Background/Purpose: LN is a serious complication of SLE that affects about 20-40% of all lupus patients and leads to kidney damage, end-stage renal disease, and patient mortality. Despite advances in therapy, progression to end stage renal disease has not been affected. Therefore, it is important to re-consider the pathogenic mechanisms involved in LN as a basis for development of more effective therapies. Here we present a multi-pronged approach to characterize LN via bioinformatic analysis of gene expression data obtained from kidney biopsies.

Methods: Genomic expression profiling data of LN patient biopsies, microdissected into glomerulus and tubulonterstitium (TI), was sourced from GSE32591 via the GEO database. Differentially expressed genes (DEGs) detected in LN-derived samples relative to samples from healthy individuals were interrogated for cell infiltrate composition using gene set variation analysis (GSVA) against a curated database of immune and non-immune cell type signatures (I-SCOPE, T-SCOPE).  Weighted gene co-expression network analysis (WGCNA) was used to generate gene modules correlated to clinical variables. DEGs were further functionally characterized using a curated immunity-specific gene functional category database (BIG-C) and IPA signaling pathway analysis software.  Queries of the perturbation database (LINCS, Library of Integrated Network-Based Cellular Signatures) were used to identify possible upstream regulators of altered gene expression patterns in LN samples as well as to identify drugs that could reverse abnormal gene expression profiles.

Results: WGCNA produced 6 gene modules (3 glomerulus, 3 TI) positively correlated with disease stage as measured by WHO class. These modules were enriched in signatures for several immune cell types, including granulocytes, pDC, DC, myeloid cells, CD4⁺/CD8⁺ T cells, and B cells. Additionally, the presence of both IG-κ and -λ as well as V_L genes and detection of pre- and post-switch PCs as indicated by IgM, IgD, and IgG1 Ig Heavy Chain genes indicate polyclonal PC infiltration. Podocyte signatures were detected as enriched in WGCNA modules negatively correlated with WHO class. Chemokines and pathways that mediate lymphocyte proliferation, organization and/or recruitment into lupus kidney tissue were detected as enriched via BIG-C and IPA analysis, including the cytokines TNF, IL1β, IL2, IL6, IL12, IL17, IL23, and IL27 and signaling pathways including CD40L, PI3K, NF-κB, NF-AT, and p70S6K.  IPA upstream regulator analysis indicated ongoing signaling by cytokines such as TNF, IFNγ, IFNα, CD40L, IL1β, IL2, IL6, and IL17.  Interestingly, connectivity analysis using LINCS elucidated high priority drug targets such as IFNβ (PF-06823859), IL12 (Ustekinumab), and S1PR (Fingolimod) that may prove to be good options for therapeutic intervention.

Conclusion: Bioinformatic analysis of LN gene expression highlights several dysregulated signaling pathways that can form the targets of novel therapeutic strategies, and further elucidation of these signatures may enhance clinical surveillance and diagnosis of LN to improve patient outcomes.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/analysis-of-lupus-nephritis-gene-expression-reveals-dysregulation-of-pathogenic-pathways-activated-within-infiltrating-cells/