ALBERT

Description: Introduction: p53 is an important tumor suppressor, and loss of p53 pathway function is a common event in human cancer. ASPP2 is a damage-inducible p53 binding protein that enhances apoptosis, at least in part, by selective stimulation of p53 pro-apoptotic target genes. Low ASPP2 expression occurs in many human tumors and as we have previously demonstrated, correlates with poor clinical outcome in patients with B-cell lymphomas. However, the mechanisms by which ASPP2 suppresses tumor formation remain unknown. Methods: To rigorously study ASPP2 in vivo function, we targeted the ASPP2 allele in a mouse model by homologous recombination by replacing exons 10–17 with a neoR gene. Aging mice were followed for spontaneous tumor formation. Additionally, six week old mice were irradiated at different doses and followed for tumor development. To explore the functional consequences of low ASPP2 expression, ASPP2+/− and ASPP2+/+ thymocytes were subjected ex vivo to 5Gy ionizing irradiation and apoptosis was assessed by Annexin-V/PI staining and flow cytometry. We also irradiated ASPP2+/+ and ASPP2+/−MEFs (mouse embryonic fibroblasts) with 5Gy and performed cell cycle analysis. Furthermore, we conducted global gene expression profiling between the unirradiated and irradiated ASPP2+/+ and ASPP2+/−MEFs using an Affymetrix GeneChip® Mouse Gene 1.0 ST Array. Moreover, phosphoproteomic analysis was performed on unirradiated and irradiated ASPP2+/+ and ASPP2+/−MEFs using 2-dimensional gel electrophoresis, fluorescent phosphoprotein dye Pro-Q Diamond staining, and liquid chromatography tandem mass spectroscopy. Results: We were unable to generate ASPP2−/− mice due to an early embryonic lethal defect. However, ASPP2+/− mice appear developmentally normal and reproduce. We observed an increased formation of spontaneous tumors in aging ASPP2+/− mice compared to ASPP2 +/+ mice (43% vs. 22%, at 115 weeks, p=0.011). Additionally we were able to show that after ionizing radiation, ASPP2+/− mice develop high-grade lymphomas in a dose-dependent manner at a significantly higher incidence at 50 weeks compared to ASPP2+/+ mice (p = 0.024 and p = 0.045, 6 Gy and 10.5 Gy respectively). Immunophenotyping demonstrated that these were high-grade T-cell lymphomas of thymic origin. Since ASPP2 is damage-inducible and promotes apoptosis, we wished to determine the extent to which reduction in ASPP2 expression attenuated the cellular damage response. We therefore gamma-irradiated ex vivo ASPP2+/+ and ASPP2+/−thymocytes in short-term culture and found a two-fold reduction in apoptosis in ASPP2+/− thymocytes compared to ASPP2+/+ thymocytes. Additionally, after 5 Gy gamma-irradiation, ASPP2+/− MEFs exhibited an attenuated G0/G1 checkpoint compared to ASPP2+/+ MEFs. Preliminary analysis of global gene expression in ASPP2+/+ and ASPP2+/− MEFs shows specific differences in gene expression patterns after damage. Likewise, preliminary analysis of phosphoproteomics between ASPP2+/+ and ASPP2+/− MEFs after damage, demonstrate differences in the phosphophorylation status of 170 proteins. Conclusions: ASPP2 is a haploinsufficient tumor suppressor, and reduction in ASPP2 expression attenuates both cell cycle checkpoints and apoptosis-induction after damage. These results suggest that reduction of ASPP2 levels modulate the cellular damage response, possibly at transcriptional as well as post-transcriptional levels, and provide a functional explanation for the increased tumor incidence in ASPP2+/− mice---since attenuated damage-response thresholds would lead to an impaired ability to eliminate cells that have accumulated tumorigenic mutations. Our study provides important insights into the p53-ASPP2 axis, and opens new avenues for investigation into its role in tumorigenesis and response to therapy.

