Recurrent inactivating mutations have been identified in various hematological malignancies in the X-linked BCOR gene encoding BCL6 corepressor (BCOR); however, its tumor suppressor function remains largely uncharacterized. We generated mice missing Bcor exon 4, expressing a variant BCOR lacking the BCL6-binding domain. Although the deletion of exon 4 in male mice (BcorΔE4/y) compromised the repopulating capacity of hematopoietic stem cells, BcorΔE4/y thymocytes had augmented proliferative capacity in culture and showed a strong propensity to induce acute T-cell lymphoblastic leukemia (T-ALL), mostly in a Notch-dependent manner. Myc, one of the critical NOTCH1 targets in T-ALL, was highly up-regulated in BcorΔE4/y T-ALL cells. Chromatin immunoprecipitation/DNA sequencing analysis revealed that BCOR was recruited to the Myc promoter and restrained its activation in thymocytes. BCOR also targeted other NOTCH1 targets and potentially antagonized their transcriptional activation. Bcl6-deficient thymocytes behaved in a manner similar to BcorΔE4/y thymocytes. Our results provide the first evidence of a tumor suppressor role for BCOR in the pathogenesis of T lymphocyte malignancies.
BCOR was originally identified as a corepressor of BCL6, a key transcriptional factor required for development of germinal center B cells (Huynh et al., 2000; Klein and Dalla-Favera, 2008). BCOR is located on chromosome X, and mutations in BCOR were initially identified in patients with X-linked inherited diseases Lenz microphthalmia and oculo-facio-cardio-dental (OFCD) syndrome (Ng et al., 2004). The mutations include stop codon gains and frame-shift insertions or deletions, indicating that they cause the loss of BCOR function. Mesenchymal stem cells isolated from a patient with OFCD exhibited increased osteo-dentinogenic potential in culture (Fan et al., 2009). However, the lack of OFCD phenotypes in Bcl6-deficient mice (Yoshida et al., 1999) suggested the involvement of non-BCL6 pathways in the developmental consequences of Bcor mutations. Recent extensive analyses of the BCOR complex revealed that BCOR also copurifies with RING1B, PCGF1, and KDM2B and functions as a component of the noncanonical polycomb repressive complex 1 (PRC1), PRC1.1, which monoubiquitinates histone H2A (Gearhart et al., 2006; Sánchez et al., 2007; Gao et al., 2012).
Recent whole-exome sequencing has identified somatic BCOR mutations in various hematological diseases. BCOR mutations have been reported in acute myeloid leukemia (AML) with normal karyotype (3.8%), secondary AML (3.5%), myelodysplastic syndrome (4.2%), chronic myelomonocytic leukemia (7.4%), and extranodal NK/T cell lymphoma (21–32%; Grossmann et al., 2011; Damm et al., 2013; Lee et al., 2015; Lindsley et al., 2015; Dobashi et al., 2016). Most of the BCOR mutations result in stop codon gains, frame-shift insertions or deletions, splicing errors, and gene loss, leading to the loss of BCOR function (Damm et al., 2013). BCOR mutations also result in reduced mRNA levels, possibly because of activation of the nonsense-mediated mRNA decay pathway (Damm et al., 2013). The closely related homolog BCOR-like 1 (BCORL1) has been implicated in AML and myelodysplastic syndrome in a manner similar to BCOR (Li et al., 2011; Damm et al., 2013). Somatic mutations in BCOR or BCORL1 have also been identified in 9.3% of patients with aplastic anemia and correlated with a better response to immunosuppressive therapy and longer and higher rates of overall and progression-free survival (Yoshizato et al., 2015). Furthermore, BCOR mutations have been found in retinoblastoma, bone sarcoma, and clear cell sarcoma of the kidney (Pierron et al., 2012; Zhang et al., 2012a; Kelsey, 2015).
BCOR has been shown to restrict myeloid proliferation and differentiation in culture using conditional loss-of-function alleles of Bcor in which exons 9 and 10 are missing. This mutant Bcor allele generates a truncated protein that lacks the region required for the interaction with PCGF1, a core component of PRC1.1, and mimics some of the pathogenic mutations observed in patients with OFCD and hematological malignancies (Cao et al., 2016). The tumor suppressor function of Bcor has recently been confirmed in vivo using Myc-driven lymphomagenesis in mice (Lefebure et al., 2017). However, limited information is available on its role in hematopoiesis and hematological malignancies. In the present study, we investigated the function of BCOR using mice expressing variant BCOR, which cannot bind to BCL6, and revealed a critical role for BCOR in restricting transformation of hematopoietic cells.
