ALBERT

Description: Abstract 3524 Background: Although Notch signaling contributes to T cell leukemogenesis, the role of Notch in human AML is unclear. We and others have found that activation of Notch signaling inhibits AML growth and survival, e.g. a tumor suppressor like effect. However it is not known what the consequences of activating or inhibiting Notch signaling are in human AML in vivo. Approach: To determine whether Notch signaling would have growth inhibiting effects in vivo, we stably-transduced ML1 human AML cells with the constitutively-active forms of Notch1 and Notch2 (ICN1, ICN2), the common Notch target gene Hairy/Enhancer of Split 1 (HES1) or the pan-Notch inhibitor dominant-negative Mastermind-like (dnMAML) and performed in vivo competitive proliferation assays. Briefly, following transduction, each vector type were sorted to 50% GFP+ (containing the gene of interest) and GFP- (untransduced control cells). Groups of NSG mice were injected with these 50:50 mixtures of ML1 cells, and peripheral blood levels of GFP- and GFP+ cells were measured with flow cytometry for anti-human CD45 and GFP. Results: Peripheral blood engraftment of human CD45+ ML1 cells by week 5 was similar (2–5%) for ICN1, ICN2 and HES1 injected mice, but was significantly higher (23%) in dnMAML injected mice (Panel A). Similar to in vitro competitive proliferation assays, ICN1, ICN2 and HES1 all led to decreased relative numbers of GFP%+ cells, with 1%, 4% and 16% respectively (Panel B). Importantly, dnMAML had little effect on AML proliferation in vitro, however led to dramatic increases in GFP% as well as early morbidity and mortality due to increased leukemia burden, with 93% GFP+ (Panel B), demonstrating a selective advantage for dnMAML-expressing ML1 in vivo. When GFP+ (transduced) ML1 peripheral engraftment was directly compared to GFP- (parental CD45+) engraftment, the GFP- control cells had similar engraftment rates (2–5%) across groups of mice, while the GFP+ engraftment rates were significantly lower in ICN1, ICN2, and HES1 groups, but significantly higher in the dnMAML group (Panel C), demonstrating enhanced engraftment/proliferation in dnMAML-expressing cells. Conclusions: This suggests a previously unreported concept, namely that endogenous Notch ligands can inhibit human AML growth in vivo. This data supports the hypothesis that Notch behaves as a tumor suppressor in AML, and suggests the potential use of Notch agonists in human AML. Disclosures: No relevant conflicts of interest to declare.

Description: Background Notch is a well-known oncogene in T-ALL, yet appears to have tumor suppressor effects in B-ALL. These cell type-specific effects of Notch signaling mirror consequences seen in early lymphocyte development and raises the question of how Notch leads to such divergent consequences in closely related cell types. In exploring these Notch mechanisms we discovered a B-ALL specific Notch-mediated reduction in the cell cycle regulator Polo-like kinase-1 (PLK1), revealing a novel targetable kinase in B-ALL. Approach To explore the consequences of Notch-mediated down regulation of cell cycle regulator kinase PLK1, we targeted PLK1 kinase function with the novel PLK1-selective inhibitor poloxin in human B-ALL lines. Results PLK1 is highly expressed in B-ALL verses normal tissues (panel A), correlates with cyclin B expression, is expressed 〉2-fold higher in B-ALL with t(1;19) than other B-ALL samples, and may predict response of ALL to methotrexate. In our panel of human B-ALL cell lines poloxin induced G2/M growth arrest and decreased cell number by 〉80% (panel B), and decreased survival in B-ALL cells (〉75% AnnexinV+, panel C). PLK1 inhibition led to tumor suppressor p53 stabilization, revealing 〉5-fold increase in p53 protein levels following poloxin treatment in B-ALL (panel D). Mechanistically, PLK1 inhibition leads to both cytoplasmic re-localization of cyclin B, disrupting the CDC2-cyclinB complex, as well as phosphorylation of p53 at Ser20, which destabilizes p53-MDM2 interaction and thus accumulation of p53. Conclusions While exploring the mechanisms of cell type-specific effects of Notch signaling in ALL, we have found a novel therapeutic target, the cell cycle regulator PLK1. Our findings reveal a novel therapeutic approach whereby PLK1-selective inhibition via poloxin induces growth arrest and apoptosis in human B-ALL via consequences on cyclin B and p53 pathways. Disclosures: No relevant conflicts of interest to declare.

