Abstract
BACKGROUND: Numerous studies were performed to illuminate mechanisms of tumorigenesis and metastases from gene expression profiles of Head and Neck Squamous Cell Carcinoma (HNSCC). The objective of this review is to conduct a network-based meta-analysis to identify the underlying biological signatures of the HNSCC transcriptome. METHODS AND FINDINGS: We included 63 HNSCC transcriptomic studies into three specific categories of comparisons: Pre, premalignant lesions v.s. normal; TvN, primary tumors v.s. normal; and Meta, metastatic or invasive v.s. primary tumors. Reported genes extracted from the literature were systematically analyzed. Participation of differential gene activities across three progressive stages deciphered the evolving nature of HNSCC. In total, 1442 genes were verified, i.e. reported at least twice, with ECM1, EMP1, CXCL10 and POSTN shown to be highly reported across all three stages. Knowledge-based networks of the HNSCC transcriptome were constructed, demonstrating integrin signaling and antigen presentation pathways as highly enriched. Notably, functional estimates derived from topological characteristics of integrin signaling networks identified such important genes as ITGA3 and ITGA5, which were supported by findings of invasiveness in vitro. Moreover, we computed genome-wide probabilities of reporting differential gene activities for the Pre, TvN, and Meta stages, respectively. Results highlighted chromosomal regions of 6p21, 19p13 and 19q13, where genomic alterations were shown to be correlated with the nodal status of HNSCC. CONCLUSIONS: By means of a systems-biology approach via network-based meta-analyses, we provided a deeper insight into the evolving nature of the HNSCC transcriptome. Enriched canonical signaling pathways, hot-spots of transcriptional profiles across the genome, as well as topologically significant genes derived from network analyses were highlighted for each of the three progressive stages, Pre, TvN, and Meta, respectively.