MLL-ENL Fusion Research Articles

Acute Myeloid Leukemia (AML) from Older HSCs Is Resistant to Chemotherapy and Has an Altered DNA Damage Response

Acute myeloid leukemia (AML) is an aggressive malignancy that leads to the accumulation of immature myeloid precursors, resulting in progressive marrow failure and death. AML arises from the hematopoietic stem cell (HSC), either directly or through an early myeloid precursor. HSCs must survive for the entire human lifespan to provide the 200+ billion blood cells that are produced daily. They are subjected to the entirety of the aging process. Further, while AML is diagnosed in every age range, it disproportionately affects older patients, with a median age at diagnosis of 68. Older age, typically defined as 60 years of age or older, is a powerful, poor prognostic factor. Clinical trials show an estimated 5-year survival of less than 10% in these patients. Ten-year survival is even worse with one study estimating 2.4% of patients over 60 years of age alive at ten years when treated with chemotherapy. Importantly, this poor outcome is the result of disease that is significantly more resistant to therapy. Even though this clinical data is well known the underlying mechanisms of therapy resistance in older AML are unknown. To address this, we generated genetically identical AML cells from fetal, 12-week-old (Young) or 24-month-old (Old) murine HSCs. Retroviral vectors were used to express the oncogenic MLL-ENL fusion protein and NRas G12D. AML was confirmed by morphology, immunophenotype and all AML cells were lethal when injected into syngeneic recipients. Old AML cells demonstrated reduced autophagy, decreased NAD levels and impaired mitochondria consistent with a significant effect of biological aging. When treated with DNA damaging chemotherapy, Old AML cells were significantly more resistant than either Young or Fetal cells to both cytarabine and the anthracycline doxorubicin. To examine the DNA damage response AML cells were treated with doxorubicin and stained for the early DNA damage marker gH2AX. Old AML cells generated significantly more gH2AX than either Fetal or Young AML cells (Figure 1). We next examined the induction of p53 in response to treatment and in contrast to the gH2AX results Old AML cells induced the least amount of p53 in response to doxorubicin treatment (Figure 2). This data demonstrates that Old AML cells were more adept at recognizing DNA damage but least likely to induce p53 in response to it suggesting a difference in DNA repair. To better understand these differences RNA sequencing was performed and differentially expressed genes (DEGs) identified. When Fetal and Young AML cells were compared, no DEGs reached the false discovery threshold for significance, demonstrating that Fetal and Young AMLs have highly similar gene expression profiles. In contrast, when Old AML cells were compared to either Young or Fetal cells, hundreds of highly significant DEGs were identified, and KEGG pathway enrichment analysis revealed that several DNA repair pathways were significantly upregulated in Old AML. These include base excision repair (FDR p= 8.60E-06), mismatch repair (FDR p=3.26E-05), and nucleotide excision repair (FDR p= 0.00043). To further validate these pathways, the Beat AML TCGA dataset was assessed for the prognostic effect of increased expression of key genes in these pathways. Patients with increased expression of these genes had a trend toward worse survival (p=0.07) consistent with these pathways being clinically relevant. These data suggest that AML arising in older patients may overexpress DNA damage repair pathways leading to chemotherapy resistance and poor outcomes. Additional work on ways to increase sensitivity to Old AML cells is in progress.

The ENL YEATS epigenetic reader domain critically links MLL-ENL to leukemic stem cell frequency in t(11;19) Leukemia.

MLL (KMT2a) translocations are found in ~10% of acute leukemia patients, giving rise to oncogenic MLL-fusion proteins. A common MLL translocation partner is ENL and associated with a poor prognosis in t(11;19) patients. ENL contains a highly conserved N-terminal YEATS domain that binds acetylated histones and interacts with the PAF1c, an epigenetic regulator protein complex essential for MLL-fusion leukemogenesis. Recently, wild-type ENL, and specifically the YEATS domain, was shown to be essential for leukemic cell growth. However, the inclusion and importance of the YEATS domain in MLL-ENL-mediated leukemogenesis remains unexplored. We found the YEATS domain is retained in 84.1% of MLL-ENL patients and crucial for MLL-ENL-mediated leukemogenesis in mouse models. Mechanistically, deletion of the YEATS domain impaired MLL-ENL fusion protein binding and decreased expression of pro-leukemic genes like Eya1 and Meis1. Point mutations that disrupt YEATS domain binding to acetylated histones decreased stem cell frequency and increased MLL-ENL-mediated leukemia latency. Therapeutically, YEATS containing MLL-ENL leukemic cells display increased sensitivity to the YEATS inhibitor SGC-iMLLT compared to control AML cells. Our results demonstrate that the YEATS domain is important for MLL-ENL fusion protein-mediated leukemogenesis and exposes an "Achilles heel" that may be therapeutically targeted for treating t(11;19) patients.

Knockout of HMGA1 Promotes ATRA-Induced Differentiation in MLL-Rearranged Leukemia

Introduction: HMGA1 is a small non-histone chromatin protein characterized by 3 AT-hooks, which modulates transcription by altering the chromatin architecture. It tends to be highly expressed in poorly differentiated cells including hematopoietic stem cells. In various kinds of solid tumors, HMGA1 has been reported to play as an oncogene, and its upregulation was reported to be associated with poorer clinical outcomes. However, although recent study demonstrated that HMGA1 was associated with progression of JAK2V617F positive myeloproliferative neoplasms to myelofibrosis or leukemia, the roles of HMGA1 in myeloid malignancies were still unclear. Therefore, we analyzed its function in myeloid malignancies using HMGA1 knockout (KO) mice. We could not detect clear differences in peripheral blood counts or bone marrow cells, and generated various kinds of models of acute myeloid leukemia (AML). Knockout of HMGA1 could not delay the onset of these leukemia, but we could demonstrate that leukemic cells with MLL-ENL fusion gene which lacked in HMGA1 gene were more susceptible to all-trans retinoic acid (ATRA)-induced expression of CD11b, which suggested that HMGA1 could be associated with the differentiation process induced by ATRA. Method: We generated Vav-iCre HMGA1flox/flox mice, and we first analyzed the fraction of peripheral blood cells and bone marrow cells of these mice using flow cytometry. Next, to evaluate roles of HMGA1 in AML, we made various murine models of AML. We collected c-kit positive fractions of bone marrow cells from the KO mice or Vav-iCre(-) HMGA1flox/flox mice and retrovirally introduced MLL-ENL, MOZ-TIF2, BCL2-MYC fusions. Primary transplantation was performed to sublethally irradiated (5Gy) mice with 4.0×105 infected cells for each fusion gene. Secondary transplantation was also performed for MLL-ENL and BCL2-MYC with 1.0×104 cells for each. Blood samples were collected weekly, and disease-free survivals were compared by Kaplan-Meier method with Vav-iCre(-) HMGA1flox/flox mice. Also, leukemic cells infected with MLL-ENL were collected, and these cells were cultured for four days with different concentrations (1μM, 5μM or 10μM) of ATRA. Flow cytometry analysis was performed to evaluate the differentiation induced by ATRA with antibodies to murine CD11b, c-kit and Gr-1. Result: There was no significant difference in birth rate between Vav-iCre(+) and Vav-iCre(-) mice, which suggested that HMGA1 was not essential for fetal hematopoiesis. Peripheral blood cell counts were measured at week 8 and week 20, and fractions of peripheral blood and bone marrow cells were assessed at week 8 using flow cytometry, but no significant differences were observed. As for AML models, there were no significant differences in survival periods after primary or secondary transplantation. Leukemic cells retrovirally induced with MLL-ENL were collected from primary recipients, and these cells were treated with ATRA. Although growth rates were not show significantly different, expression levels of CD11b after exposure to ATRA were significantly higher in Vav-Cre (+) mice (5μM: 70% vs 44%, respectively; p=0.035. 10μM: 80% vs 46%, respectively; p=0.0006) as shown in the following figure. Discussion: We established HMGA1 KO mice and assessed the function of HMGA1 both in normal and leukemic hemopoiesis. Although HMGA1 might have little effect on normal hematogenesis, it seemed that HMGA1 had inhibitory effect on ATRA-induced differentiation process of MLL-ENL leukemic cells. Though more conditions including AML types and differentiation methods should be tested and the molecular mechanisms should be scrutinized, there is a possibility that inhibition of HMGA1 is effective in combination with differentiation treatments. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Abstract 2966: The epigenetic reader function of the YEATS domain in MLL-ENL fusion critically affects leukemic stem cell frequency in MLL-ENL leukemia

