Comprehensive analysis of ceRNA Networks in UCEC: Prognostic and therapeutic implications

Abstract

Endometrial cancer (UCEC) is the most prevalent gynecological malignancy in high-income countries, and its incidence is rising globally. Although early-stage UCEC can be treated with surgery, advanced cases have a poor prognosis, highlighting the need for effective molecular biomarkers to improve diagnosis and prognosis. In this study, we analyzed mRNA and miRNA sequencing data from UCEC tissues and adjacent non-cancerous tissues from the TCGA database. Differential expression analysis was conducted using the DESeq2 package, identifying differentially expressed lncRNAs, miRNAs, and mRNAs (DElncRNAs, DEmiRNAs, and DEmRNAs). Key molecules were screened using LASSO regression, and a ceRNA network was constructed by predicting lncRNA-miRNA and miRNA-mRNA interaction, which were visualized with Cytoscape. Functional enrichment analysis elucidated the roles and mechanisms of the network. The prognostic potential of the identified RNAs was assessed through survival and Cox regression analyses, while methylation and immune infiltration analyses explored regulatory mechanisms and immune interactions. We identified a prognostic lncRNA-miRNA-mRNA ceRNA network in UCEC, centered on the CDKN2B-AS1-hsa-miR-497-5p-IGF2BP3 axis. Survival analyses confirmed the prognostic significance of this network, with univariate Cox regression demonstrating a strong association between its aberrant expression and overall prognosis in UCEC. However, multivariate Cox regression suggested that other clinical factors may modulate this relationship. Methylation analysis revealed low methylation levels of IGF2BP3, possibly contributing to its overexpression. Furthermore, immune infiltration studies highlighted significant correlations between CDKN2B-AS1, IGF2BP3, and multiple immune cell types, suggesting that this axis regulates the tumor immune microenvironment. These findings suggest that the CDKN2B-AS1-hsa-miR-497-5p-IGF2BP3 axis is a key regulatory element in UCEC and a potential therapeutic target.

Fig 1. Differential expression analysis in UCEC.

(A) Volcano plot of all DElncRNA genes in UCEC (|logFC| > 1 and p.adj < 0.05) and heatmap of the top 15 genes, (B) Volcano plot of all DEmiRNA genes in UCEC (|logFC| > 0.5 and p.adj < 0.05) and heatmap of the top 15 genes. (C) Volcano plot of all DEmRNA genes in UCEC (|logFC| > 2 and p.adj < 0.05) and heatmap of the top 15 genes.

Fig 2. Prognostic lncRNAs, miRNAs, and mRNAs in UCEC.

(A) Venn diagram showing 160 differentially expressed and OS-associated lncRNAs in UCEC. LASSO analysis selected 47 significant lncRNAs. (B) Venn diagram illustrating 21 overlapping differentially expressed and OS-associated miRNAs in UCEC. LASSO analysis selected 13 significant miRNAs. (C) Venn diagrams showing the overlap of miRNAs predicted by the 47 lncRNAs and the 13 miRNAs and the overlap of mRNAs predicted by these 5 miRNAs with differentially expressed and OS-associated mRNAs. (D) The triple regulatory network of lncRNA-miRNA-mRNA significantly associated with prognosis in UCEC. Nodes in red indicate upregulation, while nodes in green indicate downregulation. The size of each node is positively correlated with the adjusted p value.

Fig 3. Analysis of the lncRNA-miRNA-mRNA network in UCEC.

(A) GO and KEGG pathway analysis of mRNAs in the lncRNA-miRNA-mRNA network with a significance threshold of corrected p < 0.05. (B) Hub triple regulatory network identified using the Cytoscape plugin cytoHubba with a degree of > 3, including one lncRNA, five miRNAs, and ten mRNAs. (C) Differential expression of hub triple regulatory network RNAs in UCEC versus normal tissues. (D) Protein levels of the hub triple regulatory network in UCEC tissues. Notations: ns indicates no significance; **p < 0.01, ***p < 0.001.

Fig 4. Survival and predictive analysis of hub triple regulatory network molecules in UCEC.

(A) Forest plot showing the survival analysis (OS, DSS, and PFI) of hub triple regulatory network molecules in UCEC. (B) ROC analysis of hub triple regulatory network molecules.

Fig 5. Subcellular localization and regulatory network of CDKN2B-AS1 in UCEC.

(A) Subcellular localization of CDKN2B-AS1 primarily in the cytoplasm, as identified by lncLocator and A Sankey diagram illustrates the subcellular localization of CDKN2B-AS1 in various cells according to RNALOCATE. (B) Friends’ analysis identified IGF2BP3 as a key gene from the mRNAs in the hub triple regulatory network. (C) Correlation analysis involving CDKN2B-AS1, hsa-miR-497-5p and IGF2BP3 in UCEC. (D) Conceptual model of the ceRNA network CDKN2B-AS1-hsa-miR-497-5p-IGF2BP3, depicting predicted binding sites. Overexpressed genes are marked in red, and underexpressed genes are marked in blue. (E) Chromosomal location map displaying the positions of CDKN2B-AS1, hsa-miR-497-5p, and IGF2BP3 on their respective chromosomes.

Fig 6. Prognostic analysis of CDKN2B-AS1, IGF2BP3, and hsa-miR-497-5p in UCEC.

(A) Univariate and multivariate Cox regression analyses of CDKN2B-AS1 and IGF2BP3. (B) Univariate and multivariate Cox regression analyses of hsa-miR-497-5p. (C) Comparison of prognosis among different subgroups based on the expression levels of CDKN2B-AS1, IGF2BP3, and hsa-miR-497-5p.

Fig 7. Methylation analysis of IGF2BP3 in UCEC.

(A) Comparison of the expression levels of three DNA methyltransferases (DNMT1, DNMT3A, and DNMT3B) between UCEC samples with high and low IGF2BP3 expression. (B) Analysis of IGF2BP3 methylation status in UCEC using the DiseaseMeth 2.0 tool. (C) Differences in the methylation intensity of IGF2BP3 CpGs between UCEC and normal tissue samples, and the correlation between methylation values at various CpG sites and IGF2BP3 expression in UCEC. (D) Heatmap representing the clustering of CpG methylation levels with IGF2BP3 in UCEC. (E) Forest plot illustrating the impact of different CpG site methylation levels within IGF2BP3 on survival in UCEC. Notations: ns indicates no significance; *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 8. Immune infiltration and potential therapeutic analysis.

Differential enrichment scores of 24 immune cell types and the correlation between gene expression and the proportion of 22 immune cell types in UCEC between the high and low CDKN2B-AS1 (A) and IGF2BP3 (B)expression groups. (C) Venn diagram showing immune cells affected by the expression of both CDKN2B-AS1 and IGF2BP3. (D) Sankey diagram illustrating the flow between candidate drugs, genes, and immune cells. (E) 3D structure of Cisplatin, a candidate drug for treating UCEC through its impact on the ceRNA network. Notations: ns indicates no significance; *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 9. Genetic and functional analysis of IGF2BP3 in UCEC.

(A) Correlation analysis of SNPs affecting IGF2BP3 expression with the risk of endometrial cancer. (B) Effect size of significant SNPs in endometrial cancer. (C) Correlation analysis of 35 experimentally validated IGF2BP3-binding proteins in UCEC. (D) Domain enrichment analysis of IGF2BP3-binding proteins. (E) KEGG pathway analysis based on IGF2BP3-associated genes. GO enrichment analysis for IGF2BP3-associated genes distributed across Cellular Component (F), Biological Process (G), and Molecular Function (H).