Mesenchymal Stromal Cell Expansion in FLT3-ITD+ Zebrafish Treated with Gilteritinib

Internal tandem duplication mutations to the FMS-like tyrosine kinase III gene (FLT3-ITD) are common in acute myeloid leukemia (AML) and are associated with increased risk of relapse. The FDA approved tyrosine kinase inhibitor (TKI) gilteritinib is the current mainstay for treating relapsed FLT3-ITD AML, however it is non-curative in the absence of hematopoietic stem cell transplantation. In vivo mechanisms of resistance are not well-understood. Consequently, how expression of FLT3-ITD and the effect of FLT3 inhibitors such as gilteritinib remodel the hematopoietic niche to induce therapy resistance represents a critical unanswered question in improving treatment for FLT3-ITD AML.

Using a well-defined Runx1 enhancer element, we enforced expression of human FLT3-ITD by hematopoietic stem and progenitor cells (HSPCs) in zebrafish. Runx1:FLT3-ITD(+) animals had greater numbers of HSPCs compared to transgene-negative controls (2.6 ± 0.3 vs 1.7 ± 0.2, p = 0.008). Treatment with gilteritinib (100 nM), from 72-96 hours post-fertilization (hpf) reduced HSPC numbers yet expression of FLT3-ITD partially rescued HSPC numbers in Runx1:FLT3-ITD(+) transgenics. To investigate this finding more deeply, single cell RNA sequencing (scRNA-seq) was performed on FLT3-ITD(+) or FLT3-ITD(-) animals with or without gilteritinib treatment. In untreated animals, enforced expression of FLT3-ITD expanded HSPC and erythroid-precursor populations (67% and 65% relative abundance (r.a.), p = 2.8 x 10^-6 and p = 0.01, respectively), and contracted the mesenchymal stromal cell (MSC) population (35% r.a., p = 8.0 x 10^-8). Conversely, FLT3-ITD(+) animals treated with gilteritinib demonstrated a reduction in HSPC and myeloid lineage cells (40% and 20% r.a., p = 0.01 and p = 0.003, respectively), and an expansion in MSCs (57% r.a., p = 0.01).

MSCs within the human bone marrow are a transcriptionally heterogenous population of cells that support hematopoiesis through diverse mechanisms. Exploratory analysis of the single cell data revealed several potential mechanisms of early gilteritinib resistance. Lepr MSCs were top-expressing cells for growth factors, chemokines, and active signaling molecules such as kitlgb, igfbp2a, cxcl8a, WNTs and Notch ligands. Transcriptional heterogeneity within the MSC populations was noted. For instance, MSCs with strong expression of lepr also correlated with the same subpopulation which expressed jag2a. Conversely, a different subpopulation of MSCs strongly expressed cxcl8a. This is noteworthy, since notch ligands normally signal by direct cell contact, whereas cxcl8a is a soluble factor, and could point towards alternative modalities of niche remodeling. Ligand-receptor pair analysis was also performed and revealed enrichment of nectin2 by MSCs and tigit by HSPCs. While expression of Nectin2/CD112 by HSCs has been reported in the past, this inverse pattern of expression remains a novel, previously unreported interaction. Together, these data suggest that gilteritinib treatment induces expansion of MSCs expressing signaling factors that may lead to therapeutic resistance in FLT3-ITD AML.