Contribution to the Modeling of Biomolecules and Their Interactions: Inhibition of Enzymes Involved in Cancer Diseases by a New Class of Derivatives

Abstract

This dissertation presents the results of two research studies aimed at improving our understanding of breast cancer treatment drugs and potential targets. The first study used PLS regression to create QSAR models for 54 analogs of 2-phenyl-1H-indole, known for their anti-proliferative activity on MDA MB231 and MCF-7 cancer cell lines. The dataset was split into training and testing datasets 10,000 times, with 75% of the molecules used for training and the rest for external validation. The best models were selected based on the highest probability of occurrence according to the Bayesian information criterion. As a result, the PLS regression equation derived explains 6.79% and 63% of the variability in anticancer activity around its mean for model 1(MDA MB231), and model 2 (MCF-7), respectively. The leave-one-out cross-validation R2CV, the bootstrapping correlation coefficients R2boots, and the predicted R2pred indicated a high predictive power for both models. This study was accompanied by molecular docking/dynamics simulations, revealing that ligands L39, L40, and L48 fit into the pocket of estrogen-α receptor (PDB:1A52), while ligand L47 showed affinity with progesterone receptor (PDB:1A28). This affinity was confirmed by high negative score values and the establishment of several non-covalent interactions with the active site residues of both receptors. Furthermore, drug-likeness and ADME prediction analyses showed favorable absorption and oral bioavailability characteristics for ligands L39 and L48, suggesting their potential as precursor compounds for breast cancer drug development. The second study aims to identify the binding mechanism of Glutaminase C (GAC) as a potential target for triple-negative breast cancer (TNBC). Molecular docking was employed to explore the interaction of 26 Withangulatin A (WA) derivatives with the allosteric site of GAC. The molecular docking/dynamics simulation results revealed that compounds A5, A8, A13, and A18 show high affinity toward the allosteric pocket of the GAC (PDB:3UO9), as confirmed by the high negative score values. These compounds interact with the most important residues and suggest a similar binding mechanism to the native compound (BPTES) and the clinical trial drug (CB-839). The combination of MEP analysis and molecular docking/dynamics studies confirms the favorable reactive sites of these compounds. Finally, pharmacokinetics prediction showed that A8 and A13 present the best ADMET profile among the selected compounds.