Sampling and Cultivation
Based on the 95% confidence interval (CI), margin error of 5%, and prevalence of 0.65 [16], the sample size was calculated as 373. This sample size was adjusted to 410 after considering a 10% non-response rate. We decided to be conservative and collected a total of 430 samples from the provinces of Khuzestan and Fars from October 2017 to December 2018. Khuzestan Province is located in southwest of Iran, and Fars is located in south of Iran. Simple random sampling was used for the selection of subjects. Children who were referred to health centers for vaccination were randomly selected using Excel, and their stool samples were collected by sterile swap. Swaps were placed in a transport medium and were immediately transported to the laboratory on ice. The studied population included healthy infants and children under three years old who did not show any symptoms of diseases, especially gastrointestinal diseases such as vomiting, diarrhea, nausea, stomach ache, and abdominal cramps; also, they had not received antibiotics at least during the past four months. Unhealthy children or those who had received antibiotics were excluded from the study. Samples whose bacterial culture was negative, were excluded from statistical analysis. The study protocol conformed to the ethical guidelines of the Declaration of Helsinki (No: EE/99.24.3.88342/scu.ac.ir). Because the participants were children under 3 years old, parents or guardian were asked to read, accept and sign an informed consent form before any information was collected. Then cultured on MacConkey (MC) agar (Merck; Frankfurt, Germany) and Eosin Methylene Blue (EMB) agar (Merck; Frankfurt, Germany) for the isolation of E. coli strains. People usually carry a predominant E. coli strain that forms more than half of the isolated colonies from their fecal samples [17, 18] and therefore, from each cultured sample, one isolate was randomly selected for further analysis. The isolates were confirmed by standard biochemical tests and the highly specific E. coli universal stress protein A (uspA gene) was detected using PCR as described by Chen and Griffiths [19].
Antimicrobial resistance and ESBL production
The pattern of antibiotic susceptibility of the isolates was determined by the standard Kirby-Bauer disc diffusion method according to Clinical and Laboratory Standards Institute (CLSI-2016) [20]. The antibiotic discs included chloramphenicol (30 μg), nalidixic acid (30 μg), ampicillin (10 μg), tetracycline (30 μg), kanamycin (30 μg), cefotaxime (30 μg), ceftazidime (30 μg), streptomycin (10 μg), trimethoprim-sulphmetoxazole (23.75-1.25 μg), ciprofloxacin (5 μg), gentamicin (10 μg), meropenem (10 μg) and imipenem (10 μg). The minimum inhibitory concentration (MIC) of imipenem-resistant isolates was determined as described by CLSI-2016 [20]. According to the CLSI guidelines, MIC breakpoint for imipenem resistance E. coli isolates was defined ≥ 4 μg/ml. MDR was defined as resistance to three or more different antibiotic families. To evaluate ESBL production in the isolates, synergy double disc (SDD) method was performed using ceftazidime (30 µg) and cefotaxime (30 µg), alone and in combination with clavulanic acid (10 µg). A ≥ 5-mm increase in zone diameter for either cefotaxime or ceftazidime in combination with clavulanic acid versus the zone diameter of the agents when tested alone was considered as ESBL production [20].
Combined disc test (CDT) and the double disc synergy test (DDST)
The two methods of CDT and DDST were adopted to detect MBL-producing isolates. For the CDT, Mueller-Hinton agar (MHA) plate was inoculated by 0.5-McFarland test isolate. Then, two discs of imipenem and imipenem containing 10 μL of 0.5M ethylenediaminetetraacetic acid (EDTA) were placed on the occulted plate. The plates were incubated at 35 °C for 16–18 h. An increase in zone inhibition of equal or more than 7 mm around the imipenem-EDTA disc compared to imipenem alone was considered MBL-positive [12].
To perform DDST, the two discs of imipenem and blank disc with a distance of 15 mm were placed on MHA plate that was inoculated by 0.5-McFarland standard. Then, blank disc was impregnated with 10 μL of 0.5M EDTA. The plates were incubated at 35 °C for 16–18 h. The enhancement of the inhibition zone or appearance of a phantom zone between the imipenem and EDTA discs was considered positive for MBL production [12].
