Identification of novel ceRNA networks associated with PD- L1 in head and neck squamous cell carcinoma based on whole-transcriptome sequencing

Abstract Objective Head and neck squamous cell carcinoma (HNSCC) is a common and fatal cancer type worldwide. Competing endogenous RNA (ceRNA) plays an important role in tumor development and progression through circular RNAs (circRNAs). Therefore, in this study, we attempted to explore the mechanisms by which circRNA/miRNA/mRNA ceRNA networks regulate head and neck squamous cell carcinoma HNSCC. Methods The biopsy samples from patients with HNSCC were obtained intra-operatively before any therapeutic intervention. The expression profiles of circRNAs, miRNAs, and mRNAs were performed using whole-transcriptome resequencing. Then, significantly differentially expressed circRNAs, miRNAs and mRNAs were screened out. The circRNA/miRNA/mRNA ceRNA networks were constructed based on the predicted circRNA–miRNA interactions and miRNA–mRNA interactions. After that, Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses were performed to elucidate the possible functions of mRNAs contained in the ceRNA network. Furthermore, the hub network was screened among the key pathways of enrichment analysis. Finally, the expression of RNAs in hub network were verified by QRT-PCR and the association between them was revealed by Spearman correlation analysis. Results A total of 421 circRNAs, 112 miRNAs, and 1074 mRNAs with differential expression were detected. Among these, the top 9 circRNAs, 28 miRNAs, and 334 mRNAs were screened to construct a ceRNA network. The KEGG signal pathway and GO enrichment analysis of 334 mRNAs showed that cell adhesion molecules (CAMs), amino acid metabolism and other related pathways, biological processes such as extracellular matrix histogenesis were significantly enriched. Among them, CD274 and other genes were mainly enriched in CAMs pathway. Ultimately, a subnetwork including hsa_circ_0044507, hsa_circ_0044517, hsa_circ_0026774, hsa-miR-4446-3p, and PD-L1 (CD274) was screened out. QRT-PCR validated that the expression of hsa_circ_0044507, hsa_circ_0044517, hsa_circ_0026774, and PD-L1 were significantly increased, and hsa-miR-4446-3p were expressed significantly less in tumor tissue than in adjacent tissue. Spearman correlation showed that the expression of hsa_circ_0044507, hsa_circ_0044517, hsa_circ_0026774 were negatively correlated with hsa-miR-4446-3p, and positively correlated with PD-L1. Conclusion CeRNA network including hsa_circ_0044507, hsa_circ_0044517, hsa_circ_0026774, hsa-miR-4446-3p, and PD-L1 may be key regulators for HNSCC, and may be potential targets for the pathogenesis and treatment development of HNSCC.