Key Points
Leukemia-associated myeloid cells support T-ALL in the CNS via integrin signaling and differ transcriptionally from healthy counterparts.
In vivo myeloid depletion reduces leukemia burden in the CNS and enhances chemotherapy to improve mouse survival.
Abstract
T-cell acute lymphoblastic leukemia (T-ALL) is a common pediatric malignancy, and central nervous system (CNS) involvement is associated with inferior prognoses. Here, we show that in the LMO2 transgenic and LN3 mouse models of T-ALL that leukemia blasts infiltrate the CNS, where they co-localize with an altered myeloid compartment. Notably, myeloid cells from the leukemic, but not healthy, CNS were sufficient to support T-ALL survival in vitro, and depletion of myeloid cells in vivo significantly reduced CNS leukemia burden. Among myeloid subsets, macrophages and monocytes from the leukemic CNS most robustly supported T-ALL in vitro. Both T-ALL blasts and leukemia-associated myeloid cells in the CNS expressed elevated levels of adhesion molecules, and inhibition of integrin-mediated adhesion or downstream signaling prevented myeloid cells from supporting survival of T-ALL cells in vitro. Single-cell transcriptomic analyses revealed that leukemia-associated myeloid cells in the CNS undergo transcriptional changes, increasing expression of genes associated with oxidative phosphorylation, while reducing inflammatory signatures. Given the clinical importance of T-ALL relapse in the CNS, we tested if myeloid cells provide a supportive niche for chemotherapy-resistance. T-ALL cells in the CNS that rebounded post-chemotherapy retained myeloid dependence in vitro, and myeloid-depletion improved responsiveness to chemotherapy, prolonging mouse survival. Altogether, these findings demonstrate that myeloid cells support T-ALL in the CNS and suggest that targeting myeloid cells or mechanisms by which they support T-ALL, in conjunction with conventional chemotherapeutic approaches, could improve current treatment outcomes.
Author notes
Data sharing statement:
For original data, please contact “lehrlich@austin.utexas.edu”;. scRNA-seq data may be accessed at the Gene Expression Omnibus (GEO Accession Number: GSE276075). Reviewer access at: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE276075 (token: qhetsusovdwlhmf).