ALBERT

Description: Abstract 1885 Current therapies used in the treatment of AML are highly toxic and have been shown to be ineffective at targeting AML stem cells (AML-SCs), which are chemoresistant and thus likely to provide a surviving reservoir of cells that drives AML relapse. The clinical relevance of AML-SCs in relapse is supported by studies indicating that higher AML-SC proportions at diagnosis and during minimal residual disease correlate with worse outcome. Previous studies have demonstrated that survival of AML-SCs is heavily dependent on both NFkappaB activation and redox homeostasis, which are simultaneously disrupted by the known anti-AML-SC compound, parthenolide (PTL). However, PTL represents a suboptimal drug based on its poor pharmacokinetics. Derivatives of PTL have been created and are currently in clinical trials. However, the process of taking pre-clinical findings to clinic is lengthy and success rates in this process are low. As therapies amenable to rapid clinical translation are urgently needed in AML, we sought to determine whether chemical genomics could be used to predict combinations of FDA-approved drugs that target AML-SCs. Previous studies have used chemical genomics to find PTL-like drugs by interrogating the Connectivity Map and Gene Expression Omnibus databases for drugs that perturb the cancer transcriptome so as to best mimic the transcriptional signature of PTL. However, no single agent found thus far is FDA approved. We hypothesized that we could extend the use this information to identify pairs of FDA approved drugs that together optimally mimic the anti-AML-SC transcriptional signature. To this end, we partitioned transcriptome changes caused by individual chemical perturbations of the Connectivity Map database into gene networks and sought to identify pairs of drugs that maximally perturbed the same networks as drugs targeting AML-SCs. This strategy revealed a combination consisting of the off-market anti-diabetes drug, troglitazone, and the anti-nausea drug, prochlorperazine. Individually, neither troglitazone nor prochlorperazine was able to target primary AML-SCs ex vivo, consistent with the inability of each drug to achieve a favorable PTL-like signature at transcriptomic level. Indeed, ex vivo exposure of primary AML-SCs to either troglitazone or prochloroperazine produced 99% and 87% mean viability, respectively, 24 hours post-treatment (N=3). In contrast, the combination of the two drugs reduced mean viability to 16% (p 〈 0.05; N=3). We next ascertained the effects of these drugs alone or combination against AML progenitors using methylcellulose colony forming unit (CFU) assays. Individually, troglitazone and prochlorperazine each exerted a modest effect on colony formation (87% and 77% of untreated controls, respectively; N=3). In sharp contrast, the combination of both drugs proved toxic against AML progenitor/stem cells, reducing colony formation to 25% of untreated controls when measured 3 weeks post-treatment (p 〈 0.05; N=3). Toxicity to normal cord blood stem cells was modest when drugs were administered alone or in combination, preserving the selectivity of PTL. Finally, we corroborated these findings in functional studies using NOD/SCID xenotransplantation models. We found that 2 of 3 primary patient specimens showed significant eradication of AML-SC function (p 〈 0.05, 5 mice per patient sample). Together, these findings suggest that transcriptomic effects of pre-clinical drugs can be used in chemical genomic screens to define FDA approved drug combinations that can more rapidly be translated to clinic. Disclosures: No relevant conflicts of interest to declare.

Description: In this study, we characterized nuclear factor κB (NF-κB) subunit DNA binding in chronic lymphocytic leukemia (CLL) samples and demonstrated heterogeneity in basal and inducible NF-κB. However, all cases showed higher basal NF-κB than normal B cells. Subunit analysis revealed DNA binding of p50, Rel A, and c-Rel in primary CLL cells, and Rel A DNA binding was associated with in vitro survival (P = .01) with high white cell count (P = .01) and shorter lymphocyte doubling time (P = .01). NF-κB induction after in vitro stimulation with anti-IgM was associated with increased in vitro survival (P 〈 .001) and expression of the signaling molecule ZAP-70 (P = .003). Prompted by these data, we evaluated the novel parthenolide analog, LC-1, in 54 CLL patient samples. LC-1 induced apoptosis in all the samples tested with a mean LD50 of 2.8 μM after 24 hours; normal B and T cells were significantly more resistant to its apoptotic effects (P 〈 .001). Apoptosis was preceded by a marked loss of NF-κB DNA binding and sensitivity to LC-1 correlated with basal Rel A DNA binding (P = .03, r2 = 0.15). Furthermore, Rel A DNA binding was inversely correlated with sensitivity to fludarabine (P = .001, r2 = 0.3), implicating Rel A in fludarabine resistance. Taken together, these data indicate that Rel A represents an excellent therapeutic target for this incurable disease.

