Bogenberger, James M.

Abstract The lysine deacetylase inhibitor suberoylanilide acid (SAHA) exhibits promising but limited activity in the treatment of acute myeloid leukemia (AML). To identify potential targets for rational combination therapies that increase the efficacy of SAHA in AML, we performed a functional RNA-interference (RNAi) drug modifier screen to identify genes that, when inhibited, potentiate or antagonize the in vitro anti-leukemic activity of SAHA. A total of 901 kinase, phosphatase and closely associated signaling genes were silenced in TF-1, HEL and THP-1 cells, with four different siRNA sequences per gene, both alone and in combination with SAHA treatment. Screen hit lists for each cell line were over-laid on an integrated functional relationship network. A community detection algorithm was then applied to this sub-network and siRNA sensitive modules were identified. Each module represents a highly connected set of genes in the integrated network. To identify pathways represented by each module, enrichment was evaluated using the National Cancer Institute (NCI) Protein Interaction Database (PID) pathways. Both cell line-specific and universal sensitizing targets, grouped into a small number of pathways, emerged from these screens. Seven pharmacologic inhibitors interfering with these pathways have been assessed in a panel of four or five AML cell lines, including SET-2, TF-1, HEL, THP-1, and OCI-AML3. Of the seven tested inhibitors targeting SAHA sensitizing networks, one combination exhibiting universal SAHA sensitization in all AML cell lines tested was further assessed in drug dose response assays using ex vivo cultures of bone marrow mononuclear cell populations derived from patients with various myeloid malignancies. The combination demonstrated “one-sided” SAHA EC50 fold-enhancement, as well as strong synergy by “two-sided” CalcuSyn combination index determination in all ex vivo specimens examined to date (N=6), including polycythemia vera (PV), chronic myelomonocytic leukemia (CMML), myelodysplastic syndrome/myeloproliferative neoplasm (MDS/MPN) overlap syndrome, MDS transformed to AML, and de novo AML samples. Thus, this first reported large-scale SAHA RNAi modifier screen in AML has identified a specific pathway/target whose inhibition broadly synergizes with HDAC inhibition in myeloid malignancies. A novel compound putatively targeting this respective pathway/target is currently in early clinical development and has shown limited single-agent activity. With the full dataset presented, we will propose a novel rational combination to enhance the activity of HDAC inhibition in AML and other myeloid malignancies that could be translated into design of a clinical trial. Disclosures: No relevant conflicts of interest to declare.

Abstract Despite significant efforts, the clinical mechanism of action of hypomethylating agents such as 5-azacytidine (5-aza) is still poorly understood. 5-aza is currently indicated for the treatment of patients with myelodysplastic syndrome (MDS). While 5-aza has achieved good single-agent activity in acute myeloid leukemia (AML), complete response rates remain low when used as a single agent. In a recent report aimed at identifying rational therapeutic combinations with 5-aza, Bogenberger and colleagues identified multiple BCL-2 family member/BH3-containing therapeutic targets, which synergize with 5-aza when inhibited genetically or pharmacologically. The CDK9 inhibitor, alvocidib, has achieved significant improvement in complete response rates of newly diagnosed AML patients when administered before cytarabine and mitoxantrone (FLAM regimen) in a randomized multi-center Phase 2 trial when compared to 7+3 standard of care treatment. Recent reports suggest that the transcriptional repression of key anti-apoptotic proteins (eg., MCL-1) mediated by alvocidib's CDK9 inhibition, drive the pro-apoptotic activity of alvocidib in the FLAM regimen. We, therefore, hypothesized that alvocidib and 5-aza would synergize therapeutically in the treatment of AML by means of transcriptional repression of MCL-1 and sensitization to 5-aza. In this report, we demonstrate that treatment of AML cell lines with alvocidib inhibits both mRNA and protein expression of MCL-1 in a time and concentration-dependent fashion. Pre-treatment of cells with alvocidib, to repress MCL-1 expression prior to 5-aza treatment, reduced the 5-aza cell viability EC50 more than 2.5-fold, from 1.8 µM to 0.6 µM in MV4-11 cells. The alvocidib/5-aza combination also resulted in synergistic increases in caspase activity relative to either single agent within the combination, at multiple dose levels. Therefore, following reports suggesting inhibition of BCL-2 family members including MCL-1, sensitizes cells to 5-aza, our data suggest that the alvocidib/5-aza combination may constitute a viable therapeutic regimen. We also conclude that a CDK9 inhibitor/5-aza combination may be an effective clinical approach for the treatment of AML. Disclosures Kim: Tolero Pharmaceuticals: Employment. Soh:Tolero Pharmaceuticals: Employment. Bearss:Tolero Pharmaceuticals: Employment. Lee:Tolero Pharmaceuticals: Employment. Peterson:Tolero Pharmaceuticals: Employment. Whatcott:Tolero Pharmaceuticals: Employment. Siddiqui-Jain:Tolero Pharmaceuticals: Employment. Weitman:Tolero Pharmaceuticals: Employment. Bearss:Tolero Pharmaceuticals: Employment. Warner:Tolero Pharmaceuticals: Employment.

