Age modulates protein-outcome associations in deceased donor kidneys

ABSTRACT Background Organ availability limits kidney transplantation, the best treatment for end-stage kidney disease. Globally, deceased donor acceptance criteria have been relaxed to include older donors, which comes with a higher risk of inferior posttransplant outcomes. Donor age, although negatively impacts transplant outcomes, lacks granularity in predicting graft dysfunction. Better donor kidney assessment and characterization of the biological mechanisms underlying age-associated donor organ damage and transplant outcomes is key to improving donor kidney utilisation and transplant longevity. Methods 185 deceased pretransplant biopsies (from brain and circulatory death donors aged 18-78 years) were obtained from the Quality in Organ Donation (QUOD) biobank and proteomic profiles were acquired by mass spectrometry. Machine learning exploration using prediction rule ensembles guided LASSO regression modeling of kidney proteomes that identified protein signatures and biological mechanisms associated with 12-m posttransplant outcome. Data modeling was validated on held-out data and contextualised against published spatially resolved kidney injury related transcriptomes. Results Our analysis highlighted that outcomes were best modeled using combination of donor age and protein abundance signatures, revealing 539 proteins with these characteristics. Modeled age:protein interactions demonstrated stronger associations with transplant outcomes than age and protein alone and revealed mechanisms of kidney injury including metabolic changes and innate immune responses correlated with poor outcome. Comparison to single-cell transcriptome data suggests protein-outcome associations to specific cell types. Conclusions Molecular signatures resulted from integration of donor age and proteomic profiles in deceased donor kidney biopsies offer the potential to develop improved pretransplant organ assessment and aid decisions on perfusion interventions.