Description: The identification of genes and pathways that are critical for the development and maintenance of leukemia is important for designing new anti-leukemia therapeutics. However, selection of aberrantly regulated genes and pathways and the translation of these features into targeted anti-cancer drugs represent a significant challenge. Studies have characterized the various oncogenic lesions associated with leukemia, and have shown that specific combinations of mutations are sufficient to induce malignancy in model systems. Thus, we sought to identify the gene signature arising from the introduction of cooperating oncogenic mutations into non-malignant cells. We hypothesized that compounds capable of maximally antagonizing the aberrant expression of this subset of genes would represent potential anti-leukemia therapeutics. To test this hypothesis, we turned to the use of murine genetics and chemical genomics. To this end, we adapted mouse systems created by retroviral transduction of primary marrow cells with the BCR-ABL translocation alone or in combination with the NUP98-HOXA9 translocation. This approach creates authentic models of chronic and blast crisis myeloid leukemia (CML and bcCML, respectively), which has been shown to closely resemble human disease. Next, we employed flow cytometry to separate the transduced marrow into purified primitive lin- populations of either normal, BCR-ABL+, or BCR-ABL+/NUP98-HOXA9+ cells. Importantly, lin- cells were used to focus specifically on those genes most relevant to leukemic stem and progenitor populations. Microarray gene expression profiling was performed on these purified populations in six independent experiments using the Affymetrix Mouse 430 2.0 platform. Gene expression signatures for each transduced population relative to normal lin- cells were obtained. Next, in silico screening for agents that antagonize these signatures was performed comparing the resulting signatures to a local microarray compendium that included the CMap dataset. Gene expression signatures arising from the introduction of BCR-ABL alone identified traditional chemotherapeutics including cytarabine, etoposide, idarubicin – agents considered less effective against leukemic stem cell populations – as well as agents acting along the PI3 kinase pathway. Surprisingly, drugs shown to antagonize the gene signature derived by the introduction the two oncogenic mutations (BCR-ABL and NUP98-HOXA9) included 15-delta-prostaglandin-J2, celastrol, MG- 132, and parthenolide. Each of these drugs eradicate malignant stem and progenitor cells, while sparing normal cells. Thus, the murine-derived signature successfully identified drugs highly relevant to human disease. Our studies therefore demonstrate the utility of combining gene expression-based drug screening and mouse cancer genetics to reveal novel anti-leukemia compounds in humans. DCH and JMA contributed equally.

