Abstract
From lead 1, (N-(4-((4-(3-(4-(3-methoxyphenyl)-1H-1,2,3-triazol-1-yl)propyl)piperazin-1-yl)sulfonyl)-phenyl)acetamide), a S100A2-p53 protein-protein interaction inhibitor based on an in silico modelling driven hypothesis, four focused libraries were designed and synthesised. Growth inhibition screening was performed against 16 human cancer cell lines including the pancreatic cell lines MiaPaCa2, BxPC3, AsPC-1, Capan-2, HPAC, PANC-1 and the drug resistant CFPAC1. Modification of 1's phenylacetamide moiety, gave Library 1 with only modest pancreatic cancer activity. Modification of the 3-OCH3Ph moiety (Library 2) gave 4-CH3 (26), 4-CH2CH3 (27), 4-CF3 (31) and 4-NO2 (32) with sterically bulky groups more active. A 4-CF3 acetamide replacement enhanced cytotoxicity (Library 3). The 4-C(CH3)336 resulted in a predicted steric clash in the S100A2-p53 binding groove, with a potency decrease. Alkyl moieties afforded more potent analogues, 34 (4-CH3) and 35 (CH2CH3), a trend evident against pancreatic cancer: GI50 3.7 (35; BxPC-3) to 18 (40; AsPC-1) μM. Library 4 analogues with a 2-CF3 and 3-CF3 benzenesulfonamide moiety were less active than the corresponding Library 3 analogues. Two additional analogues were designed: 51 (4-CF3; 4-OCH3) and 52 (4-CF3; 2-OCH3) revealed 52 to be 10-20 fold more active than 51, against the pancreatic cancer cell lines examined with sub-micromolar GI50 values 0.43 (HPAC) to 0.61 μM (PANC-1). MOE calculated binding scores for each pose are also consistent with the observed biological activity with 52. The obtained SAR data is consistent with the proposed interaction within the S100A2-p53 bonding groove.