M6A regulator-mediated immune infiltration and methylation modification in hepatocellular carcinoma microenvironment and immunotherapy

Abstract

Introduction: Tremendous evidence indicates that N6-methyladenosine (m6A) epigenetic modification and m6A-related enzymes constitute a complex network, which jointly regulates prevailing pathological processes and various signaling pathways in humankind. Currently, the role of the m6A-mediated molecular regulatory network in hepatocellular carcinoma (HCC) remains elusive. Methods: We recruited expression and pathological files of 368 HCC patients from The Cancer Genome Atlas cohort. Four public datasets serve as external authentication sets for nearest template prediction (NTP) validation. The correlation between 35 regulators and their prognostic value was compared. Gene set variation analysis (GSVA) was used to explore the latent mechanism. Four independent algorithms (ssGSEA, xCell, MCP-counter, and TIMER) were used to calculate the ratio of tumor cells and non-tumor cells to evaluate the tumor immune microenvironment. The m6Ascore model was established by principal component analysis (PCA). Prediction of immunotherapy and potential drugs was performed using TIDE and SubMap. Results: A total of 35 m6A regulators were widely associated, most of which were risk factors for HCC patients. The m6A phenotypic-cluster revealed differences in regulator transcriptional level, gene mutation frequency, functional pathways, and immune cell infiltration abundance under distinct m6A patterns. As expected, the m6A gene cluster confirmed the aforementioned results. The m6Ascore model further found that patients in the high-m6Ascore group were associated with lower tumor purity, higher enrichment of immune and stromal cells, upregulation of metabolic pathways, lower expression of m6A regulators, and favorable outcomes. Low-m6Ascore patients were associated with adverse outcomes. Notably, low-m6Ascore patients might be more sensitive to anti-PD-L1 therapy. Conclusion: This study found that a classification model based on the m6A manner could predict HCC prognosis and response to immunotherapy for HCC patients, which might improve prognosis and contribute to clinical individualized decision-making.