3D-QSAR CoMFA studies on trypsin-like serine protease inhibitors: a comparative selectivity analysis

A series of indole/benzoimidazole-5-carboxamidines have been reported to inhibit various trypsin-like serine proteases viz. uPA, tPA, factor Xa, thrombin, plasmin, and trypsin, which are involved in various types of pathophysiological conditions such as cancer progression, thrombosis etc. Inhibition of these protease enzymes may serve as therapeutic agents in various types of cancer as well serve as anticoagulant or antithrombotic agents. The dual inhibitory action may result in poor clinical candidates. 3D-QSAR models were generated for indole/benzoimidazole-5-carboxamidines using the CoMFA technique to study their selectivity trends toward various trypsin-like serine proteases. Molecular superimposition was carried out on the template structure using atom-based RMS fit method. The CoMFA models were established from the training set of 25-29 molecules and validated by predicting the activities of seven-eight test set molecules. The CoMFA models generated using steric and electrostatic fields for tPA, fXa, thrombin, plasmin, and trypsin inhibition exhibited better statistical significance than the CoMFA models generated using ClogP as an additional descriptor. Thus, the validated CoMFA models with steric and electrostatic fields were used to generate 3D contour maps, which may provide possible modification of molecules for better selectivity/activity. The present 3D-QSAR studies emphasize the selectivity trends of indole/benzoimidazole-5-carboxamidines, which may be obliging in designing novel selective serine protease inhibitors of therapeutic interest. (c) 2005 Elsevier Ltd. All rights reserved.