Substrate-Dependent Inhibition of the Human Organic Cation Transporter OCT2: A Comparison of Metformin with Experimental Substrates

The importance of the organic cation transporter OCT2 in the renal excretion of cationic drugs raises the possibility of drug-drug interactions (DDIs) in which an inhibitor (perpetrator) drug decreases OCT2-dependent renal clearance of a victim (substrate) drug. In fact, there are clinically significant interactions for drugs that are known substrates of OCT2 such as metformin. To identify drugs as inhibitors for OCT2, individual drugs or entire drug libraries have been investigated in vitro by using experimental probe substrates such as 1-methyl-4-phenylpyridinium (MPP(+)) or 4–4-dimethylaminostyryl-N-methylpyridinium (ASP(+)). It has been questioned whether the inhibition data obtained with an experimental probe substrate such as MPP(+) or ASP(+) might be used to predict the inhibition against other, clinical relevant substrates such as metformin. Here we compared the OCT2 inhibition profile data for the substrates metformin, MPP(+) and ASP(+). We used human embryonic kidney (HEK 293) cells stably overexpressing human OCT2 as the test system to screen 125 frequently prescribed drugs as inhibitors of OCT2-mediated metformin and MPP(+) uptake. Data on inhibition of OCT2-mediated ASP(+) uptake were obtained from previous literature. A moderate correlation between the inhibition of OCT2-mediated MPP(+), ASP(+), and metformin uptake was observed (pairwise r (s) between 0.27 and 0.48, all P < 0.05). Of note, the correlation in the inhibition profile between structurally similar substrates such as MPP(+) and ASP(+) (Tanimoto similarity T = 0.28) was even lower (r (s) = 0.27) than the correlation between structurally distinct substrates, such as ASP(+) and metformin (T = 0.01; r (s) = 0.48) or MPP(+) and metformin (T = 0.01; r (s) = 0.40). We identified selective as well as universal OCT2 inhibitors, which inhibited transport by more than 50% of one substrate only or of all substrates, respectively. Our data suggest that the predictive value for drug-drug interactions using experimental substrates rather than the specific victim drug is limited.