Abstract
BACKGROUND: The exact mechanisms of PD are unclear, but Parkin-mediated mitophagy dysfunction is believed to play a key role. We investigated whether blood levels of Parkin and other biomarkers are linked to the risk of developing PD. METHODS: Baseline blood measures of Parkin and other biomarkers, including Homocysteine, carcinoembryonic antigen, Urea, total proteins, total cholesterol, creatine kinase, and albumin, were collected from 197 clinically diagnosed Parkinson's disease participants and 107 age-matched healthy controls in Wenzhou Parkinson's Biomarkers and Living Characteristics study. We conducted bioinformatics analysis using three datasets from the GEO database: GSE90514 (Cohort 1: PD = 4, HC = 4), GSE7621 (Cohort 2: PD = 16, HC = 9), and GSE205450 (Cohort 3: PD = 69, HC = 81). RESULTS: Using a bioinformatic approach, we identified dysregulated biological processes in PD patients with PRKN mutations. Compared to controls, significant abnormalities were observed in blood levels of Parkin, Hcy, total proteins, urea, albumin, and CEA in PD patients. A model incorporating Parkin, Hcy, total proteins, and urea effectively distinguished PD from healthy controls, achieving a higher accuracy (AUC 0.841) than other biomarker combinations. Gene set enrichment analysis suggested that pathways such as PINK1-Parkin-mediated mitophagy, urea cycle, cysteine degradation, and riboflavin metabolism may be involved in PRKN mutation. Additionally, the link between Parkin and PD was partially mediated by CEA and albumin, not by Hcy, total proteins, or urea. CONCLUSION: Our findings indicate that blood Parkin levels may be a minimally invasive biomarker for PD diagnosis. The model, which included Parkin, Hcy, total proteins, and urea, effectively distinguished PD from HC with greater accuracy.