|Home | About | Journals | Submit | Contact Us | Français|
The contributions of different mechanisms of resistance to carbapenems among a collection of imipenem- and meropenem-nonsusceptible Pseudomonas aeruginosa isolates were investigated. This screening included the recently reported extended-spectrum cephalosporinases (ESACs) weakly hydrolyzing carbapenems. Eighty-seven percent of the studied isolates were resistant to imipenem. Genes encoding metallo-β-lactamases or carbapenem-hydrolyzing oxacillinases were not identified. The main mechanism associated with imipenem resistance was the loss of outer membrane protein OprD. Identification of overexpressed ESACs and loss of OprD were observed for 65% of the isolates, all being fully resistant to imipenem. Resistance to meropenem was observed in 78% of the isolates, with all but one also being resistant to imipenem. Overexpression of the MexAB-OprM, MexXY-OprM, or MexCD-OprJ efflux systems was observed in 60% of the isolates, suggesting the contribution of efflux mechanisms in resistance to meropenem. The loss of porin OprD and the overproduction of ESACs were observed in 100% and 92% of the meropenem-resistant isolates, respectively. P. aeruginosa can very often accumulate different resistance mechanisms, including ESAC production, leading to carbapenem resistance.
Carbapenems remain the main antimicrobials for treating infections due to multidrug-resistant Pseudomonas aeruginosa, but the development of carbapenem resistance may significantly compromise their efficacy. In the absence of carbapenem-hydrolyzing enzymes (mainly metallo-β-lactamases [MBLs]), carbapenem resistance is usually multifactorial. Increased production of AmpC chromosome-encoded cephalosporinase, reduced outer membrane porin OprD expression, and associated factors are known to contribute to carbapenem resistance (12, 13, 21). Overexpression of the MexAB-OprM efflux system is known to affect meropenem efficacy but not that of imipenem (14, 23). In addition, the MexCD-OprJ and MexXY-OprM efflux systems may also be involved in reduced susceptibility to meropenem (1, 14, 20).
We have recently reported extended-spectrum cephalosporinases (ESACs) of P. aeruginosa with broadened hydrolytic activity toward imipenem (22). All those naturally occurring chromosome-encoded AmpC-type β-lactamases possess an alanine residue at position 105 that has a slight carbapenemase activity that is sufficient to compromise the efficacy of carbapenems when the enzyme is overexpressed (22). The isolates studied here were selected on the basis of nonsusceptibility to carbapenems, and we aimed at evaluating the contributions of the different mechanisms to their resistance to carbapenems.
Thirty-two nonduplicated P. aeruginosa clinical isolates recovered at the Bicêtre Hospital (K.-Bicêtre, France) in 2007 were included in this study (Tables (Tables11 and and2).2). Those isolates were selected on the basis of intermediate susceptibility (MIC = 4 μg/ml) or resistance (MIC ≥ 8 μg/ml) to imipenem and meropenem (3). P. aeruginosa clinical isolates were identified using the API 32GN system (bioMerieux, Marcy-l'Étoile, France).
P. aeruginosa PAO1 was used as a reference strain in susceptibility testing and β-lactamase assays. P. aeruginosa H729 lacking the oprD gene was used as a OprD-negative reference strain. The MBL VIM-2-producing P. aeruginosa isolate COL-1 (19) was used as a positive control for carbapenemase production. Bacterial cells were grown in Trypticase soy broth or on Trypticase soy agar plates (Sanofi Diagnostics Pasteur, Marnes-la-Coquette, France) supplemented with antibiotics when required.
The antimicrobial agents and their sources have been described elsewhere (16). MICs were determined by using the Etest technique (AB Biodisk, Solna, Sweden) and agar dilution method as previously described (3, 16). Results of susceptibility testing were recorded according to the guidelines of the CLSI (3). AmpC overproduction was confirmed using cloxacillin (250 μg/ml)-containing plates, since cloxacillin inhibits AmpC β-lactamase activity (15). The isolates were considering as overexpressing their blaampC genes when there was at least a twofold dilution difference between the MIC of ceftazidime and the MIC of ceftazidime plus cloxacillin.
Pulsed-field gel electrophoresis was performed according to the instructions of the manufacturer (Bio-Rad, Marnes-la-Coquette, France), as previously described (18). Results were analyzed according to the criteria of Tenover et al. (24).
