Prognostic value of CLIC3 mRNA overexpression in bladder cancer

Data sources

The gene expression data from the GSE7476 and GSE13507 datasets and related clinical characteristics and prognostic information were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). Differentially expressed genes were obtained through the “limma” package. CLIC3 was selected for further analysis. The expression data on CLIC3 mRNA was obtained from the GSE7476 and GSE13507 datasets. The clinicopathological and prognostic information of 165 BC patients was obtained from the GSE13507 dataset. The expression of CLIC3 mRNA in BC and normal tissues was verified using the Oncomine (https://www.oncomine.org) and The Cancer Genome Atlas (TCGA) database https://cancergenome.nih.gov/. Differentially expressed genes were obtained through the “DESeq” package from TCGA database. Immunohistochemical images were obtained from the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/). The relationship between CLIC3 mRNA expression and overall survival in BC patients was analyzed by the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/).

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of the genes coexpressed with CLIC3

To predict the possible functions and pathways of CLIC3 by studying the pathways of its coexpressed genes, the genes coexpressed with CLIC3 were retrieved and extracted from the Multi Experiment Matrix (MEM) database (http://biit.cs.ut.ee/mem/index.cgi) and cBioPortal (http://www.cbioportal.org/). The coexpressed genes were then taken together and displayed in a Venn diagram (http://bioinfogp.cnb.csic.es/tools/venny/). The coexpressed genes were entered into the Database for Annotation, Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov/, Version 6.8) for GO and KEGG analyses. The critical value of significance for the gene enrichment and pathway screening was set at P < 0.05.

GSEA

GSEA is a computational method used to determine whether a group of genes defined a priori show statistically significant and consistent differences between two biological states (Subramanian et al., 2005). GSEA was performed using GSEA3.0 (http://www.broad.mit.edu/gsea/) to investigate the effect of CLIC3 mRNA expression levels on biomarker gene sets in patients and to explore the possible link between CLIC3 and the pathways in BC. The BC dataset GSE7476 was divided into a high expression group and a low expression group according to the expression level of CLIC3 mRNA. The C2.cp.kegg.v6.2.symbols.gmt dataset was obtained from the Molecular Signatures Database (MsigDB) at the gene enrichment analysis website. Then, the enrichment analysis was performed according to the method of default weighted enrichment statistics, and the analysis was repeated 1,000 times at a time.

Collection of human tissue specimens

In this study, between January 2019 and June 2019, eleven samples of BC tissues and adjacent normal tissues were collected from the Department of Urology, Affiliated Haikou Hospital of Xiangya Medical College, Central South University. These tissues were used to detect the expression level of CLIC3 mRNA by qRT-PCR. Verbal consent from all patients and approval from the ethics committee of Affiliated Haikou Hospital of Xiangya Medical College, Central South University were obtained before collection. The approval number is 2016(research ethical review)-005.

qRT-PCR

Total RNA and cDNA were prepared with TRIzolReagent (Vazyme, Nanjing, China), TransScript One-Step gDNA Removal (Vazyme) and cDNA Synthesis SuperMix (Vazyme) according to the manufacturer’s instructions. cDNA was amplified on the LightCycler 480II system by using SYBR qPCR Master Mix (Vazyme). The oligonucleotide sequences of the primers used for qRT-PCR are as follows: CLIC3 (forward: 5′-CAGATCGAGGACTTTCTGGAG-3′, reverse: 5′-GGAGAACTTGTGGAAAACGTC-3′) and GAPDH (forward: 5′-CAGGAGGCATTGCTGATGAT-3′, reverse: 5′-GAAGGCTGGGGCTCATTT-3′). GAPDH was used as an internal control to standardize cycle threshold (CT) values to estimate gene expression. Relative mRNA expression was calculated using the 2^−ΔΔct method. Each sample was processed in triplicate to ensure quantitative accuracy.

Statistical analysis

Statistical analyses were performed using SPSS (version 18.0), GraphPad Prism (version 7.0) and R software (version 3.5.1). T-tests were used to compare the expression of CLIC3 mRNA between BC tissues and normal bladder tissues. The relationship between CLIC3 and the clinicopathological characteristics of BC patients was analyzed using the chi-square test. The overall survival and cancer-specific survival of BC patients were analyzed using the log-rank test with the Kaplan–Meier method. Cox univariate and multivariate analyses of the overall survival and cancer-specific survival of BC patients were performed. P < 0.05 was considered statistically significant. A gene set of with P < 0.05 and a false discovery rate (FDR) <0.25 in the GSEA was judged to be a significantly enriched gene set.

