Identification of CLIC5 as a Prognostic Biomarker and Correlated Immunomodulator for Lung Adenocarcinoma

Abstract

Background: Lung adenocarcinoma (LUAD) is one of the most common pathological types of lung cancer. The gene Chloride Intracellular Channel 5 (CLIC5) has an important role in neurophysiology, cardiovascular biology, and tumour biology. Here, we explored the prognostic value and immune infiltration of CLIC5 expression in LUAD patients.

Methods: We extracted transcriptional LUAD data from The Cancer Genome Atlas (TCGA) and the University of Alabama Cancer Database to explore CLIC5 expression profiles and their relation to CLIC5 and clinicopathological parameters. The relationship between CLIC5 and survival time was explored using Kaplan-Meier Plotter. Then, we integrated the data from TCGA and the Gene Expression Omnibus (GEO) database to perform univariate and multivariate Cox regression. We performed CLIC5 immunohistochemical staining on 167 lung adenocarcinoma samples for further verification. In addition, we analysed the Gene Ontology (GO) database, Kyoto Encyclopaedia of Genes and Genomes pathways and network analysis of protein-protein interactions in lung tissue, to explore the potential mechanism of CLIC5. To analyse the correlation between immune infiltration and CLIC5 expression, we first compared the expression of immune cells in tumour tissues and normal tissues based on the TCGA and GEO databases. We found 51 immunomodulators related to CLIC5 and structured their enrichment pathways as well as those of 50 correlated genes. We used a Cox regression model to identify multiple-gene risk prediction signatures. Finally, we assessed the prognostic accuracy of the risk scores via receiver operating characteristic curves.

Results: CLIC5 expression levels were significantly lower in LUAD tissue than in normal tissue. Lower CLIC5 expression was negatively correlated to the overall survival of LUAD patients based on survival analysis. We identified CLIC5 as an independent prognosis predictor. Functional network analysis suggested that CLIC5 is related to multiple pathways. CLIC5 expression is closely related to infiltration levels of many immune cells and immune marker sets in LUAD patients. Furthermore, the risk score based on immunomodulators related to CLIC5 was an independent prognosis predictor in the TCGA lung cohorts.

Conclusion: Our findings suggest that CLIC5 is a promising molecular marker for the prognosis and immune infiltration of LUAD patients.

Keywords: CLIC5; bioinformatic analysis; cancer immune infiltrates; immunomodulator; lung adenocarcinoma; prognosis biomarker.