Background: Tankyrases, members of PARP protein superfamily, are involved in many
cellular processes and play key roles in Wnt signaling and in several diseases including cancers.
During the past few years, there has been an increased interest in the development of selective smallmolecular
Objective: The objective was to construct a novel selective pharmacophore model for tankyrase
Methods: In this paper, the novel selective pharmacophore model was constructed using a
combination of computational methods, including molecular docking, protein-ligand interaction
fingerprint (PLIF) similarity investigation, 3D-QSAR pharmacophore merging, binding site analysis
like protein sequence alignment, SiteMap and FTMap.
Results: Within these analyses, a novel selective pharmacophore model was constructed. The two HY
features were located in the induced pocket and the sub-pocket, respectively. The hydrogen bond
acceptor (HBA) feature corresponding to Tyr1213 was located in the sub-pocket. To assess the reliability
and validity of the pharmacophore, the model was then applied to screen two validation databases for
highly selective tankyrase inhibitors. The high values of enrichment factor (32.61; 34.52) and receiver
operating characteristic (ROC) score (0.853; 0.805) indicated the model performed fairly well
at distinguishing the selective tankyrase inhibitors from putative inactive compounds. Furthermore,
12 highly selective inhibitors (10000-fold) were all ranked in the forefront of screening results.
Conclusion: The pharmacophore model reflects the most important binding requirements for the
selective tankyrase inhibitors, and the results can provide new insights for design and discovery of
novel tankyrase inhibitors with high selectivity.