1. IntroductionAntimicrobial agents have been used in humans, veterinary medicine, food security, and food safety since their discovery in the 1920s. However, due to inadequate selection, overuse, and misuse of antimicrobials have been responsible for the selection of resistant isolates, known as antimicrobial resistance (AMR) [1]. Over the past decade, AMR has become a global threat to human and animal health. Development of resistance can be the result of both chromosomal mutations and the acquisition of mobile genetic elements (MGEs), harboring AMR gene mutations [2,3]. It has been reported that, antibiotics are no longer effective against infection-causing bacteria due to increased AMR rate, as a result, every 10 min a patient dies in the USA or Europe [4,5]. However, a substantially higher prevalence of increased AMR is likely to be found in developing countries especially in Africa and Asia due to limited diagnosis facilities, unauthorized antibiotics sale, poor patient education, the inappropriate function of drug regulatory action, inappropriate prescription practices, and non-human practice of antibiotics in livestock sectors [6,7]. Due to the magnitude of the threat, the World Health Organization (WHO) recommended global surveillance programs for animal and human populations.According to WHO, the first-ever list of antibiotic-resistant “priority pathogens”, Escherichia coli is included in the most critical group of all twelve families of bacteria that carriage the greatest threat to human and animal health [8]. The level of antimicrobial resistance in E. coli has been used as an indicator of resistance dissemination in bacterial populations, and of selective pressure imposed by antimicrobials used in food animals and humans [9,10,11,12]. However, the frequency of AMR in E. coli depends on the source of the isolates. Animal origin has been reported to be the cause of drug-resistant E. coli infections in humans, and that these agents harbored the same mobile resistance genes found in diverse bacterial species from a variety of animal sources [13,14,15,16,17]. However, a high prevalence of AMR E. coli was isolated from chicken compared with other animals’ origins [18]. Additionally, AMR E. coli isolated from humans is similar to E. coli from poultry [19]. It has been reported that commensal E. coli, can serve as a good reservoir of resistance genes with the ability to transfer these genes to pathogens in the hosts as well as in the human intestinal tract after the consumption of contaminated foods of animal origin [20]. Furthermore, a number of studies have established the transfer of AMR between commensal bacteria and zoonotic pathogens in various ecological environments [21,22,23]. Poultry meat production has been increased and doubled over the past 20 years. Poultry is traded at live bird markets, and products are sold unprocessed with bigger clusters of them in city areas which presents significant public and poultry health challenges. A number of companies have already integrated their operations. Poultry and meat processing is a very new movement in the food processing industry in Bangladesh. It has been said that frozen chickens are mostly obtainable through high-end supermarkets charging premium prices and this market is growing every year. Another market segment is food preparation for the main fast-food chains. The local frozen food market is also growing, at a rate of almost 30% in 2011–2012 over the preceding year. City dwellers, are progressively becoming more conscious of their accessibility and the lifestyle they permit, as they desire to go to supermarket instead of to wet markets to buy their everyday stuff, including frozen chicken meat [24]. Tenants in the city becoming more conscious of their accessibility to safe food. Nonetheless, warranting the microbiological safety of frozen chicken meat evolves as a challenge.Few studies have already reported bacterial contamination in frozen chicken meat from different cities in Bangladesh. Customers in cities have a habit of buying frozen chicken meat along with other frozen and ready-to-cook foodstuffs as these frozen items need slight processing for cooking and, thus, they can save time [25,26,27]. Two of these studies were bacteriological along with AMR phenotype. Another study includes a few genes related to extended spectrum beta-lactamase (ESBL) and non- ESBL producing E. coli. However, recent reviews reported the uses of nineteen and ten different types of antibiotics in the broiler and layer farms, respectively in Bangladesh [28]. Therefore, further study is needed