Description: Activating mutations of the KIT, FLT3 and ABL tyrosine kinases (TKs) are found in association with hematologic malignancies including subsets of acute leukemias, systemic mastocytosis, and chronic myelogenous leukemia. TK inhibitors (TKIs) have proven clinical activity for treatment of these diseases, but responses are largely partial or incomplete. We and others have previously demonstrated that the small molecule TKIs imatinib (KIT, BCR/ABL), dasatinib (KIT, BCR/ABL), sunitinib (KIT, FLT3), and tandutinib (KIT, FLT3) are potent inhibitors of oncogenic kinases relevant to hematologic malignancies. We now report that inhibition of TK phosphorylation (a surrogate for kinase activation), potently silences STAT and MAPK1/2 signaling - but incompletely inhibits activation of the AKT pathway: Using mastocytosis, leukemia and hematopoietic cell lines with endogenous or exogenous mutant KIT (V560G, D816V), FLT3 (ITD, D835V) or BCR/ABL isoforms, oncogenic TK-dependent downstream pathways were evaluated by immunoblotting. In addition, we determined the effect of treatment with specific TKIs +/– rapamycin on cellular proliferation and the induction of apoptosis. (Auto)phosphorylation of KIT, FLT3 or BCR/ABL results in activation of downstream pathways important for cell viability and cell survival. This includes the MAP Kinase (MAPK), STAT and PI3K/AKT pathways. All tested TKIs were able to potently decrease phosphorylation of members of the MAPK1/2, STAT3 and AKT pathways in cells harboring a corresponding TKI-sensitive oncogenic kinase. Notably, TKI treatment abolished phosphorylation of AKT at the Ser-473 site, whereas the Thr-308 phospho-site remained constitutively phosphorylated. This partial inhibition of AKT activity was insufficient to completely inhibit downstream kinases, such as p70S6Kinase, a known substrate of mTOR (Target Of Rapamycin). The addition of low nanomolar doses of rapamycin completely abrogated p70S6K phosphorylation alone or in combination with TK-inhibition. To determine the biological effect of more completely inhibiting AKT-dependent signaling, we tested TKI in physiological concentrations in combination with low dosage rapamycin (5–20nM). While rapamycin had only weak anti-proliferative effects when administered as monotherapy, combining rapamycin with each TKI demonstrated a strong additive to synergistic inhibitory potency with an up to 4 fold increase of antiproliferative as well as enhanced pro-apoptotic effects when compared to TKI monotherapy. We conclude that TKI inhibition of mutant KIT, FLT3 or ABL signaling is insufficient to globally silence AKT-driven pathways that may contribute to therapy failure. The addition of inhibitors targeting the AKT pathway, such as rapamycin, may improve clinical efficacy of TKI.

Description: Activating mutations in the juxtamembrane domain or the activation loop of the receptor tyrosine kinase FLT3 occur in many cases of acute myeloid leukemia (AML), but it is not known whether genomic alterations outside these regions contribute to leukemogenesis. High-throughput DNA sequence analysis has provided insights into the mutational profiles of various cancers and represents a promising strategy for the identification of novel therapeutic targets. However, recognizing the subset of genomic alterations that are functionally relevant has proven difficult. We used a high-throughput platform to interrogate the entire FLT3 coding sequence in a cohort of 222 adult AML patients without known FLT3 mutations and experimentally tested the functional consequences of each candidate leukemogenic allele by exogenous expression in BaF3 cells. DNA sequencing detected nine non-synonymous sequence variants in six exons that were not known single-nucleotide polymorphisms. Functional assessment of these alleles identified four novel activating mutations in the extracellular domain, the juxtamembrane domain, and the activation loop that induced constitutive kinase activity, differentially activated downstream signaling pathways, and conferred varying sensitivity to pharmacologic FLT3 inhibition. In contrast, the remaining five alleles, including mutations in highly conserved, key functional domains of FLT3, were not associated with increased kinase activity and aberrant signal transduction. These findings support the concept that acquired mutations in cancer may not contribute to malignant transformation, and underscore the importance of functional validation of candidate alleles discovered using high-throughput genomic screens, to distinguish between ‘driver’ mutations underlying cancer development, and biologically neutral ‘passenger’ alterations.