Abstract Background: Hypomethylating agents (HMA) such as 5-Azacitidine (Aza) have overall responses rate of ~35-50% in acute myeloid Leukemia (AML) and higher risk myelodysplastic syndrome (MDS) patients (pts). However, eventually all pts develop resistance on therapy. To improve upfront response rates and to address HMA failure, we performed genome-scale RNAi screens targeting genes on chromosomes 5 and 7, recurrent cytogenetic abnormalities in AML/MDS. These screens identified Hedgehog pathway (Hh) genes as potentially targetable vulnerabilities with Aza. Combination experiments with smoothened (SMO) inhibitors (SMO is a central activating gene within the Hh pathway) showed pre-clinical synergy. These and additional published data support the Hh pathway as a targetable vulnerability in AML, and provide a rational for a trial. Methods: A phase I/Ib protocol of the oral SMO inhibitor LDE225 (Sonidegib) in combination with Aza in untreated and relapsed/refractory (rel/ref) AML, chronic myelomonocytic leukemia (CMML), MDS, or MDS/myeloproliferative neoplasm overlap pts. Objectives of the phase I part are to determine safety, toxicity and MTD and for the phase Ib part to assess the preliminary overall efficacy. Sonidegib was started at dose level (DL) 0 at 400mg oral daily continuously and Aza at the approved doses (75mg/m2 x 7 days or days 5-2-2) with dose adjustment per label. A classical 3+3 design was used and DLT (CTCAE v4.0) defined as > grade 3 CPK elevation or hepatic or non-hematologic adverse events (AEs) lasting > 7 days or any grade 4 event, except for hematologic toxicities. DLT was assessed after 42 days when steady-state concentrations of Sonidegib were reached. Pharmacokinetic assessments were not performed for this study. Results: Patients and Dose Levels : Between 5/22/2014 and the data cut-off date (7/20/2015), 36 pts were enrolled. Median age was 72 y (range 52-85 y) and 64% were male. Most AML pts (23 of 24) had intermediate or unfavorable cytogenetics and most MDS/CMML pts were in the int.- to high-/very high-risk group (IPSS-R). Six pts were treated at DL-0 of Sonidegib 400mg/day and one DLT of CPK elevation was observed. Another patient had grade 4 neutropenia (which at the time was included in the DLT definition, later removed). Due to overall tolerance of the combination at DL 0 and investigator assessment, despite only 1/6 pts having a formal DLT, DL -1 at Sonidegib 200mg was explored at which 0/3 pts (1 pt non-evaluable) had DLT and the phase Ib portion was opened at DL -1. To date a total of 24 AML (13 newly dx, 11 rel/ref), 11 MDS/CMML (5 newly dx, 6 rel/ref ) and 1 myelofibrosis pt have been enrolled. Safety and Tolerability: An initial safety and efficacy assessment for 25 of the 29 pts treated on the phase Ib (n=26) + the MTD cohort of 200 mg (n=3) and for who at least 2 cycles of treatment follow up was available, showed that all pts experienced at least one grade-3 or 4 hematologic AE with a lower rate in the untreated vs. treated cohort. Other, non-hematologic grade-3 and 4 AEs respectively were seen in 24 and 28% of patients-most commonly hyponatremia and fatigue, known AEs for this drug class. Treatment delays occurred in 28% of pts and Aza or Sonidegib dose reductions in 16%. 23 (79%) pts are alive and 6 have died (21%), with one death while pt was actively treated but was deemed unrelated; 11 pts remain on study, with the remainder showing disease progression (n=3) or taken-off study for other reasons (n=3 each for AEs and did not want further treatment; n=4 other). Efficacy : As of 7/20/15 with a median treatment duration of 2 cycles, 2/5 untreated MDS/CMML pts achieved CR/CRi; 1 HMA rel/ref AML had CR; 1 untreated AML pt each had PR and MLFS. 90% of pts have not progressed. Updated results will be presented at the ASH 2015 meeting; by then all 29 pts will have completed a minimum of 4-6 cycles on study and full efficacy and AE result on all pts will be reported. To date the OS at 6 mo is 77% (95% CI: 44-97%) with a median PFS of 4 mo (range: 2-NR). Many pts in the untreated cohorts have just enrolled. Conclusion: Hh pathway inhibition with the clinically well tolerated SMO inhibitor Sonidegib in combination with Aza in MDS and AML pts is well tolerated with mostly hematologic AEs as expected with early signs of clinical activity in untreated and HMA-failure pts. Full efficacy results of 29 pts and updated safety data will be reported for this first combination trial of a SMO inhibitor in combination with a HMA in MDS and AML pts. Disclosures Mesa: Gilead: Research Funding; Novartis Pharmaceuticals Corporation: Consultancy; Incyte Corporation: Research Funding; Genentech: Research Funding; Pfizer: Research Funding; Promedior: Research Funding; NS Pharma: Research Funding; CTI Biopharma: Research Funding. Al-Kali:Novartis: Research Funding.

