Molecular determination of antimicrobial resistance in Escherichia coli isolated from raw meat in Addis Ababa and Bishoftu, Ethiopia

Background Consumption of meat contaminated by E. coli causes a serious illness and even death to affected individuals. Recently the emerging of antibiotic resistant foodborne E. coli poses serious public health risks worldwide. However, little is known about the antibiotic resistance profile of E. coli in Ethiopia. This study aimed to determine the prevalence and Antimicrobial resistance (AMR) status of E. coli isolated from different type of meat. Methods Overall 292 samples were collected from December 2015 to April 2016 from slaughterhouses to determine the prevalence and AMR of E. coli isolated from raw beef, mutton, chevon and chicken meat from Addis Ababa and Bishoftu, Ethiopia. The isolates were screened for AMR against commonly used antibiotics circulating in the Ethiopian market. Both phenotypic and genotypic approach were employed for AMR detection using disc diffusion test and PCR respectively. Results The prevalence of E. coli was 63 (21.6%), indicating one sample in every five samples harbors E. coli. Among these, the highest E. coli isolates was observed in chicken meat samples (37.0%; 27), followed by mutton (23.3%; 17), chevon (20.6%; 15) and beef (5.5%; 4). Results of disk diffusion test on the 63 isolates showed that only 4.8% of them were not resistance to all antimicrobials tested. Multiple drug resistance (resistance to ≥3 drugs) was 46.0%. Significantly high resistance to ampicillin (71.4%) and tetracycline (47.6%) was observed. Identification of genes associated with AMR was also done using PCR. The prevalence of E. coli isolates harboring resistance gene responsible for tetracycline (tet(A)), beta lactams (blaCMY) and sulphanamide (sulI) antibiotics were found 65.1, 65.1 and 54.0%, respectively. Twenty-five out of the 63 (39.7% %) E. coli isolates have got antimicrobial resistance gene to three or more classes of drugs. The associations of antimicrobial resistance phenotypes and resistance genes was also determined. The detection of resistance trait against tetracycline, sulphametazole and chloramphenicol measured either phenotypically or genotypically were high. Conclusions The rising levels of resistance E. coli to multiple antimicrobial dictate the urgent need to regulate and monitor antimicrobial use in both animals and humans.


Background
Meat and meat products serve as important source of proteins for humans. However, recently the emerging antibiotic resistant foodborne pathogens combined with the injudicious use of antibiotics in animals bears considerate public health threats worldwide. Usually, meat and meat products gets contaminated by pathogens during animal slaughter and food processing. Escherichia coli (E. coli) is the frequently isolated foodborne pathogens from meat and meat products. Meat can be contaminated by E. coli during animal slaughter due to unhygienic slaughter practices, through airborne, rodents, insects, and other animals [1]. Consumption of meat contaminated by E. coli causes a serious illness and even death to affected individuals [2].
Recently published reports indicated that the E. coli strains isolated from contaminated meat and meat products have become resistant to commonly used antibiotics. This is mainly due to injudicious usage of antibiotics in both humans and animals [3]. The wide spread and imprudent use of antibiotics in food animals is thought to be accountable for the emergence and wider spreading of antimicrobial resistant (AMR) bacteria in humans [3,4]. In humans, positive selection for drug resistant bacteria have also been reported in the normal microflora of exposed individuals or populations. This indicates that antibiotic resistance can be developed in both commensal and pathogenic bacterial strains and can even be transferred to other bacterial strains, including other pathogenic and environmental bacteria [5].
Consumption of contaminated and/or uncooked meat poses the risks of acquiring foodborne E. coli strains [6] causing a serious public health concern. These bacterial strains easily harbors antibiotic resistant genes from one another. This is because genes encoding AMR determinants that are carried on mobile genetic elements such as plasmids and transposons of some bacterial strains could be transferred to other bacteria strains during contact [7], causing a threat to cure acute infections in man and animals. In Ethiopia, a review done by Alemu et al. [8] showed that several pathogen have established resistance against oxytetracycline drugs. Similarly, oxytetracycline and penstrep are the main prescribed antibiotics in Ethiopia [8,9]. So far, no study has been carried out to identify the antibiotic resistance genes of Escherichia coli in Ethiopia.
In this paper, we explain the phenotypic and genotypic sources of MDR in Escherichia coli isolates recovered from different type of meat samples taken from Addis Ababa city abattoir and Alema farm poultry slaughter slab, Ethiopia. This enquiry pursues to deliver useful evidence on the prevalence and AMR profile of E. coli isolated from beef, mutton, chevon and chicken meat.

