Purpose: An essential step in the drug development process, as indicated by the FDA, is to evaluate the various stages of absorption, distribution, metabolism, and elimination (ADME) for a particular drug. Transport proteins, specifically organic anion transporters OAT1, OAT2, OAT3, and OAT4, are important players involved in the absorption, distribution and elimination stages of numerous drugs, so much so that OAT1&3 have been specifically included in an FDA guidance to industry regarding pre-clinical investigational studies. Human OATs (hOATs) are expressed in a wide variety of tissues including the renal proximal tubule cells (hOAT1-4), hepatocytes (hOAT2), and even the placenta (hOAT4). While hOAT1&3 have been shown to transport many therapeutic agents including antivirals, antibiotics, NSAIDs, antineoplastics, and antihypertensives, little to no information is known about how these various substrates interact with OATs. Unfortunately, an experimental structure has not yet been determined for either hOAT1 or hOAT3, however, the recently solved crystal structure of the Piriformospora indica phosphate transporter (PiPT) was effectively utilized as a template to generate in silico 3-D homology models for hOAT1&3. Subsequently, the prototypical substrates para-amino hippuric acid (PAH) and estrone-3-sulfate (ES), as well as the prototypical inhibitor probenecid, were separately docked into the respective hOAT1&3 models to predict binding modes within the substrate binding pockets and identify amino acid residues involved in the substrate-transporter or inhibitor-transporter binding interactions. Residues deemed most important within each respective binding pocket were subjected to mutagenesis and are being evaluated for altered affinity of the transporter for PAH or ES. Determining the amino acids deemed critical to OAT-substrate interaction should prove beneficial to future investigations of these vitally important transporter-mediated processes, aiding in silico assessment of drug absorption and distribution, clinical drug-drug interactions, as well as the development of better patient information resources, thereby better managing possible adverse drug events. It’s predicted that substitution of hOAT1&3 residues involved in transporter-substrate interactions, identified through in silico molecular modeling and substrate docking studies will result in altered affinity (Km) of hOAT1&3 for PAH or ES, respectively.
Methods: hOAT1&3 transmembrane domain topology was predicted using the PredictProtein server (https://www.predictprotein.org/). Sequence alignments were performed using ClustalX 2.1. The resulting alignments were combined with the PiPT crystal structure (PDB ID = 4J05) using MODELLER 9.4 to generate homology models. The models were evaluated with the Discrete Optimized Protein Energy (DOPE) function and their stereoelectronic quality was assessed using PROCHECK. Automated docking of energy minimized PAH and ES was subsequently performed using GOLD 5.4. Docked solutions were visualized with SYBYL and PyMOL, and amino acids hypothesized to be most important for the recognition of PAH and ES in the OAT paralogs were identified. Conservative and non-conservative amino acid substitutions were introduced into the hOAT1&3 coding sequences via site-directed mutagenesis (QuikChange Lightning Kit). Next, mutants will be transfected into CHO cell lines, followed by functional screening assays and saturation analysis to determine affinity (Km).
Results: Many amino acid residues were found to interact with PAH (hOAT1) and ES (hOAT3). Arg15, Ile19, Tyr230, Asn439, Arg466 (hOAT1) and Tyr342, Phe426, Phe430, Leu431, Arg454 (hOAT3) were sites predicted to be most important based on the estimated strength of their interactions with PAH and ES, respectively. Conservative and non-conservative substitutions have been introduced via mutagenesis to hOAT1 and confirmed by sequencing.
It was also of interest to dock PAH, ES, and the prototypical inhibitor probenecid separately to hOAT1 and hOAT3 for potential commonality of critical amino acids to substrate/inhibitor-transporter interactions. Results showed that, for hOAT1, Tyr230 and Arg466 were common contacts for all three compounds and, for hOAT3, Tyr342 was found to significantly interact with all three compounds. Finally, several amino acid positions were conserved between the hOAT1&3 paralogs for each compound. After proper alignment of the hOAT1&3 sequences, a conserved tyrosine was identified for PAH, a phenylalanine and an arginine residue for ES, and two separate tyrosines for probenecid.
Conclusion: Amino acid sequences of hOAT1&3 were successfully aligned to a known crystal structure (PiPT) also within the MFS superfamily. The models produced were validated through the use of various scoring functions and the top models identified. Many amino acids were identified as important within each transporter’s substrate-protein binding pocket. Since OAT substrates encompass a wide variety of structures, additional docking studies were done to observe possible commonality of amino acids interacting with PAH, ES, and probenecid, and common amino acids across these compounds were found within hOAT1&3. Currently, conservative and non-conservative amino acid substitutions are being inserted at identified important positions using site-directed mutagenesis. Subsequently, mutant transporters will be subjected to saturation analysis to quantify any changes in substrate affinity (Km).