Occurrence and antimicrobial susceptibility profile of Escherichia coli O157:H7 from food of animal origin in Bishoftu town, Central Ethiopia

Background: Escherichia coli O157:H7 ( E. coli O157:H7) have frequently been associated with food borne illness and are considered as most serious of known food borne pathogens leading to severe illnesses and high mortality rates in humans. Most of outbreaks were traced to raw meat and raw milk consumption, as well as to dairy products such as yogurt and cheese derived from raw milk. Results: Out of 200 samples examined, 40 (20%) and 7 (3.5%) of the samples were positive to E. coli and E. coli O157: H7 respectively. The highest isolation of E. coli was from cheese (40%), followed by raw milk (32%), yogurt (25.71%), beef (13.84%), and pasteurized milk (0%). Among E. coli O157:H7 isolates, the highest isolation was from raw milk (12%) followed by cheese (5.71%) and meat (3.07%). However, no E. coli O157:H7 was isolated from pasteurized milk and yogurt. Antibiotic susceptibility profile showed that E. coli was resistant for vancomycin (89.74%), ampicillin (76.92%) and streptomycin (69.23%). The analysis showed that, 92.5% of isolates showed multidrug resistance comprising 2 – 4 antimicrobials. Conclusion: The occurrence of E. coli O157:H7 and its multiple antibiotic resistant profiles shows a risk for public health and food safety as well as animal production. These findings stress the need for an integrated control of E. coli O157:H7 from farm production to consumption of food of animal origin. resistant, intermediate or susceptible.


Background
Foodborne diseases and food poisoning are the widespread and great public health and well-being concerns of individuals and countries of the modern world. Especially, developing countries are largely affected by food-borne infections (Carbas et al. 2012). Among the major infectious agents, Escherichia coli O157:H7 has frequently been associated with foodborne illness. Particularly, over the past decade, E. coli O157:H7 has been reported increasingly from all parts of the world and in the worst case, it is "one of the most serious" foodborne pathogens leading to severe illnesses and high mortality rates in humans (Blanco et al. 2003;Jo et al. 2004). This consideration is in fact due to the small infectious dose of the organism because fewer than 40 cells of E. coli O157:H7 can cause illness in some people (Strachan et al. 2005).
It has been indicated that an estimated 74,000 cases and 61 deaths annually are attributable to E. coli O157:H7 in the USA, and many outbreaks (in the USA) related to foodborne illness have been connected to consumption of contaminated foods derived from cattle, especially meat and raw milk. In the 1980s, most outbreaks due to E. coli O157:H7 were associated with inadequately cooked hamburgers and raw milk. Later, outbreaks were traced to other dairy products such as yogurt and cheese (Doyle et al. 2006;Mora et al. 2007). More recently, in 2016 outbreak of E. coli O157:H7, slaughtered animals were the R E T R A C T E D A R T I C L E main sources of infection and led to illness of eleven people in the USA (CDC, 2016).
Escherichia coli O157:H7 has been found in the intestines of healthy cattle, deer, goats, and sheep. However, cattle have been identified as a major reservoir of E. coli O157:H7 and consumption of foods of bovine origin such as beef and dairy products have been associated with some of the largest food poisoning outbreaks in which this organism was identified as the etiologic agent (Acha and Szyfress, 2001; IFT (Institute of Food Technology), 2003; Perelle et al. 2007).
Due to an increased demand for animal protein, the animal production sectors in low and middle-income countries have been regularly using antimicrobials for therapy, disease prevention and growth (Van Boeckel et al. 2015). This practice could be responsible for antimicrobial resistance among commensals in the intestinal tracts of food animals, which may subsequently risk public health due to food animals' weak response to, or loss of response to, drug therapy. Hence, there should be isolation of pathogenic organisms and regular evaluation of their antimicrobial susceptibility profiles. In Ethiopia, some studies have been conducted to identify pathogenic E. coli from human and animal sources such as stool samples (Demisse, 2005), raw beef, sheep meat, goat meat (Hiko et al. 2008;Lula 2011), feces, skin of meat handlers (Mersha et al. 2010), yogurt and cheese (Tsegaye and Ashenafi 2005). However, recent and detailed information on the prevalence and multi-drug susceptibility profile of pathogenic E. coli is limited. Therefore, the present study was conducted to add current information pertaining to the occurrence and antibiotic susceptibility profiles of E. coli and E. coli O157:H7 from milk, milk products and meat in and around Bishoftu, Central Ethiopia, where food of animal origin is widely consumed.

