In silico analysis of potential Tankyrase I inhibitors for ovarian cancer

Abstract

In this study, we aimed to identify repurposed small-molecule inhibitors of Tankyrase‑1 by combining molecular docking with pharmacokinetic and drug‑likeness profiling. We screened four thousand assorted subjects including the semi-synthetic isoquinoline alkaloid Hydrastinine, the orally active P‑selectin antagonist PSI‑697, the irreversible monoamine oxidase inhibitor Benmoxin and the investigational anxiolytic α7-nicotinic receptor modulator BNC‑210. For docking, we predicted that Benmoxin is multi-point anchored within the Tankyrase-1 catalytic pocket by its interactions with residues Pro1235 (double-headed arrow in (Figures 3B and 3D), Pro1187 and Tyr1226, whilst PSI‑697 appeared to only interact with Pro1235 and Tyr1226. Hydrastinine and BNC‑210, by contrast, were predicted to bind Pro1235 and Pro1187 whilst showing a loss of the Tyr1226 interaction. Benmoxin (multi-residue binding) and PSI-697 (contacting Tyr1226) showed the best relative docking stabilities. All four of the ligands were predicted to pass Lipinski’s rule of five and have high gastrointestinal absorption according to SwissADME. For Benmoxin and BNC-210 LogP values were moderate (≈2.6), high for PSI-697 (≈4.3) and low for Hydrastinine (≈1.2). PSI-697, Benmoxin and BNC-210 were predicted to cross the blood–brain barrier (BBB), while Hydrastinine was not. Hydrastinine and BNC-210 were predicted as P-glycoprotein efflux substrates (risk of efflux), whereas PSI-697 or Benmoxin were not. Hydrastinine was found to have low predicted cytochrome P450 inhibition, Benmoxin to inhibit only CYP2D6 and PSI-697 and BNC-210 predicted broad (multiple isoforms) inhibition. Hydrastinine remained within the intestinal-absorption compartment in the BOILED-Egg model, while PSI-697, Benmoxin and BNC-210 permeated across the blood brain barrier.