Description: Introduction: PTPN11 encodes the protein tyrosine phosphatase SHP 2, which relays signals from growth factor receptors to RAS and other effectors. Germline and somatic mutations in PTPN11 are well described in the pediatric population and associated with Noonan Syndrome and Juvenile Myelomonocytic Leukemia (JMML). Pathogenesis of JMML specifically appears to be through activation of the RAS-RAF-MAP kinase pathway leading to dysregulated myeloid differentiation. There are also data to suggest that somatic PTPN11 mutations portend a poor prognosis in MDS patients (pts) receiving hypomethylating agents. The significance of PTPN11 when sporadically mutated in adults with AML remains controversial as several analyses have thus far failed to show any clinical relevance. This study evaluated the clinical significance of somatic PTPN11 mutations in a single center cohort. Methods: From 7/2015-7/2018, data from an AML database at New York Presbyterian/Weill-Cornell Medical Center was queried for the presence or absence of mutations in the PTPN11 gene as well as on all pts with TP53 mutations to use as a surrogate, given its well-known status as a poor prognostic factor. Log-rank tests were used to compare survival data, while Fisher-exact test was used to compare non-survival data (i.e. CR rates). For multivariate analysis, linear regression was performed and looked at mutational status, age, cytogenetics (CG), and controlled for age and European Leukemia Net (ELN) risk. Results: 256 AML pts with complete evaluable data. 30 were found to harbor mutations in PTPN11 at diagnosis. Within the PTPN11 mutated cohort, median age was 70, 15 were female and 15 were male. 1st cycle complete response (CR) rate was 30% (9/30) and one additional pt (4.8%) achieved a salvage CR. Hematopoietic stem cell transplantation (HSCT) was provided to 3/30 (10%) and of those, 1/3 (33.3%) relapsed, within 8 months. In the pts who achieved a CR, 4/10 (40%) relapsed. Median overall survival (OS) of the PTPN11 mutated cohort was 9 months (mo). Four patients (13.3%) are alive and in a CR 〉6 mo at time of censor. DNMT3A, NPM1, K/NRAS, RUNX1, TP53 and IDH1/2 were commonly co-mutated (n=12,9,7,7,6, and 6 respectively, table 2) with PTPN11 mutations. DNMT3A, NPM1 and PTPN11 were commonly mutated together in pts, n=8 (26.7%). The PTPN11 mutation was a single mutation in 2 pts. Common CG findings include normal (n=9), complex (n=4), trisomy 8 (n=4), chr. 3 abnormalities (n=7), chr. 5 (n=7) and chr. 7 (n=8). When comparing the PTPN11 mutated pts to all AML pts diagnosed at this center during the same time period without a PTPN11 mutation (table 1), 1st cycle CR rate (30% vs 57.5%, p=0.006), any CR (33.3% vs 71.4%, p=0.001), HSCT (10% vs 41.6%, p

Description: Abstract 388 Leukemia stem cells (LSCs) have been shown to initiate and maintain AML. Given that LSCs have also been shown to be chemoresistant relative to the bulk leukemia population, LSCs are thought to provide a surviving reservoir of cells that drive disease relapse. This concept is further supported by studies suggesting that poor prognosis is associated with leukemias presenting with a higher percentage of LSCs. Thus, identification of new therapeutic regimens that target LSCs appears to be of clinical significance. Our previous efforts to target LSCs have demonstrated the capability of parthenolide (PTL), and its water-soluble clinical derivative, dimethylamino-parthenolide (DMAPT), to impair the survival and leukemogenic activity of phenotypically- and functionally-defined LSCs, respectively. In order to improve the clinical utility of PTL–based compounds, we sought to determine rational drug combinations that would enhance the efficacy of these agents in vivo. To this end, we turned to the use of chemical genomic screening. The transcriptional signature arising when primary patient specimens are exposed to PTL for 6 h reveals a significant “protective response” (i.e. augmentation of detoxifying enzymes, antioxidant responses, and the unfolded protein response). Given this observation, we hypothesized that compounds capable of impairing the protective response would synergize with PTL and its derivatives and thereby enhance their anti-leukemia activity. To test this hypothesis, we interrogated the Connectivity Map database for instances of compounds that produced maximal inhibition of the protective response at the gene expression level. Overwhelmingly, this screen indicated compounds acting along the PI3 kinase pathway including wortmannin, LY-294002, and rapamycin. Indeed, exposure of primary AML cells to the combination of wortmannin and sub-lethal doses of PTL significantly decreased viability of AML cells (24% viable) relative to PTL (65% viable) or wortmannin (66% viable) alone. Moreover, the combination resulted in a significant decrease in colony formation (Wort=67% , PTL=87.5 and PTL+Wort = 6% CFU relative to untreated). The effect of these combinations is synergistic and not additive (Chou-Talay method). Importantly, these effects remained confined to AML cells and not their normal hematopoietic counterparts. Furthermore, these observations were corroborated with rapamycin and temsirolimus. To examine molecular mechanisms underlying the enhanced anti-leukemia activity, immunoblot analyses were performed and demonstrated that the PTL-induced protective response was abolished by the combination of PTL with wortmannin or rapamycin, consistent with the chemical genomic prediction. Finally, to assess in vivo activity of the drug combination we employed the clinical agents DMAPT and temsirolimus in a mouse xenotransplant model. Briefly, NOD/SCID mice were injected with primary human AML cells and at three weeks post transplant began a three-week course of daily treatment (vehicle control, DMAPT alone, temsirolimus alone, or DMAPT+temsirolimus). Animals were sacrificed after treatment (i.e. six weeks post transplant) and tumor burden was evaluated in the bone marrow. The mean percent of leukemic cells for the combination of temsirolimus and DMAPT treatment resulted in a four-fold decrease which was significant when compared to either treatment alone (P

