Spontaneous and Fas-induced apoptosis of low-grade MDS erythroid precursors involves the endoplasmic reticulum

Article metrics


Spontaneous apoptosis of bone marrow erythroid precursors accounts for the anemia that characterizes most low-grade myelodysplastic syndromes (MDS). We have shown that death of these precursors involved the Fas-dependent activation of caspase-8. To explore the pathway leading from caspase-8 activation to apoptosis, we transduced MDS bone marrow CD34+ cells with a lentivirus encoding wild-type (WT) or endoplasmic reticulum (ER)-targeted Bcl-2 protein before inducing their erythroid differentiation. Both WT-Bcl-2 and ER-targeted Bcl-2 prevented spontaneous and Fas-dependent apoptosis in MDS erythroid precursors. ER-targeted Bcl-2 inhibited mitochondrial membrane depolarization and cytochrome c release in MDS erythroid precursors undergoing apoptosis, indicating a role for the ER in the death pathway, upstream of the mitochondria. MDS erythroid precursors demonstrated elevated ER Ca2+ stores and these stores remained unaffected by ER-targeted Bcl-2. The ER-associated protein Bcl-2-associated protein (BAP) 31 was cleaved by caspase-8 in MDS erythroid precursors undergoing apoptosis. The protective effect of ER-targeted Bcl-2 toward spontaneous and Fas-induced apoptosis correlated with inhibition of BAP31 cleavage. A protective effect of erythropoietin against Fas-induced BAP31 cleavage and apoptosis was observed. We propose that apoptosis of MDS erythroid precursors involves the ER, downstream of Fas and upstream of the mitochondria, through the cleavage of the ER-associated BAP31 protein.


Myelodysplastic syndromes (MDS) are a group of heterogeneous diseases in which a hypercellular bone marrow contrasts with peripheral blood cytopenias. The percentage of blast cells in the bone marrow, the number of cytopenias and the presence or absence of cytogenetic abnormalities distinguish low-grade from high-grade MDS.1 A characteristic feature of low-grade MDS is the excessive apoptosis of hematopoietic precursors in the bone marrow.2 Apoptotic bone marrow cells demonstrate activated caspases, together with phosphatidylserine exposure at their surface and nuclear DNA fragmentation.3 Caspase-independent apoptosis through mitochondrial release of apoptosis-inducing factor and endonuclease G and autophagy in nutrient depletion conditions could also account for death of dysplastic cells in other lineages.4, 5

We and others have explored the mechanism of apoptosis in erythroid precursors deriving from low-grade MDS CD34+ bone marrow cells by ex vivo culture.6, 7 These cells activate a so-called ‘extrinsic pathway’ to death, starting at the level of the death receptor Fas (APO-1, CD95) at the cell surface.8 Upon stimulation, this receptor recruits the adapter protein Fas-associated death domain (FADD) that in turn recruits and activates the initiator caspase-8 in the death-inducing signaling complex (DISC). The death receptor Fas and its ligand, FasL, are overexpressed at the surface of low-grade MDS erythroid precursors.6, 8, 9 An excess of soluble Fas or the ectopic expression of a dominant-negative mutant of FADD prevents caspase activation and apoptosis in low-grade MDS erythroid precursors derived from bone marrow CD34+ progenitors by liquid culture.10

How Fas is connected to downstream apoptotic events in MDS erythroid precursors remains poorly known. Depending on the cell type, caspase-8 activated at the DISC level either directly activates effector caspase and/or cleaves Bid, a BH3-only protein of the Bcl-2 family whose truncated form connects the extrinsic to an intrinsic, mitochondria-dependent, pathway to death.11 In this latter pathway, the release of cytochrome c from mitochondria, which is negatively regulated by anti-apoptotic proteins of the Bcl-2 family, leads to the assembly of the apoptosome, a cytosolic, multimolecular platform on which caspase-9 is recruited and activated to initiate a caspase cascade leading to death. An increase in mitochondrial membrane permeability (MMP) and a cytosolic release of cytochrome c have been observed in low-grade MDS bone marrow cells, suggesting that the mitochondria were involved in their apoptosis,12 but whether these events are connected to Fas is unknown.

Another cellular organelle that can play a role in apoptosis, including Fas-mediated apoptosis, is the endoplasmic reticulum (ER).13, 14 ER-mediated apoptosis involves Ca2+-dependent14, 15, 16 and -independent mechanisms.17, 18 At physiological levels, Ca2+ released from the ER is taken up by the mitochondria to stimulate oxidative phosphorylation and adenosine-5’-triphosphate (ATP) production. In some circumstances, a rapid extrusion of Ca2+ from ER stores can initiate apoptosis through the mitochondrial Ca2+ overload.14 One of the mechanisms by which Bcl-2 and Bcl-xL, which are expressed at the ER level, protect cells from apoptosis is the depletion of ER Ca2+ stores that prevents the exit of high Ca2+ levels.15, 19, 20 ER-mediated apoptosis is also associated with changes in ER proteins such as the Bcl-2-associated protein (BAP) 3117, 21 and caspase-4.18, 22 Active caspase-8 cleaves BAP31 to generate a fragment that triggers mitochondrial cell death.21 Caspase-8-induced cleavage of BAP31 could connect Fas activation to mitochondrial events,13 indicating that the ER could participate as an intermediate organelle in death receptor-induced apoptosis in some cell types.