Description: Evasion of apoptosis is a key hallmark of cancer. BCL-2 family proteins, the central regulators of apoptosis, are often aberrantly expressed in tumors. Pro-apoptotic BCL-2 members bind and sequester anti-apoptotic BCL-2 proteins via their BH3 domains. Thus, BH3 mimetics represent a promising direction in cancer drug development. ABT-263, designed as a BH3 mimetic to inhibit BCL-2, BCL-XL, and BCL-W, has demonstrated efficacy in preclinical and clinical studies. However, thrombocytopenia is common in patients treated with ABT-263 due to the inhibition of BCL-XL, which is indispensable for platelet survival. ABT-199 (GDC-0199), a second-generation BH3 mimetic, has higher affinities for BCL-2 protein (Ki 〈 0.01 nM), which enhances the specificity of this agent to kill cancer cells without provoking unwanted thrombocytopenia (Souers, et al, Nature Med, 2013). Since BCL-2 is often overexpressed in hematological malignancies including acute myeloid leukemia (AML), we evaluated the anti-cancer effects of ABT-199 on AML cells. As a measure of BCL-2 specificity, BCL-XL overexpression in sensitive HL-60 cells resulted in complete resistance to ABT-199, while BCL-2 overexpression in these cells conferred moderate resistance to apoptosis induction. Moreover, OCI-AML3 cells with high MCL-1 levels were highly resistant to ABT-199, while knockdown of this protein greatly sensitized cells to this BH3 mimetic. Among 12 genetically diverse AML cell lines tested, seven were sensitive to ABT-199-induced apoptosis with 48-h EC50 ranging from 1.5 nM to 145 nM. In these seven sensitive, BCL-2 dependent cell lines, ABT-199 was uniformly more potent than ABT-737 (another BCL-2 inhibitor with a spectrum similar to ABT-263, p = 0.016). Next, we tested ABT-199 in 15 primary samples from relapsed/refractory AML patients. Twelve patient samples showed high sensitivity to apoptosis induction following 48-h exposure to ABT-199 (EC50 〈 10 nM). In a larger set of 23 cryopreserved AML patient samples, including AML cells with diploid cytogenetics and mutations in FLT3, NRAS, and NPM1 genes, 18 (78%) were sensitive to ABT-199 (100 nM). However, samples from patients with complex cytogenetics, t(8;21) and JAK2 mutation (n = 12) were largely insensitive to ABT-199 (17% response rate). Interestingly, in five of six primary AML samples with high blast counts, ABT-199 induced marked apoptosis in CD34+/CD38- AML stem/progenitor cells compared to bulk AML blasts (p = 0.01). Quantitative Western blot was used to determine BCL-2 protein levels in AML cell lines. Spearman analysis showed that EC50 of ABT-199 correlated negatively with BCL-2 protein expression (r = -0.605, p = 0.0143) and correlated positively with BCL-XL protein expression (r = 0.633, p = 0.0101). Similar correlations were also observed in primary AML samples, suggesting that pre-treatment cellular BCL-2 and BCL-XL levels might have utility as predictive markers of ABT-199 sensitivity. We next examined the in vivo anti-leukemic efficacy of ABT-199 in NOD SCID gamma (NSG) mice injected with luciferase-labeled MOLM-13 cells. The mice were treated with ABT-199 by daily oral gavage (a 2-wk treatment at dose of 100 mg/kg). Bioluminescence imaging showed that ABT-199 treatment significantly inhibited leukemia burden, which was also manifested by smaller spleen size and prolonged overall survival (p = 0.0004) when compared to the vehicle-treated mice. Furthermore, a 2-wk ABT-199 treatment significantly reduced leukemia burden (〉 50%) in bone marrows of NSG mice engrafted with primary FLT3-mutated AML cells (i.e., a mean of 70 ± 16% human CD45+ cells in bone marrow of control mice (n = 9) versus 32.7 ± 12% in ABT-treated mice (n = 11), p = 0.00002). Conclusions: the in vitro and in vivo efficacy data indicates that ABT-199 is a selective BCL-2 inhibitor, a potent apoptogenic agent, and hence a promising candidate for AML BCL-2-targeted therapy. Disclosures: Leverson: AbbVie, Inc.: Employment, Equity Ownership. Konopleva:AbbVie: Research Funding.