In Silico Predictions for Improving Permeability Properties of Principal Anticancer Anthracyclines through Structural Modification

Abstract

A method, probably the best one, to improve anticancer drugs permeability is structural modification on the base of “drug-likeness” fundamental concepts in the way of reducing ionizability and polarity, reducing number of hydrogen bond donors or acceptors and increasing lipophilicity. The present work looks forward to modify structure of major anticancer anthracyclines in order to improve their permeability, with a final aim to increase absorption and oral bioavailability. SmiLib v2.0 was used first to build a combinatorial library, by virtual reactions of building blocks (from antimetabolite and/or antineoplastic drugs) with scaffold molecules (the tetracene-5,12-dione moiety of anthracyclines). Second, ADME-Tox web-based software tool (hosted on the server of the Ressource Parisienne en Bioinformatique Structurale) was used to mass-computation of the most relevant physicochemical properties for “drug-likeness” and oral bioavailability for all anthracycline virtual derivates from combinatorial library. Qualified derivates as potential drugs were tested for genotoxicity and acute toxicity with Tox Boxes version 2.0. Sixteen from 160 derivates created with SmiLib v2.0 were qualified as “drug-like” compounds with good oral bioavailability. In silico toxicity tests showed that all derivates have good values for oral administration. One derivate presented a smaller genotoxicity and much better values for acute toxicity at oral, subcutaneous and intraperitoneal administration than any of anthracyclines. Only at intravenous administration this derivate showed certain acute toxicity, but the aim of the study was to improve the oral bioavailability.