In Silico Discovery and Pharmacological Analysis of Potential Ligands Targeting the Serotonin Transporter

Abstract

The serotonin transporter (SERT) protein is responsible for terminating serotonergic signaling by actively removing serotonin from the synapse through a sodium‐ and chloride‐dependent uptake mechanism into the neuron. The SERT is the primary target of the widely prescribed antidepressant drugs known as the Selective Serotonin Reuptake Inhibitors (SSRIs), which include Citalopram, Fluoxetine, and Paroxetine. Consequently, interest in SERT as a target for drug discovery is substantial due to the limited efficacy and side effects of current antidepressants. Computational approaches have previously been used with homology models of SERT with success, and the recent publication of the human SERT crystal structures has provided an opportunity to revisit virtual screening. For this study, our approach was to modify the SERT structure (PDB id. 5I71), which had 13 mutations introduced to crystallize it, back to its wildtype structure by replacing the 13 mutations with the wildtype amino acids using the Molecular Operating Environment (MOE) software. Also, the backbone of the protein was relaxed using molecular dynamics with ACEMD software. The S1 binding pocket was identified using the Site Finder module in MOE, and the Maybridge Hit Discoverer library, which contained 45,000 compounds, was virtually screened through the Dock module that incorporated a pharmacophore filter on aspartic acid 98. The screen scored and ranked 33,000 compounds of which 450 were selected as the top hits based on predicted affinity (S‐score under −17). The 450 compounds were then narrowed down to 44 by filtering the dataset with the Lipinski rule of 5, and predicted affinity (S‐score < −18). The 44 compounds were subjected to fingerprinting and chemical similarity analysis using R, Python, and Pearl scripts so the compounds could be placed into various groups. Of the 44 compounds, 10 were chosen for further pharmacological analysis based on the available chemical space. These compounds represent the most unique structures found within the dataset as measured by their Tanimoto coefficient in relationship to each compound utilizing scripts in the MayaChemTools software collection. The identified compounds had characteristics ranging from 6 to 10 hydrogen bond acceptors, 1 to 4 hydrogen bond donors, and a −6 to 4.87 cLogP. A stably transfected SERT cell line was used to pharmacologically evaluate the compounds using in vitro assays. Radioligand competition assays were used to determine experimental binding affinity to the SERT. Efficacy as an inhibitor was determined using an uptake inhibition assay using radioactive serotonin.In conclusion, computational methods were utilized to identify potential ligands targeted at the serotonin transporter. From this virtual screen, 10 compounds were selected based on their chemical uniqueness to be evaluated pharmacologically.This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.