Laboratory Surveillance for Prospective Plasmid-Mediated AmpC β-Lactamases in the Kinki Region of Japan

Bacterial isolates.This laboratory surveillance was conducted with the cooperation of 17 institutions (16 clinical laboratories of various hospitals and 1 commercial laboratory) in the Kinki region, which is located in Western Japan (21, 30). Specimens from inpatients and outpatients were collected during the following periods: November 2002 to April 2003 (first study period), November 2003 to April 2004 (second study period), November 2004 to April 2005 (third study period), November 2005 to April 2006 (fourth study period), November 2006 to April 2007 (fifth study period), and November 2007 to April 2008 (sixth study period). A total of 22,869 isolates of Gram-negative bacilli, including Escherichia coli (13,995 isolates), K. pneumoniae (5,970 isolates), Klebsiella oxytoca (1,722 isolates), and Proteus mirabilis (1,182 isolates), were isolated from various clinical specimens and were then tested. A single isolate was selected from each patient. These isolates were identified by the clinical procedures routinely used in each laboratory, and the identifications were confirmed using the WalkAway system (Siemens Healthcare Diagnostics, Tokyo, Japan).

β-Lactamase assays.A cefazolin (CFZ) MIC of >16 μg/ml was used for the initial screening of isolates. Strains that met this criterion were tested using the three-dimensional (3D) test with cefoxitin (FOX) (30 μg per disk) as previously described (8). In brief, the crude enzyme extracts were prepared by freezing and thawing cell pellets from centrifuged Trypticase soy broth cultures. A Mueller-Hinton agar plate (Becton Dickinson Co., Ltd., Tokyo, Japan) was inoculated with E. coli (ATCC 25922) according to the recommendations of the Clinical and Laboratory Standards Institute (CLSI) (6), and a Sensi-Disk (Becton Dickinson Co., Ltd.) containing FOX was placed on the inoculated agar. A slit was cut in the agar with a sterile scalpel blade, and the crude enzyme preparation was dispensed into the slit. After overnight incubation at 35°C, the observation of enhanced bacterial growth at the location where the slit intersected the zone of inhibition was interpreted as a positive 3D test for that strain. Coexistence of AmpC and ESBLs was also investigated by the double-disk synergy (DDS) test with amoxicillin-clavulanic acid (20 and 10 μg per disk, respectively), cefotaxime (CTX) (30 μg per disk), ceftazidime (CAZ) (30 μg per disk), and cefepime (FEP) (30 μg per disk) according to previously published methods (12, 30).

Multiplex PCR amplification and sequencing of PABLs.Strains positive for the 3D test were investigated further for genotype determination with a previously described multiplex PCR protocol (24). The template DNA was replaced with water in negative-control reactions. The positive-control templates, kindly provided by N. D. Hanson, were 6 DNA fragments of bla _MOX-1, bla _LAT-1, bla _DHA-1, a gene native to Hafnia alvei, bla _ACT-1, and bla _FOX-1. DNA sequencing for PABLs was carried out by PCR amplification with consecutive primers for bla _MOX-1 (11), bla _CMY-8 (32), bla _DHA-1 (33), and bla _CMY-2 (10). The amplified products were sequenced in an automated DNA sequencer (ABI 3100; Applied Biosystems, Foster City, CA). The genotypes of strains found positive by the DDS test were also investigated using PCR methods, as described previously (2, 23, 30). The primers used for multiplex PCR or for the sequencing of PABL genes are listed in Table 1.

Transfer of resistance.The transmissibility of PABLs was tested by conjugation analysis using an E. coli CSH2 strain that was resistant both to nalidixic acid and to rifampin (metB F⁻ Rif^r Nal^r) as the recipient. The E. coli CSH2 recipient strain was kindly provided by Y. Arakawa. The donor-to-recipient ratio was 1:4, and mating plates were incubated at 35°C for 4 h. Transconjugants were selected on modified Drigalski agar (Eiken Chemical Co., Ltd., Tokyo, Japan) supplemented with 50 μg/ml rifampin (Sigma-Aldrich, St. Louis, MO), 50 μg/ml nalidixic acid (Sigma), and 50 μg/ml ampicillin (AMP) (Sigma). The presence of CMY-2 genes in the transconjugants was confirmed by PCR.

