products. Conventional culture‐based, serologic, molecular, disk diffusion, PCR, and RAPD‐PCR methods were used to determine the prevalence rate, phenotypic and genotypic antibiotic resistance profile, and genetic diversity of the Shigella isolates from food samples including vegetable salad, ground meat, and raw cow's milk (405 samples). The prevalence rate of Shigella in food samples was 4.44%. The incidence of S. sonnei (3.7%) was higher than that of S. flexneri (0.74%). S. dysenteriae and S. boydii were not detected in food samples examined. Also, no Shigella were recovered from raw cow's milk. This study showed that the Shigella isolates were resistant to sulfamethoxazole/trimethoprim (83.3%), amoxicillin (66.6%), streptomycin (66.6%), tetracycline (61.1%), ampicillin (50%), amoxicillin–clavulanic acid (50%), azithromycin (50%), and chloramphenicol (50%) and completely sensitive to cefoxitin, cefepime, amikacin, and gentamicin. All Shigella isolates were multidrug‐resistant. We detected bla SHV resistance gene in all isolates; however, no isolate harbored bla TEM gene. RAPD‐PCR categorized the Shigella isolates into five main clusters. The highest antibiotic resistance was observed in the isolates of cluster R4. The finding of this study also indicated an association between antimicrobial resistance profiles and genotyping properties of the isolates. Novel food monitoring systems, including surveillance of multidrug‐resistant foodborne pathogens, especially in developing countries, are required to control the foodborne diseases.


| INTRODUC TI ON
Foodborne pathogens (including bacteria, viruses, and parasites) have become one of the most critical public health concerns around the world. When a pathogen is ingested with the contaminated food or drink and establishes itself in the human host or the toxin of a toxigenic pathogen is released and ingested, a foodborne illness occurs (Nyachuba, 2010). WHO estimated more than 600 million cases and 420 000 deaths each year caused by 22 major foodborne diseases worldwide (Organization, 2021). The most severe foodborne illnesses tend to occur in immunocompromised, very young, or very old patients. Foodborne pathogens (mainly bacterial agents), leading to intestinal and extraintestinal diseases, are causing significant adverse effects on human health and economic hardship (Hanson et al., 2012). The most common foodborne pathogenic bacteria include Escherichia coli, Campylobacter spp., Salmonella spp., Clostridium perfringens, Staphylococcus aureus, Shigella spp., and Listeria monocytogenes. More than 200 intestinal and extraintestinal disorders have been identified to be caused by foodborne pathogens (Pakbin et al., 2020;Smith & Fratamico, 2018).
Shigellosis, caused by Shigella spp., is characterized by invasion of the epithelial cells lining the colon and rectum. Shigella is a nonmotile, rod-shaped, non-spore-forming, and gram-negative bacteria belonging to the Enterobacteriaceae family. This pathogen includes four species: S. dysenteriae (serogroup A), S. flexneri (serogroup B), S. boydii (serogroup C), and S. sonnei (serogroup D). Detection of S. sonnei and S. flexneri is more frequent in developed and developing countries, respectively. It is also one of the most critical causes of diarrhea-related mortality and morbidity around the world. It is estimated that 55,000 and 110,000 deaths and hospitalizations, respectively, in children under five years old and older children or adults worldwide, are caused by Shigella annually (Bennish & Ahmed, 2020).
Shigella spp., as the causative agent of bacillary dysentery, has been involved in several foodborne and waterborne outbreaks (Kotloff et al., 2018).
Shigella is the third most reported bacterial foodborne pathogen.
Foods are mainly contaminated with these pathogens by infected food handlers with poor personal hygiene (Nataro et al., 2011).
However, foodborne outbreaks caused by Shigella mainly occur while foods are subjected to preparation or processing by hands, exposed to insufficient thermal processing, and delivered or served raw to the consumers. Shigella can survive in acidic and salty conditions, be transmitted to the human host and cause disease (Baker & The, 2018). Fresh vegetables, deli meats, and unpasteurized milk are more susceptible to be contaminated with Shigella spp. (Warren et al., 2006). Some researches also demonstrated that Shigella spp. had been isolated from ground beef, raw oysters, bean dip, raw vegetables, potato salad, and fish (Ahmed & Shimamoto, 2014, 2015Bantawa et al., 2019;Cetinkaya et al., 2008).
In addition to the prevalence rate, the multidrug resistance of Shigella spp. has recently been considered a significant concern in food safety. Antibiotic therapy reducing the likelihood of complications, death, and hastening clinical recovery is the primary treatment of shigellosis (Ingle et al., 2020). The emergence of multidrugresistant (MDR) Shigella spp. makes the treatment of shigellosis more difficult and highlights the problem of antimicrobial resistance (Ma et al., 2018). However, it is essential to determine the appropriate antibiotics for the treatment of shigellosis via a deep understanding of the current changes in resistance patterns by studying the phenotypic and genotypic antimicrobial resistance profiles of the Shigella isolates. Molecular characterization of MDR strains has been used to analyze isolated pathogens. Shigella isolates have also been used to determine the genetic relatedness and diversity between the isolates (Zamanlou et al., 2018). Random amplified polymorphic DNA (RAPD) is one of the PCR-based molecular techniques for genotyping and determining genetic diversity between the pathogenic isolates (Ben Braïek et al., 2018). Several studies have been implemented to characterize the antibiotic resistance in Shigella isolates from clinical specimens; however, limited literature provides the molecular characterization of MDR Shigella isolates from food samples (Shahin et al., 2019). Therefore, the purpose of this study was to investigate the genetic relatedness and molecular characterization of MDR Shigella isolated from fresh vegetable, unpasteurized milk, and ground meat samples collected in a survey in Iran.

