First report on the structural exploration and prediction of new BPTES analogs as glutaminase inhibitors

Glutaminase is one of the important key enzymes regulating cellular metabolism, growth, and proliferation in cancer. Therefore, it is being explored as a crucial target regarding anticancer drug design and development. However, none of the potent and selective glutaminase inhibitors is available in the market though two prototype glutaminase inhibitors are reported namely DON as well as BPTES. Due to severe toxicity in clinical trials, the use of DON is restricted. However, BPTES is an allosteric glutaminase inhibitor with less toxic profile and, therefore, lead optimization of BPTES may be a good option to develop newer drug candidates. In this study, a multi-QSAR modeling is carried out on a series of BPTES analogs. A significant connection between different descriptors and the glutaminase inhibitory activities is noticed by employing multiple linear regression, artificial neural network and support vector machine techniques. The classification-based QSAR such as linear discriminant analysis and Bayesian classification modeling are also performed to search important molecular fingerprints or substructures that may help in classifying the probability of finding 'active' and 'inactive' BPTES analogs. Moreover, HQSAR and Topomer CoMFA analyses are also performed. In addition, the SAR observations are interpreted with all these validated computational models along with the structure based contours. Finally, new twenty two compounds are designed and predicted for their probable glutaminase inhibitory activity. (C) 2017 Elsevier B.V. All rights reserved.