Lipid metabolism dysregulation in Parkinson's disease: A Mendelian randomization and transcriptomic analysis.

Background: Parkinson's disease (PD) is a progressive neurodegenerative disorder in which mechanisms linking metabolic dysregulation to neuronal vulnerability remain incompletely understood. Increasing evidence suggests that dysregulation of lipid metabolism plays an important role in modulating mitochondrial function, protein homeostasis, and neuroinflammation. In this study, we aimed to identify lipid metabolism-associated molecular drivers of PD and explore their potential contribution to the disease. Methods: Gene expression datasets GSE20141 and GSE20292 were obtained from the Gene Expression Omnibus (GEO) public database. Weighted Gene Co-expression Network Analysis (WGCNA) was performed to identify PD-associated modules, which were intersected with differentially expressed genes and curated lipid metabolism gene sets to define candidate lipid metabolism-related genes (LMRGs). Two machine learning algorithms, LASSO regression and Boruta, were applied to identify potential biomarkers. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. Gene Set Enrichment Analysis (GSEA), immune infiltration analysis, and Mendelian randomization (MR) were performed to assess biological functions and causality. Drug-gene interactions were predicted using the DGIdb. Results: Three genes: DGKD, LEP, and GGPS1, were consistently identified as candidate PD-associated LMRGs with reproducible diagnostic performance (AUC > 0.7). Among these, GGPS1, a key enzyme in the mevalonate pathway regulating protein prenylation, showed enrichment in the neurotrophin signaling pathway and was associated with central memory T cells. These findings suggest a mechanistic link between lipid metabolism, small GTPase prenylation, and neuronal homeostasis. Drug-gene interaction analysis identified compounds targeting GGPS1, including bisphosphonate derivatives, as potential therapeutic candidates for further investigation. Conclusion: This integrative analysis identified DGKD, LEP, and GGPS1 as lipid metabolism-associated molecular signatures of PD and highlights GGPS1-mediated prenylation pathways as a potential mechanistic axis linking metabolic dysregulation to neurodegeneration. These findings provide a framework for future experimental validation and therapeutic exploration targeting lipid-driven pathways in PD. Ongoing studies are evaluating the functional impact of GGPS1 modulation on mitochondrial homeostasis and neuronal vulnerability in human neuronal models.