Antimicrobial-resistant Shiga-toxin producing Escherichia coli Isolated from Ready-to-Eat Meat Products and Fermented Milk Sold in the Formal and Informal Sectors in Harare, Zimbabwe

This study aimed to: (1) assess Escherichia coli contamination in polony, beef burgers and traditionally fermented cow milk from the formal and informal markets in Harare, Zimbabwe, (2) determine the antibiotic sensitivity of Escherichia coli isolates, and (3) identify Shiga-toxin producing Escherichia coli isolates using the presence of virulence genes, namely, intimin, enterohemolysin A and Shiga toxins 1 and 2. Ninety-six samples comprising 32 beef polony slices, 32 beef burger patties, and 32 fermented milk specimens were obtained from the informal and formal outlets of the central business district. Escherichia coli occurred in 20 (21%) of the samples, being more prevalent in the informal (29%) than in the formal (13%) market. Of the 20 E. coli isolates, 6 (30%) were Shiga-toxin producing E. coli, and the rest (70%) were negative for virulence genes. The predominance of Escherichia coli was greater in meat products (25%) than in fermented milk (13%). Total Escherichia coli counts were not substantially different between formal and informal markets (t-test: p=0.08). All the E. coli isolates were multidrugresistant with antimicrobial resistance prevalence ranging from 25% for Sulphamethoxazole to 100% for Penicillin and Erythromycin. The presence of E. coli in food indicates faecal contamination and probable existence of other enteric pathogens. The presence of virulent and antimicrobial-resistant E. coli strains in food threatens food safety and public health. We conclude that ready-to-eat animal products from both informal and formal sectors could result in the dissemination of antimicrobialresistant Escherichia coli species if corrective measures are not taken.


Journal of Pure and Applied Microbiology iNtROduCtiON
Food contaminated with pathogenic microorganisms is unsafe and represents a danger to food safety and community well-being, especially in developing countries. Escherichia coli (E. coli) is among the most prevalent food-borne pathogenic agents 1 . A large proportion of resistant and pathogenic E. coli causing human infections are mainly derived from meat and milk products 2 . Escherichia coli O157:H7 and several other E. coli strains including Shiga-toxin-producing E. coli (STEC) strains (serogroups O45, O121, O26, O111, O103, O145 etc.) are important pathogenic infectious agents causing many morbidities and mortalities in people of both the developed and developing world 3 .
Pathogenicity of several Escherichia coli species is generally founded on their virulence properties such as enterotoxigenic E. coli (ETEC), enteropathogenic E. coli (EPEC), enterohaemorrhagic E. coli (EHEC) and STEC 4 . Virulence in Escherichia coli results from genomic or plasmid virulence genes 5 . Genes exist individually or polygenically in clinical isolates of varying frequency 6 . The presence of virulence genes such as Shiga-toxins (stx1 and stx2), hemolysin (hlyA) and intimin (eaeA) in E.coli strains of ready-to-eat animal products pose a serious food safety and public health concern. The virulence factors are responsible for E.coli adhesion, colonization and invasion into the gastric epithelial cells 7 .
Even if isolation and antimicrobial susceptibility testing are done on food samples, some bacterial infections are treated without the establishment of the causative agent, antibiogram and/or assessing virulence of the strains 8 . Environmental contamination leads to the perpetuation of pathogenic E. coli as well as acquiring of antimicrobial resistance (AMR) and virulence genes by non-pathogenic Escherichia coli. Lack of information on antibiotic resistance and prevalent virulence gene combination profiles of pathogenic microbes compromises the effectiveness of antibiotic therapy 9 . Several types of research found that insufficient evaluation in the use of antibiotics during the treatment of human / animal infections, prophylactic and growth stimulation in animal production can confer resistance in various bacteria 10-14 . The emergence of AMR and multidrug-resistant strains render the management of pathogenic bacteria increasingly challenging. Throughout the decades, resistance to cephalosporins among Escherichia coli species has multiplied primarily due to the prevalence of Extended-spectrum β-Lactamases (ESBL).
Virulence and resistance factors are transmitted by horizontal gene transfer among organisms 15 . The exchange of DNA fragments is possibly the primary genetic mechanism for the spread and co-selection of resistance and virulence genes, though some modes, like the compensatory or adaptive mutations, could also be involved 15 . Mobile genetic components, like plasmids and transposons, have frequently been correlated with genes that encode AMR and may be interchanged across microbes belong to different evolutionary phenotypes 16 .Antimicrobial resistance and virulence are acquired by organisms through different evolutional routes but both processes are important for bacterial survival under hostile environments 15 . Virulence factors are needed for overcoming the host's immune system and acquisition of AMR is vital for resisting antimicrobial treatments, adaptation and survival in adverse conditions 15 .
Food safety is really a human health concern in Zimbabwe, as demonstrated by repeated outbreaks of food-related diseases 17 . Like in other developing countries, there is little up-todate statistics on the incidence, type, and virulence of food borne pathogens causing infections in humans, creating a missing link in developing strategies to control food borne diseases. It is against this background that this work seeks to document the existence of antimicrobial-resistant STEC strains in ready-to-eat animal products from the informal and formal sectors in Harare, Zimbabwe. This information is potentially valuable in developing effective intervention strategies and in educating the food industry and consumers.
The general purpose of this work was to evaluate the microbial safety of ready-to-eat meat products and sour milk from the formal and informal markets in Harare, Zimbabwe,with a focus on Shiga-toxin producing E.coli. To achieve this objective the following specific objectives were addressed:(1) to evaluate the occurrence of Escherichia coli contamination in polony, beef burgers and traditionally fermented cow milk from the formal and informal markets, (2) to assess the susceptibility of Escherichia colis trains against eight commonly used antibiotics (Ampicillin, Gentamycin, Penicillin G, Erythromycin, Tetracycline, Neomycin, Ceftriaxone, and Sulfamethoxazole) and (3) to determine STEC isolates by the existence of four virulence genetic factor, namely, enterohemolysin A (hly A), intimin (eae A) and Shiga toxins 1 and 2 (stx1 and stx2)

