EXPLORING INFLAMMATORY BIOMARKER, DRUG SCREENING, AND MOLECULAR MECHANISM FOR TREATING NEURODEGENERATIVE DISEASES

Abstract

Neurodegenerative diseases such as Parkinson’s and Huntington’s disease arise from a web of pathological events, including protein aggregation, mitochondrial impairment, chronic oxidative stress, and sustained inflammatory activation within the nervous system. Although these mechanisms are increasingly well documented, there are still few therapies that meaningfully alter the disease course. In this thesis, a layered computational workflow was used that combined ligand‑based virtual screening with structure‑guided docking, molecular dynamics simulations, and MM/PBSA binding‑free‑energy calculations. Within this framework, candidate regulatory compounds were identified that target Sirtuin‑1 (SIRT1) in HD and DJ‑1 (PARK7) in PD, and parallel transcriptomic analyses were used to characterize immune‑related changes that may drive or modulate neurodegenerative progression. For HD, Selisistat-guided similarity searching across the LOTUS natural-products repository (276,518 compounds; ≥68% Tanimoto) yielded 1,401 structural analogues that were refined through stringent Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET), Blood-brain barrier (BBB)- permeability, and drug-likeness filters prior to docking into the high-resolution SIRT1 catalytic domain (PDB 4I5I), prepared using CHARMM force-field optimization and validated via re-docking (RMSD < 2 Å). Subsequent 100-ns all- atom MD simulations (GROMACS/CHARMM36/TIP3P) demonstrated that top candidates, particularly LTS0217483, exhibit exceptional conformational stability, low-fluctuation RMSD/RMSF profiles, persistent H-bonding, and favorable compactness (Rg), while MM/PBSA decomposition revealed pronounced van der Waals and electrostatic contributions, confirming high-affinity NAD⁺-competitive ix inhibition driven by interactions with catalytic residues His363, Phe414, Val445, and Arg446. Parallel repurposing of FDA-approved anti-inflammatory drugs against the PD-associated mutant DJ-1 (2R1T) employed SwissADME pre- screening followed by CB-Dock and Webina docking, identifying Oxatomide and Levocabastine as high-binding, BBB-permeable modulators that form stable interactions with residues essential for DJ-1’s redox-chaperone and glyoxalase functions (e.g., Glu163, Glu170, Arg27, Ala56), thereby potentially mitigating oxidative and inflammatory perturbations characteristic of PD. An overview of the key molecular pathways involved in neuroinflammatory signaling is presented in this thesis to provide a biological context for the computational findings. Toll-like receptors, STAT3, p38 MAPK, and the NLRP3 inflammasome are discussed alongside protective regulators, such as SIRT1, SOCS protein, YY1, and MEF2. It has been shown that disruption of the balance between these pathways contributes to progressive neuronal damage, highlighting potential therapeutic targets. A robust in silico framework for CNS-targeted drug discovery is shown in this multidisciplinary study, and mechanistically validated lead scaffolds are identified for HD (LTS0217483) and PD, as well as a systems-level neuroinflammatory model that informs future translation strategies and precision therapeutic development for NDDs.