Transcriptome analysis reveals the link between lncRNA-mRNA co-expression network and tumor immune microenvironment and overall survival in head and neck squamous cell carcinoma

• Main Authors: Zhaoming Zhong, Min Hong, Xiao Chen, Yan Xi, Yuanyuan Xu, Deyu Kong, Jun Deng, Yun Li, Rui Hu, Chuanzheng Sun, Jin Liang
• Format: Article
• Language: English
• Published: BMC 2020-03-01
• Series: BMC Medical Genomics
• Subjects: HNSCC; lncRNA; Gene expression; Tumor microenvironment; Immune
• Online Access: http://link.springer.com/article/10.1186/s12920-020-0707-0

Abstract Background As the sixth most common cancer worldwide, head and neck squamous cell carcinoma (HNSCC) develops visceral metastases during the advanced stage of the disease and exhibits a low five-year survival rate. The importance of tumor microenvironment (TME) in tumor initiation and metastasis is widely recognized. In addition, accumulating evidence indicates that long non-coding RNA (lncRNA) is involved in crosstalk between TME and tumor cells. However, the lncRNA-associated regulators modulating the HNSCC microenvironment and progression remain largely unknown. Methods The publicly available transcriptome data and matched clinical HNSCC data were collected from The Cancer Genome Atlas (TCGA). Immune scores (ISs) and stromal scores (SSs) of HNSCC TME were calculated using ESTIMATE algorithm. Weighted gene co-expression network analysis (WGCNA) was conducted to determine the co-expressed lncRNAs and protein-coding mRNAs. Results Results showed that the high IS HNSCC male patient subgroup exhibited improved survival. Additionally, we identified 169 lncRNAs and 825 protein-coding mRNAs that were differentially expressed in high IS HNSCC samples, with the up-regulated mRNAs displaying enrichment in immune-related biological processes. Notably, we identified a high co-expression lncRNA-mRNA module (i.e., purple module) that showed strong correlation with ISs. This module contained 79 lncRNAs and 442 mRNAs, including 26 lncRNAs and 215 mRNAs showing association between expression and male HNSCC survival. Consistently, 207 of the 215 mRNAs were up-regulated in high IS HNSCC group and were enriched in immune-related signaling pathways. Based on bioinformatics analyses and previous functional assays, certain lncRNAs (e.g., AL365361.1 and PCED1B-AS1) in the purple module likely contributed to the modification of tumor immune microenvironment (TIME) in the high IS HNSCC patients, achieved by regulating transcription of abundant immune-related genes (e.g., CCR7 and TLR8). Conclusions In summary, we ascertained a HNSCC male patient subgroup that displayed high ISs and good survival probability. We identified hundreds of genes with specific expression patterns in this HNSCC subgroup as well as a highly co-expressed lncRNA-mRNA module with great potential for the modulation of TIME of HNSCC. Our study provides evidence of a link between the lncRNA-associated gene network, TIME, and HNSCC progression, and highlights potential therapeutic targets for this disease.