Synergistic blood-based diagnostic value of AP3B1 and BMPR2 in Parkinson's disease.
Xiyan Zhao, Li Yang, Yumin Luan, Tao Ding, Xinglong Yang, Xin Geng, Tuo Zhang, Jigang Pan, Ziwen Xiao, Wei Pan, Tengxiang Chen
Abstract
Open AccessReliable blood-based biomarkers for Parkinson's disease (PD) are needed for minimally invasive diagnosis. We identified a synergistic mRNA biomarker pair, AP3B1 and BMPR2, detectable in blood through an integrative multi-omics workflow. DEGs from a meta-analysis of PD versus healthy controls (HCs) were intersected with DEG-enriched pathway genes and analysed via three-step SMR to identify PD risk candidates, from which machine learning (SVM-RFE and random forest) prioritized AP3B1 and BMPR2. Knockdown of each gene in SH-SY5Y-derived neurons reproduced Parkinsonian phenotypes, with protein docking and co-immunoprecipitation suggesting a direct interaction. An XGBoost model built on PPMI blood RNA-seq (n = 2585) using 25 established PD biomarkers (baseline AUC ~ 0.595) improved to 0.745 with addition of both AP3B1 and BMPR2. qRT-PCR in a cohort of clinical blood samples confirmed their downregulation in PD. These findings support AP3B1 and BMPR2 as a synergistic biomarker pair with speculative biological relevance and possible translational potential.