t(17;19)-acute lymphoblastic leukemia (ALL) shows extremely poor prognosis. E2A-HLF derived from t(17;19) blocks apoptosis induced by the intrinsic mitochondrial pathway and has a central role in leukemogenesis and chemoresistance. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is expressed on cytotoxic T cells and natural killer cells and binds with death receptors (DR4/DR5), inducing apoptosis by dual activation of intrinsic and extrinsic pathways, and TRAIL mediates the graft-versus-leukemia (GVL) effect after allogeneic stem cell transplantation (allo-SCT). We found that cell lines and patients' samples of t(17;19)-ALL expressed death receptors for TRAIL, and recombinant soluble TRAIL immediately induced apoptosis into t(17;19)-ALL cell lines. E2A-HLF induced gene expression of DR4/DR5, which was dependent on the DNA-binding and transactivation activities of E2A-HLF through the 5′ upstream region of the start site at least in the DR4 gene. Introduction of E2A-HLF into non-t(17;19)-ALL cell line upregulated DR4 and DR5 expression, and sensitized to proapoptotic activity of recombinant soluble TRAIL. Finally, a newly diagnosed t(17;19)-ALL patient underwent allo-SCT immediately after induction of first complete remission, and the patient has survived without relapse for over 3–1/2 years after allo-SCT. These findings suggest that E2A-HLF sensitizes t(17;19)-ALL to the GVL effect by upregulating death receptors for TRAIL.
The graft-versus-leukemia (GVL) effect after allogeneic stem cell transplantation (allo-SCT) is mediated by cytotoxic factors on cytotoxic T cells and natural killer cells1 and is a critical response for curing patients with leukemia. It was previously demonstrated that death receptor ligands including Fas ligand (FasL)2, 3 and tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)4, 5 have an important role in cytotoxic T-cell- and natural killer cell-mediated antitumor immunity. TRAIL induces apoptosis in a variety of tumor cells6 upon binding with its death-inducing receptors, DR4(ref. 7) and DR5.8 In contrast, DcR1(refs 8,9) and DcR2(ref. 10) lack a functional death domain and compete with DR4 and DR5 for TRAIL binding as ‘decoy receptors’. The binding of TRAIL to death-inducing receptors induces the formation of a death-inducing signaling complex that includes death receptors, adaptor proteins and procaspase-8.11, 12 Activation of procaspase-8 mediated through an autocatalytic mechanism cleaves procaspase-3 and thereby activates the extrinsic pathway. Simultaneously, caspase-8 cleaves Bid, thereby activating the intrinsic mitochondrial pathway. Owing to activation of these dual proapoptotic pathways, TRAIL effectively induces apoptosis of tumor cells and seems to be a promising candidate for cancer treatment. The potential of TRAIL as an anticancer agent is supported by studies in animal models showing selective toxicity to human tumor xenografts but not normal tissues.13, 14 Moreover, in bone marrow transplantation models using TRAIL-deficient mice, TRAIL was required for optimal graft-versus-tumor activity by donor T cells, whereas it had little or no role in the development of graft-versus-host disease.6 We previously demonstrated that Philadelphia chromosome (Ph1)-positive leukemia cells, which have been reported to be clinically sensitive to the GVL effect,15, 16 frequently express DR4 and/or DR5 and are sensitive to the antileukemic activity of recombinant human soluble (rhs) TRAIL.17 In contrast, mixed-lineage leukemia (MLL)-rearranged leukemia cells, which have been reported to be rather clinically resistant to the GVL effect,18 generally express very low or undetectable levels of DR4 and DR5 and are resistant to rhsTRAIL.19 Accordingly, DR4 and DR5 expression could be one of the critical factors in the susceptibility of leukemia cells to TRAIL and consequently to the GVL effect.
t(17;19)(q21-q22;p13) is a relatively rare translocation present in <1% of childhood acute lymphoblastic leukemia (ALL) cases,20 and its associations with the B-precursor phenotype,21 aberrant expression of CD33,22 hypercalcemia23 and acquired coagulation abnormalities24, 25 have been noted.26 The E2A-HLF fusion gene generated by t(17;19)(q21-q22;p13) encodes a chimeric protein in which the transactivation domain of E2A links to the basic leucine-zipper dimerization and DNA-binding domain of HLF.27, 28 As a result, E2A-HLF binds to a consensus sequence of nucleotides (HLF-CS; 5′-IndexTermGTTACGTAAT-3′) as a dimer and transactivates downstream target genes.28, 29, 30 E2A-HLF promotes anchorage-independent growth of murine fibroblasts31 and protects cells from apoptosis due to growth factor deprivation32, 33, 34 and p53 activation.35 Although recent chemotherapeutic regimens have remarkably improved the long-term outcome of ALL in childhood and adolescence, the prognosis of t(17;19)-ALL still remains extremely poor.23 Among 14 cases treated with chemotherapy, 13 cases relapsed or died within 2 years from diagnosis with the median period of the first complete remission (CR) being 9 months.23 In contrast, one reported patient who underwent allogeneic bone marrow transplantation during the first CR exceptionally maintained CR for over 3 years,36 suggesting that allo-SCT performed early in the first CR would prolong the disease-free survival of t(17;19)-ALL patients even if they are not cured. Therefore, to address the clinical importance of allo-SCT for improving the prognosis of t(17;19)-ALL, it is urgent to confirm whether t(17;19)-ALL cells are sensitive to cytotoxic ligands.
In the present study, we found that t(17;19)-ALL cells expressed high levels of DR4 and DR5 and showed remarkably high sensitivity to the cytotoxic activity of rhsTRAIL. Of note, E2A-HLF induced gene and cell surface expression of DR4 and DR5, resulting in the sensitization of non-t(17;19) ALL cell line to the antileukemic activity of rhsTRAIL. These observations indicate that E2A-HLF directly induces expression of DR4/DR5 and sensitizes leukemia cells to the antileukemic activity of TRAIL.