Abstract Abstract 1432 Previously we demonstrated that inhibition of anti-apoptotic BCL-2 family members sensitize leukemic cells to 5-Azacytidine (5-Aza), using siRNA and pharmacological inhibition with the BH3-mimetic ABT-737, both in vitro and ex vivo (Bogenberger, JM., et. al. ASH Annual Abstracts 2011;118:Abstract 3513). Crucially, several anti-apoptotic BCL-2 members (e.g. BCL-2 and BCL-XL) required concurrent inhibition for potent and universal sensitization to 5-Aza. Anti-apoptotic BCL-2 proteins block the execution of programmed cell death (apoptosis) by binding to and counteracting two types of pro-apoptotic BCL-2 family proteins: the “BH3-only” proteins, including both activators (BIM and BID) and sensitizers (e.g. NOXA, PUMA, HRK), and the multi-domain effector proteins (BAX and BAK). Cells dependent on anti-apoptotic BCL-2 family members for survival have been defined as “primed for death” (Certo, M., et. al. Cancer Cell 2006 May;9(5):351-65). Importantly, the priming status reflects a dependence on anti-apoptotic BCL-2 family proteins and can be characterized with the BH3 profiling functional assay. This assay measures induction of mitochondrial outer membrane permeabilization (MOMP) in response to treatment with peptides derived from BH3-only proteins (Ni Chonghaile, T., et. al. Science 2011 Nov 25;334(6059);1129-33). Thus, the unique BH3 profile associated with a specific malignant cell population, is a function of the anti-apoptotic BCL-2 family member/s contextual expression in that cell population. Based on our observations of potent 5-Aza sensitization in combination with anti-apoptotic BCL-2 protein family inhibition, we hypothesized that BH3-profiling would predict response to 5-Aza. To address the potential utility of BH3 profiling in predicting response to 5-Aza in myeloid malignancies, we assayed a broad panel of AML-derived cell lines (N=13) by BH3 profiling and correlated BH3 metrics with 5-Aza response. Identical cell line passages were assayed by BH3 profiling and in 5-Aza drug dose response assays. The cell panel comprised lines derived from AML FAB subtypes M7, M6, M5, M4, M2 and M0, as well as diverse cytogenetics such as t(11;21), t(9;11), t(4;11), t(6;11), del 5q, del 7, and a broad spectrum of mutations such as FLT3, N-RAS, CDKN2A, NPM1 and DNMT3A. The panel included a normal karyotype AML cell line (CG-SH) (Munker, R., et. al. Leuk Res 2009 Oct:33(10):1405-8) and a blastic cell line derived from a patient with MDS (MDS-L) (Tohyama, K., et. al. Br J Haematol 1994 Jun;87(2):235-42). Partition modeling using several BH3 metrics discriminates 5-Aza response with statistical significance (N=13, Mann-Whitney p<0.01) between more sensitive (EC50<2uM) and less sensitive (EC50>2uM) AML lines. For instance, the Puma BH3 peptide alone distinguishes more sensitive from less sensitive cell lines (N=13, Mann-Whitney p=0.0046). Several more complex parameters, such as “Puma+Hrk”, “Puma+Noxa,” and “Bim+Puma+Noxa+Hrk,” also significantly discriminate 5-Aza response (Mann-Whitney p=0.0011). Using continuous variable analysis, a Puma+Hrk metric exhibited the strongest correlation with 5-Aza response (R2=0.85, p<0.0001), while an individual Puma metric yielded the second strongest correlation (R2=0.70, p=0.0004). In conclusion, BH3 profiling discriminates 5-Aza response and confirms a central role for anti-apoptotic BCL-2 members in 5-Aza response. Furthermore, HRK priming indicates that BCL-XL is an important, although not the sole, anti-apoptotic component determining 5-Aza response in myeloid cells. These BH3 profiling results are in agreement with our functional 5-Aza RNAi and BCL-XL/BCL-2 inhibitor data in combination with 5-Aza. Based on these results, we are currently evaluating BH3 profiles from 5-Aza responding and non-responding AML and MDS patients to validate BH3 profiling-derived metrics in predicting clinical outcomes to 5-Aza. Disclosures: Pierceall: Eutropics Pharmaceuticals: Employment, Equity Ownership. Lena:Eutropics Pharmaceuticals: Employment. Mesa:Sanofi: Research Funding; Incyte: Research Funding; Lilly: Research Funding; NS Pharma: Research Funding; YM Bioscience: Research Funding. Cardone:Eutropics Pharmaceuticals: Employment, Equity Ownership.

