Assessing Extracellular Vesicle Proteins as Predictive Biomarkers for Developing Type 1 Diabetes

Area of Science:

• Biochemistry
• Immunology
• Proteomics

Background:

• Type 1 diabetes (T1D) arises from autoimmune destruction of insulin-producing beta cells.
• Predictive biomarkers are crucial for developing preventive therapies and understanding T1D's autoimmune trigger.
• Plasma extracellular vesicles (EVs) are valuable sources for biomarker discovery due to their disease-specific cargo.

Purpose of the Study:

• To evaluate plasma EV proteomics for identifying predictive biomarkers of T1D development.
• To assess the potential of EV protein cargo to distinguish individuals with islet autoimmunity from controls.

Main Methods:

• Plasma EVs were captured using strong anion exchange beads (Mag-Net) from individuals with islet autoimmunity (AAB+) and controls.
• EV protein cargo was analyzed using mass spectrometry, identifying and quantifying thousands of proteins.
• Machine learning (random forest) was employed to test the predictive capacity of the identified proteins.

Main Results:

• Over 5,480 proteins were identified in plasma EVs, significantly increasing proteome coverage.
• 448 proteins were found to be differentially abundant between AAB+ individuals and controls, including 69 verified EV proteins.
• A random forest model achieved an area under the receiver operating characteristic curve of 0.81 for predicting AAB+ status.

Conclusions:

• Plasma EV proteomics is a promising approach for identifying predictive biomarkers of T1D.
• The identified differentially abundant proteins and pathways offer insights into T1D pathogenesis.
• This methodology supports further research into early T1D detection and therapeutic strategies.