Results and discussion
Generation of mice expressing BCOR that cannot bind to BCL6
To understand the physiological role of BCOR as a BCL6 corepressor, we generated mice harboring a Bcorfl mutation in which exon 4 encoding the BCL6-binding site (Ghetu et al., 2008) was floxed (Fig. 1 a), and then crossed Bcorfl mice with Rosa26::Cre-ERT (Cre-ERT) mice. We transplanted BM cells from Cre-ERT control (WT) and Bcorfl/y;Cre-ERT CD45.2 male mice (Bcor is located on the X chromosome) without competitor cells into lethally irradiated CD45.1 recipient mice and deleted Bcor exon 4 by intraperitoneal injections of tamoxifen at 4 wk posttransplantation. We hereafter refer to the recipient mice reconstituted with WT and BcorΔE4/y cells as WT and BcorΔE4/y mice, respectively. We confirmed the efficient deletion of Bcor exon 4 in hematopoietic cells from BcorΔE4/y mice by genomic PCR (Fig. 1 b). RNA-sequence analysis of lineage-marker (Lin)− Sca-1+ c-Kit+ (LSK) hematopoietic stem and progenitor cells (HSPCs) revealed the specific deletion of Bcor exon 4 (Fig. 1 c). Bcor lacking exon 4 generates a short form of BCOR protein (BCORΔE4) that lacks the BCL6 binding site but still retains the binding site for PCGF1, a component of PRC1.1 (Fig. 1 d). Western blot analysis detected a short form of BCOR in thymocytes from BcorΔE4/y mice (Fig. 1 e). To test physical interactions between BCL6 and BCORΔE4, we cotransfected 293T cells with plasmids encoding HA-tagged BCL6 and Flag-tagged BCOR and performed immunoprecipitation. Full-length BCOR readily coimmunoprecipitated with BCL6, but BCORΔE4 scarcely did. In contrast, full-length BCOR and BCORΔE4 both retained binding to PCGF1 and RING1B, components of PRC1.1, suggesting that the deletion of Bcor exon 4 does not compromise the function of PRC1.1 (Fig. 1 f). Interestingly, the stabilization of BCL6, which may be induced by interaction with exogenous BCOR, was observed in cells transfected with full-length Bcor but not BcorΔE4 (Fig. 1 f). Western blot analysis revealed that BcorΔE4/y BM c-Kit+ progenitors and CD4+CD8+ double-positive (DP) thymocytes had polycomb histone modifications (H2AK119ub1 and H3K27me3) at levels similar to WT (Fig. 1 g).
Bcor exon 4 impairs the repopulating capacity of hematopoietic stem cells
We first examined hematopoiesis in BcorΔE4/y mice. These mice exhibited leukopenia that was mainly attributed to impaired B lymphopoiesis (Fig. 2, a and b). 6 mo after the deletion of Bcor exon 4, BcorΔE4/y mice showed reductions in the numbers of total BM, CD34−LSK hematopoietic stem cells (HSCs), CD34+LSK multipotent progenitors, spleen cells, thymocytes, and DP thymocytes (Fig. 2, c–e). No significant difference was observed in the frequency of apoptotic cells between control and BcorΔE4/y LSK cells (Fig. 2 f). Moreover, the proportion of HSPCs in the G0 stage of the cell cycle did not change (Fig. 2 f). To avoid the proliferative stress caused by BM transplantation, we also deleted Bcor exon 4 in primary Bcorfl/y mice. The mice showed hematopoietic phenotypes similar to those observed in BM transplant mice with the hematopoietic cell-specific deletion (not depicted). We then performed competitive repopulation assays using the same number of test cells and BM competitor cells and found that BcorΔE4/y BM cells were gradually outcompeted by WT BM cells in both peripheral blood (PB) and BM (Fig. 2 g), indicating the impaired repopulating capacity of BcorΔE4/y HSCs. We next purified CD34−LSK HSCs and LSK HSPCs from the BM of WT and BcorΔE4/y mice and evaluated their proliferative capacity in vitro. Although BcorΔE4/y HSCs showed significantly impaired cell growth under culture conditions that supported stem cell proliferation rather than differentiation (stem cell factor [SCF] and thrombopoietin [TPO]), BcorΔE4/y LSK HSPCs demonstrated moderately higher proliferation rates than WT under myeloid culture conditions (SCF+TPO+IL-3+GM-CSF; Fig. 2 h), as reported previously for mice lacking Bcor exons 9 and 10 (Cao et al., 2016). However, a serial methylcellulose replating colony assay did not show any enhancement in the serial replating capacity of BcorΔE4/y LSK cells (not depicted). These results indicate that the deletion of Bcor exon 4 compromises the repopulation capacity of HSCs both in vitro and in vivo, but not the proliferation and differentiation of hematopoietic progenitor cells.
