Identification of A Novel Biomarker Using Microarray Analysis and Investigation of Its Molecular Mechanisms in Esophageal Squamous Cell Carcinoma: A Translational Cancer Study

Background: The incidence of esophageal cancer has increased year by year to become one of the ten most common and lethal cancers in Taiwanese males. Most patients are diagnosed with initial presentation of dysphagia and therefore have poor prognosis due to the locally advanced disease. Furthermore, the majority of patients have a histology of squamous cell carcinoma (SCC), which quite different in pathogenesis, risk factors, outcomes, and treatment responses, from the western countries, where adenocarcinoma of the lower esophagus and esophagogastric junction is predominant. For patients with locally advanced esophageal cancer, prospective randomized trials have shown that combined modality therapy, including chemotherapy, radiotherapy, and radical esophagectomy, provides the best treatment results. However, median progression free survival remains unsatisfactorily around 20 to 25 months. Concurrent chemoradiotherapy (CCRT), either as neoadjuvant or definitive therapy, improves the overall survival in patients with esophageal SCC (ESCC). It has been recognized that pathological complete response to CCRT is the most important prognostic factor. At present, there is no reliable factor which predicts response to chemoradiation, and thus enables the individualization of treatment strategies. Furthermore, there is no molecular target therapy to improve outcomes for ESCC. In the present study, we utilized microarray analysis to identify novel biomarkers in association with outcomes for ESCC and investigated underlying mechanisms in tumorigenicity. Our aim in conducting the present translational research was to discover biomarkers with the potential to predict outcomes and which may be used as therapeutic targets, to help improve the prognosis of ESCC. Methods: In the first part of this study, we collected blood samples in ESCC patients before and after neoadjuvant CCRT and extracted the messenger RNA. We performed microarray experiments and identified candidate circulating biomarkers using bioinformatics and biostatistics tools. We validated our findings using reverse-transcriptase polymerase chain reaction and proximal ligation assay to detect serum protein levels in an independent cohort. We also investigated whether the identified biomarker was overexpressed in cancerous tissues using immunohistochemistry (IHC) analysis, and in ESCC cell lines using western blot. In the second part, we used established ESCC cell lines to study the biological properties of the candidate gene by manipulating its expression levels. We performed in vitro cell assay and in vivo subcutaneous xenograft animal models. We also used a gene expression microarray analysis to identify candidate genes associated with our gene of interest. Results: In the first part, we showed that circulating expression profiles were significantly altered before and after CCRT. We therefore hypothesized that changes in gene expression may be associated with treatment outcome and successfully identified a novel candidate biomarker, human FAM84B. Expression levels of FAM84B were significantly down-regulated after CCRT and its fold change was predictive of pathological response to CCRT. Furthermore, we demonstrated that the human FAM84B protein was overexpressed in ESCC cancerous tissue and cell lines but not in the normal esophageal epithelium and epithelial cells. In the second part, we discovered that FAM84B knockdown by short hairpin RNA impaired the clonogenic survival and migration of ESCC cell lines, while FAM84B knock-in promoted clonogenic survival. Inhibition of FAM84B in ectopic xenografts also delayed tumor growth with reduced Ki-67 proliferation index, mitotic index, and promoted cell differentiation. We showed that the tumorigenicity of FAM84B may be associated with expression of the squamous cell differentiation protein KRT15. In addition, microarray analysis revealed that CCND2 expression was significantly associated with FAM84B expression. We demonstrated that inhibition of FAM84B delayed G1/S phase progression in the cell cycle by downregulating CCND2. In the last part, we showed that FAM84B was associated with sensitivity to ionizing radiation and cytotoxic cisplatin in ESCC cell lines. Discussion Unlike similar research, we identified a novel circulating biomarker with biological properties using high-throughput technology. Human FAM84B is located at chromosome 8q24.21, which is known to associate with different cancer types due to copy number alterations. The function of FAM84B is understudied due to its proximity to the well-known oncogene, MYC. However, we and other independent investigators have shown that FAM84B is associated with pathogenesis in ESCC in the Han population. While our earlier findings showed the oncogenic properties of FAM84B in ESCC, its molecular mechanisms required further investigation. In our preliminary research, we found that copy number variations in FAM84B may have a more important role in ESCC. Furthermore, inhibition of FAM84B in 8q24.21 amplified cancers, such as colorectal cancer cells, and significantly impaired cell proliferation. Our finding is only the beginning, and we believe it worth further investigation in both clinical specimens and laboratory work to study the role of FAM84B in cancer biology.