| Food samples collection
Food samples (N = 405), including 135 raw cow's milk, 135 vegetable salads, and 135 ground meat samples, were randomly collected from different local supermarkets and restaurants in various areas of Qazvin city, Iran, between July and December 2018. All samples were aseptically collected in sterile containers and tubes, labeled and immediately transferred to ice-boxes, and transported to the laboratory for microbiological analysis.

| Isolation, biochemical, and serogroup identification of Shigella spp
Shigella spp. were isolated and identified from the food samples according to the method for isolation of Shigella from food described previously (Mokhtari et al., 2012) (Figure 1). 25 g of each vegetable salad and ground meat samples and 25 ml of raw cow milk samples were mixed with Shigella Broth (Merck, Germany) containing 0.5 µg/ ml novobiocin (Sigma-Aldrich, NSW, Australia), homogenized for 1 min at 320 rpm in a Stomacher BagMixer (InterScience, France), and incubated anaerobically overnight at 42°C. 0.1 ml of enriched samples was streaked onto the MacConkey agar (Merck, Germany) plate and incubated anaerobically at 42°C for 48 hr. Presumptive colonies (convex colorless to slightly pink) on MacConkey agar were selected and subjected for Gram staining and biochemical evaluation, including indole production, citrate utilization, methyl red, Voges-Proskauer, triple sugar iron, motility, urea, and oxidase (Merck, Germany) (Shahin et al., 2019). All isolates which biochemical tests had confirmed were subjected to serologic tests. Genus and species of the isolates were confirmed and identified by slide agglutination method using Shigella Genus and Shigella Species Difco antisera kits (BD-Difco Co., USA) according to the manufacturers` instructions.

