Glioma has been considered as the most dangerous tumor among intracranial tumors. Because curative treatments for glioma are lacking, numerous molecular-level studies have been carried out to understand this disease, revealing the molecular mechanisms of glioma from different perspectives. Studies of regulation of the blood-brain barrier [27, 28], competing endogenous RNA mechanism [29, 30], and glioma stem cell mechanisms [31, 32] have provided feasible methods for treating glioma and improving patient prognosis. m6A modification has also been reported to interfere with glioma development and invasion. Various studies [33, 34] showed that m6A is associated with glioma; however, most of these studies determined the relationship between a single m6A regulator and glioma, and comprehensive studies of the relationship between m6A-related genes and glioma are lacking.
It was previously thought that the central nervous system (CNS) has an immune-privileged status. However, researchers’ studies revealed lymphatic vessels in the CNS [35, 36], suggesting that the CNS is under surveillance by the immune system, with the mechanism determined in 2021 . CNS-derived antigens in the cerebrospinal fluid flowing through lymphatic vessels enter and accumulate in the dural sinus. Antigen-presenting cells capture these antigens and present them to T cells. Thus, immune surveillance does not occur within the brain tissue directly but rather at the margin of the brain.
Immune cells can lead to intracranial disease; however, the relationship of m6A and immune cell infiltration and how these factors regulate the tumor microenvironment are not well-understood. Here, we evaluated three modification patterns of m6A and analyzed immune cell infiltration. We found that in glioma, most m6A regulators showed reduced copy numbers, which may have led to their low expression. The mutation frequency of m6A regulators was low, suggesting that in the brain, most m6A regulators are in a steady state, whereas in other cancers such as gastric cancer, prominent gene  mutations are observed. However, even low-frequency gene mutations can lead to heterogeneity. Based on survival analysis, the different expression status of m6A regulators can lead to different prognoses, and thus m6A regulators can be divided into oncogenes and tumor suppressor genes. Studies [39, 40] showed that changes in the expression level of oncogenes and gene suppressor genes can affect the growth and invasion of glioma cells.
In this research, we constructed three different m6A modification patterns, clusters A–C, which were used to analyze immune cell infiltration. Unexpectedly, most immune cells including B cells, CD4 T cells, CD8 T cells, dendritic cells, CD56 bright natural killer cells, CD56 dim natural killer cells, eosinophils, gamma delta T cells, immature B cells, immature dendritic cells, myeloid-derived suppressor cells, macrophages, mast cells, monocytes, natural kill T cells, neutrophils, plasmacytoid dendritic cells, regulatory T cells, T follicular helper cells, type 17 T helper cells, and type 1 T helper cells were highly infiltrated in cluster C. Cluster A was considered to suppress immunity and to be related to an immune-desert phenotype. Cluster B can be considered to activate innate immunity, which is related to an immune-excluded phenotype, and cluster C can be considered to activate adaptive immunity, which is related to an immune-inflamed phenotype. However, unexpectedly, cluster C was not associated with a better prognosis corresponding to its high immune cell infiltration, whereas cluster A, which lacked immune cell infiltration, was associated with the best prognosis.
Immune cells can kill tumor cells via their cytotoxic effects or phagocytosis. Traditional analysis also showed that high immune infiltration is effective for combating glioma; however, immune infiltration may act as a double-edged sword in glioma. Cluster C was highly related to autoimmune diseases such as lupus erythematosus, allograft rejection, and autoimmune thyroid disease. This result may be explained the high immune infiltration in cluster C, and clusters B and C were found to closely interact with extracellular matrix (ECM) receptors, whereas cluster A did not. Thus, to kill tumor cells, immune cells must pass through the collagen network of the ECM, the killing effect of immune cells is highly related to their migration speed; a dense collagen network may weaken immune cell function . Research  has shown that remodeling and deposition of the ECM can create a stiffer environment in which tumor cells reside. Tumor cells are protected by fibrocytes and collagen cells; to eliminate these structures, immune cells must gather and attack (high infiltration status), during which the ECM is consistently compressed, causing nucleus and membrane damage and immune cell depletion . This progress may damage the normal brain tissue and accelerate the invasion of glioma cells, explaining the poor prognosis in cases with high immune cell infiltration. Furthermore, the unique microenvironment may be regulated by the blood-brain barrier, and ECM stiffness in glioma may be higher than that for other tumors in the body. Thus, high levels of immune infiltration may contribute to glioma growth and invasion by interfering with the ECM.
Moreover, quantifying the m6Ascore helped us evaluated patient prognosis, gene mutation frequencies, and immune cell correlations and can be considered as an independent prognostic biomarker for glioma. Additionally, based on the m6Ascore, we predicted that anti-CD52/HE5 can be used to treat glioma.
There were some limitations to this study. In m6Ascore-related survival analysis, because patients grouped by age and sex were imbalanced, survival differences differed between the male and female groups and in patients older and younger than 65 years. Moreover, samples from Gene Expression Omnibus (GEO) were astrocytoma-related, and thus the description of relationship between m6A and glioma may be incomplete. Data regarding microsatellite instability in the glioma tumor microenvironment were also lacking and require further analysis. The process of the application of our results in the clinic to improve patient survival needs to also be examined.