BRD3/4 inhibition and FLT3-ligand deprivation target pathways that are essential for the survival of human MLL-AF9+ leukemic cells

In the present work we aimed to identify targetable signaling networks in human MLL-AF9 leukemias. We show that MLL-AF9 cells critically depend on FLT3-ligand induced pathways as well as on BRD3/4 for their survival. We evaluated the in vitro and in vivo efficacy of the BRD3/4 inhibitor I-BET151 in various human MLL-AF9 (primary) models and patient samples and analyzed the transcriptome changes following treatment. To further understand the mode of action of BRD3/4 inhibition, we performed ChIP-seq experiments on the MLL-AF9 complex in THP1 cells and compared it to RNA-seq data of I-BET151 treated cells. While we could confirm a consistent and specific downregulation of key-oncogenic drivers such as MYC and BCL2, we found that the majority of I-BET151-responsive genes were not direct MLL-AF9 targets. In fact, MLL-AF9 specific targets such as the HOXA cluster, MEIS1 and other cell cycle regulators such as CDK6 were not affected by I-BET151 treatment. Furthermore, we also highlight how MLL-AF9 transformed cells are dependent on the function of non-mutated hematopoietic transcription factors and tyrosine kinases such as the FLT3-TAK1/NF-kB pathway, again impacting on BCL2 but not on the HOXA cluster. We conclude that BRD3/4 and the FLT3-TAK1/NF-kB pathways collectively control a set of targets that are critically important for the survival of human MLL-AF9 cells.


Introduction
In the last two decades, our understanding of the molecular mechanism underlying human malignancies has greatly improved [1]. Progress in DNA-sequencing technologies has reinforced the notion that cancer is initiated and maintained by alterations in the genome and it has also become more evident that epigenetic regulators are among the most frequent aberrancies in hematopoietic malignancies [2]. Furthermore, changes in the chromatin state can also PLOS

Primary cell isolations
Neonatal CB samples were obtained after informed consent from healthy full-term pregnancies from the obstetrics departments of the University Medical Center in Groningen (UMCG) and Martini Hospital Groningen. BM mononuclear cells from untreated patients were studied after informed consent and protocol approval by the Medical Ethical Committee of the UMCG, in accordance with the Declaration of Helsinki. All studies were approved by the Intitutional Review Board. After ficoll separation of mononuclear cells, CD34 + cells were enriched using a magnetically activated cell-sorting CD34 progenitor kit or automatically by using auto Macs (Miltenyi Biotech) as described previously [20,21] and cryopreserved until further use.

Lentiviral transductions
CB CD34 + cells were pre-stimulated and transduced as described previously [20][21][22]. One round of transduction was performed and cells were harvested at day 2 after transduction. For MLL-AF9 transformation of CB CD34 + cells UMG LV6 MLL-AF9 lentiviral vectors were used [23]. THP1 cells were transduced with a lentiviral vector expressing human BCL2 in which the EGFP cDNA was exchanged for mBlueberry2 cDNA which was a kind gift of Robert Campbell, University of Alberta, Canada [24].

Cell culture
Leukemic cell lines MOLM13, THP1 and K562 (all obtained from the American Type Culture Collection, ATCC) were cultured in RPMI 1640 (Lonza, Leusden, The Netherlands) supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin. CB transduced MLL-AF9 cells in the liquid and MS5 myeloid co-culture experiments were grown in Gartner's medium consisting of αMEM (Fisher Scientific Europe, Emergo, The Netherlands) supplemented with 12.5% heat-inactivated fetal calf serum (Lonza, Leusden, The Netherlands), 12.5% heat-inactivated horse serum (Invitrogen, Breda, The Netherlands), 1% penicillin and streptomycin, 2mM glutamine (all from PAA Laboratories), 57.2 μM β-mercaptoethanol (Merck Sharp & Dohme BV, Haarlem, The Netherlands) and 1 mM hydrocortisone (Sigma-Aldrich Chemie B.V., Zwijndrecht, The Netherlands). CB MLL-AF9 myeloid restricted cultures were supplemented with 20 ng/ml IL-3, SCF and FLT3-L (R&D Systems). Lymphoid permissive co-cultures contained the same components as the myeloid cultures with the exception of hydrocortisone and horse serum but with the presence of 50 μg/ml ascorbic acid (Sigma). In lymphoid CB MLL-AF9 co-cultures IL-3 was replaced with 10 ng/ml IL-7 (R&D Systems). AML cells in myeloid restricted cultures were supplemented with 20 ng/ml IL-3, SCF, FLT3-L, granulocyte colony-stimulating factor (Rhone-Poulenc Rorer, Amstelveen, The Netherlands), and thrombopoietin (Kirin, Tokyo, Japan) as previously described [25][26][27]. All the cultures were kept at 37˚C and 5% CO2. During demi-depopulation of cultures, it was made sure that stromal cells remained attached and that only hematopoietic/leukemic cells were included in the analyses (as also validated by flow cytometry for EGFP in case of transduced cells or by staining with huCD45 when appropriate). For all experiments (except for data shown in Fig 4D) week 6-10 immortalized CB MLL-AF9 cultures were used, or where indicated cell lines or primary patient samples, and the molecular consequence upon inhibition of BRD3/4, NF-kB or TAK1, or upon deprivation of FLT3-L were analysed within 1-6 days as indicated in the respective figures. I-BET151 (GSK GSK1210151A) was kindly provided by Nicholas Smithers (GSK R&D, UK), the TAK1 inhibitor 5z-7-oxozeanol was obtained from Sigma and the IKK VII inhibitor was obtained from Abcam (ab216471, Abcam).