Bcor exon 4 induces acute T-cell lymphoblastic leukemia
During an observation period of 300 d after the deletion of Bcor exon 4, 50% of BcorΔE4/y mice developed lethal acute T-cell lymphoblastic leukemia (T-ALL), whereas only one mouse developed B-ALL/lymphoma (Fig. 3 a). Moribund mice with T-ALL showed expansion of lymphoblasts, which were mostly DP or CD8 single-positive (SP; Fig. 3, a–c and e). T-ALL cells were also detected in the PB, BM, thymus, and spleen to varying degrees (Fig. 3, b–e). All the T-ALL mice showed high chimerism of donor-derived cells (PB, 89.0 ± 4.6%; BM, 92.3 ± 4.3%; thymus, 98.8 ± 0.5%; and spleen, 97.4 ± 0.8%; n = 8–9). Most T-ALL involved the thymus, but 20% had no or minimal involvement of the thymus, suggesting BM origin of the disease. It was previously reported that loss of p53 induces the transformation of thymocytes, leading to the development of thymic T cell lymphoma characterized by the expansion of DP or CD8 SP lymphoblasts in mice (Donehower et al., 1992; Jacks et al., 1994), and promotes transformation of cells with insufficient tumor suppressor functions (Jonkers et al., 2001; Celeste et al., 2003). To test whether the loss of p53 accelerates leukemogenesis in BcorΔE4/y mice, we generated BcorΔE4/yp53−/− mutants. Although all p53−/− mice developed thymic T cell lymphoma, BcorΔE4/yp53−/− mice developed T-ALL (60%) in addition to thymic T cell lymphoma (40%; Fig. 3 f). BcorΔE4/yp53−/− T-ALL showed phenotypes that were similar to those of BcorΔE4/y T-ALL (not depicted), albeit with markedly earlier onset (Fig. 3, a and f). Although BcorΔE4 did not accelerate the development of thymic T cell lymphoma induced by the loss of p53, these results suggest that BcorΔE4 and inactive p53 alleles cooperate in the development of T-ALL in mice.
Myc is activated in BcorΔE4/y DP thymocytes and T-ALL
To elucidate the molecular mechanisms responsible for BcorΔE4/y T-ALL, we performed RNA sequence analysis of BcorΔE4/y DP leukemic blasts from the thymus of moribund mice (T-ALL no. 1) and WT DP thymocytes. Gene set enrichment analysis (GSEA) revealed the stronger enrichment of MYC target genes in BcorΔE4/y DP leukemic blasts than in WT DP thymocytes (Fig. 4 a). Myc is one of the potent oncogenes implicated in T-ALL (Belver and Ferrando, 2016). It plays an important role in the control of cell growth downstream of NOTCH1 and pre-TCR signaling during early T cell development. NOTCH1 directly binds to a long-range distal Myc enhancer and activates Myc, which is crucial for T cell development and also the initiation and maintenance of NOTCH-dependent T-ALL (Yashiro-Ohtani et al., 2014; Herranz et al., 2015). Although Myc is completely transcriptionally repressed in DP thymocytes that lose proliferative capacity, it was strongly up-regulated in BcorΔE4/y DP leukemic blasts (T-ALL no. 1; Fig. 4 b) and DP and CD8 SP blasts from T-ALL nos. 2 and 3, respectively (not depicted).