MATERIAlS ANd METHOdS Study area
Harare, the commercial and administrative capital of Zimbabwe, does indeed have a populace of around two million citizens 18 . Apart from the typical formal food outlets (e.g. supermarkets and butcheries), foods like fruits, vegetables, and ready-to-eat meat products, are sold at informal markets dotted around the city centre. Food vendors sell their foods from vending stalls and open space. The socio-economic challenges facing Zimbabwe have compromised adherence to food safety regulations and hygiene practices by both formal and informal food outlets in Harare, potentially exposing consumers to unsafe food.

Sample collection
This study is an observational crosssectional study in Harare Central Business Centre assessing the presence of Shiga-toxin producing and other E. coli strains from ready-to-eat animal products sold formally and informally. A total of 96 samples of ready-to-eat meat products and sour milk were randomly obtained from informal and formal traders in the central business centre, over four weeks beginning 05 January 2020. A total of 32 beef polony, 32 beef burger patties, and 32 fermented milk samples were collected from both informal (48 samples) and formal (48 samples) markets at bus termini and shopping centres in the city centre. The specimens were placed inside an ice cooler as well as immediately ferried to the Central Veterinary Laboratory for analysis.

detection of Escherichia coli
A mass of 1 gram of each specimen was applied to 9 milliliters of peptone water broth, homogenized and incubated for 24 hours at 44°C for 24. A culture loop was inoculated separately on Blood agar and MacConkey agar plates and incubated aerobically at for 24 hours. Colonies were examined for their cultural/ morphological characteristics on the media. Large, round, red or pink colonies surrounded by a hazy boundary on MacConkey agar indicated E. coli. Gram staining and various biochemical tests such as catalase, oxidase, indole, motility, citrate, triple iron sugar, lysine, urease and carbohydrates (sugar) fermentation (inositol, lactose, xylose, inulin, glucose, maltose, mannose, sorbitol, mannitol and sucrose) were conducted to confirm whether the bacterial strains were E. coli strains or not.

Escherichia coli enumeration/ Total Escherichia coli count
A mass of 1g sample was suspended in 9ml phosphate-buffered saline, homogenized and serially diluted up to 10 -6 . A volume of 0.1 ml of each serial dilution was spread over Cefixime Tellurite Sorbitol-MacConkey agar and incubated for 24 h at 44°C. Suspected colonies of Escherichia coli were counted and confirmed using gram staining and biochemical tests as described above.