for genotyping which shows higher diversity than phenotypes and consequently allows for more accurate comparisons between resistant bacterial populations [29,30]. The aim of this study was to determine the prevalence of E. coli in frozen chicken and phenotypic AMR profile as well as the detection of ARG. 3. DiscussionE. coli is recognized as a common inhabitant of the vertebrate intestinal tract which frequently causes contamination in retail meat products. It is one of the most common food-borne pathogens associated with mortality in commercial poultry as well as condemning the carcasses in slaughterhouses and has been considered a significant public health threat and economic burden [31]. It has been reported that resistant strains from the gut readily contaminate poultry carcasses at slaughter, and consequently, poultry meats are often contaminated with resistant E. coli [32]. Antibiotics have been widely used for preventing economic losses caused by E. coli and increasing production efficiency [33]. However, with increased consumption of these drugs may lead to scattering them into manure and other poultry wastes and transferring them to humans by their residues in carcasses and can be the origin of bacterial resistance, mortality, and increase in human hospitalization [34]. In this study, the overall prevalence of E. coli in frozen chicken was found 31.25% which is lower than the prevalence (76.1%) reported from frozen chicken [25]. Our findings are also lower than the findings of the previous study [35]. It has been reported about a 63.5% prevalence of E. coli in raw chicken meat covering both layer and broiler swab samples. We have taken about 10 g of meat from the surface of the breast and thigh muscles of each of the broilers. However, processed meat samples for E. coli isolation from various parts of the body of layer, broiler and cockerel has been examined [25,35]. On the other hand, the present study was limited only to Dhaka city. Moreover, sampling time, season, etc. were also different. All of these factors may contribute to the differences of E. coli prevalence in frozen meat samples. Moreover, broilers sold in supermarkets especially come from contract farms that manage their farms more hygienically than the general farmers may also contribute to lower occurrences of E. coli. A contract farm is defined as a farm where farmers have a contract with the company (supermarket authority) that the company provides the chicks, the feed, veterinary care, and technical advice, etc. while the farmers provide the day-to-day care of the birds, land, and housing, as well as utilities/maintenance of the housing and finally share benefits as per contract [25]. Additionally, the prevalence may not show actual prevalence as we have examined a portion of muscle sample from the surface of the frozen chicken. The source of E. coli may be the chicken itself or it comes from contamination during the dressing and packing of chicken. It is to be mentioned that we have ensured the aseptic handling of samples in the laboratory to avoid laboratory-acquired contamination.From Bangladesh, many studies have been reported on AMR and the majority of them concentrate on the isolation and investigation of the antibiotic resistance patterns of E. coli by disc diffusion technique [36,37]. Although the conventional method is most widely used for determining AMR because of convenience, efficiency, and cost; it has some limitations. Results may be unexpected or borderline in addition to some other limitations such as its inapplicability to many fastidious organisms and anaerobes [38], unable to obtain minimal inhibitory concentration (MIC) values [39], labor-intensive and time-consuming [40]. In this study, we have used both phenotypic detection of AMR as well as detection of ARGs from the same isolates. It has been reported that tetracycline resistance E. coli was found more frequently [28]. Besides, resistances were also found against almost all antibiotics used in this study. ESBL E. coli isolates from frozen meat displayed resistance to oxytetracycline and amoxicillin (91.9%), ampicillin and trimethoprim–sulfamethoxazole (89.2%), pefloxacin (87.8%), cefepime (81.1%), piperacillin–tazobactam (73.0%), and doxycycline (70.3%) [25]. A recent review [28] reported that nineteen and ten different types of antibiotics are used in the broiler and layer farms, respectively in Bangladesh. The most commonly used antibiotics included ciprofloxacin, ampicillin, amoxicillin, trimethoprim, oxytetracycline, tylosin tartrate, tiamulin, norfloxacin, enrofloxacin, doxycycline, and