The study was carried out in Addis Ababa, the capital city of Ethiopia. The city covers 540 km² and is divided into 10 subcities. The city lies at an elevation of 2,355 m above sea level and is located at 9°1′48″N 38°44′24″E. The city has a minimum, maximum, and average temperature of 14, 21, and 17.5°C, respectively. Addis Ababa has bimodal rainfall with an annual average of 1,800 mm with the highest of rain falls during the long rainy season from June to September. The short rainy season is from February to April. The capital city has an estimated human population of 3.15 million (CSA, 2010).

A cross-sectional study was conducted from October 2018 to December 2019 to determine the prevalence and antimicrobial susceptibility of E. coli O157:H7 serotype in retail raw lettuce samples obtained from selected vegetable retail outlets, in Addis Ababa.

The sample size required was calculated according to the formula in Thrusfield (2018):

where Z = z statistic for level of confidence; n = required sample size; P_exp, the expected prevalence and; d the desired absolute precision. We used an expected E. coli O157:H7 prevalence of 50, 5% precision, and level of confidence of 95%. The calculated sample size required was 384.

Addis Ababa has 10 subcities. Across the 10 subcities, 390 vegetable retail outlets were conveniently included in the study. A lettuce sample was purchased from each of the outlets in the same fashion as they are sold to the buyers.

Each lettuce sample was placed in a sterile individual plastic bag. The sample was identified by its exclusive sample identification number, which was written on the plastic bag, alongside the subcity and date of sampling. Finally, the sample was transported to the Microbiology Laboratory of the Aklilu Lemma Institute of Pathobiology, Addis Ababa University, at cold temperature in a cool box. Upon arrival to the laboratory, the samples were stored in a refrigerator at ±4°C. The samples were processed within 6–12 h from arrival. Detection of E. coli O157:H7 was done using an ISO-16654:2001 standard (ISO, 2001).

The inner and outer leaves of the lettuce were cut by a sterile scalpel to form a 25-g sample and put in a sterile Stomacher bag. Then, 225 ml of modified Tryptone Soya Broth (TSB) supplemented with Novobiocin (mTSB + N) (1:9) was added to the lettuce and homogenized using a homogenizer (Stomacher 400, Seward Medical, England) at high speed for 2 min. The enrichment sample was then incubated aerobically at 41.5°C for 24 h.

Detection of E. coli O157:H7 was done using ISO-16654:2001 standard (Manyi-Loh et al., 2018). All enriched broths were plated onto cefixime tellurite sorbitol MacConkey agar (CT-SMAC) that was sorbitol MacConkey agar (Oxoid, England), supplemented with 0.05 mg/L cefixime and potassium 2.5 mg/L tellurite (Oxoid, England) (CT-SMAC) (Oxoid, England) and incubated at 37°C for 24 h. Following the end of the incubation period, the CT-SMAC agar plates were examined for the presence of non-sorbitol fermenter colorless colonies, and subsequently they were subcultured on Rainbow agar O157 (Hayward, United States). The plates were then incubated for 20–4 h, at 37°C and observed for the presence of typical black or gray coloration on Rainbow agar O157, which shows pure colonies (Biolog, 2008).

Five typical colonies from each Rainbow agar O157 plate were subcultured on nutrient agar (Oxoid, England) for biochemical confirmation by indole formation. The agar plates were incubated at 37°C for 18–24 h. One colony from the pure culture on the nutrient agar was inoculated into a tube of tryptone/tryptophan medium (Oxoid, England) and incubated at 37°C for 24 h. Then, 1 ml of Kovack’s (Oxoid, England) was added, and the tube was allowed to stand at room temperature for 10 min. The formation of red color indicates a positive reaction (Manyi-Loh et al., 2018).

Latex agglutination was performed for confirmation of E. coli O157:H7 using an RIM E. coli O157:H7 latex test kit (Oxoid, England) (Manyi-Loh et al., 2018). The latex kit has four compartments: latex test reagent, latex control reagent, the positive controls, and negative controls. The particle in each reagent was coated with a different antibody against E. coli serotype O157 or E. coli serotype H7, and therefore the control was coated with normal rabbit globulin. The positive and negative controls are suspensions of formalin-killed E. coli O157:H7 cells and non-specific E. coli cells, respectively.

