1 Introduction

Antimicrobial resistance (AMR) is a growing global concern and poses a grave threat to public health (Robinson et al. 2016). The void in the development of new therapeutic substances has exacerbated the problem, and as a result, very few antimicrobials are available to treat infections caused by multidrug-resistant (MDR) pathogens (Robinson et al. 2016). Numerous reports have described the presence of large amounts of antimicrobial resistant bacteria and antimicrobial resistant genes (ARGs) in various food products, such as meat, fish, vegetables, ready-to-eat (RTE) foods, and fruits (Rajaei et al. 2021; Vuthy et al. 2017; Amer et al. 2018). This presence of MDR bacteria in the food chain is a rising public health concern. Indeed, besides contributing to the spread of foodborne pathogens, an additional concern is that food represents a vehicle for the transmission of AMR determinants to humans, which may occur through contact or consumption of contaminated food products (Amer et al. 2018).

Escherichia coli (E. coli) is a ubiquitous rod-shaped Gram-negative bacterium abundant in the gastrointestinal tract of humans and animals, where it plays a crucial role in digestion and absorption (Kaushik et al. 2018). However, E. coli can infect animals and humans, causing diseases ranging from opportunistic wound infections to severe systemic infections. Furthermore, E. coli can easily spread through the food chain and function as ARG carriers/reservoirs, sharing them with other bacteria of the same or other species (Kaushik et al. 2018). As a result, E. coli strains are frequently used as an appropriate indicator bacterium in AMR surveillance, as changes in these species can serve as an early warning of the emergence of resistance in similar pathogenic bacteria (Guo et al. 2019). Resistant E. coli strains containing various ARGs have been extensively investigated, mainly in industrialized nations (Balbin et al. 2020; Li et al. 2020a). Most of the AMR research in Burkina Faso has focused on clinical samples from humans (Ouedraogo et al. 2016; Dissinviel et al. 2017; Kpoda et al. 2018; Dembélé et al. 2020). There is limited information on the antimicrobial resistance and molecular characteristics of E. coli strains isolated from food sources throughout our country. Accordingly, in this study, we performed antimicrobial susceptibility and molecular characterization of resistance genes from E. coli strains obtained from various foods controlled by Burkina Faso’s National Food Safety Agency. All isolates were tested for antimicrobial susceptibility using the disk diffusion method, followed by a polymerase chain reaction (PCR) assay to screen 11 ARGs and integrons. To our knowledge, this report is the first to investigate the role of E. coli in the propagation of AMR in the food chain in Burkina Faso. The results call for strengthening the monitoring of foodborne E. coli in the Burkina Faso food chain to protect human health.

2 Material and methods

2.1 Bacterial isolates and storage

This study included 155 E. coli isolates from food samples obtained from the Burkina Faso National Food Safety Agency surveillance activities between 2018 and 2021. All samples received were processed for isolation and identification of E. coli on MacConkey agar (HiMedia, Mumbai, India) and Eosin methylene blue (EMB) agar (HiMedia, Mumbai, India) followed by biochemical characterization by standard procedures. Isolated strains were stored at -80 °C in brain heart infusion (BHI), (HiMedia, Mumbai, India) medium containing 15% (v/v) glycerol (HiMedia, Mumbai, India). For the purpose of this study, all frozen E. coli isolates were reactivated from cryovials on Nutrient Agar (HiMedia, Mumbai, India) at 37 °C for 18–24 h. The identification of E. coli was confirmed by Gram staining and biochemical characterization using API 20E kits (BioMérieux, Marcy l’Etoile, France). The isolates were recovered from 7 types of food samples, including dairy foods (29), fish-based dishes (7), mango juices (4), lettuces (35), RTE salads (21), sandwiches (32) and sesame (27) (Table 1).

