Comprehensive structure-activity relationship (SAR) investigation of -aryl glycoside derivatives for the development of SGLT1/SGLT2 dual inhibitors

• Published in: New journal of chemistry Vol. 45; no. 31; pp. 14193 - 1421
• Main Authors: Xin, Yunting, Zhou, Shuhao, Wang, Huibin, Hu, Baichun, Zhang, Zhigang, Wang, Jian, Sun, Tiemin
• Format: Journal Article
• Published: 09.08.2021
• ISSN: 1144-0546; 1369-9261
• DOI: 10.1039/d1nj02510d

Human sodium dependent glucose co-transporters (SGLTs) are attractive targets for diabetes treatment, and the development of SGLT1/SGLT2 dual inhibitors has received substantial attention in recent years due to their better glycemic control and fewer side effects. In this study, multiple computational strategies were employed to reveal the three-dimensional qualitative structure-activity relationship (3D-QSAR) of SGLT1/SGLT2 dual inhibitors. The homology models of SGLT1 and SGLT2 were constructed, which were refined by molecular dynamics (MD) simulation and validated by Ramachandran plots, ERRAT and the ProSA program. Molecular docking was performed to generate the pharmacophoric conformers and investigate the interactions between ligands and SGLT proteins. MD simulations combined with MM-GBSA calculation were used to evaluate the binding stability, and alanine scanning mutagenesis (ASM) was used to explore the role of specific amino acids in protein-ligand interactions at the same time. The molecular field analysis (MFA) was carried out on 35 SGLT1 inhibitors and 46 SGLT2 inhibitors through a genetic partial least-squares (G/PLS) method to generate statistically significant 3D-QSAR models ( R 2 SGLT1 = 0.999 and R 2 SGLT2 = 0.963), with cross validated coefficients ( Q 2 CV = 0.742 and 0.753). Furthermore, new compounds were designed according to the summarized SAR information, which were predicted to be more potent with satisfactory pharmacokinetic properties and expected binding modes. In brief, our results will greatly support the development of SGLT1/SGLT2 dual inhibitors. Multi-combined computational approaches were used to explore the SAR and design novel potential SGLT1/SGLT2 dual inhibitors.