Leukemia-associated myeloid cells support distinct T-ALL subtypes via unique signals converging on metabolic alterations.

Abstract

Although intensified chemotherapy regimens have improved survival of T-cell acute lymphoblastic leukemia (T-ALL) patients, treatment-related toxicities and poor outcomes following relapse highlight the need for alternative therapeutics. Our previous studies showed that leukemia-associated myeloid cells support T-ALL progression, suggesting myeloid cells or associated signals could serve as therapeutic targets. It remains unknown whether leukemia-associated myeloid cells support multiple molecular subtypes of T-ALL, and if so, whether subtype-specific mechanisms are involved. We demonstrate that tumor-associated myeloid cells support survival of both Early T-cell Progenitor (ETP)-like and non-ETP-like T-ALL subtypes from the LMO2 mouse model in vitro and in vivo. Transcriptional profiling and in vitro assays of mouse and human T-ALL reveal that myeloid cells support distinct T-ALL subtypes via different signaling pathways: IL6ST/STAT3 signaling supports ETP-like T-ALL, while growth factor receptor signaling supports non-ETP-like T-ALL. Notably, both subtypes require AKT activation for myeloid-mediated support, and acute myeloid depletion in vivo induces a common metabolic shift towards oxidative phosphorylation (OxPhos). These findings suggest that myeloid cells promote T-ALL survival via subtype-specific signals that converge on a common pathway regulating metabolism. The shared metabolic adaptation to myeloid cell loss suggests a compensatory mechanism enabling T-ALL persistence under stress. Consistent with this possibility, combining myeloid depletion with OxPhos inhibition reduces survival of mouse and patient T-ALL cells and prolongs leukemic mouse survival more than either single treatment. Our data highlight unique and shared mechanisms by which myeloid cells support T-ALL subtypes and implicate tumor-myeloid interactions and downstream metabolic reprogramming as promising therapeutic targets.