Methylation-related differentially expressed genes as potential prognostic biomarkers for cervical cancer

Abstract

Aim: To discover novel methylation-related differentially expressed genes (MRDEGs) for cervical cancer, with a focus on their potential for early diagnosis and prognostic assessment.

Materials & methods: We integrated data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. TCGA-MRDEGs were identified by analyzing differentially methylated genes (DMGs) and their correlation with gene expression. We examined GEO datasets GSE39001, GSE9750, and GSE46306 for GEO-MRDEGs. Overlapping MRDEGs were subjected to overall survival (OS) analysis to identify prognostic markers. The expression and methylation levels of these genes were validated in a total of 30 tissue samples, comprising 20 from cervical cancer patients and 10 from normal cervical tissues, using qRT-PCR and MassARRAY EpiTYPER Assay.

Results: A total of 314 TCGA-MRDEGs and 40 GEO-MRDEGs were identified. Intersection analysis yielded 10 overlapping MRDEGs. Notably, NOVA1, GSTM5, TRHDE, and CXCL12 were found to have reduced expression and increased methylation in cervical cancer, which correlated with poor prognosis. The methylation status and expression levels of these genes were confirmed in tissue specimens.

Conclusion: We identified four MRDEGs as potential prognostic biomarkers for cervical cancer. Their clinical utility is highlighted, but further validation in larger cohorts is required to establish their clinical significance.

Keywords: Biomarker; Cervical cancer; DNA methylation; Differentially expressed genes; Prognosis.

Fig. 1

Overview and flowchart of the study.

Fig. 2

Venn diagrams of the DMGs mining process. (A) Venn diagram of DEGs, DMGs and methylation-expression correlated genes in the TCGA database. “Spearman” denotes genes obtained through Spearman analysis (P < 0.001), and “Pearson” indicates those gained via Pearson analysis (P < 0.001). (B) Venn diagram of DMGs and DEGs between GSE39001, GSE9750, and GSE46306. (C) Intersection of MRDEGs between TCGA and GEO databases.

Fig. 3

Overall Survival (OS) analysis for four prognostic MRDEGs. (A–D) Kaplan-Meier curves for NOVA1, GSTM5, TRHDE, and CXCL12.

Fig. 4

Methylation status (A–D) and mRNA expression level (E–H) of four prognostic MRDEGs in DNMIVD database. P values are indicated on each figure.

Fig. 5

The methylation status (A–D), mRNA expression level (E–H) and their respective pan-cancer analysis (I–L) of four prognostic MRDEGs in normal tissues and cancer tissues in UALCAN database.

Fig. 6

The mRNA expression of four prognostic MRDEGs in TCGA and GEO databases. The mRNA expression of four MRDEGs in GEPIA (based on TCGA, A-D), GSE39001 (E–H) and GSE9750 (I–L). “*” represents a significant difference in mRNA expression between normal tissues and tumor tissues (P < 0.05).

Fig. 7

Verification of mRNA and methylation expression levels of four prognostic MRDEGs. (A–D) The qRT-PCR results of four MRDEGs in normal specimens and cervical cancer specimens. “****” represents a significant difference in mRNA expression between normal tissues and tumor tissues (P < 0.0001). (E–H) Comparison of total methylation levels in the promoter region of MRDEG between tumor and normal tissues. P-values are indicated on each figure.

Fig. 8

Comparative analysis of CpG site methylation levels in MRDEG between tumor and normal tissues. Symbols “*”, “**”, and “***” denote statistical significance with P-values less than 0.05, 0.01, and 0.001, respectively.