Abstract MLL1 (KMT2A) translocations are found in ~10% of acute leukemia and give rise to an aggressive form of leukemia in infant, pediatric and adult patients. MLL1 fusion driven acute leukemia is characterized by deregulated activity of the Super Elongation Complex (SEC) and the H3K79 methyltransferase DOT1L, which alter the epigenetic landscape and transcription of pro-leukemic MLL1 fusion targets like HoxA9 and Meis1. The Eleven-Nineteen-Leukemia gene (ENL or MLLT1) is a common MLL1 fusion partner and a SEC component. The ENL protein contains a highly conserved N-terminal epigenetic reader YEATS domain that recognizes acetylated H3K9/K18/K27 (H3Kac hereinafter). Wild type ENL was recently found to be essential for leukemic cell growth, which is dependent on its YEATS domain interaction with H3Kac. While this finding highlighted the YEATS domain importance in wild type ENL function in leukemic cells, the inclusion and importance of the YEATS domain in MLL-ENL fusion protein remain to be elucidated. Here, we investigated the clinical relevance and importance of the ENL YEATS domain in MLL-ENL leukemias. We analyzed >300 t(11;19) MLL-ENL leukemia patients for the breakpoint location within the ENL gene and found that the YEATS domain is retained in the resultant MLL-ENL fusion protein in 84.1% of t(11;19) leukemia patients. We tested the importance of the YEATS domain in MLL-ENL mouse models and found that the YEATS domain and downstream sequence is required for MLL-ENL leukemogenesis in vivo. YEATS deletion decreased expression of pro-leukemic targets such as Meis1, an important factor for leukemic stem cells (LSC). To interrogate the contribution of the YEATS epigenetic reader function in MLL-ENL leukemogenesis, we introduced YEATS point mutations rendering the domain defective in interacting with H3Kac and found that this significantly increased leukemia latency in vivo. Further investigation revealed that YEATS point mutations disrupting H3Kac binding significantly decreased MLL-ENL LSC frequency while not affecting homing to the bone marrow. We attribute this LSC frequency change to altered Meis1 expression. Additionally, disruption of the YEATS epigenetic reader function in MLL-ENL leukemia cells does not induce differentiation, apoptosis nor cell cycle arrest. Therapeutically, we predicted the YEATS domain in MLL-ENL would sensitize MLL-ENL leukemia to YEATS domain inhibitors. Indeed, MLL-ENL leukemia cells are more sensitive to the ENL/AF9 YEATS domain inhibitor, SGC-iMLLT, compared to acute leukemia cells driven by other fusions. Together, our results demonstrate that YEATS-H3Kac binding plays an important role in MLL-ENL fusion mediated leukemogenesis. Our data establishes a strong rationale for future exploration of small molecules aimed at disrupting the YEATS-H3Kac interaction as a targeted therapeutics for treating t(11;19) leukemia patients. Citation Format: Hsiangyu Hu, Nirmalya Saha, Ejaz Ahmad, Yuting Yang, Lili Chen, Lauren Lachowski, Blaine Teahan, Sierrah Grigsby, Rolf Marschalek, Zaneta Nikolovska-Coleska, Andrew G. Muntean. The epigenetic reader function of the YEATS domain in MLL-ENL fusion critically affects leukemic stem cell frequency in MLL-ENL leukemia [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2966.

A somatic mutation in moesin drives progression into acute myeloid leukemia.

Acute myeloid leukemia (AML) arises when leukemia-initiating cells, defined by a primary genetic lesion, acquire subsequent molecular changes whose cumulative effects bypass tumor suppression. The changes that underlie AML pathogenesis not only provide insights into the biology of transformation but also reveal novel therapeutic opportunities. However, backtracking these events in transformed human AML samples is challenging, if at all possible. Here, we approached this question using a murine in vivo model with an MLL-ENL fusion protein as a primary molecular event. Upon clonal transformation, we identified and extensively verified a recurrent codon-changing mutation (Arg295Cys) in the ERM protein moesin that markedly accelerated leukemogenesis. Human cancer-associated moesin mutations at the conserved arginine-295 residue similarly enhanced MLL-ENL–driven leukemogenesis. Mechanistically, the mutation interrupted the stability of moesin and conferred a neomorphic activity to the protein, which converged on enhanced extracellular signal–regulated kinase activity. Thereby, our studies demonstrate a critical role of ERM proteins in AML, with implications also for human cancer.

Reciprocal Regulation of BCL6 and MLL Expression As a Vulnerability in MLL -Rearranged Acute Lymphoblastic Leukemia