Materials and methods
Leukemia cell lines and patients' samples
Four t(17;19)-ALL cell lines,21, 22 UOC-B1, HAL-O1, YCUB-2 and Endo-kun, were used in this study. MLOT4 was used as a control for FasL sensitivity. A total of 9 MLL-rearranged ALL (MLL+ALL) cell lines (KOPN1, KOPB26, KOCL33, -44, -45, -50, -51, -58 and -69),19 6 Ph1-positive ALL cell lines (KOPN30bi, -57bi, -66bi, -72bi, YAMN73 and -91)17 and 13 B-precursor ALL cell lines19 including 7 with t(1;19) (697, KOPN34, -36, -60, -63, YAMN90 and -92), 1 with t(12;21) (Reh) and 5 with others (KOPN35, -61, -62, -79 and -84) were also used. Analysis of samples from patients with t(17;19)-ALL was approved by the Ethical Review Board of University of Yamanashi. Mononuclear cells of bone marrow separated by Ficoll–Hypaque centrifugation were stored in liquid nitrogen and used for experiments.
Cytotoxic activity assays
RhsTRAIL (Killer TRAIL) and rhsFasL (SUPER FasL) were purchased from Alexis Biochemicals (San Diego, CA, USA). For the 3H-thymidine uptake assay, cells (5 × 104 cells/well) were cultured in triplicate in a flat-bottomed 96-well plate. The plates were incubated for the indicated periods of time, pulsed for the last 6 h of the incubation with 3H-thymidine (1 μ Ci/well) and harvested onto glass-fiber filters. The level of radioactivity incorporated into DNA was measured by liquid scintillation counting. The effect of rhsTRAIL or rhsFasL was determined in a 42-hour incubation in the absence or presence of rhsTRAIL or rhsFasL (11, 33, 100 ng/ml). The % inhibition by TRAIL or FasL was calculated as follows: (1−((cpm of treated well)/(cpm of untreated well))) × 100. RIK-2 (10 μg/ml), a neutralizing anti-TRAIL monoclonal antibody (mAb)17 and z-VAD-fmk (20 μM), a caspase inhibitor with broad spectrum (Enzyme Systems Products, Livermore, CA, USA), were used to block the activity of rhsTRAIL and caspases, respectively. To detect the early apoptotic event, cells (4 × 105 cells/ml) were cultured in the absence or presence of rhsTRAIL (100 ng/ml) for 12 h, stained with fluorescein isothiocyanate-conjugated Annexin-V (MBL, Nagoya, Japan) and analyzed by flow cytometry (FACSCalibur, BD Biosciences, San Jose, CA, USA). Viability of the primary samples was determined by staining with trypan blue.37 The Δ viability (%) of samples from the patients with Ph1-positive ALL, MLL+ALL, T-ALL or T-non-Hodgkin lymphoma, calculated as: (% viability in the absence of rhsTRAIL)−(% viability in the presence of rhsTRAIL), was determined when percent viability of background was >60%.37
Cell lysates were prepared in Nonidet P-40 (50 mM Tris-HCL, pH 7.5, 150 mM NaCl, 1.0% Nonidet P-40, 5 mM EDTA, 0.05% NaN3, 1 mM phenylmethylsulfonyl fluoride, 100 mM sodium vanadate). Total cellular proteins were separated on a SDS-polyacrylamide gel electrophoresis under reducing conditions and then transferred to polyvinyl difluoride membranes. After blocking with 5% non-fat dry milk in 0.05% Tween-20 Tris-buffered saline, membranes were incubated with the primary antibodies in 5% milk in Tris-buffered saline at 4 °C overnight. MAbs against human PARP and caspase-3 were purchased from BD Transduction Laboratories (Lexington, KY, USA) and mAb against caspase-8 from MBL. Rabbit antisera against human E2A and HLF(c) were established as previously reported29 and antiserum against Bid was purchased from Cell Signaling Technology (Beverly, MA, USA). Membranes were incubated with horseradish peroxidase-conjugated goat anti-mouse or rabbit IgG (1:1000 dilution; MBL) for 1 h and were then developed using an enhanced chemiluminescence kit (Amersham Pharmacia Biotec, Buckinghamshire, UK).
Cell surface expression of TRAIL receptors
MAbs specific for DR4, DR5, DcR1 or DcR2 were originally generated as previously reported.17, 19 A total of 1 × 106 cells were incubated with 1 μg of biotinylated control mouse IgG1 or mAb on ice for 30 min. After washing, the cells were incubated with phycoerythrin-conjugated streptavidin (Biomeda, Foster City, CA, USA) on ice for 30 min and then analyzed by flow cytometry. Relative fluorescence intensities (RFIs) were calculated as the ratio of the mean fluorescence intensity of specific staining to that of control staining.
Real-time PCR analysis
Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA). Reverse transcription was performed with random hexamer (Amersham Bioscience, Uppsala, Sweden) and Superscript II reverse transcriptase (Invitrogen) and the product was then incubated with 1 μl of RNase (Invitrogen) at 37 °C for 20 min. For quantitative real-time PCR of DR4 and DR5 transcripts, triplicated samples containing 9 μl of cDNA with 10 μl of Taqman Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) and 1 μl of 20 × Assays-on-demand Gene Expression Product (DR4; Hs 00269492_m1, DR5; Hs 00366272_m1, Applied Biosystems) were preincubated at 50 °C for 2 min and subsequently at 95 °C for 10 min.37 Amplification was performed by 40 cycles of reaction at 95 °C for 15 s and 60 °C for 1 min. Fluorescence data were quantitatively analyzed on an ABI Prism 7500 Sequence Detection System (Applied Biosystems). Nalm1, a TRAIL-sensitive CML-BC-derived cell line that expresses DR4 and DR5,17 was used as a control. As an internal control, quantitative real-time PCR for GAPDH (Hs 99999905_m1, Applied Biosystems) was performed.