| Phenotypic and genotypic antibiotic resistance profile
Antimicrobial susceptibility phenotypes of Shigella isolates were determined by the Kirby-Bauer disk diffusion method according to the interpretive criteria and standards previously described by the Clinical and Laboratory Standards Institute (Weinstein & Lewis, 2020). The These phenotypic resistances were selected to be evaluated because they were frequently observed mainly among the Shigella and other Enterobacteriaceae family strains isolated from clinical and food samples previously. All results were recorded and interpreted based on CLSI guidelines (Weinstein & Lewis, 2020). To investigate the genotypic profile of antibiotic resistance, beta-lactamase resistance genes including bla TEM , bla OXA , bla SHV , bla CT X-M-1 , bla CT X-M-2 , bla CT X-M-8 , and bla CT X-M-9 were detected among 18 Shigella isolates by conventional PCR using the specific primers and thermal cycling

| DNA extraction
All Shigella isolates were grown anaerobically on BHI broth overnight at 42°C. 1 ml of the bacterial suspension was suspended in 1 ml of Phosphate-Buffered Saline (PBS, Promedia Spain) and centrifuged at 10,000 g for 5 min. After removing the supernatant, the bacterial pellet was subjected to DNA extraction using the Sinaclon Gram-negative DNA extraction kit (Sinaclon Co., Iran) according to the kit manufacturers` instructions. The quality and quantity of the extracted genome were determined using NanoDrop 8000 spectrophotometer (Wilmington, DE, USA). All extracted DNA concentrations were adjusted to 50 μg/ml with PBS before PCR reactions.

| RAPD-PCR genotyping
The genetic relatedness among the Shigella isolates was evaluated using the RAPD-PCR method as previously described. As previously recommended, the arbitrary primer UBC245 (5'-CGCGTGCCAG −3') (Berthold-Pluta et al., 2017) was used in this study for RAPD-PCR.
The total volume of PCR reaction was 25 µl, including 12.5 µl of PCR Master Mix (Ampliqon, Denmark), 1 µl of RAPD primer (5 pmol/µL), 2 µl of DNA template, and sterile deionized water to reach the final reaction volume. The PCR was performed as follows: 94°C for 7 min, 36°C for 1 min, 72°C for 4 min and 40 cycles of 94°C for 1min, 36°C for 1 min and 72°C for 4 min. The amplified products were characterized using electrophoresis through a 1.5% (w/v) agarose gel F I G U R E 1 Flowchart of culture-based methods for detection and identification of presumptive Shigella spp. isolated from food samples stained with DNA safe stain at 100 V for 1.5 hr. Gels were visualized, and the patterns were recorded using the NovinPars Gel Doc system (NovinPars Co., Iran). RAPD genetic markers were analyzed using PyElph software (Pavel & Vasile, 2012). The dendrogram was constructed using the Unweighted Pair Group Method with Arithmetic averages (UPGMA) method and Dice coefficient using NTSYS-pc software version 2.1 (Rohlf, 2001). The RAPD-PCR pattern types of the isolates with a high similarity index (≥ 0.6) were considered the closely related RAPD pattern type.

| Statistical analysis
Chi-square and Fisher's exact tests were used to evaluate significant differences (p <.05) between the contamination rates using IBM SPSS software, version 21.0.1 (IBM Corp., Armonk, NY, USA). All experimental and measurements were performed in triplicate (Table 1).

| Isolation and identification of Shigella spp. in food products
Prevalence rates of Shigella spp. in food products, including vegetable salad and ground meat samples, are shown in Figure 2; however, no Shigella spp. were isolated from raw cow's milk samples. In total, eighteen isolates of Shigella spp. (4.44%) were detected and molecularly identified in 405 vegetable salad, raw cow milk, and ground meat samples tested in this study. ipaH gene was detected in all 18 Shigella isolates. Among the isolates, 13 (3.20%) and 5 (1.23%) isolates were detected in vegetable salad and ground meat samples, respectively. Notably, 15 (3.70%) and 3 (0.74%) out of all Shigella isolates were identified as S. sonnei and S. flexneri, respectively. No S. boydii and S. dysenteriae were recovered from the food samples.
All isolates were primarily identified using biochemical methods and then confirmed by serologic tests. S. sonnei strains showed a significantly (p =.01) higher prevalence than other species. Also, Shigella spp. were detected significantly (p =.01) higher in vegetable salad samples than the other food products assessed.