FACS analysis and cell sorting
All FACS analyses were performed on a FACS Calibur (Becton Dickinson) or LSR-II (Becton Dickinson) and data were analysed using Flow Jo (Tree Star, Inc.). Cell sorting was performed on a MoFlo-Astrios (Beckman Coulter). All antibodies were obtained from IQ products. The following clones were used: Annexin V R-PE (IQP-120R), Annexin V APC (IQP-120A) and Annexin V FITC (IQP-120F). Cells were washed in calcium buffer (10 mM HEPES/NaOH, pH 7.4, 140 mM NaCl) and incubated with antibodies at 4˚C for 30 min.

In vivo treatment study
The ectopic bone model was established as described previously [28,29]. Briefly, four hybrid scaffolds consisting of three 2-3 mm biphasic calcium phosphate particles loaded with human MSCs were implanted subcutaneously into 6 to 8 weeks old female NOD.Cγ-Prkdcscid Il2rγtm1Wjl/SzJ (NSG) mice. Six to eight weeks after scaffold implantation, 0.5x10 6  were used as gene annotation database. Quantified reads were normalization using the DESeq package. Reads with failed vendor QC, reads with average quality less than 24, reads with mapping quality below 50 and reads with length less than 20 were all filtered out.

I-BET151 inhibits MLL-fusion leukemic cell lines, cord blood models and patient derived cells in vitro and in vivo
First, we verified the therapeutic potential and specificity of the BRD3/4 inhibitor I-BET151 (GSK1210151A) for leukemic cell lines harbouring MLL-fusion oncogenes. Annexin V levels were measured after 72 hours incubation with dimethyl sulfoxide (DMSO) or different dosages of I-BET151. Significantly higher and dose-dependent levels of Annexin V positive cells were detected for MOLM13 and THP1 treated cell lines, while the K562 cell line (harbouring the BCR-ABL translocation) was not affected by the treatment (Fig 1A and 1B). Next, immortalized CB-transduced MLL-AF9 cells were grown for two weeks under myeloid MS5 co-culture conditions in the absence or presence of I-BET151 and reduced expansion was observed in a dose-dependent manner, while normal CB CD34 + cells were not sensitive ( Fig 1C and S1A  Fig). When CB transduced MLL-AF9 cells were transformed along the B-ALL lymphoid lineage [22] cells were also sensitive to I-BET151 ( Fig 1C and S1B Fig). In addition to CB MLL-AF9 transduced models, I-BET151 also strongly impaired proliferation of primary MLL-AF9 patient cells ( Fig 1C). Intriguingly, when we evaluated the response to I-BET151 under liquid culture conditions, we observed a much faster and stronger response to I-BET when compared to MS5 cocultures ( Fig 1D). In line with these observations, Annexin V levels were already increased within 72 hrs of I-BET151 treatment in CB MLL-AF9 cells grown under liquid culture conditions but not when grown under MS5 stromal coculture conditions ( Fig 1E). These data suggest that the presence of a bone marrow microenvironment provides protective signals reducing I-BET151 sensitivity. Lastly, in order to evaluate the therapeutic potential of I-BET151 in vivo, we took advantage of a humanized bone marrow xenograft model (huBM-sc) of secondary MLL-AF9-driven B-ALL that we have previously published [29]. 5 × 10 5 cells harvested from a primary MLL-AF9-driven B-ALL generated in huBM-sc mice were injected into three huBM scaffolds of secondary recipients. Five weeks after injection of the cells, we performed daily intra-scaffold (i.sc.) injections of the inhibitor for a total of 9 days. One control group (n = 3) was injected with a solution of phosphate-buffered saline and 1% DMSO, one group (n = 3) with I-BET151 100 μM and another group (n = 3) with I-BET151 500 μM. The tumor volume of each scaffold was measured every 2 days, allowing us to monitor the progression of the disease in time. All mice were sacrificed at the same time 44 days after injection. Mice treated with 500 μm I-BET151 displayed a significant reduction in tumor volume (p = 0.04) (Fig 1F and 1G). No significant differences were found in spleen weight or blood and BM chimerism. The effects on overall survival could not be assessed since mice needed to be sacrificed due to tumor size in order to adhere to the predefined human endpoints of the experiment. Since mice needed to be sacrificed Together, these data indicate that I-BET151 treatment significantly delays lymphoid leukemia progression in vivo but is not sufficient for a complete eradication of the leukemic cells, at least not in this experimental setting.

