1. Abnormal expression of m6A RNA methylation regulators is associated with tumour stages in patients with HCC
The expression of m6A RNA methylation regulators and PD-L1 in HCC was analysed using ONCOMINE and UALCAN. A total of 50 healthy tissues and 371 HCC tissues were examined using the UALCAN database. As shown in Figure 1, the expression of all genes, except PD-L1, was significantly upregulated in HCC tissues. In addition, in the ONCOMINE database, analysis of multiple HCC datasets showed that the expression of m6A RNA methylation regulators was significantly higher in HCC tissues than in healthy tissues (Table S2), suggesting that m6A RNA methylation regulators play an important role in the occurrence and development of HCC and may serve as molecular markers for early diagnosis and prognosis of HCC. Furthermore, the relative expression levels of each m6A RNA methylation regulator and PD-L1 were compared in HCC tissues. As shown in Figure 2, m6A RNA methylation regulators were generally significantly upregulated in HCC tissues.
Furthermore, the relationship between the expression of m6A RNA methylation regulators and clinical stages of HCC was evaluated using the GEPIA database. As shown in Figure 3, the expression of YTHDF1, YTHDF2, YTHDC1, RBM15 and METTL3 was found to be significantly associated with HCC staging, whereas other genes did not have a significant relationship with the clinical stages of HCC. Moreover, the expression of m6A methylation regulators was downregulated in the late stage of HCC (stage 4), suggesting that m6A modification mainly occurs in the early stages of HCC. This finding may also be related to the small sample size of patients with stage 4 disease; therefore, the sample size should be expanded for further verification. In conclusion, the abovementioned results suggest that the expression level of m6A methylation regulators is significantly correlated with individual tumour stage and is higher in patients with early-stage disease.
2. Prognostic characteristics of m6A RNA methylation regulators in patients with HCC
To investigate the effects of m6A RNA methylation regulators on the prognosis of HCC, the relationship between the expression of m6A regulators and PD-L1 and OS, relapse-free survival (RFS), PFS and DSS was examined using the K-M plotter. As shown in Table S3, patients in each cohort were divided into the low- and high-risk groups based on the cut-off values. The cut-off values for OS associated with m6A RNA methylation regulators are shown in Figure S2. As shown in Figure 4, upregulated mRNA expression of KIAA1429 was associated with poor OS (p = 0.045) but was not associated with RFS, PFS or DSS. Upregulated mRNA expression of METTL3 was associated with poor OS (p = 0.003), RFS (p = 0.0014), PFS (p = 0.0072) and DSS (p = 0.0068), suggesting that patients with HCC with upregulated METTL3 expression had a poor prognosis. In addition, upregulated mRNA expression of RBM15 was associated with PFS (p = 0.027), and upregulated mRNA expression of WTAP was correlated with RFS (p = 0.0092) and PFS (p = 0.0047). Furthermore, downregulated mRNA expression of ZC3H13 was associated with poor OS (p = 0.00033), RFS (p = 0.025), PFS (p = 0.019) and DSS (p = 0.00082), whereas downregulated mRNA expression of METTL14 was associated with poor OS (p = 0.00039), RFS (p = 0.021) and DSS (p = 0.0039) (Figure 5). Low expression of ZC3H13 and METTL14 as m6A ‘writer’ genes has been associated with a poor prognosis in breast and colon cancers[26-28] and may lead to downregulation of m6A RNA modification in tumours, thereby reducing the level of immune cell infiltration and resulting in a poor prognosis, which is consistent with the results of this study. Upregulated mRNA expression of HNRNPC was associated with poor OS (p = 0.028) and DSS (p = 0.012) (Figure 5). Upregulated mRNA expression of YTHDC1 was associated with poor PFS (p = 0.04) and good DSS (p = 0.047). In addition, upregulated mRNA expression of YTHDF1 was associated with poor OS (p = 0.0042), RFS (p = 0.00012), PFS (p = 0.0017) and DSS (p = 0.0072), whereas mRNA expression of YTHDF2 was associated with poor OS (p = 0.017) and RFS (p = 0.015). Other m6A RNA methylation regulators associated with the prognosis of HCC are shown in Figure S1. In conclusion, abnormal expression of m6A RNA methylation regulators plays an important regulatory role in the prognosis of HCC, with clinical significance. For example, upregulated expression of METTL3, WTAP, YTHDF1 and YTHDF2 indicates a poor prognosis of HCC. However, low expression of ZC3H13 and METTL14 as tumour suppressor genes can also lead to a poor prognosis of HCC, indicating that m6A RNA methylation regulators play an extremely important role in the prognosis of HCC.