Construction of eukaryotic expression vectors and transfection
Expression plasmids containing wild-type and mutated E2A-HLF cDNA were constructed with the pMT-CB6+ eukaryotic expression vector (a gift from F Rauscher III, Wistar Institute, Philadelphia, PA, USA).32 ΔAD1/ΔLH mutant lacks both of the two transactivation domains of E2A, whereas basic lesion mutant (BX) contains substitutions of six critical basic amino acids in the DNA-binding portion of the HLF bZIP domain as previously reported.32, 33 Transfectants of 697 cell line were generated by electroporation with a gene pulser (Bio-Rad, Hercules, CA, USA)31, 33 and selected in the presence of the neomycin analog G418 for 4 weeks. The induction of protein expression in G418-resistant cells was confirmed by immunoblot analysis of cells cultured in the presence of zinc (ZnSO4) at 100 μM, and several independent clones expressing the expected protein at comparable levels were selected for further experiments. Transfectants or wild type of 697 cells were preincubated with a proteasome inhibitor N-acetyl-Leu-Leu-norLeu-al (LLnL) (Sigma-Aldrich, Tokyo, Japan)17 at a concentration of 2.5 μM for 3 h. To determine the susceptibilities to rhsTRAIL, transfectants or wild type of 697 cells preincubated with or without zinc at 100 μM for 24 h were cultured in the absence or presence of rhsTRAIL (3.3 or 10 ng/ml) for 12 h and induction of cell death was monitored by flow cytometry using the propidium iodide (PI) and Annexin-V double staining.
−820/−2 (DR4pro) and −820/+620 (DR4int1) fragments of the DR4 gene generated by PCR using genomic DNA of UOC-B1 cells as a template were cloned into the pGL3 basic vector (Promega, Madison, WI, USA) and subjected to nucleotide sequence analysis. The SacI/NcoI (DR5pro) and PstI/MluI (DR5int1) fragments of the DR5 gene were cloned into the pGL3 basic vector (a gift from T Sakai, Kyoto Prefectural University of Medicine, Japan).38 The PstI/MluI fragment with mutated nuclear factor-kappaB (NF-κB) site (DR5int1mNF; 5′-IndexTermggaGGGAATTCCCgag-3′ was mutated as 5′-IndexTermggaTTTAATTCCCgag-3′) was generated by PCR mutagenesis (underlined sequences indicate the mutated nucleotids). pRL-TK vector (Promega) was used as parameters for transfection efficiency. For transfection, cells (5 × 105 cells/well) were plated in a 24-well plate and 5 μg of luciferase reporter plasmid, 1 μg of pRL-TK and 2 μl of lipofectamine (Invitrogen) in 50 μl of serum-free medium (Opti-MEM I, Invitrogen) were mixed and added to the cells. In some experiments, 0.1 or 0.2 μg of RC/RSV expression vector (Invitrogen) in which E2A-HLF cDNA had been subcloned, was cotransfected. Cells were harvested 48 h after transfection, and the activities of firefly and Renilla luciferases in each lysate were measured sequentially using the Dual-Luciferase reporter assay system (Promega).
Electrophoretic mobility shift assay
Nuclear extract of UOC-B1 cells was prepared and binding reactions was performed as previously reported.29 Briefly, a 32P-end-labeled oligonucleotide probe containing the HLF-binding site sequence (HLF-CS probe; 5′-IndexTermGCTACATATTACGTAATAAGCGTT-3′) was incubated in 10 μl of binding buffer and 5 μl of nuclear lysates in the presence of 1 μg of shared calf thymus DNA as carriers (underlined sequence indicates the position of HLF-CS). In the competitive inhibition experiments, nuclear lysates were preincubated with excess of unlabeled oligonucleotides.
Sensitivity of t(17;19)-ALL cells to TRAIL
As clinical observation suggested that t(17;19)-ALL cells might be sensitive to the antileukemic activities of cytotoxic factors that mediate the GVL effect, we examined the susceptibilities of four t(17;19)-ALL cell lines, whose expression of E2A-HLF was shown in Supplementary Figure 1, to rhsTRAIL and rhsFasL by 3H-thymidine uptake assay (Figure 1a). rhsTRAIL markedly inhibited cell growth in a dose-dependent manner, and the % inhibition upon incubation with 100 ng/ml of rhsTRAIL was nearly 100% in all four t(17;19)-ALL cell lines. In contrast, t(17;19)-ALL cells were relatively resistant to rhsFasL (Figure 1a): the % inhibition upon incubation with 100 ng/ml of rhsFasL was <50% in three of the four cell lines and median % inhibition was approximately 30%. When compared with various B-precursor ALL cell lines (Figure 1b), the % inhibition upon incubation with 100 ng/ml of rhsTRAIL in 4 t(17;19)-ALL cell lines (median 100%) was significantly higher than that in 9 MLL+ALL cell lines (median 5%, P=0.0055),19 6 Ph1-positive ALL cell lines (median 50%, P=0.0142)17 and 13 other B-precursor ALL cell lines (median 25%, P=0.0362) by the Mann–Whitney test. The antileukemic activity of TRAIL was specific and dependent on the activation of caspases, as RIK-2, a neutralizing antibody against TRAIL17 and z-VAD-fmk, a broad caspase inhibitor, almost completely blocked the activity of rhsTRAIL (Figure 1c). Upon 12-hour incubation with rhsTRAIL (100 ng/ml), YCUB-2 cells underwent Annexin-V-binding/PI staining-double-positive cell death and the other three cell lines underwent apoptotic cell death positive for Annexin-V-binding (Figure 1d). As ligation of death receptors by TRAIL induces autoproteolytic activation of caspase-3 both directly through activation of caspase-8 (extrinsic pathway) and indirectly through activation of Bid (intrinsic mitochondrial pathway),11, 12 we subjected t(17;19)-ALL cell lines (UOC-B1 and Endo-kun) treated with rhsTRAIL to immunoblot analysis of caspases-8 and -3, Bid and PARP, a substrate of caspase-3. These molecules were processed within 2 h after rhsTRAIL treatment (Figure 1e).
Expression of TRAIL receptors on t(17;19)-ALL cells
To verify that the antileukemic activity of TRAIL against t(17;19)-ALL cells is mediated by expression of death receptors, we analyzed cell surface expression of TRAIL receptors by flow cytometry using specific mAbs for each receptor.17, 19, 37 Cell surface expression of DR4 and DR5 was clearly detectable on all four t(17;19)-ALL cell lines except for HAL-O1, which showed marginal expression of DR4, whereas that of DcR1 and DcR2 was almost undetectable (Figure 2a). Strong correlations were observed between the sensitivity of a cell line to rhsTRAIL and its cell surface expression of DR4 (r=0.666, P<0.0001) and DR5 (r=0.539, P=0.0012) (Figure 2b and Supplemental Figure 2). When compared with various B-precursor ALL cell lines (Figure 2c), the RFI of DR4 on t(17;19)-ALL cell lines (median 1.9) was significantly higher than that on MLL+ (median 1.0, P=0.0055),19 Ph1-positive (median 1.05, P=0.0330)17 and other B-precursor ALL cell lines (median 1.0, P=0.0127)37 by Mann–Whitney analysis. In addition, the RFI of DR5 was higher than 1.4 in all 4 t(17;19)-ALL cell lines, whereas it was ⩽1.3 in 5 of 9 MLL+, 2 of 6 Ph1-positive, and 8 of 13 other B-precursor ALL cell lines.