| Genetic relatedness among the Shigella isolates
In this study, genetic relatedness and genotypic polymorphism of eighteen Shigella spp. isolates from vegetable salad and ground meat samples were characterized by the RAPD-PCR method using the UBC245 arbitrary primer. Amplification of the Shigella spp. isolates with the UBC245 primer resulted in polymorphic patterns composed of three to eight bands ranging in size from 300 to >1,500 bp ( Figure 3). As it is illustrated in Figure 3, the UBC245 oligonucleotide discriminated the eighteen Shigella isolates into five clusters (R1-R5) of genetically identical Shigella isolates with more than 60% RAPD-PCR profile similarity (at 60% similarity cut-off value). The t test compared all the groups, and the significant differences were not observed (p >.01). Genetic diversity within Shigella isolates was calculated using Simpson's diversity index. The average diversity index for the isolates was 0.82. Consequently, a high level of genetic diversity was observed among the Shigella spp. isolates. Clusters R1, R3, R4, and R5 contained 3 Shigella isolates; however, only cluster R2 included six isolates. All S.
flexneri isolates were grouped in cluster R2 (Table 3). Vegetable salad Shigella spp. isolates were not grouped in the same cluster; ground meat isolates were also included in different clusters. As shown in Table 3, the highest and lowest antibiotic-resistant isolates were profiles of the isolates in the same cluster were significantly close to each other. There is also a strong association between the presence of beta-lactamase genes and the resistance phenotype profile of the isolates. More beta-lactamase genes were detected in the Shigella isolates which were resistant to beta-lactam antibiotics, including amoxicillin, ampicillin, and amoxicillin-clavulanic acid.

