Synthesis and chemoinformatics analysis of N-aryl-β-alanine derivatives

Abstract

Carbohydrazides of N-substituted β-amino acids exhibit a variety of different biological activities including antibacterial, antiviral, fungicidal, antihelminthic, anticancer, antiinflammatory, etc. New potentially biologically active N-(4-iodophenyl)-β-alanine derivatives, N-(4-iodophenyl)-N-carboxyethyl-β-alanine derivatives, and their cyclization products were designed and synthesized. To determine the most propitious directions for further investigation of the obtained compounds, we tried to appraise their biological activity in silico using the ChemSpider and chemical structure lookup service (CSLS), chemical similarity assessment (Integrity and SuperPred), and machine learning methods [prediction of activity spectra for substances (PASS)]. No useful hints on potential biological activity of the obtained novel compounds were delivered by ChemSpider, CSLS, Integrity or SuperPred. In contrast, PASS predicted some biological activities that could be verified experimentally. Neither antibacterial nor antifungal activity was predicted for the compounds under study despite these actions being known for compounds from this chemical class. Evaluation of antibacterial (Escherichia coli B-906, Staphylococcus aureus 209-P, and Mycobacterium luteum B-91) and antifungal (Candida tenuis VKM Y-70 and Aspergillus niger F-1119) activities in vitro did not reveal any significant antimicrobial action, which corresponds to the computational prediction. Advantages and shortcomings of chemical similarity and machine learning techniques in computational assessment of biological activities are discussed. Based on the obtained results, we conclude that academic organic chemistry studies could provide a significant impact on drug discovery due to the novelty and diversity of the designed and synthesized compounds; however, practical utilization of this potential is narrowed by the limited facilities for assaying biological activities.