Study area
The study was conducted in Bishoftu town. Bishoftu town is located at 9°N latitude and 40°E longitudes at an altitude of 1850 m above sea level in central high lands of Ethiopia. It has an annual rainfall of 866 mm of which 84% is in the long rainy season (June to September). The dry season extends from October to February. The mean annual maximum and minimum temperatures are 26°C and 14°C respectively, with mean relative humidity of 61.3% (ADARDO 2007). The livestock production system in the area is both intensive and extensive type (CSA 2015).

Study design and sampling strategy
A cross-sectional study was conducted from November 2016 to April 2017 to determine the occurrence and antimicrobial resistance profile of E. coli O157:H7 in/for milk, milk products (cheese and yogurt) and beef samples. In the present study 200 samples (milk = 65, cheese = 35, yogurt = 35 and meat = 65) were collected on a voluntary basis (owner's willingness to provide the samples). Cafeterias, restaurants, open markets and supermarket that had a high level of consumers were included in the study.

Collection and transportation of samples for laboratory analysis
About 20 ml of milk (both pasteurized and raw), cheese and yogurt samples were collected aseptically in sterile disposable corked plastic tubes. The pasteurized milk, cheese and yogurt obtained from the cafeterias, restaurants, and supermarket were kept under refrigerator until used for consumption by customers. The pasteurized milk was packaged using a disposable small plastic bag, whereas the cheese and yogurt were kept in silver/glass vessels until used for consumption. The raw milk samples were obtained from milk sellers found in open markets (the streets of the town). Milk found on the open markets was handled with a plastic container of up to 3 litters' capacity and with no cooling facility. About 25 g of beef meat sample was taken from carcass hanged inside the houses of restaurants and placed in a disposable plastic bag. The entire collected samples were labeled appropriately, placed in a box containing ice and transported immediately to Microbiology Laboratory, College of Veterinary Medicine and Agriculture, Addis Ababa University. Then the samples were placed in a refrigerator at +4°C and subjected to culture within 24 h of sampling.
Isolation and identification of Escherichia coli and Escherichia coli O157:H7 Detection of E. coli and E. coli O157:H7 was carried out according to the protocol of ISO-16654: 2001 standard. A loopful of milk, cheese and yogurt aseptically taken from all of the sample bottles and a swab from the surface of about 25 g portions of meat dissected by sterilized blade from all of the meat samples collected were individually inoculated on MacConkey agar for primary isolation of E. coli (Difco laboratories, USA) and incubated aerobically at 37°C for 24 h. The plates were observed for the growth of E. coli (pink colony; lactose fermenter). A single, isolated colony was picked and sub-cultured on Eosin Metyline Blue (EMB) agar for formation of metallic sheen. Simultaneously another single colony with similar characteristics was picked and stained with Gram's stain. The isolate was examined for stain and morphological characteristics using bright-field microscopy. KOH test was then employed to confirm the Gram's reaction (Quinn et al. 2004). Suspected colonies of E. coli (pinkish color appearance on MacConkey agar and metallic sheen on EMB) (Figs. 1 and 2) were then sub-cultured onto blood agar to appreciate colony characteristics and then pure colonies taken from blood agar were inoculated on nutrient agar (OXOID) (non-