Abstract Introduction/Background Recurrent deletions of chromosome 5 and 7 are characteristic of the myeloid malignancies MDS and AML, determine prognosis and influence therapeutic decision. The pathogenetic contribution of individual genes to MDS/AML development has been shown for several of these genes. However, how and if these genetic losses may relate to therapeutic interventions or represent haploinsufficient targets has remained elusive for the most part. We hypothesized that genes found within the commonly deleted regions (CDRs) of chromosome 5/7 (chr 5/7) represent targetable vulnerabilities in malignant myeloid cells. Therefore we generated a custom RNAi library to functionally interrogate these genomic regions alone or under 5-Azacitidine (5-Aza) treatment, to identify individual genes whose silencing modulate 5-Aza responsiveness. Methods/Materials Custom siRNA plates against ~270 genes (3x sequences/gene, Qiagen) on CDRs of chr 5/7 were assembled. Genes were silenced by siRNA for 48 hours followed by 48-72 hour 5-Aza treatment after which cell viability was determined. Four cell lines (TF1, THP1, MDS-L and HEL) were investigated in parallel with combined siRNA/5-Aza incubations and appropriate 5-Aza and siRNA only control. Hits were selected as > 2 standard deviation changes in viability from the median log2 value of the ratio [(siRNA + 5-Aza)/(siRNA only)], normalized per plate and across the entire screen. Duplictae RNAis screens were performed. Synergy between 5-Aza and the SMO inhibitors LDE225 (Sonidegib) and GDC0449 (Vismodegib) was assessed with CalcuSyn in a panel of AML cell lines. Ex vivo viability and clonogenic re-population experiments in primary patient derived AML and MDS cells were performed with Sonidegib. Results Several genes within the Hedgehog (HH) pathway emerged as sensitizers to 5-Aza when silenced by siRNA in three of four cell lines examined. Specifically Shh siRNA sensitized to 5-Aza in both TF-1 and THP-1 cells, SMO in HEL and GLI3 in THP-1 cells. Shh silencing alone (siRNA only) resulted in reduction of viability only in TF-1 cells but not in THP-1 cells, while SMO and GLI3 siRNA alone did not result in significant reductions in viability but sensitized to 5-Aza, indicating true sensitization and an interactions between the HH pathway and 5-Aza treatment. The SMO inhibitor Vismodegib synergized with 5-Aza in drug dose response assays in a panel of four AML cell lines (TF-1, THP-1, ML-2, HL-60) with Combination Index (CI) values of around 0.6 or lower at low doses of both drugs. Similar results were also observed with Sonidegib in vitro in 5 molecularly heterogeneous AML cell lines. Given stem cell modulation capacity of SMO inhibition, we examined ex vivo proliferation and more importantly clonogenic growth capacity. Highly interesting, ex vivo viability was reduced and synergy observed by the Sonidegib/5-Aza combination to a much greater extend in CD34+ selected primary MDS and AML cells as compared to bulk or CD34 depleted (normal) MNCs. Clonogenic growth was reduced in the combination over single agent 5-Aza or Sonidegib in ~ 50% of samples assessed to date. No direct correlations to cytogenetics were observed. A clinical Phase 1b trial was designed based on these results and is currently accruing in the first cohorts. Conclusions RNAi screening of CDRs of chr 5/7 yielded potential targetable therapeutic vulnerabilities with several genes within the HH pathway sensitizing to 5-Aza. SMO inhibitors pharmacologically validate this concept in vitro and ex vivo with the potential to more selectively affect leukemia stem/progenitor cells. Clinical data will show if this is a specific effect involving chr 5/7 or a general concept in malignant myeloid cells. Thus, SMO inhibition in combination with 5-Aza may be a novel rational combination in AML and MDS. Disclosures Off Label Use: LDE225/Sonidegib on a clinical trial. Mesa:Incyte, CTI, NS Pharma, Inc., Gilead, Celgene: Research Funding.