We then examined whether the active NOTCH1 pathway is involved in this process. Although NOTCH1 target gene sets were not significantly enriched in BcorΔE4/y T-ALL in GSEA (not depicted), many of the representative direct targets of NOTCH1 were up-regulated in BcorΔE4/y T-ALL in addition to Myc, such as Hes1, Ptcra, Ccnd1, Skp2, and Il7r (Fig. 4 c), suggesting that NOTCH1 signaling is augmented in BcorΔE4/y T-ALL. Sanger sequencing of exons 26–28 and 34 of Notch1, in which the majority of activating NOTCH1 mutations are found in human T-ALL, revealed a mutation in exon 34 in one of five T-ALLs (T-ALL no. 1; Fig. 4 d). This mutation caused frameshift, which resulted in truncation of the PEST domain in the C-terminal region of the protein. We also found Notch1 deletions in two of five T-ALLs (T-ALL nos. 3 and 4), which remove exon 1 and the proximal promoter (Fig. S1 a), leading to ligand-independent NOTCH1 activation (type 1 deletions; Ashworth et al., 2010). None of the five T-ALLs had activating Notch1 transcripts, in which exon 1 is spliced out of frame to 3′ Notch1 exons (type 2 deletions; Fig. S1 b; Ashworth et al., 2010). Western blotting for activated NOTCH1 revealed accumulation of cleaved NOTCH1 in CD8 SP and CD4 SP T-ALL cells from the spleens of T-ALL nos. 4 and 5, respectively (Fig. 4 e). Correspondingly, immunohistochemical analysis of the spleen of T-ALL no. 4 showed nuclear accumulation of cleaved NOTCH1 (Fig. S2). Finally, we treated leukemic cells (T-ALL no. 4) with N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester (DAPT), a γ-secretase inhibitor, and confirmed marked inhibition of cell growth in culture (Fig. 4 f). All these findings, which are summarized in Fig. S1 b, suggest that BCORΔE4 collaborates with active NOTCH1 to induce T-ALL.
During the differentiation of T lymphocytes in the thymus, Bcl6 is expressed in a manner that is reciprocal to Myc. Bcl6 is turned on in DP thymocytes in which Myc is shut off. In contrast, Bcor is expressed throughout differentiation (Fig. 4 g). We generated Bcl6Δ/Δ mice reconstituted with Bcl6fl/fl;Cre-ERT BM cells followed by treatment with tamoxifen. The Bcl6Δ mutant allele lacks exons 7–9, which encode the C-terminal zinc finger domains ZF1 to ZF5. The resultant protein cannot bind to DNA (Kaji et al., 2012). RNA sequence analyses of DP thymocytes from BcorΔE4/y and Bcl6Δ/Δ mice demonstrated significantly overlapping genes derepressed from those in WT, including Myc (Fig. 4 h and Table S1 a). Myc was moderately up-regulated in both BcorΔE4/y and Bcl6Δ/Δ DP thymocytes, albeit at lower levels than those in BcorΔE4/y T-ALL DP cells (Fig. 4 i). GSEA revealed that the MYC target gene signature was positively enriched in BcorΔE4/y and Bcl6Δ/Δ DP thymocytes (normalized enrichment score [NES] 1.66, nominal p-value [NOM] 0, and false discovery rate [FDR] 0.3; NES 2.06, NOM 0, and FDR 0, respectively).
We next purified CD4CD8 double-negative 1 (DN1) and DN2 thymocytes from BcorΔE4/y and Bcl6Δ/Δ thymocytes and cultured them on TSt-4/DLL stromal cells expressing the NOTCH ligand, delta-like 1 (DLL1). BcorΔE4/y and Bcl6Δ/Δ DN1/2 thymocytes showed higher proliferative rates under conditions that support thymocyte proliferation (Fig. 4 j). Furthermore, under conditions that promote thymocyte differentiation, BcorΔE4/y and Bcl6Δ/Δ thymocytes preferentially differentiated into CD8 SP thymocytes (Fig. 4 k). These differentiation profiles of BcorΔE4/y and Bcl6Δ/Δ thymocytes correlated well with that of BcorΔE4/y T-ALL cells, which preferentially showed the DP to CD8 SP phenotype.
