Wei HongYu WangZhe ChangYanhui YangJing PuTao SunSargit KaurJames C. SacchettiniHunmin JungWee Lin WongLee Fah YapNgeow Yun FongIan C. PatersonHao Wang2024-11-112024-11-112015-10-1610.1038/srep15328https://dspace-cris.utar.edu.my/handle/123456789/6650<jats:title>Abstract</jats:title><jats:p>It is an urgent need to develop new drugs for <jats:italic>Mycobacterium tuberculosis</jats:italic> (Mtb) and the enzyme, dihydrofolate reductase (DHFR) is a recognised drug target. The crystal structures of methotrexate binding to <jats:italic>mt</jats:italic>- and <jats:italic>h</jats:italic>-DHFR separately indicate that the glycerol (GOL) binding site is likely to be critical for the function of <jats:italic>mt</jats:italic>-DHFR selective inhibitors. We have used <jats:italic>in silico</jats:italic> methods to screen NCI small molecule database and a group of related compounds were obtained that inhibit <jats:italic>mt</jats:italic>-DHFR activity and showed bactericidal effects against a test Mtb strain. The binding poses were then analysed and the influence of GOL binding site was studied by using molecular modelling. By comparing the chemical structures, 4 compounds that might be able to occupy the GOL binding site were identified. However, these compounds contain large hydrophobic side chains. As the GOL binding site is more hydrophilic, molecular modelling indicated that these compounds were failed to occupy the GOL site. The most potent inhibitor (compound <jats:bold>6</jats:bold>) demonstrated limited selectivity for <jats:italic>mt</jats:italic>-DHFR, but did contain a novel central core (7H-pyrrolo[3,2-f]quinazoline-1,3-diamine), which may significantly expand the chemical space of novel <jats:italic>mt</jats:italic>-DHFR inhibitors. Collectively, these observations will inform future medicinal chemistry efforts to improve the selectivity of compounds against <jats:italic>mt</jats:italic>-DHFR.</jats:p>The identification of novel Mycobacterium tuberculosis DHFR inhibitors and the investigation of their binding preferences by using molecular modellingjournal-article