Blood Exosome miRNA Signature Shows Strong Accuracy for Detecting Parkinson’s Disease

Scientists have identified a molecular fingerprint in tiny blood-borne vesicles that distinguishes people with Parkinson’s disease with high accuracy, according to a study published in npj Parkinson’s Disease. The research combined microRNA profiling of blood exosomes with proteomic and single-cell analyses, drawing on multiple public datasets for discovery and an in-house cohort for independent validation.

Exosomes are small membrane-bound particles released by cells that can carry central nervous system-derived cargo across the blood-brain barrier into the bloodstream, making them a promising liquid biopsy platform for brain diseases that currently lack reliable non-invasive diagnostic biomarkers. The team applied consensus feature selection integrating five different statistical methods to identify robust biomarker panels, which were then evaluated using four classification models.

The resulting 69-microRNA signature achieved strong classification performance, with Random Forest showing the best generalization at an area under the curve of 0.952. By contrast, a 63-protein signature built from exosomal proteomics showed limited reproducibility across studies, with a best AUC of 0.611, a shortfall the authors attribute to heterogeneity between measurement platforms.

Tissue origin profiling indicated that the signature microRNAs derive from both the central nervous system and peripheral immune compartments. Functional enrichment analyses across the omics layers consistently pointed to two converging pathological themes: failure of mitochondrial energetics and suppression of complement and coagulation cascades.

To trace where the exosomal cargo originates, the researchers used deconvolution methods and validated their findings with independent single-cell RNA sequencing data. This pointed to a distinct immune-metabolic axis involving B cells and CD4-positive T cells as contributors to the exosomal signal.

The study’s authors describe the work as establishing a foundation for exosome-based liquid biopsy and mechanistic biomarker development in Parkinson’s disease, while also framing mitochondrial dysfunction and peripheral immune dysregulation as core pathological features supported by converging multi-omics evidence. Because the microRNA signature reproduced well across cohorts whereas the protein signature did not, the findings suggest that nucleic acid cargo in exosomes may currently offer a more stable route to blood-based disease classification than proteomic measurements, which remain sensitive to platform differences.

The research was published as an open-access article on 22 October 2026, shared early to provide faster access to the peer-reviewed, accepted version, which is subject to further edits before the final Version of Record. As with any biomarker study, the reported classification performance reflects the analyzed cohorts and models, and the authors’ stated aim is a foundation for future diagnostic development rather than an already approved clinical test.

Subject of Research: Exosomal microRNA biomarkers for Parkinson's disease diagnosis and multi-omics pathology

Article Title: Cross-cohort exosomal miRNA profiling identifies a robust Parkinson’s disease signature with multi-omics evidence for mitochondrial and immune pathology

Article References: Liu, S., Wang, S., Ye, X., Zeng, L., Wang, M., Wang, X., & Chen, C. (2026). Cross-cohort exosomal miRNA profiling identifies a robust Parkinson’s disease signature with multi-omics evidence for mitochondrial and immune pathology. npj Parkinson's Disease. https://doi.org/10.1038/s41531-026-01566-7

Image Credits: AI Generated

DOI: 10.1038/s41531-026-01566-7

Keywords: Parkinson's disease, exosomes, microRNA, biomarkers, liquid biopsy, mitochondrial dysfunction, immune dysregulation, proteomics, single-cell RNA sequencing, machine learning, Random Forest, blood-brain barrier