Abstract Hypomethylating agents have changed treatment and outcomes in MDS and are active in AML patients. However a significant number of patients do not respond to 5-Azacytidine or Decitabine and there is no clinically validated assay that predicts response to 5-Azacytidine (5-Aza) or Decitabine (DAC). Clinical parameters and recently TET2 mutations have been identified as genetic predictors of response to 5-Aza. Several recent papers report on the possible predictive value of anti-apoptotic proteins, for example BCL2L10. In previous genome-scale RNA-interference (RNAi) screens anti-apoptotic BCL-2 family members, most potently BCL-XL, were identified as top targets whose down-regulation sensitize to 5-Aza. Results were confirmed with the BCL-XL and BCL-2 inhibitor ABT-737, as well as the BCL-2-specific inhibitor ABT-199, both of which sensitized myeloid cells to 5-Aza. Although it is expected that interfering with anti-apoptotic genes would sensitize to hypomethylating similar to cytotoxics, few of the other >900 genes from our RNAi sensitizer screens, including few of the 571 kinases, sensitized to 5-Aza as potently as inhibition of BCL-XL. Thus BCL-2 family members constitute an especially potent context to modulate the activity of 5-Aza. To assess if protein expression of the top targets BCL-XL, BCL-2 or MCL-1 correlate with and may explain a preferential sensitization, 577 primary AML samples were examined by Reverse Phase Protein Array (RPPA). There was substantial overlap of expression across and within FAB subgroups and cytogenetics, and expression of neither of these genes/proteins could explain the observed effects. This may be expected due to functional redundancies amongst the different anti-apoptotic BCL-2 family members, as well as distinct, yet overlapping binding affinities between anti- and pro-apoptotic BCL-2 proteins and BH3 peptides. Thus, we next aimed to functionally interrogate the overall balance of pro- and anti-apoptotic BCL-2 family members, also described as “primedness” by BH3 profiling. This assay uses peptides derived from pro-apoptotic BH3-only proteins to determine the capacity of cells to initiate apoptosis in response to pro-apoptotic BH3 molecules [anti-apoptotic BCL-2 molecules like BCL-XL and BCL-2 bind and inhibit BH3-only molecules]. This assay is a broad functional readout that incorporates many parameters including the consequences of varying BCL2L10 expression. To establish a proof-of-concept that the functional interrogation of specific apoptotic thresholds may be an optimal readout for 5-Aza response, we first assessed a broad panel of 13 AML-derived cell lines by BH3 profiling and in parallel 5-Aza drug dose response experiments. Several BH3 profiling metrics, including NOXA plus BIM, significantly correlated with 5-Aza sensitivity in vitro (Figure 1). Next, to correlate actual clinical responses to 5-Aza with BH3 profiling, specimens from 28 AML, MDS and MDS/MPN overlap patients treated with 5-Aza or DAC-based regimens and for whom clinical outcome was available, were assayed in BH3 assays. In the clinic, the best BH3 metrics were combined values from NOXA and BIM peptides similar to cell lines. NOXA plus BIM discriminated clinical responses defined as achieving any response (sensitive) versus a patient being resistance/refractory to 5-Aza/DAC based regimens with statistical significance (Mann-Whitney two-tailed p = 0.001). This was true for the entire group of patients (5-Aza- and DAC- regimens) with a Receiver Operating Characteristic (ROC) of an AUC=0.875, with a further increase for patients treated only with 5-Aza based regimens with an AUC=0.95.Figure 1BH3 metrics distinguish cell lines with higher EC 50 (> 2uM) from those with lower EC50 (generally < 1uM).Figure 1. BH3 metrics distinguish cell lines with higher EC 50 (> 2uM) from those with lower EC50 (generally < 1uM). In conclusion, due to significant expression overlap and functional redundancies of the BCL-2 family proteins, expression does not correlate with 5-Aza clinical or in vitro responses. However, BH3 profiling as a general functional readout of apoptotic primedness, significantly discriminated responses in patients as well as in vitro. Thus, BH3 profiling incorporates the entirety of pro- and anti-apoptotic molecules (including BCL2L10) and is promising to predict responses to hypomethylating agents. We propose to prospectively validate BH3 profiling as a predictive biomarker assay for 5-Aza or DAC based regimens. Disclosures: Pierceall: Eutropics: Employment, Equity Ownership. Lena:Eutropics: Employment, Equity Ownership. Cardone:Eutropics: Employment, Equity Ownership. Elashoff:Eutropics: Employment. Blake:Noel Blake: Employment, Equity Ownership.

