Rapid Flow Cytometry Test for Identification of Different Carbapenemases in Enterobacteriaceae

Abstract

A flow cytometry test was developed to identify carbapenemase production by Enterobacteriaceae and to discriminate between the different types of carbapenemases (classes A, B, and D). It is based on the detection of meropenem activity against bacteria, coupled with different carbapenemase inhibitors, which is assessed by flow cytometry. It represents a convenient, fast, and reliable approach (100% sensitivity and 100% specificity) for the detection and characterization of different carbapenemases.

TEXT

In the past decade, the prevalence of resistance to β-lactam antibiotics mediated by β-lactamases has increased at an alarming rate, especially among Enterobacteriaceae (1), originally because of the spread of extended-spectrum β-lactamases (ESBLs) and more recently due to carbapenemases, which are enzymes able to hydrolyze carbapenems. These enzymes have also spread rapidly throughout the world (2) and are an alarming problem in some countries, especially in Greece and India (3). Carbapenemases include class A (such as Klebsiella pneumoniae carbapenemase [KPC]), class B (metallo-β-lactamases [MBLs], such as VIM and NDM), and class D (OXA-48-like, such as OXA-181) enzymes. Invasive infections with carbapenemase-producing Enterobacteriaceae (CPE) are associated with high morbidity and mortality (4, 5). Accurate detection of CPE is relevant at the individual patient level due to not only the impact on antibiotic therapy but also infection control purposes, especially in outbreak settings. The European Committee on Antimicrobial Susceptibility Testing (EUCAST) recently developed guidelines for carbapenemase detection in Enterobacteriaceae isolates (6). These recommendations must be adopted whenever decreased susceptibility to imipenem, meropenem, and/or ertapenem is detected (i.e., MIC values of >1, >0.125, or >0.125 μg/ml, respectively). However, 48 h is usually required after sample collection to determine MIC values, and an additional 18 to 24 h is needed to confirm the presence of carbapenemase. The combination disk test is a unique convenient phenotypic method for discriminating among different carbapenemases (7–9). In brief, disks or tablets containing meropenem with or without various inhibitors, such as aminophenylboronic acid (APBA) as a class A inhibitor and EDTA as a class B inhibitor, are used. There is no currently available inhibitor for class D carbapenemases, although they have shown high-level resistance to temocillin (MICs of >32 μg/ml) (9). Decreased susceptibility of Enterobacteriaceae to carbapenems may also be caused by the production of either ESBLs or AmpC β-lactamases coupled with decreased permeability due to alteration or downregulation of porins (10). Cloxacillin, which inhibits AmpC β-lactamases, should be added to the test to differentiate between AmpC hyperproduction plus porin loss and carbapenemase production. Because it is based on growth in the presence of the drugs, the main disadvantage of such a procedure is that it requires a long incubation period. The same drawback is seen with Etest KPC and MBL strips, which associate a carbapenem with boronic acid or EDTA, respectively (11). Molecular identification of carbapenemase genes (12, 13) is a possibility, but because there are several genes that must be searched and the cost associated with the search is high, it can result in false-negative results when mutations or new genes emerge. The chromogenic culture medium tests currently available, although easy to perform, are growth dependent (14). More recently, different biochemical tests for detecting carbapenemase production, such as Carba NP, were described (15); the Carba NP test seems to be rapid, sensitive, and specific but cumbersome and expensive. Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), introduced recently in laboratory routines, seems promising for the detection of antibiotic modifications due to degradation by enzymes, including carbapenemases (16).

