Identification and Analysis of An Epigenetically Regulated Five-lncRNA Signature Associated With Outcome and Chemotherapy Response in Ovarian Cancer.

Hao Yang,Xiaomei Li,Meiling Zhang,Ning Ning,Di Wu,Lin Gao,Yan Wang

doi:10.3389/fcell.2021.644940

Abstract

The deregulation of long non-coding RNAs (lncRNAs) by epigenetic alterations has been implicated in cancer initiation and progression. However, the epigenetically regulated lncRNAs and their association with clinical outcome and therapeutic response in ovarian cancer (OV) remain poorly investigated. This study performed an integrative analysis of DNA methylation data and transcriptome data and identified 419 lncRNAs as potential epigenetically regulated lncRNAs. Using machine-learning and multivariate Cox regression analysis methods, we identified and developed an epigenetically regulated lncRNA expression signature (EpiLncRNASig) consisting of five lncRNAs from the list of 17 epigenetically regulated lncRNAs significantly associated with outcome. The EpiLncRNASig could stratify patients into high-risk groups and low-risk groups with significantly different survival and chemotherapy response in different patient cohorts. Multivariate Cox regression analyses, after adjusted by other clinical features and treatment response, demonstrated the independence of the DEpiLncSig in predicting survival. Functional analysis for relevant protein-coding genes of the DEpiLncSig indicated enrichment of known immune-related or cancer-related biological pathways. Taken together, our study not only provides a promising prognostic biomarker for predicting outcome and chemotherapy response but also will improve our understanding of lncRNA epigenetic regulation mechanisms in OV.

Highlights

Ovarian cancer (OV) is one of the most lethal gynecologic cancers and is the eighth leading cause of cancer-related deaths in women (Coburn et al, 2017; Momenimovahed et al, 2019)
To identify DEpiLncRNAs in ovarian cancer (OV), we calculated the Pearson correlation coefficient to evaluate the association between Long non-coding RNAs (lncRNAs) expression and CpG levels and identified 1,497 lncRNA-CpG pairs, including 419 lncRNAs, which were defined as DEpiLncRNAs
We first conducted the univariate analysis for 419 lncRNAs with OS and identified 17 lncRNAs that were significantly associated with OS (p < 0.01)

Summary

Introduction

Ovarian cancer (OV) is one of the most lethal gynecologic cancers and is the eighth leading cause of cancer-related deaths in women (Coburn et al, 2017; Momenimovahed et al, 2019). Increasing efforts in studying molecular omics have improved our understanding of the molecular mechanisms of OV carcinogenesis and progression and contributed to the identification and development of novel molecular biomarkers and specific therapies (Cancer Genome Atlas Research Network, 2011; Lu et al, 2014; Clifford et al, 2018; Li et al, 2020). Molecular profiles have been extensively investigated and characterized during the past years, leading to the identification of a number of dysregulated molecules associated with development, progression, recurrence, metastasis, and therapeutic response of OV (Adib et al, 2004; Barrett et al, 2015; Sallinen et al, 2019; Wang et al, 2019; Zhao H. et al, 2020). The epigenetically regulated lncRNAs and their association with clinical outcome and therapeutic response in OV remain poorly investigated

Methods

Results

Conclusion