Next, the levels of DR4 and DR5 mRNA expression were quantitatively analyzed by real-time RT-PCR. Significant correlations were observed between cell surface expression and gene expression of DR4 (r=0.517, P=0.0021) and DR5 (r=0.35, P=0.048) (Figure 2d). In the case of DR4 gene expression, although there were limited statistically significant differences between the t(17;19)-ALL cell lines and other B-precursor ALL cell lines, relatively high levels of transcripts were detected in all four t(17;19)-ALL cell lines (Figure 2e). In the case of DR5 gene expression, all 4 t(17;19)-ALL cell lines and 18 of 28 other B-precursor ALL cell lines expressed comparable or relatively higher level of transcripts in comparison with Nalm1 cell line, which was a TRAIL-sensitive CML-BC-derived cell line.17
Induction of DR4 and DR5 and sensitization to antileukemic activity of TRAIL by E2A-HLF
As E2A-HLF chimeric transcription factor derived from t(17;19) has an essential role in the leukemogenesis of t(17;19)-ALL,27, 28 E2A-HLF might be involved in the distinctive gene expression of DR4 and DR5 in t(17;19)-ALL cells. To directly test this possibility, we transfected E2A-HLF into 697 cells, a B-precursor ALL cell line with t(1;19), using a zinc-inducible vector, and studied the induction of DR4 and DR5 gene expression. An obtained clone of 697 cells expressed E2A-HLF at equivalent levels to UOC-B1 cells within 4 h of culture in the presence of zinc (Figure 3a). Of note, the gene expression levels of DR4 and DR5 in the 697 cells transfected with E2A-HLF were also remarkably upregulated within 4–8 h after the addition of zinc (Figure 3b). In contrast, the gene expression levels of DR4 and DR5 in the wild-type cells remained almost unchanged, suggesting that an influence of cell damage by zinc over DR4 and DR5 gene expression was almost negligible. The basal gene expression level of DR4 was almost above the threshold for cell surface expression (Supplemental Figure 3). In contrast, the basal gene expression level of DR5 was under the threshold for cell surface expression, but it was upregulated above the threshold for cell surface expression (approximately 40-fold of background) by the addition of zinc (Supplemental Figure 3). The basal gene expression levels of DR4 and DR5 in the 697 cells transfected with E2A-HLF were almost two-fold and ten-fold higher than those in wild-type 697 cells, respectively, probably due to a leaky expression of E2A-HLF. Consistent with the gene expression, cell surface expression of DR4 and DR5 in the 697 cells transfected with E2A-HLF was significantly upregulated by the addition of zinc (Figures 3c and d), whereas it was almost unchanged in the wild-type 697 cells.
We further analyzed the susceptibility of E2A-HLF-transfected 697 cells to the antileukemic activity of rhsTRAIL. After 24 h preincubation with or without zinc, the cells were cultured for 12 h in the absence or presence of lower concentrations rhsTRAIL (3.3 and 10 ng/ml), and analyzed by Annexin-V/PI staining (Figures 3e and f). The wild-type 697 cells showed resistance to lower concentrations of rhsTRAIL. In contrast, E2A-HLF sensitized leukemia cells to antileukemic activity of rhsTRAIL, as the viability of the E2A-HLF-transfected 697 cells not preincubated with zinc was decreased from 91.5% to 84.7% and 61.9% by the treatment with rhsTRAIL at 3.3 and 10 ng/ml, respectively, and that preincubated with zinc was further decreased from 85.8% to 72% and 40.6%, respectively.
Transactivation of DR4 and DR5 gene promoters by E2A-HLF
To clarify the mechanisms for E2A-HLF-inducing DR4 and DR5 expression, we tested the activity of two types of E2A-HLF mutants (Figure 4a); BX contains substitutions of six critical basic amino acids in the basic region of HLF to abolish DNA-binding ability, whereas ΔAD1/ΔLH lacks both E2A transactivation domains to abolish transactivation ability but retain DNA-binding ability.32, 33 Obtained clones of 697 cells expressed BX and ΔAD1/ΔLH mutant proteins in the presence of zinc at nearly equivalent levels as E2A-HLF (Figure 4b), but the gene expression levels of DR4 and DR5 remained almost unchanged (Figure 4c), indicating that E2A-HLF upregulates the gene expression of DR4 and DR5, which is dependent on both the DNA-binding and transactivation abilities of E2A-HLF. We next analyzed the promoter activity of the DR4 and DR5 genes. Owing to high structural homology in gene configuration and tight cluster on 8p21-22, the genes encoding DR4 and DR5 are considered to be generated from a common ancestral gene. p53(refs 39, 40) and NF-κB41, 42 binding sites are commonly identified in the 5′ end of intron1 of the DR4 and DR5 genes (Figure 4d), and their involvement in p53-induced and NF-κB-induced DR4 and DR5 expression has been well clarified.39, 40, 41, 42, 43 In contrast, the basic structures of the 5′ upstream region of the transcription start site of the DR4(ref. 43) and DR5(ref. 38) genes are largely different from each other (Figure 4d). First, the activity of the 5′ upstream region of the transcription start site of the DR4 gene (DR4pro) and DR5 gene (DR5pro) (Figure 4d) was tested by luciferase assay in YCUB-2, one of the t(17;19)-ALL cell lines, and KOPT5, a T-ALL cell line that lacks cell surface expression of DR4 and DR5(ref. 37) as a negative control. As shown in Figure 4e, PGL3-control, which contains CMV promoter and was served as a marker for transfection efficiency, was active in both YCUB-2 and KOPT5, and both DR4pro and DR5pro were active in YCUB-2 but almost silent in KOPT5. In contrast, the activity of DR4int1 that also contains the 5′ end of intron1 of the DR4 gene (Figure 4d) was almost silent in YCUB-2 (Figure 4e), suggesting that exon1 and/or the 5′ end of intron1 might contain a negative regulatory region. Although the activity of DR5intl that contains the 5′ end of intron1 of the DR5 gene (Figure 4d) was relatively high in comparison with that of DR5pro in YCUB-2 (Figure 4e), the activity of DR5int1mNF that has a mutation in the NF-κB site of the 5′ end of intron1 (Figure 4d) was almost unchanged in comparison with that of DR5intl (Figure 4e). Next, considering that E2A-HLF induced both DR4 and DR5 gene expression (Figure 3b), the involvement of the NF-κB site in the induction of DR4 and DR5 expression by E2A-HLF was tested using proteasome inhibitor LLnL.17 Although the induction of DR5 gene expression was relatively weak in this experimental setting, real-time RT-PCR analysis demonstrated that the induction of DR4 and DR5 was not affected by LLnL in E2A-HLF-transfected 697 cells (Figure 4f), suggesting that activation of NF-κB is unlikely to be involved in the E2A-HLF-mediated DR4 and DR5 expression. Finally, we tested the activation of DR4pro and DR5pro by E2A-HLF in 697 cells using transient transfection (Figure 4g). E2A-HLF activated DR4pro in 697 cells, suggesting that E2A-HLF induces gene expression of DR4 through the 5′ upstream region of the transcription start site in the DR4 gene in the absence of binding sites for p53 and NF-κB. In contrast, E2A-HLF did not significantly activate DR5pro in 697 cells (Figure 4g). Considering that E2A-HLF upregulated DR5 expression in 697 cells (Figure 3) and that DR5pro was active in YCUB-2 cells (Figure 4e), the inactivity of DR5pro in 697 cells suggests that the 5′ upstream region of the transcription start site in the DR5 gene alone was insufficient for effective gene expression of DR5 by exogenously introduced E2A-HLF in 697 cells.
