Identification and Validation of IFI44 as Key Biomarker in Lupus Nephritis

Abstract

Lupus nephritis (LN) is a common and severe organ manifestation of systemic lupus erythematosus (SLE) and is a major cause of SLE related deaths. Early diagnosis is essential to improve the prognosis of patients with LN. To screen the potential biomarkers associated with LN, we downloaded the gene expression profile of GSE99967 from the Gene Expression Omnibus (GEO) database. Weighted gene co-expression network analysis (WGCNA) was utilized to construct a gene co-expression network and identify gene modules associated with LN. Gene Ontology (GO) analysis was also applied to explore the biological function of genes and identify the key module. Differentially expressed genes (DEGs) were identified and Maximal Clique Centrality (MCC) values were calculated to screen hub genes. Furthermore, we selected promising biomarkers for real-time PCR (qRT-PCR) and enzyme-linked immunosorbent assay (ELISA) validation in independent cohorts. Our results indicated that five hub genes, including IFI44, IFIT3, HERC5, RSAD2, and DDX60 play vital roles in the pathogenesis of LN. Importantly, IFI44 may considered as a key biomarker in LN for its diagnostic capabilities, which is also a promising therapeutic target in the future.

Keywords: biomarker; hub genes; lupus nephritis; type-I interferon; weighted gene co-expression network analysis.

Figure 1

Overall study design. WGCNA, weighted gene co-expression network analysis; GO, gene ontology; MCC, maximal clique centrality; DEGs, differentially expressed genes; GEO, gene expression omnibus; qRT-PCR, quantitative real-time PCR; ELISA, enzyme-linked immunosorbent assay.

Figure 2

Construction of weighted gene co-expression network. (A) Analysis of the scale-free topology model fit index for soft threshold powers (Left) and analysis of the mean connectivity for soft threshold powers (Right). The soft-thresholding power was set as four. (B) Clustering of module eigengenes. The red line represents cut height (0.5). (C) The cluster dendrogram of genes. (D) The cluster dendrogram and adjacency heatmap of eigengenes.

Figure 3

Key module identification. (A) Heatmap of module-trait relationships. The correlation coefficients and p-value are shown in each cell. Red presents a positive correlation and blue presents a negative correlation. (B) Gene significance of all genes in 12 modules related to LN. Different colors of columns present different modules. (C) Scatter plot for correlation between module membership and gene significance in darkgreen module (Left) and purple module (Right). The top 20 significant GO terms (biological process) of the purple module (D) and darkgreen module (E). The color and size of each point indicate the p-value and the number of genes in the corresponding term, respectively. LN, lupus nephritis; GO, gene ontology.

Figure 4

Validation of five hub genes in datasets. (A–C) The blue box refers to the normal group and the orange one refers to the LN group. Box represents mean ± SD by an unpaired t-test. *P < 0.05, **P < 0.01, ***P < 0.0001, ns, no significance; LN, lupus nephritis.

Figure 5

ELISA validation and ROC analysis. (A) Serum IFI44 (Left) and IFIT3 (Right) were measured in 51 healthy controls and 73 LN patients by ELISA. (B) Receiver operator characteristic (ROC) curves show the diagnostic performance of serum IFI44 (blue) and IFIT3 (green) in identifying LN patients. (C) 73 LN patients were divided into two groups (Inactive LN=22, Active LN=51) according to renal biopsy and the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI). (D) ROC curves of IFI44, anti-dsDNA, C3, and C4 in differentiating active LN from inactive LN patients. (E) Serum IFI44 was tested in 12 healthy controls and 25 patients with IgA nephropathy. Data shown are mean ± SD by an unpaired t-test. *P < 0.05, **P < 0.01, ****P < 0.0001, ns, no significance; HC, healthy control; LN, lupus nephritis; IgAN, immunoglobulin A nephropathy.