Myc and other NOTCH1 targets in DP thymocytes
To identify BCOR targets in thymocytes, we performed a chromatin immunoprecipitation/DNA sequencing (ChIP-seq) analysis of BCOR using total thymocytes containing DP thymocytes as the major cell population. We detected 2,509 significant binding peaks of BCOR, the majority of which were located on the promoter region around the transcriptional start site (TSS; TSS ± 2.0 kb; Fig. 5 a and Table S1 b). Mapping of NOTCH1 peaks detected in T-ALL cells (Yashiro-Ohtani et al., 2014) revealed that a large number of BCOR peaks were located in close proximity to NOTCH1 peaks (±1.0 kb) at the promoter regions but not in the intergenic regions that contain enhancer elements (Fig. 5 a). One of the representative NOTCH targets, Myc, was identified as a BCOR target in DP thymocytes (Fig. 5, a and b). A manual ChIP analysis confirmed the binding of BCOR to the Myc promoter (Fig. 5 c). BCL6 has also been reported to transcriptionally repress Myc in pre-BII cells in which Bcl6 is strongly up-regulated downstream of the pre-B cell receptor signal (Nahar et al., 2011). Interestingly, BCOR appeared to reside on the Myc promoter in close proximity to NOTCH1 detected in T-ALL cells (Fig. 5 b), but BCOR did not bind to the long-range distal Myc enhancer, which binds NOTCH complexes and physically interacts with the Myc promoter in T-ALL cells (Fig. 5 a; Yashiro-Ohtani et al., 2014; Herranz et al., 2015). Further comparisons of the binding sites of BCOR in DP thymocytes and NOTCH1 in DP T-ALL cells (Yashiro-Ohtani et al., 2014) revealed that a large number of gene promoters had both BCOR and NOTCH1 peaks in close proximity (≤1.0 kb; Fig. 5, a and d), suggesting that BCOR and NOTCH regulate many genes reciprocally in the thymocytes including the representative NOTCH1 targets Myc and Hes1 (Fig. 5, b and d; and Table S1 c). These results suggest that BCOR antagonizes the transcriptional activation of T-ALL–related oncogenes by NOTCH1. However, most genes bound by BCOR at their promoters were not significantly derepressed in BcorΔE4/y DP thymocytes (Fig. 5 e and Table S1 c), indicating that the loss of BCOR binding to BCL6 has minimal impact on the transcription of BCOR targets in a physiological setting in T lymphocytes. BCL6 may function as a repressor by recruiting histone deacetylases SMRT and NCOR, even in the absence of BCOR. Furthermore, the BCORΔE4/y protein stayed at nearly half of the WT BCOR targets in BcorΔE4/y thymocytes (Fig. 5, b and f; and Table S1, c and d), suggesting that the majority of BCOR is recruited to its target gene loci independently of physical binding to BCL6, possibly as a component of PRC1.1. Because global levels of the polycomb histone marks, H2Aub1 and H3K27me3, did not change in BcorΔE4/y thymocytes (Fig. 1 g), BCOR target genes in BcorΔE4/y thymocytes may still remain under the control of repressive polycomb histone modifications.
In this study, we provide the first evidence of a tumor suppressor function for BCOR in T-lymphocytes in vivo. Loss-of-function mutations in BCOR have been identified in myeloid malignancies. Furthermore, recent studies also identified BCOR mutations in lymphoid malignancies, such as extranodal NK/T cell lymphoma (21–32%) and chronic lymphocyte leukemia (2.2%; Landau et al., 2015; Lee et al., 2015; Dobashi et al., 2016). Although BCOR mutations are detected in a significant portion (5–8%) of patients with T cell prolymphocytic leukemia, an aggressive neoplasm of mature T lymphocytes (Kiel et al., 2014; Stengel et al., 2016), they are not common in patients with pediatric T-ALL (1.2%; Seki et al., 2017). Our results indicate that the part of BCOR encoded by exon 4 mediates a tumor suppressor function in T lymphocytes. In germinal center B cells, BCOR functions as a corepressor of BCL6 in the context of a PRC1 complex. EZH2-containing PRC2 and BCL6 cooperate to assemble BCL6-BCOR-PRC1.1-PRC2 repressive complexes to promote lymphomagenesis (Béguelin et al., 2016). In contrast, our results indicate a tumor suppressor function for BCOR in T lymphocytes. Nevertheless, it is important to note that many of the components of PRC2 are inactivated by somatic gene mutations in T lymphocytes (Ntziachristos et al., 2012; Zhang et al., 2012b; Kiel et al., 2014; Stengel et al., 2016). These findings suggest that BCL6 cooperates with the BCOR-containing PRC1 in T lymphocytes in a manner similar to B lymphocytes, but targets different genes. In this regard, it would be intriguing to test whether the deletion of Bcl6 or the genes encoding other components of the BCOR-containing PRC1 also induces T-ALL in mice.