Statistical analysis
The collected data for bacterial contamination analysis were entered and analyzed using SPSS version 17 computer software. Accordingly, descriptive statistics such as percentages and frequency distribution were used to describe/present bacterial isolates and antimicrobial susceptibility which was expressed as percent of resistant, intermediate or susceptible.
Multidrug resistance analysis showed that, 37/40 (92.5%) of tested E. coli isolates were resistant to different combinations of two or more antimicrobials (Table 4) and the proportion was higher in milk and milk products (28.4%) than meat samples (15.4%) ( Table 5). A multidrug resistance pattern consisting of four drugs was seen in 3/40 (7.5%) isolates. Moreover, the majority of the isolates 16/40 (40%) showed multidrug resistance to Ampicillin, Vancomycin and Streptomycin. All isolates of E. coli O157:H7 were resistance to at least two drugs and 14.4% of them showed resistance to Ampicillin, Vancomycin, Streptomycin and Tetracycline.

Discussion
The present study revealed that E. coli was isolated from 20% of ready to eat foods of animal origin (milk, milk products and meat). Meanwhile, the study confirmed that E. coli and E. coli O157:H7 were not found in pasteurized milk. The presence of E. coli in pasteurized milk doesn't reflect the survival of the organism to the appropriate level of pasteurizing temperature. Rather, it   (2014) showed that 23.7% samples from food of bovine origin harbored E. coli. Furthermore, 32% of raw milk samples were found to harbor E. coli, which is somewhat in agreement with the report of 33.9% by Disassa et al. (2017). However, the prevalence is far lower when compared to the reports of Shunda et al. (2013) from Mekelle town (44%) and far higher when compared to 26% prevalence reported by Farhan et al. (2014) and 23.3% by Elbagory et al. (2016). In the present study, the isolation rate of E. coli O157:H7 from raw milk was 12%, which is comparable to prevalence report of 10.4% by Mekuria and Beyene (2014). Whereas, the highest occurrence of E. coli O157:H7 were found by Chye et al. (2004) (33.5%) and Lye et al. (2013) (18.75%) in Malaysia. This might be due to differences in animal management, milking systems, and milk handling practices among different countries.
In the present study, 5.71% isolation rate of E. coli O157:H7 was recorded from cheese sample. This rate is slightly higher than the report of Sancak et al. (2015) with 2% prevalence. In the study of Zelalem et al. (2015), E. coli O157:H7 was found to survive the manufacturing of Ayib (Ethiopian cottage cheese). In Ethiopian cottage cheese, complete inactivation of the organism occurred after 20 and 40 min of cooking at 70°C, indicating that if there is under treatment of heat, the cheese can act as source of Escherichia coli O157:H7 (Zelalem et al. 2015). Furthermore, Spano et al. (2003) stated that, cheese could be free of E. coli O157:H7 if high temperature is used during milk processing. Furthermore, in some types of cheese like Cheddar cheese, E. coli O157:H7 has the ability to grow during the manufacture of the cheese and it could be detected by enrichment after 60 days of   (Reitsma and Henning 1996). In addition, Ramsaran et al. (1998) observed a significant increase in the number of E. coli O157:H7 during the manufacture of Camembert cheese, and stabilized number of colony forming units can be found after 75 days, indicating the potential for survival in this type of cheese. The other finding of the present study is that Escherichia coli O157:H7 was not isolated from yogurt (Ethiopian naturally fermented milk) samples. Contrarily, Vahedi et al. (2013) reported 9% prevalence of Escherichia coli O157:H7 in yogurt samples and Zelalem et al. (2015) indicated that E. coli O157:H7 was found to survive the manufacturing of Ergo (Ethiopian naturally fermented milk). However, the absence of E. coli O157:H7 from yogurt is partly supported by the study of Osaili et al. (2013), who found that E. coli O157:H7 increased during fermentation and the population of E. coli O157:H7 decreased slightly during cooling. In connection to this, Osaili et al. (2013) indicated that lowering the temperature during cooling may lead to the increased susceptibility of E. coli O157:H7 to an acid environment and the population of E. coli O157:H7 during storage at +4 oC decreased sharply. It was evident that almost all cafeterias in the study area had refrigeration, and this could partly contribute for the absence of the isolates in the yogurt samples. Overall, the variation in the prevalence reports of the organism from cheese and yogurt samples could be due to differences in procedures followed during preparation of the dairy products, as well as improved enrichment and isolation procedures.
The use of antibiotics in the treatment of E. coli O157:H7 infection is controversial, since antimicrobial therapy may increase the risk of development of hemolytic uremic syndrome (Molbak et al. 2002). Although some studies do not advise antibiotic treatment for infections caused by such bacteria, others suggest that disease progression may be prevented by administrating antibiotics during the early stage of infection (Schroeder et al. 2002). Thus, for the better response, an antimicrobial susceptibility test is necessary (Quinn et al. 2011). Hence, on the basis of this necessity, antimicrobial susceptibility testing was conducted on the isolates recovered from all the samples.
The present study showed that E. coli isolates were highly sensitive to ceftriaxone, gentamicin, ciprofloxacin, chloramphenicol and tetracycline. Meanwhile, the majority of the isolates were resistant to ampicillin, streptomycin and vancomycin. Similarly, Hiko et al. (2008) and Bekele (2012) from Ethiopia and Magwira et al. (2005) from Botswana revealed that the resistance of E. coli does exist mainly to ampicillin and streptomycin. However, various authors reported that E. coli is resistant to tetracycline (Hiko et al. 2008;Bekele 2012;Mude et al. 2017), which is contrary to the results of the present study. But in Dire Dawa, Mohammed et al. (2014) reported that E. coli was susceptible to tetracycline, which is in line with the present study finding.
Multidrug resistance analysis showed that 37/40 (92.5%) of tested isolates were resistant to different combinations of two to four tested antibiotics. This is in agreement with the report of Mude et al. (2017), who showed 92.3% of isolates were multidrug resistant. Moreover, various authors (Bekele et al. 2014;Iweriebor et al. 2015;Atnafie et al. 2017) from the country and abroad reported multidrug resistance patterns. Moreover, the present study revealed that the prevalence of multidrug resistant isolates was  higher in milk and milk products (28.4%) as compared to meat (15.4%) samples. This higher occurrence in dairy products could be related to the greater emphasis given to dairy production compared to beef production in the study district. Multidrug resistance usually occurred either due to indiscriminate utilization of antimicrobial agents or genetic mutation, which was difficult to elucidate with the present study methodology.

