YK-4-279 effectively antagonizes EWS-FLI1 induced leukemia in a transgenic mouse model

Abstract

Ewing sarcoma is an aggressive tumor of bone and soft tissue affecting predominantly children and young adults. Tumor-specific chromosomal translocations create EWS-FLI1 and similar aberrant ETS fusion proteins that drive sarcoma development in patients. ETS family fusion proteins and over-expressed ETS proteins are also found in acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) patients. Transgenic expression of EWS-FLI1 in mice promotes high penetrance erythroid leukemia with dense hepatic and splenic infiltrations. We identified a small molecule, YK-4-279, that directly binds to EWS-FLI1 and inhibits its oncogenic activity in Ewing sarcoma cell lines and xenograft mouse models. Herein, we tested in vivo therapeutic efficacy and potential side effects of YK-4-279 in the transgenic mouse model with EWS-FLI1 induced leukemia. A two-week course of treatment with YK-4-279 significantly reduced white blood cell count, nucleated erythroblasts in the peripheral blood, splenomegaly, and hepatomegaly of erythroleukemic mice. YK-4-279 inhibited EWS-FLI1 target gene expression in neoplastic cells. Treated animals showed significantly better overall survival compared to control mice that rapidly succumbed to leukemia. YK-4-279 treated mice did not show overt toxicity in liver, spleen, or bone marrow. In conclusion, this in vivo study highlights the efficacy of YK-4-279 to treat EWS-FLI1 expressing neoplasms and support its therapeutic potential for patients with Ewing sarcoma and other ETS-driven malignancies.

Keywords: ETS fusion proteins; EWS-FLI1; YK-4-279; erythoid leukemia; ewing sarcoma.

Figure 1. EWS-FLI1 is expressed in E/F; Mx1-cre mouse model

A. RT-qPCR comparing EWS-FLI1 mRNA level in spleens of wt (C57BL/6) (n = 3), E/F; control (n = 2), vs. E/F; Mx1-cre (n = 3) mice. B. Western blot showing protein levels of EWS-FLI1 in spleens of E/F; Mx1-cre transgenic mice compared to wt (C57BL/6) mice and E/F; control mice that lack cre required for EWS-FLI1 activation. Anti-FLI1 antibody epitope is retained in EWS-FLI1. Endogenous FLI1 is highly expressed in the normal mouse spleens. Anti-HA antibody was used to detect the protein level of EWS-FLI1 that is HA tagged. Actin was used as a loading control. Each lane represents an individual animal.

Figure 2. CD43+, CD71+, CD117+, and CD45-, cells are enriched in spleens and peripheral blood of mice with EWS-FLI1 induced leukemia

Cells from A. spleen and B. peripheral circulation of mice with EWS-FLI1-induced leukemia vs. control were evaluated for surface antigen expression. The cells were stained with fluorescently labeled CD43, CD45, CD71, and CD117 antibodies, subsequently analyzed by flow cytometry.

Figure 3. Spleen cells of mice with EWS-FLI1 induced leukemia display significantly increased clonogenic potential, which is inhibited by YK-4-279

A. The number of colony forming cells (CFCs) in 1 × 10⁵ spleen cells of leukemic E/F; Mx1-cre vs. E/F; control mice was evaluated in an in vitro colony forming unit (CFU) assay. Effect of YK-4-279 on colony forming potential of cells from EWS-FLI1 induced leukemic mice was assessed in the presence of vehicle (DMSO) vs. 3 μM or 10 μM YK-4-279. Colonies were counted on the 4^th day following plating. ***; p < 0.0001. B. Representative phase contrast images of CFU colonies from each experimental condition are given.

Figure 4. YK-4-279 antagonizes EWS-FLI1 driven erythroleukemia

A. Effect of YK-4-279 on white blood cell count of leukemic E/F; Mx1-cre mice at time of randomization (Day 0), one week after treatment (Day 7), and two weeks following treatment or at the time of euthanasia (Day 14) was assessed. B. and C. Enrichment of erythroblasts (CD43⁺, CD71⁺, CD117⁺, and CD45⁻ white blood cells) in peripheral circulation at time of randomization to YK-4-279 or DMSO treatment (Day 0) and after two weeks of treatment (Day 14) was evaluated. Erythroblast enrichment in the peripheral blood of pIpC injected E/F; Mx1-cre mice was monitored and was compared to healthy E/F; control mice to confirm leukemia induction. Mice were assigned to treatment following an average enrichment of CD43⁺/CD45⁻ and CD71⁺/CD117⁺ WBC to 1.41% and 1.75% compared to 0.02% and 0.19% in E/F; control mice. The amount of erythroblasts in the peripheral blood after the two weeks long treatment regimen with YK-4-279 vs. DMSO was compared to assess efficacy of YK-4-279 in inhibiting EWS-FLI1 driven erythroleukemia. Additionally post-mortem D. spleen, E. liver, and F. body weights of YK-4-279 or vehicle treated mice were compared to those of healthy E/F; control mice of the same age at Day 14 to show the effect of treatment in reducing hepatomegaly and splenomegaly. Representative spleen and liver images of healthy E/F; control mice and DMSO or YK-4-279 treated leukemic E/F; Mx1-cre mice are given. *; p < 0.05, **; p < 0.001, ***; p < 0.0001, ns; not-significant.