Expression level of CLIC3 mRNA in BC

Figure 1 shows the flow chart of our research. To explore the differential expression of CLIC3 mRNA in BC tissues and normal tissues, two GEO datasets (GSE13507 and GSE7476) were used. GSE13507 includes 165 tumor tissue samples and 68 normal tissue samples, and GSE7476 includes 9 tumor tissue samples and 3 normal tissue samples. The results showed that the expression of CLIC3 mRNA in BC tissues was significantly higher than that in normal tissues (Figs. 2A–2B, P < 0.01). In Oncomine, the ‘Sanchez-Carbayo Bladder 2’ dataset contains 48 normal tissue samples, 81 infiltrating BC samples and 28 superficial BC tissue samples. The ‘Dyrskjot Bladder 3’ dataset contains 9 normal tissue samples, 5 bladder mucosa samples and 13 infiltrating BC samples. The ‘Lee Bladder’ dataset contains 68 bladder mucosa samples, 62 infiltrating BC samples and 126 superficial BC tissue samples. Data from the Oncomine database indicated that CLIC3 mRNA expression was higher in BC tissue than in normal tissue (Figs. 2C–2G, P < 0.01). Data from the TCGA database also indicated that CLIC3 mRNA expression in BC tissues was higher than that in normal tissues (Fig. 2H, P = 0.024). The above studies indicated that the expression of CLIC3 mRNA in BC tissues was significantly higher than that in normal tissues. The immunohistochemical images from the HPA database showed that CLIC3 protein was expressed in the nuclear and cytoplasmic or cytomembrane and no all the tumors are positive for CLIC3 (Fig. S1).

Correlation between CLIC3 mRNA expression and the clinical characteristics of BC patients

To study the correlation between CLIC3 mRNA expression and the clinical characteristics of BC patients, clinical data of 165 patients with BC were collected from the GSE13507 dataset. The results showed that CLIC3 mRNA expression was significantly correlated with age (P = 0.021, Table 1) and grade (P = 0.045, Table 1).

Relationship between CLIC3 mRNA expression and the survival time of patients with BC

To elucidate the relationship between CLIC3 and prognosis in patients with BC, overall survival in BC patients was analyzed with the GEPIA database. The results showed that the overall survival time of patients with high CLIC3 mRNA expression was significantly shorter than that of patients with low CLIC3 mRNA expression (Fig. 3A, P = 0.044). The relationship between the expression of CLIC3 mRNA and tumor-specific survival in 165 BC patients was analyzed. The results showed that patients with high CLIC3 mRNA expression had shorter tumor-specific survival time (Fig. 3B, P = 0.042). These results indicated that CLIC3 mRNA expression may be prognostic factor for poor outcome in patients with BC.

Cox univariate and multivariate analyses of overall survival and cancer-specific survival in BC patients

Cox univariate analysis showed that invasiveness, N stage and systemic chemo were associated with overall survival (Table 2); CLIC3 mRNA expression, gender and invasiveness were associated with tumor-specific survival (Table 3). The multivariate analysis showed that invasiveness, N stage, and M stage were associated with overall survival (Table 2); CLIC3 mRNA expression, gender, invasiveness, and N stage were associated with tumor-specific survival (Table 3). These results indicated that CLIC3 was an independent factor influencing the tumor-specific survival of BC patients.

GO and KEGG analyses of the genes coexpressed with CLIC3

A total of 2,976 genes with at least two independent gene probes coexpressed with CLIC3 were extracted from the MEM database (P < 0.05). There were 2,550 genes with a Spearman correlation coefficient greater than 0.2 in cBioPortal (P < 0.0001). The coexpressed genes obtained from the above databases were intersected in the Venn diagram, and 564 genes were obtained for further analysis (Fig. 4).

To explore the possible mechanisms in which CLIC3 may be involved in BC, a cluster analysis of the genes coexpressed with CLIC3 was performed to predict the function of CLIC3. A total of 564 coexpressed genes were analyzed by GO and KEGG. GO analysis revealed that these coexpressed genes mainly enriched in the biological processes of cell adhesion, the positive regulation of cell migration, Rho signaling transduction, the protein kinase B signaling pathway, cell migration and protein transport (Fig. 5A). The coexpressed genes mainly enriched in the cellular components of focal adhesion, intercellular adhesion and intercellular junction (Fig. 5B). The coexpressed genes mainly enriched in the molecular functions of integrin binding, protein binding and receptor binding (Fig. 5C). The coexpressed genes mainly enriched in focal adhesion, the extracellular matrix (ECM) receptor interaction, the P53 signaling pathway and the TNF signaling pathway according to KEGG analysis (Fig. 5D).

Signal transduction pathways associated with CLIC3 mRNA expression

To further study CLIC3-associated signaling pathways in BC, GSEA was performed. The results showed that the cell cycle, focal adhesion, the ECM receptor interaction and the P53 signaling pathway were significantly enriched in the high CLIC3 mRNA expression phenotype (Figs. 6A–6D, Table 4). These results suggested that CLIC3 may be significantly associated with cell cycle, focal adhesion, the ECM receptor interaction and the P53 signaling pathway. The GSEA results provide a good basis for understanding how CLIC3 plays a role in BC.

Validation of CLIC3 mRNA expression in BC by qRT-PCR

To further verify the expression of CLIC3 mRNA in BC, 11 pairs of BC tissues and adjacent normal tissues were selected for detection by qRT-PCR. The results showed that the expression level of CLIC3 mRNA in BC tissues was significantly higher than that in normal tissues (Fig. 7, P = 0.0086).