A Small Molecule Inhibitor of the CBFβ-SMMHC/RUNX Interaction Attenuates Inv(16) Leukemia in Vivo

Abstract 286

Emerging literature suggests that inability to cure cancers with current therapies may be attributed to a population of so-called cancer stem cells or cancer initiating cells that have long term self-renewal potential and can fully recapitulate tumor phenotype at time of relapse. The use of targeted therapies that inhibit the activity of oncoproteins directly present the potential for specificity in ablating tumor cells, including cancer stem cells, while having minimal impact on normal tissues.

The gene encoding CBFβ (CBFB) is disrupted by the chromosome 16 inversion [inv(16)(p13q22)], associated with ∼12% of acute myeloid leukemia (AML) in humans, resulting in a fusion protein containing most of CBFβ fused to the coiled-coil tail region of smooth muscle myosin heavy chain (SMMHC). The CBFβ-SMMHC fusion protein acts as a dominant repressor of CBF function, binding RUNX1 and dysregulating the expression of multiple target genes required for normal hematopoiesis. Current AML treatment utilizing cytotoxic chemotherapy results in 45–65% five year overall survival but only 20% for patients older than 60. These data clearly indicate that the development of targeted therapy that can improve the therapeutic response for inv(16) AML patients is essential.

The targeting of transcription factors in cancer therapy is a relatively new approach with tremendous potential. We developed a small molecule inhibitor, AI-10-49, that binds to the aberrant transcription factor CBFβ-SMMHC and disrupts its interaction with RUNX proteins. We have taken advantage of the oligomeric nature of CBFβ-SMMHC by developing dimeric compounds to achieve selectivity toward the fusion protein. Optimization of the linker length in these dimeric compounds shows a clear linker length dependence and a dramatic enhancement in activity for the dimeric versus monomeric inhibitors, providing validation for this approach. AI-10-49 is a result of this optimization and modifications to improve pharmacokinetic properties, resulting in a potent inhibitor with a half-life of ∼3 hours in mice. This compound displays selective toxicity to human leukemia cell lines with the inv(16) at a submicromolar dose (IC₅₀=0.4μM). We have also shown no effect on growth of normal mouse or human bone marrow mononuclear cells. Using RT-PCR, we have shown dramatic derepression of the well-validated target genes CSF1R (9-fold) and RUNX3 (18-fold) in ME-1 (inv(16)) cells and little to no effect in Kasumi-1 (t(8;21)) and U937 cells. We have also shown increased apoptosis of the mouse preleukemic myeloid progenitor cells expressing CBFβ-SMMHC upon treatment ex vivo.

The toxicity of AI-10-49 was tested in mice at the 25 mg/kg dosage employed for efficacy studies. Mice were dosed twice a day at 12 hour intervals for 1 week. No change in counts of white blood cells, red blood cells, or platelets was observed compared to formulation control. No changes were observed in granulocytes, monocytes, T cell, and B cell frequency by flow cytometry. Analysis of tissues after treatment showed no defects in spleen, kidney, liver, lungs, GI tract, heart, and bone marrow. A maximum tolerated dose (MTD) study showed no toxicity at doses up to 500 mg/kg, indicating the compound is very well tolerated.

The efficacy of AI-10-49 was tested in vivo using the Cbfb^+/MYH11/Mx1Cre/Nras^G12D mouse model of inv(16) AML. Mice were transplanted with 5×10⁵ leukemic cells, allowed 5 days for engraftment, and treated between days 5 and 15 post transplantation with 2 doses of vehicle or of AI-10-49 per day (25mg/kg/day). The control group (n=4) developed leukemia with a median latency of 4.6 weeks and full penetrance, while the test group (n=4) developed leukemia with a median latency of 8.4 weeks and incomplete penetrance. One mouse (25%) from the test group showed progressive decrease of leukemic cells to background levels at week 12 in peripheral blood, and remains healthy with no evidence of disease to date (week 16). Current efforts are focused on testing the compound efficacy of AI-10-49 with first line therapy drug Ara-C in mice.

This work presents strong evidence of the identification of a small molecule that specifically targets CBFβ-SMMHC function and ablates cells expressing the leukemia fusion protein in vitro and in mice. This study provides proof of principle that oncogenic transcription factors can be targeted and may have significant impact on the treatment of inv(16) acute myeloid leukemia.