Identification of Selective Mtb-DHFR Inhibitors as Antitubercular Agents: A Fragment Merging Approach.

INTRODUCTION: Tuberculosis (TB) is the world's deadliest disease, with 8.2 million new cases reported in 2023. There are many druggable targets for tuberculosis, out of which Dihydrofolate reductase from M. tuberculosis (Mtb-DHFR) is a validated target essential for folate metabolism and protein synthesis.

METHODS: A unique Fragment-Based Drug Design (FBDD) approach was used to get selective Mtb-DHFR inhibitors (PDBID: 1DF7) over human DHFR. The active site was bifurcated into acidic- and basic-rich regions to generate dual docking grids. The fragment database (Enamine) was virtually screened against both acidic and basic grids, yielding 258 fragments per sites followed by constructing a focused library of 2000 molecules against each grid, which were then screened against the whole Mtb-DHFR active site. The top 104 hits were evaluated for selectivity through cross-docking against human DHFR (PDB ID: 1OHJ).

RESULTS: From Virtual screening, we have identified 258 fragment hits against per sub pocket, enabling the construction of 2000 full-length molecules. A total of 104 molecules were identified, from which 20 compounds demonstrated improved selectivity against Mtb-DHFR over hDHFR. Compound Z1 showed a high binding affinity and good ADMET property indicated favorable pharmacokinetic properties along with good stability by molecular dynamics simulations.

DISCUSSION: The dual-grid FBDD strategy worked well and achieved selective inhibition of Mtb-DHFR. Compound Z1 showed a stable binding interaction along with structural differences between Mtb-DHFR and h-DHFR, to impart excellent selectivity and off-target effects.

CONCLUSION: Upon computational workflow, Z1 was successfully identified as a selective MtbDHFR inhibitor with promising drug-like characteristics. These findings support Z1 as a strong lead scaffold for further optimization in antitubercular therapeutics.