Description: Introduction Activating mutations of the KIT class III receptor tyrosine kinase (TK) are associated with the pathophysiology of acute leukemia, especially core binding factor leukemia (CBFL), and systemic mastocytosis (SM). Despite considerable antiproliferative and proapoptotic activity of several KIT TK inhibitors in vitro, clinical efficacy in AML and SM is generally moderate. We hypothesized that resistance to therapy is promoted by activation of alternative signaling pathways. Previously we reported that KIT TK inhibition results in significantly increased phosphorylation of heat shock protein (HSP) family members and that KIT is a client protein of phosphorylated (p)HSP90 putatively stabilizing KIT protein function in the presence of KIT TK-inhibitors (Kampa-Schittenhelm et al., ASH annual meeting 2010). This prompted us to further test the HSP inhibitor IPI-504 in KIT dependent CBFL, including the leukemic stem cell fraction. Methods Protein expression levels of (p)HSPs in leukemic blasts of high-risk and CBFL patients were studied by flow cytometry focusing on the CD34+/CD38- leukemia stem/progenitor cell fraction. Cellular proliferation and induction of apoptosis in leukemia cells treated with the HSP90 inhibitor IPI-504 was determined by XTT- and annexin V-based assays. Results (p)HSP90/60 levels were preferentially upregulated in CBFL associating with KIT dysregulation. Consequently, HSP90 inhibition with IPI-504 potently degraded KIT expression causing a direct antiproliferative and antiapoptotic effect in CBFL in in vitro and ex vivo models. Efficacy of IPI-504 was potentiated when combined with TK inhibitors. Importantly, high expression of (p)HSP90 and HSP60 was particularly observed in the CD34+/CD38- putative leukemia stem cell fraction arguing for a function as protection mechanism to conserve (leukemic) progenitor cell function upon cell stress such as antileukemic treatment. Conclusion HSPs are upregulated in CBFL, and IPI-504 induces antiproliferative and proapoptotic effects in primary leukemia samples. Importantly, in particular the putative malignant progenitor cell pool in KIT-associated acute leukemia expresses high levels of (p)HSP90, identifying HSP90 inhibition as an attractive novel strategy to overcome the therapy-refractory behavior of malignant stem/progenitor cells. Our results provide a rationale for the evaluation of HSP90 inhibitors such as IPI-504 in CBFL. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2472 Introduction: Using a global phosphoproteome screen, we have recently identified heat shock proteins (HSPs) to be crucial effectors to stabilize KIT oncoprotein function in the presence of tyrosine kinase inhibitors in mutant-KIT leukemia models (Kampa-Schittenhelm et al., ASH annual meeting 2010). We now show that HSP60 and (p)HSP90 are frequently expressed in acute leukemia, arguing for an integrative function of HSPs to stabilize leukemia-driving oncoproteins. Importantly, HSPs are highly expressed in leukemia stem cells (LSCs), which provides a rationale for therapy insensitivity. Targeting HSPs may therefore be an attractive strategy to target the LSC clone. Methods: Intracellular protein levels of (p)HSPs in leukemic blasts were studied by flow cytometry. Three dimensional detection of the putative CD34+/CD38-/HSP+ leukemia stem/progenitor cell fraction was performed using -FITC, -PE and APC-coupled secondary antibodies using standard techniques. Cellular proliferation and induction of apoptosis in leukemia cells treated with the HSP90 inhibitor IPI-504 was determined by XTT- and annexin V-based assays. Immunoblot experiments to assess IPI-504 interaction on the chaperone level were set up using standard protocols including a LICOR imaging system. Results: Of 16 evaluated patients with AML, 50% and 56% demonstrated significant HSP90 and HSP60 expression levels (with a threshold level set as 〉 3× of the relative mean fluorescence compared to IgG controls), respectively. Interestingly, HSP expression was predominantly seen in patients with core–binding factor leukemia (associated with KIT mutations). Moreover, five evaluable freshly harvested patient samples were used to assess the putative CD34+/CD38-/HSP90 and HSP60 population by flow cytometry. Tantalizingly, all tested CD34+/CD38- samples