Cross-species validation of a 6-miRNA blood signature for Parkinson's disease: from MPTP mice to human PBMC and serum exosomes.

INTRODUCTION

Early detection of Parkinson's disease (PD) remains challenging due to the lack of reliable blood-based biomarkers. While microRNAs (miRNAs) show promise as circulating biomarkers, translating preclinical discoveries to clinically applicable panels requires rigorous validation across platforms and populations.

METHODS

We performed temporal miRNA profiling in an acute MPTP mouse model (day 0 vs. day 5, n = 4 per group) using limma differential expression analysis with FDR correction. To address high-dimensional small-sample challenges, we employed global permutation testing and stability selection with elastic net regularization over 2,000 iterations. A compact miRNA panel was derived and validated in three independent human cohorts: GSE16658 (PBMC, n = 32), GSE269776 (serum exosomes 2021, n = 76), and GSE269775 (serum exosomes 2020, n = 100). Performance was assessed using ROC analysis with permutation-based p-values.

RESULTS

Seventeen miRNAs showed significant time-dependent changes in MPTP-treated mice (FDR <0.05), with 15 down-regulated and 2 up-regulated at day 5. Stability selection identified a 6-miRNA panel comprising miR-92b, miR-133a, miR-326, miR-125b, miR-148a, and miR-30b. External validation demonstrated consistent discriminative performance across platforms: GSE16658 AUC = 0.696 (p = 0.060), GSE269776 AUC = 0.791 (p < 0.001), and GSE269775 AUC = 0.725 (p < 0.001).

DISCUSSION: The signature showed platform-agnostic stability, performing comparably in PBMC and serum exosomes despite biological and technical differences. A 6-miRNA signature derived from acute MPTP response translates effectively to human blood samples, demonstrating reproducible PD discrimination across multiple platforms. The compact panel size and cross-platform compatibility support its potential for clinical biomarker development. By integrating AI-enhanced feature selection and permutation-based validation, this study provides a reproducible framework for biomarker discovery and a foundation for future early detection and precision medicine in Parkinson's disease.