A Computational Drug Design approach for optimization of PHGDH Inhibitors in Cancer metabolism

Introduction Phosphoglycerate dehydrogenase (PHGDH) , an enzyme, diverts glycolytic intermediates toward serine production and is upregulated in various cancers leading to therapy- resistant disease, underscoring the need for improved PHGDH targeted treatments. Methods Reported PHGDH inhibitors were studied by using ligand-based pharmacophore modelling and QSAR analysis to assess correlations among molecular descriptors and the inhibition necessary to inhibit PHGDH. Ten pharmacophore models were then produced and further validated. The ZINC database was screened using the best-performing model to select new hit compounds. Through the studies of molecular docking with PHGDH target (7VA1), the promising hits were further refined. A 500 ns molecular dynamics simulation was used to determine the stability of the ligand-protein interactions. Moreover, the electronic properties and reactivity of the compounds were also studied by using density functional theory (DFT). Lastly, ADMET analysis was conducted to determine the drug-likeness and drug-pharmacokinetic suitability of the discovered compounds. Results Combined virtual screening, molecular docking, molecular dynamics simulations and MM-GBSA analysis have determined a number of potential PHGDH inhibitors. A large database was screened using the optimal pharmacophore model (Hypothesis 2), which had ROC value 0.893 resulting in compound 39 and ZINC84927964 being identified as promising hits with acceptable ADMET properties. Discussion The current integrated in silico approach identified compound 39 and ZINC84927964 as potentially useful PHGDH inhibitors with good drug-like and safety properties. These results demonstrate that they can be optimized and experimentally proven as therapeutic agents in PHGDH-targeted cancer therapy.