revealed high co-expression of HSPs. Consequently, an HSP90 inhibitor (IPI-504) potently inhibited cellular proliferation in in vitro leukemia models as well as freshly harvested leukemia cells at highly variable doses starting in low nanomolar ranges up to lower micromolar concentrations with regard to IC50s. Several cell lines with defined leukemia-driving oncogenes (MOLM14 – FLT3 ITD, HMC1.1 – KIT V560G, Kasumi1 – KIT N822H) were treated with IPI-504, and degradation of FLT3 and KIT was confirmed by immunoblotting. Consequently, oncogene-degradation led to induction of apoptosis in the nanomolar range which was confirmed in ex vivo blasts. Noteworthy, treating isolated freshly harvested CD34+/CD38- progenitor cells lead to potent reduction of the viable cell fraction. Conclusion: HSPs are frequently expressed in acute leukemias, including LSCs, arguing for a function as mechanism to protect and uphold LSC function in the presence of cell stress such as antileukemic treatment. Targeting the HSP-expressing stem cell pool may be an attractive novel therapeutic strategy in acute leukemia. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction Constitutive phosphorylation of AKT is frequently found in acute leukemia. In a proportion of patients, activation of the PI3K/AKT pathway can be linked to gain-of-function tyrosine kinase mutations. We have previously shown that TKI only incompletely inactivate AKT signaling. PI3K/AKT pathway inhibitors are attractive for targeting this pathway. However, most inhibitors lead to G1/G0 arrest. Likely for this reason, response rates to PI3K inhibitor treatment were moderate in various neoplasms. We here demonstrate that the dual PI3K-MTORC1/2 inhibitor NVP-BEZ235, which is currently under clinical investigation in relapsed acute leukemia, mediates a profound G1/G0 arrest in cells harboring leukemia-driving FLT3 ITD or BCR-ABL1 mutations, impairing induction of apoptosis. Furthermore, combination with TKI or chemotherapy overcomes cellular resistance. Methods Proliferation and apoptosis assays were performed in leukemia cell lines, an isogenic Ba/F3 cell model harboring mutant-KIT, FLT3 or ABL1 isoforms and primary leukemic cells as well as healthy donor cells. Immunoblots were performed to study effects on AKT signaling pathways including downstream effectors of autophagy (p-ULK1), proliferation, cell cycle (p70S6Kinases, cyclinD, pRB) and apoptosis (caspases). Isobolograms were created to compute combination indices. Results NVP-BEZ235 profoundly inhibited phosphorylation of AKT and p70S6K, leading to a potent antiproliferative effect in the low nanomolar range in most cell lines as well as patient blasts. However, no meaningful induction of apoptosis was observed. Instead, a sustained G1/G0 cell cycle arrest with dephosphorylation of RB and upregulation of cyclinD, induction of protective autophagy pathways (via ULK1) and absence of cleaved caspase 3 was detected. An isogenic Ba/F3 model confirmed the low proapoptotic efficacy especially for FLT3 ITD and BCR-ABL1. In an attempt to overcome NVP-BEZ235 induced cell cycle arrest, we co-treated leukemic cells with specific TK inhibitors or chemotherapeutics such as daunorubicin in a fixed dose-dilution ratio. Combination indices revealed additive to superadditive proapoptotic effects for all tested combinations. Conclusion We provide evidence that the antileukemic activity of NVP-BEZ235 in acute leukemia with mutations in tyrosine kinases is abrogated due to induction of profound G1/G0 cell cycle arrest which can be overcome by combination with TKI or chemotherapy. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction Inactivation of the p53 pathway is a frequent event in human cancers promoting tumorigenesis and resistance to chemotherapy. Inactivating p53 mutations are uncommon in non-complex karyotype leukemias suggesting that the p53-pathway must be inactivated by other mechanisms. ASPP2 is a damage-inducible p53-binding protein that enhances apoptosis at least in part through a p53-mediated pathway. We have previously shown that ASPP2 is an independent haploinsufficient tumor suppressor in vivo and that ASPP2 expression is attenuated in a proportion of acute leukemia patients. We now provide evidence that attenuation of ASPP2 is exclusively linked to patients