Oncogenic role and therapeutic targeting of ABL-class and JAK-STAT activating kinase alterations in Ph-like ALL

Abstract

New therapies for Philadelphia chromosome-like acute lymphoblastic leukemia (Ph-like ALL) patients are urgently needed. The genetic landscape of Ph-like ALL is characterized by a diverse array of kinase-activating alterations (including rearrangements, sequence mutations, and copy number alterations), suggesting that patients with Ph-like ALL are candidates for targeted therapy, similar to BCR-ABL1 ALL. We sought to investigate the functional role and targetability of the spectrum of kinase-activating alterations identified in Ph-like ALL. We demonstrate cytokine-independent growth and activation of JAK-STAT signaling pathways in Ba/F3 cells by all alterations tested. The development of murine Arf^-/- pre-B ALL expressing RCSD1-ABL2 or SSBP2-CSF1R was accelerated with the presence of IK6, a dominant negative isoform of Ikaros common in Ph-like ALL, providing evidence that these fusions are leukemogenic. In vitro screening using a panel of tyrosine kinase inhibitors against 14 different kinase alterations identified the ABL1-inhibitor, dasatinib, as a potent inhibitor of ABL-class fusions (ABL1, ABL2, CSF1R, PDGFRB), whereas the JAK1/JAK2 inhibitor ruxolitinib, was most effective against JAK-STAT-activating alterations (JAK1, JAK2, JAK3, IL7R, IL2RB), but not TYK2. Evaluation of dasatinib or ruxolitinib against patient-derived xenograft models demonstrated superior antileukemic efficacy when combined with dexamethasone compared with either agent alone. These data provide the foundation for rationally designed clinical trials that assess the efficacy of targeted therapy in patients with Ph-like ALL.

Graphical abstract

Figure 1.

Expression of 5′ and 3′ fusion partner genes during normal ontogeny. Unsupervised clustering of gene expression data from (A) mouse and (B) human hematopoietic cells. Kinase genes are highlighted in red. CLP, common lymphoid progenitor; CMP, common myeloid progenitor; DC, dendritic cell; GMP, granulocyte monocyte progenitor; HPC, hematopoietic progenitor cell; HSC, hematopoietic stem cell; MEP, megakaryocyte-erythroid progenitor; mono, CD14+ monocyte; NK, natural killer.

Figure 2.

Cytokine independence and constitutive activation of signaling pathways in Ba/F3 cells. (A) Kinase alterations investigated in this study. Red and yellow shapes denote sequence mutations. (B) Growth of Ba/F3 cells in the absence of IL-3. Cell number was determined using trypan blue. Each point represents the mean ± standard deviation (SD; n = 3). (C) Phosphoflow cytometric analysis of signaling phosphoproteins in Ba/F3 cells. Cells were grown in the absence of IL-3, harvested, and assessed for the phosphorylation of pSTAT5, pCRKL, pERK1/2, and PS6. A-J, ATF7IP-JAK2; E-N, ETV6-NTRK3; KD, kinase domain; M-I, MYH9-IL2RB; P-A, PAG1-ABL2; P-J, PAX5-JAK2; R-A, RCSD1-ABL1; R-ABL2, RCSD1-ABL2; RAN-A, RANBP2-ABL1; S-C, SSBP2-CSF1R; Z-A, ZMIZ1-ABL1.

Figure 3.

Expression of RCSD1-ABL2 and SSBP2-CSF1R in Arf^−/−pre-B cells induces B-ALL in C57Bl/6 mice. (A) Immunoblot of Arf^−/− pre-B cells showing expression of ABL2, CSF1R, IKAROS, and actin. (B) Kaplan-Meier curve of sublethally irradiated C576Bl/6 mice receiving Arf^−/− pre-B cells by tail vein injection (5 × 10⁵ cells/mouse). (C) Spleen weight of C576Bl/6 mice transplanted with Arf^−/− pre-B cells (n = 5). MIG, MSCV-ires-GFP. (D) Bone marrow, spleen, and brain sections from a representative mouse transplanted with Arf^−/− pre-B cells stained with hematoxylin and eosin (H&E) or a B220-specific antibody. (E) Analysis of bone marrow of diseased animals by flow cytometry. Immunophenotype is consistent with arrest at the pre-B stage of development.

Figure 4.

In vitro sensitivity of Ba/F3 cells to tyrosine kinase inhibitors. Response of 14 Ba/F3 cell lines expressing Ph-like ALL alterations to targeted agents. Cell viability was measured after 48 hours using CellTitre-Blue viability assay. IC₅₀ values were calculated using serial dilutions and linear regression. The heat map shows IC₅₀ values (log₁₀ scale) according to the color key. Cell lines (columns) and drugs (rows) were ordered by unsupervised clustering.

Figure 5.

Inhibition of signaling pathways by tyrosine kinase inhibitors in Ba/F3 cells. (A) Ba/F3 cells were treated with dasatinib or ruxolitinib for 1 hour, harvested, and assessed for phosphorylation of STAT5 and CRKL. Heat map indicates mean fluorescent intensity and was generated using Cytobank. (B) Phosphorylation of extracellular signal-regulated kinase 1/2 in Ba/F3-ETV6-NTRK3 cells at basal levels or after treatment with crizotinib, dasatinib, or ruxolitinib (1 hour). (C) Dose-response curve of Ba/F3 cells after 48 hours’ treatment with JAK3i. Values are normalized to dimethyl sulfoxide controls for each cell line and represent mean ± SD (n = 3). (D) Phosphorylation of STAT5 in Ba/F3 cells at basal levels or after JAK3i treatment (1 hour).

Figure 6.

Dasatinib is effective in vivo against ABL-class fusions. (A) Levels of human CD45/CD19⁺ blasts in the peripheral blood of NSG mice injected with primary B-ALL cells. (B) Spleen weight of NSG mice at harvest. (C) NSG mice were injected with primary leukemic cells. Upon peripheral engraftment of 5% human CD45, animals (n = 5) were treated with vehicle, dasatinib (20 mg/kg per day), dexamethasone (4 mg/mL ad libitum), or dasatinib and dexamethasone combined. Treatment length is indicated by the gray shaded area; animals were euthanized at the last time point. Analysis of slope was measured by linear regression and unpaired 2-way t test comparing vehicle with dasatinib (ALL1, ALL9, ALL13), or 1-way analysis of variance (ANOVA) with Tukey posttest for multiple comparisons (ALL14). (D) Spleen weights were measured at the time of euthanasia. Each point represents the mean ± SD (n = 5). (E) Representative H&E and human CD45 staining of bone marrow and brain harvested from vehicle or dasatinib-treated mice of ALL1. *P < .05, ***P < .001, ****P < .0001. Das, dasatinib; dex, dexamethasone.

Figure 7.

Superior efficacy of ruxolitinib combined with dexamethasone in JAK2-rearranged ALL. NSG mice were injected with primary leukemic cells (A) ALL19 or (B) ALL21. When engraftment reached 10¹⁰ p/sec per cm²/steridian on bioluminescent imaging, animals (n = 5) were treated with vehicle, ruxolitinib (provided as chow), dexamethasone (4 mg/mL ad libitum), or ruxolitinib and dexamethasone combined. Treatment length is indicated by the gray shaded area; animals were euthanized at the last time point. Analysis of slope was measured by linear regression and 1-way ANOVA with Tukey posttest for multiple comparisons. Spleen weights were measured at the time of euthanasia. Each point represents the mean ± SD (n = 5). Representative H&E staining of bone marrow and spinal cord of treated mice are shown. *P < .05, **P < .01, ***P < .001, ****P < .0001. Ruxo, ruxolitinib.