Flow cytometry (FC) represents an accurate and fast approach for the analysis of cell architecture and its functional parameters, and it offers considerable advantages over conventional methods. Similar to ESBL detection by FC (17), a novel protocol for carbapenemase detection was developed and evaluated using a comprehensive set of phenotypically and molecularly well-characterized strains. Included in this study was a panel of 30 carbapenemase-producing Enterobacteriaceae strains, including 8 KPC, 7 VIM, 1 IMP, 4 NDM, 1 IMI, 1 GIM, 6 OXA-48-like, and 2 KPC-plus-VIM strains. As noncarbapenemase producers, 30 Enterobacteriaceae strains, including several strains that give false-positive results on phenotypic tests (18), such as ESBL or AmpC producers coupled to impermeability, were included. A novel flow cytometric test for carbapenemase detection (provided by FASTinov S.A., Porto, Portugal) as a proof of concept was used. A bacterial suspension in brain heart medium (Merck) of approximately 10⁶ cells/ml at initial log phase (optical density at 600 nm ≥ 0.2, which takes approximately 1.5 h) was incubated with meropenem (2 and 8 μg/ml) with or without the inhibitors APBA (2.5 and 5 mM), EDTA (6.25 and 12.5 mM), APBA plus EDTA (2.5 plus 6.25 mM and 5 plus 12.5 mM), and cloxacillin (250 and 500 μg/ml). Temocillin (Eumedica, Brussels, Belgium) alone at 32 and 64 μg/ml was also included. A fluorochrome dye was disposed in all tubes. Cells were incubated for 1 h at 37°C and analyzed in a flow cytometer apparatus (FACSCalibur; Becton Dickinson). The intensity of fluorescence (IF) of the cells treated with meropenem alone was compared with that of those exposed to meropenem plus inhibitors. An increase (of at least twice) in the IF of cells in the presence of meropenem with inhibitors, when compared with meropenem alone, indicates that meropenem was active against bacterial cells; this result was observed with APBA for all KPCs, which correspond to class A carbapenemases. All VIM, IMP, NDM, IMI, and GIM enzymes showed an increase in IF when cells were treated with meropenem with EDTA, which corresponds to class B carbapenemases. Strains with both KPC and MBL enzymes had an increase in IF when cells were treated with meropenem combined with only APBA plus EDTA. However, for class D carbapenemases, no increase was observed. They were resistant to temocillin; the IF of the cells exposed to temocillin was similar to that of nonexposed cells. For AmpC associated with porin loss, the increase in the IF of the cells was observed with APBA but also with cloxacillin. In the case of ESBL associated with porin loss, no increase in IF was observed; such strains were susceptible to temocillin, increasing the IF of the cells exposed to this drug. Figure 1 shows the median ratios of the IF of each type of carbapenemase producers to that of a group of nonproducers. A typical example of flow cytometry data of the different types of carbapenemases is shown in Fig. 2. This FC test combines excellent sensitivity (100%) and specificity (100%) and a short time to results. Moreover, dedicated software that makes it very easy to use was developed. In addition, this test can be applied directly to bacterial colonies or directly to positive blood cultures (data not shown). The estimated costs for this new test, not including the equipment, are rather low (around 2.5 euros). FC equipment is found in most hematology/immunology laboratories. The early detection of carbapenemases can have a real impact on treatment (19, 20) and outbreak management, preventing its spread. The improvement in detection and surveillance of CPE might contribute to their epidemiological control.

FIG 1.

Mean stain indexes (SIs) of each group of strains, including KPCs, MBL carbapenemases, OXA-48-like carbapenemases, and nonproducers. The SI is the ratio of the intensity of fluorescence at FL1 (530/30 nm) of cells treated with meropenem (MEM) plus the inhibitor(s) (aminophenyl boronic acid [APBA], EDTA, and cloxacillin) to that of cells treated with only meropenem. For temocillin, the SI is the ratio of the fluorescence intensity of cells treated with temocillin to that of nontreated cells. For each strain, the lower concentrations of MEM with inhibitors and of temocillin that exhibited an increase in intensity of fluorescence were chosen to perform the calculation. For the nonproducers, the maximum drug concentrations were considered. The dotted line represents an SI of 2, which indicates an increase of two times the intensity of fluorescence.

FIG 2.

Representative histograms obtained by flow cytometry for different types of carbapenemases. For each strain (Klebsiella pneumoniae KPC-2, Enterobacter cloacae VIM-4, and K. pneumoniae OXA-48), the cells not exposed to drugs, cells exposed to 8 μg/ml meropenem (MEM) and with the inhibitors 5 mM aminophenylboronic acid (APBA), 12.5 mM EDTA, 5 mM APBA plus 12.5 mM EDTA, or 500 μg/ml cloxacillin, and cells exposed to 32 μg/ml temocillin are shown. IF, intensity of fluorescence.