# Proteomic Analysis Identifies Dysregulated Proteins in Albuminuria: A South African Pilot Study

## TL;DR

This study identifies proteins in urine that could help detect kidney disease early, before traditional markers show issues.

## Contribution

The study identifies 80 differentially abundant urinary proteins and an 80-protein model with high predictive accuracy for early kidney disease detection.

## Key findings

- 80 proteins were found to be differentially abundant in albuminuric individuals compared to controls.
- An 80-protein model predicted cases with 91.3% accuracy.
- Key pathways include insulin growth factor functions and extracellular matrix organization.

## Abstract

Chronic kidney disease remains a global health priority, only detected at relatively advanced stages by current markers. Identifying alternative markers for early detection is imperative. In this study, we profiled the urinary proteome in patients with albuminuria and well-preserved eGFR. We identified 80 proteins that were differentially abundant between the cases (albuminuria) and controls (normoalbuminuria). Among these, 12 proteins (SERPINA1, ALB, SERPINC1, AFM, PIGR, A1BG, COL6A1, MYG, LV39, MUC1, ICOSLG, and UMOD) had the highest discriminating abilities (area under curve > 0.8) between the cases and controls. When differentially abundant proteins were combined into an 80-protein model, the model was able to predict cases from controls with a predictive accuracy of 91.3%. The top five enriched biological pathways associated with the differentially abundant proteins included insulin growth factor functions, innate immunity, platelet degranulation, and extracellular matrix organization.

Albuminuria may precede decreases in the glomerular filtration rate (GFR) and both tests are insensitive predictors of early stages of kidney disease. Our aim was to characterise the urinary proteome in black African individuals with albuminuria and well-preserved GFR from South Africa. This case-controlled study compared the urinary proteomes of 52 normoalbuminuric (urine albumin: creatinine ratio (uACR) < 3 mg/mmol) and 56 albuminuric (uACR ≥ 3 mg/mmol) adults of black African ethnicity. Urine proteins were precipitated, reduced, alkylated, digested, and analysed using an Evosep One LC (Evosep Biosystems, Odense, Denmark) coupled to a Sciex 5600 Triple-TOF (Sciex, Framingham, MA, USA) in data-independent acquisition mode. The data were searched on SpectronautTM 15. Differentially abundant proteins (DAPs) were filtered to include those with a ≥2.25-fold change and a false discovery rate ≤ 1%. Receiver–operating characteristic curves were used to assess the discriminating abilities of proteins of interest. Pathway analysis was performed using Enrichr software. As expected, the albuminuric group had higher uACR (7.9 vs. 0.55 mg/mmol, p < 0.001). The median eGFR (mL/min/1.73 m2) showed no difference between the groups (111 vs. 114, p = 0.707). We identified 80 DAPs in the albuminuria group compared to the normoalbuminuria group, of which 59 proteins were increased while 21 proteins were decreased in abundance. We found 12 urinary proteins with an AUC > 0.8 and a p < 0.001 in the multivariate analysis. Furthermore, an 80-protein model was developed that showed a high AUC ˃ 0.907 and a predictive accuracy of 91.3% between the two groups. Pathway analysis found that the DAPs were involved in insulin growth factor (IGF) functions, innate immunity, platelet degranulation, and extracellular matrix organization. In albuminuric individuals with a well-preserved eGFR, pathways involved in preventing the release and uptake of IGF by insulin growth factor binding protein were significantly enriched. These proteins are indicative of a homeostatic imbalance in a variety of cellular processes underlying renal dysfunction and are implicated in chronic kidney disease.