Modified carbapenem inactivation method (mCIM) and EDTA-CIM (eCIM)
mCIM was performed to detect carbapenemases in E. coli isolates and then eCIM was used together with mCIM to differentiate MBLs from serine carbapenemases. Briefly, 1 μL loopful of bacteria was emulsified in 2 mL of TSB; then, one meropenem (10 μg) disc was added to the emulsion and incubated at 35 °C for 4 h. An MHA plate was inoculated with a 0.5-McFarland suspension of meropenem-susceptible E. coli ATCC25922. Meropenem disc was removed from TSB-meropenem disc suspension and placed on the MHA plate inoculated with the E. coli ATCC 25922 indicator strain. Following the incubation of MHA plate at 37 °C for 18–24 h, the zone of inhibition was measured. The isolates that showed a zone inhibition below 15 mm or the presence of pinpoint colonies within a 16–18 mm zone were considered as carbapenemase-positive. For eCIM, all the steps were similar to mCIM, except that EDTA was added to TSB to obtain a final concentration of 5 mM EDTA. The isolates that produce an increase of ≥ 5 mm in zone diameter for eCIM compared to mCIM were considered MBL-positive [21].
Modified Hodge test (MHT)
The MHT test was performed for the phenotypic detection of carbapenemase production. Briefly, a 0.5-McFarland suspension of E. coli ATCC25922 was prepared and diluted 1:10 in saline. An MHA plate was inoculated with the indicator E. coli; then one meropenem (10 µg) disc was placed at the center of the plate. Subsequently, test organisms were cultured on the plate in a straight line from the edge of the disc in a length of 25 mm. The enhanced growth of the indicator E. coli strain towards the carbapenem disc was considered as a positive result for carbapenemase production [22].
Phenylboronic acid (PBA) disc test
PBA disc test was performed for the phenotypic differentiation of MBLs and class A KPC carbapenemases. PBA was dissolved in dimethyl sulfoxide (DMSO) to obtain a final concentration of 20 mg/ml. Then, 20 µl (400 µg PBA) of the solution was dispensed onto the meropenem disc. The standard disc diffusion method was performed for the isolates by meropenem (10 µg) disc with and without PBA. After incubation of the plates at 37 °C for 18 h, the diameter of the inhibition zones was measured. An increase of ≥ 5-mm in zone diameter for PBA-meropenem compared to plain meropenem was considered KPC carbapenemase [8, 23].
PCR amplification
Total DNA was isolated using the boiling method. MBL genes including blaIMP-1, blaIMP-2, blaVIM-1, blaVIM-2, blaSPM-1, blaNDM-1, blaSIM, and blaGIM were amplified. PCR conditions were followed as described previously [13, 24-27]. Additionally, strains were tested for blaKPC, oxa-23, and oxa-48, the three main genes of carbapenemases as described in the corresponding references (Table 1) [28-30]. The PCR products were subjected to Sanger sequencing (BIONEER Company; Daejeon; Korea). Their nucleotide sequences were analyzed with software available from the National Centre for Biotechnology Information (www.ncbi.nlm.nih.gov).
Plasmid profiling
The resistant imipenem isolates were investigated as to their plasmid content. Plasmid DNA was extracted by an alkaline lysis method [31]. The products were electrophoresed on 1% agarose gels. The size of the plasmid bands were determined using a molecular weight marker, made from a lambda/Hind III digest.
Conjugation Experiment
The resistant imipenem isolates were conjugated with an imipenem sensitive, lactose-negative, enteroinvasive E. coli (EIEC) plasmid-free strain. The overnight cultures of the donors and recipient E. coli isolates were mixed in a ratio of 1:10 in nutrient broth and incubated for 48 hours at 37 °C. Then, the mixtures were spread on MacConkey agar containing imipenem (4 µg/mL) and incubated overnight at 37 °C. The lactose-negative, imipenem resistant isolates were analyzed for the presence of plasmids [31].
Statistical analysis
SPSS software (v.22.0) was used for data analysis. A χ2 test or Fisher’s exact test was used to determine the statistical significance of the data. A P value of < 0.05 was considered as statistically significant. The relations between the AMR profiles and the age of children were described using logistic regression model with 95% CI.