For P‐glycoprotein mediated transport through a confluent cell monolayer, the IC50 of a P‐gp substrate depends inversely upon the passive permeability coefficient of the probe substrate

The inhibition of transport of one drug by another is quantified by the IC50. This analysis was developed for water-soluble enzymes which bind drug, and/or inhibitor, directly from the aqueous phase to which the drug is added initially. However, some transporters, e.g. P-gp and MRP, which pick up drugs from the inner monolayer of the cell plasma membrane. This means that their binding site is at least one passive permeation barrier away from where the drug can bind to the efflux transporter. Hence passive permeability through the plasma membrane may affect correlation between the inhibitor’s KD and it’s IC50. This question cannot be examined by experiments alone. The current best approach is exhaustive computer simulations of drug transport in the presence of “virtual inhibitors” whose rigorous elementary kinetic rate constants lie within physiologically and pharmaceutically relevant values (Bentz et al., 2005). Using an MDCKII-hMDR1 confluent cell monolayer, we have fitted the elementary rate constants for drug binding to and efflux from P-gp for amprenavir, quinidine and loperamide. The simulations show that the fitted IC50 values will overestimate the intrinsic dissociation constant by a factor of roughly 11,000/Pa, where Pa (nm/s) is the passive permeability coefficient of the substrate through the confluent cell monolayer, obtained by inhibiting P-gp activity, e.g. with GF120918.