Description: Abstract 3966 Poster Board III-902 Since oncogenic activation of HoxA9 is induced by multiple chromosomal translocations affecting MLL1 (11q23)(e.g. MLL-Af9) or Nup98 (11p15)(e.g. Nup98-HoxA9 or Nup98-Nsd1), the function of the HoxA9 transcription factor is of critical interest in human acute myeloid leukemia (AML). HoxA9 forms heterodimeric DNA binding complexes with members of the Pbx and/or Meis family of homeodomain proteins. Importantly, the direct transcriptional targets of endogenous HoxA9 that mediate transformation remain largely unknown. The Growth factor independent-1 (Gfi1) transcriptional repressor is known to induce granulopoiesis and inhibit myeloid progenitor proliferation. GFI1 is mutated in patients with severe congenital neutropenia (SCN). SCN patients are at increased risk for AML. In a transcriptional circuit conserved to Drosophila, we have recently shown that Gfi1 represses HoxA9, Meis1 and Pbx1 expression, that Gfi1 and HoxA9 demonstrate dramatic epistatic relationships, and that Gfi1 loss of function is potently preleukemic. Our new bioinformatic, biochemical and expression data reveal microRNA genes to be targets of endogenous HoxA9 versus Gfi1 antagonism. Moreover, these miR are activated by Hox-signaling leukemia oncoproteins. Next, in both murine leukemia models and primary human AML samples, antagomir-mediated inhibition of microRNA function specifically disrupts oncogenic signaling by HoxA9, Nup98-HoxA9 and MLL-Af9 (but not AML-ETO which does not signal through HoxA9). In vivo, antagomir treatment blocked MLL-Af9-initiated leukemia lethality. These data establish microRNA genes as functional downstream targets of endogenous HoxA9, and implicate epigenetic signaling as critical client/mediators of Hox-based leukemia oncoproteins. Disclosures: No relevant conflicts of interest to declare.

Description: Previous studies indicate that human acute myelogenous leukemia (AML) arises from a rare population of leukemic stem cells. Cells of this nature can initiate and maintain leukemic cell growth in both long-term cultures and nonobese diabetic/severe combined immune-deficient mice. To characterize the biology of primitive AML cells, gene expression screens were performed with 7 primary AML and 3 normal specimens. For each sample, stem cell populations (CD34+/CD38−) were isolated and used to synthesize radiolabeled complementary DNA (cDNA). AML vs normal probes were then hybridized to cDNA arrays containing genes related to cancer and apoptosis. Of approximately 1400 genes analyzed, 2 tumor-suppressor genes were identified that were overexpressed in all 7 of the AML CD34+/CD38−cell populations: death-associated protein kinase and interferon regulatory factor 1. Expression of each gene was confirmed by reverse-transcription polymerase chain reaction and immunoblot analysis. It is proposed that tumor-suppressor proteins play a role in the biology of primitive AML cells.