In intron1 of the DR4 (Figure 4h) and DR5 (Figure 4i) genes, there were a couple of candidate sequences for the potential E2A-HLF-binding sites with 2–3 bases mismatch to HLF-CS. Thus, we next analyzed the binding of E2A-HLF to these sequences by electrophoretic mobility shift assay. Among these candidate sequences, site b and site d of the DR4 gene showed weak and moderate competition, respectively (Figure 4j), whereas none of those sites of the DR5 gene showed significant competition (Figure 4k). These observations suggested that several binding sites for E2A-HLF in intron1 of the DR4 gene might be also involved in the E2A-HLF-medaited DR4 expression in t(17;19)-ALL cells.
DR4 and DR5 gene expression and TRAIL sensitivity in clinical samples and effect of allo-SCT
We next analyzed the gene expression of DR4 and DR5 in the primary sample of t(17;19)-ALL patient, whose frozen sample was available. As shown in Figures 5a and b, the gene expression levels of DR4 and DR5 in the patient's leukemia cells (case 1) were as high as those in the t(17;19)-ALL cell lines. We also analyzed the sensitivity of the other patient's sample (case 2) to rhsTRAIL. As shown in Figures 5c and d, although basal viability of the sample in the absence of rhsTRAIL was relatively low (% viability in trypan blue exclusion assay and Annexin-V/PI staining was 58% and 26%, respectively), decreased viability (Figure 5c) and induction of apoptotic cell death (Figure 5d) in the presence of rhsTRAIL were confirmed. Considering the exceptionally long first CR in the t(17;19)-ALL case who underwent allo-SCT early in the first CR36 and higher sensitivity of t(17;19)-ALL cell lines to the proapoptotic activity of TRAIL, performing allo-SCT in the first CR would be effective for t(17;19)-ALL. Although the incidence of t(17;19)-ALL is <1% of childhood ALL cases, we previously reported that over 20% of childhood ALL cases who presented with hypercalcemia have t(17;19).23 Thus, we prospectively performed RT-PCR analysis of E2A-HLF in newly diagnosed ALL patients with hypercalcemia. Among five patients, an 8-year-old boy, whose karyotyping was later confirmed to have t(17;19), had E2A-HLF by RT-PCR analysis. Of note, the patient's leukemia cells demonstrated comparable gene expression levels of DR4 and DR5 (Figures 5a and b, case 3). On the basis of these results, the patient received unrelated-BMT (u-BMT) in the first CR 7 months after diagnosis with a conditioning regimen consisting of total body irradiation (12 Gy) and melphalan (90 mg/m2/dose × 2 days). The patient has no sign of relapse at 43 months after u-BMT (50 months after diagnosis), and minimal residual disease analyzed by nested RT-PCR for E2A-HLF, in which sensitivity to detect minimal residual disease was 10−5 level, was negative in the BM obtained 12 and 18 months after u-BMT.
In the present study, we showed that t(17;19)-ALL cell lines are extremely sensitive to the antileukemic activity of rhsTRAIL due to the high levels of DR4 and DR5 expression. High levels of DR4 and DR5 gene expression were also confirmed in samples from patients with t(17;19)-ALL. As E2A-HLF induced gene expression of DR4 and DR5 in 697 cells, which was dependent on the transactivation and DNA-binding activities of E2A-HLF, E2A-HLF seems to have a central role in the high levels of DR4 and DR5 expression in t(17;19)-ALL cells. To our knowledge, this is the first study to demonstrate that a fusion transcription factor derived from chromosomal translocation induces DR4 and DR5 expression. There are no sequences for the potential binding site of E2A-HLF in the 5′ upstream regions of the DR4 and DR5 genes, suggesting that E2A-HLF indirectly transactivates the DR4 and DR5 gene promoters. The involvement of p53 and NF-κB binding sites in the DR4 and DR5 genes in tumor-specific expression was reported,39, 40, 41, 42 but E2A-HLF transactivated at least a fragment of the DR4 gene that lacks the p53 and NF-κB binding sites in the luciferase assay. Moreover, LLnL, an inhibitor of NF-κB, failed to attenuate the induction of DR4 and DR5 gene expression by E2A-HLF in 697 cells. These findings suggest that E2A-HLF indirectly induces the gene expression of DR4 and DR5 through activation of a transcription factor(s) other than p53 and NF-κB. As there were several E2A-HLF-binding sites in intron1 of the DR4 gene, there is a possibility that E2A-HLF also induces the gene expression of DR4 and DR5 cooperatively through the binding sites located outside the adjacent promoter region.
