S-27-3: hypertensive nephropathy: glomerular proteomic profiling of kidney biopsies reveals a signature of disease progression

Objective: Hypertensive nephropathy, a major cause of chronic kidney disease (CKD) worldwide, requires kidney biopsy as the diagnostic gold-standard. However, histological findings are unspecific and specific prognostic markers are missing. Accordingly, our objective was to detect candidate markers based on glomerular protein signatures from the diagnostic kidney biopsy that predict subsequent disease progression. Design and method: We studied adult patients (n = 17) with an eGFR > 30 ml/min/1.73m 2 and proteinuria < 3 g/d from the Norwegian Kidney Biopsy Registry, including stable non-progressing patients (n = 9) and patients progressing (n = 8) to end-stage kidney disease (ESKD) within 20 years. Glomerular cross-sections from archival kidney biopsy sections were microdissected and processed for protein extraction. Proteomic analyses were performed using Q-exactive HF mass spectrometer and relative glomerular protein abundances were compared between progressive and non-progressive patients. Results: Amongst n = 1870 quality filtered proteins, n = 58 were differentially expressed in progressive and non-progressive glomerular samples, with absolute fold changes ≥ 1.5, p ≤ 0.05. However, only by using n = 17 glomerular proteins with absolute fold changes ≥ 2 and p ≤ 0.05, the hierarchical clustering and principal component analysis (PCA) effectively separated progressor and non-progressor patient samples. Supervised K nearest neighbor (KNN) analysis with leave-one-out internal cross validation was employed to develop a prognostic five protein classifier signature of hypertensive nephropathy progression. Thereby, Gamma-butyrobetaine dioxygenase (BBOX1, O75936) and Cadherin 16 (CDH16, O75309), overexpressed in progressors, and Eosinophil peroxidase (EPX, P11678), DnaJ homolog subfamily B member 1 (DNAJB1, P25685) and Alpha-1-syntrophin (SNTA1, Q13424), overexpressed in non-progressive glomeruli, correctly classified 16/17 samples into progressors and non-progressors. Respective classifier results, including area under the ROC curve of 0.99, hierarchical cluster analysis and PCA, are depicted in Figure 1 A-D (P = progressors; NP = non-progressors). Immunohistochemistry confirmed results of BBOX1 and CDH16. Geneset Enrichment Analysis (GSEA) showed that metabolic pathways were enriched in progressors, and structural cell pathways in non-progressors. Pathway analysis identified Epithelial Adherens Junction Signaling as most affected canonical pathway. We then evaluated protein signatures associated with hypertensive nephropathy progression with those associated with IgA nephropathy progression from our published dataset (Paunas T. Clin Proteom 2017) . The expression of only three proteins were altered in a similar direction in both datasets. KNN classifier proteins of progression of HN were either not detected in IgAN or were not significantly altered during progression of IgAN. Conclusion: Glomerular proteomic profiling from the initial diagnostic kidney biopsy can be used to discriminate progressive from non-progressive patients with hypertensive nephropathy.