Isolation and Characterization of Integron-Containing Bacteria without Antibiotic Selection

Bacterial strainsThree control strains of Escherichia coli carrying class 1 integron-containing plasmids R388 and R46 and class 2 integron-containing transposon Tn7, respectively, were kindly provided by Hatch Stokes (Macquarie University). Plasmid R388 carries a class 1 integron with two gene cassettes, dfrB2 (which encodes trimethoprim resistance) and orfA. Plasmid R46 also carries a class 1 integron with four gene cassettes. R46 has two identical copies of oxa2 (which encodes ampicillin resistance) and one copy each of aadA1 (which encodes spectinomycin and streptomycin resistance) and orfD. Transposon Tn7 carries a class 2 integron with three gene cassettes, drfA1 (which encodes trimethoprim resistance), sat (which encodes streptothricin resistance), and aadA1 (which encodes spectinomycin and streptomycin resistance). The control strains were maintained on Luria-Bertani agar containing ampicillin or streptomycin at 100 μg/ml.

Sample collection and processingFive fecal samples were collected from each of 10 herds of cattle over a 1-month period. All fecal samples were collected from cattle that had been held overnight at a local abattoir. Each herd of cattle originated from independent geographic locations separated by at least 50 km. A 10-g quantity of feces was placed in 90 ml of buffered peptone water (BPW; Oxoid, Basingstoke, England), treated with a stomacher for 1 min, and subsequently enriched for 6 h at 37°C without agitation. A 100-μl aliquot of enrichment was reenriched in 100 ml of BPW for a further 18 h at 37°C. DNA templates were prepared from the resulting enrichments by boiled cell lysis. Briefly, 1 ml of overnight enrichment was pelleted by centrifugation, resuspended in 200 μl of sterile distilled water, and boiled for 10 min. The boiled cell suspensions were centrifuged, and the resulting lysate was used for PCR.

PCR detection of integrase genesBoiled cell lysates were tested by PCR for the presence of integrase genes intI1 and intI2 by using specific primers designed to amplify conserved regions of the respective genes (Table 1). Primers HS 463a and HS 464 were used to detect intI1-positive samples. Primers RB 201 and RB 202 were used to detect intI2-positive samples. The PCRs were performed in 25-μl reaction mixtures containing 2 μl of DNA, 1.0 μM each oligonucleotide primer, 200 μM deoxynucleoside triphosphates (Roche Diagnostics, Mannheim, Germany), 10 mM Tris-HCl, 1.5 mM MgCl₂, 50 mM KCl (Roche Diagnostics), and 1 U of Taq polymerase (Roche Diagnostics). PCR was performed for 30 cycles of denaturation at 94°C for 30 s, annealing at 65°C (intI1) and 62°C (intI2) for 30 s, and extension at 72°C for 45 s. The PCR products were visualized by ethidium bromide staining after agarose gel electrophoresis.

Isolation of integron-containing bacteriaEnrichments positive for intI1 and/or intI2 were serially diluted to 10^−5.5 in 0.1% peptone-1% Tween 80. One milliliter of the dilution was filtered through a hydrophobic grid membrane filter (HGMF) by using a spread filter (Filtaflex, Almonte, Ontario, Canada) and a vacuum manifold. HGMFs were placed onto modified hemorrhagic colitis (mHC) agar (21), which did not include methylumbelliferyl glucuronide, and were incubated overnight at 37°C. Colonies grown overnight were replicated onto a fresh HGMF by use of an HGMF replicator-inoculator system (Filtaflex) and incubated as described above. The bacterial colonies on the original HGMF were lysed by the method of Nizetic et al. (11) after wetting of the membranes with a pretreatment solution for 30 min (24). The DNA was cross-linked to the HGMF by exposure to UV light for 5 min.

