Integrated computational approach on sodium-glucose co-transporter 2 (SGLT2) Inhibitors for the development of novel antidiabetic agents

Nowadays, diabetes mellitus has become a global health concern. Various marketed drugs are available to treat diabetes but many of them produced undesirable side effects. From past few decades, sodium dependent glucose co-transporter 2 (SGLT2) is regarded as a hot target for type 2 diabetes mellitus due to its novel mechanism of action with less side effects. A majority of computational studies have been performed by many researchers to find out novel inhibitors to the target SGLT2. In the present study, we performed computational analysis using different drug designing tools to ascertain the best possible molecules from a dataset of 90 C-aryl glucoside analogues obtained from literature. The pharmacophore mapping study was performed to determine the key functional features responsible for the activity. The 3D-QSAR (CoMFA, CoMSIA and atom based QSAR) analysis was carried out by using 63 molecules as a training set & rest of the 27 as test sets to know the contribution of different fields & atoms for model development. The MFA with PLS method was used to generate a statistically significant model (q(2) = 0.7387, r(2) = 0.8670). The QSAR model was validated using LOO cross-validation (r(CV)(2) = 0.5305) & external validation (r(pred)(2) = 0.985). The homology model of SGLT2 was constructed using co-crystallized structure (PDB: 2XQ2_A) as templates & validated by Ramachandran plot. The docking study of potent C-aryl glucoside analogues (22, 23 and 26) showed highest SP docking scores (-7.22, -7.012 & -6.867) with the involvement of critical interactions with amino acid residues Asn75, His80, G1u99, Ser287, Tyr290 & Trp291. ADME analysis provides valuable information about drugability of newly designed compounds. Correlating the docking results with the 3D-QSAR and binding free energy calculations by MMGBSA analysis can suggest the active conformation of SGLT2 inhibitors and the nature of interactive fields important for binding and activity. (C) 2020 Elsevier B.V. All rights reserved.