Additional characteristics of Bacteroides fragilis carbapenemases belonging to group 3a.

In their excellent review of the classification of carbapenemhydrolyzing metallo-b-lactamases, Rasmussen and Bush (6) proposed that such enzymes produced by Bacteroides fragilis should be assigned to a new functional subgroup, 3a. The characteristics of b-lactamases from three B. fragilis strains were cited and included those from B. fragilis QMCN3 and B. fragilis QMCN4 (7, 8). To avoid confusion in the literature, we wish to point out that these strains were first isolated at the Queen’s Medical Centre, Nottingham, United Kingdom (not Queen’s Medical College, London, United Kingdom, as stated in earlier reports by Rasmussen et al. [7, 8]) and were originally designated B. fragilis 57 and B. fragilis 97, respectively (5). In tabulating features of these enzymes for comparison with other group 3 b-lactamases, Rasmussen and Bush indicated that data on inhibition profiles, pI values, effect of zinc ions, and the relative hydrolysis of meropenem and imipenem were incomplete. However, these and other data, which would be useful when considering the characteristics of this group of enzymes, are available. Resistance (50% inhibitory concentration of .100 mM) of b-lactamases from these strains to six b-lactamase inhibitor compounds, including clavulanic acid and sulbactam, was reported in 1986 (5). The effect of cations on their activity has also been examined in detail (2). Zinc ions were shown to enhance activity and to be required for restoration of .90% of the activity after inhibition by EDTA. Partial restoration of activity (39 to 45%) was achieved with cobalt ions, although manganese, magnesium, and iron had no or minimal effect. The pI value of the metallo-b-lactamase of B. fragilis 97 (QMCN4) was found to be 4.6 (5) and that of B. fragilis 57 (QMCN3), initially reported as 4.2 (5), has been found to be 4.7 after repeated testing (3). Also, b-lactamase-mediated hydrolysis of meropenem has been shown to be at least as efficient as hydrolysis of imipenem (1). Unlike Rasmussen et al. (8), we have not detected coproduction of other b-lactamases with B. fragilis 57 (QMCN3), either by isoelectric focusing following pretreatment of b-lactamase extracts with EDTA or by demonstration of residual cephalosporinase activity in the presence of the chelator (3). Interestingly, neither B. fragilis 57 nor B. fragilis 97 exhibits high-level resistance to carbapenems. Indeed the MICs of imipenem and meropenem for B. fragilis 97 (QMCN4) are 0.5 and 2 mg/liter, respectively (1), within the normal breakpoints for susceptibility. However, the conversion of B. fragilis from lowto high-level resistance, which has been observed in a patient undergoing therapy with imipenem (9), underlines the potential clinical importance of these enzymes. Strains exhibiting low-level resistance to imipenem currently comprise about 7% of isolates of anaerobic gram-negative bacilli in Nottingham (4). They are easily overlooked by routine susceptibility testing and may pose a more significant threat than has been supposed.