Abstract Background: Histone deacetylase inhibitors (HDIs) have shown clinical activity in myeloid malignancies, albeit insufficient to justify single-agent therapy. Thus, we have sought targets/therapies that synergistically enhance the activity of HDIs for combination therapy. We conducted RNA-interference (RNAi) drug modifier screens with the HDI SAHA, and have previously reported convergence of multiple screen hits on different points of the p38-SAPK/JNK signaling pathway. However, in subsequent pharmacological studies, well-characterized p38α/β inhibitors SB2012190 and LY2228820, and JNK inhibitor SP600125 did not modulate SAHA anti-leukemic activity. Nonetheless, a putative MK2 substrate-selective p38 inhibitor known as CMPD1 (Biochemistry 2004 Sep 21; 43(37):11658-71) was found to potently synergize with SAHA in all AML cell lines tested (N=8) in a dose-dependent manner. Further, using ex vivo cultures of primary myeloid malignancies (N=14), CMPD1 showed strong, selective synergy with SAHA in malignant CD34+ isolated cells, as compared to little or no synergy in CD34-depleted cells. Furthermore, CMPD1 specifically synergized with SAHA and a similar HDI panobinostat, but was found to interact antagonistically with cytarabine, and only additively with azacitidine. Herein, we present additional data from our investigation to better understand the mechanism of CMPD1 synergy with HDIs for future clinical translation. Results: Concurrent cell cycle and apoptosis measurements by flow cytometry show that single-agent CMPD1 results in a potent G2/M arrest that is resolved over time without a significant induction of apoptosis, whereas the combination of SAHA + CMPD1 results in a similar or increased level of G2/M arrest, which conversely culminates in a significant induction of apoptosis. The synergistic dose of single-agent SAHA used in these flow cytometry studies did not cause any significant cell cycle changes, and did not induce significant apoptosis alone in the AML cell lines studied. CMPD1 is known to be a MK2 substrate-selective p38 inhibitor; however, we find that CMPD1 potently increases the phosphorylation of MK2 at doses strongly synergistic with SAHA. This is not a universal activity of p38 inhibitors in our in vitro AML models, as the p38 inhibitor SB202190 does not increase MK2 phosphorylation alone or in combination with SAHA. We further hypothesized that CMPD1 synergizes with SAHA through a transcription-based mechanism, thus we measured mRNA expression by next-generation sequencing after SAHA and CMPD1 treatment alone, and in combination, in an AML cell line TF-1. We find that SAHA + CMPD1 treatment induces a synergistic increase in transcript levels of: i) several chemokines such as IL-8, CXCL1, CXCL2, and CXCL3, ii) inhibitor of DNA binding proteins ID2 and ID3, and iii) cell cycle regulator p21. Transcript levels of the Hippo pathway transcription factor YAP1 were also found to be increased by CMPD1 treatment in TF-1. We confirm an increase of total YAP1 protein levels by western blot. We also find an increase of YAP1 phosphorylation at serine 127 in CMPD1 or CMPD1 + SAHA treated samples in several AML cell lines, yet we find that Ser-127 phosphorylated YAP1 is located in the nucleus. Consistent with YAP1 Ser-127 phosphorylation, we find that upstream signaling kinase LATS1 is activated/phosphorylated at serine 909 in AML cell lines after treatment with CMPD1 or SAHA + CMPD1. We also confirm a synergistic increase in p21 protein levels upon SAHA + CMPD1 treatment by western blot in several AML cell lines. Changes in the protein levels of ID2 and ID3 were confirmed in TF-1 by western blot; however, ID protein levels appear to be uniquely affected by CMPD1 or the combination in a cell line-dependent manner. Conclusion: CMPD1 selectively synergizes with histone deacetylases inhibitors in myeloid malignancies. The synergistic activity of this combination highlights the potential for selectively targeting malignant CD34+ cells. The phosphorylation of LATS1 observed is consistent with i) the known roles for LATS1 in cell cycle regulation, ii) the G2/M arrest observed with CMPD1, and iii) the downstream phosphorylation of YAP1 observed. These findings linking CMPD1 with LATS1/YAP1 await further study and characterization. Although CMPD1 is not a clinically-relevant compound, CMPD1 may be used as a scaffold or tool compound to enable future development and research. Disclosures Mesa: Pfizer: Research Funding; Novartis Pharmaceuticals Corporation: Consultancy; Genentech: Research Funding; NS Pharma: Research Funding; Promedior: Research Funding; Gilead: Research Funding; CTI Biopharma: Research Funding; Incyte Corporation: Research Funding.