colistin sulfate. Information regarding the use of antimicrobials in broiler was not available to the research team to draw further insights.MDR bacteria are an emerging clinical challenge in the poultry sector as well as the livestock sector. In this study, about 90.0% of the E. coli isolates were found MDR, and 52% of the isolates showed resistance against 4–7 different antimicrobials. Our findings are within the findings of recent reports regarding MDR phenotypes of E. coli. It has been reported eighty-six E. coli isolates from frozen chicken meat against sixteen antimicrobials and found that all the isolates are MDR [25] and as suggested by other literature reported 49.23 and 51.09% MDR E. coli isolates from broiler and layer meat samples [35]. AMR pattern (streptomycin-gentamicin-tetracycline-ampicillin-erythromycin-chloramphenicol-sulfonamide) of one E. coli isolates ( Supplementary Table S1) indicates the necessity of prudent use of antibiotics. AmpC beta-lactamase-producing gene (blaCITM) and the gene coding for extended-spectrum SHV beta-lactamases (bla SHV) were detected in broiler chicken E. coli isolates in the present study. It has been also reported that 12.8% of broiler chicken E. coli isolates carried blaSHV and 4.56% of isolates possess blaCITM genes [35]. Differences in findings might reflect the sources and number of samples etc. In Bangladesh, blaCTX-M-1 (94.4%) and blaTEM (50–91.3%) ESBL-producing E. coli were reported in the droppings of chickens [41,42]. Strong correlations between most of the antimicrobial-resistant phenotypes and genotypes were observed among the investigated E. coli isolates that the similar findings are reported earlier [35].In E. coli, the AMR phenotypic-genotypic agreement of 33–85% [29] has been reported for different antimicrobial agents and related genes. In the present study, it was found that few isolates with resistance phenotypes lacked the corresponding ARGs tested, indicating the occurrence of multi-gene mediated AMR. On the other hand, some isolates carry the resistance genes but phenotypically not resistant to the corresponding antibiotics used in this study. The occurrence of similar AMR phenomena was also reported previously [29]. Sometimes, the phenotype or the genotype alone is unable to accurately predict the outcome of the other, as molecular mechanisms of AMR are multifaceted. Thus, the presence or absence of a specific gene corresponding to a particular phenotype does not necessarily infer that the particular strain is resistant or susceptible [43]. The differences between the genotype and phenotype observed in this study might be due to not testing for all possible resistance genes, or genes not being turned on, or the presence of ‘silent gene cassettes’ in certain isolates. It is established that the use of a specific antibiotic can result in its own resistance. It can also play a role as a co-selection marker for other antibiotics. This may happen in completely unrelated drug classes [44,45]. The use of chloramphenicol in the poultry sector is very rare. However, about 22.0% of the isolates showed resistance to chloramphenicol. Moreover, chloramphenicol resistance genes viz. catA1 and cmlA were detected in 36.0 and 34.0% of the E. coli isolates, respectively. Resistance to chloramphenicol might be due to the co-selection dynamics among chloramphenicol, oxytetracycline, and sulfonamide [30,44]. A non-significant poor association between tet(A) and tet(B) resistance genes among E. coli isolates (Supplementary Table S3) was observed which may be due to the incompatibility of plasmids carrying the tetracycline resistance determinants [30]. However, further study is required to enumerate the relationships among the resistance gene(s) and the probable link to antimicrobial exposure.

The findings of this study indicated that more caution are required for personnel hygiene in the processing and handling of poultry and poultry products to prevent the transfer of AMR E. coli from frozen poultry sold in supermarkets in Bangladesh. Present findings also highlighted the necessity of cautious use of antimicrobials in chickens to minimize the development of antibiotic-resistant bacterial strains. The study has limitations and these include a small sample size and fewer antibiotic-resistance genes tested. Further detailed investigation using a large number of samples, targeting more antibiotics including latest antibiotics as well as more ARGs, etc. would provide broader insights into the AMR patterns, prevalent ARGs, etc. among clinically important pathogens from food producing animals.