Downmodulation of BCL2 and C-MYC but not HOXA genes upon I-BET151 treatment
In order to gain more insight in the transcriptional pathways regulated by I-BET151, genomewide transcriptome analysis was performed in myeloid immortalized CB MLL-AF9 cells after treatment with DMSO or I-BET151 in an MS5 co-culture setting. Since cells cultured on the MS5 stromal feeder displayed a slower response to BRD3/4 inhibition, RNA was isolated after 6 days, following two rounds of treatment. Cells were treated with 1 μM, 2 μM, 4 μM or with DMSO controls. Thus we identified 306 genes that were commonly downregulated at least 2 fold compared to DMSO treated cells and that responded to the inhibitor in a dose-dependent manner (Fig 2A). In accordance to previously published observations, we found a consistent downmodulation of transcription factors such cMYC and anti-apoptotic genes like BAX and BCL2 ( Fig 2B). Surprisingly, well-known MLL-AF9 targets such as the HOXA gene cluster and MEIS1 were not significantly affected by the treatment, as we also confirmed by independent Q-RT-PCR experiments (Fig 2B and 2C).
BCL2 is a crucial anti-apoptotic gene and its role in MLL-AF9 driven leukemia has already been investigated [32]. To assess whether we could rescue the apoptotic phenotype upon I-BET151 treatment, we overexpressed BCL2 in THP1 cells. After 72 hours incubation with different I-BET151 dosages, a significant reduction of Annexin V cells was observed in the THP1 cells overexpressing BCL2 (Fig 2D).
Since the direct MLL-AF9 target genes HOXA9 and MEIS1 were not downregulated upon I-BET151 treatment we investigated whether BRD3/4 inhibition would directly affect the MLL-AF9 oncoprotein complex and the activity of the methyl transferase DOT1L by performing ChIP-qPCRs for H3K79me2. While we observed a slight reduction in H3K79 dimethylation  on all the tested loci upon treatment with 2 μM I-BET151, the loss of H3K79me2 marks was not very strong (Fig 2E), suggesting that indeed MLL-AF9 itself is not particularly affected by I-BET151. In order to confirm that I-BET151 was targeting BRD3/4 we performed ChIP-PCRs for BRD4 and H3K27-ac on the BCL2 locus in MOLM13 and CB MLL-AF9 cells and as shown in Fig 2F a strong reduction in both BRD4 binding as well as acetylation of H3K27 was observed, in line with the strong reduction in BCL2 expression.

