FLT3/ITD-driven noncanonical STAT1 S727 phosphorylation upregulates CD276, mediating CD8positive- T cell exhaustion and immune evasion in AML

Abstract

Abstract FMS-like tyrosine kinase 3 internal tandem duplications (FLT3/ITD) mutation promotes leukemogenesis and confers poor prognosis in acute myeloid leukemia (AML). The direct impact and underlying mechanisms of FLT3/ITD on CD8+ T cell function remain poorly understood. Here we comprehensively analyze the CD8+ T-cell landscape in FLT3/ITD AML through integrated clinical observations and mechanistic investigations. First, integrated scRNA-seq of 43 AML bone marrow samples revealed an altered immune microenvironment, with 7 cellular compartments identified. FLT3/ITD samples showed significantly reduction in CD8+ T cells, which also exhibited elevated exhaustion signatures. This was validated by multicolor flow cytometry in 120 de novo AML cases, showing reduced CD8+ T cell frequency and higher exhaustion marker expression in FLT3/ITD patients. Functional co-culture assays with primary bone marrow mononuclear cells and activated healthy donor CD8+ T cells further demonstrated that FLT3/ITD cells exhibited lower CD8+ T cell-mediated killing efficiency. Isogenic models in FLT3/WT cell lines were engineered. FLT3/ITD lines exhibited constitutive FLT3 phosphorylation and hyperactivation of downstream effectors STAT5 (Y694) and AKT (S473). In co-culture assays with activated healthy donor CD8+ T cells, compared to FLT3/WT or NC, FLT3/ITD cells conferred stronger resistance to CD8+ T cell cytotoxicity, suppressed T cell proliferation and IFN-γ secretion, while Granzyme B and Perforin were unchanged. FLT3/ITD cells also upregulated exhaution signatures on co-cultured CD8+ T cell, effects not seen in FLT3/WT or NC. These findings demonstrate that FLT3/ITD mutations intrinsically drive CD8+ T cell dysfunction and immune evasion. Then we analyzed 452 primary AML samples from the BeatAML cohort and found significant enrichment of T-cell receptor signaling and marked upregulation of the co-inhibitory ligand CD276 in FLT3/ITD blasts, confirmed by multispectral flow cytometry in a real-world validation cohort (n=32). Ectopic FLT3/ITD expression increased both CD276 mRNA and surface protein levels, while pharmacologic disruption of FLT3/ITD signaling with cycloheximide or HSP70 inhibitor QL47 reduced CD276 expression. shCD276 or anti-CD276 mono-antibody blockade (MGA271) reversed FLT3/ITD-mediated immunosuppression, restoring CD8+ T cell cytotoxicity, proliferation, IFN-γ secretion, and reducing exhaustion markers. Neither STAT5 nor FLT3 kinase inhibition affected CD276. By integrating FLT3/ITD-interacting partners (CO-IP/MS dataset) with CHIP dataset-predicted CD276 transcription factors, STAT1 was identified as the lead candidate. Molecular docking and co-IP identified STAT1 as the key interactor, uniquely binding FLT3/ITD at two interfaces and inducing specific S727 phosphorylation. Spatial colocalization of pS727-STAT1 and FLT3-FLAG was observed in FLT3/ITD cells. Multiplex immunohistochemical confirmed elevated pS727-STAT1+CD276+ blasts and reduced CD8+ T cells in FLT3/ITD patient samples, linking FLT3/ITD–STAT1 signaling to CD276 transactivation and immune suppression. Pharmacologic (fludarabine) and genetic approaches showed dose-dependent suppression of CD276 mRNA and protein in FLT3/ITD models. JASPAR analysis identified 4 STAT1-binding motifs in the CD276 promoter, with the 338–348 bp region being critical. Dual-luciferase reporter assays revealed STAT1-dependent transactivation strictly requiring the 338-348 bp promoter segment and FLT3/ITD co-expression. ChIP-qPCR validated STAT1 enrichment at this locus. Further, STAT1-S727 phosphorylation—specific to FLT3/ITD and not Y701 was essential: STAT1-S727mut abolished CD276 transactivation, while STAT1-Y701mut retained partial activity. In vitro, the CD276 bispecific engager MG009 enhanced CD8+ T cell cytotoxicity, similar to CD276 monoclonal antibody MGA271. In humanized NSG mice with FLT3/ITD, quizartinib + MGA271/MG009 nearly eradicated tumors, with functional CD8+ T cell recovery. In FLT3/ITD PDX models, quizartinib + MGA271/MG009 achieved a 99% complete remission rate, with excellent tolerability, validating CD276 blockade as a potent synergistic strategy with FLT3 inhibition. In summary, FLT3/ITD drives CD8+ T cell exhaustion via tumor-intrinsic CD276 upregulation, mediated by STAT1 S727 phosphorylation. Co-targeting FLT3 and CD276 synergistically restores T-cell function and eradicates leukemia in vivo, offering a translatable immunotherapy strategy for AML.