Antimicrobial resistance test (AMR test)
A standardized disc diffusion technique was used for antibiotic susceptibility testing of the E. coli isolates on Mueller Hinton agar (Thermo Fisher Scientific, UK) without any  Table 1 and then run on a PCR thermo-cycler (Gene PCR System 24, Perkin Elmer, USA).
The components of the reaction included a commercial master mix kit (2x concentration) (Fermentas) (containing-a PCR buffer, dNTPs, MgCl 2 , Taq DNA Polymerase) and primers dissolved in nuclease-free water. A total of 35 thermo-cycler reaction cycles were set up with initial denaturation of 3 min at 94°C, denaturation at 94°C for 1 min, annealing at 56°C for 50 s, elongation at 72°C for 1 min and final elongation at 72°C for 1 min. Two percent agarose gel with 5µl of ethidium bromide (10mg/ml) submerged to 4mm depth of 0.5xTBE buffer in a gel electrophoresis tank was used. A 5µl volume of PCR products (Amplicons) and negative control was mixed with 2 µl of the loading/tracking dye and loaded into the wells on the gel. Hyper ladder IV 100bp DNA (0.1 µg/µl, 50 µg) (Bioline) was employed as a molecular weight marker or ladder.
Electrophoresis was run for 1 hour at 120V on a BRL Horizontal Gel Electrophoresis Apparatus (Horizon 11.14 Life Technologies Gibco, USA) and all the gel results were viewed and photographed using the Gel Logic 100 Imaging System (Kodak, EEC) under UV trans-illumination. Analysis of the gel picture was done on the computer to generate the DNA profiles and draw conclusions about the sample by estimating the band size from the DNA ladder/maker.

Statistical analysis
The antimicrobial-resistant test results were analysed using the CLSI breakpoints and WHONET breakpoints 20 . The independent samplest-test was done in R ver. 3.6.2 to test if total E. coli counts differed significantly between the two trading sectors.

Escherichia coli identification and enumeration
A total of 20(21%) morphologically and biochemically confirmed isolates of E. coli were isolated from ready-to-eat animal products. Escherichia coli was higher in the informal sector with a prevalence rate of 29% (14 out of 48) than the formal sector with a prevalence rate of 13% (6 out of 48). There was no statistically significant difference in total E.coli count between formal and informal sectors (Student t-test: p= 0.08).Polony had a higher prevalence rate of about 56% (9 out of 16) in the informal market and about 19% (3 out of 16) prevalence rate from the formal sector. Total E.coli counts ranged from 4.0 x 10 1 to 1.0 x 10 3 in the informal market and from 2.0 x 10 1 to 1.2 x 10 2 in the formal market. Escherichia coli occurred in 3(19%) of burgers from the informal sector and 1(6%) of those from the formal sector. Total E.coli counts for burgers ranged from 1.6 x 10 2 to 3.0 x 10 2 in the informal market and from less than 1.0 x 10 1 to 9.0 x 10 1 in the formal market. Escherichia coli was equally prevalent in fermented milk from the informal and formal markets occurring in 2(13%) of the samples. Total E. coli count for fermented milk ranged from 3.0 x 10 3 to 6.0 x 10 3 in the informal market and from 3.0 x 10 1 to 2.0 x 10 3 in the formal market.

Antibiotic sensetivity testing of Escherichia coli isolates
The Zone of inhibition was interpreted as resistant (R), Intermediate (I) and susceptible (S) using CLSI and WHONET breakpoints. All of the 20 Escherichia coli isolates were resistant to PEN G and ERY. About 35%, 40%, 80%, 30%, 70% and 25% of Escherichia coli strains demonstrated resistance to Ampicillin, T, CN, CRO, NEO and SXT, respectively (Fig 1). Escherichia coli isolates showed low resistance rate of 25% to CRO and SXT. Isolates susceptibility to SXT, CRO, AMP and T was 75%, 60%, 65%and 45%, respectively. SXT, CRO, AMP and T showed overlapping zones of inhibition diameter or synergy. Extended-spectrum β-lactamase producing strains constituted 43% prevalence. From the antibiotic sensitivity patterns of the Escherichia coli strains, it was observed that all the isolates were multidrug-resistant and were resistant to 3 or more antibiotics tested. informal sector and a general lack of compliance to standards enshrined in the National Food Laws & Regulations of Zimbabwe 21,22 . Some of the cutlery used such as knives for cutting polony may not be cleaned at all or cleaned after very long intervals resulting in the accumulation and persistence of E. coli. Such unhygienic practices may account for high food prevalence and total counts of E. coli in the informal sector. It has been confirmed that Escherichia coli could even remain alive outside the usual hosts in other environments such as soil and water for more than 90 days 23 . The high prevalence of E. coli indicates high chances of occurrence of E. coli associated outbreaks.
In general, the total Escherichia coli counts of samples from the formal sector were lower, probably owing to better food handling practices in the sector. Hygienic practices in the formal sector include the provision of sanitary wipes to customers at the entrance and several points within the shops, short intervals for cleaning equipment and utensils, which prevent E. coli from multiplying. Total E. coli count was lower in burgers than polony probably due to the heating of burgers during preparation, which might reduce bacterial load. Though at lower prevalence, the existence of Escherichia coli in some of the samples from the formal sector indicates possible faecal and urine contamination from food handlers.
The fermented milk samples generally showed low contamination by E. coli and this is attributed to the gram-negative lactose fermenters, which produce a bacteriocin-like inhibitory substance with a broad spectrum of antimicrobial activity against enteric pathogens 24 . The occurrence of Escherichia coli in animal products is a well-thought-out indicator of faecal contamination as well as the presence of other enteric pathogenic bacteria such as Salmonella species, Shigella, Yersinia and many others in the ready-to-eat animal products 25 . Some of the E. coli isolates were positive for ESBL, therefore, authorities need to act to combat the risk of these serotypes.