3. Mutations in m6A RNA methylation regulators and their relationship with survival in patients with HCC
Epigenetic changes play a crucial role in the occurrence and development of tumours. The effects of mutated m6A RNA methylation regulators on the progression and prognosis of HCC remain unknown. In this study, mutations in m6A RNA methylation regulators in patients with HCC were analysed using the online tool cBioPortal for Cancer Genomics (TCGA, Firehose Legacy). Mutated m6A RNA methylation regulators were found in 331 (89%) of the 371 patients with HCC (Figure 6A). Among these regulators, the mutation rates of KIAA1429, YTHDF3, ALKBH5, WTAP, YTHDF1, YTHDF2 and FTO were 43%, 29%, 27%, 23%, 21%, 18% and 17%, respectively. In addition, the mRNA expression of these regulators and PD-L1 (RNA Seq V2 RSEM) in LIHC (TCGA, Firehose Legacy) was analysed using cBioPortal, followed by Pearson correlation analysis. The correlation between RNA methylation regulators and PD-L1 was observed (Figure 6B), and the results showed that KIAA1429 was positively correlated with PD-L1 and YTHDF3; METTL3 was positively correlated with PD-L1, YTHDF1, HNRNPC, METTL14, MTEEL16 and RBM15; RBM15 was positively correlated with YTHDC2, YTHDC1, HNRNPC and METTL14; ZC3H13 was positively correlated with ALKBH5; METTL16 was positively correlated with YTHDC2 and METTL14; METTL14 was positively correlated with FTO and YTHDC2 and HNRNPC was positively correlated with PDCD1 and YTHDF1. In addition, YTHDC1 was positively correlated with PD-L1; YTHDF1 was positively correlated with PDCD1; YTHDF2 was positively correlated with YTHDF3 and PD-L1 was positively correlated with PDCD1. In addition, the relationship between mutated m6A RNA methylation regulators and OS and disease-free survival (DFS) in patients with HCC was investigated. The K-M plot and log-rank test results showed that mutated m6A RNA methylation regulators were associated with shorter OS (Figure 6C, p = 5.339E-3) and DFS (Figure 6C, p = 4.411E-3) in patients with HCC. These results suggest that mutations in m6A RNA methylation regulators and the PD-L1 gene play an important role in the prognosis of HCC.