Plasmid DNAs extracted according to the procedure described by Kado and Liu (13) were used to transform E. coli strain DH5α (Nippon Gene Co., Ltd., Toyama, Japan). Transformants were selected on Mueller-Hinton agar (Becton Dickinson Co., Ltd.) supplemented with 50 μg/ml AMP and were subjected to the conjugation experiment. Plasmid DNAs extracted from the transformants and from the resulting transconjugants were first electrophoresed on 1% agarose (Nippon Gene Co., Ltd.) at 4°C and then transferred to a nylon membrane (Roche Diagnostics, Tokyo, Japan). To detect CMY-2 gene-containing plasmids, the membrane was probed with digoxigenin (DIG)-labeled DNA probes prepared using PCR DIG Labeling Mix (Roche Diagnostics), with a DNA template prepared from a bla _CMY-2-producing E. coli strain and primers CITMF and CITMR.

Testing for susceptibility to antimicrobial agents.MICs for strains with PABLs were determined by a broth microdilution method with the MicroScan Neg MIC 3.31E and Neg Combo 6.11J panels (Siemens Healthcare Diagnostics) as described by the NCCLS (now the CLSI) (20). The following concentrations of antimicrobial agents were used: piperacillin, 8 to 64 μg/ml; CTX, 0.5 to 128 μg/ml; CAZ, 0.5 to 128 μg/ml; FEP, 1 to 32 μg/ml; FOX, 2 to 32 μg/ml; cefmetazole, 0.5 to 32 μg/ml; imipenem (IPM), 1 to 8 μg/ml; meropenem (MEM), 0.25 to 16 μg/ml; amikacin (AMK), 4 to 32 μg/ml; gentamicin (GEN), 1 to 8 μg/ml; minocycline (MIN), 1 to 8 μg/ml; levofloxacin (LVX), 1 to 4 μg/ml; and sulfamethoxazole-trimethoprim (SXT), 38 and 2 μg/ml. AMK, GEN, MIN, LVX, and SXT were used for antibiograms. E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as quality controls. The quality control criteria used for the dilution methods were those of the CLSI (7).

RAPD pattern analysis.Isolates confirmed to be positive for the PABL gene by PCR were subjected to chromosomal DNA typing by random amplification of polymorphic DNA (RAPD) analysis in order to generate RAPD fingerprints (22). Genomic DNAs were prepared using the QIAamp DNA Mini kit (Qiagen, Tokyo, Japan). RAPD amplification was performed with primer ERIC2 (0.8 μM) in a final volume of 50 μl containing 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 0.2 mM each deoxynucleoside triphosphate, 2.5 mM MgCl₂, 1.25 U of Taq DNA polymerase (Toyobo Co., Ltd., Osaka, Japan), and 4 μl template DNA. The cycle conditions were as follows: initial denaturation at 94°C for 5 min; 35 cycles of denaturation for 1 min at 94°C, annealing for 1 min at 35°C, and extension for 1 min at 72°C; and a final extension at 72°C for 10 min. RAPD band patterns were compared visually, and samples differing in at least two bands were considered nonidentical.

PFGE.Isolates showing identical RAPD patterns were analyzed by pulsed-field gel electrophoresis (PFGE) using the GenePath system (Bio-Rad Laboratories). Genomic DNAs were prepared using the CHEF genomic DNA plug kit (Bio-Rad Laboratories), digested with XbaI (Nippon Gene Co., Ltd.), and electrophoresed through 1% agarose (Bio-Rad Laboratories) under the following conditions: 19.7 h at 6 V/cm (200 V) and 14°C, with a pulse time of 5.3 to 49.9 s and a nonlinear ramp with a half-height of 21%. After electrophoresis, gels were stained with ethidium bromide. PFGE patterns were interpreted according to the criteria proposed by Tenover et al. (27).