Sample collection and isolation of E. coli
The study was carried out from December 2015 to April 2016 to collect meat sample from Addis Ababa city abattoir enterprise and Alema farm slaughter slab found in Bishoftu town, Ethiopia. A total of 292 meat sample including beef meat (n = 73), sheep meat (n = 73), goat meat (n = 73) were collected from Addis Ababa municipal abattoir and chicken meat (n = 73) from Alema farm slaughter slab located in Bishoftu city. Sample were collected by swabbing from carcass gluteal muscle and inserted into 10 ml peptone water containing test tube. The samples were taken following the guidelines of the International Organization for Standardization [10]. The sample were transported to National Agricultural Biotechnology Research Center in an ice box for further processing. The E. coli isolates were identified by using standard bacteriological methods, comprising of colony structure determination, using different culture media and biochemical tests as previously indicated [11]. Isolates were preserved at −80 °C in a trypticase soy broth with 10% glycerol for further analysis.

DNA extraction and PCR amplification of resistance genes
Bacterial strains were grown overnight in nutrient agar (HIMEDIA, India) at 37 °C. A loop full of the colonies was added to 100 μl of sterile water. Bacterial DNA was extracted by boiling a bacterial suspension in water. After boiling the suspension for 13 min, the suspension were frozen for 5 min in ice and centrifuged at 14,000 rpm for 15 min to pellet the cell debris [15]. The supernatant from the centrifuged tubes was transferred to new 1.5 ml clean plastic tube and used as a template for PCR amplification. The purified DNA were detected by electrophoresis in 1.5% agarose gel and then kept at −20 °C for further use. Primers for AMR genes such as streptomycin (aadA1), tetracycline [tet(A)], gentamicin [aac(3)-IV], sulfonamides (sul1), beta-lactams (blaSHV, blaCMY), erythromycin [ere(A)] and chloramphenicol (catA1, cmlA) were used from published article. The specific primer sequences (Bioneer, South Korea) and the estimated size of the amplified products for different resistance gene coding regions are found in Table 1. Amplification of antimicrobial resistance gene from E. coli isolates were performed as described by Fode et al. [16]. The amplification products were then separated by electrophoresis on a 1.5% agarose gel stained with gel red (Biotium Inc, USA) as described by Huang et al. [17] and visualized using UV illumination. A 100 bp DNA molecular marker (Bioneer, South Korea) were used to determine the size of the PCR product.

Statistical analysis
The prevalence of E. coli infection was quantified and compared among meat samples of different livestock species. Prevalence of AMR was quantified along with resistance patterns. The data were analyzed by using SPSS software version 20 and P value was calculated using Chi square and Fisher's exact tests to determine any significant correlation. P value less than 0.05 was considered statistically significant.

Prevalence of E. coli
In this study, a total of 63 (21.6%) E. coli isolates were identified from the 292 raw meat samples examined ( Table 2). Of these positive cases, chicken meat had the highest (37.0%) whilst beef meat had the lowest prevalence (5.5%) of E. coli. A significance difference in E. coli prevalence (P < 0.05) was observed among meat samples of different livestock species. In this regard, meat of chevon 2.8 times, mutton 3.2 times and chicken meat
Of 63 isolates, only 3 isolates (4.8%) were pan-sensitive and the rest were single drug resistance (12 isolates; 20.4%), double drug resistance (19; 30.2%) and multidrug (≥3 or more drugs) resistance (29; 46.0%). Two isolates demonstrated resistance to six different drugs. The AMR patterns and AMR frequencies were shown in Table 4. The most common AMR pattern found in the multiresistant isolates was ampicillin, tetracycline and erythromycin which jeopardizes the use of these drugs in the study area.