Background: Chromosomal translocations involving the Mixed Lineage Leukemia gene MLL account for ~70% of infant acute lymphoblastic leukemias (ALLs; Ayton and Cleary, 2001) and are associated with dismal prognosis. These findings underscore the importance of understanding the molecular mechanisms underlying disease progression of MLL-rearranged (MLLr) leukemia for the development of novel therapeutics. Recently, an integrative epigenomic analysis has identified the transcriptional repressor and proto-oncogene BCL6 as a transcriptional target in MLL-AF4 -driven pre-B ALL (Geng et al., 2012). BCL6 functions as a mediator of drug resistance to tyrosine kinase inhibitors in Ph+ ALL (Duy et al., 2011). Here, we tested the hypothesis that BCL6 is required for MLLr leukemia and represents a therapeutic target in MLLr pre-B ALL.Results: Western blot analysis of a panel of patient-derived pre-B ALL cells (MLLr : n= 7; ETV6-RUNX1 : n= 2; BCR-ABL1 : n= 1; not otherwise specified: n= 4) and normal CD19+ B cells (n= 3) revealed that BCL6 levels were elevated in MLL r ALL samples. Furthermore, immunohistochemistry staining in bone marrow biopsies from ALL patients (n= 72) showed that BCL6 levels were upregulated in MLLr cases. Expression of MLL-ENL or MLL-AF4 fusion protein increased Bcl6 protein levels (~10-fold) in murine pre-B cells. Moreover, ChIP-seq analysis of human MLLr pre-B ALL cell lines using antibodies against MLL, AF4 and ENL demonstrated direct binding to the BCL6 promoter. Studying methylation status revealed that the BCL6 promoter was hypomethylated in patients with MLLr pre-B ALL compared to patients with other subtypes of pre-B ALL (including BCR-ABL1, ETV6-RUNX 1 and hyperdiploid) as well as normal pre-B cells from healthy donors (St. Jude, n= 132, Figueroa, 2014). Consistent with these observations, multivariate analysis of high risk pre-B ALL patients showed that high BCL6 levels in combination with MLL rearrangements were associated with the worst clinical outcome (COG P9906). Collectively, our findings demonstrated that BCL6 levels are elevated in MLLr ALL, and that expression levels of MLL and BCL6 can serve as predictors for risk stratification in ALL.Interestingly, expression of wild-type MLL was positively correlated with that of BCL6 in both pediatric (COG P9906) and adult (ECOG E2993) pre-B ALL. Furthermore, multivariate analysis comparing MLLHighBCL6High patients with MLLLowBCL6Low patients showed that high expression levels of both genes correlated with poor clinical outcome (COG P9906). These findings prompted us to study whether MLL regulates BCL6 expression, and vice versa . Cre-mediated deletion of Mll abrogated imatinib-mediated induction of Bcl6 in BCR-ABL1 -driven pre-B ALL, indicating that Mll is required for upregulation of Bcl6 expression. While genetic inactivation of Bcl6 as well as treatment with the BCL6-specific retro-inverso peptide-inhibitor (RI-BPI) decreased Mll expression, doxycycline-induced expression of Bcl6 upregulated Mll expression. Combining chromatin immunoprecipitation and DNA microarray (ChIP-on-chip), we found that BCL6 was recruited to the MLL promoter in pre-B ALL. Taken together, these findings support our proposed scenario of reciprocal regulation of MLL and BCL6 expression.Notably, ChIP-seq analyses of MLLr pre-B ALL cells showed that there are 268 common gene targets of both MLL and BCL6. Among the common targets include cell cycle checkpoint regulators (CDKN1B, ETS2, BMI1, RB1), apoptosis facilitator BIM, and tumor suppressor TET2, indicating that BCL6 is a key regulator of the transcriptional program in MLLr pre-B ALL. In support of an essential role of BCL6 in MLLr ALL, expression of a dominant negative BCL6 mutant in patient-derived MLLr pre-B ALL cells induced rapid depletion from cell culture in competitive-growth assays. Furthermore, RI-BPI treatment resulted in cell cycle arrest, impaired colony forming ability and inhibited leukemia initiation in transplant recipient mice. Importantly, RI-BPI in combination with standard chemotherapy Vincristine exerted more potent effects on killing MLLr pre-B ALL cells than either agent alone.Conclusions: Our findings suggest that aberrant BCL6 expression levels are important for disease progression, and that pharmacological inhibition of BCL6 augments responsiveness to standard chemotherapy in the treatment of MLLr pre-B ALL. DisclosuresHurtz:Incyte: Research Funding. Armstrong:Cyteir: Consultancy; Imago: Consultancy; C4 Therapeutics: Consultancy; Epizyme: Consultancy.

Tip60 activates Hoxa9 and Meis1 expression through acetylation of H2A.Z, promoting MLL-AF10 and MLL-ENL acute myeloid leukemia.

Chromosome translocations involving the MLL gene are common rearrangements in leukemia. Such translocations fuse the MLL 5'-region to partner genes in frame, producing MLL-fusions that cause MLL-related leukemia. MLL-fusions activate transcription of target genes such as HoxA cluster and Meis1, but the underlying mechanisms remain to be fully elucidated. In this study, we discovered that Tip60, a MYST-type histone acetyltransferase, was required for the expression of HoxA cluster and Meis1 genes and the development of MLL-fusion leukemia. Tip60 was recruited by MLL-AF10 and MLL-ENL fusions to the Hoxa9 locus, where it acetylated H2A.Z, thereby promoting Hoxa9 gene expression. Conditional deletion of Tip60 prevented the development of MLL-AF10 and MLL-ENL leukemia, indicating that Tip60 is indispensable for the leukemogenic activity of the MLL-AF10 and MLL-ENL-fusions. Our findings provide novel insight about epigenetic regulation in the development of MLL-AF10 and MLL-ENL-fusion leukemia.

Unlike Its Paralog LEDGF/p75, HRP-2 Is Dispensable for MLL-R Leukemogenesis but Important for Leukemic Cell Survival.

HDGF-related protein 2 (HRP-2) is a member of the Hepatoma-Derived Growth Factor-related protein family that harbors the structured PWWP and Integrase Binding Domain, known to associate with methylated histone tails or cellular and viral proteins, respectively. Interestingly, HRP-2 is a paralog of Lens Epithelium Derived Growth Factor p75 (LEDGF/p75), which is essential for MLL-rearranged (MLL-r) leukemia but dispensable for hematopoiesis. Sequel to these findings, we investigated the role of HRP-2 in hematopoiesis and MLL-r leukemia. Protein interactions were investigated by co-immunoprecipitation and validated using recombinant proteins in NMR. A systemic knockout mouse model was used to study normal hematopoiesis and MLL-ENL transformation upon the different HRP-2 genotypes. The role of HRP-2 in MLL-r and other leukemic, human cell lines was evaluated by lentiviral-mediated miRNA targeting HRP-2. We demonstrate that MLL and HRP-2 interact through a conserved interface, although this interaction proved less dependent on menin than the MLL-LEDGF/p75 interaction. The systemic HRP-2 knockout mice only revealed an increase in neutrophils in the peripheral blood, whereas the depletion of HRP-2 in leukemic cell lines and transformed primary murine cells resulted in reduced colony formation independently of MLL-rearrangements. In contrast, primary murine HRP-2 knockout cells were efficiently transformed by the MLL-ENL fusion, indicating that HRP-2, unlike LEDGF/p75, is dispensable for the transformation of MLL-ENL leukemogenesis but important for leukemic cell survival.

