Identification of potential superoxide dismutase allosteric modulators by structure-based computational approaches: homology modelling, molecular dynamics and pharmacophore-based virtual screening.

The visceral form of Leishmaniasis, also known as kala-azar, caused by is the main etiological agent of this form in Brazil responsible for 30,000 annual deaths. Despite its epidemiological impact, treatment of the disease is limited by resistance, species-dependent efficacy and serious adverse effects. The application of computational tools to prioritize potential bioactive molecules based on 3D structural of biological target is a viable alternative. Among the validated targets, Fe + 2 superoxide dismutase B2 (LcFeSODB2) is the first parasite enzyme against oxidative stress and it is involved in essential metabolic processes for its survival. Due to substrate binding-site volume (superoxide ion) and consequent difficulty in its active site modulation for small molecules, the search for allosteric sites at LcFeSODB2 3D structure is a promising strategy. As there are no 3D structures of LcFeSODB2, comparative modeling was applied to build 3D models by SWISS-MODEL and MODELLER version 9.19. Next, the best 3D model was used in molecular dynamics (MD) routines with multiple probes on GROMACS version 5.1.2. In addition, potential allosteric sites predicted by FTMap and Metapocket web servers were used with probe occupancy maps from MD to select an allosteric binding site and propose a pharmacophore model. Next, it was used as a template in virtual screening by UNITY module available on SYBYL-X version 2.1.1 at Sigma-Aldrich CPR™ subset of ZINC12 database. The pharmacophore-based virtual screening resulted in the selection of two potential allosteric LcFeSOD compounds with partial pharmacophoric requirements, drug-like properties and commercial availability for enzymatic assays. Communicated by Ramaswamy H. Sarma.