| D ISCUSS I ON
Shigella spp. is one of the most prominent foodborne bacterial pathogens worldwide, especially in developing countries such as India, Tunisia, Iran and Egypt, and 99% of shigellosis infections occur in these countries (Kotloff et al., 2018). Shigellosis is estimated to cause more than 80 million cases of hospitalizations and 700 000 deaths worldwide annually (Bennish & Ahmed, 2020). In the current study, we isolated and estimated the prevalence rate of Shigella spp. in food products, including vegetable salad, raw cow milk, and ground meat samples collected in Qazvin, Iran, by using conventional culturing and serologic methods and TA B L E 3 Resistance phenotype, presence of ipaH gene, RAPD genotype groups, incidence of beta-lactamase genes and serogroups in Shigella isolates from vegetable salad and ground meat samples  determining antimicrobial susceptibility pattern and genetic diversity among the isolates using disk diffusion and RAPD-PCR assays, respectively. From 405 food samples, 18 Shigella spp. were isolated and molecularly identified, indicating that a total of 4.4% of the samples were contaminated with different species of Shigella.
The prevalence rate of Shigella spp. in this study is significantly more than the previous reports from Egypt (1.7%; 27 out of 1,600 samples) (Ahmed & Shimamoto, 2015), Tunisia (2.14%; 6 out of 280 samples) (Mokhtari et al., 2012), and Iran (1.4%; 19 out of 1,400 samples) (Shahin et al., 2019); and less than a report from Ethiopia (7.4%; 10 out of 135 samples) (Garedew et al., 2016). However, a study in Turkey reported that Shigella spp. were not detected in any food samples (Cetinkaya et al., 2008). These differences in prevalence rates between the studies may be due to the differences in geographical locations, sample size, source of sampling, and the level of public hygiene and health services. In addition to personto-person transmission, foods are one of the primary vehicles for human infection with different species of Shigella. Consequently, travel-and trade-associated Shigella may also cause a higher prevalence rate of this foodborne pathogen in different countries.
Several epidemiological studies in these countries revealed various outbreaks caused by foodborne shigellosis were associated with consuming fresh food products, which are served raw, processed by hand, and exposed to a poor heat treatment (Ahmed & Shimamoto, 2015). Notably, the spread of Shigella contamination in foods may occur via flies, fingers, cutting surfaces, and utensils (Mokhtari et al., 2012).
However, recently in Iran, Shahin et al. (2019) reported a higher incidence of Shigella spp. in vegetables (2.2%) than in ready-to-eat (2.0%) and meat samples (0.8%) (Shahin et al., 2019). Mokhtari et al. (2012) also demonstrated a higher prevalence rate of Shigella spp. in raw salad samples than in other food products in Tunisia (Mokhtari et al., 2012). Our results showed that Shigella spp. is not often recovered from dairy products comparable with the report from Turkey.
Few publications reported Shigella spp. in dairy products or outbreaks of shigellosis in humans via consumption of these products (Shahin et al., 2019). Because humans and apes are regarded as the foremost and primary sources of Shigella spp., Shigella can be transferred to raw vegetables, vegetable salads, meat and dairy products by hand of food handlers, food processing workers, or food processing workers contaminated equipment (Ahmed & Shimamoto, 2015).
A higher incidence of Shigella spp. in raw vegetable salads indicates the high potential of these products as Shigella carriers because of the utilization of sewage and wastewater for irrigation and human waste-based fertilizers during the cultivation of edible vegetables (Ahmed & Shimamoto, 2014).
In our study, the frequency of S. sonnei (3.7%; 15 isolates) was higher than S. flexneri (0.74%; 3 isolates); however, S. dysenteriae and S. boydii were not detected in any sample. S. flexneri is the predominant species of Shigella caused shigellosis in developing countries.
However, recent clinical and food contamination reports showed that S. sonnei had become the most prevalent species of Shigella in Iran (Karimi-Yazdi et al., 2020). This revealed a correlation between the prevalence rate of frequent species of Shigella in food products and humans in Iran (Shahin et al., 2019). This correlation also was previously seen between clinical and food isolates of Shigella in Egypt (Ahmed & Shimamoto, 2015). It is worth noting that industrialization and improving the public hygiene level in developing countries such as Iran and Egypt may contribute to S. sonnei becoming the most prevalent species of Shigella in both food and clinical samples (Ahmed & Shimamoto, 2015;Shahin et al., 2019).
One of the most critical challenges in food safety is the emergence of antimicrobial-resistant bacterial strains. Resistance genes could be transferred horizontally from animal food resources and environments to humans' normal flora and pathogens through food and drink chains (Caniça et al., 2019). A broad range of antibiotics is currently used in food animals for disease prevention, veterinary treatment, and growth promotion, allowing MDR foodborne bacterial pathogens (Ebmeyer et al., 2021 Shahin et al., 2019). However, the communities and industries' reckless and unsupervised use of antibiotic agents to improve animal health may lead to differences between the antibiotic resistances patterns observed in studies in different countries (Ahmed & Shimamoto, 2014, 2015. The current study also indicated that all Shigella isolates showed an MDR phenotype. However, in previous studies on clinical isolates in Iran, the incidence rate of MDR Shigella spp. ranged from 45% to 100% (Karimi-Yazdi et al., 2020).
The prevalence rate of MDR Shigella spp. isolated from food samples was reported 90% in Egypt (Ahmed & Shimamoto, 2015). It is worth noting that the progressive increase in multidrug resistance among Shigella strains of food and clinical origins is considered a serious global concern and a threat to public health (Shahin et al., 2019). In this study, we detected beta-lactamase genes including bla SHV , bla OXA , bla C T X-M-1 , bla C T X-M-1 , bla C T X-M-8 , and bla CTX- and dairy products. They found higher beta-lactamase coding genes in meat than in dairy products (Ahmed & Shimamoto, 2015); however, we observed no significant difference between antibiotic resistance genes in the isolates from vegetable salads and ground meat samples. Beta-lactamases are considered as the main enzymes responsible for resistance of gram-negative bacteria such as E. coli, Shigella, etc., against the beta-lactam antibiotics including cefoxitin, imipenem, amoxicillin, ampicillin, cefepime, and amoxicillin-clavulanic acid. The bla SHV gene, which was the dominant beta-lactamase gene in the Shigella isolates, encodes resistance against amoxicillin-clavulanic and amoxicillin antibiotics.
The fluoroquinolones class of antibiotics, especially levofloxacin, norfloxacin, nalidixic acid, and ciprofloxacin, has been strongly considered the main treatment for shigellosis to the WHO guidelines for children and adults (Jafari et al., 2009). In the current study, 22.2%, 16.6%, and 16.6% of the MDR Shigella isolates were resistant to norfloxacin, nalidixic acid, and levofloxacin, respectively. Our results were lower than a similar study by Ahmed and Shimamoto (2015), which reported 95.8% of the MDR Shigella isolates resistant to nalidixic acid (Ahmed & Shimamoto, 2015). However, due to the development of antibiotic resistance, fluoroquinolone antibiotics are no longer effective for shigellosis (Karimi-Yazdi et al., 2020).
Colistin resistance phenotype has been observed in Shigella isolates, mostly in S. flexneri strains . In this study, we detected three isolates resistant to colistin, 2 of them were S. flexneri, and one was S. sonnei. The transfer of plasmid-mediated colistin resistance into clinical bacterial strains can lead to colistin resistance in clinical Shigella isolates and other Enterobacteriaceae pathogens. It may be of concern that resistance against colistin was transferred, and other mobile genetic elements and other resistance and virulence factor encoding genes present in other parts of the plasmid or different plasmids may be transferred . However, this problem and its mechanism need more investigation in the future.
Several DNA fingerprinting methods have been developed and used to assess the genetic diversity and clonal relatedness of foodborne bacterial pathogens such as RAPD, ERIC, and BOX-PCR methods (Kumar et al., 2018). We have used the RAPD-PCR method to characterize the genetic diversity of Shigella isolates from vegetable salad and ground meat samples and assess the clonality of these isolates as other studies have previously demonstrated for clinical bacterial strains and foodborne pathogens (Staji et al., 2018). The result of RAPD-PCR analysis showed a significant cluster (cluster R2) with six members constituting 33.3% of total Shigella isolates. All S. flexneri isolates recovered in this study were included in this cluster. All of the isolates in cluster