Investigating the Roles of the Yeats Domain in MLL-ENL Mediated Leukemogenesis

Approximately 10% of acute leukemia involves rearrangement at chromosome 11q23, giving rise to a relatively aggressive form of acute leukemia characterized by MLL1 (KMT2A) fusion proteins. Despite the identification of >100 MLL1 fusion partners, the majority are members of several similar transcriptional activation complexes including: The Super Elongation Complex (SEC), AEP and EAP (SEC used hereafter). MLL fusion-driven acute leukemia is characterized by deregulated activity of the SEC and the H3K79 methyltransferase DOT1L. This leads to altered epigenetic landscapes at and deregulated transcription of pro-leukemic MLL1-fusion target genes like HoxA9 and Meis1. Thus, targeting these transcriptional and epigenetic complexes has become an attractive therapeutic strategy for treating MLL-fusion leukemia. Eleven-Nineteen-Leukemia (ENL or MLLT1) is the third most common MLL1 fusion partner and a component of the SEC. Recently, wild type ENL was identified as an essential factor for leukemic cell growth. The ENL protein possesses a C-terminal ANC-homology domain (AHD) necessary for SEC recruitment and is essential for MLL-fusion mediated leukemogenesis. In addition, ENL contains a highly conserved N-terminal YEATS domain that functions as an epigenetic reader for acetylated H3K9, H3K18 or H3K27, which is essential for leukemic cell growth. Additionally, the ENL YEATS domain directly interacts with the Polymerase Associated Factor 1 complex (PAF1c), an epigenetic regulator protein complex essential for MLL-fusion mediated leukemogenesis. These studies highlight the importance of the YEATS domain in regulating wild type ENL function in leukemic cells. However, the importance of the YEATS domain in the context of MLL-ENL mediated leukemia remains to be elucidated. In this study, we investigate the clinical relevance and leukemic importance of the ENL YEATS domain in MLL-ENL leukemias. We first analyzed t(11;19) (MLL-ENL) patient data to determine the sites of chromosomal translocation within the ENL gene. We found that the YEATS domain (coded by exons 2 through 4) is retained in 84.1% of MLL-ENL patients (n=302). Specifically, 50.7% (n=153) of these patients possess breakpoints located 5' of the first exon of the ENL gene, while 33.4% (n=101) of the patients display breakpoints within the first intron of ENL gene. These data point towards a tendency for YEATS domain retention in MLL-ENL fusion proteins in t(11;19) patients. We next tested whether the YEATS domain was functional in MLL-ENL mouse leukemia models. Our data shows the YEATS domain is required for MLL-ENL leukemogenesis in vivo, as deletion of the YEATS domain destroys MLL-ENL leukemogenesis and increases apoptosis in cell culture. Transcriptionally, deletion of the YEATS domain decreased expression of pro-leukemic genes such as Meis1 and the anti-apoptotic gene Bclxl. To dissect the contribution of different YEATS domain functions in MLL-ENL leukemogenesis, we engineered YEATS domain mutants defective in interacting with PAF1 or acetylated H3K9/K18/K27. Disrupting the YEATS-PAF1 or YEATS-H3Kac interaction decreased MLL-ENL mediated colony formation exvivo and significantly increased leukemia latency in vivo. The MLL-ENL YEATS domain mutants will be used in future studies to determine how the YEATS domain affects 1) MLL-ENL fusion localization, 2) key protein complexes localization (i.e. SEC and PAF1c) and 3) the epigenetic landscapes (i.e. H3K79me2/3 and H3K4me3) at pro-leukemic targets. To further interrogate the YEATS-PAF1 interaction in MLL-ENL mediated leukemia, we identified the minimal region of the PAF1 protein required for the YEATS-PAF1 interaction. This PAF1 protein fragment will be used to biochemically characterize the structure of the PAF1-YEATS interaction, which might aid in therapeutically targeting specific YEATS interactions in MLL-ENL leukemia. Our results demonstrate for the first time, to our knowledge, an essential role for the YEATS domain in MLL-ENL mediated leukemogenesis. Additionally, our genetic studies elucidate the importance of the YEATS domain interaction with either the PAF1c or H3Kac in MLL-ENL leukemias. Taken together, our study establishes a rationale for exploring the effectiveness of small molecule development aimed at disrupting either the YEATS-H3Kac or the YEATS-PAF1 interaction as a therapeutic intervention for treating MLL-ENL leukemia patients. Disclosures No relevant conflicts of interest to declare.

3088 – DISSECTING THE ONTOGENIC REMODELING OF THE PROTEOMIC LANDSCAPE AND FUNCTIONALITY OF HEMATOPOIETIC PROGENITOR CELLS IN NORMAL DEVELOPMENT AND LEUKEMIA

The process of hematopoiesis is subject to extensive ontogenic remodeling that is accompanied by alterations in cellular fate both during normal development and disease. Although the functional differences between fetal and adult hematopoiesis are well established, the responsible molecular mechanisms have long remained largely unexplored on the proteomic level. In our latest work, we characterized and compared the proteomic makeup of fetal and adult HSPCs (Jassinskaja et al., 2017, Cell Reports). Here, we have utilized state-of-the-art mass spectrometry to gain deep coverage of the proteome of 100,000 fetal and adult lymphoid-primed multipotent progenitors (LMPPs), common lymphoid progenitors (CLPs) and granulocyte-monocyte progenitors (GMPs). Our analysis resulted in the identification and quantification of over 4000 proteins, with 200-300 proteins per cell type displaying differential expression between the fetus and the adult. We uncovered ontogenic changes in expression of several transcription factors with critical roles in hematopoiesis. Together with other proteins involved in lineage commitment and differentiation, we show that protein expression in fetal and adult hematopoietic progenitor cells is predictive of differential lineage potential. Our subsequent functional analyses further emphasize that the proteomic differences affect cell fate both in the context of normal differentiation and leukemic hematopoiesis driven by the MLL-ENL fusion oncogene. Collectively, our work represents a significant advancement in the understanding of the molecular programs that govern ontogenic differences in hematopoiesis and provides a solid foundation for future investigation of which factors are responsible for the difference in susceptibility and outcome of different leukemias in infants and in adults.

Rationale for Targeting BCL6 in MLL-Rearranged B-ALL

Background and significance: MLL-gene rearrangements occur in ~10% of B-cell acute lymphoblastic leukemia (B-ALL) and 70% of infant B-ALL and define a group of patients with dismal outcomes. MLL belongs to the family of histone methyltransferases and plays a critical role in maintaining hematopoietic stem cells (Jude et al. 2007). Analyzing gene expression data from 207 children with high-risk ALL, we found that BCL6 is a predictor of poor clinical outcome in B-ALL, particularly in cases with MLL rearrangements. Thus, while the mechanisms of drug-resistance in MLL-rearranged B-ALL are largely unknown, we studied here a potential role of BCL6 in promoting the aggressive and refractory phenotype in this subtype of B-ALL. Results: Immunohistochemical staining of bone marrow biopsies from 10 of 11 MLL-rearranged B-ALL samples studied revealed aberrant expression of BCL6, a transcription factor that promotes oncogenic B-cell transformation and drug-resistance in B-ALL (Duy et al. Nature 2011). Our genetic and ChIP-seq analyses showed that MLL-AF4 and MLL-ENL fusions directly bound to the BCL6 promoter and upregulated BCL6 expression. While oncogenic MLL-fusions strongly induced aberrant BCL6 expression in B-ALL cells, germline MLL was required to upregulate Bcl6 in response to physiological stimuli during normal B-cell development. Inducible expression of Bcl6 increased MLL mRNA levels, which was reversed by genetic deletion and pharmacological inhibition of Bcl6, suggesting a positive feedback loop between MLL and BCL6. Highlighting the central role of BCL6 in MLL-rearranged B-ALL, conditional deletion and pharmacological inhibition of BCL6 compromised leukemogenesis in transplant recipient mice and restored sensitivity to vincristine chemotherapy in MLL-rearranged B-ALL patient samples. Oncogenic MLL-fusions strongly induced transcriptional activation of the pro-apoptotic BH3-only molecule BIM, while BCL6 was required to curb MLL-induced expression of BIM. Notably, peptide (RI-BPI) and small-molecule (FX1) BCL6-inhibitors derepressed BIM and synergized with the BH3-mimetic ABT-199 in eradicating MLL-rearranged B-ALL cells. These findings uncover MLL-dependent transcriptional activation of BCL6 as a previously unrecognized requirement of malignant transformation by oncogenic MLL-fusions and identified BCL6 as a novel target for the treatment of MLL-rearranged B-ALL. Discussion: Pharmacological inhibition of BCL6 using a retro-inverso peptide inhibitor (RI-BPI) compromised MLL-rearranged B-ALL leukemia-initiation and subverted vincristine-resistance. A central mechanistic aspect of BCL6-function in MLL-rearranged B-ALL involves transcriptional repression of the pro-apoptotic BH3-only molecule Bim (BCL2L11). Profiling for BH3-only proteins in various B-ALL subtypes revealed that MLL-rearranged B-ALL is associated with increased expression of the pro-apoptotic protein BIM (Benito et al. 2015). ABT-199 (venetoclax) is a BH3-mimetic that disrupts the interaction between BCL2 and BIM, leading to BIM release and induction of apoptosis. Previous studies have found that MLL-rearranged leukemia cells are sensitive to the BCL2 inhibitor ABT-199. Besides BCL2, we here identified BCL6 as a central antagonist of pro-apoptotic BIM function in MLL-rearranged B-ALL cells. In genetic experiments, we showed that oncogenic MLL-fusions strongly activated BCL2L11 transcription reinforcing the notion that constitutively high Bim expression levels represents an important and selective vulnerability in MLL-rearranged B-ALL. Both BCL2 and BCL6 represent crucial antagonists of BIM in MLL-rearranged B-ALL, BCL2 mediating BIM-sequestration and BCL6 being required for transcriptional repression of BIM. In support of this scenario, peptide (RI-BPI) and small molecule (FX1) inhibitors of BCL6 strongly synergized with blockade of BCL2-mediated BIM sequestration (ABT-199) in killing MLL-rearranged B-ALL cells. Figure 1 Disclosures Armstrong: AstraZeneca: Research Funding; Epizyme, Inc.: Consultancy, Equity Ownership; Imago Biosciences, Inc.: Consultancy, Equity Ownership; Cyteir Therapeutics: Consultancy, Equity Ownership; Janssen: Research Funding; Novartis: Research Funding; Accent Therapeutics: Consultancy, Equity Ownership; Mana Therapeutics: Consultancy, Equity Ownership; C4 Therapeutics: Consultancy, Equity Ownership; Syros Pharmaceuticals: Consultancy, Equity Ownership; OxStem Oncology: Consultancy, Equity Ownership. Melnick:Janssen: Research Funding; Constellation: Consultancy; Epizyme: Consultancy. Milne:OxStem Ltd.: Equity Ownership.

