Analysis of j8-lactam hydrolysis by intact cells of several strains of enteric bacteria suggests that a significant fraction of P-lactamase molecules is present at a location directly accessible for the substrates in the medium.We propose a method to correct for the error introduced by these "surface-exposed" enzymes in the Zimmermann-Rosselet assay of outer membrane permeability.In gram-negative bacteria, the outer membrane barrier very often slows the influx of antibiotic molecules and constitutes one of the major factors that determine the degree of antibiotic resistance of the organism (4). Thus, the precise determination of outer membrane permeability is important in the quantitative analysis of the efficacy of antibiotics, especially those that are inactivated within bacterial cells, for example 1-lactams (5). Among the several methods available for the measurement of outer membrane permeability, the one proposed by Zimmermann and Rosselet (8) appears to be most rigorously correct. In this method, hydrolysis of 1-lactam antibiotics by intact cells of gramnegative bacteria is analyzed as the consequence of two processes: the first, diffusion across the outer membrane following Fick's first law of diffusion, and the second, hydrolysis by periplasmic P-lactamase molecules following 8).In the actual application of this method, some interference is caused by ,-lactamase molecules that have leaked into the medium. This is corrected by measuring P-lactam hydrolysis catalyzed by the centrifugal supernatant of the cell suspension (6). However, if there are cells with damaged, "leaky" outer membranes but that still contain periplasmic ,-lactamase molecules or if there are enzyme molecules that are exposed on the cell surface but remain associated with the cell, the contribution of these enzymes will be difficult to correct for, as was pointed out earlier (3). Indeed, Hewinson et al. (1) reported that the Zimmermann assay of outer membrane permeability of a Pseudomonas aeruginosa strain produced different permeability coefficients depending on the concentration of the substrate, cephalosporin C, used, and it was pointed out by R. E. W. Hancock that these results may be due to the presence of surface-associated 3-lactamase (cited in reference 2). The potential problems arising from the presence of cell-surface-associated enzyme have not, however, been quantitatively studied. We propose here a method for estimating this contribution and show that it may introduce significant errors in the determination of outer membrane permeability.If cell-surface-exposed enzyme hydrolyzes the P-lactam * Corresponding author.antibiotic with a Vmax value of V., then the total rate of hydrolysis of the drug by a suspension of intact cells, vcells, is as follows: Vcells = Vs X Col(Km+ CO) + P x A x (Co-Cp) (1) where CO, Cp, P, and A represent the external concentration of the drug, its periplasmic concentration, the coefficient of outer membrane permeability to that drug, and the area of the outer membrane per unit of cell mass, respecti...