Abstract
BACKGROUND: Recently, many studies have been conducted to examine immune response modification at epigenetic level, but the candidate effect of RNA 5-methylcytosine (m(5) C) modification on tumor microenvironment (TME) of acute myeloid leukemia (AML) is still unknown at present. METHODS: We assessed the patterns of m(5) C modification among 417 AML cases by using nine m(5) C regulators. Thereafter, we associated those identified modification patterns with TME cell infiltration features. Additionally, stepwise regression and LASSO Cox regression analyses were conducted for quantifying patterns of m(5) C modification among AML cases to establish the m(5) C-score. Meanwhile, we validated the expression of genes in the m5C-score model by qRT-PCR. Finally, the present work analyzed the association between m(5) C-score and AML clinical characteristics and prognostic outcomes. RESULTS: In total, three different patterns of m(5) C modification (m(5) C-clusters) were identified, and highly differentiated TME cell infiltration features were also identified. On this basis, evaluating patterns of m(5) C modification in single cancer samples was important for evaluating the immune/stromal activities in TME and for predicting prognosis. In addition, the m(5) C-score was established, which showed a close relation with the overall survival (OS) of test and training set samples. Moreover, multivariate Cox analysis suggested that our constructed m(5) C-score served as the independent predicting factor for the prognosis of AML (hazard ratio = 1.57, 95% confidence interval = 1.38-1.79, p < 1e(-5) ). CONCLUSIONS: This study shows that m(5) C modification may be one of the key roles in the formation of diversity and complexity of TME. Meanwhile, assessing the patterns of m(5) C modification among individual cancer samples is of great importance, which provides insights into cell infiltration features within TME, thereby helping to develop relevant immunotherapy and predict patient prognostic outcomes.