Identifcation of the FGF family as therapeutic targets and prognostic biomarkers in the microenvironment of head and neck squamous cell carcinoma.

Abstract

Background: Almost 90% of head and neck malignancies are malignant squamous cell cancers, making it the sixth most common malignancy in the developing countries, with an overall five-year overall survival rate about 40%-50%. Early diagnosis and treatment can bring a better prognosis. Fibroblast growth factor (FGF) is an important polypeptide in vivo. Studies have found that FGF signal has carcinogenic potential and participates in a variety of carcinogenic behaviors. Some experiments have proved that FGF signal has the function of tumor inhibition in some cases, and the role of FGF signalling in tissue repair and homeostasis suggest a role for FGF in targeted therapy and prognosis. However, its manifestation and predictive role in HNSC have not been clearly defined.

Methods: Genome-wide expression analysis of Oncomine evaluated the evaluation of FGF family expression in HNSC. Expression analysis and HNSC data set were used to obtain FGF family expression data and T statistic was applied for analysis. The differential mRNA expression levels in tumor versus normal tissues, as well as the correlation with pathological staging and prognosis, were examined using the GEPIA single-gene analysis tool for the FGF family.FGF family altered CO expression and network modules were obtained from cBioportal and analyzed in 520 HNSC samples.Pro-protein interaction (PPI) flow network is performed on the differentially ordered FGF clusters using STRING, Gene Operating System (GO) domain domain enrichment as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis are performed on the FGF cluster and its neighbouring genes using DAVID6.8, key transcriptional factors (TF) of FGF family was analyzed by TRRUST, correlation between FGF family level and autoimmune cell migration was evaluated by TIMER, and biological analysis of FGF family kinase target enrichment was performed using LinkInterpreter.

Results: Only the expression of FGF6 in HNSC was down-regulated in all FGF family(FC=2),Transcriptional level of FGF1, FGF2, FGF5, FGF7-14, FGF17-19, FGF21 and FGF22 was upregulated in HNSC .In terms of the relative level of FGF family in HNSC, the greatest amount of FGF11. In different pathological stages of HNSC, the expression of FGF was meaningless (P>0.05), and FGF3-6, FGF8-10, FGF14, FGF16, FGF17, FGF19 -21, FGF23 showed no significant difference in different HNSC stages. Low expression of FGF5 and high expression of FGF22 had low overall survival(OS) rate of HNSC(P =0.012, P =0.0015). In addition, enrichment analysis of FGF family in HNSC showed that it was highly abundant in PI3K-Akt signaling pathway, MAPK and rasper pathway. Our data showed that ATF4, STAT, RELA, NFKB1 are key transcription target of the FGF family, NLK, LOCK1, LYN, ZAP70, MAP2K3, RPS6KA4, AURKB, ATR, ROCK1, MYLK2, CAMK2A, EGFR, MAPK3, MAP3K8, SYK, LCK, HCK, PKN2, RPS6KA1, BUB1, CDK5, ITK, FYN, TBK1, ATM, CDK2, PTK2 are kinase targets of the FGF family. We identified a relationship between the modulation of FGF expression and cellular infiltration, such as B lymphocytes, CD4+ T cells and macrophages dendritic cells.

Conclusions: Our data may shed new light on the choice of immunotherapeutic targets and predictive biomarkers in HNSC.