BRD3/4 inhibition downregulates MLL-AF9-dependent and independent target genes
While BRD3/4 inhibition significantly impaired proliferation and survival of MLL-AF9 driven leukemia in both in vitro and in vivo studies, the involved targets did not seem to completely overlap with the MLL-AF9-induced oncogenic program. In order to gain more insights in the mechanism of action of the I-BET151 inhibitor, we compared MLL-AF9-occupied loci in THP1 cells using ChIPseq data [31] with RNA-seq data upon 24h of I-BET151 treatment. A total of 766 MLL-AF9 occupied loci were identified using an association rule of ± 10 kb from the transcription start site (Fig 3 and S1 Table) [31]. As shown in Fig 3A, the overlap of these loci with the genes down regulated in THP1 cells upon 24h I-BET151 treatment was very limited. While we identified 911 downregulated genes in THP1 upon 2 μM I-BET151 treatment in a dose-dependent manner, only 66 of these were also bound by MLL-AF9 indicating that I-BET151 targets only few MLL-AF9 direct target genes. Gene ontology (GO) annotation for the non-overlapping downmodulated genes upon I-BET151 treatment on THP1 were significantly enriched for 'Transcription' and 'DNA replication' involving genes such as MYB, FOXM1 and various members of the E2F family of transcription factors. As expected, instead, the MLL-AF9 specific and non-overlapping genes were enriched for GOs such 'chromatin regulator' and including genes such the HOXA cluster, MEIS1, BM1, CBX8 and other known MLL-AF9 specific targets. Examples of MLL-AF9 specific and non-specific binding sites are displayed in Fig 3B, together with H3K79me2, H3K4me3, H3K27Ac and H3K27me3 chromatin modifications tracks.
Interestingly, we did observe MLL-AF9 binding and H3K79-me2 deposition on BCL2 and MYC, genes that were also downregulated upon I-BET151 treatment. Clearly, MLL-AF9 binding might contribute to expression of these loci, but the sensitivity to I-BET151 is particularly mediated via BRD3/4 independent of MLL-AF9.
Interestingly, among the upregulated genes upon I-BET151 treatment we have found genes that negatively control cell cycle such as CDKN1A and CDKN1B and again a rather limited overlap with MLL-AF9 occupied loci was observed (S2A Fig and S1 Table). The upregulated genes enriched for GO terms such as R-SMAD binding, G2/M transition of mitotic cell cycle and negative regulation of cell proliferation (S2A Fig) and it is quite possible that these upregulated cell cycle inhibitors further contribute to the negative phenotype imposed on MLL-AF9 cells by BRD3/4 inhibition.

MLL-AF9 cells critically depend on FLT3
In our previous studies aimed at better understanding molecular mechanisms sustaining MLL-AF9-induced transformation we identified FLT3 as a strongly upregulated MLL-AF9 H3K79-me2 antibodies. The distance from the transcription start site (in kb) is indicated. Bar graphs are represented as the mean enrichment relative to input and error bars reflect standard deviation of results derived from technical triplicate experiments. F) ChIP-PCR experiments as in E, but now antibodies against BRD4 or H3K27-Ac were used.
https://doi.org/10.1371/journal.pone.0189102.g002 target [22]. Here, we further confirmed this upregulation at the RNA and protein level (Fig 4A  and 4B). We also observed direct MLL-AF9 binding and H3K79-me2, H3K4-me3 and H3K27-Ac deposition at the FLT3 locus (Fig 4C). To functionally study the role of the FLT3 receptor in CB MLL-AF9 cells during the initiation phase of transformation, freshly isolated CD34 + CB cells were transduced with the MLL-AF9 fusion gene and cultured in the presence or absence of the FLT3 ligand (FLT3-L) on MS5 stroma under myeloid-restricted conditions. After two to three weeks of FLT3-L-deprivation a reduction in proliferation of CB MLLAF9 cells was observed, which became progressively more evident after the third week of culture ( Fig 4D). These data indicate that FLT3-L is required for the initiation of transformation induced by MLL-AF9, which usually starts to occur around week 3 [22]. After 4 weeks in culture, no differences in Annexin V levels where detected between cells supplied with or without the FLT3-L, indicating that cells are not entering apoptotic program (data not shown).
We also questioned whether FLT3-L was required for the maintenance of MLL-AF9 transformed cells. To evaluate this, we used week 10 or week 18 immortalized CB MLL-AF9 cultures, and as shown in Fig 4E, removal of FLT3-L also had detrimental effects on immortalized cultures.
Next, we sought to analyse the changes in gene expression upon deprivation of FLT3-L under leukemia maintenance conditions. RNA from immortalized CB MLL-AF9 cells grown in MS5 co-culture or liquid culture conditions were isolated after 48 hours. Interestingly, down-regulation of genes was stronger in liquid culture conditions compared to the MS5 cocultures, possibly due to the fact that stromal cells are producing low levels of FLT3-L. 113 genes were commonly downregulated after 48 hours in both liquid and co-culture settings which were significantly enriched for GO terms such as Cell Cycle, Cell Division and G1/S transition of mitotic cell cycle (Fig 4F and S1 Table). Indeed, genes involved in cell cycle regulation such CND2, CHEK1, CDC25C and CDC25A were found to be downregulated upon FLT3-L deprivation, and a marked reduction of~25% in cells in S/G2/M phase was found in cells deprived of the FLT3-L after 24 hours (Fig 4H).
As observed for the I-BET151 treated cells, a component of the PcG complex SCMH1 and the anti-apoptotic BCL2 were also found to be consistently downregulated (Fig 4I), suggesting that FLT3-L and BRD3/4 pathways converge at a set of target genes including BCL2. Moreover, and once again as observed for the I-BET151 treated cells, none of the HOXA cluster of genes was affected by the ligand deprivation as confirmed also by qPCR for HOXA9 (Fig 4I). Taken together, these data reinforce the notion that although certain expression programs are directly regulated by the MLL-AF9 oncoprotein complex, leukemic transformation still critically depends on additional pathways.

