Relevance of resistance levels to carbapenems and integron-borne blaIMP-1, blaIMP-7, blaIMP-10 and blaVIM-2 in clinical isolates of Pseudomonas aeruginosa

Pseudomonas aeruginosa is an opportunistic pathogen and shows high intrinsic resistance to many antimicrobial agents. Carbapenems are one of the first-choice antibiotics for the treatment of P. aeruginosa infections (Gilbert et al., 2005). However, acquired metallo-β-lactamases (MBLs), including the types IMP, VIM, SPM and GIM, have been reported in a variety of clinical isolates of Enterobacteriaceae worldwide (Bush, 1998; Livermore & Woodford, 2000; Walsh et al., 2005). MBLs are becoming increasingly important clinically because the enzymes hydrolyse almost all β-lactams, including carbapenems, and to date there is no clinically available inhibitor (Walsh et al., 2005).

IMP- and VIM-type enzymes are two dominant groups of MBLs, and the nucleotide sequences of 24 IMP variants and 22 VIM variants have been determined (Lahey Clinic, 2008). The bla_IMP-1 gene was the first MBL determinant reported, initially in a clinical isolate of Serratia marcescens (Ito et al., 1995; Osano et al., 1994). The bla_IMP-7 gene is a bla_IMP allele and IMP-7 shares 91 % amino acid identity with IMP-1 (Gibb et al., 2002). The bla_IMP-10 allele is a point mutation derivative of bla_IMP-1 with a single base replacement of G by T at nt 145 leading to an amino acid alteration of Val49 to Phe (Iyobe et al., 2002). The bla_VIM-2 gene is a bla_VIM allele and VIM-2 shares 90 % identity with VIM-1, but only 31 % identity with IMP-1 (Poirel et al., 2000).

Five classes of integron are known to play a role in the dissemination of antibiotic resistance, and class 1 integrons are the most extensively studied (Mazel, 2006). MBL-encoding genes are usually found as gene cassettes in class 1 integrons (Shibata et al., 2003). Typical class 1 integrons contain two conserved segments (CSs), the 5'-CS and the 3'-CS (Stokes & Hall, 1989; Bennett, 1999). The 5'-CS includes the intI1 gene encoding integrase, the attI site for addition of the inserted gene cassette and a promoter(s). The 3'-CS is composed of the qacEΔ1 and sul1 genes, which are responsible for resistance to quaternary ammonium compounds and sulphonamides, respectively. Four different P_ant promoters have been described (Bunny et al., 1995; Stokes & Hall, 1989), and their relative strengths have been compared with the Escherichia coli tac promoter (Collis & Hall, 1995; Lévesque et al., 1994). TTGACA-N₁₇-TAAACT and TGGACA-N₁₇-TAAGCT have been described as strong and weak P_ant promoters, respectively, whilst TTGACA-N₁₇-TAAGCT and TGGACA-N₁₇-TAAACT are described as hybrid promoters. In this study, we analysed the relevance of resistance determinants (bla_IMP-1, bla_IMP-7, bla_IMP-10 and bla_VIM-2) and resistance levels to carbapenems in imipenem (IPM)-resistant clinical isolates of P. aeruginosa, and identified the associated integrons.

Bacterial strains and reagents. Twenty-seven IPM-resistant strains of P. aeruginosa (MIC ≥8 µg ml^–1) and one susceptible strain of P. aeruginosa P01 were used. They were isolated from unrelated inpatients at Showa University Hospital, Tokyo, Japan, from 2005 to 2008. Mueller–Hinton (MH) broth (Becton Dickinson) supplemented with 25 mg Ca²⁺ l^–1 and 12.5 mg Mg²⁺ l^–1 was used for bacterial cell culture and antibiotic-susceptibility testing. The following reagents were purchased from commercial sources: IPM (US Pharmacopeia); meropenem (MEPM) and biapenem (BIPM) (both from LKT Laboratories); panipenem (PAPM; Sankyo Organic Chemicals); ceftazidime (CAZ) discs (Becton Dickinson) and sodium mercaptoacetic acid (SMA) discs (Eiken Chemical).

PFGE. PFGE of SpeI-digested genomic DNA from P. aeruginosa isolates was carried out using a CHEF-DRIII system (Bio-Rad) according to the manufacturer's instructions. The restriction patterns were evaluated according to a standard, as described previously (Tenover et al., 1995).