AmpC β-lactamase-specific activity (μmol of cephalothin [cefalotin] hydrolyzed per minute and per milligram of protein) was determined by UV spectrophotometry with crude culture extracts, as described previously (8, 11, 22). Extracts showing >90% reduction of β-lactamase activity after cloxacillin addition were considered to produce AmpC-type β-lactamase as a major contributor of β-lactamase activity.
The carbapenemase activity was determined as previously described (19). Briefly, hydrolyses of imipenem and meropenem (100 μM solution in 100 mM sodium phosphate [pH 7.0]) by crude cell extracts obtained after sonication in phosphate buffer were measured spectrophotometrically at 297 nm and 298 nm, respectively.
Additionally, the presence of genes coding for MBL-type (blaIMP, blaVIM, blaSPM, and blaAIM) and carbapenem-hydrolyzing class D (blaOXA-23, blaOXA-40, blaOXA-48, and blaOXA-58) β-lactamases was searched out by PCR amplification using previously described primers (Table (Table3).3). Whole-cell DNA was extracted as described previously (2).
Outer membrane protein (OMP) profiles were examined using previously reported methods (7). Following sonication, membranes were collected by ultracentrifugation at 50,000 × g for 35 min. OMPs were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
For each PCR experiment, 500 ng of total DNA was used in a standard PCR. The primers used have been previously described (Table (Table3)3) (22). Sequence analyses were performed using software available on the Internet at www.ncbi.nlm.nih.gov and http://www.ebi.ac.uk/clustalw/.
PCR amplification of oprD was performed by using the primers described in Table Table3.3. Sequences were compared with that of reference strain P. aeruginosa PAO1.
The levels of expression of oprD, mexB, mexY, and mexD were quantified using the one-step real-time quantitative reverse transcription PCR (RT-PCR), as recommended by Fey et al. (6). Expression level results were standardized according to the transcription level of the constitutively expressed rpsL ribosomal gene (Table (Table3)3) (8). Transcript quantification was performed by using the QuantiFast SYBR green RT-PCR kit on a LightCycler 1.0 instrument (Roche Diagnostics, Neuilly-sur-Seine, France). Isolates were considered to be MexAB-OprM, MexXY-OprM, MexCD-OprJ hyperproducers when the level of expression of mexB, mexY, or mexD were at least two-, four-, or twofold higher than that of PAO1, respectively, according to previously defined criteria (10, 21). Reduced oprD expression was considered relevant when it was ≤30% compared to that of the P. aeruginosa PAO1 reference strain (21).
Thirty-two (6.7%) out of the 474 nonduplicate P. aeruginosa isolates that had been recovered at the Bicêtre Hospital in 2007 were categorized as nonsusceptible to imipenem and to meropenem and therefore retained for this study. Overall, 87.5% and 78.1% of the selected isolates were resistant to imipenem and meropenem, respectively (Table (Table1).1). Genotyping performed using pulsed-field gel electrophoresis revealed high clonal diversity, with 22 distinct clones identified among the 32 isolates. One genotype (clone E) included eight isolates with two different subclones, E1 (two isolates) and E2 (six isolates).
All the isolates had reduced oprD expression (Table (Table2).2). Twenty-four isolates showed relative oprD expression of ≤10% compared to that of P. aeruginosa PAO1 (Table (Table2).2). Reduced expression of OprD was also assessed by OMP analysis (data not shown).
In order to evaluate the possible impact of qualitative modifications of OprD, the sequence of the oprD gene was also determined. Twenty-four isolates (17 distinct clones) out of the 32 isolates contained inactivating mutations of the oprD gene (Tables (Tables22 and and4).4). The most frequent mechanisms of oprD inactivation resulted from 1-bp insertions/deletions or point mutations, leading to premature stop codons identified in 11 and 5 isolates, respectively. In two isolates, the OprD inactivation was due to a 4- or 5-bp insertion inside the coding sequence (Table (Table4).4). Finally, the inactivation of OprD was the result of a disruption of the coding sequence by a novel 1.3-kb insertion sequence (IS) element, named ISPa27 (submitted to the IS database at the Internet site http://www-is.biotoul.fr/), in six clonally related isolates (clone E2).
We also identified a shortened putative loop L7 of the OprD porin in the clones B2, O, and W (Table (Table4).4). This shortening may open the porin channel to allow optimal penetration of meropenem and increase its activity (5). Conversely, this alteration might not modify the susceptibility to a smaller-sized carbapenem molecule, such as imipenem (5). Interestingly, downregulated expression of the oprD gene observed in these clones may modulate the phenotype of imipenem and meropenem resistance (Table (Table22).