Description: In the bone marrow, osteoblastic cells have recently been shown to be an important component of the hematopoietic stem cell (HSC) niche. We demonstrated that activation of the PTH Receptor (PTHR1) in osteoblastic cells results in expansion of HSC in vitro and in vivo. In addition, PTH treatment of mice undergoing myeloablative bone marrow transplantation using limiting numbers of donor cells dramatically improved their survival. Activation of the Notch signaling pathway was necessary to mediate the PTH-induced expansion of HSC (Calvi L.M., Adams G.B. et al., Nature 425, 841–846). This pathway, through cell-cell interactions plays a fundamental role in HSC self-renewal. Since normal primary murine stromal cells treated with PTH (1–34) had improved ability to support HSC, we analyzed the expression of the Notch ligand Jagged1 in these cells. Stromal cell Jagged1 has been shown to promote HSC self-renewal. Realtime PCR analysis of total RNA from primary stromal cells showed no changes in Jagged1 message in PTH compared to vehicle-treated cells. To assess whether this was due to rare and heterogeneous expression of Jagged1 in the stromal cell population, we performed immunocytochemical analysis of primary murine stromal cells treated with PTH(1–34). A small subpopulation of PTH-treated primary murine cells had a dramatic increase of Jagged 1 protein, while this subpopulation was not identified in vehicle-treated stromal cells. Since the PTHR1 is expressed in stromal osteoblastic cells, rat osteosarcoma UMR106 cells were chosen in order to study PTH-dependent Jagged1 expression in osteoblastic cells. A 10 fold time and dose-dependent increase in Jagged1 expression was measured by realtime PCR in RNA from UMR106 cells treated with PTH compared with vehicle. Western blot analysis using two distinct anti-Jagged1 antibodies confirmed the PTH-dependent increase in Jagged1 protein in UMR106 cells. In osteoblasts, PTH(1–34) is known to activate both the adenylate cyclase (AC) and the protein kinase C (PKC) signaling cascades downstream of the PTH1R. We independently determined the effect of activating either pathway on Jagged1. Forskolin, an AC activator, dramatically increased Jagged1 expression in a time and dose dependent fashion. Jagged1 protein was also increased by Forskolin. In contrast, Jagged1 protein levels were not increased by treatment with TPA, a PKC activator. In summary, osteoblastic cell treatment with PTH increases expression of the Notch ligand Jagged1 through activation of the AC signaling pathway downstream of the PTH1R. UMR106 provide a useful model to study the molecular mechanisms underlying osteoblastic PTH-dependent increases in Jagged1 levels, which are likely to play an important role in the PTH-induced enhanced osteoblastic support of HSC. Further definition of this pathway is likely to provide targets for pharmacologic manipulation of the HSC microenvironment.

Description: CLL-1 (C-type Lectin-Like Molecule-1) is an inhibitory receptor expressed on myeloid cells which was previously shown to be expressed on AML cancer stem cells. To further validate the potential therapeutic against CLL-1, we generated a series of monoclonal antibodies (mAbs) against CLL-1 and assessed their cytotoxic and anti-cancer activities. While expression of CLL-1 was restricted to myeloid cells, it was expressed in 80% (37/46) of AML blasts but not in ALL blasts (n=5). Expression on AML CD34+/CD38- stem cells was detected in 71% (12/17) of cases. Selected anti-CLL-1 mAbs mediated dose-dependent. Complement Dependent Cytotoxicity (CDC) against various AML-derived cell lines with no detectable cytotoxic activity against lymphoid derived cell lines. Moreover, human embryonic kidney 293 cells became susceptible to anti-CLL-1 mAb mediated killing only after transfection with CLL-1, demonstrating that cytotoxic activity is mediated through a specific interaction with CLL-1. Furthermore, anti-CLL-1 mAbs showed CDC activity against all AML blasts tested in ex vivo assays (n=13), while no activity was observed against ALL blasts. CLL-1 efficiently internalizes upon antibody binding in both 293 transfected cells as well as AML cell lines, demonstrating the potential therapeutic opportunity for toxin-conjugation of anti-CLL-1 mAbs. In vivo anti-cancer activity of the lead chimeric mAb against CLL-1 was tested in an HL-60 xenograft model. In this model, growth of established tumors was reduced by 38% at 5 mg/kg twice weekly dosing. Our results demonstrate CLL-1 as an attractive target for AML with restricted expression on cells from myeloid origin, AML blasts and leukemic stem cells, as well as specific cytotoxic activity in in vitro, ex vivo and in vivo assays. We are currently undertaking additional xenograft models to evaluate the therapeutic potential of these mAbs against primary AML engrafted cells and in combination with chemotherapy in vivo.

Description: Key Points GLI3R inhibits Hh signaling and is required for response to SMO antagonist in AML. GLI3 is silenced in AML, and decitabine restores GLI3 expression and leads to modulation of Hh signaling.