INVESTIGATING THE ROLE OF PROTEIN MISFOLDING IN NEURODEGENERATIVE DISEASES: DEVELOPING NEW BIOCHEMICAL MARKERS FOR EARLY DIAGNOSIS

Abstract

Protein misfolding and aggregation underlie the pathogenesis of Alzheimer’s disease (AD) and Parkinson’s disease (PD), yet sensitive, non-invasive markers for preclinical detection remain lacking. In this study, we developed and validated a multimodal biomarker panel by combining real-time quaking-induced conversion (RT-QuIC), multiplex immunoassays, metal-binding stoichiometry, biophysical characterization, and machine learning. Cerebrospinal fluid (CSF) and plasma from 30 AD patients, 30 PD patients, and 30 age-matched controls underwent RT-QuIC, revealing significantly shorter lag times (11.8 ± 1.2 h in AD; 13.5 ± 1.4 h in PD vs. 19.7 ± 1.6 h in controls) and elevated maximum fluorescence (14 200 ± 850 RFU in AD; 12 800 ± 780 RFU in PD vs. 9 000 ± 600 RFU in controls). Multiplex assays confirmed elevated CSF Aβ₄₂ (512 ± 45 pg/mL), total tau (112 ± 12 pg/mL), and p-Tau₁₈₁ (62 ± 8 pg/mL) in AD, and increased plasma α-synuclein (340 ± 30 pg/mL) in PD. Metal analyses showed higher Cu:protein (1.2 ± 0.1 mol/mol in AD; 1.1 ± 0.1 in PD vs. 0.8 ± 0.1 in controls) and Zn:protein ratios (0.9 ± 0.1 in AD; 0.85 ± 0.1 in PD vs. 0.6 ± 0.1). Dynamic light scattering and circular dichroism revealed increased aggregate size and altered secondary structure in disease cohorts. Integrating these features, XGBoost achieved 91.7 % accuracy, 92.3 % sensitivity, 91.0 % specificity, and AUC = 0.95, with RT-QuIC kinetics and metal-binding ratios as top predictors. Our workflow enables early, high-fidelity detection of protein misfolding, paving the way for preclinical screening and timely therapeutic intervention in neurodegenerative diseases.