Membranes containing cross-linked DNA were hybridized with digoxigenin (DIG)-labeled probes capable of detecting intI1 or intI2 generated with primers HS 463a and HS 464 (intI1-specific probe) and primers RB 210 and RB 202 (intI2-specific probe) and the PCR DIG labeling kit (Roche Diagnostics) according to the instructions of the manufacturer. Hybridization was carried out at 68°C by the methods outlined in the DIG System User's Guide to Filter Hybridization (Roche Diagnostics). Presumptive intI1- and intI2-positive colonies were picked from replicate HGMFs and streaked onto nutrient agar plates (Oxoid).

Confirmation and preliminary characterization of integronsPresumptively integrase-positive isolates were confirmed to be integrase positive by PCR. All isolates containing intI1 were tested for the presence of intI2, and likewise, all intI2-positive isolates were tested for the presence of intI1. Primers able to amplify the inserted gene cassette regions of class 1 and class 2 integrons were used to determine the sizes of the inserted cassette regions. Primers RB 317 and RB 320 were used to amplify the class 1 integron gene cassette region (Table 1). PCR cycling conditions for the class 1 integron gene cassette region PCR were 30 cycles of denaturation at 94°C for 30 s, annealing at 59°C for 30 s, and extension at 72°C for 3 min. The PCR of White et al. (23) was used to amplify the class 2 integron gene cassette region.

Purification of PCR products was achieved by isopropanol precipitation. Briefly, 45 μl of each PCR mixture was added to 180 μl of 75% isopropanol. The samples were incubated at room temperature for 15 min to allow precipitation of the PCR products. The samples were then placed in a microcentrifuge and spun at 17,000 × g for 20 min. The resulting supernatant was carefully aspirated, and then 250 μl of 75% isopropanol was added and the samples were centrifuged at 17,000 × g for 5 min. Following centrifugation, the supernatants were carefully aspirated and the resulting DNA pellets were dried at 90°C for 1 min. The pellets were finally resuspended in 60 μl of TE (Tris-EDTA) buffer (pH 8.0). Quantification of the DNA concentration of each sample was performed by using gel electrophoresis and undigested bacteriophage λ DNA quantification standards of 2, 4, 6, 8, 10, and 20 ng/μl (Gibco BRL, Rockville, Md.).

Sequencing reactions were performed by the use of BigDye terminator (version 2.0; Applied Biosystems) chemistry and a 9700 thermal cycler (Perkin-Elmer, Norwalk, Conn.). Each reaction mixture contained 8 μl of the BigDye terminator (version 2.0) ready reaction mixture and 3.2 pmol of primer in a 20-μl reaction mixture. The primers used for sequencing were the same as those used to generate the PCR amplicon. The PCR program for all sequencing reactions included initial denaturation at 94°C for 5 min, followed by 25 cycles of denaturation for 10 s at 96°C, primer annealing for 5 s at 50°C, and extension for 4 min at 60°C. The sequencing products were purified by the isopropanol precipitation protocol outlined above and were transported on ice to the Australian Genome Research Facility (University of Queensland, Brisbane, Australia) for sequencing. The resulting sequences were analyzed by searching the GenBank database of the National Center for Biotechnology Information via the BLAST network service (1). Sequences were aligned and compared with the sequences in the GenBank database by using the ContigExpress component of the Vector NTI Suite (version 5.5; InforMax, Inc., Frederick, Md.).

Isolate identification and antibiotic susceptibility testingIdentification of integron-containing organisms and the MICs of antibiotics for the organisms were determined by using the VITEK Junior system (BioMerieux, Hazelwood, Mo.). VITEK cards for identification (GNI cards) and antibiotic susceptibility testing (GNS-424 cards) were inoculated and incubated according to the recommendations of the manufacturer. The following antibiotics were tested: amikacin, amoxicillin-clavulanic acid, ampicillin, cefotaxime, ceftazidime, cephalothin, ciprofloxacin, gentamicin, imipenem, meropenem, nitrofurantoin, norfloxacin, ticarcillin-clavulanic acid, tobramycin, trimethoprim-sulfamethoxazole, and trimethoprim. A test for the detection of extended-spectrum β-lactamases (ESBLs) was also carried out, and interpretation of the results was based on comparison of the reduction in growth caused by cefotaxime-clavulanate and ceftazidime-clavulanate and that caused by the cephalosporins alone. The outcome of the test was either ESBL positive or ESBL negative, as determined by the VITEK Junior apparatus. E. coli ATCC 25922 was used as a control organism.

