Gene Expression Analysis Reveals Novel Shared Gene Signatures and Candidate Molecular Mechanisms between Pemphigus and Systemic Lupus Erythematosus in CD4+ T Cells

Abstract

Pemphigus and systemic lupus erythematosus (SLE) are severe potentially life-threatening autoimmune diseases. They are classified as B-cell-mediated autoimmune diseases, both depending on autoreactive CD4⁺ T lymphocytes to modulate the autoimmune B-cell response. Despite the reported association of pemphigus and SLE, the molecular mechanisms underlying their comorbidity remain unknown. Weighted gene co-expression network analysis (WGCNA) of publicly available microarray datasets of CD4⁺ T cells was performed, to identify shared gene expression signatures and putative overlapping biological molecular mechanisms between pemphigus and SLE. Using WGCNA, we identified 3,280 genes co-expressed genes and 14 co-expressed gene clusters, from which one was significantly upregulated for both diseases. The pathways associated with this module include type-1 interferon gamma and defense response to viruses. Network-based meta-analysis identified RSAD2 to be the most highly ranked hub gene. By associating the modular genes with genome-wide association studies (GWASs) for pemphigus and SLE, we characterized IRF8 and STAT1 as key regulatory genes. Collectively, in this in silico study, we identify novel candidate genetic markers and pathways in CD4⁺ T cells that are shared between pemphigus and SLE, which in turn may facilitate the identification of novel therapeutic targets in these diseases.

Keywords: CD4+ T cells; autoimmunity; gene expression analysis; pemphigus; systemic lupus erythematosus; weighted gene co-expression analysis.

Figure 1

PCA plot illustrating the normalization procedure. (A) PCA plot showing clustering of the samples based on the gene expression profiling, before and (B) after batch correction on raw data. (C) PCA plot showing clustering of the samples after using identical background correction and normalization methods, before and (D) after batch correction. The X- and Y-axes represent the first and the second principal components and the associated percentage of variation.

Figure 2

Boxplots of eigengene values across modules. Boxplots depicting different identified modules on the X-axis and the corresponding module eigengene values for each group of samples on the Y-axis. The significance among the groups was calculated using Kruskal–Wallis test. *P < 0.05; **P < 0.01. PF, pemphigus foliaceus; PV, pemphigus vulgaris; SLE, systemic lupus erythematosus.

Figure 3

Gene–gene interaction network for the “magenta” module. De novo network generated by C3NET algorithm for the “magenta” module. The figure shows statistically significant (α < 0.05) edges predicted by the algorithm. Fully colored nodes represent the “magenta” module-associated genes. Empty nodes represent the regulatory genes (degree ≥ 5).

Figure 4

Gene–gene interaction network for the “salmon” module. De novo network generated by C3NET algorithm for the “salmon” module. The figure shows statistically significant (α < 0.05) edges predicted by the algorithm. Fully colored nodes represent the “salmon” module-associated genes. Empty nodes represent the regulatory genes (degree ≥ 4).

Figure 5

Interactions among genome-wide-associated genes and module-derived genes. Direct curated gene–gene interactions between modular genes and genes identified from SLE GWAS. Hub genes are represented by empty blue nodes. Common genes between SLE GWAS and the “magenta” module are denoted in blue nodes with red contour. SLE, systemic lupus erythematosus; GWAS, genome-wide association study.