Abstract Background: Histone deacetylase (HDAC) inhibitors (HDIs) vorinostat (SAHA) and romidepsin are approved for the treatment of cutaneous T-cell lymphoma, while belinostat has recently been approved for the treatment of peripheral T-cell lymphoma. SAHA exhibits clinical activity in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML), although insufficient to justify single-agent therapy for these indications. The precise clinical mechanism of SAHA and other HDIs remains incompletely characterized, impeding the design and implementation of rational HDI-based therapeutic combinations for myeloid malignancies. To circumvent this lack of mechanistic understanding, we conducted RNA-interference (RNAi) modifier screens of the kinome and phosphatome to identify signaling pathways that modulate SAHA (HDI) anti-leukemic activity in AML cell lines TF-1, HEL and THP-1 for rapid translation of findings from ex vivo validation to clinical application. Results: Multiple screen hits (both kinases and phosphatases) converge on distinct points of the p38-SAPK/JNK signaling pathways, suggesting involvement of these stress-activated pathways in SAHA anti-leukemic activity. However, the most well characterized p38α/β inhibitors SB202190 and LY2228820 do not augment SAHA anti-leukemic activity at any dose across the broad dose ranges tested. Similarly, the JNK inhibitor SP600125 does not modulate SAHA activity at doses that inhibit JNK. However, the putative MK2a substrate-selective p38 inhibitor CMPD1 (Boehringer-Ingelheim; Davidson, W., et. al., Biochemistry 2004 Sep 21;43(37):11658-71) selectively potentiates the activity of SAHA in all AML cell lines tested (N=8), as well as in ex vivo cultures of primary myeloid malignancies (N=14), including polycythemia vera, chronic myelomonocytic leukemia, MDS/myeloproliferative neoplasm overlap syndrome, MDS transformed to AML, and de novo AML. The observed dose-dependent synergy occurs at nanomolar doses consistent with the reported apparent inhibitory constant for CMPD1 inhibition of p38-dependent phosphorylation of MK2a. Further, SAHA + CMPD1 synergy is significantly greater in CD34+-selected as compared to CD34+-depleted cell populations ex vivo. Furthermore, CMPD1 is similarly synergistic with the HDI panobinostat in vitro, but interacts antagonistically with cytarabine and additively with 5-azacitidine. This suggests that CMPD1 interacts specifically with SAHA and panobinostat in AML. The synergistic interaction between CMPD1 and SAHA is independent of ROS induction, as pretreatment with N-acetylcysteine does not abrogate synergy, despite reducing single-agent SAHA activity. While CMPD1 is reported to be an MK2a substrate-selective p38 inhibitor, CMPD1 does not affect the regulation of several canonical MK2 targets at doses that are potently synergistic in AML cell lines, including the phosphorylation of HSP27 or AATF. To begin to probe the mechanism of CMPD1 and SAHA synergy more extensively, mRNA expression was measured by next-generation sequencing after SAHA and CMPD1 treatment alone and in combination in the AML cell line TF-1. Sequencing data is currently being analyzed and findings will be presented. Conclusion: RNAi knockdown of several p38-SAPK/JNK pathway components modulates SAHA activity in malignant myeloid cells; however, common pharmacological sensitizing targets remain elusive. We hypothesize that MK2 and/or a yet unrecognized pathway modulated by CMPD1 can be targeted for HDI combination therapy in malignant myeloid cells. Current and ongoing results from this study will provide validated targets for rational HDI combination in myeloid malignancies. Disclosures Off Label Use: LDE225/Sonidegib as investigational agent. Mesa:Incyte, CTI, NS pharma, Gilead, Celgene: Research Funding.