Indole-positive colonies were examined for their serological reaction with antiserum to E. coli O157:H7 using an RIM E. coli O157:H7 latex test (Oxoid, England). Indole-positive colonies were subcultured from the nutrient agar to the sorbitol MacConkey agar (Oxoid, England). For every isolate to be tested, one drop of test latex was dispensed into a well of the test slide. In like manner, one drop of E. coli control latex was dispensed into a separate well of the test slide. Using a plastic stick, a portion of the non-sorbitol fermenting colony (NSFC) was removed from the sorbitol MacConkey agar-SMAC (Oxoid, England) plate, emulsified in E. coli O157 test latex on the slide, and spread over the reaction area. Using a fresh plastic stick, the process was repeated with the remaining NSFC and emulsified in E. coli control, latex on the slide. The slide was rotated using circular motions for up to 1 min or until agglutination appeared. For E. coli O157 positives that agglutination occurs with the E. coli O157 test latex and the control latex is negative, the isolate was streaked from sorbitol MacConkey agar (Oxoid, England) to a blood agar (Oxoid, England) plate and incubated at 37°C for 18–24 h. After 18–24 h of incubation, the sweep of growth from the blood agar plate was emulsified in a drop of E. coli H7 test latex. Colonies giving an agglutination reaction were confirmed as E. coil O157:H7 positive.

Isolates were confirmed by polymerase chain reaction (PCR) for the O157 lipopolysaccharide O-antigen synthesis rfbEO157 gene. The reactions were performed using oligonucleotide primer set specific for rfbEO157 gene (Thermo Fisher Scientific, United States). Primers for the amplification of the rfbEO157 genes are rfbEO157-F 5′-AAG ATT GCG CTG AAG CCT TTG-3′and rfbEO157 R-5′-CAT TGG CAT CGT GTG GAC AG-3′ rfbEO157 producing 497 bp of rfbEO157 gene (Desmarchelier et al., 1998).

The PCR was performed using Illustra^TM PuReTaq^TM Ready-To-Go^TM PCR Beads (GE Healthcare Bio-Sciences, United States) in 25 μl reaction using SimpliAmp Thermal Cycler (Applied Biosystems, Singapore). The PCR reaction mixtures were subjected to the following conditions in SimpliAmp Thermal Cycler: an initial denaturation of one cycle at 95°C for 5 min, followed by 35 cycles, each consisting of 94°C for 30 s, 66°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 10 min (Desmarchelier et al., 1998).

Amplification products were resolved by electrophoresis in 2% agarose gels, stained with ethidium bromide. The gel was run for at 100 V for 40 min and then visualized with UV light using AlphaImager instrument and FluorChem HD2 software. Product sizes were determined by comparison with a 100-bp gene ruler DNA ladder (Thermo Fisher Scientific, United States).

The antimicrobial susceptibility was performed following the standard agar disk diffusion method consistent with CLSI (2014) using commercial antimicrobial disks (Table 1).

Each isolated bacterial colony from pure fresh culture was transferred into a tube of 5 ml TSB (Oxid, England) and incubated at 37°C for 6 h. The turbidity of the culture broth was adjusted using sterile saline solution or added more colonies to get turbidity, usually comparable with that of 0.5 McFarland standards (approximately 3 × 10⁸ CFU per ml). Mueller–Hinton agar (Oxiod, England) plates were prepared consistent with the manufacturer. A sterile cotton swab was immersed into the suspension and rotated against the side of the tube to get rid of the sulplus fluid and then swabbed in three directions uniformly on the surface of Mueller–Hinton agar plates. After the plates dried, antibiotic disks were sited on the inoculated plates using sterile forceps. The antibiotic disks were gently pressed onto the agar to ensure firm contact with the agar surface, and incubated at 37°C for 24 h. Following this the diameter of the inhibition zone formed around each disk was measured using a black surface, reflected light, and transparent ruler by lying it over the plates. The isolates were classified as sensitive, intermediate, and resistant according to the breakpoints of the clinical interpretative criteria recommended by CLSI guidelines (CLSI, 2014). E. coli (ATCC 25922) was used as a positive control.

The study protocol was ethically approved by the Institutional Review Board of Aklilu Lemma Institute of Pathobiology, Addis Ababa University (Minutes Ref NO: ALIPB IRB/006/2011/2018).

The data were entered and coded in MS Excel and then analyzed using IBM SPSS version 25.0 (SPSS, 2017). The prevalence was determined by dividing the number of positive samples by the total number of samples examined. Descriptive statistics such as frequency and percentages were used to describe the proportion of resistant, intermediate, or susceptible strains. The difference in prevalence by subcity was determined using chi-square (χ²) test. A p-value < 0.05 was considered indicative of a statistically significant difference.

A total of 390 retail lettuces samples were collected from vegetable retailer outlets in Addis Ababa, 39 in each of the 10 subcities. Out of 390 lettuce samples examined, two (0.51%) (95% CI = −0.2–1.23) were positive to E. coli O157:H7.

The positive samples were from outlets in Bole and Yeka subcities. However, the prevalence of E. coli O157:H7 was not statistically significant (χ2 = 8.041; df = 9, p = 0.530 (p > 0.05) among the subcities of vegetable retail outlets (Table 2).

Both isolates were resistant to ampicillin. One of the isolates, the one from Bole subcity, showed additional resistance to tetracycline. Both isolates were susceptible to five antibiotics: amoxicillin-clavulanic acid, ceftriaxone, cefoxitin, nalidixic acid, and sulfamethoxazole-trimethoprim (Table 3).