In silico characterization of coding and non-coding SNPs of the Androgen Receptor gene

Abstract The Androgen Receptor (AR), an important ligand-dependent nuclear transcription factor, encoded by the AR gene, functions in regulating the cell proliferation, migration, invasion, and apoptosis pathways. Single nucleotide polymorphisms (SNPs) of AR have been implicated in playing a role in several diseases. This study sought to explore the molecular consequences of the disease susceptible germ-line missense and non-coding SNPs of AR gene, by using different computational methods. 586 Missense SNPs were retrieved from dbSNP, DisGeNET, ClinVar, and LitVar databases, and then analyzed using nine different functionality prediction servers such as SIFT, PolyPhen-2, PROVEAN, PhD-SNP, etc. These sieved the most deleterious SNPs, which was then followed by the analysis of conservation, surface exposure, post-translational modification, stability, secondary structure, protein-ligand docking and molecular dynamics simulation. Among the missense SNPs, 21 SNPs were identified as most deleterious by all nine filtration tools. These high-risk disease-causing mutations mostly belong to highly conserved regions (buried or exposed) and affect protein stability and post-translational modifications. Based on the phenotypic effect predicted by ModPred and HOPE, and integrating the result with modeled mutated structures, two variants S866P and P893Q were considered to have the most severe functional consequences on protein 3D structure. Furthermore, based on the structural variations in mutant structure modeling, 98 noncoding pathogenic SNPs were identified as having an effect on the transcription factors’ binding and gene expression regulation. In contrast, 39 AR-SNPs were investigated using the PolymiRTS database. It was found that they affect miRNA target sites as well as destroying or creating miRNA target seeds. Significantly, these SNPs, distributed in random coil, alpha helix, extended sheet, and beta-turn, are most likely to affect the phenotypic characteristics of AR by altering hydrophobicity, flexibility, charges, H-bonds, salt bridges, and spatial structure. This current study is a comprehensive analysis, using an integrated in-silico approach on AR gene variants. It may provide a platform to conduct large-scale experiments involving AR polymorphism.