4. GO and KEGG enrichment analyses of m6A RNA methylation regulators, PD-L1 and 170 co-expressed genes in patients with HCC
After analysing the gene mutations of m6A RNA methylation regulators and their prognostic value in patients with HCC, the ‘co-expression’ module of cBioPortal was used to analyse the top 10 co-expressed genes (among 170 genes) significantly associated with the mutations of each m6A RNA methylation regulator (see Table S4 for specific co-expressed genes). Subsequently, a PPI network was established using the STRING database. As shown in Figure 7A, the HNRNP family (HNRNPA1/3, HNRNPL and HNRNPU), SNRPD1/3, SRSF3/7, PHF5A and PRPF38A were found to be closely associated with mutated m6A RNA methylation regulators. The HNRNP family is a class of RNA-binding proteins with various key cellular functions. Cell dysfunction, including selective splicing, translation and RNA processing, plays an important role in tumorigenesis. Therefore, the HNRNP family has attracted increasing attention owing to its association with cancer progression. In addition, SNRPD1/3 plays a regulatory role in breast cancer through cell cycle regulation, and SRSF3/7 is an RNA-binding protein associated with metastasis and recurrence. PHF5A and PRPF38A play an important role in regulating tumour proliferation and migration. Therefore, in this study, GO and KEGG analyses were performed using DAVID to analyse the potential role of m6A RNA methylation regulators and their 170 co-expressed genes in HCC (Table S5). As shown in Figure 7B–E, mutated m6A RNA methylation regulators were found to significantly regulate biological processes (BPs), such as regulation of signal transduction by a p53 class mediator, cell proliferation, T-cell co-stimulation, mRNA processing and positive regulation of T-cell proliferation (Figure 7B). Furthermore, molecular functions included protein, poly(A) RNA and protein kinase binding (Figure 7C), whereas cellular components included the cytoplasm, nucleus and nucleoplasm (Figure 7D). KEGG analysis indicated that mutated m6A RNA methylation regulators were associated with the activation of the T-cell receptor signalling pathway, RNA transport and spliceosomes (Figure 7E). Moreover, the relationship between mutated m6A RNA methylation regulators and the clinical characteristics of patients with HCC was also analysed (Figure S3, Table S6). The results suggested that mutated m6A RNA methylation regulators were associated with vascular invasion. In addition, the cancer type, neohistological grade and primary tumour site were significantly correlated with the mutated regulators, suggesting that the regulators were associated with a poor prognosis.
5. Immune infiltration analysis of m6A RNA methylation regulators in the tumour microenvironment of patients with HCC
Tumour-infiltrating immune cells, as an important part of the tumour microenvironment, are closely related to the occurrence, progression or metastasis of tumours. Previous studies have shown that differences in tumour tissue expression, survival and gene mutations were highly significant among genes in the YTHDF family. In addition, studies have shown that m6A RNA methylation regulators play an important role in the regulation of tumour immunity. Therefore, the underlying mechanisms and role of the YTHDF family in immune infiltration within the tumour microenvironment should be further investigated. In this study, the CIBERSORT, CIBERSORT-ABS, QUANTISEQ, XCELL, MCPCOUNTER, TIDE and EPIC algorithms were used to investigate the potential relationship between the infiltration levels of various immune cells and the expression of the YTHDF family in TCGA-HCC cohort (371 patients). The role of the YTHDF family in the infiltration of tumour-associated fibroblasts and Tregs was determined using the MCPCOUNTER, TIDE and EPIC algorithms, which showed that the expression of the YTHDF family was significantly positively correlated with the infiltration of tumour-associated fibroblasts in HCC (Figure 8A). In addition, analyses performed using the CIBERSORT, CIBERSORT-ABS, QUANTISEQ and XCELL algorithms (Figure 8B) showed that the expression of the YTHDF family was significantly correlated with Treg infiltration in HCC, suggesting its important role in the regulation of tumour immunity and immune escape.
6. qRT-PCR and WB validated the upregulation of m6A RNA methylation regulators in HCC
Analyses performed using UALCAN, TCGA and other databases revealed that the expression of m6A RNA methylation regulators was upregulated in HCC. To further verify these results, surgical specimens were collected from 6 patients clinically and pathologically diagnosed with HCC. qRT-PCR and WB were used to verify changes in the expression of m6A RNA methylation regulators. As shown in Figure 9A, the results of qRT-PCR showed that the expression of m6A RNA methylation regulators was upregulated in HCC, and the results of WB (Figure 9B) showed that the expression of the YTHDF family was upregulated in HCC. These results are consistent with those mentioned above. In addition, to examine the potential role of m6A modification in HCC, the EpiQuik m6A RNA Methylation Quantification Kit was used to detect m6A modification levels in the total RNA of tumour and para-cancerous tissues. The results revealed that m6A modification levels were lower in HCC tissues than in the corresponding adjacent tissues (Figure 9C). These results suggest that the overall m6A modification levels in HCC are mainly mediated by eraser and reader proteins, which warrants further study.