Detection and characterisation of extended-spectrum and plasmid-mediated AmpC β-lactamase produced by Escherichia coli isolates found at poultry farms in Bosnia and Herzegovina

Between September and October 2019, we took 108 faecal samples (cloacal swabs) from 25 poultry farms located in the Zenica-Doboj Canton (Bosnia and Herzegovina) for diagnostic laboratory investigations, including bacterial isolation and identification. Samples were cultured on blood and MacConkey agar supplemented with 3 mg/L of cefotaxime (Oxoid, Besingstoke, UK) to detect cefotaxime-resistant isolates (17). Using the standard microbiological methods and observing colony morphology and biochemical reactions with indole, Kligler Iron Agar (KIA), citrate agar, phenyl-alanine agar, and urease agar (18) all 27 non-copy isolates were identified as E. coli. They were further evaluated at the University Hospital Centre Zagreb, Department for Clinical and Molecular Microbiology using molecular methods.

Antimicrobial susceptibility of the 27 cefotaxime-resistant E. coli isolates was established with the Kirby-Bauer disk-diffusion and broth microdilution methods according to the Clinical and Laboratory Standards Institute (CLSI) recommendations (19). The same antibiotics were used for both disk-diffusion and broth dilution test, except that some antibiotics such as sulphamethoxazole/trimethoprim, ertapenem, and cefoxitin were tested only with the disk method.

Minimum inhibitory concentrations (MICs) of amoxicillin alone and in combination with clavulanate, piperacillin/tazobactam, cefazoline, extended-spectrum cephalosporins (ESCs; ceftazidime, cefotaxime, and ceftriaxone), cefepime, imipenem, meropenem, gentamicin, or ciprofloxacin were determined with the broth dilution test. The range of tested concentrations was 0.06 to 128 μg/mL. E. coli ATCC 25922 and K. pneumoniae ATCC 700603 were used as quality control strains.

ESBLs were screened for using the double-disk synergy test (DDST) as described elsewhere (20). Briefly, a disk containing amoxicillin with clavulanic acid is placed in the centre of the plate, and disks containing ceftazidime, cefotaxime, ceftriaxone, and cefepime 25 mm apart from the central disk. The test is considered positive if, after overnight incubation at 37 °C, the inhibition zone around cephalosporin disks extends towards the central disk with clavulanic acid.

ESBL production was confirmed with the combined disk test with cephalosporins and clavulanic acid according to CLSI (19). Briefly, the overnight broth culture of the test isolate was diluted to McFarland 0.5 turbidity and swabbed on Mueller-Hinton agar. Disks containing ceftazidime (30 μg), cefotaxime (30 μg), ceftriaxone (30 μg), and cefepime (30 μg) were placed on the surface of the agar plate, and 10 μL of clavulanic acid (10 g/L) was dropped on the disks. The control disks contained the same antibiotics but without clavulanate. ESBL production was confirmed if the inhibition zones around ceftazidime, cefotaxime, ceftriaxone, and cefepime disks with clavulanic acid were at least 5 mm wider in diameter than around control disks without it.

Cefotaxime hydrolysis by ESBL was tested with the cephalosporin inactivation method (CIM) as originally described for carbapenem inactivation testing (21). Briefly, disks containing cefotaxime (10 mg) were placed in a heavy suspension of the test strains and the samples were incubated at 37 °C for 2 h. Disks were then taken out and placed on the Mueller-Hinton agar previously inoculated with E. coli ATCC 25922 susceptible to cefotaxime. The test was considered positive if there was no inhibition zone, if it was <14 mm in diameter, or if the colonies grew within the inhibition zone.

Four cefoxitin-resistant isolates were tested for plasmid-mediated AmpC β-lactamases using the combined disk test with cephalosporin and 3-aminophenylboronic acid (PBA) disks as described elsewhere (22). Briefly, overnight broth culture of the test isolate was diluted to the 0.5 McFarland turbidity and swabbed on Mueller-Hinton agar. Disks containing ceftazidime (30 μg), cefotaxime (30 μg), ceftriaxone (30 μg), and cefepime (30 μg) were placed on the surface of the agar plate and 10 μL of pAmpC-inhibiting PBA was dropped on the disks. The control disks contained the same antibiotics without PBA. The test was considered positive if the inhibition zone around PBA was at least 5 mm longer in diameter than the respective control disk.

The conjugation experiment was performed by “mating” the experimental E. coli isolates (donors) with the J65 strain resistant to sodium azide (recipient) according to Elwel and Falkow (23). Overnight broth cultures of donor and recipient strains were mixed in the ratio of 1:2 v/v in Brainheart infusion broth (BHI) and incubated without shaking at 37 °C for another night. Mating mixtures were then placed on combined plates containing cefotaxime (2 mg/L) or ciprofloxacin (1 mg/L) to inhibit the growth of recipient strain and sodium azide (100 mg/L) to inhibit the donor strains. The frequency of conjugation was determined relative to the number of donor cells. The co-transfer of resistance to non β-lactam antibiotics such as tetracycline, chloramphenicol, sulphamethoxazole/trimethoprim and gentamicin was tested as well.