In the present study, the results of ChIP-seq analysis on thymocytes revealed that BCOR targets a significant portion of NOTCH1 targets detected in T-ALL cell lines, including Myc and Hes1. These results suggest that BCOR counteracts the oncogenic active NOTCH to restrain transformation. Therefore, the loss of BCOR may potentiate the transactivating ability of active NOTCH, thereby promoting the development of T-ALL. Indeed, BCORΔE4 appeared to collaborate with active NOTCH1 to induce T-ALL. Collectively, our results suggest that BCL6-BCOR-PRC1.1-PRC2 repressive complexes all collaborate to restrain the excessive activation of NOTCH1 target genes in thymocytes.
Materials and methods
Mice and generation of hematopoietic chimeras
The conditional Bcor allele (Bcorfl), which contains LoxP sites flanking Bcor exon 4 containing the BCL6 binding domain, was generated by homologous recombination using R1 embryonic stem cells according to a conventional protocol. Bcorfl mice were backcrossed at least six times onto a C57BL/6 (CD45.2) background. Bcl6 and p53 conditional knockout mice (Bcl6fl/fl and p53fl/fl, respectively) have been described (Jonkers et al., 2001; Kaji et al., 2012). In the conditional deletion of Bcor and Bcl6, mice were crossed with Rosa26::Cre-ERT mice (TaconicArtemis). To generate hematopoietic chimeras, we transplanted WT, Bcorfl/y;Cre-ERT, or Bcl6fl/fl;Cre-ERT BM cells into lethally irradiated CD45.1+ recipient mice and deleted Bcor or Bcl6 4 wk posttransplantation by intraperitoneally injecting 100 µl tamoxifen dissolved in corn oil at a concentration of 10 mg/ml for five consecutive days. Littermates were used as controls. C57BL/6 mice congenic for the Ly5 locus (CD45.1) were purchased from Sankyo-Lab Service. All animal experiments were performed in accordance with our institutional guidelines for the use of laboratory animals and approved by the Review Board for Animal Experiments of Chiba University (approval ID 25-104).
Flow cytometry and antibodies
Monoclonal antibodies recognizing the following antigens were used in flow cytometry and cell sorting: CD45.2 (104), CD45.1(A20), Gr-1 (RB6-8C5), CD11b/Mac-1 (M1/70), Ter-119, CD127/IL-7R (A7R34), B220 (RA3-6B2), CD4 (L3T4), CD8 (53-6.7), CD34 (RAM34), CD117/c-Kit (2B8), Sca-1 (D7), CD16/32/FcγRII-III (93), γδTCR (GL3), NK1.1 (PK136), CD11c (N418), CD25 (PC61; APC), and CD44 (IM7; PE). Monoclonal antibodies were purchased from BD, BioLegend, eBioscience, or Tonbo. In Annexin V staining, cells were suspended in 1× Annexin binding buffer (BD) and stained with FITC Annexin V (BD) according to the manufacturer’s protocol. To analyze the cell-cycle status, cells were incubated with 1 µg/ml Pyronin Y (Sigma-Aldrich) at 37°C for 45 min. Dead cells were eliminated by staining with 1 µg/ml propidium iodide (Sigma-Aldrich). All flow cytometric analyses and cell sorting were performed on FACSAria III or FACS Canto II (BD).
Growth and differentiation assays
CD34−LSK HSCs were cultured in S-Clone SF-O3 (Sanko Junyaku) supplemented with 0.1% BSA, 50 µM 2-mercaptoethanol, 1% glutamine/penicillin/streptomycin, 20 ng/ml SCF, and 20 ng/ml TPO (Peprotech). LSK HSPCs were cultured in the presence of 20 ng/ml SCF, TPO, IL-3, and GM-CSF (Peprotech). Using a biotinylated lineage mixture (Gr-1, Mac-1, Ter-119, B220, CD4, CD8α, γδTCR, CD3ε, CD11c, and NK1.1), CD25, and CD44, Lin− CD25−CD44+ DN1 and Lin−CD25+CD44+ DN2 thymocytes were sorted on FACSAria III and then cultured on TSt-4/DLL stromal cells. SCF, IL-7, and Flt3L (Peprotech) were added to cultures at concentrations of 10 ng/ml and 2 ng/ml for growth and differentiation assays, respectively. Primary T-ALL cells from leukemic mice were cultured on TSt-4 stromal cells in the presence of DAPT (Calbiochem).