Prevalence of PABL-positive isolates.Nine hundred (6.4%) of 13,995 E. coli isolates, 106 (1.8%) of 5,970 K. pneumoniae isolates, 140 (8.1%) of 1,722 K. oxytoca isolates, and 60 (5.1%) of 1,182 P. mirabilis isolates met the criterion of a CFZ MIC of >16 μg/ml. Of these 1,206 isolates, 290 (2.1%) E. coli isolates, 13 (0.2%) K. pneumoniae isolates, 6 (0.3%) K. oxytoca isolates, and 1 (0.1%) P. mirabilis isolate were positive by the 3D test. The numbers of isolates positive for the PABL gene were 17 (0.13%) for E. coli, 8 (0.13%) for K. pneumoniae, 3 (0.17%) for K. oxytoca, and 1 (0.08%) for P. mirabilis. Of these 29 PABL-positive isolates, 1 E. coli and 1 K. pneumoniae isolate were positive for ESBLs by the DDS test.

Characterization of PABL-producing isolates.Some characteristics and selected clinical associations of PABL-producing isolates are shown in Table 2. PABL-positive isolates were detected in 13 of 17 laboratories. With respect to the PABL genotypes, 20 (69.0%), 6 (20.7%), 2 (6.9%), and 1 (3.5%) isolate carried CMY-2, DHA-1, CMY-8, and MOX-1 genes, respectively. CMY-2 PABL genes were detected in 17 E. coli isolates from 8 hospitals in 5 prefectures, in 2 K. oxytoca isolates from 2 hospitals in 2 prefectures, and in 1 P. mirabilis isolate from 1 hospital. DHA-1 PABL genes were detected in 6 K. pneumoniae isolates from 4 hospitals in 3 prefectures. CMY-8 genes were detected in 2 K. pneumoniae isolates from 2 hospitals in 2 prefectures. The MOX-1 gene was detected in 1 K. oxytoca isolate from 1 hospital.

The multiplex PCR results for PABLs are shown in Fig. 1.

• Open in new tab
• Download powerpoint

FIG. 1.

Multiplex PCR results for PABL. (a) Lanes 1 to 6, positive controls for MOX-1, LAT-1, DHA-1, a gene native to H. alvei, ACT-1, and FOX-1, respectively; lane 7, negative control; lane 8, a CMY-2-producing E. coli strain; lane 9, a DHA-1-producing K. pneumoniae strain; lane 10, a CMY-8-producing K. oxytoca strain; lanes M, 100-bp DNA ladder. (b and b′) Lanes 1 and 13, positive control for LAT-1; lanes 2 and 14, negative control; lanes 3 to 12 and 15 to 21, the 17 CMY-2-producing E. coli isolates; lanes 22 and 23, CMY-2-producing K. oxytoca strains; lane 24, a CMY-2-producing P. mirabilis strain; lanes M, 100-bp DNA ladder. (c) Lane 1, positive control for DHA-1; lane 2, negative control; lanes 3 to 8, the 6 DHA-1-producing K. pneumoniae isolates; lane 9, positive control for MOX-1; lane 10, negative control; lane 11, the single MOX-1-producing K. pneumoniae isolate; lanes 12 and 13, the 2 CMY-8-producing K. pneumoniae isolates; lanes M, 100-bp DNA ladder.

