Targeting CCL3+mono like LSC overcomes chemotherapy resistance in NPM1 mutant AML

Abstract Background: Nucleophosmin 1 mutation (NPM1mut) represents one of the most frequent genetic alterations in acute myeloid leukmia (AML). Although is generally associated with a favorable prognosis, nearly 40% of patients eventually relapse, highlighting the need to elucidate the specific mechanisms underlying this prognostic variability. Methods:We enrolled forty NPM1mut AML patients exhibiting divergent clinical outcomes for single-cell RNA-sequencing (scRNA-seq) analysis to characterize the cellular heterogeneity of AML subsets and immune microenvironment features associated with poor prognosis. Flow cytometry was employed to validate and functionally characterize the chemoresistant subpopulation identified through scRNA-seq. Subsequent cell-cell communication analysis and in vitro assays were performed to delineate interactions between malignant clones and immune cells. To investigate the molecular basis of treatment resistance, we conducted pathway enrichment analysis followed by shRNA-mediated knockdown experiments in NPM1mut AML cell lines. Finally, we implemented high-throughput screening of clinical compound libraries and evaluated promising inhibitors using in vivo models. Results:Patients were stratified into favorable and adverse prognosis groups based on clinical outcomes, followed by comprehensive single-cell transcriptomic profiling. Subgroup analysis of AML uncovered a monocyte-like CCL3+ leukemia stem cell (LSC) population that was strongly correlated with poor prognosis in NPM1mut AML. Flow cytometry validation in bone marrow samples from 45 NPM1mut AML patients further confirmed that this monocyte-like CCL3+ LSC subset serves as a robust prognostic biomarker for disease progression. Notably, comparative analysis of the immune microenvironment revealed a marked reduction in both the proportion and functional activity of CD8+ T cells in the adverse prognosis group relative to the favorable group. Leveraging cell-cell communication studies and in vitro functional assays, we demonstrated that monocyte-like CCL3+ LSCs in the adverse group actively suppress T cell function through TGF-β1 secretion. Mechanistic investigations revealed that cytoplasmic-localized mutant NPM1 (NPM1c) physically interacts with ERK, driving its phosphorylation and subsequent nuclear translocation. This signaling cascade culminates in the transcriptional upregulation of the anti-apoptotic gene BCL2 and immune regulatory genes, including TGFB1, thereby promoting therapeutic resistance and facilitating immune evasion in NPM1mut AML cells. To identify potential therapeutic vulnerabilities, we conducted high-throughput screening of clinical drug libraries, which revealed several promising targeted compounds. Further preclinical evaluation demonstrated that MAPK pathway inhibitors significantly enhance chemotherapy response in NPM1mut AML models in vivo, offering a potential strategy to overcome treatment resistance. Conclusion: This study provides a comprehensive analysis of the intrinsic heterogeneity in NPM1mut AML and identifies novel therapeutic targets with clinical potential. These findings not only advance our understanding of AML pathogenesis but also offer promising translational opportunities to improve patient outcomes. We are confident this work will shed new light on AML heterogeneity and inspire future therapeutic development. We sincerely appreciate your time and consideration of our manuscript for publication.