Antimicrobial resistance pattern of E. coli isolates
Overall, all Escherichia coli strains were resistant to penicillin and erythromycin. It is notable that no single antibiotic tested was 100% effective against the isolates and all the isolates were multidrug-resistant. These findings disCussiON

Prevalence of E. coli
One of the purposes of this research was to evaluate the prevalence of Escherichia coli in ready-to -eat meats and sour milk purchased throughout the informal and formal markets of Harare. Our findings indicate a higher frequency of Escherichia coli in ready-to-eat meat products sold in the informal market, especially polony, than those sold in the formal market. The high prevalence of E. coli in the informal sector can be attributed to poor food handling practices in the corroborate those of the previous studies 8; 26 . The resistance might have arisen due to the prevalent use of antibiotics, which is proportional to the development of antimicrobial resistance 27 . The resistance patterns of T, AMP, NEO, ERY and PEN G observed are not astounding. Since its approval in 1948, tetracycline has been commonly used for the management of disease and growth stimulation in various animal production processes 28 . In recent studies, Escherichia coli resistant to ESBL has been isolated from food and UTI clients in several parts of Zimbabwe 26 . Due to the variability of the affinity of ESBLs for different substrates, the resistance of isolates to CRO with specific zones of inhibition diameter indicates ESBL production 29 . Penicillin G displays a fairly low antibiotic efficacy towards Gram-negative bacteria, as the development of ESBL denotes resistance to all beta-lactam antibiotics except certain carbapenems and cephamycins 30 . ESBL / ampC encoding plasmids may also bear genes that encode resistance to other classes of antibiotics, such as fluoroquinolones, aminoglycosides and sulphonamides 31 , hence the high resistance to PEN G, ERY, NEO, and AMP, as well as resistance to T, SXT and CRO in some of the isolates 32 . The synergistic effect observed for SXT, CRO, AMP and T suggests the effectiveness of their combination. Incorporation of plant essential oils in the treatment administration/ regiment together with the conventional antibiotics is required if antimicrobial resistance is to be reduced 33 . detection of STEC using virulence genes STEC is the most toxic of all DEC (Diarrheagenic Escherichia coli) species and the most dangerous STEC is O157:H7 serotype. Even so, in recent years , the incidence of many non-O157 serotypes in humans associated with the intake of tainted food has increased 36 . The presence of the Aeromonas species variant of the stx2 (Shiga toxin expressing gene: 406bp) in E. coli indicates gene exchange between E. coli and Aeromonas species 9 . In several studies,the species of Escherichia coli and Aeromonas were simultaneously isolated from food, faeces and urine samples of patients with uncomplicated UTIs. Despite the presence of stx gene, most of the isolates from this study, are unlikely to cause any serious infection due to lack of hly A and eaeA virulence factors except one Shiga-toxin E.coli with eaeA virulence gene positive. Some virulent genes in addition to stxare needed to cause severe human diseases 34,35 . The six (30%) strains of STEC identified indicated low prevalence.

CONClusiONs
Presence of E. coli in ready-to-eat meat products and fermented milk from the informal sector suggests poor observance of standard food handling practice, lack of personal hygiene and microbial contamination of the environment. Buying ready-to-eat meaty products and fermented milk from the informal sector poses a higher risk of E. coli infections to consumers. The proportion of antimicrobial-resistant strains of Shiga-toxin producing E. coli and other E. coli isolates from ready-to-eat animal products suggests a major threat to food safety and public health in Zimbabwe. Our findings highlight that there is a higher risk of contracting food-borne E. coli infections from consuming ready-to-eat meat products and fermented milk from the informal market than from the formal market.

Recommendations
Rigorous food safety checks by the food regulatory authorities are critical in both sectors to reduce public health risks.
There is a need to discontinue the use of antibiotics to which E. coli has developed resistance and replace them with their derivatives or natural plants with antimicrobial properties.