Longitudinal blood microsampling and proteome monitoring facilitate timely intervention in experimental type 1 diabetes

Abstract

Symptoms of immune-mediated diseases (IMIDs) typically appear after irreversible tissue damage, making early interventions based on pre-symptomatic indicators crucial. Current efforts to identify molecular markers of early disease lack the resolution, convenience and cost efficiency required to prevent irreversible tissue damage. Analyzing frequently self-collected samples, such as dried blood spots (DBS), could enable the earlier detection of diseases, identify disease-predictive markers and facilitate tailored interventions. To test this, we regularly microsampled a mouse model infected with a type 1-diabetes (T1D)-associated virus. This longitudinal DBS sample collection was analyzed for 92 circulating proteins, revealing transient molecular changes in virus-infected animals that would have been missed with less frequent sampling. Machine learning predicted infection status after day 2 post-infection with >90% accuracy, enabling well-timed treatment of virus-infected animals and diabetes prevention. Our study demonstrates the utility of frequent blood microsampling to monitor disease during the pre-symptomatic phase, allowing for timely interventions. TeaserFrequent blood microsampling detects early biomarkers, enabling timely intervention in immune-mediated diseases