Mesothelin Targeting Bites for Pediatric AML: In Vivo Efficacy and Specificity

Acute myeloid leukemia (AML) remains the type of pediatric leukemia with poorest outcome. Despite maximally intensive therapy, approximately 20% of patients experience recurrent disease. Novel targeted therapies are needed to improve survival. We recently identified that mesothelin, a well-validated target in some cancers, is also highly expressed in a subset of pediatric AML samples (Tarlock et al., Blood, 128:2873, 2016). Considering that it is not expressed in normal tissues in children (Fan et al., Blood, 130:3792, 2017), MSLN is a viable target for immunotherapies such as Bispecific T-cell Engaging antibodies (BiTEs) that combine antibody single chain variable (scFv) regions targeting a cancer antigen and the T-cell co-receptor CD3. We designed and tested the efficacy and specificity of BiTEs targeting MSLN in disseminated xenograft models of pediatric AML. Using scFv sequences derived from Amatuximab, which recognizes the N-terminal domain of the GPI-linked ectodomain of MSLN, targeting region 1 of MSLN, and from Blinatumomab/AMG-330 targeting CD3, we engineered and expressed two kinds of BiTE molecules - a canonical BiTE and an IgG BiTE, a larger molecule with improved serum half life in vivo. To evaluate the specificity and efficacy of canonical BiTEs, MV4;11-MSLN cell line was generated by lentiviral transduction of parental MV4;11 cells which do not constitutively express MSLN (Fig. 1A, B). These two cell lines were injected i.v. into NSG-SGM3 mice. Once engraftment was confirmed, purified human T cells (3 x 106) were injected to act as effector cells. Mice were then treated with the canonical αMSLN-αCD3 BiTE at a dose of 3 mg/kg/day daily for 6 days. A cohort of mice that were untreated or received BiTE or T-cell infusion only served as controls. Mice from both treated and untreated groups had to be euthanized when they presented with distended abdomens due to myeloid sarcomas and no significant differences in survival were observed. Post euthanasia, bone marrows were flushed and evaluated for the percentage of AML cells (human CD45+CD33+) and T cells (human CD45+CD3+). We observed that the αMSLN-αCD3 BiTE was effective in promoting T-cell activation (based on high T-cell counts compared to mice injected with T-cells alone) and greatly reducing leukemic burden in mice injected with MV4;11 cells engineered to express MSLN (Fig. 1C, D). Similar results were obtained using BiTEs targeting a different MSLN epitope. No T-cell expansion and anti-leukemic effect was observed in mice engrafted with parental MV4;11 cells. Although, there were no significant differences between the median survival of untreated and treated miceThese data highlight the specificity and efficacy of the aMSLN-CD3 BiTEs. Among a panel of 8 AML patient-derived xenograft (PDX) lines generated in the laboratory, NTPL-146 bearing MLL-ENL fusion was found to have endogenous MSLN expression (Fig. 1E). We evaluated the efficacy of αMSLN-αCD3 canonical BiTE (3 mg/Kg Qdx6) against NTPL-146 PDX line in NSG-B2m mice by transfusing human CD3+ T-cells to act as effector cells. A Kaplan-Meier survival plot based on the time when each mouse reached experimental end-point (reduced body weight greater than 20%, impaired mobility, hind limb paralysis) showed that the survival benefit for mice receiving BiTE in the presence of human T-cells (4/6 mice survived at the end of experiment) greatly exceeded the efficacy of T-cells alone (22-day improvement in median survival with no surviving mice), or BiTE treatment alone (no improvement in survival) compared to untreated mice (Fig. 1F, P<0.001). These data validate the efficacy of MSLN targeting BiTEs in a PDX model with endogenous MSLN expression. The efficacy of canonical vs IgG BiTEs was evaluated in MV4;11-MSLN xenografted mice. Mice were dosed Qd5x3 for canonical BiTE and Q7dx3 for IgG BiTE as shown (Fig. 1G). IgG BiTE treatment along with T-cell infusion significantly prolonged survival in mice transplanted with MV4;11-MSLN (Fig. 1H), suggesting that IgG BiTE was far more efficacious than canonical BiTEs (P<0.01). Taken together, these data indicate that MSLN-targeting BiTEs could be used as novel immunotherapy for pediatric AML with MSLN expression. Figure 1 Disclosures Kaeding: Celgene: Employment.

