Preferred β-lactone synthesis can explain high rate of false-negative results in the detection of OXA-48-like carbapenemases

Abstract

The resistance to carbapenems is usually mediated by enzymes hydrolyzing β-lactam ring. Recently, an alternative way of the modification of the antibiotic, a β-lactone formation by OXA-48-like enzymes, in some carbapenems was identified. We focused our study on a deep analysis of OXA-48-like-producing Enterobacterales, especially strains showing poor hydrolytic activity. In this study, well characterized 74 isolates of Enterobacterales resistant to carbapenems were used. Carbapenemase activity was determined by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), liquid chromatography/mass spectrometry (LC-MS), Carba-NP test and modified Carbapenem Inactivation Method (mCIM). As meropenem-derived β-lactone possesses the same molecular weight as native meropenem (MW 383.46 g/mol), β-lactonization cannot be directly detected by MALDI-TOF MS. In the spectra, however, the peaks of m/z = 340.5 and 362.5 representing decarboxylated β-lactone and its sodium adduct were detected in 25 out of 35 OXA-48-like producers. In the rest 10 isolates, decarboxylated hydrolytic product (m/z = 358.5) and its sodium adduct (m/z = 380.5) have been detected. The peak of m/z = 362.5 was detected in 3 strains co-producing OXA-48-like and NDM-1 carbapenemases. The respective signal was identified in no strain producing class A or class B carbapenemase alone showing its specificity for OXA-48-like carbapenemases. Using LC-MS, we were able to identify meropenem-derived β-lactone directly according to the different retention time. All strains with a predominant β-lactone production showed negative results of Carba NP test. In this study, we have demonstrated that the strains producing OXA-48-like carbapenemases showing false-negative results using Carba NP test and MALDI-TOF MS preferentially produced meropenem-derived β-lactone. We also identified β-lactone-specific peak in MALDI-TOF MS spectra and demonstrated the ability of LC-MS to detect meropenem-derived β-lactone.

Figure 1

LC–MS analysis of meropenen before and after exposure to the different OXA-48 strains. Extracted Ion Chromatograms (EIC) of m/z = 384.15 (± 0.2Th) show retention behavior of meropenem and meropenem-derived β-lactone isomer. Insets show mass spectra from the apex of each chromatographic peak. (A) EIC of pure meropenem standard; (B) negative control, EIC of meropenem exposed only to E. coli ATCC 14,169; (C) positive control, EIC of meropenem exposed to strain P759. There is no chromatographic peak corresponding to meropenem, the inset mass spectrum shows degradation product at m/z = 358 at the retention time of meropenem standard; (D) meropenem exposed to strain P370 and partially converted to β-lactone isomer; EIC shows chromatographic resolution of original meropenem (earlier peak) and newly formed β-lactone isomer of meropenem (later peak). Inset mass spectrum for the early peak is identical with the standard meropenem spectrum and with the negative control spectrum, while mass spectrum of newly formed isomer also contains peak at m/z = 340; (E) meropenem exposed to strain P370 and fully converted to the β-lactone, as can be seen from the shift in retention time. Mass spectrum contains peaks at m/z = 340 and 384.

Figure 2

MALDI-TOF MS analysis of carbapenemase activity. Panel (A) shows a native meropenem (m/z = 384.5) and its sodium adduct (m/z = 406.5). Panel (B) shows a positive control—KPC-2-producing Klebsiella pneumoniae ST258. Panel (C) represents the sample P370 with a visible sodium adduct of meropenem-derived β-lactone (m/z = 362.5). Panel (D) shows the strain P759 with decarboxylated hydrolyzed meropenem (m/z = 358.5) and its sodium adduct (m/z = 380.5).

Figure 3

Meropenem, meropenem-derived β-lactone and their decarboxylated molecules with an average molecular weight (MW).