failing to obtain complete remission after induction chemotherapy. Furthermore, we link attenuated ASPP2 expression to hypermethylation of the promoter and 5’UTR regions of ASPP2. Methods ASPP2 levels were determined using standard qRT-PCR assays as well as FACS and immunoblot analyses. Expression levels in bone marrow samples of 51 leukemia patients were normalized to a control cohort comprising bone marrow samples of 17 different healthy donors. Statistical analyses were performed using the non-parametric Wilcoxon rank-sum test and Kruskal-Wallis test. To silence ASPP2 expression, siRNA experiments were performed using leukemia cell lines and primary leukemic patient blasts. gDNA methylation arrays were performed using patient derived leukemia cells. Results ASPP2 mRNA expression was found to be significantly lower (p=0.02) in acute myeloid (AML) and lymphoid leukemia (ALL) samples (n=51) compared to mononuclear cells from healthy bone marrow donors (n=17). Notably, lower ASPP2 mRNA and protein expression levels were statistically significantly (p=0.02) associated with clinical unfavorable disease according to ELN. We further tested whether ASPP2 levels differ between patients responding or not to induction chemotherapy and found significantly lower overall ASPP2 expression levels in the non-responder cohort (p=0.01). Importantly, 27% of the patients in the non-responding cohort displayed ASPP2 levels below 0.8 (on a normalized scale with the levels in healthy donors set to 1). Such low levels were not observed in good risk patients or in higher-risk patients achieving complete remission after the first cycle of induction chemotherapy. Receiver operating characteristic (ROC) analysis determined

Description: ASPP2 is an independent haploinsufficient tumor suppressor which initiates induction of apoptosis after cellular damage in a TP53-dependent manner. We recently reported the identification of two C-terminally truncated splice variants of ASPP2 with high frequency in acute leukemias. Tantalizingly, these variants lack the functionally relevant TP53 binding sites, suggesting a dominant-negative, oncogenic function. The most prevalent isoform lacking exon 17 (ASPP2k) is detected in up to 60% of acute leukemias. A cell sort for the CD34+ stem cell fraction confirmed ASPP2k-expression in the leukemia-initiating clone. Immunoblotting revealed translation of ASPP2k into a genuine protein isoform. Generation of isoform-specific antibodies allows to discriminate ASPP2k from the wildtype isoforms. Interestingly, ASPP2k is the sole isoform expressed in some native leukemia samples. Other samples reveal a heterozygous expression pattern suggesting a dominant-negative function of the C-terminally truncated ASPP2k isoform. Forced expression of ASPP2k using a HisMax plasmid vector in murine IL3-dependent pro-B Ba/F3 cells induced a more aggressive phenotype with mitotic failure and perturbed cellular proliferation compared to the empty vector controls. Of note, IL3 was successfully weaned in the ASPP2k cells - indicating that autoactivating genomic alterations must have occurred upholding cellular integrity and viability. Fractionated gamma irradiation accelerated this process. In contrast, the empty vector strains did not survive IL3 weaning. Chromosomal analyses of IL3-independent Ba/F3 ASPP2kcells revealed structural alterations including monosomies and additional marker chromosomes. Furthermore, we demonstrate that ASPP2k is stress-inducible. Changing culture conditions (e.g. temperature) of cell lines or ex vivo native AML samples resulted in a dramatic increase of the ASPP2k isoform. Consecutive treatment of cells with daunorubicin lead to relative resistance of pre-stressed cells compared to their counterparts cultured under standard conditions. Vice versa, specific knock-down of ASPP2k resulted in a significant increase in induction of apoptosis upon chemotherapy compared to empty vector controls. Our data demonstrate that ASPP2k plays a distinctive role as an anti-apoptotic regulator of the TP53 checkpoint rendering cells to a more aggressive phenotype as evidenced by proliferation and apoptosis rates - and facilitates acquisition of genomic mutations, a first initiating step in leukemogenesis. Disclosures No relevant conflicts of interest to declare.