Table 1 Percentages of E. coli isolates resistant to antimicrobials from various food sources
Full size table

2.2 Antimicrobial susceptibility tests

The disk diffusion method on Mueller-Hinton (MH) agar (HiMedia, Mumbai, India) was used for antimicrobial susceptibility testing, as recommended by the Clinical and Laboratory Standards Institute (CLSI, 2017). In brief, a single bacterial colony from nutrient agar was randomly selected and suspended in 0.85% sterile saline (HiMedia, Mumbai, India) to achieve an inoculum equivalent to 0.5 McFarland (HiMedia, Mumbai, India). A sterile cotton swab (HiMedia, Mumbai, India) was immersed in the 0.5 McFarland suspension and spread on MH plate. The plate was then allowed to dry completely at room temperature (no more than 15 min). Antimicrobial disks (Condalab, Madrid, Spain) were then applied to the MH (HiMedia, Mumbai, India) agar surface with sterile fine curved stainless-steel splinter forceps at a distance to avoid overlapping of the inhibition zones. The following antibiotics were used: cefotaxime (CTX, 30 µg), ceftazidime (CAZ, 30 µg), cefepime (FEP, 30 µg), aztreonam (ATM, 30 µg ), amoxicillin/clavulanic acid (AMC, 30 µg), ampicillin (AMP, 10 µg), cefoxitin (FOX, 30 µg), chloramphenicol (C, 30 µg), trimethoprim-sulfamethoxazole (STX, 1.25/23.75 µg), nalidixic acid (NA, 30 µg), tetracycline (TE, 30 µg), imipenem (IMP, 10 µg ), meropenem (MEM, 10 µg), gentamicin (CN, 10 µg), amikacin (AK, 30 µg), and ciprofloxacin (CIP, 10 µg). The inoculated plates were incubated at 37 °C for 24 h and the inhibition zones were measured in millimeters across the diameters. The standard strain of E. coli ATCC 35218 was used as a quality check. Isolates exhibiting resistance to 3 or more classes were considered multidrug-resistant (MDR) (Magiorakos et al. 2012). Multiple antimicrobial resistance (MAR) indexes were calculated using Krumperman’s approach (Krumperman 1983).

2.3 DNA extraction

The heat lysis method (Yang et al. 2019) was used to extract genomic DNA from resistant bacteria, which was then used as a PCR template for all assays. Briefly, a single pure bacterial colony from a Nutrient agar (HiMedia, Mumbai, India) plate was subcultured in 1 ml of BHI (HiMedia, Mumbai, India) overnight. The next day, each culture broth was centrifuged at 13,000 rpm for 5 min at 4 °C, and the cell pellets were resuspended in 500 µl of sterile distilled water and heated at 100 °C in a water bath for 10 min. The cell suspension was immediately frozen at -20 °C for 10 min and then centrifuged at 13,000 rpm for 5 min at 4 °C. The supernatant was transferred to a new Eppendorf tube and conserved at -20 °C until use.

2.4 Identification of resistance genes and integrons

The presence of antimicrobial resistance genes among the resistant isolates was determined using the uniplex PCR tests to amplify ARGs. 11 pairs of oligonucleotide primers (Genecust, Paris, France) were used to target antimicrobial resistance genes that confer resistance to antimicrobial agents, including ampicillin (blaTEMblaSHV, temA, temB), tetracycline, [tet(A), tet(B)], aminoglycosides (StrA, aadA), quinolone (GyrA) and trimethoprim-sulfamethoxazole (sul1, sul2). The PCR cycling conditions for reactions were set at 94 °C for 3 min, followed by 30 cycles of 1 min at 94 °C, 30 s at annealing temperature and 1 min at 72 °C and final step for 7 min at 72 °C. Furthermore, uniplex PCR was conducted to detect the presence of intI1 integrase genes in the 65 isolates resistant to at least 2 antimicrobials, to detect the presence of Class 1 integrons. The thermal cycling conditions included preincubation at 94 °C for 1 min, followed by 35 cycles of denaturation at 98 °C for 30 s, annealing at 60 °C for 30 s, and polymerization at 72 °C for 30 s, and final extension at 72 °C for 10 min. Subsequently, the variable regions (VRs) of the strains that were positive for the intI1 gene were evaluated by uniplex PCR using the primers hep58 and hep59 with the following cycling conditions: preincubation at 94 °C for 1 min, followed by 35 cycles of denaturation at 98 °C for 30 s, annealing at 55 °C for 30 s, and polymerization at 72 °C for 4 min and final extension at 72 °C for 10 min. The primer (Genecust, Paris, France) sequences, the annealing temperatures, expected product sizes, and references are presented in Table 2. The PCR was done using a TECHNE Prime Elite thermal cycler (Bibby scientific, Staffordshire, United Kingdom). Appropriate positive and negative controls for amplifications were used in each PCR procedure. All PCR amplifications were performed in a reaction mixture comprising of 4 µl FIREPol® Master Mix (Solis BioDyne, Tartu, Estonia), 0.4 µl of each primer, 3 µl of DNA template and nuclease-free water to make volume up to 20 µl. Amplified PCR products were run on a 1.5% agarose gel containing 0.5 µg/ml of ethidium bromide (ThermoFisher, Waltham, United States) for 60 min at 100 V, then visualized using the gel documentation system E-BOX VX5 (VILBER, Marne-la-Vallée, France).

