Plasma EV Analysis Identifies Proteins Associated With T1D Risk, With Ernesto Nakayasu, PhD

A proteomic analysis of plasma extracellular vesicles (EVs) identified 448 proteins associated with islet autoimmunity and generated a predictive model for type 1 diabetes (T1D) risk, highlighting a potential approach for identifying biomarkers before clinical disease onset.

The findings were presented at the Endocrine Society (ENDO) 2026 Annual Meeting in Chicago, Illinois by Ernesto S. Nakayasu, PhD, a senior researcher at Pacific Northwest National Laboratory. In an interview with HCPLive, he discussed the potential of EVs as biomarkers, including a 3.2-fold increase in proteomic coverage compared with the team's previous plasma proteomics analysis.

By the time T1D reaches clinical onset, the immune system has destroyed an estimated 80%-90% of insulin-producing pancreatic beta cells, resulting in elevated blood glucose levels that require lifelong insulin therapy. Earlier identification of individuals at risk for T1D could help facilitate a smoother transition to treatment and potentially expand opportunities for disease-modifying immunotherapies before substantial beta-cell loss occurs.

“Extracellular vesicles are particles that cells secrete into the circulation, and they can be very good biomarkers because they carry signatures of the cells that secreted them, as well as signatures of disease,” Nakayasu explained. “That's why we're moving in that direction.”

Because plasma EV proteomics analysis can be complicated by co-isolated contaminants, Nakayasu and colleagues built upon a previously developed strategy that separates EV characterization and protein quantification into different experiments. According to the investigators, this approach allows researchers to prioritize EV recovery while maximizing quantification of individual proteins.

The team enriched plasma EVs using a method based on strong anion exchange beads (Mag-Net) followed by mass spectrometry analysis. Samples were obtained from 19 donors with islet autoimmunity, defined by the presence of circulating autoantibodies against islet proteins, and 17 control participants without autoantibodies.

Investigators identified and quantified 5482 proteins, representing a 3.2-fold increase in coverage compared with their previous plasma proteomics analysis for identifying T1D biomarker candidates. The analysis included 1315 of 1717 previously validated EV proteins (77%).

Statistical analysis identified 448 proteins that were quantitatively or qualitatively differentially abundant between autoantibody-positive (AAB+) participants and controls, including 69 proteins previously validated as EV-associated proteins.

In a functional enrichment analysis, investigators identified 25 pathways enriched among the differentially abundant proteins, including pathways related to autoimmune responses and lipid metabolism.

Machine learning analysis using a random forest model identified a multivariate protein panel capable of predicting autoantibody positivity with a receiver operating characteristic area under the curve (ROC-AUC) of 0.805.

Although the findings underscore the promise of plasma EV proteomics for identifying biomarkers associated with early T1D development, Nakayasu noted that technical challenges remain before the approach can be translated into routine clinical practice.

“The challenge is always that there are contaminants. It's very hard to purify,” Nakayasu said. “The way that we're trying to go around that is really understanding better the composition of the EVs and then quantifying the components rather than trying to get very pure fractions.”

Editor’s Note: Nakayasu reports no relevant disclosures.

References

1. Nakayasu ES, Dakup P, Bramer LM, et al. Plasma extracellular vesicles as predictive biomarkers for developing type 1 diabetes. Poster SUN-539. Presented at: ENDO 2026, Endocrine Society Annual Meeting; June 13-16, 2026; Chicago, IL.

2. Lucier J, Weinstock RS. Diabetes Mellitus Type 1. National Library of Medicine. Published October 5, 2024. Accessed June 13, 2026. https://www.ncbi.nlm.nih.gov/books/NBK507713/

3. Los E, Wilt AS. Type 1 Diabetes in Children. National Library of Medicine. Published June 26, 2023. Accessed June 13, 2026. https://www.ncbi.nlm.nih.gov/books/NBK441918/