E2A-HLF blocks apoptosis of cells induced by growth factor deprivation32, 33, 34 and p53 activation,35 which activate the intrinsic mitochondrial pathway, and, thus, it has a central role in leukemogenesis and might be involved in the poor prognosis of t(17;19)-ALL. In the present study, however, t(17;19)-ALL cell lines showed the highest sensitivity to the proapoptotic activity of rhsTRAIL among B-precursor ALL cell lines. Moreover, introduction of E2A-HLF to 697 cells increased their sensitivity to lower concentrations of rhsTRAIL through upregulation of DR4 and DR5 expression. Considering that binding of TRAIL to a death receptor induces both the mitochondria-dependent intrinsic pathway and the mitochondria-independent extrinsic pathway,11, 12 TRAIL could bypass the antiapoptotic activities of E2A-HLF against the intrinsic mitochondrial pathway by directly activating downstream effector caspases through the extrinsic pathway (Figure 6).
As it was reported that TRAIL has a role in innate immunity for immune surveillance against tumor development,4, 5 upregulated DR4 and DR5 expression by E2A-HLF might sensitize leukemia cells (or preleukemia cells) to innate immunity, resulting in suppression of leukemia clone expansion. In this scenario, how is induction of DR4 and DR5 involved in the leukemogenesis of t(17;19)-ALL? To explain a paradox that TRAIL-sensitive leukemia evades innate immunity, chemical carcinogen methylcholanthrene (MCA)-induced mouse sarcoma model44 may be worth noting: neutralization of endogenous TRAIL by the anti-TRAIL antibody promoted development of sarcomas in mice treated with low-dose MCA but not significantly in mice treated with high-dose MCA. In this model, it was noted that high-dose MCA preferentially generated TRAIL-sensitive sarcomas whereas low-dose MCA generated TRAIL-resistant sarcomas, suggesting that endogenous TRAIL-mediated immune surveillance could be overwhelmed by an emergence of high-potential tumors.44 These findings may be relevant to the paradoxical finding that TRAIL-sensitive t(17;19)-ALL shows clinically aggressive feature and poor therapeutic outcome in chemotherapy. Moreover, in some studies,45, 46 TRAIL has potential to stimulate cell growth of tumor cells through DR4 and DR5 in the context of receptor-proximal apoptosis defects such as loss of caspase-8 and overexpression of FLIP. As expression levels of caspase-8 and FLIP in tumor cells were reported to be regulated by interferon-γ47 and interferon-α,48 respectively, TRAIL might have dual activities to stimulate proliferation and cell death pathways of tumor cells depending on the cellular environment. Thus, E2A-HLF might stimulate proliferation of t(17;19)-ALL cells by the induction of DR4/DR5 expression in a certain context.
Alternatively, high sensitivity of t(17;19)-ALL to TRAIL suggests that an enhancement of TRAIL-mediated surveillance by donor-derived cytotoxic T cells or natural killer cells after allo-SCT could overcome the poor therapeutic outcome of t(17;19)-ALL. In fact, it was reported that one exceptional t(17;19)-ALL patient receiving allogeneic bone marrow transplantation early in the first CR avoided relapse for 42 months,36 in contrast to 14 other reported cases treated with chemotherapy alone who relapsed within 24 months from diagnosis.23 Accordingly, we performed u-BMT in the newly diagnosed t(17;19)-ALL case 7 months after diagnosis, and the patient did not have sign of relapse 43 months after u-BMT with negative minimal residual disease in BM. These observations suggest that the TRAIL-sensitive ALL could be sensitive to the GVL effect in vivo once allogeneic immunity is initiated toward leukemia cells. Moreover, our results indicated that t(17;19)-ALL patients seems to be a candidate for the treatment modality that activates the DR4/DR5 death pathways. Recent results of a phase 1 clinical trials of the humanized agonistic mAbs against DR4(ref. 49) or DR5(ref. 50) in patients with solid tumors demonstrated that their administration was well tolerated and showed signs of clinical activity. Therefore, although there is a possibility that humanized anti-DR4/DR5 agonistic mAbs might stimulate growth of t(17;19)-ALL in a certain situation as discussed before, therapy of humanized anti-DR4/DR5 agonistic mAbs would be a good therapeutic modality for overcoming chemoresistance in patients with t(17;19)-ALL.
Kolb HJ, Schmid C, Barrett AJ, Schendel DJ . Graft-versus-leukemia reactions in allogeneic chimeras. Blood 2004; 103: 767–776.
Kagi D, Vignaux F, Ledermann B, Burki K, Depraetere V, Nagata S et al. Fas and perforin pathways as major mechanisms of T cell-mediated cytotoxicity. Science 1994; 265: 528–530.
Lowin B, Hahne M, Mattmann C, Tschopp J . Cytolytic T-cell cytotoxicity is mediated through perforin and Fas lytic pathways. Nature 1994; 370: 650–652.
Smyth MJ, Cretney E, Takeda K, Wiltrout RH, Sedger LM, Kayagaki N et al. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) contributes to interferon gamma-dependent natural killer cell protection from tumor metastasis. J Exp Med 2001; 193: 661–670.
Takeda K, Hayakawa Y, Smyth MJ, Kayagaki N, Yamaguchi N, Kakuta S et al. Involvement of tumor necrosis factor-related apoptosis-inducing ligand in surveillance of tumor metastasis by liver natural killer cells. Nat Med 2001; 7: 94–100.
Schmaltz C, Alpdogan O, Kappel BJ, Muriglan SJ, Rotolo JA, Ongchin J et al. T cells require TRAIL for optimal graft-versus-tumor activity. Nat Med 2002; 8: 1433–1437.
Pan G, O'Rourke K, Chinnaiyan AM, Gentz R, Ebner R, Ni J et al. The receptor for the cytotoxic ligand TRAIL. Science 1997; 276: 111–113.
Pan G, Ni J, Wei YF, Yu G, Gentz R, Dixit VM . An antagonist decoy receptor and death domain-containing receptor for TRAIL. Science 1997; 277: 815–818.
Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, Baldwin D et al. Control of TRAIL-induced apoptosis by a family of signaling and decoy receptors. Science 1997; 277: 818–821.
