Novel In-Vitro Approaches to Investigate Membrane Partitioning and pH-Dependent Passive Permeation of Small Molecule Drugs

Abstract

Passive permeability of drug molecules through biological membranes is a fundamentally important process that involves the partitioning of molecules into the lipid bilayer membrane and passive permeation across the membrane. The majority of the drug molecules are either weak bases or acids. For these drugs, along with lipophilicity and other physicochemical properties, the acid/base nature and ionization constant (pKa) are very important for predicting their passive permeation across biological barriers. Depending on the pKa of ionizable groups, drug molecules may coexist in various biological fluids both their charged and uncharged forms to varying extents. According to the widely known pH-partition hypothesis, only the uncharged form of ionizable molecules contributes to the passive permeability. Therefore, permeability and absorption rate should be proportional to the molar fraction of the uncharged form in the bulk medium. However, there is ample evidence of deviations from this linear relationship manifesting in nonlinear passive permeability vs fraction unionized plots. It is possible that the assumption of a constant pKa for ionizable drugs contributes to many of these cases of apparent deviations which cannot be explained by other factors such as paracellular or active transport. Studies have provided evidence for pKa shifts of ionizable drugs when partitioned into different bilayer lipid membrane systems.It is thought that electrostatic interaction between the membrane partitioned charged solutes and phospholipid charged headgroups contributes to the pKa shift (apparent pKa) which can change the uncharged fraction in the membrane interface causing a proportional change in the passive permeation rate from the expected rate predicted with an aqueous pKa. This would appear as a violation of the pH-partition hypothesis when other contributing factors can be ruled out. Therefore, a surrogate membrane phospholipid system (DAPC/n-hexane) was employed to investigate potential pKa shifts following interaction between phospholipid charged headgroups and ionizable drugs. Several probe drugs of acidic, basic and neutral nature were investigated with this biphasic surrogate system with controls at different aqueous pH levels to determine their apparent pKa values. The partitioning of compounds into membranes can be thought of as an integral part of membrane permeability. Extensive partitioning of drug molecules into the intracellular membranes which constitute the majority of all the cell membranes potentially results in the observed lag phase in monolayer permeability assays. Hepatocyte drug binding correlates strongly with microsomal drug binding suggests that binding to cells is presumably a result of non-specific drug partitioning into membranes. For this reason, cellular drug distribution equilibrium in the cells might not be achieved instantaneously, as is often assumed in permeability models and perfusion-limited PBPK models. Therefore, characterization of this early distribution phase by observing the time course of the free drug concentration in drug-spiked cell suspensions (MDCK & rat hepatocyte) was pursued with a novel in vitro microdialysis approach followed by in silico investigations with mathematical modeling. Depending on the drug, the apparent pKa values in the surrogate membrane system at different pH levels indicate significant shift from the inherent pKa values of the ionizable probe drugs. The in vitro microdialysis technique allowed the determination of the time course of free drug concentration in MDCK and rat hepatocyte cell suspensions spiked with probe drugs with high temporal resolution. Two mechanistic models developed with intracellular membrane compartment along with lysosomal or mitochondrial ion-trapping compartment resulted in excellent model fitting for the observed time-course data in MDCK cell and rat hepatocyte suspensions for tolbutamide and metoprolol, respectively.