In this present research, we attempted to demonstrate the expression patterns, prognostic values, and effects on the TME of m6A regulators in GEA. Differential expression analysis found that the majority of m6A RNA methylation regulators were significantly differently expressed between adjacent normal and tumor samples. This should preliminarily suggest that those m6A regulators have a distinct role of influencing GEA initiation and could be applied for the subsequent analysis. Compare to normal tissues, GEA is locally infiltrated with higher immune cell subgroups, including B cells naive, B cells memory, Plasma cells, T cells gamma delta, Macrophages M0, Macrophages M1, Dendritic cells resting, Dendritic cells active. Meanwhile, ESTIMATE algorithm-derived immune scores, stromal scores and ESTIMATE scores were applied to facilitate the quantification of the non-tumor components in malignancy [22]. Stromal, immune and ESTIMATE scores for tumor tissue were found to be significantly associated with the clinicopathologic features of the tumor such as age, differentiation grade and TNM stage. Importantly, most m6A regulators’ expression was significantly associated with immune/stromal scores. Furthermore, we constructed a comprehensive single-cell transcriptome atlas of 18070 cells from 2 samples of GEA and 1 sample of adjacent normal mucosa and revealed m6A regulators exert an important role in maintaining aberrant immune infiltration. Therefore, these results demonstrated that aberrant immune infiltration and m6A regulators expression in GEA as a tightly regulated process might play important roles in tumor development and this process had clinical importance.
We characterized the effects of distinct m6A methylation modification on different GEA subtypes by clustering m6A regulators. The two subtypes showed significant differences in patient prognosis, PD-1 expression, immune cell infiltration and RS. This suggests that two subtypes’ differences are essential and reflect the heterogeneity of the immune microenvironment of GEA, which is worthy of further study. In order to investigate the expression characteristics of m6A methylation regulators in tumor, many studies clustered the tumor samples into different groups using consensus clustering analysis. For instance, Jing Chen et al. [23] identified two clusters of clear cell renal carcinoma with significant differences in OS and tumor stage between them based on the expression pattern of m6A RNA methylation regulators by means of consensus clustering. Similarly, Lilan Yi et al. [24] showed that two molecular subtypes were identified by consensus clustering for 15 m6A regulators, and two subtypes were distinct in prognosis, PD-L1 expression, immunoscore, and immune cell infiltration. However, so far, the expression of m6A regulators has remained elusive for typing research by consensus clustering analysis in GEA. Our research had once again identified the special relationship between m6A modification and tumor immune cell infiltration, which has important clinical value.
To elucidate the potential association between the genome, m6A modification pattern and TME in GEA, we systematically clustered the co-expressed genes by WGCNA analysis. This approach allowed us to identify gene modules most related to cancer immunological phenotypes. COL4A1 and COL5A2, the two hub genes in the module, in the collagen family can encode pro-alpha collagen chains that are assembled into collagen. They have a relatively high expression in tumor cells and related to the prognosis and clinicopathological factors of patients. Désert et al. [25] reported that elevated expression of COL4A1 was significantly correlated with tumor stage and worse overall survival in patients with hepatocellular carcinoma. Zhang et al. [26] demonstrated the abnormally high expression of COL4A1 in GC, and high expression of COL4A1 was closely correlated with primary tumor size, lymph node metastasis and distant metastasis, with the silencing of COL4A1 significantly inhibiting cell proliferation of GC cells in vitro. Meanwhile, elevated COL4A1 gene expression has been found to be associated with trastuzumab resistance in GC [27]. Several studied had reported that COL5A2 might function as crucial role in the initiation and progression of tumor by using bioinformatics technologies [28, 29]. More importantly, COL5A2, was correlated with stromal scores in GC, promoted the recruitment of circulating monocytes into the TME and facilitated their differentiation into tumor-associated macrophages [30]. Similarly, in our research, we found that COL4A1 and COL5A2 were significantly related to prognosis of GEA patients and regulate TME infiltration characteristics. Intriguingly, the expression of COL5A2 and COL4A1 was significantly correlated with ICB (PD-1/L1and CTLA-4) expression. Notably, COL5A2 expression was also linked to enhanced response to anti-PD-1 immunotherapy. Above results suggest that distinct m6A modification patterns of COL5A2 may establish totally different TME, which influence immunotherapy response.