Degli-Esposti MA, Dougall WC, Smolak PJ, Waugh JY, Smith CA, Goodwin RG . The novel receptor TRAIL-R4 induces NF-kappaB and protects against TRAIL-mediated apoptosis, yet retains an incomplete death domain. Immunity 1997; 7: 813–820.
Sprick MR, Weigand MA, Rieser E, Rauch CT, Juo P, Blenis J et al. FADD/MORT1 and caspase-8 are recruited to TRAIL receptors 1 and 2 and are essential for apoptosis mediated by TRAIL receptor 2. Immunity 2000; 12: 599–609.
Kischkel FC, Lawrence DA, Chuntharapai A, Schow P, Kim KJ, Ashkenazi A . Apo2L/TRAIL-dependent recruitment of endogenous FADD and caspase-8 to death receptors 4 and 5. Immunity 2000; 12: 611–620.
Walczak H, Miller RE, Ariail K, Gliniak B, Griffith TS, Kubin M et al. Tumoricidal activity of tumor necrosis factor-related apoptosis-inducing ligand in vivo. Nat Med 1999; 5: 157–163.
Ashkenazi A, Pai RC, Fong S, Leung S, Lawrence DA, Marsters SA et al. Safety and antitumor activity of recombinant soluble Apo2 ligand. J Clin Invest 1999; 104: 155–162.
Faderl S, Talpaz M, Estrov Z, O'Brien S, Kurzrock R, Kantarjian HM . The biology of chronic myeloid leukemia. N Engl J Med 1999; 341: 164–172.
Arico M, Valsecchi MG, Camitta B, Schrappe M, Chessells J, Baruchel A et al. Outcome of treatment in children with Philadelphia chromosome-positive acute lymphoblastic leukemia. N Engl J Med 2000; 342: 998–1006.
Uno K, Inukai T, Kayagaki N, Goi K, Sato H, Nemoto A et al. TNF-related apoptosis-inducing ligand (TRAIL) frequently induces apoptosis in Philadelphia chromosome-positive leukemia cells. Blood 2003; 101: 3658–3667.
Pui CH, Gaynon PS, Boyett JM, Chessells JM, Baruchel A, Kamps W et al. Outcome of treatment in childhood acute lymphoblastic leukaemia with rearrangements of the 11q23 chromosomal region. Lancet 2002; 359: 1909–1915.
Inukai T, Zhang X, Goto M, Hirose K, Uno K, Akahane K et al. Resistance of infant leukemia with MLL rearrangement to tumor necrosis factor-related apoptosis-inducing ligand: a possible mechanism for poor sensitivity to antitumor immunity. Leukemia 2006; 20: 2119–2129.
Look AT . Oncogenic transcription factors in the human acute leukemias. Science 1997; 278: 1059–1064.
Hirose K, Inukai T, Kikuchi J, Furukawa Y, Ikawa T, Kawamoto H et al. Aberrant induction of LMO2 by the E2A-HLF chimeric transcription factor and its implication in leukemogenesis of B-precursor ALL with t(17;19). Blood 2010; 116: 962–970.
Akahane K, Inukai T, Inaba T, Kurosawa H, Look AT, Kiyokawa N et al. Specific induction of CD33 expression by E2A-HLF: the first evidence for aberrant myeloid antigen expression in ALL by a fusion transcription factor. Leukemia 2010; 24: 865–869.
Inukai T, Hirose K, Inaba T, Kurosawa H, Hama A, Inada H et al. Hypercalcemia in childhood acute lymphoblastic leukemia: frequent implication of parathyroid hormone-related peptide and E2A-HLF from translocation 17;19. Leukemia 2007; 21: 288–296.
Kurosawa H, Goi K, Inukai T, Inaba T, Chang KS, Shinjyo T et al. Two candidate downstream target genes for E2A-HLF. Blood 1999; 93: 321–332.
Matsunaga T, Inaba T, Matsui H, Okuya M, Miyajima A, Inukai T et al. Regulation of annexin II by cytokine-initiated signaling pathways and E2A-HLF oncoprotein. Blood 2004; 103: 3185–3191.
Hunger SP . Chromosomal translocations involving the E2A gene in acute lymphoblastic leukemia: clinical features and molecular pathogenesis. Blood 1996; 87: 1211–1224.
Inaba T, Roberts WM, Shapiro LH, Jolly KM, Raimondi SC, Smith SD et al. Fusion of the leucine zipper gene HLF to the E2A gene in human acute B-lineage leukemia. Science 1992; 257: 531–534.
Hunger SP, Ohyashiki K, Toyama K, Cleary ML . Hlf a novel hepatic bZIP protein, shows altered DNA-binding properties following fusion to E2A in t(17;19) acute lymphoblastic leukemia. Genes Dev 1992; 6: 1608–1620.
Inaba T, Shapiro LH, Funabiki T, Sinclair AE, Jones BG, Ashmun RA et al. DNA-binding specificity and trans-activating potential of the leukemia-associated E2A-hepatic leukemia factor fusion protein. Mol Cell Biol 1994; 14: 3403–3413.
Hunger SP, Brown R, Cleary ML . DNA-binding and transcriptional regulatory properties of hepatic leukemia factor (HLF) and the t(17;19) acute lymphoblastic leukemia chimera E2A-HLF. Mol Cell Biol 1994; 14: 5986–5996.
Inukai T, Inaba T, Yoshihara T, Look AT . Cell transformation mediated by homodimeric E2A-HLF transcription factors. Mol Cell Biol 1997; 17: 1417–1424.
Inaba T, Inukai T, Yoshihara T, Seyschab H, Ashmun RA, Canman CE et al. Reversal of apoptosis by the leukaemia-associated E2A-HLF chimaeric transcription factor. Nature 1996; 382: 541–544.
Inukai T, Inaba T, Ikushima S, Look AT . The AD1 and AD2 transactivation domains of E2A are essential for the antiapoptotic activity of the chimeric oncoprotein E2A-HLF. Mol Cell Biol 1998; 18: 6035–6043.
Inukai T, Inaba T, Dang J, Kuribara R, Ozawa K, Miyajima A et al. TEF, an antiapoptotic bZIP transcription factor related to the oncogenic E2A-HLF chimera, inhibits cell growth by down-regulating expression of the common beta chain of cytokine receptors. Blood 2005; 105: 4437–4444.