Molecular detection of antimicrobial resistance
All of the 63 E. coli isolates were confirmed by PCR to determine the existence of any AMR gene. The erythromycin esterase gene ere(A) and the aminoglycoside adenylyl transferases genes aadA1 were not identified in any of the 63 E. coli isolates (Fig. 1). The ere(A) gene confers resistance to erythromycin whilst aadA1 gene confer resistance to streptomycin. The most commonly detected AMR genes were blaCMY (65.1%), tetA (65.1%) and sul1 (54.0%). The distribution of resistance genes among E. coli isolates are summarized in Table 5. The bla SHV , β-lactamase gene and bla CMY genes were identified in 4.8 and 65.1% of the isolates, respectively. The aac(3)-IV gene which codes resistance against gentamycin was found also in 14.3% of isolates.

Relationship between drug resistance by disc diffusion and PCR methods
The most commonly drugs to which isolates demonstrated resistance by disk diffusion method were Table 3   consistent with the most common resistance genes detected (Table 7). Tetracycline, sulfamethoxazole and chloramphenicol were top three drugs recognized as being the most common for resistance measured either disk diffusion or molecular method. However, the phenotypic and genotypic resistance characterization of the isolates to erythromycin, streptomycin and gentamycin were not correlated.

Discussion
The magnitude of E. coli prevalence and its AMR patterns were investigated in raw meat samples of four livestock species at Addis Ababa and Bishoftu slaughter house. In our study, in raw meat, the prevalence of E.
coli was 21.6%. The study indicated that E. coli was 2.8 times, 3.2 times, and 5.7 times more prevalent in chevon, mutton and chicken meat, respectively, than in beef meat. It implies that meat of livestock species types has a difference in magnitude of E. coli infection, therefore, has implications on the risk to public health. It has been reported that poor hygienic practices is responsible for contamination of meat by E. coli [20]. The magnitude of E. coli prevalence in this study was similar to the previous studies in different part of Ethiopia 22.2, 26.6 and 20.3% by Haileselassie et al. [21], Haimanot et al. [22] and Bitew et al. [23] respectively. However, Mekonen et al. [24] and Taye et al. [25] also reported higher prevalence of carcass contamination with an overall isolation rate of 40 and 30.97%, respectively. The current study deals with the magnitude of E. coli only at one point (i.e. slaughterhouse) of the food chain until it reaches the fork. It shouldn't be overlooked that there could be ample chances for E. coli to infect the meat released from the slaughterhouse during transport, handling and processing chains before it reaches the consumer (fork) as hygienic practices are substandard in Ethiopia. It has been reported that poor hygienic practices is responsible for contamination of meat by E. coli [20].
In the current study, the prevalence of E. coli was significantly higher in chicken meat than in chevon, mutton and beef meat. This might be attributed to husbandry and production systems where, small holder farmers prefer to rear their chickens in backyard to scavenge. Thus, E. coli could be transmitted from intestinal normal flora of  livestock to the chicken via the food chain [26]. Over all, the variations observed among the reported prevalence could be emanated from difference in hygiene, breed, geographical origins of animals and history of treatment with antimicrobial. Emergence and dissemination of antimicrobial resistance is on the increase trend among enteric bacteria [27]. In this research the overall result showed that significantly high resistance rate to ampicillin (71.4%) and tetracycline (47.6%). More importantly, in the current report, chicken meat harbored the highest number of drug resistant E. coli isolates compared to the other meat origins. As expected, the most common resistance were found to older drugs such as ampicillin (Introduced in 1961) followed by tetracycline (Introduced in 1948) [28]. Similarly, Momtaz et al. [29] observed that tetracycline, sulfamethoxazole, chloramphenicol and trimethoprim resistance was the most common finding with the prevalence rate of 91.2, 45.6 and 29.8% respectively. Likewise, studies conducted in Korem, Ethiopia on diarrheic patients showed that the highest E. coli resistance were found against ampicillin followed by chloramphenicol, and tetracycline [30]. Hiko et al. [31] also found that the isolated E. coli is highly resistant to streptomycin, cephlaothin, tetracycline, ampicillin and trimethoprim. Another study revealed that all E. coli isolates from beef were found 100% resistance to ampicillin and amoxicillin and 33.33% to tetracycline [25]. The observed higher level of antimicrobial resistance might be attributed to the widespread and indiscriminate use of antibiotics in animals for medication and other prophylaxis purposes. Although tetracycline has the second most resistance in this study, it is one of the most commonly used antibiotics for the treatment of different infections, including E. coli in Ethiopia.