Single Cell Transcriptomics Predicts Relapse in Infants with Acute Lymphoblastic Leukemia

Care of infants (<1 year of age) diagnosed with MLL (KMT2A)-rearranged acute lymphoblastic leukemia (MLLr-iALL) suffers from two major drawbacks. First, a poor survival rate due to a high rates of early relapse and chemo-resistance. Additionally, the approximately 30-50% of patients that do survive, suffer from life-long, debilitating side-effects of current treatment. While almost all MLLr-iALL patients show an initial promising response to treatment, two-third of the patients relapse, typically within the first year from diagnosis and while still on treatment. Accurate relapse prediction would allow implementation of treatment strategies that take relapse risk into account, with great potential benefit for all patients. Here, we show that Single-cell RNA sequencing (scRNA-seq) can be valuable for risk stratification and that the abundance of chemo-resistant cells within the diagnosis sample might be a powerful indicator of the likelihood of relapse. We have used scRNA-seq to analyze the response to treatment of leukemic cells in bone marrow biopsies of seven MLLr-iALL patients, expressing either the oncogenic MLL-AF4 or MLL-ENL fusion gene, at the time of initial diagnosis. Three of these patients successfully underwent treatment and remained disease-free during 7 years of follow-up, while in the remaining four cases the disease returned within a year from diagnosis. All samples were subjected to scRNA-seq by FACS index sorting with the aim of identifying differences between early relapsers and long-term survivors. Quantification of the proportion of cells classified by single cell transcriptomics, categorized as either chemo-resistant or chemo-sensitive, accurately predicts the occurrence of future relapse in individual patients. Strikingly, the single cell-based classification is even consistent with the order of relapse timing. Additionally, leukemic cells associated with high relapse risk are typified by a small phenotype, which coincides with an apparent quiescent gene expression pattern. This study clearly and, to the best of our knowledge, for the first time shows how disease classification and treatment management can directly benefit from single cell genomics. It demonstrates how classification based on a pivotal functional characteristic of single cells can be performed, despite individual patient variation. Our results shed light on the subpopulation from which leukemic relapse originates, and opens up opportunities for strong, risk-based strategies for future MLLr-iALL treatment regimens. Disclosures Pieters: medac: Consultancy; jazz farmaceuticals: Consultancy.

Rationale for targeting BCL6 in MLL-rearranged acute lymphoblastic leukemia.

Chromosomal rearrangements of the mixed lineage leukemia (MLL) gene occur in ∼10% of B-cell acute lymphoblastic leukemia (B-ALL) and define a group of patients with dismal outcomes. Immunohistochemical staining of bone marrow biopsies from most of these patients revealed aberrant expression of BCL6, a transcription factor that promotes oncogenic B-cell transformation and drug resistance in B-ALL. Our genetic and ChIP-seq (chromatin immunoprecipitation [ChIP] combined with high-throughput sequencing) analyses showed that MLL-AF4 and MLL-ENL fusions directly bound to the BCL6 promoter and up-regulated BCL6 expression. While oncogenic MLL fusions strongly induced aberrant BCL6 expression in B-ALL cells, germline MLL was required to up-regulate Bcl6 in response to physiological stimuli during normal B-cell development. Inducible expression of Bcl6 increased MLL mRNA levels, which was reversed by genetic deletion and pharmacological inhibition of Bcl6, suggesting a positive feedback loop between MLL and BCL6. Highlighting the central role of BCL6 in MLL-rearranged B-ALL, conditional deletion and pharmacological inhibition of BCL6 compromised leukemogenesis in transplant recipient mice and restored sensitivity to vincristine chemotherapy in MLL-rearranged B-ALL patient samples. Oncogenic MLL fusions strongly induced transcriptional activation of the proapoptotic BH3-only molecule BIM, while BCL6 was required to curb MLL-induced expression of BIM. Notably, peptide (RI-BPI) and small molecule (FX1) BCL6 inhibitors derepressed BIM and synergized with the BH3-mimetic ABT-199 in eradicating MLL-rearranged B-ALL cells. These findings uncover MLL-dependent transcriptional activation of BCL6 as a previously unrecognized requirement of malignant transformation by oncogenic MLL fusions and identified BCL6 as a novel target for the treatment of MLL-rearranged B-ALL.

Abstract 2040: Acute myeloid leukemia cells induce new bone formation by expanding osteoprogenitor niche in the bone marrow

Abstract Genetic alterations in osteoprogenitor cells have been shown to induce myeloid leukemia in mouse models. We reported recently that acute myeloid leukemia (AML) cells induce osteogenic differentiation in mesenchymal stromal cells (MSC) in the bone marrow (BM) to facilitate faster AML engraftment in mice (Battula et al., JCI Insight, 2017). However, specifics of this osteogenic niche generated by AML are not known. Here, we hypothesize that AML cells induce new bone formation by expanding osteoprogenitor-rich niche in the BM. To investigate the effect of AML cells on osteoprogenitor cells and mature osteoblasts, we generated transgenic mice by crossing Osx-CreERt2 mice with Ocn-GFP; ROSA-tdTomato mice. The resulting triple transgenic mice had the genotype of Osx-CreERt2;Ocn-GFP;ROSA-tdTomato. In these mice the tdTomato (red) positive cells represented cells that originated from Osterix-expressing (Osx+) cells, whereas a GFP+ (green) cell represented an osteocalcin-expressing (Ocn+) mature osteoblast. We implanted murine AML cells with MLL-ENL fusion proteins into these mice. Three weeks after implantation of AML cells, the femurs and tibia of these mice were subjected to histological evaluation using whole-tissue fluorescence microscopy. Interestigly, in the BM of mice implanted with AML cells, we found a 3-4 fold increase in Osx+ cells compared to control animals. However, no significant difference in the number of GFP+ cells on the endosteum and trabecular bone surface was observed, suggesting that AML cells expand osteoprogenitor cells in the BM. Next, to investigate AML-induced alterations in bone, we implanted AML patient-derived xenograft (PDX) cells with the FLT3-ITD translocation into non-obese diabetic scid interleukin-2Rγnull (NSG) mice. When the percentage of AML blasts in peripheral blood increased over 95%, we performed micro-computed tomography (µCT) to analyze changes in the bone architecture. Of note, we observed a massive increase in cortical bone thickness and new medullary bone formation in the diaphysis area. Quantitative analysis revealed that the bone volume and bone mineral content in AML mouse femurs were about twofold higher than those in healthy mouse femurs. To validate these findings, we stained femurs from mice with and without AML with Masson-Goldner's trichrome reagents. Notably, staining of the femurs from mice with AML revealed massive web-like bone formation within the medullary cavity, which is usually not seen in normal BM. Also, bone that was not fully formed, but in the process of maturation, stained light blue. In conclusion, our data suggest that AML cells expand osteoprogenitor cells resulting in thickening of the cortical bone and new bone formation in mouse models. In-depth analysis of bone remodeling in AML patients could result in new insights into the pathobiology of the disease and provide novel therapeutic avenues for AML. Citation Format: Bin Yuan, Stanley Ly, Ivo Veletic, Kiersten Maldonado, Xin Zhou, Benoit deCrombrugghe, Michael Andreeff, Venkata Lokesh Battula. Acute myeloid leukemia cells induce new bone formation by expanding osteoprogenitor niche in the bone marrow [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2040.

