Abstract
Esophageal squamous cell carcinoma (ESCC) has a dismal prognosis because of atypical early symptoms and heterogeneous therapeutic responses. 5-methylcytosine (m(5)C) modification plays an important role in the onset and development of many tumors and is widespread in long non-coding RNA (lncRNA) transcripts. However, the functions of m(5)C and lncRNAs in ESCC have not been completely elucidated. Herein, this study aimed to explore the role of m(5)C-related lncRNAs in ESCC. The RNA-seq transcriptome profiles and clinical information were downloaded from the TCGA-ESCC database. Pearson analysis was used to identify m(5)C-related lncRNAs. Then we established the m(5)C-related lncRNAs prognostic signature (m(5)C-LPS) using univariate Cox and least absolute shrinkage and selection operator (LASSO) regression analysis. Then, the prognostic value of m(5)C-LPS was evaluated internally and externally using the TCGA-ESCC and GSE53622 databases through multiple methods. We also detected the expression of these lncRNAs in ESCC cell lines and patient tissues. Fluorescence in situ hybridization (FISH) was used to detect the prognostic value of specific lncRNA. In addition, clinical parameters, immune status, genomic variants, oncogenic pathways, enrichment pathways, and therapeutic response features associated with m(5)C-LPS were explored using bioinformatics methods. We constructed and validated a prognostic signature based on 9 m(5)C-related lncRNAs (AC002091.2, AC009275.1, CAHM, LINC02057.1, AC0006329.1, AC037459.3, AC064807.1, ATP2B1-AS1, and UBAC2-AS1). The quantitative real-time polymerase chain reaction (qRT-PCR) revealed that most lncRNAs were upregulated in ESCC cell lines and patient tissues. And AC002091.2 was validated to have significant prognostic value in ESCC patients. A composite nomogram was generated to facilitate clinical practice by integrating this signature with the N stage. Besides, patients in the low-risk group were characterized by good clinical outcomes, favorable immune status, and low oncogenic alteration. Function enrichment analysis indicated that the risk score was associated with mRNA splicing, ncRNA processing, and DNA damage repair response. At the same time, we found significant differences in the responses to chemoradiotherapy between the two groups, proving the value of m(5)C-LPS in treatment decision-making in ESCC. This study established a novel prognostic signature based on 9 m(5)C-related lncRNAs, which is a promising biomarker for predicting clinical outcomes and therapeutic response in ESCC.