Abstract
The incidence of cholangiocarcinoma (CCA) has risen in many countries, but there is still no appropriate screening and treatment available. The growing number of microarray data published todays can be a powerful resource for the discovery of biomarkers to tackle challenges in the management of CCA. This study analyzed multiple microarray datasets to identify the common transcriptional networks in CCA and select a possible biomarker for functional study in CCA cell lines. A systematic searching identified 4 microarray datasets from Gene Expression Omnibus (GEO) repository and PubMed articles. Differential expression analysis between tumor and normal tissues was performed in each dataset. In order to characterize the common expression pattern, differentially expressed genes (DEGs) from all datasets were combined and visualized by hierarchical clustering and heatmap. Gene enrichment analysis performed in each cluster revealed that over-expressed DEGs were enriched in cell cycle, cell migration and response to cytokines while under-expressed DEGs were enriched in metabolic processes such as oxidation-reduction, lipid, and drug. To explain tumor characteristics, genes enriched in cell migration and response to cytokines were further investigated. Among these genes, CCL20 was selected for functional study because its role has never been studied in CCA. Moreover, its signaling may be regulated by disrupting its only receptor, CCR6. Treatment with recombinant CCL20 induced higher cell migration and increased expression of N-cad. In contrast, knockdown of CCR6 by siRNA reduced cell migration ability and decreased N-cadherin level. Altogether, these results suggested the contribution of CCL20/CCR6 signaling in cell migration through epithelial-mesenchymal transition process. Thus, CCL20/CCR6 signaling might be a target for the management of CCA.