Susceptibility testing. The susceptibilities of bacteria to the various antibiotics were expressed in terms of their MICs. MICs were determined by a broth microdilution method according to Clinical and Laboratory Standards Institute (formerly the National Committee for Clinical Laboratory Standards) guidelines (NCCLS, 2000). Bacterial cells (5x10⁵ ml^–1) were inoculated into 100 µl MH broth containing twofold serial dilutions of the carbapenems, and MICs were determined after incubation at 35 °C for 18 h. All experiments were performed at least three times to establish reproducibility.

Detection of MBL production. Double-disc synergy testing was performed to detect the production of MBLs by using discs containing CAZ, a third-generation cephalosporin, and SMA, a specific inhibitor of MBLs (Arakawa et al., 2000). Bacterial cells (2x10⁷) were inoculated on an MH agar plate (Becton Dickinson) and discs were then set on the plate with a distance of 1.5 cm between the CAZ and SMA discs. After incubating the plates at 35 °C for 24 h, strains that had an inhibition zone (>12 mm) around the CAZ disc towards the SMA disc were considered to be MBL-producing strains.

Identification of resistance gene cassettes and associated integrons by PCR and DNA sequencing. Bacterial genomic DNA was extracted using a commercial kit (Qiagen). The primers used for detecting resistance genes and other associated genes are listed in Table 1. The primer pairs intI1-F/qacEΔ1-R and intI1-F/sul1-R were used for amplification of the integrons containing bla_IMP, and the primer pair intI1-F and vim2-R was used for amplification of the integron containing bla_VIM-2. The amplicons were purified using a QIAquick PCR purification kit and sequenced on an Applied Biosystems 3730xl DNA analyser. Sequences were compared with those in GenBank by using the BLAST network service.

Table 1. Primers used for detecting resistance genes and sequence identification of the amplified genes

Relevance of resistance levels and resistance determinants
As all of the strains were from the same hospital, we performed PFGE to exclude the possibility that the isolates might be identical. Only isolates confirmed as being different were used in this study. The different patterns of MICs, growth speeds and abilities to produce the pigments, and the different gene cassettes in integron 1, as described below, also indicated that the studied isolates were clonally unrelated.

Of 27 IPM-resistant clinical isolates of P. aeruginosa, 15 strains were putative MBL producers. Of these 15 MBL producers, the bla_IMP-1, bla_IMP-7, bla_IMP-10 and bla_VIM-2 genes were harboured by 2, 5, 7 and 1 of the strains, respectively. The 27 isolates could be divided into three groups according to their resistance levels to carbapenems (Table 2). Group I was highly resistant to carbapenems with MICs of 512 to 2048, 2048 to >4096, 1024 to 4096 and 128 to 2048 µg ml^–1 for IPM, MEPM, PAPM and BIPM, respectively. By comparison, group II showed an intermediate level of resistance with MICs of 32 to 256, 128 to 512, 256 to 512 and 16 to 512 µg ml^–1, respectively. Group III showed a low level of resistance with MICs of 8 to 32, 4 to 64, 8 to 64 and 4 to 32 µg ml^–1, respectively. Double-disc (CAZ and SMA) synergy testing confirmed that the strains in group I (n=7) and II (n=8) produced MBLs, whilst the strains in group III (n=12) did not. Furthermore, it was confirmed by PCR and DNA sequencing analysis that all of the strains in group I harboured bla_IMP-10, whilst those in group II harboured bla_IMP-1, bla_IMP-7 or bla_VIM-2. MBL-encoding genes were not detectable in group III. Of the group II strains carrying bla_IMP-1, bla_IMP-7 or bla_VIM-2, no significant differences were observed in their resistance to carbapenems.

Table 2. Phenotypes and genotypes of IPM-resistant P. aeruginosa clinical isolates

The above results indicated that the production of MBLs was a crucial factor in conferring high and intermediate levels of resistance and that the bla_IMP-10 allele was responsible for the observed cases of extremely high resistance to carbapenems. These data were also supported by the experiment using the specific MBL inhibitor SMA, which greatly reduced the resistance level of MBL-producing strains. For example, the MIC of IPM against strain P0701 decreased from 512 to 32 µg ml^–1 in the presence of 0.5 mM SMA. The low-level resistance to carbapenems exhibited by the non-MBL-producing strains was probably caused by some non-MBL factors, such as the low permeability of the outer membrane due to loss of OprD, the activity of the efflux system due to overexpression of the MexAB–OprM and MexEF–OprN pumps, or the production of chromosomal AmpC β-lactamases, as well as the cooperation of such factors (Li et al., 2000; Masuda et al., 1999; Nikaido, 1989, 1996; Pai et al., 2001; Quale et al., 2006). Further experiments are needed to clarify the details of the underlying mechanism(s).

