Prognostic Role and Potential Mechanisms of N6-methyladenosine-related Long Noncoding RNAs in Hepatocellular Carcinoma

Abstract

Background and aims: Numerous studies have explored the important role of N6-methyladenosine (m⁶A) in cancer. Nonetheless, the interaction between m⁶A and long noncoding RNAs (lncRNAs) is poorly investigated. Herein, we systematically analyzed the role and prognostic value of m⁶A-related lncRNAs in hepatocellular carcinoma (HCC).

Methods: The m⁶A-related lncRNAs were identified based on the correlation coefficients with m⁶A-related genes in HCC from The Cancer Genome Atlas. Subsequently, a novel risk score model was determined using the least absolute shrinkage and selection operator Cox regression analyses. Univariate and multivariate Cox analyses were used to identify independent prognostic factors for overall survival (OS) of HCC; thereafter, a prognostic nomogram was constructed.

Results: A total of 259 lncRNAs showed significant correlations with m⁶A in HCC, while 29 lncRNAs had prognostic significance. Further, six critical m⁶A-related lncRNAs (NRAV, SNHG3, KDM4A-AS1, AC074117.1, AC025176.1, and AL031985.3) were screened out to construct a novel risk score model which classified HCC patients into high- and low-risk groups. Survival analyses revealed that patients in the high-risk group exhibited worse OS, both in the training and validation groups. The risk score was also identified as an independent prognostic factor of OS, and a nomogram was established and verified with superior prediction capacity. Besides, the risk score significantly correlated with the expression of immune checkpoint genes and immune subtypes.

Conclusions: These findings indicated the significant role of m⁶A-related lncRNAs in HCC and the potential application of the novel risk score model for prognostic prediction.

Keywords: Hepatocellular carcinoma; Immune checkpoints; Long noncoding RNAs; N6-methyladenosine; Prognosis.

Fig. 1. Flowchart of the data analysis procedures.

Fig. 2. Identification of m⁶A-related lncRNAs with prognostic significance in HCC.

(A) Bar plot for exhibition of the m⁶A-DEGs expression in HCC tumors and normal controls. (B) The correlation plot of the m⁶A-related genes (n=24) with the lncRNAs (n=259). (C) Univariate analysis for m⁶A-related lncRNAs with prognostic significance in HCC (n=29). (D) Heatmap of the expression of m⁶A-related lncRNAs in HCC tumors and normal controls. (E) The correlation among these m⁶A-related lncRNAs in HCC. *p<0.05. HCC, hepatocellular carcinoma; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; m⁶A-DEGs, differentially expressed m⁶A-related genes.

Fig. 3. Consensus clustering of HCC based on the m⁶A-related lncRNAs.

(A–C) The CDF, relative changes in area under the CDF curves, and tracking plots show with the index from 2 to 9. (D) The distribution of different clusters with the clustering index was 2. (E) Survival curves in different clusters. (F) The heatmap with visualization of the expression of 29 m⁶A-related lncRNAs in HCC patients and the correlation of the risk score with other clinical factors. (G) Pathway analyses by GSEA in different clusters. (H) The expression profiles of immune checkpoint genes (PD-1, PD-L1, CTLA-4, LAG3, TIM3, and TIGIT in the different clusters. ns, p>0.05, *p<0.05, **p<0.01, ***p<0.001. CDF, cumulative distribution function; CTLA-4, cytotoxic T lymphocyte-associated antigen-4; GSEA, gene set enrichment analysis; HCC, hepatocellular carcinoma; LAG3, lymphocyte-activation gene 3; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; PD-1, programmed cell death 1; PD-L1, PD-1 ligand 1; TIGIT, T cell immunoreceptor with immunoglobulin and ITIM domain; TIM3, T-cell immunoglobulin and mucin domain 3.

Fig. 4. Identification of critical m⁶A-related lncRNAs for the novel risk score model and prognostic evaluation in HCC.

(A–B) Screening of the critical m⁶A-related lncRNAs by LASSO Cox regression. (C–D) Survival analyses of high- and low-risk groups in the training and validation cohorts. (E–F) Time-dependent ROC curves based on the risk score for 1-, 3-, and 5-year OS of HCC in the training and validation groups. AUC, area under the curve; HCC, hepatocellular carcinoma; LASSO, least absolute shrinkage and selection operator; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; OS, overall survival; ROC, receiver operating characteristic; TCGA, The Cancer Genome Atlas.

Fig. 5. Distribution of the risk score, survival status, and expression of the six m⁶A-related lncRNAs in the training and validation groups of HCC.

(A–B) Proportions of death in the high- and low-risk groups of the training and validation cohorts. (C–D) Risk maps of the risk score, survival status, and expression of lncRNAs in the training and validation cohorts. HCC, hepatocellular carcinoma; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; TCGA, The Cancer Genome Atlas.

Fig. 6. Evaluation of the independent prognostic significance of the novel risk score model based on m⁶A-related lncRNAs in HCC.

(A–B) Univariate and multivariate Cox analyses with risk score and other clinical factors for OS of HCC in the training group. (C–D) Univariate and multivariate Cox analyses with risk score and other clinical factors for OS of HCC in the validation group. (E–L) Survival analyses of the risk score in different subgroups of various clinical factors: age (≤/>60 years), sex (female/male), pathological grade (1–2/3–4), and clinical stage (I–II/III–IV). HCC, hepatocellular carcinoma; HR, hazard ratio; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine;

Fig. 7. Clinical correlation of the risk score model and the prognostic nomogram for OS of HCC established based on the risk score model.

(A) Heatmap with visualization of the correlations between risk score and other clinical characteristics and the expression of the six m⁶A-related lncRNAs in HCC. (B) Distribution of risk score in subgroups of different pathological grades, clinical stages, and clusters. (C) Prognostic nomogram established based on the risk score and clinical stage for prediction of 1-, 3-, and 5-year OS of HCC. (D) Calibration curves for evaluation the prognostic accuracy of the nomogram. *p<0.05, **p<0.01, ***p<0.001. HCC, hepatocellular carcinoma; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; OS, overall survival.

Fig. 8. Correlations between the risk score and immune-related genes, immune subtypes, tumor stemness, and drug susceptibility in HCC.

(A–B) Correlation between the risk score and the expression of PD-1, CTLA-4, and TIM3 in the training and validation groups. (C–D) Distribution of the risk score and immune subtypes in HCC. *p<0.05, **p<0.01, ***p<0.001. CTLA-4, cytotoxic T lymphocyte-associated antigen-4; HCC, hepatocellular carcinoma; lncRNAs, long noncoding RNAs; m⁶A, N6-methyladenosine; PD-1, programmed cell death 1; TCGA, The Cancer Genome Atlas; TIM3, T-cell immunoglobulin and mucin domain 3.