Comprehensive analysis of m6A reader YTHDF2 prognosis, immune infiltration, and related regulatory networks in hepatocellular carcinoma

Abstract

Background: N6-Methyladenosine (m6A) RNA modification is the most prevalent internal modification pattern in eukaryotic mRNAs and plays critical roles in diverse physiological and pathological processes. However, the expression of m6A regulator YTHDF2, its prognostic value, its biological function, its correlation with tumor microenvironment (TME) immune infiltrates, and related regulatory networks in hepatocellular carcinoma (HCC) remain determined.

Methods: TCGA, GTEx, and GEO databases were used to investigate the expression profile of YTHDF2 in HCC. We performed differentially expressed genes (DEGs) analysis and constructed a PPI network to explore the biological processes of YTHDF2 in HCC. Kaplan-Meier curves and Cox regression analysis were used to assess the prognostic value of YTHDF2 and then a clinical prognostic nomogram was constructed. Additionally, ssGSEA was performed to assess the correlation between YTHDF2 and immune infiltration levels. The TISIDB database was applied to explore the expression of YTHDF2 in immune and molecular subtypes of HCC. GSEA identifies the YTHDF2-related signaling pathways. Finally, we utilized miRNet and starBase database to construct regulatory networks for HCC based on lncRNA-miRNA and miRNA-YTHDF2 interactions.

Results: YTHDF2 was significantly upregulated in HCC tumor tissues compared with the adjacent normal tissues. HCC patients in the high YTHDF2 expression group had poorer survival. Multivariate Cox analysis suggested that YTHDF2 may be a new independent prognostic indicator for HCC patients, with the prognostic nomogram exhibiting satisfactory results. YTHDF2 expression was significantly correlated with TME immune cell-infiltrating characteristics. Strong correlations were also shown in immune subtypes, molecular subtypes and immune checkpoints. Further analysis revealed that the combination of YTHDF2 expression and immune cell score was considerably associated with survival outcome in HCC patients. GESA analysis demonstrated that high YTHDF2 expression is associated with multiple biological processes and oncogenic pathways. Moreover, 14 possible regulatory networks were constructed, which are associated with HCC progression.

Conclusion: Our findings revealed that YTHDF2 may serve as a promising prognostic biomarker for HCC and may regulate the tumor immune microenvironment to provide effective therapeutic strategies.

Keywords: Epitranscriptomics; Hepatocellular carcinoma; Immune infiltration; Prognostic signature; m6A methylation.

Fig. 1

Expression level of YTHDF2 in HCC. (A) The expression of YTHDF2 in pan-cancer data from TCGA and GTEx database. (B) The expression levels of YTHDF2 in all HCC samples from TCGA database. (C) The expression levels of YTHDF2 in paired tumor and adjacent normal tissues in HCC from TCGA database. (D, E and F) showed the expression levels of YTHDF2 in all HCC samples from GSE14520, GSE45267and GSE64041. ns P ≥ 0.05; *P < 0.05; **P < 0.01 and ***P < 0.001.

Fig. 2

Diagnostic and prognostic value of YTHDF2 in HCC. (A) The expression levels of YTHDF2 in normal liver and HCC were visualized by IHC in HPA. (B) Diagnostic ROC curve of YTHDF2 in HCC. (C) Time-dependent ROC curve of YTHDF2 in HCC. (D) OS curves of YTHDF2. (E) PFI curves of YTHDF2.

Fig. 3

DEGs analysis, enrichment pathway analysis and construction of PPI network. (A) The volcano plot of DEGs. (B) PPI network comprising YTHDF2 and its most closely associated genes. Functional enrichment analysis in (C) GO Biological Process, (D) GO Molecular Function, (E) GO Cellular Component, (F) KEGG.

Fig. 4

Association between the prognosis value of YTHDF2 and clinicopathologic characteristics. Kaplan–Meier survival analysis showed that high expression of YTHDF2 was associated with worse OS in (A) ≤ 60 years old, (B) male, (C) Asian, (D) G3 and G4, (E) stage Ⅲ and Ⅳ, (F) stage T2, (G) stage T3 and T4, (H) stage N0, and (I) stage M0.

Fig. 5

Construction and validation of nomogram based on YTHDF2 expression. (A) The nomogram for predicting survival probability at 1-, 3- and 5-years in HCC patients. (B–D) Calibration curves of the nomogram. (E–G) Decision curve analysis of the nomogram.

Fig. 6

Correlation between YTHDF2 expression and immune infiltration. (A–B) Correlation between 24 immune cell infiltration. (C) The immune cell infiltration in the high- and low-expression groups of YTHDF2.

Fig. 7

Impact of immune cell infiltration on prognosis in HCC patients. Kaplan–Meier survival analysis using combinations YTHDF2 expression and Immune cells score, including (A) B cells, (B) CD8 T cells, (C) cytotoxic cells, (D) eosinophils cells, (E) iDC cells, (F) macrophages cells, (G) mast cells, (H) neutrophils cells, (I) NK CD56bright cells, (J) NK CD56dim cells, (K) NK cells, (L) pDC cells, (M) T cells, (N) T helper cells, (O) Tcm cells, (P) Tem cells, (Q) Tgd cells, (R) Th17 cells, and (S) TReg cells.

Fig. 8

Expression of YTHDF2 in immune and molecular subtypes and its correlation with immune checkpoint genes. (A) The relationship between YTHDF2 expression and HCC immune subtypes and molecular subtypes. (B) The heat map of the 50 immune checkpoint genes correlated to YTHDF2. *P < 0.05; **P < 0.01 and ***P < 0.001.

Fig. 9

Enrichment plots from GSEA. (A) FCGR activation, (B) FCGR3a mediated IL10 synthesis, (C) CD22 mediated BCR regulation, (D) role of phospholipids in phagocytosis, (E) retinoblastoma gene in cancer, (F) immunoglobulin complex, (G) homophilic cell adhesion via plasma membrane adhesion molecules synapse assembly, (H) intrinsic component of synaptic membrane, and (I) appendage development. ES, Enrichment score; FDR, false discovery rate; NES, normalized ES.

Fig. 10

Prediction of the upstream lncRNAs potentially binding to YTHDF2 and lncRNA-miRNA-m6A regulator (YTHDF2) regulatory networks construction in HCC. (A) The correlation between miRNA and the target lncRNA. (B) The Sankey diagram shows the lncRNA-miRNA-m6A regulator (YTHDF2) regulatory network based on lncRNA-miRNA and miRNA-YTHDF2 co-expression patterns.