Clinicopathologic correlations of urinary proteomic and metabolomic analysis in patients with renal aa amyloidosis and membranous nephropathy

Abstract

Background and Aims AA Amyloidosis is a multisystemic amyloidosis subtype that develops on the background of various chronic inflammatory etiologies. Urinary omics studies have become a promising tool for elucidating pathophysiology and prognosis of glomerular diseases. However, no urinary omics analysis has been performed focusing on renal AA amyloidosis in literature to the best of our knowledge. Our main aim in this study is to perform a comparative urine proteomic and metabolomic analysis of recently diagnosed renal AA amyloidosis and to investigate the correlation of bioinformatic results with clinical and pathological data.

Method Urine samples of 8 recently diagnosed AA amyloidosis (AA), 8 membranous nephropathy (MN) and 6 healthy control group patients were collected before kidney biopsy procedure. Proteomic analyzes were performed with nLC/Q-TOF MS/MS and metabolomic analyzes were performed by GC/MS in all patients. Biopsy specimens were scored according to glomerulosclerosis (G), tubular atrophy (TA) and interstitial fibrosis (IF) grades by two pathologists. Raw spectroscopic data was analyzed using MaxQuant and MS-DIAL programs for proteomic and metabolomic studies, respectively. Statistical analysis of the differences in molecules between study groups were performed with ANOVA and HSD-Tukey tests. Principal component (PCA) and heatmap analyzes were made in R language, while gene ontology (GO), network and functional enrichment analysis of bioinformatic results were performed with PANTHER, STRING and MetaboAnalyst databases.

Results In comparison between AA and MN groups, median eGFR values tend to be lower in the AA group (67.6 vs 112 ml/min/1.73 m2 respectively, p = 0.08). Median 24-hour urine protein levels did not show statistically significant difference (9499 vs 9512 mg/day respectively, p = 0.9). Percentage of patients with moderate/severe IF/TA was higher and G score was tend to be in AA group compared to MN group (p values 0.02 and 0.07 for IF/TA and G scores, respectively). As a result of proteomic analysis, a total of 859 proteins were determined. Statistical analysis showed 51 proteins that were significantly differ in AA group compared to the control group. GO and functional enrichment analyzes showed that statistically most significant sub-domains were mainly related with cell-cell adhesion (Figure 1 & 2). In comparative analysis between AA and MN patients, uromodulin (UMOD) was lower in the AA group than in the MN group (log2FC -3.37), whereas ribonuclease 1 (RNASE1) and α-1-microglobulin/bikunin precursor protein (AMBP) were higher in the AA group (log2FC 3.41 and 3.07, respectively). In Spearman correlation analyzes, significant negative correlations were demonstrated between UMOD-proteinuria (r = -0.48, p = 0.03) and between AMBP-eGFR (r = -0.69, p = 0.003) variables. Metabolomic analysis showed 9 metabolites that were significantly different between AA and other study groups. Myo-inositol and urate were higher in AA group compared to MN group, while D-mannitol and N-acetylglutamate were higher in AA group compared to the control group. Significant positive correlation independent of GFR was detected between RNASE1 and urate (r = 0.63, p = 0.01).