Conclusion
The presence of E. coli O157:H7 in foods of animal origin may originate from infected animals or unhygienic conditions during processing, handling and distribution. Importantly, the occurrence of E. coli O157:H7 and its multiple antibiotic resistant profiles shows a risk for public health and food safety, as well as animal health and production (Ulukanli et al. 2006). The higher prevalence of multidrug resistant E. coli isolates in dairy products is especially alarming. Proper handling and cooking foods of animal origin are probably as important in preventing E. coli O157:H7 infections.

Funding
There is no fund for this activity. The research was initiated by the staffs (or was it faculty?) of the College of Veterinary Medicine and the College of Veterinary Medicine and Agriculture and an externship student (for fulfillment of the DVM degree), with the help of Haramaya University and Addis Ababa University, which provided materials and reagents (bacterial media, biochemical test kits, antibiotic discs).

Availability of data and materials
All necessary data supporting our findings can be found in the repository.
Authors' contributions SB, DS and AA developed the research concept and designed the methodology, data analysis and interpretation and preparation of the manuscript for publication. TM provided critical comments on the proposal methodology and reviewed the manuscript for publication. SB and TM carried out the sample collection, laboratory work and revision of the manuscript. All authors read and approved the final manuscript. Ethics approval and consent to participate There was no involvement of animals or humans for sample taking, as this study was conducted on milk samples taken from containers which were ready for sale in non-standardized market systems.

Consent for publication
In our study, we don't have any images or videos, etc. of individual participants.