Description: Inactivation of tp53 by chromosome/gene loss or loss-of function mutation is frequently observed in therapy-associated acute myeloid leukemia. Tp53 mutations associate with complex and monosomal karyotypes and provide the strongest predictive risk factor in this subgroup. ASPP2 is an independent tumor suppressor. We have previously shown that ASPP2 expression is attenuated in acute leukemia via methylation by unknown mechanisms. We now provide evidence that attenuation of ASPP2 expression may act as an early leukemia-initiating event - facilitating acquisition of structural chromosomal aberrations and loss of its binding partner p53. We found low ASPP2 mRNA and protein expression levels in complex karyotype AML, arguing for a role of ASPP2 loss in leukemogenesis. To evaluate for malignant transformation we engineered IL3-dependant murine Ba/F3 cells and stably silenced ASPP2 expression using retroviral-induced shRNA, Interestingly, only ASPP2(ko) cell strains could be successfully weaned from IL3 dependence - suggesting that loss of ASPP2 expression promoted cellular transformation. In contrast, empty vector control cells did not survive the weaning process. As expected, cellular proliferation was enhanced in the ASPP2(ko) cell lines as measured by standard XTT and CSFE protocols. To study how loss of ASPP2 could promote transformation, we performed cytogenetic analysis of the chromosome set of parental (+IL3) vs. gamma-irradiated empty vector controls (+IL3) and ASPP2(ko) cell strains (-IL3) with and without irradiation (+/- stress enforcement using gamma irradiation, 2 x 5Gy). Gamma irradiation did not significantly alter cytogenic analysis between the empty vector controls compared to the parental cells. Tantalizingly however, we found a consistent chromosomal loss of chr8, der8 and der19 - and, most intriguingly, loss of chromosome 11 (which locates tp53 in mice), and a monosomal karyotype in the ASPP2-attenuated cells. In addition, several marker chromosomes (6-8 in ASPP2i vs. 0-2 in empty vector controls were identified. We are now in the process of verifying our findings in knockout native human bone marrow blasts. Conclusions: Our findings provide novel evidence that loss of ASPP2 contributes to malignant transformation by promoting genomic instability as measured by increased cytogenetic abnormalities after gamma-irradiation. These findings may have important relevance for the pathogenesis of therapy-associated AML - but also of hypodiploid acute lymphoblastic leukemia, which is as well associated with tp53 loss-of-function and predicted for a dismal outcome. Studies are ongoing to elucidate the mechanisms of how loss of ASPP2 expression promotes human AML, as well as restore ASPP2 function as a novel therapeutic pathway to exploit in treating AML. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 1416 Introduction: Inactivation of the p53 pathway is a universal event in human cancers and promotes tumorigenesis and resistance to chemotherapy by impaired induction of apoptosis. Unlike in solid tumors, inactivating p53 mutations are uncommonly found in de novo acute leukemias, and the p53-pathway thus must be inactivated by other mechanisms instead. ASPP2 is a damage-inducible p53-binding protein that enhances apoptosis at least in part through a p53-mediated pathway. Moreover, ASPP2 is a haploinsufficient tumor suppressor (Kampa et al. PNAS 2009) that is suppressed in many human cancers, correlating with poor clinical outcomes. Using a global gene-chip approach evaluating attenuated ASPP2 expression in mouse embryonic fibroblasts, we found evidence that loss-of ASPP2 alters pathways involved in hematologic tumorigenesis. We now show that ASPP2 is attenuated in high-risk leukemias, which may have implications for further therapy decisions. Methods: Quantitative RT-PCR was performed to analyze ASPP2 mRNA levels in leukemia patients and healthy volunteers. Protein expression levels were evaluated in treatment naïve as well as anthracycline-treated patient samples using flow cytometry and immunoblotting. Small-interference (si)RNA experiments were employed to reveal involvement of ASPP2 in therapy response and tumorigenesis in functional viability assays using primary patient samples. Results: Analysis of ASPP2 mRNA levels of 50 patients diagnosed with acute leukemia revealed wide expression variance ranging from 0.03 to 97–fold (normalized) compared to GAPDH as house keeping gene control. Notably, patients considered as prognostic good risk (n=24) showed a trend to higher levels (mean 8.5, range 0.8–97.1) as compared to the high risk cohort (n=26, mean 6.1, range 0.03–31.5). Even more, within the high risk cohort 31% revealed expression levels below 1 (27% below 0.8) – compared to 13% in the good risk population (0% below 0.8). Subanalysis of the high risk cohort revealed that none of the patients achieving early complete remission after initial induction chemotherapy belonged to those patients presenting with attenuated ASPP2 expression levels. In accordance, intracellular ASPP2 protein expression levels were higher in the good risk patient cohort. Additionally, we found that low ASPP2 protein expression levels in ex vivo cultured patient-derived blasts with high-risk features did not increase after chemotherapy-induced damage, while upregulation of ASPP2 levels – indicating apoptosis induction – was observed in blasts from good-risk patients. Failure to upregulate ASPP2 upon chemotherapy may thus contribute to therapy resistance and indeed, ASPP2 attenuation in in vitro and ex vivo leukemia models using siRNA lipofection resulted in increased cellular proliferation and resistance towards anthracyclines and tyrosin kinase inhibition. Conclusions: Our results suggest that attenuated ASPP2 expression levels contribute to leukemogenesis of high-risk leukemia and therapy failure. Thus, ASPP2 expression should be further explored as a potential prognostic and/or predictive biomarker in AML. Disclosures: No relevant conflicts of interest to declare.