Whether m6A RNA methylation regulators have prognostic value in cancer is of great significance [31]. We performed univariate and LASSO Cox regression analyses to construct a prognostic related risk signature with three m6A RNA methylation regulators, including KIAA1429, HNRNPA2B1 and FMR1, which divided the GEA patients into low- and high-risk groups. In the m6A methyltransferase complex, KIAA1429 acts as a scaffold in bridging the catalytic core components of methyltransferase complex and RNA substrates, which affect the installation of m6A at specific locations [32]. Ran Miao et al. [33] found that KIAA1429 could serve as an oncogene in gastric cancer by stabilizing c-Jun mRNA in an m6A-independent manner. HNRNPA2B1 is a nuclear reader of the m6A mark and has important effects on primary microRNA processing and alternative splicing. Barceló et al. [34] reported that HNRNPA2B1 acts as a regulator of KRAS-dependent tumorigenesis through the critical pancreatic ductal adenocarcinoma cells signaling pathway PI3K/AKT. FMR1 gene and consequently lack of synthesis of FMR protein (FMRP) is associated with fragile X syndrome, and FMRP plays a critical role in chromatin dynamics, RNA binding, mRNA transport, and mRNA translation [35, 36]. Jianhao Li et al. [37] indicated that high expression of KIAA1429 and HNRNPA2B1 were significantly associated with poor prognosis in osteosarcoma, and m6A regulators might be involved in osteosarcoma progression through humoral immune response. Francesca Zalfa et al. [38] reported that there was an association between FMRP levels and the invasive phenotype in melanoma. In accordance with previous results, we found the three-gene risk signature, KIAA1429, HNRNPA2B1 and FMR1, showed a good performance for predicting GEA patients’ prognosis and immune cells infiltration characteristics. Importantly, our further study firstly revealed that the three m6A regulators’ expression in T cells and macrophages may play a crucial role in tumor progression. Moreover, COL5A2 expression was significantly related to KIAA1429 and FMR1 expression. Therefore, we speculate that KIAA1429 and FMR1 regulate the number of macrophages and T cells in TME by affecting the m6A modification of COL5A2, which in turn affects the clinicopathological characteristics of patients.
The tumor microenvironment plays an essential regulatory role in tumorigenesis, and its heterogeneity can lead to multiple dimensions, including patient prognosis and therapeutic response [39–41]. Here, we generated a single-cell transcriptome atlas of GEA sample to validate the role of m6A modification pattern on TME. 13 different cell types were identified in the GEA microenvironment. Notably, CD8+ T cells mostly originated from normal mucosal tissues, while Macrophages, CD4+ T cells and Treg cells were enriched in GEA tissues. Therefore, we indicated that the downregulated immunogenicity of cancer cells potentially contributes to the formation of an immunosuppressive microenvironment. Hanjie Li et al. [42] reported that a large population of CD8+ T cells showing continuous progression from an early effector "transitional" into a dysfunctional T cell state and the intensity of the dysfunctional signature is related to tumor reactivity. Moreover, CD4+T cell help in the form of IL-21 may potentially be harnessed to bolster the formation of protective cytolytic CX3CR1+CD8+T cells and improve control over tumor progression [43]. Our monocle analysis verified again that CD4+ T cells regulate CD8+ T cell differentiation in tumor progression of GEA. m6A RNA modification controls the differentiation of naive T cells and sustains the suppressive functions of Tregs [5, 44]. Then, our work revealed that three m6A regulators are highly expressed in CD4+ T cells, CD8+ T cells, Tregs and Macrophages, which was consistent with previous works. In short, we first discovered that KIAA1429, HNRNPA2B1 and FMR1 regulate T cell differentiation in the GEA microenvironment, which may provide new targets to optimize immunotherapy.