Efficacy of ABT-737, a BCL-2 Inhibitor, in an NRAS/BCL2 Mouse Model of High Risk Myelodysplasia (HR-MDS) By Targeting Pathways Identified By Gene Expression Profiling

Background: MDS lack therapeutic approaches in many instances and very few animal models of MDS available for preclinical testing of new drugs have been reported. We have created transgenic animal models of HR-MDS and acute myelogenous leukemia(AML) post-MDS using mutant NRAS and BCL-2 overexpression (Omidvar Cancer Res.2007;67:1657; Rassool Cancer Res. 2007;67:8762) and reported the efficacy of azacitidine in the HR-MDS model (Gorombei Blood 2013;122:1514) and of the BCL2 inhibitor, ABT-737, on the AML post-MDS model (Beurlet Blood 2013;122:2864). The objective of this study was to determine the effects of this inhibitor on our mouse model of HR-MDS. Methods: The mice were followed for disease and once the platelets fell (<1000x10³/mm³) at around 3 months, ABT-737 (75mg/kg) was administered every other day for 15 injections. Responses were measured by survival, peripheral blood (PB) counts and Mac1hi/Gr1lo estimations by flow cytometry (FCM) as a measure of circulating blasts. A subset were sacrificed and examined for BM myeloid colony forming units (CFU), lineage negative/ Sca1+/KIT+ (LSK) by FCM and secondary transplants for leukemia initiating cells (LICs). Apoptosis was measured using technicium (^99mTc)-labelled annexin V and imaging by SPECT and TUNEL of liver sections. The RAS:BCL-2 complex was visualized by pull down assays and immunofluorescence confocal microscopy. GEP was conducted on Sca1+ enriched spleen cells from the diseased mice before and after treatment using exon specific arrays. Results: ABT-737 significantly extended survival of the HR-MDS mice from a median survival of 25 days in untreated (n=25) to 80 days in treated mice (n=31, p<0.0001), with reductions of BM blasts (median 6± 2.8% in treated, 5± 2.1% in untreated, p<0.05) and LSK cells (1.8±0.6% normal; 8.5± 3.6% untreated, 4.5± 2.4 in treated, p<0.05). Colony growth was restored to near normal range (33± 10 in normal mice, 51± 10 in untreated, 38± 8 in treated mice). Secondary transplants of spleen cells from untreated mice showed deaths within 50 days; treated cells remained alive for 90-115 days (p<0.05). Increased apoptosis was observed by SPECT and TUNEL of liver sections. BCL2 expression and RAS activity decreased in treated HR-MDS while they increased in our treated AML post-MDS mice (Beurlet Blood 2013;122:2864). The RAS:BCL-2 complex remained in the plasma membrane in the HR-MDS mice whereas it localized from the mitochondria to the plasma membrane in the AML mice. Exon specific GEP of HR-MDS mice showed 1952 differentially regulated genes upon treatment, including genes important for the regulation of stem cells, differentiation, proliferation and apoptosis. When compared with the signatures of the AML post-MDS arrays of 974 genes regulated post treatment, 143 genes in common were regulated, identifying potential targets of ABT- 737; of these, 133 were coordinately regulated and 10 were discordantly regulated. Conclusion: The BH3 mimetic inhibitor ABT-737 is effective in our animal model of HR-MDS, as in our AML model, but the majority of genes regulated by ABT-737 were not common to both mouse models, reflecting differences in the biology of the different stages of these diseases. Given the importance of BCL2 overexpression in the progression of MDS (Parker Blood 2000;96:393) this drug and other BCL2 inhibitors, including the more specific ABT-199 that is showing promise in clinical trials in CLL and lymphoma (Souers Nat.Med.2013;19:202) may prove effective in MDS.