Summary Leukemia stem cells (LSC) play a pivotal role in chronic myeloid leukemia (CML) tyrosine kinase inhibitor (TKI) resistance and progression to blast crisis (BC), in part, through alternative splicing of self-renewal and survival genes. To elucidate splice isoform regulators of human BC LSC maintenance, we performed whole transcriptome RNA sequencing; splice isoform-specific qRT-PCR, nanoproteomics, stromal co-culture and BC LSC xenotransplantation analyses. Cumulatively, these studies show that alternative splicing of multiple pro-survival BCL2 family genes promotes malignant transformation of myeloid progenitors into BC LSC that are quiescent in the marrow niche and contribute to therapeutic resistance. Notably, a novel pan-BCL2 inhibitor, sabutoclax, renders marrow niche-resident BC LSC sensitive to TKIs at doses that spare normal progenitors. These findings underscore the importance of alternative BCL2 family splice isoform expression in BC LSC maintenance and suggest that combinatorial inhibition of pro-survival BCL2 family proteins and BCR-ABL may eliminate dormant LSC and obviate resistance.

3776 Cumulative evidence suggests that dormant self-renewing leukemia stem cells (LSC) contribute to relapse and blast crisis transformation by evading therapies that target cycling cells. Previously, sonic hedgehog (Shh) signaling was shown to modulate cell cycle regulation and self-renewal in normal mouse hematopoietic stem cells. However, its role in human LSC regeneration and quiescence had not been elucidated. Here we investigated the role of Shh signaling in maintenance of dormancy. We show that, compared to chronic phase CML and normal progenitors, human blast crisis LSC harbor enhanced expression of the Shh transcriptional activator, GLI2, and decreased expression of a transcriptional repressor, GLI3. Treatment of human blast crisis LSC engrafted RAG2−/−gc−/− mice with a selective Shh inhibitor, PF-04449913, reduced leukemic burden in a niche-dependent manner commensurate with GLI downregulation. Full transcriptome RNA sequencing performed on FACS-purified human progenitors from PF-04449913 treated blast crisis LSC engrafted mice demonstrated greater Shh gene splice isoform concordance with normal progenitors than vehicle treated controls. In addition, RNA sequencing revealed significantly decreased cell cycle regulatory genes expression and splice isoform analysis demonstrated reversion towards a normal splice isoform signature for many cell cycle regulatory genes. Moreover, cell cycle FACS analysis showed that selective Shh inhibition permitted dormant blast crisis LSC to enter the cell cycle while normal progenitor cell cycle status was unaffected. Finally, PF-04449913 synergized with BCR-ABL inhibition to reduce blast crisis LSC survival and self-renewal in concert with increased expression of Shh pathway regulators. Our findings suggest that selective Shh antagonism induces cycling of dormant human blast crisis LSC, rendering them susceptible to BCR-ABL inhibition, while sparing normal progenitors. Implementation of novel LSC splice isoform detection platforms to assess efficacy of Shh inhibitor-mediated sensitization to molecularly targeted therapy may inform dormant cancer stem cell elimination strategies that ultimately avert relapse. Disclosures: Levin: Pfizer Oncology: Employment; Pfizer Oncology: Equity Ownership.

3739 Leukemia stem cells (LSC) represent a frequently dormant self-renewing population integral to the initiation, maintenance, and progression of human chromic myeloid leukemia (CML). The current standard of care dasatinib, a BCR-ABL targeted tyrosine kinase inhibitor (TKI), effectively eradicates the bulk of CML cells but frequently fails to affect the LSC population that is thought to drive CML relapse. Members in the BCL2 family are proteins that regulate apoptosis, 6 of which regulate cell survival. Each of these 6 members has a long and short isoform with opposing functions; generally, long isoforms promote cell survival while the short isoforms promote apoptosis. Previously, we demonstrated that upregulation of pro-survival BCL2 proteins in CML LSC contributes to chemotherapy resistance and LSC quiescence in protective hematopoietic niches. LSC found in different hematopoietic niches differ in their response to TKI treatment. Niche affects LSC cell cycle, either by maintaining quiescence or by promoting rapid cell cycling. Quiescent cells are a hurdle for traditional chemotherapy, which usually targets rapidly cycling cells, leaving the quiescent LSC untouched. We hypothesize that the inhibition of pro-survival BCL2 protein family members will sensitize LSC to dasatinib therapy and therefore prevent CML relapse. We tested a novel pan pro-survival BCL2 family protein inhibitor, sabutoclax, delivered by intravenous injection either alone or in combination with oral dasatinib in immunodeficient RAG2−/-gc−/- mice engrafted with BC CML patient samples. After treatment, LSC burden, self-renewal, and cell cycle status were quantified using FACS. Our results showed a reduction in the LSC burden in combination treated mice when compared to mice that received either drug alone. Mice treated with the combination regimen were found to have fewer quiescent human leukemic cells than their counterparts that received single agent treatments. Immunofluorescence staining confirmed the reduction of quiescent cells in the bone marrow after combination treatment when compared to single agent or vehicle treatments. We validated the molecular targets by using human specific splice isoform primers to perform RT-qPCR on FACS sorted LSC and showed a reduction in the BCL2 long to short isoform ratio in sabutoclax versus vehicle treated animals, indicating a skewing towards the pro-apoptotic splice variant. Together, these results indicate that the combination strategy with a pan pro-survival BCL2 family inhibitor and a tyrosine kinase inhibitor may be the foundation for a promising clinical strategy to effectively eliminate LSC and prevent cancer progression and relapse. Disclosures: No relevant conflicts of interest to declare.

