Abstract
Acute myeloid leukemia (AML) is a highly heterogeneous malignant disease of the blood cell. The current therapies for AML are unsatisfactory and the molecular mechanisms underlying AML are unclear. 5-methylcytosine (m5C) is an important posttranscriptional modification of mRNA, and is involved in the regulation of mRNA stability, translation, and other aspects of RNA metabolism. However, based on our knowledge of published literature, the role of the m5C regulators has not been explored in AML till date. In this study, we clarified the expression and gene variants of m5C regulators in AML and found that most m5C regulators were differentially expressed and correlated with disease prognosis. We also found that the methylation status of certain m5C regulators (e.g., DNMT3A, DNMT3B) affects the survival of AML patients. Two m5C modification subtypes, and high- and low-risk subgroups identified based on the expression of m5C regulators showed significant differences in the prognosis as well as immune cell infiltration. In addition, most of the m5C regulators were found to be correlated with miRNA expression in AML, as well as IC50 values of many drugs. The miRNA and GSVA analysis were used to identify the different miRNAs and KEGG or hallmark pathways between high- and low-risk subgroups. We also built a prognostic model based on m5C regulators, which was validated by two GSE databases. To verify the reliability of our analysis and conclusions, qPCR was used to identify the expressions of m5C regulators between normal and AML. In summary, we comprehensively explored the molecular characteristics of m5C regulators and built a prognostic model in AML. We proposed new mechanistic insights into the role of m5C in multiple databases and clinical data, which may pave novel ways for the development of therapeutic strategies.
Introduction
Acute Myeloid Leukemia (AML) is characterized by high molecular heterogeneity, making effective treatment a significant clinical challenge. While genetic mutations and histone modifications have been extensively studied, the "epitranscriptome"—specifically chemical modifications to RNA—represents a critical but understudied layer of regulation. This study focuses on 5-methylcytosine (m5C), a modification that influences mRNA stability and translation. Despite its known role in other cancers, the impact of m5C regulators in AML has remained largely unexplored until now.
Methods and Data Integration
The researchers employed a multi-omic computational approach, integrating transcriptomic data from TCGA-LAML and GTEx databases. They identified 17 key m5C regulators, categorized into "Writers," "Erasers," and "Readers." Using LASSO Cox regression, they developed a 6-gene prognostic signature (NSUN3, NSUN4, NSUN6, NSUN7, DNMT1, ALKBH1). The model was validated using external datasets (GSE12417, GSE37642) and single-cell RNA sequencing. Finally, RT-qPCR was performed on clinical samples to confirm the expression patterns of these regulators.
Results
Differential Expression and Molecular Landscape
The study revealed that most m5C regulators are significantly differentially expressed in AML compared to normal tissues. For instance, regulators like ALYREF, NSUN2, and NSUN5 showed distinct expression profiles, often being overexpressed in normal cells relative to leukemic ones. Furthermore, DNMT3A, a well-known DNA methyltransferase, was found to have the highest mutation rate (19%) among m5C regulators, highlighting a potential crosstalk between DNA and RNA methylation.
Prognostic Significance and Risk Modeling
The 6-gene risk model demonstrated robust predictive power, achieving an AUC of 0.744 for 1-year survival. Survival analysis showed that patients in the low-risk group lived significantly longer than those in the high-risk group. The prognostic value of individual regulators was quantified via hazard ratios, identifying several as significant risk factors for poor outcomes.
Immune Microenvironment and Drug Sensitivity
"High-risk" m5C patterns were associated with an "immune-cold" environment, characterized by increased macrophages and CD8+ T cells but decreased CD4+ T cells, suggesting a mechanism for immune evasion. Additionally, m5C regulators correlated with cancer cell "stemness" and drug sensitivity. Notably, high DNMT3A expression was linked to increased sensitivity to Nelarabine, suggesting potential for personalized therapeutic strategies.
Discussion
This study provides the first comprehensive map of m5C RNA modification regulators in AML. The developed 6-gene signature offers a promising tool for risk stratification. However, the authors note that the study is primarily retrospective and computational. Future research utilizing m5C MeRIP-seq is necessary to map actual modification sites and establish direct mechanistic links to leukemic transformation. Prospective clinical trials are also required to validate the "RiskScore" in a hospital setting.
Summary updated