Toward a pharmacophore for anti-tumor based tubulin inhibitors: insight from a catalyst hypogen mapping of cryptophycin analogues

ABSTRACT

The invention of a new class of anti-tumour tubulin inhibitors based on a pharmacophore hypothesis for small molecule analogues of cryptophycin is discussed. A training set of 20 compounds was selected from 45 crypiophycin inhibitors [1C 50 value ranges from 0.022 nm to 1860 nm] of Periyar Bio in-house data bases to generate hypogen model. All structures were built and minimized within the CATALYST and conformational analysis was implemented using the pooling algorithm. The maximum number of conformers generated was 250 and a range of 10 kcal/mol was chosen. Training set consists of 20 compounds tested against human tubulin was used to develop pharmacophore hypothesis. Among the generated 10 hypotheses, the best pharmacophore model [Hypothesis 1] feature one hydrogen bond acceptor, one hydrogen bond donor, one hydrophobic aliphatic and one ring aromatic with the correlation coefficient of 0.948, RMS deviation of 0.774, and a cost difference of 57.71. The obtained pharmacophore models were validated on 18 test molecules. The mapping of hypothesis 1 models on to a highly active compound 6S [1C 50 = 0.58 nm] and the mapping of hypothesis 1 model on to a highly inactive compound 7a [1C 50 = 2000 nm], was successfully designed using CATALYST Hypogen. For the test set, the accuracy in predicting active compounds was greater than 95%, while 8% and 1% representing both false positive and negative, respectively