516 Introduction: Several studies have demonstrated the role of leukemia stem cells (LSC) in the development and maintenance of human chronic myeloid leukemia (CML). These cells, which first develop in chronic phase CML (CP CML) with acquisition of the BCR-ABL fusion protein, are often quiescent and can be highly resistant to apoptosis induced by drugs and radiotherapy that target rapidly dividing cells. Data has also shown that CML LSC become increasingly resistant to BCR-ABL inhibition with progression to blast crisis CML (BC CML). Bcl-2 family proteins are key regulators of apoptosis and have been shown by numerous studies to regulate cancer resistance to chemotherapy. This family of proteins has also been implicated in the development of BC CML, however most studies have focused on CML cell lines and their expression of Bcl-2 family proteins in vitro. Thus, there is relatively little data on expression of Bcl-2 family proteins in primary CML LSC and on the role of these proteins in regulating chemotherapy resistance in CML LSC in vivo. As Bcl-2 family proteins are known regulators of chemotherapy resistance we hypothesized that human BC CML LSC may overexpress these proteins compared to normal hematopoietic stem cells. We analyzed Bcl-2 family mRNA and protein expression in CP CML and BC CML LSC and compared this expression to normal cord blood stem and progenitor cells. We also analyzed whether these cells were sensitive to chemotherapy treatment in vitro. Finally, we tested whether a high potency pan-Bcl-2 inhibitor, 97C1, could effectively kill CML LSC in vitro and in vivo. Methods: Bcl-2 and Mcl-1 protein expression was measured in primary CP CML, BC CML, and normal cord blood cells using intracellular FACS. We also measured Bcl-2, Mcl-1, Bcl-X, and Bfl-1 mRNA expression in FACS sorted CD34+CD38+lin− cells (LSC) from these samples. For all drug studies we used either serially transplanted CD34+ cells derived from primary BC CML patient samples or primary CD34+ normal cord blood cells. In vitro drug responses were tested by culturing CD34+ cells either alone or in co-culture with a mouse bone marrow stromal cell line (SL/M2). Effects on colony formation and replating were also tested by culturing sorted CD34+CD38+lin− cells in methylcellulose in the presence and absence of drug. For in vivo testing of 97C1 we transplanted neonatal RAG2-/-yc-/- mice with CD34+ cells from 3 different BC CML and cord blood samples. Transplanted mice were screened for peripheral blood engraftment at 6–8 weeks post-transplant and engrafted mice were then treated for 2 weeks with 97C1 by IP injection. Following the treatment period the mice were sacrificed and hemotapoietic organs were analyzed for human engraftment by FACS. Results: BC CML progenitors expressed higher levels of Bcl-2 and Mcl-1 protein compared to normal cord blood and chronic phase CML cells. mRNA expression of Mcl-1, Bcl-X, and Bfl-1 was also increased in BC CML progenitors compared to CP CML progenitors. While BC CML LSC cultured in vitro were resistant to etoposide and dasatinib-induced cell death, 97C1 treatment led to a dose-dependent increase in cell death along with a dose-dependent decrease in the frequency of CD34+CD38+lin− cells compared to vehicle treated controls. While cord blood progenitor cells were also sensitive to 97C1 treatment they had an IC50 around 10 times higher than that for the BC CML cells (100nM versus 10nM). Importantly, 97C1 treatment did not inhibit cord blood colony formation or colony replating in vitro. Mice transplanted with BC CML LSC developed CML in 6–8 weeks post-transplant with diffuse myeloid sarcomas and engraftment of human CD34+CD38+lin− cells in the peripheral blood, liver, spleen, and bone marrow. In vivo treatment with 97C1 led to a significant reduction in both total human engraftment and engraftment of CD34+CD38+lin− cells in all hematopoietic organs analyzed. Conclusion: Our results demonstrate that BC CML LSC are resistant to conventional chemotherapy but are sensitive to 97C1 in vitro and in vivo. Broad-spectrum inhibition of Bcl-2 family proteins may help to eliminate CML LSC while sparing normal hematopoietic stem and progenitor cells. Disclosures: Jamieson: CoronadoBiosciences: Research Funding; CIRM: Research Funding.

