In Silico Studies to Demonstrate Potential Drug‐Drug Interaction (DDI) By Disrupting Cellular Homeostasis of Organic Anion Transporting Proteins (OATP)

Abstract

Membrane transporters OATP1B1 and OATP1B3 mediate hepatic uptake of multiple drugs (e.g., statins, antibiotics, antidiabetics). Altered OATP levels or activities cause DDI. The 2012 FDA recommendation for DDI investigations focused on competitive inhibition of OATP using co‐incubation of victim drug (VD) and perpetrator drug (PD). This method has yielded false‐negatives and underestimated severe/fatal adverse events in patients, indicating other yet unknown causes of DDI. The DDI research community named unexpected schedule‐dependent (long pre‐incubation vs. short co‐incubation) DDI as a priority, whereas FDA recommends, in 2017, additional pre‐incubation of VD and PD.We used in silico studies to evaluate if disrupting cellular transporter homeostasis is sufficent to induce clinically significant DDI (defined by FDA as causing >20–25% deviations in VD uptake, measured as area‐under‐concentration‐time curve or AUC), and if the resulting DDI is schedule‐dependent.The figure shows a cellular homeostasis model including biosynthesis and trafficking of OATP in endocytic organelles (from membrane to early endosome (EE) to recycling endosome (RE) and back to membrane, or from EE to late endosome (LE) followed by degradation). The model assumes (a) rapid OATP recycling, (b) rapid OATP degradation after entering LE, (c) slow, zero‐order OATP biosynthesis, (d) VD enters cells through transporter, (e) PD enters or exit cells via passive diffusion, (f) PD stimulates or inhibits the first‐order rate constants of OATP transfer among endocytic organelles, and (g) negligible OATP exocytosis. We used ordinary differential equations (ODE) to depict (a) OATP homeostasis, (b) intracellular pharmacokinetics of VD and PD, and (c) pharmacodynamics of PD‐mediated perturbations (IC50 and Hill coefficient n). Model parameter values were taken from literature reports on other membrane transporters (e.g. internalization of human organic anion transporter 1), calculated, or arbitrarily set based on current knowledge of cellular and endosomal pathways. ODE were solved in MATLAB (ode45 algorithm) to compute the intracellular VD AUC as functions of (a) distribution of OATP in membrane vs. cytosol at baseline (i.e., without PD), (b) extracellular concentrations of VD and PD, (c) OATP‐mediated uptake of VD, and (d) potency of PD on inhibiting or stimulating selected endocytic transfer processes.The in silico results show disruptions of OATP homeostasis induced clinically significant DDI. The extent of DDI increased with VD+PD co‐incubation time, baseline cytosol‐to‐membrane ratio of OATP, extent of PD‐perturbations of kEE and kRE, and difference between kEE and kRE. These results provide a theoretical proof‐of‐concept that perturbing OATP homeostasis may cause clinically significant DDI in a time‐dependent manner.Support or Funding InformationSupported in part by research grants from NCI and NIBIB (R01CA163015, R01EB015253) and FDA (2091IPAAU, 2091IPAWIENTJES), and by Mosier Chair of Pharmaceutical Sciences.Figure 1