Total RNA was extracted using an RNeasy Mini kit (Qiagen) and reverse-transcribed by the ThermoScript RT-PCR system (Invitrogen) with an oligo-dT primer. Real-time quantitative PCR was performed with a StepOnePlus Real-Time PCR System (Life Technologies) using FastStart Universal Probe Master (Roche Applied Science) and the indicated combinations of the Universal Probe Library (Roche Applied Science). All data are presented as relative expression levels normalized to Hprt expression. Primer sequences used were as follows: Bcor forward, 5′-ATGGCGACGCTTCAAAAG-3′, reverse 5′-GGGCTCCACTGGTATCCAC-3′, probe 19; Bcl6 forward, 5′-CCCCACAGCATACAGAGATG-3′, reverse 5′-TTGCAGAAGAAGGTCCCATT-3′, probe 3; Myc forward, 5′-CCTAGTGCTGCATGAGGAGA-3′, reverse 5′-TCCACAGACACCACATCAATTT-3′, probe 77; and Hprt forward, 5′-TCTTCCTCAGACCGCTTTT-3′, reverse 5′-CCTGGTTCATCATCGCTAATC-3′, probe 95.
Total RNA was extracted using an RNeasy Plus Micro kit (Qiagen), and cDNA was synthesized using a SMART-Seq v4 Ultra Low Input RNA kit for Sequencing (Clontech). ds-cDNA was fragmented using S220 Focused ultrasonicator (Covaris), then cDNA libraries were generated using a NEBNext Ultra DNA Library Prep kit (New England Biolabs). Sequencing was performed using HiSeq1500 (Illumina) with a single-read sequencing length of 60 bp. TopHat (version 2.0.13; default parameters) was used to map the reads to the reference genome (UCSC/mm10 or UCSC/hg19) with annotation data from iGenomes (Illumina). Levels of gene expression were quantified using Cuffdiff (Cufflinks version 2.2.1; default parameters).
ChIP assays and ChIP-seq
In ChIP assays, total thymus cells were fixed with 1% PFA at 25°C for 10 min, lysed in RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, pH 8.0, 1% NP-40 substitute, 0.5% sodium deoxycholate, and 0.1% SDS), and sonicated using a homogenizer (NR-50M; Micro-tec Co.). After centrifugation, supernatants were subjected to immunoprecipitation using an anti-BCOR antibody (Gearhart et al., 2006). Sheep anti–rabbit IgG Dynabeads were used to capture the anti-BCOR antibody. In the ChIP assay, quantitative PCR was performed with an ABI StepOnePlus thermal cycler with SYBR Premix Ex Taq II (Takara Bio) and the following primers: −3033 from TSS forward, 5′-TCTCCCTCCCCTTTTTCAGT-3′, reverse 5′-TGGCGTGTCATGAAACAGAT-3′; −296 from TSS forward, 5′-CAGGGCAAGAACACAGTTCA-3′, reverse 5′-GCTCCGGGGTGTAAACAGTA-3′; and 558 from TSS forward, 5′-GAGCTCCTCGAGCTGTTTG-3′, reverse 5′-ACACAGGGAAAGACCACCAG-3′.
ChIP-seq libraries were prepared using a ThruPLEX DNA-seq kit (Rubicon Genomics). Bowtie2 (version 2.2.6; default parameters) was used to map the reads to the reference genome (UCSC/mm10 or UCSC/hg19). Peaks were called using MACS2 v2.1.0 with a q-value of <0.2 (BCOR) or <10−10 (NOTCH1). ChIP peaks that overlapped with those of corresponding input (distance between centers <10 kb) were removed. Reads per million mapped reads (RPM) values of the sequenced reads were calculated every 1,000-bp bin, with a shifting size of 100 bp using Bedtools. To visualize with Integrative Genomics Viewer (Broad Institute), the RPM values of the immunoprecipitated samples were normalized by subtracting the RPM values of the input samples in each bin and converted to a Bigwig file using the Wigtobigwig tool.
Target sequencing of
Targeted sequencing of Notch1 was performed using previously described primers (Mayle et al., 2015).