Inhibition of the TAK1-NFkB axis phenocopies FLT3-L deprivation
Next, we aimed to further identify the pathways downstream of the FLT3 receptor that are critically important for the survival and proliferation of MLL-AF9 cells. Recently, it was shown that MLL oncoproteins, such as MLL-AF10, are dependent on IKK/NF-kB activity for their survival [33]. Furthermore, there is experimental evidence showing that FLT3 activation either by its ligand or by oncogenic mutations results in activation of NF-kB to support cell survival [33][34][35][36][37]. We analysed whether inhibition of IKK/NF-kB or one of its upstream activators TAK1 would phenocopy the effects observed upon FLT3L deprivation in MLL-transformed cells. Treatment with either the TAK1 inhibitor AZ-TAK1 or the IKK VII inhibitor for 24 hrs resulted in strongly enhanced apoptosis of immortalized CB MLL-AF9 transformed cells ( Fig  5A). Cell proliferation was also severely impaired on MS5 stromal cocultures by AZ-TAK1 (Fig 5B) or the IKK VII inhibitor (data not shown). Importantly, also a concentration- dependent reduction in cobblestone formation and maintenance was seen upon treatment with AZ-TAK1 (Fig 5C) implying that the primitive transformed leukemic cells were also targeted by the TAK1 inhibitor. To study whether the TAK1/NF-kB axis would act independently of MLL-AF9 we determined the expression of HOXA9 and MEIS, and we observed that the expression of both transcripts was not affected upon TAK1 or IKK VII inhibition (Fig 5D and  5E). In contrast, these inhibitors resulted in a marked reduction of BCL2 expression (Fig 5E  and data not shown) indicating that, similarly to what observed upon FLT3-L deprivation, the anti-proliferative effect observed upon NF-kB inhibition mostly relies on a limited number of genes directly regulated by MLL-AF9.