Acute Myeloid Leukemia Expands Osteoprogenitor Rich Niche in the Bone Marrow but Resorbs Mature Bone Causing Osteopenia/Osteoporosis in Animal Models

Acute myeloid leukemia (AML) is one of the most aggressive hematological malignancy that originates in the bone marrow (BM). Despite advances in the molecular characterization of AML, factors regulating its progression are still not known. Among several BM niches that support AML growth in the BM, the osteogenic niche has gained attention in recent years owing to its potential role in leukemogenesis. Genetic alterations in osteoprogenitor cells have been shown to induce myeloid leukemia in mouse models. We reported recently that AML cells induce osteogenic differentiation in mesenchymal stromal cells (MSCs) in the BM to facilitate faster AML engraftment in mice (Battula et al., JCI Insight, 2017). However specifics of this osteogenic niche generated by AML are not known. Here we hypothesize that AML expands osteo-progenitor rich niche in the BM, but that the mature bone is reduced.To determine the type of AML-induced osteo-lineage differentiation in the BM, we generated transgenic reporter mice by crossing Osx-CreERt2 mice with Ocn-GFP; ROSA-tdTomato mice. The resulting triple transgenic mice has the genotype of Osx-CreERt2;Ocn-GFP;ROSA-tdTomato. In these mice the tdTomato (red) positive cells represents osteo-lineage cells that originate from Osterix expressing (Osx+) cells, whereas a GFP+ cell represents an osteocalcin-expressing (Ocn+) mature osteoblast. Seven day old triple transgenic mice were injected with tamoxifen to activate Osx-CreERT2 to mark the Osx+ cells with tomato reporter. To investigate the osteogenic cell type that is induced by AML cells in the bone marrow, we implanted murine AML cells with MLL-ENL fusion proteins into Osx-CreERt2;Ocn-GFP;ROSA-tdTomato mice. Three weeks after implantation of AML cells, the femurs and tibia of these mice were dissected and subjected to histological evaluation using fluorescence microscopy. In control BM without AML, the GFP+ (Ocn+) cells were found in the trabecular bone surface as well as the periosteum of the bone, whereas the tdTomato+ (Osx+)cells were found in the marrow and the bone matrix; this suggests that some of the osteocytes originated from tamoxifen-induced Osx+ osteoprogenitor cells. Interestigly, in mice implanted with AML cells, we found a 3-4 fold increase in Osx+ cells in the marrow compared to normal BM (Fig 1A). However, the number of GFP+ cells on the endosteum and trabecular bone surface was reduced, suggesting that AML cells might expand osteoprogenitor cells but not fully differentiated mature osteoblasts.Next, to investigate whether AML cells affect the mature bone, AML PDX cells developed in our laboratory were implanted into NSG mice. The PDX models usually take 12-14 weeks to achieve >90% engraftment in the peripheral blood which provides ample time to observe alterations in bone composition. At this stage, the mice were subjected to computed tomography imaging to measure bone architecture, volume (BV), mineral density (BMD) and bone volume fraction (BVF). Interestingly, we observed large bone cavities close to epiphysis and metaphysis areas in the femur and tibia of mice with AML (Fig 1B). In addtion, BMD and BVF in these mice were reduced by 20-30% compared to control mice without leukemia. To validate the bone resorption in these mice, bone histomorphometric analysis was performed on femurs and tibias from mice with and without AML. Masson-Goldner's Trichrome staining revealed a 5- to 10-fold decrease in the trabecular and cortical bone thickness in AML femurs compared to normal femurs. Moreover, measurements of osteoclast activation by tartrate-resistant acidic phosphatase (TRAP) revealed positive staining for osteoclasts on the endosteal surface and massive bone resorption in AML bone compared to normal bone. Mechanistic studies showed that AML cells inhibit osteoprotegerin (OPG) ~10 fold in MSCs, a factor that inhibits the RNAK ligand which in turn activates osteoclasts that breakdown the bone.In conclusion, our data suggest that bone homeostasis is dysregulated in AML by induction of osteogenic and osteolytic activities simultaneously. AML cells induce an osteoprogenitor niche but also activate osteoclasts resulting in osteopenia/osteoporosis in mouse models. In-depth analysis of bone remodeling in AML patients could result in new insights into the pathobiology of the disease and provide therapeutic avenues for AML. DisclosuresAndreeff:Amgen: Consultancy, Research Funding; Oncolyze: Equity Ownership; United Therapeutics: Patents & Royalties: GD2 inhibition in breast cancer ; Daiichi-Sankyo: Consultancy, Patents & Royalties: MDM2 inhibitor activity patent, Research Funding; Celgene: Consultancy; Astra Zeneca: Research Funding; Jazz Pharma: Consultancy; Oncoceutics: Equity Ownership, Membership on an entity's Board of Directors or advisory committees; SentiBio: Equity Ownership; Eutropics: Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Aptose: Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Reata: Equity Ownership. Battula:United Therapeutics Inc.: Patents & Royalties, Research Funding.

ZNF521 sustains the differentiation block in MLL-rearranged acute myeloid leukemia.

Zinc finger protein 521 (ZNF521) is a multiple zinc finger transcription factor and a strong candidate as regulator of hematopoietic stem cell homeostasis. Recently, independent gene expression profile studies have evidenced a positive correlation between ZNF521 mRNA overexpression and MLL-rearranged acute myeloid leukemia (AML), leaving open the question on the role of ZNF521 in this subtype of leukemia. In this study, we sought to analyze the effect of ZNF521 depletion on MLL-rearranged AML cell lines and MLL-AF9 xenograft primary cells. Knockdown of ZNF521 with short-hairpin RNA (shRNA) led to decreased leukemia proliferation, reduced colony formation and caused cell cycle arrest in MLL-rearranged AML cell lines. Importantly, we showed that loss of ZNF521 substantially caused differentiation of both MLL-rearranged cell lines and primary cells. Moreover, gene profile analysis in ZNF521-silenced THP-1 cells revealed a loss of MLL-AF9-directed leukemic signature and an increase of the differentiation program. Finally, we determined that both MLL-AF9 and MLL-ENL fusion proteins directly interacted with ZNF521 promoter activating its transcription. In conclusion, our findings identify ZNF521 as a critical effector of MLL fusion proteins in blocking myeloid differentiation and highlight ZNF521 as a potential therapeutic target for this subtype of leukemia.

MerTK Receptor Tyrosine Kinase Inhibition As a Potential Strategy to Augment Immune-Mediated Clearance of Acute Myeloid Leukemia