Abstract Abstract 1851 Background: In previous work we performed RNAi sensitizer screening of epigenetic modulation with 5-Azacytidine and the human kinome. Few kinase hits were identified. Therefore we asked if phosphatases might be involved in modulating epigenetic therapies. Thus in order to identify potentiators of the histone deacetylase inhibitor suberoylanilide hydroxamic acid (SAHA) we performed high-throughput RNA-interference (RNAi) lethality screening using the myeloid cell lines HEL, THP-1 and TF-1 in combination with SAHA. Methods: A small-interfering RNA (siRNA) library targeting 206 phosphatases with 4 sequences per target gene was used for this study. siRNA were transfected in a single-siRNA-per-well format in 384-well plates using commercial lipid-based transfection reagents. Non-silencing siRNA were included as negative controls to access non-specific siRNA toxicity and universal lethal siRNA were included as positive controls to indirectly assess functional transfection efficiency. After a 96 hour incubation period, cell viability/proliferation was measured with Cell Titer Glo. For analysis of screen data, log2 values of (SAHA + siRNA treatment/siRNA treatment only) were calculated. Potentiation was defined as two standard deviations from the median log2 ratio of all siRNA (excluding positive controls) on a plate-by-plate basis. Hits were defined as ≥ 2/4 siRNA sequences meeting potentiation criteria. Results: Transfection efficiency was 95, 90 and 70% for HEL, THP-1 and TF-1, respectively. All three cell lines were similarly sensitive to SAHA with IC50 values of 0.3, 0.4 and 0.6 μM for HEL, THP-1 and TF-1, respectively. Resulting SAHA concentrations for screens were between the IC5 to IC30. For THP-1, 5 phosphatases qualified according to the hit selection criteria applied indicating strong potentiation to SAHA. Of these 5 hits, 3 did not exhibit strong reductions in viability with siRNA alone, suggesting a SAHA-dependent relationship. For HEL and TF-1, although no phosphatases met the strict selection cutoff, there were several phosphatases exhibiting potentiation, suggesting biological relevance. The primary screening data and preliminary analysis will be validated with rigorous secondary experimentation and top hits for all cell lines will be investigated and advanced. Conclusions: While there are numerous kinase inhibitors in the clinic and in preclinical development, phosphatases are much less explored as therapeutic targets. Validated potentiating targets from this RNAi screen can facilitate the development of rational SAHA combination therapies. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 180 Background: Proliferative signals in myeloid malignancies/leukemias are often driven by kinase activating events. Phosphatases often counteract these effects, though some phosphatases are oncogenes as well. To characterize these opposing signaling networks and build a loss of function intervention map, we performed the first phosphatome RNAi screen in conjunction with a kinome RNAi screen in myeloid leukemia cells. Methods: Four myeloid cell lines, TF-1, THP-1, HEL and MDS-L were incubated with siRNA against 206 phosphatases and 711 kinases with 4 different siRNA sequences (4x) per gene. After 96 hours, viability/proliferation was assessed by luminescence with Cell Titer Glo. Tranfection efficacy was determined by the ratio of the median of an universal lethal control vs. all siRNA. Normalization was done calculating the percent of sample median according to % proliferation = [siRNA signal × 100] / [median all siRNA signals]. Hits were defined using the median absolute deviation (MAD) of > 2 standard deviation above or below the individual siRNA signal (Birmingham et al, Nature Methods 2009) for at least 2 of 4 siRNA per gene. Results: Transfection efficacy was around 70, 95, 95 and 80 percent for TF-1, THP-1, HEL and MDS-L cells respectively. In TF-1, five kinases met hit selection criteria reducing proliferation whereas surprisingly no phosphatase was a hit. In THP-1, five kinases each reduced or increased proliferation respectively and only one phosphatase was a hit altogether increasing proliferation upon siRNA silencing. A different trend was seen in HEL cells: siRNA silencing of 16 kinases decreased and silencing of one kinase increased proliferation. Interestingly, inhibition of six phosphatases reduced proliferation in HEL whereas none increased proliferation. In MDS-L cells, siRNA inhibition of 9 kinases reduced and of 1 kinases increased viability respectively; inhibition of two phosphatases increased and of one decreased viability. The most common universal hit was PLK1 across 3 of 4 cell lines. No phosphatase was found to be a common hit across several lines. Conclusion: RNAi kinome-phosphatome profiling of 917 phosphatases and kinases reveals distinct signaling patterns among representative myeloid leukemia cell lines. Some cell lines, like HEL, which harbors a Jak2 mutation, are possibly more driven by kinase activating events but also show putative phosphatases as therapeutic targets. Candidate kinases whose siRNA silencing reduces viability/proliferation may uncover synthetic lethal interactions between known mutational and genomic abnormalities. The identified siRNA hits are potential novel targets for drug development. This combined functional genomics approach also highlights the complexity of signaling and that inhibition of some kinases can accelerate growth at least in vitro, e.g. inhibition of PIM3 in THP-1 cells. Phosphatases are an emerging network of cellular regulators with occasional oncogenic potential. A “loss-of-phosphatome/ase” RNAi approach is a fast way to gain new insights into the complex and little researched field. Disclosures: No relevant conflicts of interest to declare.