Role of P‐glycoprotein‐mediated secretion in absorptive drug permeabiity: An approach using passive membrane permeability and affinity to P‐glycoprotein††Dedicated to Prof. B. C. Lippold on the occasion of his 60th birthday.

It has been shown in vivo and in vitro that P-glycoprotein (P-gp) may be able to influence the permeability of its substrates across biological membranes. However, the quantitative contribution of the secretion process mediated by P-gp on the overall permeability of membranes has not been determined yet. In particular, observations need to be clarified in which substrates showing high affinity to P-glycoprotein, e.g., verapamil, apparently do not seem to be greatly influenced by P-gp in their permeability and consequently also with respect to their extent of GI-absorption after oral administration, whereas weaker substrates of P-gp, e.g., talinolol, have clearly shown P-gp-related absorption phenomena such as nonlinear intestinal permeability and bioavailability. Experiments with Caco-2 cell monolayers and mathematical simulations based on a mechanistic permeation model should aid in clarifying the underlying mechanism for these observations and quantifying the influence of passive membrane permeability and affinity to P-gp to the overall transmembrane drug flux. In addition, the concentration range of drug at which P-glycoprotein-mediated transport across the biological membrane is relevant should be examined. The permeability of various drugs in Caco-2 monolayers was determined experimentally and modeled using a combination of passive absorptive membrane permeability and a Michaelis-Menten-type transport process in the secretory direction. The passive permeabilities were experimentally obtained for the apical and basolateral membrane by efflux experiments using Caco-2 monolayers in the presence of a P-gp inhibitor. The Michaelis-Menten parameters were determined by a newly developed radioligand-binding assay for the quantification of drug affinity to P-gp. The model was able to accurately simulate the permeability of P-glycoprotein substrates, with differing passive membrane permeabilities and P-glycoprotein affinities. Using the outlined approach, permeability vs donor-concentration profiles were calculated, and the relative contribution of passive and active transport processes to the overall membrane permeability was evaluated. A model is presented to quantitatively describe and predict direction-dependent drug fluxes in Caco-2 monolayers by knowing the affinity of a compound to the exsorptive transporter P-gp and its passive membrane permeability. It was shown that a combination of high P-gp affinity with good passive membrane permeability, e.g., in the case of verapamil, will readily compensate for the P-gp-mediated reduction of intestinal permeability, resulting in a narrow range in which the permeability depends on the apical drug concentration. On the other hand, the permeability of compounds with low passive membrane permeability (e. g., talinolol) might be affected over a wide concentration range despite low affinity to P-gp.