Covalent Trapping and Bacterial Resistance to Ceftazidime

Levitt et al. (1) suggest that “covalent trapping” explains the increased ceftazidime resistance of cells producing the R164S mutant of KPC-2. This hypothesis appears to be unrealistic. The rate of diffusion of the antibiotic into the periplasm at an external concentration corresponding to the minimum inhibitory concentration (MIC) is v_in = PA (MIC − Ipl) where P (cm s⁻¹) is the permeability coefficient, A is the surface area (132 cm²/mg dry weight), and Ipl is the “lethal antibiotic periplasmic concentration.” With Escherichia coli, the value of Ipl is well approximated by the MIC of a strain devoid of β-lactamase. MIC and Ipl are in μm and v_in is in nanomoles s⁻¹ (mg dry weight)⁻¹. The value of P for ceftazidime has been measured by two different methods yielding 16 and 96 × 10⁻⁷ cm s⁻¹ (for a detailed review see Docquier et al. (2)).

The MICs for the nonproducing and the R164S-producing strains are 0.4 and 200 μm, respectively. With an intermediate value of P (40 × 10⁻⁷ cm s⁻¹), v_in = 0.132 nmol s⁻¹ (mg dry weight)⁻¹. If covalent trapping contributes to the MIC, the rate of β-lactamase synthesis should be similar to v_in, and the cells corresponding to 1 mg of dry weight must synthesize about 1 mg of β-lactamase over a period of 5 min. Moreover, at the lethal periplasmic concentration (0.4 μm), the acyl-enzyme represents at most 0.22% of the total enzyme because the K_m of the R164S mutant for ceftazidime is 180 μm. Clearly, the amount of β-lactamase necessary for covalent trapping to be significant is impossible to reach.