LINKING TRANSCRIPTOMIC PROFILES OF KIDNEY AND BLOOD SAMPLES PROVIDES INSIGHT INTO IDENTIFICATION OF LUPUS NEPHRITIS

O058 / #380 Topic: AS12 - Genetics, Epigenetics, Transcriptomics ABSTRACT CONCURRENT SESSION 10: INTEGRATING PROTEOMIC & TRANSCRIPTOMICS IN SLE 24-05-2025 10:40 AM - 11:40 AM Background/Purpose Current clinical methods to diagnose and evaluate the severity of lupus nephritis (LN) rely on identification of kidney dysfunction followed by invasive kidney biopsies. Here, we sought to identify molecular profiles of LN in the kidney tissue that would be reflected in blood gene expression. Methods Gene expression was analyzed from 46 kidney biopsies for which renal disease classification had been carried out by a blinded clinical pathologist and 91 blood samples from lupus patients with or without biopsy documented LN. For blood samples from patients with LN, biopsies were taken at the time of blood draw. Each dataset was analyzed by Gene Set Variation Analysis (GSVA) for enrichment of gene modules identifying immune cells/pathways, metabolism pathways, and kidney tissue cells. Samples were clustered using k-means and ordered based on ISN/RPI classification of disease involvement. Results GSVA and unsupervised k-means clustering of kidney tissue identified 4 subsets of LN (Figure 1A). Subsets 1-3 exhibited molecular profiles indicative of increasing severity, including upregulation of immune/inflammatory modules, decrease in metabolism modules and decrease in kidney tissue modules. Subset 4 was characterized by de-enrichment of immune modules and restored enrichment of metabolism and kidney tissue modules, but retained a loss of the podocyte and proximal tubule gene modules indicative of a post-inflammatory state and end organ damage. The molecular profile of each subset was associated with the ISN/RPI histologic classification (Figure 1B-D). Subset 1 was dominated by class II mesangial LN and class V membranous LN with low activity and chronicity indices. Subset 2 contained the majority of class III, focal, proliferative LN patients and increased activity and chronicity indices compared to Subset 1. The most active disease cluster, Subset 3, largely consisted of class IV, diffuse proliferative LN patients with the highest overall activity and chronicity scores. Finally, Subset 4 was dominated by class V LN patients. Gene modules used to separate LN kidney biopsies were able to stratify blood gene expression from lupus patients with or without LN (Figure 2A). Blood samples from LN patients were largely in Subsets 3-4 and mean SLEDAI was significantly increased in Subset 3 (Figure 2B). Among the LN patients, Subset 3 had the highest activity index and Subsets 1 and 4 had the highest chronicity indices (Figure 2C-D). (a) GSVA heatmap of 46 LN patients with histological classification for enrichment of immune, metabolism, and kidney tissue gene modules. (b) ISN histological kidney classification for patients in each subset. Average renal activity (c) and chronicity (d) indices for kidney biopsies from each subset. *p<0.05 (a) GSVA heatmap of 91 lupus patients with or without LN for enrichment of immune, metabolism, and kidney tissue gene modules. (b) Average SLEDAI for patients in each subset. Average renal activity (c) and chronicity (d) indices for LN patients with kidney biopsies in each subset. *p<0.05 Figure 1: Gene expression based clustering and clinical evaluation of LN kidney biopsies. Figure 2: Gene expression based clustering and clinical evaluation of blood from lupus patients. Conclusions Gene expression analysis of LN kidney biopsies revealed 4 subsets with distinct profiles of gene module enrichment indicative of immune involvement, metabolic dysfunction, and tissue damage that aligned with histologic class. Although there was greater heterogeneity between subsets in the blood as compared to the kidney, distinct gene profiles associated with higher disease severity and LN activity were identified.