Synergy tests performed with clavulanic acid-ticarcillin and ceftazidime-containing disks did not evidence any inhibition of ceftazidime resistance for all the P. aeruginosa isolates, ruling out the production of extended-spectrum β-lactamases. Using cloxacillin-containing plates, the MICs of ceftazidime were decreased by at least twofold for 25 out of 32 isolates (78% of total), suggesting overproduction of AmpC that was confirmed by determination of β-lactamase activities (Table (Table2).2). Twenty-one out of those 25 isolates overexpressed the AmpC β-lactamase and had reduced MICs of ceftazidime but also of imipenem and cefepime after cloxacillin addition (Table (Table1),1), suggesting the presence of an ESAC (Tables (Tables11 and and2).2). PCR and sequencing identified 10 AmpC β-lactamase variants, with 8 of those variants possessing an alanine residue at position 105 (Table (Table2).2). This residue had been previously shown to be the key factor for an ESAC property (Table (Table2)2) (22). All isolates were negative for blaIMP-, blaVIM-, blaSPM-, blaAIM-, and blaOXA-type genes by PCR screening.
Overexpression of the MexAB-OprM, MexXY-OprM, and MexCD-OprJ efflux systems was demonstrated in 28%, 37.5%, and 31% of the carbapenem-nonsusceptible isolates, respectively, suggesting the contribution of efflux in the resistance (Table (Table2).2). Seven isolates overexpressed only one efflux pump, and 12 isolates overexpressed two efflux pumps. Sixty percent of the isolates overexpressed at least one efflux pump. No isolate overexpressed all three efflux pumps.
Whereas OprD inactivation alone is the source of intermediate susceptibility or resistance to imipenem, the mechanisms leading to meropenem resistance seem to be more complex and are very likely multifactorial, involving overproduction of AmpC or overexpression of the efflux pumps MexAB-OprM, MexXY-OprM, and MexCD-OprJ (4, 21). In our collection, clonal spread played a rather minor role in the epidemiology of infections caused by P. aeruginosa, and the main mechanism associated with reduced susceptibility or resistance to imipenem was the absence or weak expression of OprD found in all the isolates.
Our data show that both oprD gene downregulation and OprD protein inactivation contributed to resistance to imipenem and reduced susceptibility to meropenem. Noteworthy, the inactivating mutations identified in the oprD gene of isolates Paeβ-05 and Paeβ-23 (corresponding to clones U and J, respectively) (Tables (Tables22 and and4)4) had been previously identified in carbapenem-resistant P. aeruginosa isolates from Spain, Portugal, and the United States (8, 17).
Noteworthy, the prevalence of the ESAC-encoding genes in our collection was very high, with 28 out of the 32 isolates. In addition, overproduction of those ESAC enzymes was evidenced in most of the isolates (75%). Basically, when considering the high rate of OprD-deficient isolates, it might be hypothesized that ESAC overproduction could play a secondary but additive role toward reduced susceptibility or resistance to imipenem. Of note, the additive presence of overexpressed ESACs and loss of OprD was correlated with resistance to imipenem in all cases (Tables (Tables11 and and22).
On the other hand, the main mechanism associated with reduced susceptibility or resistance to meropenem was probably overexpression of efflux pumps identified in 60% of the isolates, together with the loss of OprD. ESAC overexpression (65% of the isolates) could play an additive role in reduced susceptibility or resistance to meropenem. Thirteen isolates did not overexpress any efflux pump, suggesting that the lack of OprD, overexpression of AmpC, or production of ESAC β-lactamases (or other unknown mechanisms) may explain meropenem resistance (9). The additive presence of overexpressed ESACs and overexpression of at least one efflux pump correlated with resistance to meropenem in 87.5% of the cases (14 out of 16 isolates with this combination of mechanisms of resistance) (Tables (Tables11 and and22).
Our results show that P. aeruginosa may very often accumulate different resistance mechanisms (overproduction of AmpC cephalosporinase, ESACs, increased drug efflux, and deficient production or loss of porin OprD). One of the most interesting and original results obtained from the present study is that P. aeruginosa-type ESACs probably contribute to carbapenem resistance.
This work was funded by a grant from the Ministère de l'Education Nationale et de la Recherche (UPRES-EA3539), Université Paris XI, Paris, France, and mostly by grants from the European community (DRESP2, LSHM-CT-2005-018705; TROCAR, HEALTH-F3-2008-223031) and the INSERM. J.-M.R.-M. was funded by a postdoctoral grant from the Ministerio de Educación y Ciencia from Spain (2007/0292).
Published ahead of print on 8 September 2009.