Evaluation of Combinational Therapy of MDM2-Antagonist Nutlin-3 and HDAC-Inhibitor Valproic Acid in Acute Myeloid Leukemia in Vitro and in Vivo

Nutlin-3 is a small-molecule antagonist of MDM2 that induces non-genotoxic stabilization and activation of the tumor suppressor protein p53, resulting in therapeutic effects in tumor models comprising wild type TP53. Valproic acid (VPA) is an anti-convulsive drug with histone deacetylase (HDAC) inhibitor activity that induces differentiation and apoptosis in acute myeloid leukemia (AML) cells. Both therapeutic compounds indirectly affect the regulation of p53; nutlin-3 by inhibiting MDM2, the main negative regulator of p53, and VPA by inhibiting HDACs that participate in p53 deacetylation and destabilization. While mutations in TP53 occur in less than 10% of AML, over-expression of MDM2 is frequently observed. Furthermore, aberrant recruitment of histone deacetylases (HDACs) is seen in AML, leading to block of myeloid differentiation. We therefore hypothesized that concomitant inhibition of MDM2 and HDACs would synergistically induce p53-mediated apoptosis and inhibit tumor growth. We examined the anti-leukemic effects of nutlin-3 in combination with low doses valproic acid in AML cell lines, primary AML cells, and in an in vivo xenograft model using optical imaging. To evaluate the efficacy of the combination of nutlin-3 and VPA in AML cells expressing wild type TP53, the AML cell line MOLM-13 was treated with various concentrations of the two drugs both alone and in combinations (nutlin-3; 0.5–10 μM, VPA; 50–1000 μM, nutlin-3:VPA; 1:100) over different time periods (24–72h, nutlin-3 only for the 24 last hours). Synergistic or additive effects were detected in three distinct viability assays; 3H-thymidine incorporation was used to examine effect on proliferation, WST-1 was used to determine number of metabolic active cells in culture, and DNA specific staining with Hoechst 33342 was used to determine apoptosis after drug-treatment. In addition, MOLM-13 cells treated with the combination showed super additive induction of p53 and target genes. The optimal combination and time period found in MOLM-13 cells were tested in 40 different primary AML samples using the three different viability assays. Toxicity of the combination treatment was tested in normal peripheral blood lymphocytes, while preliminary toxicity of nutlin-3 and VPA alone and in combination on healthy NOD/SCID IL2γnull mice permitted determination of treatment regime. We developed an optical imagable model for in vivo evaluation of the combinational therapy by injecting NOD/SCID IL2γnull mice with MOLM-13 cells transfected with a tetracycline activated luciferase expressing construct (termed L192). Bioluminescent imaging was performed using a TD-SAMI (Time-domain small animal molecular imager). The efficacy of the combinational therapy was tested in 20 NOD/SCID IL2γnull mice injected with MOLM-13 L192 cells and divided into four groups; control, VPA (50 mg/kg b.i.d), nutlin-3 (200 mg/kg b.i.d) and nutlin-3 + VPA (200 mg/kg + 50 mg/kg) b.i.d. Treatment was scheduled for three weeks. The combination of nutlin-3 and VPA significantly inhibited disease development after one week, as monitored by in vivo imaging. Limitations due to nutlin related toxicity prevented further evaluation of continuous combinational therapy after 14 days. However, all treatment groups showed a significant increase in survival compared to the control group, with the combination group demonstrating decreased leukaemic burden as visualized by optical imaging and longer mean average survival time. VPA effect on survival was also tested in a BNML rat leukemia model, in which VPA-treatment (170 mg/kg) resulted in significant longer mean survival compared to the control. Together, the results suggest combined targeting of MDM2 and HDACs as a promising therapeutic approach in AML. Future studies will apply the established bioluminescent MOLM-13 AML xenograft model for further evaluation of the combinational therapy, using a different dosing regimen and scheduling. In addition, we will evaluate combinations of differing classes of HDAC inhibitors and MDM2 antagonists.