Table 2 Primer sequences, annealing temperature, product size and source of primers used during this study
Full size table

3 Results and discussion

In this study, we characterized the antimicrobial resistance profiles and ARGs of a collection of 155 E. coli isolates recovered from sandwich, dairy food, fish, sesame, lettuce and RTE salad at the Burkina Faso National Food Safety Agency. To our knowledge, this is the first time a study was conducted with many E. coli isolates from various food sources in our country.

3.1 Antimicrobial susceptibility of E. coli isolates

Among the E. coli isolates used in the present study, 105 (67.7%) were resistant to at least 1 of the 16 antimicrobials used (Table 1). A similarly high level of resistance (74.2%) was reported in isolates collected from RTE foods in 2017 in Shaanxi Province, Republic of China (Baloch et al. 2017). In our study, resistances to tetracycline, ampicillin, gentamycin, amoxicillin-clavulanate acid, nalidixic acid, chloramphenicol, and trimethoprim-sulfamethoxazole were generally observed. These findings are consistent with previously published studies showing that E. coli isolates from foods were resistant to several older antimicrobials, including tetracycline, streptomycin, sulfonamides, ampicillin, gentamicin, and nalidixic acid. Vuthy et al. (2017), reported high resistances to tetracycline, amoxicillin, and sulfamethoxazole (63.1–76.1%) of E. coli isolated from chicken food chains in Phnom Penh (China). In addition, E. coli isolated from RTE foods (Singapore) during a study of Guo et al. (2019) showed resistances to tetracycline (17.2%), ampicillin (15.2%) and chloramphenicol (10.1%). In a study conducted by Ayamah et al. (2021) in Ghana, all isolates obtained from kebabs (beef, chevon, and gizzard) were resistant to ampicillin, tetracycline, and gentamicin. Furthermore, Somda et al. (2018) found that E. coli isolated from RTE chickens in Burkina Faso were resistant to tetracycline (64.3%), ampicillin (42.8%), amoxicillin + clavulanic acid (46.3%), and trimethoprim-sulphamethoxazole (32.14%). These traditional antimicrobials have been overused and abused throughout the years by consumers, veterinarians, farmers, physicians, and patients due to their low prices and easy availability over the counter, rendering them empirically ineffective (Guo et al. 2019). Interestingly, our study did not find an E. coli isolate resistant to cefotaxime, ceftazidime, cefepime, and aztreonam (ESBL-producing strain). On the contrary, resistance to these antimicrobials was observed among E. coli isolated from street foods in India (Sivakumar et al. 2021). Additionally, a study of street food in Tamale (Ghana) showed that 55.4% (41/74) of the isolated E. coli were ESBL producers (Karikari et al. 2022). This disparity may be attributed to the sample sources, that is, foods of plant origin in our case and street foods in the others. In addition, bacteria found in meat (especially meat poultry origin) can be resistant to a wide range of antibiotics, including quinolones and the third generation of cephalosporins (Guo et al. 2019). It is important to note that none of the isolates in this study were resistant to carbapenems (imipenem and meropenem). Until recently, carbapenems, such as meropenem, imipenem, ertapenem, and doripenem, were the antibi-otics of last choice for treatments of MDR pathogenic bacteria (Osei et al. 2016). As a result, prescribers should not abuse these antibiotics to avoid the emergence and development of resistances. In our study, the E. coli isolates also showed low resistance to ciprofloxacin (8.4%). This finding is of public health significance, as they are critically important for the treatment of serious human infections. In developed countries, the use of fluoroquinolones has been banned for use in animal husbandry; therefore, E. coli isolated from this source showed very little resistance to these antibiotics (Boerlin et al. 2005).