Discussion
Over the last years, a large body of reports has highlighted the important role of epigenetic regulators in the pathogenesis of various types of cancer, including leukemias. One example of the latter is represented by MLL-fusion leukemias that, although characterized by an adverse prognosis, possess a relatively simple genetic background where the MLL-fusion gene is in most cases the only driver mutation [38]. Therefore, MLL-rearranged leukemias have been studied extensively to evaluate the influence of epigenetic regulators on the oncogenic transcription programs.
Wild-type MLL proteins contain a SET domain that harbors H3K4 methyltransferase activity allowing transcription initiation by Polymerase II [17]. Within MLL fusion proteins this SET domain is lost together with its catalytic activity, but MLL fusion proteins gain the ability to methylate H3K79 via recruitment of the H3K79 methyltransferase DOT1L [18,39]. As a consequence, the epigenetic landscape around promoters is changed resulting in transcriptome alterations with important consequences for leukemic transformation [18,39]. It is therefore no surprise that DOT1L inhibitors are currently extensively evaluated for their efficacy in MLL-rearranged leukemias [40][41][42][43].
In the present work, we aimed to identify targetable signaling networks in human MLL-AF9 leukemias. We show that MLL-AF9 cells critically depend on FLT3 ligand-induced pathways for their survival. Both the initiation of transformation as well as the maintenance of transformed leukemic cells critically requires the presence of FLT3-ligand. We also find that the FLT3-receptor is strongly upregulated by MLL-AF9, coinciding with a direct binding of MLL-AF9 to the FLT3 locus, thereby facilitating FLT3 signaling. Murine MLL-ENL-transformed cells have been shown to depend on FLT3-ligand, while MLL-AF9-induced transformation of murine cells did not depend on FLT3-ligand [44], in contrast to our findings. Possibly, differences in the role of FLT3 signaling in the murine and human hematopoietic compartments might play a role. The FLT3 receptor is for instance expressed in the most primitive human long-term HSCs, but not in murine LT-HSCs [45][46][47], and it is possible the FLT3 signaling is less important in the GMP-driven MLL-AF9 leukemia models in the mouse compared to more HSC-driven MLL-AF9 leukemia in human CB model systems [22]. In line with our findings, FLT3-L dependency has also been shown in human CB MLL-AF9 models by the Mulloy lab [48].
Downstream, we find that CB MLL-AF9 cells heavily depend on the activity of the serine/ threonine kinase TAK1 and NF-kB. Previously, we and others identified that AML cells frequently display constitutive NF-kB activity [37,49,50] and heavily depend on TAK1 for their survival [35], and here we show that FLT3-ligand stimulated MLL-AF9 cells also do not tolerate inhibition of either TAK1 of NF-kB activity. An important role for NF-kB for MLL oncoproteins was also shown earlier [33]. Transcriptome studies revealed that deprivation of FLT3 ligand resulted in loss of genes that enriched for the GO terms DNA replication and G1/S cell cycle transition, coinciding with a loss of S/G2/M and accumulation of cells in G1 in cell cycle analysis studies, and also expression of the survival gene BCL2 depended on the presence of FLT3 ligand.
These data show strong overlap with a dependency of MLL-AF9 cells on BRD3/4 activity. We evaluated the in vitro and in vivo efficacy of the BRD3/4 inhibitor I-BET151 in various human MLL-AF9 (primary) models and patient samples and in line with what was published previously [8,51] we find good efficacy upon treatment with I-BET151. MLL-AF9 can transform CB CD34 + cell along the myeloid as well as the lymphoid lineage, and we find that both are strongly dependent on the presence of BRD3/4 activity. Interestingly, many of the genes regulated by BRD3/4 were not directly bound by MLL-AF9 as determined by ChIP-seq experiments. These data suggest that BRD3/4 does not directly control MLL-AF9 chromatin recruitment but rather suggest that it controls independent, but critically important, pathways for the survival of MLL-AF9 cells. While it was initially proposed that BRD4 would directly recruit MLL-fusion proteins to the chromatin it has now become clear that MLL-fusion proteins/DOT1L and BRD4 act independently, whereby DOT1L results in methylation of H3K79 followed by transcription factor-mediated recruitment of EP300 that deposits an acetyl mark on H4K5 [52]. This in turn forms a docking site for BRD4 thereby facilitating the target gene expression and co-inhibition of DOT1L and BRD4 has been shown to act synergistically in targeting MLL-rearranged leukemias [52].
Another interesting finding is that, although MLL-AF9 cells were sensitive to I-BET151, the sensitivity under liquid culture conditions was much stronger compared to the sensitivity observed under stromal co-culture conditions. Although the exact underlying mechanisms are not yet clear, it is quite possible that the bone marrow niche can provide protective signals for leukemic cells. For instance, survival pathways controlling BCL2 can also be activated by various (secreted) factors arising from stromal cells, a phenomenon that will certainly be further investigated in our future studies. Possibly, this also explains why we do not observe a complete eradication of MLL-AF9 cells in our in vivo humanized niche xenograft model in which leukemias are grown in the presence of a microenvironment composed of human mesenchymal stromal cells (Fig 1 and [32]). While we performed all the in vitro experiments on murine MS5 stroma supplemented with human cytokines in order to have a reproducible standardized system, it would certainly be of interest to evaluate drug dependencies in vitro in the presence of primary human MSCs as well.
In conclusion, a concept emerging from all our experiments is that although MLL-AF9 drives an aberrant gene expression program regulating for instance the HOXA cluster, it still critically relies on non-mutated transcription factors or tyrosine kinases. We show that BRD3/ 4 and the FLT3-TAK1/NF-kB pathways collectively control a set of targets that are critically important for the survival of human MLL-AF9 cells.
Supporting information S1 Table. Transcriptome  technical assistance with cell sorting; Jenny Jaques, and Mariana Pliego for technical assistance with the experiments and Vincent van den Boom for help with ChIP experiments. This work was supported by the EU (FP7-PEOPLE-2010-ITN-264361 EuroCancer StemCell Training network). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. I-BET151 was kindly provided by Nicholas Smithers (GSK R&D, UK).