Figure 5. YK-4-279 reduced nucleated erythroblast cells in peripheral blood smears of mice with EWS-FLI1 induced leukemia

Representative images of peripheral blood smears of E/F; Mx1-cre mice before pIpC induction are given on the left column. Images of blood smears after one week (middle) and two weeks (right) of treatment with YK-4-279 or vehicle are shown. Induction of EWS-FLI1 expression causes a high white blood cell count originating from an erythroblast expansion. This expansion could be efficiently blocked by YK-4-279 treatment.

Figure 6. Erythroblast infiltration and expansion was efficiently blocked by YK-4-279

E/F; Mx1-cre mice after two weeks treatment with vehicle (middle row) or YK-4-279 (bottom row) were euthanized and spleen and liver samples were processed for histo-pathology analysis. E/F; control mice served as healthy controls (top row). E/F; control mice do not display any symptoms due to the lack of Mx1-cre transgene that allows for cre mediated recombination leading to subsequent EWS-FLI1 activation. Hematoxylin and Eosin (H&E) staining displays disrupted spleen and liver architecture due to massive infiltration and expansion of erythroblasts. Erythroblast infiltration and expansion was efficiently blocked by two weeks YK-4-279 treatment improving splenic architecture (germinal center formation) and restoring normal liver architecture.

Figure 7. Gata1⁺ erythroblasts were effectively diminished in liver and spleen upon YK-4-279 treatment

E/F; Mx1-cre mice after two weeks treatment with vehicle (middle row) or YK-4-279 (bottom row) were euthanized and spleen and liver samples were processed for histo-pathology analysis. E/F; control mice served as healthy controls (top row) since they lack cre required for EWS-FLI1 activation. EWS-FLI1⁺ erythroblasts stain positive for the erythroid lineage differentiation factor Gata1. A much weaker Gata1 staining in the spleen and liver of YK-4-279 treated mice was observed which is indicative of efficient leukemic disease blockade.

Figure 8. YK-4-279 caused little apoptosis in liver or spleen after two weeks of treatment

E/F; Mx1-cre mice after two weeks treatment with vehicle (middle row) or YK-4-279 (bottom row) were euthanized and spleen and liver samples were processed for histo-pathology analysis. E/F; control mice served as healthy controls (top row) which did not display any symptoms of disease or any peculiar apoptosis or proliferation. Cleaved caspase 3 was utilized as a marker of apoptosis for immunohistochemistry. Improved splenic and liver architectures were noted at the analysis time of two weeks after YK-4-279 treatment. At that time, there were no significant apoptosis detectable in the spleen or liver suggesting a healthy organ.

Figure 9. Decreased proliferation was observed in spleens and livers of YK-4-279 treated leukemic mice

E/F; Mx1-cre mice after two weeks treatment with vehicle (middle row) or YK-4-279 (bottom row) were euthanized and spleen and liver samples were processed for histo-pathology analysis. E/F; control mice that lack cre required for EWS-FLI1 activation served as healthy controls (top row) which did not display any symptoms of disease or any peculiar apoptosis or proliferation. Ki67 was used as a marker of proliferation. Ki67 staining of spleen and liver tissues of leukemic E/F;Mx1-cre mice treated with YK-4-279 show a decreased proliferation compared to vehicle treated mice.

Figure 10. YK-4-279 inhibited genes deregulated by EWS-FLI1 in E/F; Mx1-cre mice

A. Mest (Mesoderm specific transcript), Cpne7 (Copine VII), c-Myc, Car8 (Carbonic anhydrase 8), Gata2 (GATA binding protein 2), and Gata1 (GATA binding protein 1) are up-regulated when EWS-FLI1 is expressed through cre mediated recombination of Rosa26-loxP-STOP-loxP-EWS-FLI1 site in E/F; Mx1-cre mice. The expression of these genes was significantly reduced following two weeks of treatment with YK-4-279 (n = 12) over control (n = 12). B. EWS-FLI1 level did show significant change after two weeks of treatment with YK-4-279 at mRNA and protein level. Each lane in the western blot represents an individual animal. *; p < 0.05, **; p < 0.001, ***; p < 0.0001, ns; not-significant.

Figure 11. YK-4-279 improved overall survival and disease burden of mice with EWS-FLI1 driven erythroleukemia

A. Treatment of the transgenic mice daily, five times a week for the first 2 weeks with 150 mg/kg YK-4-279 or vehicle improved overall survival of mice with EWS-FLI1 induced leukemia (p = 0.0003). B. WBC counts at time of randomization to treatment (Day 0), one week (Day 7), and two weeks (Day 14) following randomization to treatment are given. YK-4-279 treatment regimen led to significant reduction of WBC count. Disease profile of C. vehicle vs. D. 150 mg/kg YK-4-279 treated individual E/F; Mx1-cre mice is presented as weekly WBC counts. **; p < 0.001, ***; p < 0.0001, ns; not-significant.