Structure of integrons associated with MBL-encoding genes from IPM-resistant strains of P. aeruginosa
By sequencing the PCR products, the class 1 integron gene cassettes were identified from 15 MBL-producing strains; no class 2 or class 3 integrons were detected. Based on GenBank sequences (Table 3), five of the integrons were a typical class 1 integron, containing a 5'-CS intI1, a 3'-CS qacEΔ1 and sul1, and a variable region (Fig. 1a, b, c, d, e). One integron carried only part of qacEΔ1, the 132 bp at the 3'-end of the total 348 bp gene (Fig. 1f). The integron carrying bla_VIM-2 did not possess the 3'-CS (Fig. 1g). The bla_IMP-1, bla_IMP-7 and bla_IMP-10 gene cassettes were preceded by a hybrid P_ant promoter, TGGACA-N₁₇-TAAACT, located at the 5' end of the integrase 1-encoding gene. The bla_VIM-2 gene cassette was preceded by a weak P_ant promoter, TGGACA-N₁₇-TAAGCT. Except for in the integron shown in Fig. 1(a), the bla_IMP and bla_VIM-2 genes were linked to resistance genes encoding aminoglycoside-modifying enzymes, such as aac(6')Iae, aac(6')II, aacA7, aacC4, aadA1, aadA2 and aadA6 (Fig. 1).

Table 3. GenBank sequences used for identification of the amplified genes

(20K): Fig. 1. Schematic representation of the structures of integrons containing blaIMP or blaVIM. The class 1 integrons were identified from 15 MBL-producing clinical isolates of P. aeruginosa. Arrows represent genes and their transcriptional direction. Short lines represent attI sites. Circles represent attC sites (e.g. the 59 base element).

We have reported previously that S. marcescens strains harbouring bla_IMP-1 and bla_IMP-10 genes showed quite different patterns of MICs, as with the strains of P. aeruginosa described in this study. Using E. coli transconjugants and transformants, it was also observed that the bla_IMP-10-positive strains showed much higher levels of resistance to IPM, MEPM and PAPM than the strains harbouring bla_IMP-1 (Hu & Zhao, 2009). The E. coli transconjugants and transformants are more suitable models than wild-type strains of S. marcescens and P. aeruginosa because factors such as the outer-membrane barrier, the efflux system and the chromosomal AmpC β-lactamases that are possibly possessed by these wild-type strains can be excluded. A strain of bla_IMP-10-positive P. aeruginosa has also been reported to show extremely high-level resistance against IPM (Takeda et al., 2008).

The mechanism for this phenomenon is unclear at present. One possible explanation is that IMP-10 MBL might be more potent than IMP-1 MBL in hydrolysing IPM, MEPM and PAPM. In a published paper, the kinetic parameters of purified IMP-1 and IMP-10 MBL in hydrolysing IPM, MEPM and other β-lactams were described (Iyobe et al., 2002). Compared with the rate of hydrolysis (K_cat) of IMP-1 MBL (130 and 13 s^–1 for IPM and MEPM, respectively), IMP-10 MBL showed higher values of K_cat (220 and 64 s^–1). The higher K_cat values of IMP-10 MBL compared with those of IMP-1 MBL could partly explain the phenomenon described in this study. However, no significant differences of the specificity constant (K_cat/K_m) were observed between IMP-1 and IMP-10 MBLs. Further experiments are therefore needed to reveal the detailed mechanism(s) involved.

Conclusion
The molecular detection and surveillance of resistance genes harboured by P. aeruginosa is becoming increasingly important for assessment and control of their spread and colonization in hospitals, and for guiding treatment of infections. In the IPM-resistant clinical isolates of P. aeruginosa, the MBL-producing strains showed much higher resistance to carbapenems than non-MBL-producing strains. Of the MBL producers, the strains bearing bla_IMP-10 showed extremely high resistance compared with strains bearing bla_IMP-1, bla_IMP-7 and bla_VIM-2. In addition to the MBL-encoding genes, the class 1 integrons generally harboured one or two aminoglycoside-resistance genes, highlighting the multidrug-resistant properties of these clinical isolates.

This study was funded in part by a research grant from Showa University, Tokyo, Japan.