Recently, among gram negative bacteria, multidrug resistant (MDR) phenotypes are spreading widely [32]. In the current study, the overall rate of multiple drug resistance was 46.0% and only 4.8% of the isolates were found sensitive to all antimicrobials tested (Table 4). This is in agreement with the findings of Schroeder et al. [33] and Zhao et al. [34] from USA that reported that multiple resistance to tetracycline, kanamycin, streptomycin, ampicillin, and sulfamethoxazole. Similarly, among sulfonamide resistant E. coli isolates, Wu et al. [35] reported that ampicillin and streptomycin are the two   most frequently co-transferred resistance phenotypes. However, Daniel et al. [28] reported a resistance profile of tetracycline (80%) and streptomycin (74%), among the sulfonamide-resistant E. coli isolates. The transfer of resistance among microorganisms has long been recognized as a serious threat, contributing to the evolution and emergence of antibiotic resistant bacteria, thereby reducing the therapeutic potential against pathogens [36]. After phenotypic screening, genes associated with antimicrobial resistance was analyzed using PCR. In this study, the most commonly detected resistance genes were blaCMY (65.1%), tetA (65.1%) and sul1 (54.0%)) for cephalosporin, tetracycline and sulfonamide respectively. However, streptomycin and erythromycin resistance genes were not detected. In Iran, among the E. coli isolates, 52.6% of resistance to tetracycline as well as 47.4% for both sulfonamides and erythromycin was reported [31]. The prevalence of different resistance genes from pediatric patient E. coli isolate were also reported 85.06, 60.38, 57.79, 90.25, 40.25 and 54.54% positive for tetA, cmlA, SHV β-lactamase, CITM, sul1 and aac(3)-IV resistance genes respectively [37].
The phenotypic and genotypic resistance patterns with in the same E. coli isolates might be strongly correlated [38]. In our research, detection of tetracycline resistance by disk diffusion test and identification of its resistance gene tet(A) by PCR test, sulfamethoxozole-trimethoprim and sulI, chloramphenicol and cmlA were strongly associated. But there was a lack of association between phenotypic erythromycin resistance and molecular detection of ere(A), streptomycin and aadA1, chloramphenicol and catA1 resistance gene. Similarly, reports from Thailand showed that, E. coli isolates from chickens were found to be resistant to tetracycline (90%) and erythromycin (73.3%) in Agar disk diffusion assays and these resistance properties were associated with tet(A) and ere(A) resistance genes respectively [4]. However, the presence of resistance phenotypes might not represent all the underlying resistance genes. Alternatively, the presence or absence of a resistance gene might not indicate the particular isolate is resistant or susceptible to an antimicrobial [39].
The observed discrepancy in the resistance level of genotypes and phenotypes might be attributed to either of or combination of the reasons including but not limited to the presence of unexpressed genes in the bacterial isolates, not containing the possible resistance genes in the test, use of incorrect cut points of test results used for resistance and susceptibility classification, or some of resistance phenotypes are caused by due to point mutations rather than gene transfer or acquisition [40].

Conclusions
This study revealed that raw meat available for consumers in Ethiopia was often contaminated with E. coli. We reported on a comprehensive study of the phenotypic and molecular basis for Multi Drug Resistance (MDR) in meat E. coli isolates recovered from Ethiopian Abattoir. Furthermore, a high rate of resistance to Ampicillin and resistance to more than one class of antibiotics among E. coli isolates was found. Additionally, several strains were found positive for the bla CMY , tet(A), and sul1 antimicrobial resistance genes. All these findings suggest that the consumption of undercooked meat or food cross-contaminated with E. coli may pose a serious threat to consumer health.   cmlA (n = 7; 11.1)