The DNA was extracted using the heat lysis protocol as described elsewhere (24). Briefly a heavy suspension of the tested strains prepared in 500 μL of ultrapure water was boiled at 95 °C for 15 min and then spun at 6720 g for 2 min. The obtained clear supernatant was used as template DNA for polymerase chain reaction (PCR).

Genes coding for broad and extended-spectrum β-lactamases (bla_SHV, bla_TEM, bla_CTX-M, and bla_PER-1), plasmid-mediated AmpC β-lactamases, and quinolone resistance (qnr) were detected as described earlier (25–29). Table 1 shows the primers used in this study. The CTX-M β-lactamase cluster was detected with multiplex PCR including five primer pairs: CTX-M-1, CTX-M-2, CTX-M-8, CTX-M-9, and CTX-M-25. For genes coding for CTX-M-8 and CTX-M-25 clusters we used the common reverse primer (30). Amplicons were visualised after electrophoresis in 1 % agarose by staining it with ethidium bromide (Sigma Aldrich, St. Louis, MO, USA). A 100 bp ladder was used as standard to determine the size of the products. The expected size of the amplicons is shown in Table 1. Amplicons were purified using the QIAquick PCR purification kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions (31). DNA was sequenced on both strands by the Eurofins Genomics (Ebersberg, Germany) using the same primers that generated amplified products.

The genetic context of bla_CTX-M genes was determined by PCR mapping with forward primer for insertion sequences ISEcp1 and IS26 combined with primer MA-3 (universal reverse primer for bla_CTX-M genes) (32). Positive control strains producing TEM-1, TEM-2, SHV-1, and SHV-2 were kindly provided by Professor Adolf Bauernfeind (Max von Pettenkofer Institute, Munich, Germany), and those producing CTX-M-3 and CTX-M-15 by Professor Neil Woodford (Health Protection Agency, London, UK).

Plasmids were extracted with the Qiagen Plasmid Mini Kit according to the manufacturer’s instructions (33). To identify plasmids coding for ESBLs we relied on PCR-based replicon typing (PBRT) as described by Carattoli et al. (34) updated to identify and distinguish between IncL and IncM plasmids (35). Five multiplex (I, II, III, IV, and V) and two simplex (B and K) reactions were used to determine the incompatibility group based on the size of the product after gel electrophoresis and staining with ethidium bromide. Plasmid extractions from donor and transconjugant strains were subjected to PCR for detection of bla_ESBL genes in order to determine their location. Positive control strains were kindly provided by Dr Alessandra Carattoli (Istituto Superiore di Sanità, Rome, Italy).

Two randomly selected isolates were genotyped with multilocus sequence typing (MLST) as described by Wirth et al. (36). Seven housekeeping genes were amplified with PCR, namely adk, fumC, gyrB, icd, mdh, purA, and recA (37).

PCR products were detected with agarose gel electrophoresis and then purified and sequenced using the Eurofins service. The obtained sequences were deposited by above mentioned online service to obtain sequence types (STs).

Eighteen of the 27 isolates transferred cefotaxime resistance to the E. coli recipient strain with the frequency ranging from 5.6×10^-6 to 5×10^-4. Alongside cefotaxime, tetracycline resistance was cotransferred from six, cotrimoxazole from two, and gentamicin resistance from one isolate (Table 4). Ciprofloxacin resistance was not transferred from any of the tested strains.

Gm – gentamicin; Smx – sulphamethoxazole-trimetoprim; Tet – tetracycline

PCR identified bla_CTX-M-1 cluster genes in 21 isolates, of which seven tested positive for bla_CTX-M-15. Fourteen isolates tested positive for bla_TEM and none for bla_SHV genes. ISEcp was found in only one isolate, whereas IS26 tested negative. The same bla genes were found in donor isolates and their respective transconjugants. Six isolates did not harbour any bla genes in spite of being phenotypically positive for ESBLs, as shown in Table 2. All but six isolates harboured bla_CTX-ML genes, and 14 harboured additional bla_TEM gene. Since bla_TEM genes were not sequenced, it was not possible to distinguish between broad spectrum TEM-1 and TEM-2 β-lactamases and their ESBL variants. PCR for qnr genes yielded no amplicons.

The most frequent plasmid incompatibility group was IncFIB, identified in eight isolates, followed by IncFIA (n=3) and Inc HI1 (n=2) (Table 2). Plasmids were not found in nine isolates.

Two different STs were identified: ST117 (E. coli 4) and ST155 (E. coli 23) (Table 2).