Plasmids, Western blot analysis, and immunoprecipitation
3xFlag-mouse Bcor and 3xFlag-Bcor delta exon 4 were subcloned into the lentivirus vector, CSII-EF1-MCS-IRES-Venus. HA-mouse Bcl6 was subcloned into the pcDNA3 vector. To detect nonhistone proteins using Western blot analysis, lysates were prepared as follows to minimize the fragmentation of proteins by sonication. Cells were lysed in 0.1% NP-40 lysis buffer (300 mM NaCl) and centrifuged. The resulting supernatants were kept on ice (solution A). Pellets were resuspended in SDS sample buffer and sonicated using a Bioruptor (Cosmo Bio; solution B). Mixtures of solutions A and B were incubated at 95°C for 10 min. To detect histone proteins, cells were lysed in 2× SDS sample buffer, sonicated, and incubated at 95°C for 10 min. Proteins were separated by SDS-PAGE, transferred to a PVDF membrane, and detected by Western blotting using the following antibodies: anti-BCOR (Gearhart et al., 2006), anti-BCL6 (sc-7388; Santa Cruz Biotechnology), anti-actin (clone C-4, SC-47778; Santa Cruz Biotechnology), anti-Flag (clone M2, F3165; Sigma-Aldrich), anti-HA (clone 3F10, 11867423001; Roche), anti–cleaved NOTCH1 (4147; Cell Signaling Technology), anti-PCGF1 (183499; Abcam), anti-RING1B (D139-3; MBL), anti-H2AK119ub (8240S; Cell Signaling Technology), and anti–histone H2A (ab18255; Abcam). Protein bands were detected with enhanced chemiluminescence reagent (Immobilon Western; Millipore). The sequential reprobing of membranes with antibodies was performed after the removal of primary and secondary antibodies from membranes in 0.2 M glycine-HCl buffer (pH 2.5) and/or the inactivation of HRP by 0.1% NaN3. In immunoprecipitation, cells were lysed in 0.1% NP-40 lysis buffer (300 mM NaCl) as above. The pellets were resuspended in 0.1% NP-40 lysis buffer (300 mM NaCl) and sonicated (solution C). Mixtures of solution C and supernatant kept on ice were diluted with 0.1% NP-40 lysis buffer (0 mM NaCl) until the final NaCl concentrations reached 150 mM. After centrifugation, the resulting supernatants were used as input lysates for immunoprecipitation, which was performed using anti-Flag M2 Affinity Gel (A2220; Sigma-Aldrich). Immunoprecipitated proteins were subjected to Western blot analysisas described above.
Statistical tests were performed using GraphPad Prism version 5. The significance of differences was measured by a Student’s t test. Data are shown as the mean ± SEM. Significance was taken at values of *, P < 0.05; **, P < 0.01; and ***, P < 0.001.
Deposition of data
RNA-seq and ChIP-seq data were deposited in the DNA Data Bank of Japan (accession no. DRA005459).
Online supplemental material
Fig. S1 provides data of Notch1 status in BcorΔE4/y T-ALL. Fig. S2 shows immunohistochemical data of active NOTCH1 in BcorΔE4/y T-ALL. Table S1, included as a separate Excel file, lists genes up-regulated in BcorΔE4/y and Bcl6Δ/Δ DP thymocytes, Bcor ChIP-seq peaks in DP thymocytes, Notch1 ChIP-seq peaks in DP T-ALL, and BcorΔE4 ChIP-seq peaks in DP thymocytes.
We thank Drs. Toshitada Takemori and Anton Berns for providing us with Bcl6 and p53 mutant mice, respectively; Vivian Bardwell for providing BCOR antibodies and reviewing the manuscript; Noriko Yamanaka for technical assistance; Motoo Kitagawa for valuable suggestions; and Ola Mohammed Kamel Rizq for the critical review of our manuscript.
This work was supported in part by the Ministry of Education, Culture, Sports, Science and Technology, Japan, through Grants-in-Aid for Scientific Research (15H02544) and Scientific Research on Innovative Areas “Stem Cell Aging and Disease” (25115002) and grants from the Uehara Memorial Foundation, Yasuda Memorial Medical Foundation, and Tokyo Biochemical Research Foundation.
The authors declare no competing financial interests.
Author contributions: T. Tanaka, Y. Nakajima-Takagi, and K. Aoyama performed the experiments, analyzed results, made the figures, and wrote the manuscript; S. Tara, M. Oshima, A. Saraya, S. Koide, S. Si, I. Manabe, M. Sanada, M. Masuko, and H. Sone assisted with the experiments; M. Nakayama and H. Koseki generated mice; and A. Iwama conceived of and directed the project, secured funding, and wrote the manuscript.
acute myeloid leukemia
gene set enrichment analysis
hematopoietic stem cell
hematopoietic stem and progenitor cell
Lin- Sca-1+ c-Kit+
stem cell factor
acute T-cell lymphoblastic leukemia
transcriptional start site
T. Tanaka, Y. Nakajima-Takagi, and K. Aoyama contributed equally to this work.