Hepatic organic anion transporting polypeptides—OATP1B1,
OATP1B3, and OATP2B1—are expressed at the basolateral membrane
of hepatocytes, being responsible for the uptake of a wide range of
natural substrates and structurally unrelated pharmaceuticals. Impaired
function of hepatic OATPs has been linked to clinically relevant drug–drug
interactions leading to altered pharmacokinetics of administered drugs.
Therefore, understanding the commonalities and differences across
the three transporters represents useful knowledge to guide the drug
discovery process at an early stage. Unfortunately, such efforts remain
challenging because of the lack of experimentally resolved protein
structures for any member of the OATP family. In this study, we established
a rigorous computational protocol to generate and validate structural
models for hepatic OATPs. The multistep procedure is based on the
systematic exploration of available protein structures with shared
protein folding using normal-mode analysis, the calculation of multiple
template backbones from elastic network models, the utilization of
multiple template conformations to generate OATP structural models
with various degrees of conformational flexibility, and the prioritization
of models on the basis of enrichment docking. We employed the resulting
OATP models of OATP1B1, OATP1B3, and OATP2B1 to elucidate binding
modes of steroid analogs in the three transporters. Steroid conjugates
have been recognized as endogenous substrates of these transporters.
Thus, investigating this data set delivers insights into mechanisms
of substrate recognition. In silico predictions were complemented
with in vitro studies measuring the bioactivity of a compound set
on OATP expressing cell lines. Important structural determinants conferring
shared and distinct binding patterns of steroid analogs in the three
transporters have been identified. Overall, this comparative study
provides novel insights into hepatic OATP-ligand interactions and
selectivity. Furthermore, the integrative computational workflow for
structure-based modeling can be leveraged for other pharmaceutical
targets of interest.