| CON CLUS IONS
In this study, we demonstrated that the prevalence rate of Shigella spp. including S. sonnei and S. flexneri is more frequent in vegetable salad than other food samples. The prevalence rate of S. sonnei was higher than that of S. flexneri. However, S. dysenteriae and S. boydii were not recovered from the food samples. Also, no Shigella were detected in raw cow's milk samples. Shigella isolates were highly resistant to sulfamethoxazole/trimethoprim, amoxicillin, streptomycin, tetracycline, ampicillin, amoxicillin-clavulanic acid, azithromycin, and chloramphenicol; and highly susceptible to cefoxitin, cefepime, amikacin, and gentamicin antibiotics. All Shigella isolates were MDR to three or more classes of antibiotics. We detected bla SHV resistance gene in all Shigella isolates; however, bla TEM gene was not found in any isolate. High genetic diversity was observed among the Shigella isolates. A significant association was seen between antimicrobial resistance phenotypes and genotyping of Shigella isolates. The novel food control and monitoring systems, including surveillance of foodborne pathogens, especially MDR bacterial strains in developing countries, are required to detect, control, and prevent foodborne diseases such as shigellosis.

ACK N OWLED G EM ENT
We thank our colleagues from the Department of Food Safety and Hygiene, Qazvin University of Medical Science, who assisted us in this scientific research.

CO N FLI C T O F I NTE R E S T
All authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
We confirm that all data supporting the findings of this study are available within the article.