The MerTK receptor tyrosine kinase is aberrantly expressed in 80-100% of acute myeloid leukemia (AML), and inhibition of MerTK prolongs leukemia-free survival in vivo, making it an appealing therapeutic target. MerTK is normally expressed on macrophages (MMs) and dendritic cells (DCs) and facilitates engulfment of apoptotic debris. Through this process, MerTK promotes tolerogenicity and anti-inflammatory cytokine production. We hypothesized that inhibition of MerTK will also mediate anti-leukemia effects through relief of microenvironment immunosuppression and stimulation of innate MM and DC activity leading to enhanced T-cell immunity.To test this hypothesis, we transplanted wild-type C57BL/6 mice with syngeneic murine MLL-ENL fusion AML or primary mouse MLL-AF9 AML. We evaluated the effects of MerTK inhibition using MRX-2843, a MerTK kinase inhibitor. Mice were treated daily and survival was monitored. Treatment with MRX-2843 prolonged median survival (43 days) compared to vehicle treatment (17 days) in mice transplanted with MLL-ENL AML (p<0.0001), and MLL-AF9 primary murine AML (42 days with MRX-2843, 30 days with vehicle; p<0.05).To explore the immune effects of MerTK inhibition in AML, we evaluated in vivoMMs and DCs by flow cytometry. Compared to non-leukemic controls, vehicle-treated leukemic mice demonstrated a significant infiltration of MMs and DCs into the primary sites of leukemic burden (spleen, bone marrow). These MMs and DCs expressed high levels of co-inhibitory ligands PD-L1 and PD-L2. Treatment with MRX-2843 significantly decreased PD-L1 and PD-L2 expressing MM and DCs in the spleen (p<0.01) and bone marrow (p<0.01) down to non-leukemic levels. Similar results were not obtained when mice were treated with the non-MerTK targeted FLT3 inhibitor AC220, suggesting that these effects are likely specific to MerTK inhibition.We next evaluated effects of MerTK inhibition on the leukemia microenvironment. We confirmed that T-cells do not express MerTK in-vivowith or without AML, nor +/- MRX-2843. Compared to non-leukemic mice, leukemic mice significantly upregulated co-inhibitory receptors PD-1 and Tim-3 on CD4 and CD-8 T-cells in the spleen and bone marrow. Treatment with MRX-2843 in leukemic mice significantly decreased CD4 and CD8 T-cell PD-1 and Tim-3 expression in the spleen (p<0.05) and bone marrow (p<0.05) down to non-leukemic levels. We next evaluated the role of functional T-cells in the anti-leukemia effects of MerTK inhibition. Wild-type mice and mice harboring a T-cell receptor mutation (TCRa-/-) were transplanted with MLL-ENL AML, treated with MRX-2843 or vehicle, and monitored for symptoms of leukemia. Median survival of all vehicle-treated mice with leukemia was 20-22 days. Median survival of TCRa-/- mice treated with MRX-2843 was 30 days, whereas median survival of MRX-2843 treated wild-type mice was significantly extended to 40 days (p<0.01). To confirm these findings, leukemic wild-type mice were CD8 depleted and treated as above. Median survival of CD8-depleted MRX-2843 treated mice was significantly diminished (30 days) compared to isotype (control) + MRX-2843 treatment (40 days, p<0.01). These data suggest an additional non-cell autonomous role of T-cell anti-tumor immunity in mice treated with a MerTK inhibitor.We next evaluated peripheral serum cytokines using Luminex magnetic bead analysis. Treatment with MerTK inhibition significantly increased serum [IL-18] (p<0.01) and [MCP-3/CCL7] (p<0.05) by 2-fold compared to vehicle-treated mice. Additionally, there was no correlation between co-inhibitor expression and serum [IFN-g]. These changes appeared to be mediated through MerTK and not solely attributable to tumor burden given that MLL-ENL cells in culture did not exhibit these changes +/-MRX-2843. However, MLL-ENL cells did secrete significantly less IL-4 (p<0.01) and LIF (p<0.05) with MerTK inhibition in vitrocompared to vehicle treatment.In conclusion, inhibition of MerTK decreased infiltration of immunosuppressive co-inhibitory ligand-expressing MMs and DCs into primary sites of leukemic engraftment. Furthermore, inhibition of MerTK altered the leukemia microenvironment to decrease T-cell co-inhibitory receptor expression and improve anti-leukemia effects. These preliminary studies demonstrate the potential of MerTK inhibition as an immunotherapeutic strategy in AML. DisclosuresWang:Meryx, Inc: Equity Ownership, Patents & Royalties: MRX-2843. Earp:Meryx: Employment, Patents & Royalties: MRX-2843. Frye:Meryx, Inc: Equity Ownership, Patents & Royalties: MRX-2843.

Leukemogenic MLL-ENL Fusions Induce Alternative Chromatin States to Drive a Functionally Dichotomous Group of Target Genes.

MLL fusions are leukemogenic transcription factors that enhance transcriptional elongation through modification of chromatin and RNA Pol II. Global transcription rates and chromatin changes accompanying the transformation process induced by MLL-ENL were monitored by nascent RNA-seq and ChIP-seq, revealing 165 direct target genes separated into two distinct clades. ME5 genes bound MLL-ENL at the promoter, relied on DOT1L-mediated histone methylation, and coded preferentially for transcription factors, including many homeobox genes. A distinct ME3 group accumulated MLL-ENL beyond the termination site, was dependent on P-TEFb-mediated phosphorylation of RNA Pol II for transcription, and translated mainly into proteins involved in RNA biology and ribosome assembly. This dichotomy was reflected by a differential sensitivity toward small molecule inhibitors, suggesting the possibility of a combinatorial strategy for treatment of MLL-induced leukemia.

Differential Regulation of c-Myc/Lin28 Discriminates Subclasses of Rearranged MLL Leukemia

Rearrangements of the 11q23 locus account for ~70% of infant ALL and ~50% of infant AML1 and about 10% of leukemia overall. The prognosis for 11q23 patients is generally poor, however, outcomes vary depending on the fusion partner2. Rearrangements fuse the N-terminus of MLL with one of >70 different partner genes that includes both nuclear and cytoplasmic proteins. Despite the different intracellular localization of these partner proteins, to date, all studied MLL fusion proteins (MLL-FPs) localize to chromatin in the nucleus and drive aberrant transcriptional activation. Recent seminal work by a number of groups has revealed that several of the most common nuclear translocation partner genes (including AF9, ENL, AFF1 (AF4), AFF4 (AF5q31), AF10, AF17 and ELL) assemble into a transcriptional activation complex that includes p-TEFb and/or the histone H3K79 methyltransferase DOT1l. Translocation of MLL with members of this complex results in deregulated transcriptional activation of target genes. Conversely, oligomerization motifs are necessary for transformation following 11q23 translocation with a cytoplasmic partner. However, these mechanisms fail to explain different survival outcomes observed in patients. Further, the transcriptional programs induced in these diverse 11q23 leukemias are currently not well understood. In this study, we examined the genome wide expression profiles in leukemic cells transformed by several MLL-FPs representative of nuclear translocations [t(9;11) (MLL-AF9), t(10;11) (MLL-AF10) and t(11;19) (MLL-ENL)] or cytoplasmic translocations [t(1;11) (MLL-AF1p), t(6;11) (MLL-AF6) and t(11;17) (MLL-Gas7)]. Leukemia cell lines established from mouse bone marrow cells expressing these MLL-FPs proliferated at different rates and mice transplanted with the transformed cells develop leukemia with different latencies remarkably consistent with differences observed in patients harboring different MLL translocations. To elucidate differences in the gene programs induced by different MLL-FPs we performed genome wide expression profiling by RNA-sequencing. These data demonstrated that while the MLL-AF9 and MLL-ENL fusion proteins induce very similar gene programs, the cytoplasmic fusion proteins (MLL-AF6, MLL-AF1p and MLL-GAS7) all possess unique gene signatures. We then performed a pathway analysis comparing nuclear fusion proteins and cytoplasmic fusion proteins and discovered the Myc transcription factor program as one of the top distinguishing features. Myc overexpression significantly increased the growth rate of slow-growing cells that also had low intrinsic Myc, while the growth rate change of more highly proliferative cells was minimal. While all leukemic cell lines were sensitive to the BET inhibitor JQ1 which regulates c-Myc expression, greater sensitivity was observed in those with low c-Myc expression demonstrating the universal importance of this gene program. The Myc target and micro RNA binding protein Lin28B is also differentially expressed between nuclear and cytoplasmic fusions. Negative regulation of miR-150 by Lin28B was observed in all MLL-FP cell lines, which is necessarily downregulated in 11q23 leukemias. We then investigated another Lin28B microRNA target let-7. Interestingly, let-7g expression was significantly increased in MLL-FP transformed cells associated with the longest disease latency. These data demonstrate that differential activation of the c-Myc/Lin28 program accounts for changes in let-7g expression and is associated with MLL-FP disease latency. These data also suggest that patients harboring different 11q23 rearrangements will respond differentially to therapeutic targeting of c-Myc expression dependent on fusion partner.