PROTEOMIC PROFILING OF GLOMERULI FROM KIDNEYS WITH HYPERTENSIVE NEPHROPATHY REVEALS SIGNATURE OF DISEASE PROGRESSION

Abstract Background and Aims Hypertensive nephropathy (HN) often represents an unspecific clinical diagnosis applied to non-diabetic, chronic kidney disease (CKD) patients with low-level proteinuria and elevated blood pressure. Kidney biopsy-based findings are the diagnostic gold standard, but they are not entirely specific and there is a need for prognostic markers. We aimed at defining candidate markers that predict disease progression based on protein signatures. Method We included adult patients (n=17) with an eGFR >30 ml/min/1.73m2 and proteinuria <3g/d from the Norwegian Kidney Biopsy Registry (NKBR). Subjects were divided into two groups: stable patients (n=9) and subjects with HN progression (n=8) leading to end-stage renal disease (ESRD) within 20 years of follow-up. Glomerular cross-sections were microdissected from 10 μm whole archival kidney biopsy sections and processed for protein extraction. Proteomic analyses were performed at our local facility PROBE in Bergen using a Q-exactive HF mass spectrometer. Abundances of glomerular proteins were compared between the two groups. Results Amongst a total of 1870 quality filtered proteins, we identified 58 proteins with absolute fold change (FC) ≥1.5, p≤0.05, including 17 proteins with absolute FC ≥2, indicative of HN progression (highest FC: Cadherin 16 and UDP-glucuronosyl-transferase 2B7). Hierarchical cluster and principal component analysis (PCA) with the 17 proteins showed clear separation of samples into these two disease clusters of HN progressors and non-progressors. To find proteins as biomarkers for the identification of progressors, we employed unsupervised K nearest neighbors validation algorithm coupled with leave-one-out internal cross-validation. Thereby, a set of five proteins (incl. cadherin 16) performed best and separated the two groups in a PCA, as shown in Figure 1. This classifier classified 16 of 17 samples correctly (AUC 0.993), misclassifying only one progressor sample. Applying Geneset Enrichment Analysis (GSEA; The Broad Institute, USA), in general metabolic pathways were up-regulated in progressors and structural cell pathways up-regulated in non-progressors. Ingenuity Pathway Analysis (IPA; Qiagen, USA) identified Epithelial Adherens Junction Signaling as the most affected canonical pathway (p=1.95E-06); five of six member proteins were down-regulated in progressors. Conclusion Glomerular proteomic profiling can be used to distinct progressors from non-progressors in HN.