Using in silico tools for primary screening and prediction of pharmacological activity of the range new theophyline derivatives

Introduction. Pharmacotherapy of various diseases requires the use of new highly effective drugs. At the initial stages of drug development, it is advisable to use in silico methods for the their screening. This saves time and money in selecting the most promising compounds for synthesis and in vitro and in vivo pharmacological studies. It is known that 7,8-disubstituted theophylline derivatives have a high potential for biological action. We chose 7-R-8-hydrazine derivatives of theophylline for the search of drug-like substances, since it is known that functional hydrazine derivatives are widely used in medical practice as drugs for the pharmacotherapy of depression, infectious diseases, hypertension, diabetes, etc. The aim of the work was primary screening and evalution of potential pharmacological activity of model molecules using virtual computer programs. Materials and methods. The research materials are derivatives of 4-(2-(1,3-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)hydrazineylidene)heptanedioic acid with linear and branched alkyl substituents at the 7th position of the basic heterocycle (12 compounds): methyl-, ethyl-, n-propyl-, n-butyl, i-butyl, n-amyl, i-amyl, n-hexyl, n-heptyl, n-octyl, n-nonyl, n-decyl. Research methods – freely available web tools SwissADME and SuperPred 3.0 for screening pharmacologically inactive molecules and selecting potential drug candidates. Results and discussion. The virtual design of compounds (7,8-disubstituted theophyllines) was substantiated and implemented, and their screening was carried out using selected in silico methods. Using SwissADME, the physicochemical and pharmacokinetic parameters of the molecules were determined, drug-likeness, potential for oral administration and prospects for medicinal chemistry were established. SuperPred 3.0 webserver made it possible to establish the possible belonging of the studied compounds to certain ATC-classes in accordance with the WHO classification of drugs and target prediction of compounds. The biotargets with the highest probability of binding (˃ 90 %) and model prediction accuracy (˃ 80 %) were taken into account. The common predicted targets for them are Cathepsin D і Nuclear factor NF-kappa-B p105 subunit. The probability of the binding of the compounds 1–12 with the first target is in the range of 95,74–98,38 %, and with the second – 96,12–96,72 %. The accuracy of target prediction models ranges from 96,09–98,95 %. The obtained results were compared with the results of the prediction for theophylline and basic structure – 4-(2-(1,3-dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)hydrazineylidene)heptanedioic acid. The results of in silico studies indicate the drug-like nature of molecules for which a fairly easy synthesis is predicted. Conclusions. A virtual screening of the studied compounds was carried out and some relationships «structure – biological activity» were established. Further chemical modification of the the structures should not involve the elongation of the alkyl substituent at position 7 of the basic heterocycle, because this does not significantly change their pharmacological potential, but negatively affects the ADME profile and limits oral biovailability. The results of in silico prediction confirmed the prospects of further research for biologically active compounds among 7,8-disubstituted theophyllines.