ABSTRACT
Matrix metalloproteinases (MMP) are a group of enzymes related to extracelular matrix remodeling. Some types of MMP are overexpressed by malignant tumors, mainly the MMP-2 subtype, and have been associated to cancer invasiveness and metastasis. A receptor-independent (RI) 4D-QSAR formalism was applied, herein, to a set of forty ß-N-biaryl ether sulfonamide hydroxamates, previously reported as potent MMP-2 inhibitors, in order to map 3D-pharmacophore models and predict novel antitumor agents. The best RI 4D-QSAR model was statistically significant (N=30, r(2) =0.93, q(2) =0.88, five occupancy descriptors (GCOD), LSE=0.04, LOF=0.11, outliers=0), robust and not obtained by chance. The external predictability was 75 % (test set; N=8). A different orientation (binding mode) in the MMP-2 catalytic site was suggested regarding the most hydrophobic portion (R1 ) of the compounds' structure. Compounds were predicted and their inhibitory activity against MMP-2 was calculated by using the optimum RI 4D-QSAR model. The findings have provided interesting information to drive the designing and synthesis of novel potentially MMP-2 inhibitors against melanoma invasion.