Introduction Chronic myeloid leukemia (CML) was one of the first malignancies shown to be initiated in hematopoietic stem cells by the BCR-ABL1 oncogene and sustained in blast crisis (BC) by progenitor cells that co-opt stem cell properties and behave like leukemia stem cells (BC LSCs). The BCR-ABL fusion oncogene encodes a constitutively active tyrosine kinase BCR-ABL. Although tyrosine kinase inhibitor (TKI) therapy targeting BCR-ABL suppresses CML during the chronic phase (CP), progenitors undergo expansion as a consequence of subsequent genetic and epigenetic alterations that fuel blast crisis transformation, BC LSC generation and TKI resistance. Self-renewing human BC LSCs harbor increased expression of Inflammation responsive adenosine deaminase acting on RNA (ADAR1), which can alter transcript as well as microRNA (miRNA) maturation, splicing and translation by Adenosine (A)-to-Inosine (I) editing of double stranded RNA. miRNAs are a family of small non-coding RNA molecules that regulate gene expression at a post-transcriptional level by inhibiting protein translation and/or reducing mRNA stability. Eukaryotic cells employ miRNAs in diverse biological processes including cell proliferation, differentiation, pluripotency and self-renewal. The stem cell pluripotency RNA binding protein LIN28B plays critical roles in BC transformation of CML. In this study we sought to characterize CML related-oncogenes, such as BCR-ABL, JAK2 and ADAR1, alone or in stromal co-culture in terms of their ability to regulate LSC self-renewal through modulation of let-7 /LIN28B stem cell transcriptional regulatory axis. Methods MiRNAs were extracted from purified CD34+ cells derived from CP and BC CML patient samples as well as cord blood by RNeasy microKit (QIAGEN) and let-7 expression was evaluated by qRT-PCR using miScript Primer assay (QIAGEN). CD34+ cord blood (n=3) were transduced with lentiviral human BCR-ABL, JAK2, let-7a, wild type ADAR1 and ADAR1 mutant, which lacks a functional deaminase domain. Then, 72 hours after transduction, lentivirally transduced cells were plated on irradiated SL/M2 cells. After 5 days of culture, cells were collected for RNA and microRNA extraction. Transduction efficiency and LIN28B levels were evaluated by qRT-PCR and let-7 expression was quantified by qRT-PCR using miScript primer assay. Hematopoietic Progenitor and Replating assaywere performed on lenti-let-7a-overexpressing CB cells to assess differentiation, survival and self-renewal capacity. Results Lentiviral overexpression of human BCR-ABL in CD34+ CB did not induce any significant change in let-7 family members and LIN28B expression in absence of stromal co-culture. However, stromal co-culture of BCR-ABL overexpressing CB led to the significant downregulation of members of the let-7 family as well as to upregulation of their target gene LIN28B, thus suggesting that extrinsic microenvironmental cues are necessary for modulating let-7 family levels in presence of BCR-ABL. Notably, qRT-PCR of CB transduced with JAK2 showed significant upregulation of ADAR1 in the absence of stroma, thus suggesting that JAK2 might be a mediator of inflammatory cytokine-driven ADAR1 activation. Lentiviral overexpression of both human JAK2 and ADAR1 significantly reduced the expression of let-7 family members and induced up-regulation of LIN28B. Interestingly, lentiviral overexpression of ADAR1 mutant did not induce any significant change in most let-7 family members. Finally, lentiviral overexpression of let-7a induced significant reduction in survival and self-renewal. Conclusion These finding suggest that BCR-ABL requires extrinsic signals from the niche to modulate self-renewal of BC LSCs. Conversely, lentiviral JAK2 overexpression induces activation of aberrant RNA editing and subsequent reduction of let-7 family members in the absence of the niche. Interestingly, experiments with ADAR1 mutant, suggest that ADAR1 downregulates most of the let-7 family members in a RNA–editing dependent way manner. In summary these findings suggest a novel mechanism for BC LSC generation that may have utility in prognostication and selective LSCs targeting. Disclosures Jamieson: J&J: Research Funding; Sanofi: Research Funding, Travel Support, Travel Support Other; Roche: Research Funding.