Identification of novel inhibitors of rat Mrp3

Abstract

Multidrug resistance-associated protein (MRP; ABCC gene family) mediated efflux transport plays an important role in the systemic and tissue exposure profiles of many drugs and their metabolites, and also of endogenous compounds like bile acids and bilirubin conjugates. However, potent and isoform-selective inhibitors of the MRP subfamily are currently lacking. Therefore, the purpose of the present work was to identify novel rat Mrp3 inhibitors. Using 5(6)-carboxy-2',7'-dichlorofluorescein diacetate (CDFDA) as a model-(pro)substrate for Mrp3 in an oil-spin assay with primary rat hepatocytes, the extent of inhibition of CDF efflux was determined for 1584 compounds, yielding 59 hits (excluding the reference inhibitor) that were identified as new Mrp3 inhibitors. A naive Bayesian prediction model was constructed in Pipeline Pilot to elucidate physicochemical and structural features of compounds causing Mrp3 inhibition. The final Bayesian model generated common physicochemical properties of Mrp3 inhibitors. For instance, more than half of the hits contain a phenolic structure. The identified compounds have an AlogP between 2 and 4.5, between 5 to 8 hydrogen bond acceptor atoms, a molecular weight between 260 and 400, and 2 or more aromatic rings. Compared to the depleted dataset (i.e. 90% remaining compounds), the Mrp3 hit rate in the enriched set was 7.5-fold higher (i.e. 17.2% versus 2.3%). Several hits from this first screening approach were confirmed in an additional study using Mrp3 transfected inside-out membrane vesicles. In conclusion, several new and potent inhibitors of Mrp3 mediated efflux were identified in an optimized in vitro rat hepatocyte assay and confirmed using Mrp3 transfected inside-out membrane vesicles. A final naive Bayesian model was developed in an iterative way to reveal common physicochemical and structural features for Mrp3 inhibitors. The final Bayesian model will enable in silico screening of larger libraries and in vitro identification of more potent Mrp3 inhibitors.

Keywords: 5(6)-Carboxy-2′,7′-dichlorofluorescein diacetate; Computational modeling; Efflux transporter; Inhibitor screening; Inside-out membrane vesicles; Mrp3; Naive Bayesian model; Rat hepatocytes.