Prevalence of integron-containing bacteria in bovine fecesA total of 50 bovine fecal samples from 10 geographically independent cattle herds were tested for the presence of class 1 and class 2 integrases. Overall, 43 of 50 (86%) samples tested positive for the class 1 integrase and 47 of 50 (94%) were positive for the class 2 integrase. Analysis of each herd for the presence of the class 1 integrase identified six herds for which five of five samples were positive, two herds for which four of five samples were positive, and one herd each for which three of five and two of five samples were positive. The analysis of each herd for the presence of the class 2 integrase also identified eight herds for which five of five samples were positive, one herd for which four of five samples were positive, and another herd for which three of five samples were positive.

Isolates containing class 1 integrons were isolated from 25 of 50 (50%) samples, and isolates containing class 2 integrons were isolated from 14 of 50 (28%) samples. Nine samples had both a class 1 integron-containing organism and a class 2 integron-containing organism, and five samples contained more than one class 1 integron-containing organism. No isolates containing both class 1 and class 2 integrons were found.

Isolate identification and antibiotic susceptibility testingAll isolates confirmed to carry either a class 1 or a class 2 integron were identified to the species level with the VITEK Junior system. Class 1 integrons were found in Aeromonas caviae, Aeromonas veronii biovar sobria, and E. coli. Class 2 integrons were found in urea-positive Providentia stuartii isolates, E. coli, Proteus vulgaris, Proteus mirabilis, Shewanella putrefaciens, and Morganella morganii. Each isolate was given an Abr identifier number, and the antibiotic susceptibilities of the integron-containing isolates were determined by using the GNS-424 cards, which determine the susceptibilities of gram-negative bacteria. The antibiotic resistances of the strains tested are shown in Table 2.

Characterization of inserted gene cassette regions of class 1 and class 2 integronsTo validate the success of our approach to detecting bacterial isolates carrying integrons encoding antibiotic resistance, preliminary characterization of the inserted gene cassette regions was performed. The inserted gene cassette regions of class 1 and class 2 integrons were amplified by PCR with primers RB 317 and RB 320 for class 1 integron-containing organisms and primers Hep 74 and Hep 51 for class 2 integron-containing organisms (Table 1). The PCR amplicons produced by primers RB 317 and RB 320 ranged in size from 900 to 2.1 kb and included variable regions with the capacity to code for between one and three inserted gene cassettes. A few isolates did not produce an amplicon when they were tested for the presence of an inserted gene cassette region. This was probably due to the lack of a 3′ conserved segment in these organisms. The PCR amplicons produced by primers Hep 74 and Hep 51 were 2.2 kb and were the same size as the amplicon produced from the Tn7 control strain. Urea-positive P. stuartii isolates that were intI2 positive failed to produce an amplicon of any size when they were tested.

Partial direct sequencing of the gene cassette regions of 15 randomly selected class 1 integrons was performed to validate the preliminary evidence for the presence of antibiotic resistance gene cassettes. This approach determined the presence of open reading frames with identities to genes encoding dihydrofolate reductases (dfr2D, dfrA1, dfrXII, and dfrV), aminoglycoside adenyltransferases (aadA1 and aadA2), and chloramphenicol acetyltransferase (catB8 and catB3) (Table 2). Partial sequencing of class 2 integrons indicated the presence of sequences orthologous to the trio of antibiotic resistance genes (drfA1, sat, and aadA1) associated with Tn7 (Table 2).