Two novel classes of fused azaisocytosine-containing congeners as promising drug candidates: Design, synthesis as well as in vitro, ex vivo and in silico studies

Searching for new less toxic anticancer drug candidates is a big challenge from a medical point of view. The present investigation was aimed at describing two independent synthetic approaches based on isosteric replacements, spectroscopic characteristics, in vitro anticancer and ex vivo antihaemolytic activities of novel molecules (9-22) and correlations between their standardised lipophilicity indices, computed log P-average values and pharmacokinetic descriptors. Two novel protocols for annelation of the triazinone template on hydrazinylideneimidazolidines (1-8) (showing a high reactivity towards electrophilic reagents, such as ethyl trifluoropyruvate and ethyl 3-methyl-2-oxobutyrate) were developed for the first time, giving rise to two original classes of highly conjugated azaisocytosine-containing molecules (9-16 and 17-22). Both syntheses proceeded under basic conditions to yield the most probable intermediates (e.g. hemiaminals and imines), which in refluxing two-component solvent mixtures or a suitable solvent cyclised through closing the triazinone ring on functionalised imidazolidines in both cases. All fused azaisocytosine-containing congeners were investigated with the purpose of preselecting possible drug candidates with a better selectivity that could be suitable for further more detailed drug development studies. The majority of test molecules revealed strong antiproliferative effects in most tumour cell cultures and they were more cytotoxic against tumour cells than anticancer drug - pemetrexed. These cytotoxicities may be associated with the activation of initiator and executioner caspases (confirmed for compound 12) which are inducers of apoptosis. Simultaneously, three bioisosteres bearing the trifluoromethyl moiety at the C-3 and the ortho substitution at the phenyl ring (10, 12 and 13) proved to be the most promising in terms of selectivity as they were less or equally toxic to normal cells as pemetrexed. It was shown that isosteric replacement of the ethyl group in antitumour active congeners by the trifluoromethyl or isopropyl group was favourable for the selectivity of the designed drug-like molecules. Almost all new compounds revealed the protective effects in an ex vivo model of oxidatively stressed rat erythrocytes (better or comparable than that of ascorbic acid/Trolox), proving that they are safe to red blood cells. The statistically significant and predictive QSAR equations were derived that describe relationships between some pharmacokinetic descriptors (such as log K-a,K- HSA, f(u, brain), Caco-2, log K-p) and lipophilicity parameters of test molecules. Among all molecules with anticancer profile, the possible drug candidates seem to be 10, 12, 13, 19 and 21 which are the least toxic for normal cells, deprived of haemolytic effects on oxidatively-stressed red blood cells and have the optimum pharmacokinetic descriptors in terms of their lipophilicity parameters. Because of a high development potential they should be utilised in further more extended in vivo investigations aimed at developing novel less toxic anticancer agents.