3.2 Profile of MDR E. coli isolates

Our study revealed a moderate percentage of MDR strains among E. coli (26.5%). MDR strains are isolates resistant to 3 or more different classes of antibiotics (Magiorakos et al. 2012). Previous studies have reported that multidrug resistance is common in poultry and other food animals (Li et al. 2020a; Abatcha et al. 2018; Adzitey et al. 2020). The presence of MDR bacteria in foods (meat, vegetables, fruit and RTE foods) is a public health concern since they can be transmitted to humans directly or indirectly by eating contaminated RTE foods. The MDR bacteria identified in the present study may result from environmental contamination or transmission by food handlers due to poor personal hygiene and inadequate sanitation in the processing plant. Therefore, the microbiological quality of food should be monitored to ensure public health safety. As shown in Table 3, this study also revealed that the multiple antibiotic resistance (MAR) index ranged from 0.13 (resistant to 2 antibiotics) to 0.47 (resistant to 7 antibiotics). 19 resistant isolates (18.7%) had a MAR index > 0.2. Bacteria that have a MAR index > 0.2 originate from a high-risk source of contamination where antibiotics or growth promoters are frequently used (Krumperman 1983). Overall, the extensive use by humans and livestock production, particularly in poultry, may have led to the rise of resistance (Rajaei et al. 2021). Therefore, One Health strategies which require multidisciplinary collaborations, adequate surveillance systems, and strong laboratory capacity are needed in order to reduce antibiotic misuse.

Table 3 Antimicrobial patterns and multiple antimicrobial resistance index of E. coli isolates
Full size table
Table 4 Occurrence of resistance genes among E. coli
Full size table

3.3 PCR-based detection of antimicrobial resistance genes

The result of PCR screening for the presence of resistance genes among phenotypically resistant E. coli isolates is shown in Table 4. Among drug resistant isolates, PCR tests detected 9 genes (blaTEM, blaSHV, temB, tet(A), tet(B), aadA, GyrA, sul1, and sul2). In the present study, detection of beta-lactamase genes among ampicillin resistant isolates revealed the predominance of blaTEM and blaSHV genes (37.3% each), followed by the temB genes (33.3%). Beta-lactamase, which deactivates beta-lactam antibiotics by hydrolysis of the beta-lactam ring, is one of the most significant enzymes implicated in bacterial resistance to antibiotics (Shaikh et al. 2015). Although numerous beta-lactamases have been identified, TEM and SHV are reported to be the main causes of bacterial resistance to beta-lactam antibiotics (Hassani et al. 2022). All beta-lactamase genes detected during our work have been widely documented in the community, clinical isolates, and food-producing animals (Guo et al. 2019). The tetracycline resistance genes tet(A) and tet(B) were found in 46.2% and 17.3% of the isolates tested in this study, respectively. This finding supports previous research that identified tet(A) genes as the most common tetracycline resistance genes (Vuthy et al. 2017; Guo et al. 2019; Dissinviel et al. 2017; Abdelwahab et al. 2022). Currently, 30 different tetracycline resistance genes (tet genes) have been identified (Eliopoulos et al. 2003). All tet efflux genes encode membrane-associated energy-dependent proteins, which export tetracycline out of the cell. Among our isolates resistant to trimethoprim-sulfamethoxazole, 44.4% carried sul1 genes, while 22.2% harbored sul2 genes. These 2 genes that express highly resistant dihydropteroate synthases to sulfamethoxazole are carried by transposons and plasmids (Roberts 1996). However, almost 20 different resistance genes that express drug-insensitive dihydrofolate reductase are spread as cassettes in integrons, transposons, and plasmids (Nelson et al. 2019). In this study, 86% of E. coli resistant to quinolones carried the GyrA gene. Overexpression of naturally occurring efflux pumps, resistance encoded by the chromosome, mutations in molecular target DNA gyrase and topoisomerase IV, and resistance mediated by plasmids are molecular mechanisms of bacterial resistance to quinolone drugs (Abdelwahab et al. 2022). However, the most common mechanism of resistance to fluoroquinolones are mutations of target genes such as GyrA, parE, and parC that encode type II topoisomerases (Guo et al. 2019; O’Bryan et al. 2018).

