Implications of In Vitro System Conditions for CYP2C9 and CYP3A4: Impact of pH on Clearance and Enzyme Inhibition

Abstract

Common practice in drug discovery is to use buffer systems witha pH of 7.4 for drug metabolism and drug‐drug interaction studies despite evidence that the intracellular pH of hepatocytes is approximately 7.0. Systemp H will govern the ratio of ionized (charged) and non‐ionized (non‐charged)species for ionizable drugs. As a result of the pH gradient between the extracellular plasma and the intracellular cytosol of the hepatocyte, anunbound drug gradient, dependent on the physiochemical properties of the drug exists. An ionization correction factor [Berezhkovskiy (2011)]was proposed to account for this phenomenon when scaling in vitro findings at pH 7.4 to in vivo predictions of hepatic clearance. However, this correction ignores the potential changes in enzyme function at the lower pH environment.In the current study we evaluated, in human liver microsomes, the impact of buffer pH on the unbound fraction, Km, Vmax and intrinsic clearance (CLint=Vmax/Km) of probe substrates for two CYP enzymes, CYP2C9 (diclofenac and S‐warfarin) and CYP3A4 (midazolam, dextromethorphan, testosterone for CYP3A4) over a pH range of 6.0 to 9.0. Variable changes in the enzyme kinetic factors were observed over the pH range. These changes did not coincide with a particular species, ionizedor non‐ionized, for any of the substrates. When comparing the physiologically relevant pHs (7.0 and 7.4) for CYP2C9, a consistent trend was observed for both substrates, with lower Km, higher Vmax and higher CLint observed at 7.0 compared to 7.4. For the CYP3A4 enzyme, the inverse trend for observed, with higher Km, lower Vmax (Vmax was unchanged for midazolam) and lower CLint observed at 7.0 compared to7.4 for the dextromethorphan and midazolam substrates only. When testosterone was used as the probe substrate, both Km and Vmax were lower at pH 7.0 compared to pH 7.4, resulting in only a slight decrease in CLint at pH 7.0.Changes in the inhibition parameter, Ki, for the CYP2C9 reversible inhibitors amiodarone, desethylamiodarone, clozapine and nicardipine were consistent between the two probe substrates, with pH 7.0 Ki values consistently higher (less potent) than pH 7.4. Fold changes between pH 7.0 and pH 7.4 varied between the inhibitors and ranged from 1.25 to2.46 for the probe substrate diclofenac and 1.4 to 1.65 for S‐warfarin. These results suggest that changes as a result of pH vary from enzyme, compound and inhibitor combinations. In order to improve predictive performance of hepatic clearance, we recommend that in vitro parameters be determined at pH 7.0 to provide appropriate starting inputs for in vitro to in vivo extrapolations for ionizable drugs.