Identification of Biomarkers Related to Systemic Sclerosis With or Without Pulmonary Hypertension Using Co-expression Analysis

Abstract

Systemic sclerosis (SSc), also known as scleroderma, is an autoimmune disease with multiple system involvement, and pulmonary complications, including pulmonary hypertension (PH), are leading causes of death. This study aimed to develop early biomarkers to distinguish SSc with or without PH from normal population using bioinformatics approaches. The gene expression profile GSE22356, which contains 10 SSc samples with PH, 10 SSc samples without PH, and 10 normal samples, was obtained from the Gene Expression Omnibus database. First, we constructed co-expression networks and identified critical gene modules using the weighted gene co-expression network analysis. Then, functional enrichment analysis of significant modules was performed. Finally, the “real” hub gene was screened out by intramodule analysis and protein–protein interaction networks, and the receiver operating characteristic analysis was conducted. A total of 5046 genes were screened out to construct co-expression networks, and 18 modules were identified. Of these modules, the turquoise module had a strong correlation with SSc only, whereas the midnightblue module showed an obvious positive correlation with SSc with PH. Functional enrichment analysis indicated that the turquoise module was mainly enriched in transcription of DNA template and its regulation and protein ubiquitination and involved in apoptosis and pyrimidine metabolism pathway. The midnightblue module was significantly associated with inflammatory and immune response and pathways in Staphylococcus aureus infection and Chagas disease. The “real” hub genes in the turquoise module were WDR36, POLR1B, and SRSF1, and those in midnightblue were TLR2 and TNFAIP6.