RAPD typing demonstrated that each of the 17 E. coli isolates (from 8 hospitals) carrying the genes for CMY-2 PABLs had a unique chromosomal DNA pattern. The RAPD patterns of the PABL-positive K. pneumoniae and K. oxytoca isolates are shown in Fig. 2. Three K. pneumoniae isolates from 1 hospital (strains F-1, F-2, and F-3) carrying a gene for PABL DHA-1 demonstrated identical RAPD patterns. Two K. pneumoniae isolates from 2 hospitals (strains I-1 and J-1) carrying a gene for PABL CMY-8 showed identical RAPD patterns. Three K. oxytoca isolates from 3 hospitals carrying genes for MOX-1 (strain M-1) or CMY-2 (strains C-1 and F-4) were distinct. The PFGE profiles of 3 DHA-1-producing K. pneumoniae isolates were identical or clearly related. However, 2 CMY-8-producing K. pneumoniae isolates were nonidentical. The results of PFGE analyses for PABL-positive K. pneumoniae isolates with identical RAPD patterns are shown in Fig. 3.

• Open in new tab
• Download powerpoint

FIG. 2.

RAPD fingerprinting of isolates investigated in this work, carried out with primer ERIC2. Lanes 1 and 2, the 2 CMY-8-producing K. pneumoniae isolates; lanes 3 to 8, the 6 DHA-1-producing K. pneumoniae isolates; lane 9, the single MOX-1-producing K. oxytoca isolate; lanes 10 and 11, the 2 CMY-2-producing K. oxytoca isolates; lanes M, 100-bp ladder.

• Open in new tab
• Download powerpoint

FIG. 3.

PFGE of PABL-positive K. pneumoniae isolates that had identical RAPD patterns. Lanes 1 and 2, the 2 CMY-8-producing K. pneumoniae isolates; lanes 3 to 5, the 3 DHA-1-producing K. pneumoniae isolates; lanes M, PFGE marker.

Two PABL-positive isolates (strains C-2 and H-2) tested positive by the DDS test, and these isolates were positive for the CTX-M-type ESBL gene as determined by PCR.

Transfer of resistance.Among the 17 CMY-2-producing E. coli strains, resistance was transferred from 12 isolates, and a CTX-M-type ESBL was cotransferred from the single strain expressing this resistance. One K. oxytoca strain and 1 P. mirabilis strain that produced CMY-2 transferred resistance. Resistance did not transfer from 1 CMY-2-producing K. oxytoca strain, 1 MOX-1-producing K. oxytoca strain, 2 CMY-8-producing K. pneumoniae strains, or 6 DHA-1-producing K. pneumoniae strains. Southern blotting was used to confirm the presence of CMY-2 genes on plasmids. In the conjugation experiment, one large plasmid was transferred to E. coli from each transformant. Southern blotting with the bla _CMY-2-specific probe showed the presence of bla _CMY-2 on this plasmid (Fig. 4).

• Open in new tab
• Download powerpoint

FIG. 4.

Results of agarose gel electrophoresis and Southern blotting of plasmid DNAs from CMY-2-producing isolates. (A) Agarose gel electrophoresis. (B) Southern blotting with the bla _CMY-2-specific probe. Lanes 1, strain B-1, a CMY-2-producing E. coli isolate; lanes 2, its corresponding E. coli DH5α transformant; lanes 3, E. coli CSH2 transconjugant derived from the transformant; lanes 4, strain C-1, a CMY-2-producing K. oxytoca isolate; lanes 5, its corresponding E. coli DH5α transformant; lanes 6, E. coli CSH2 transconjugant derived from the lane 5 transformant; lanes M, 2.5-kbp DNA ladder.

Susceptibilities of PABL-producing isolates.The types of β-lactamases produced by the 29 PABL-producing strains and the MICs of β-lactams for these strains are shown in Table 3. The susceptibilities of these strains to several β-lactams differed, with the exception of IPM and MEM. Although 2 PABL-positive strains coproducing CTX-M-type ESBLs demonstrated resistance to FEP, the remaining 27 PABL-positive strains were susceptible to FEP. The MICs for E. coli ATCC 25922 and P. aeruginosa ATCC 27853 fell within CLSI control ranges.