Distinct proteomic signatures of urinary extracellular vesicles link to albuminuria and treatment responses in diabetic kidney disease

INTRODUCTION: Urinary extracellular vesicles (uEVs) offer a noninvasive approach to gain pathophysiological insight into kidney disease and treatment effects. Our study aimed to investigate the impact of albuminuria and antialbuminuric therapies on the uEV proteome in diabetic kidney disease.

METHODS: uEVs were isolated from urine samples of 20 patients with diabetes and albuminuria (age 64 years, estimated glomerular filtration rate 85 ml/min per 1.73 m2, urinary albumin creatinine ratio 130 mg/g) who participated in a crossover trial. Participants received telmisartan, empagliflozin, linagliptin, and baricitinib in random order across four 4-week treatment periods. Quantitative proteomics was performed using liquid chromatography-tandem mass spectrometry. Results were compared with two datasets (one single-nucleus, one single-cell) analyzing the effects of proteinuria and sodium-glucose cotransporter 2 inhibition (SGLT2i) on the kidney transcriptome.

RESULTS: A total of 2,441 proteins were identified in uEVs, of which 355 were significantly associated with albuminuria. Treatment with telmisartan, empagliflozin, linagliptin, and baricitinib resulted in, respectively, 86, 313, 80, and 47 differentially abundant uEV proteins with minimal overlap between treatments. Of the albuminuria-associated uEV proteins, 100 were significantly altered by at least one treatment. The overall proteomic signature was characterized by enrichment of podocyte-associated proteins, complement proteins, serpins, galectins, G-protein coupled receptors, lysosomal proteins, solute carriers, tubular injury proteins, and EV-related proteins such as Rab GTPases and tetraspanins. Comparison with the transcriptomic datasets showed that 15% of the albuminuria-associated uEV proteins and 62% of the sodium-glucose transport protein 2 inhibitor-responsive uEV proteins also exhibited changes at the transcript level in kidney cells, usually in a concordant manner.

CONCLUSIONS: Our study identified distinct uEV proteomic signatures linked to albuminuria and treatment responses in diabetic kidney disease, with minimal overlap across drugs and substantial concordance with kidney transcriptomic changes. These findings highlight the potential of uEVs as carriers of biomarkers for kidney injury and therapeutic effects.