Altura RA, Inukai T, Ashmun RA, Zambetti GP, Roussel MF, Look AT . The chimeric E2A-HLF transcription factor abrogates p53-induced apoptosis in myeloid leukemia cells. Blood 1998; 92: 1397–1405.
Devaraj PE, Foroni L, Sekhar M, Butler T, Wright F, Mehta A et al. E2A/HLF fusion cDNAs and the use of RT-PCR for the detection of minimal residual disease in t(17;19)(q22;p13) acute lymphoblastic leukemia. Leukemia 1994; 8: 1131–1138.
Akahane K, Inukai T, Zhang X, Hirose K, Kuroda I, Honna H et al. Resistance of T-cell acute lymphoblastic leukemia to tumor necrosis factor--related apoptosis-inducing ligand-mediated apoptosis. Exp Hematol 2010; 38: 885–895.
Yoshida T, Maeda A, Tani N, Sakai T . Promoter structure and transcription initiation sites of the human death receptor 5/TRAIL-R2 gene. FEBS Lett 2001; 507: 381–385.
Wu GS, Burns TF, McDonald ER, Jiang W, Meng R, Krantz ID et al. KILLER/DR5 is a DNA damage-inducible p53-regulated death receptor gene. Nat Genet 1997; 17: 141–143.
Liu X, Yue P, Khuri FR, Sun SY . p53 upregulates death receptor 4 expression through an intronic p53 binding site. Cancer Res 2004; 64: 5078–5083.
Shetty S, Graham BA, Brown JG, Hu X, Vegh-Yarema N, Harding G et al. Transcription factor NF-kappaB differentially regulates death receptor 5 expression involving histone deacetylase 1. Mol Cell Biol 2005; 25: 5404–5416.
Mendoza FJ, Ishdorj G, Hu X, Gibson SB . Death receptor-4(DR4) expression is regulated by transcription factor NF-kappaB in response to etoposide treatment. Apoptosis 2008; 13: 756–770.
Guan B, Yue P, Lotan R, Sun SY . Evidence that the human death receptor 4 is regulated by activator protein 1. Oncogene 2002; 21: 3121–3129.
Takeda K, Smyth MJ, Cretney E, Hayakawa Y, Kayagaki N, Yagita H et al. Critical role for tumor necrosis factor-related apoptosis-inducing ligand in immune surveillance against tumor development. J Exp Med 2002; 195: 161–169.
Baader E, Toloczko A, Fuchs U, Schmid I, Beltinger C, Ehrhardt H et al. Tumor necrosis factor-related apoptosis-inducing ligand-mediated proliferation of tumor cells with receptor-proximal apoptosis defects. Cancer Res 2005; 65: 7888–7895.
Belyanskaya LL, Ziogas A, Hopkins-Donaldson S, Kurtz S, Simon HU, Stahel R et al. TRAIL-induced survival and proliferation of SCLC cells is mediated by ERK and dependent on TRAIL-R2/DR5 expression in the absence of caspase-8. Lung Cancer 2008; 60: 355–365.
Ruiz-Ruiz C, Muñoz-Pinedo C, López-Rivas A . Interferon-gamma treatment elevates caspase-8 expression and sensitizes human breast tumor cells to a death receptor-induced mitochondria-operated apoptotic program. Cancer Res 2000; 60: 5673–5680.
Kanetaka Y, Hayashida M, Hoshika A, Yanase N, Mizuguchi J . Interferon-alpha induces transient upregulation of c-FLIP through NF-kappaB activation. Exp Cell Res 2008; 314: 246–254.
Tolcher AW, Mita M, Meropol NJ, von Mehren M, Patnaik A, Padavic K et al. Phase I pharmacokinetic and biologic correlative study of mapatumumab, a fully human monoclonal antibody with agonist activity to tumor necrosis factor-related apoptosis-inducing ligand receptor-1. J Clin Oncol 2007; 25: 1390–1395.
Plummer R, Attard G, Pacey S, Li L, Razak A, Perrett R et al. Phase 1 and pharmacokinetic study of lexatumumab in patients with advanced cancers. Clin Cancer Res 2007; 13: 6187–6194.
We thank H Goto (Department of Pediatrics, Yokohama City University, Japan) for YCUB-2 cell line, F Rauscher III (Wister Institute, PA) for the pMT-CB6+ expression vector and T Sakai (Department of Molecular-Targeting Cancer Prevention, Kyoto Prefectural University of Medicine, Japan) for the vectors for DR5 reporter assay. This work was supported in part by research grants from the Ministry of Education, Science and Culture in Japan.
XZ performed experiments and analyzed the data; TI designed the project, analyzed the data and wrote the manuscript; KH, KA, IK, HH, KK, KG, performed experiments; KN, MK, ME provided critical tools for analysis; HY, HK, HH provided critical tools for analysis and wrote the manuscript; ATL designed and supervised the project; HH provided critical tools for analysis; TI designed the project and wrote the manuscript; SN supervised the project; KS supervised the project and wrote the manuscript.
The authors declare no conflict of interest.
Supplementary Information accompanies the paper on the Leukemia website
About this article
Cite this article
Zhang, X., Inukai, T., Hirose, K. et al. Oncogenic fusion E2A-HLF sensitizes t(17;19)-positive acute lymphoblastic leukemia to TRAIL-mediated apoptosis by upregulating the expression of death receptors. Leukemia 26, 2483–2493 (2012). https://doi.org/10.1038/leu.2012.139
- graft-versus-leukemia effect
- acute lymphoblastic leukemia
- TNF-related apoptosis-inducing ligand
Leukemia & Lymphoma (2019)
Eya2 is critical for the E2A‑HLF‑mediated immortalization of mouse hematopoietic stem/progenitor cells
International Journal of Oncology (2019)
Successful preemptive donor lymphocyte infusions from a haploidentical donor in a boy with E2A-HLF-positive ALL
Leukemia & Lymphoma (2018)
TNF-related apoptosis-inducing ligand (TRAIL) as the potential therapeutic target in hematological malignancies
Biomedicine & Pharmacotherapy (2018)
NGS-based methylation profiling differentiates TCF3-HLF and TCF3-PBX1 positive B-cell acute lymphoblastic leukemia