Table 5 Distribution of antimicrobial resistance, class1 integrons among the 65 isolates resistant to at least 2 antimicrobials
Full size table

In summary, it is worth noting that several genes can be involved in AMR. In this study, we did not evaluate the entire suite of resistance genes. Therefore, some resistant isolates did not carry any of the AMR genes tested, indicating that other AMR genes may be present or there are other novel genetic resistant determinants (Abdelwahab et al. 2022). Therefore, isolates that were phenotypically resistant but negative for PCR should be screened using novel techniques such as metagenomic analysis based on high-throughput sequencing (HTS) to provide more comprehensive information on their antibiotic resistomes (Li et al. 2020b).

3.4 PCR-based detection of class 1 integrons genes

The present study has shown the presence of MDR among E. coli isolates, as mentioned above, suggesting the existence of mobile genetic elements such as integrons, plasmids, and transposons (Deng et al. 2015). Of these, integrons play a central role in the spread of ARGs among gram-negative bacteria (Kaushik et al. 2018). Integrons capture gene cassettes using an integron-integrase and then express cassettes using an integron encoded promoter. There are many types of integrons based on the integrase gene (intI), but most of the integrons found in clinical isolates belong to Class 1 (Deng et al. 2015). Table 5 shows the distribution of antimicrobial resistance, class-1 integrons in the 65 isolates resistant to at least 2 antimicrobials. All the isolates were positive for class 1 integrons, which is consistent with previous findings (Zhang et al. 2019; Malek et al. 2015). Particularly in Burkina Faso, class 1 integrons were found in 65% of Salmonella isolated from environment and diarrhea stool samples (Somda et al. 2021). Subsequently, strains harboring intI genes were elaborated for the amplification of variable regions (VRs). The occurrence of VRs within class 1, integron positive isolates were 32.3% (21/65 strains). The integrons that did not carry gene cassettes are called empty integrons and can have the potential to rapidly convert themselves into MDR strains (Mohamed et al. 2020; Fang et al. 2019).

3.5 Limitations

This study has some limitations due to a lack of financial resources. Firstly, appropriate tests were not done to confirm the PCR product identity. The confirmation tests could be: (i) DNA sequencing of the PCR product; (ii) hybridization of the PCR product with specific DNA probes; (iii) or restriction analysis of the PCR product. Secondly, we have focused exclusively on the antimicrobial resistance in E. coli isolates, and the pathogenicity of the collected E. coli was not assessed.

4 Conclusion

This study generated preliminary data on the resistance of E. coli identified in the food chain of Burkina Faso. Phenotypic characterization revealed resistance to tetracycline, ampicillin, cefoxitin, gentamicin, clavulanic acid, and sulfamethoxazole/trimethoprim. In addition, isolates displayed multidrug resistance characteristics and revealed possible molecular resistance mechanisms. Therefore, a prospective study with a well-designed geographic distribution of the samples is required.