Modeling Gilteritinib Resistance By Enforced cxcl8a Expression in Zebrafish HSPCs

In Acute Myeloid Leukemia (AML), the FMS-like tyrosine kinase III Internal Tandem Duplication (FLT3-ITD) is common and increases the risk of relapse which is only partially mitigated by tyrosine kinase inhibitor (TKI) therapy. Gilteritinib is the current mainstay for treating relapsed FLT3-ITD AML, but therapy is non-curative, and the duration of remission is transient. Mutations in the RAS pathway are known to promote gilteritinib resistance, however non-mutational mechanisms of resistance remain under-explored and could provide new avenues for combination therapy. Our previous work has associated AML cell-autonomous expression of chemokines such as CXCL8 (IL-8) as a mechanism of resistance to gilteritinib therapy. Here we enforce expression of cxcl8a (cxcl8, ENSDARG00000104795) by hematopoietic stem and progenitor cells (HSPCs) in transgenic zebrafish embryos using a defined Runx1 enhancer element. In the absence of gilteritinib, Runx1:cxcl8(+) animals had greater numbers of HPSCs compared to Runx1:cxcl8(-) animals (4.0 vs 2.1, p = 0.04). When treated with gilteritinib at 100 nM from 48-72 hours post-fertilization (hpf), HSPC numbers were decreased in both groups; however, Runx1:cxcl8(+) animals still had more HSPCs compared to Runx1:cxcl8(-) animals (1.6 vs 0.4, p = 0.02). This suggests that cxcl8 may act as a resistance factor for gilteritinib in vivo and that understanding regulators of cxcl8 expression may provide insight into gilteritinib resistance. Reanalysis of published single cell RNA-seq data in AML patients revealed an association between expression of CXCL8 and BTK and BMX but not other Tec-family kinases such as ITK and TEC itself. Zebrafish do not have an annotated BMX gene; BTK is expressed in HSPCs and neutrophils but not the erythroid lineage. CHMFL is a specific and potent inhibitor of human BMX with an unknown spectrum of activity in zebrafish. In Runx1:cxcl8(-) Runx1:GFP zebrafish, CHMFL treatment reduced HSPC numbers, but this was not rescued in Runx1:cxcl8(+) animals (untreated vs treated, cxcl8(-): 2.9 vs 1.5, p = 0.02, untreated vs treated, cxcl8(+): 4.0 vs 1.4, p = 0.0006). This contrasts with our prior work with the non-specific Tec family kinase inhibitor, ibrutinib, where HSPC numbers were rescued in Runx1:cxcl8(+) animals. Ongoing studies are focused on understanding the targets of ibrutinib and CHMFL in the zebrafish and how these pathways may affect gilteritinib resistance.