Analysis of Exosome-Related Differential Gene Expression Profiles in Acute Myeloid Leukemia and Exploration of Therapeutic Targets

This bioinformatics investigation systematically interrogates leukemia stem cell (LSC) transcriptomes from acute myeloid leukemia (AML) patients, focusing on exosome-related biomarkers to delineate their mechanistic involvement and therapeutic potential in AML pathogenesis. We curated two LSC gene expression datasets (GSE17054, GSE24395) from GEO, implementing rigorous normalization and batch effect correction using R-based bioinformatics pipelines (ComBat algorithm via sva package). Harmonized datasets underwent differential expression analysis through empirical Bayes moderated t-tests (limma package), cross-referenced with exosome-associated genes from GeneCards to identify AML-specific candidates. Subsequent DAVID functional annotation revealed critical enrichment in vesicle-mediated transport (FDR<0.05) and ferroptosis pathways (p=3.2×10⁻⁴). STRING database analysis coupled with network topology analysis in Cytoscape identified six hub genes (ANXA2, TNFRSF1A, THY1, CD63, IDH1, S100A11) exhibiting significant connectivity (degree centrality ≥8), including membrane trafficking regulators (ANXA2, CD63) and metabolic modulators (IDH1). These findings establish an exosome-related gene signature with therapeutic implications, particularly highlighting IDH1's dual role in 2-hydroxyglutarate metabolism and extracellular vesicle biogenesis. Our multi-omics integration strategy provides a novel framework for understanding AML pathogenesis through the exosome-LSC axis, serving as potential diagnostic biomarkers and therapeutic targets, ultimately advancing therapeutic development for AML.