The present study explores the role of the ER in apoptosis of low-grade MDS erythroid precursors. By contrast to Bcl-xL,23 the expression of Bcl-2 strongly decreases along erythroid differentiation in normal and MDS erythroid precursors. We transduced CD34+ bone marrow progenitors with either wild-type Bcl-2 (WT-Bcl-2) construct, whose product is expressed at both the mitochondria and the ER, or an ER-targeted construct (ER-Bcl-2),24 before inducing their erythroid differentiation in liquid medium. We show that WT-Bcl-2 and the ER-targeted construct both inhibit spontaneous and Fas-induced apoptosis in MDS erythroid precursors. We demonstrate that the ER controls MDS erythroid cell apoptosis upstream of the mitochondria. Further exploration of the Fas death pathway identifies the caspase-8-mediated cleavage of BAP31 as a characteristic event in MDS erythroid precursors undergoing spontaneous apoptosis.

Patients, materials and methods

Patients and controls

A total of 28 patients with low-grade MDS were tested, including 17 with refractory anemia (RA) or refractory cytopenia and 11 with RAEB with less than 10% blast cells, as referred to the WHO classification. All patients and volunteers were informed and gave their consent according to the recommendations of the local ethics committee. RA with ringed sideroblasts were excluded because of severe impairment of mitochondria due to iron overload.7, 12 All patients were at diagnosis or received only supportive care with blood transfusion. MDS and normal bone marrow samples from volunteers (n=10) were obtained by sternal aspiration. Normal CD34+ cells were also purified from peripheral blood cytapheresis (n=5). Due to limited cell availability, all parameters could not be analyzed in each sample.


The reagents used for culture were Iscove's modified Dulbecco's medium (IMDM), serum-free substitute BIT 9500 (Stem Cell Biotechnology, Vancouver, Canada), β-mercaptoethanol, dexamethasone, glutamine, streptomycin and penicillin (all from Sigma-Aldrich, St Louis, MO, USA). Recombinant human erythropoietin (rHuEpo) and stem cell factor (SCF) were generous gifts from Amgen (Thousand Oaks, CA, USA), insulin was purchased from Bayer (Leverkunsen, Germany) and rHu insulin-like growth factor (IGF)-1 was from Sigma-Aldrich. Lonidamine, thapsigargin, β1,1′,3,3,3′,3′-hexamethylindodicarbo-cyanine iodide (HIDC), propidium iodide (PI), carbonyl cyanide m-chlorophenylhydrazone (CCmClP), and 4,6-diamidino-2-phenyl indole (DAPI) were purchased from Sigma-Aldrich, and 7-aminoactinomycin D (7-AAD), mitotracker Red, Fura-2/acetomethyl ester (AM) were from Molecular Probes (Irvine, CA, USA). Agonist Fas antibodies (clone CH11) were from Beckman Coulter (Miami, FL, USA). Rabbit polyclonal antibodies to Bcl-2 were from Santa-Cruz Biotechnology (Santa-Cruz, CA, USA) and to calreticulin from Affinity Bioreagents (Golden, CO, USA). The mouse monoclonal antibody to cytochrome c was from BD Pharmingen (Franklin Lakes, NJ, USA). Chicken and rabbit polyclonal antibodies to BAP31 were a generous gift from Dr GC Shore (McGill University, Montreal, Quebec, Canada). The mouse monoclonal antibody to Hsc70 was from Santa-Cruz Biotechnology. The mouse monoclonal antibodies to actin, to FLAG and the human control immunoglobulin M (IgM) were from Sigma-Aldrich. Cell permeant inhibitor of caspase-8 z-Iso-Glu-Thr-Asp (IETD)-fmk was from Calbiochem (Darmstadt, Germany).

Cell culture and stimulation

For erythroid cell expansion, CD34+ cells were isolated from the mononuclear cell fraction of bone marrow samples or cytapheresis products using the MidiMacs system (Miltenyi Biotec, Bergisch Gladbach, Germany). CD34+ cells, which purity was higher than 85%, were cultured at 0.5 × 106 per ml for 14 days in IMDM containing 20% BIT 9500, β-mercaptoethanol 10−4M, penicillin/streptomycin, sodium bicarbonate 1.9 mM, rHuEpo 1 UI/ml, SCF 50 ng/ml, IGF-1 40 ng/ml and dexamethasone 10−6M. Cells were diluted every 2 days in the same medium until day 10. From days 10 to 14, cells were switched to Epo (1 UI/ml) and insulin (1 UI/ml) to obtain terminal erythroid differentiation.8 At day 5 of the liquid culture, cells (10 000 per ml) were seeded in methylcellulose medium containing Epo (1 UI/ml), SCF (50 ng/ml), IL-3 (0.3 UI/ml), IL-6 (10 ng/ml) and granulocyte colony-stimulating factor (5 ng/ml). Erythroid progenitors of burst forming unit-erythroid (BFU-E) type were counted at day 10 of the semisolid culture. In some experiments, erythroid cells (1 × 106 per ml) harvested between days 10 and 14 of the culture were stimulated with 500 ng/ml Fas agonist antibody CH11 or control IgM in the presence or absence of z-IETD-fmk at indicated times. As a control of ER targeting, hematopoietic cell line U937 was used and treated with 5 μM thapsigargin for 16 h as an inducer of ER stress-dependent apoptosis.

Lentiviral constructs and cell transduction

Flag-tagged human WT-Bcl-2 and flag-tagged human Bcl-2 selectively localized to the ER by exchanging the COOH-terminal transmembrane sequence of Bcl-2 for the transmembrane domain of the ER protein cytochrome b5 (ER-Bcl-2) were kindly provided by Dr VM Dixit (Genentech, San Francisco, CA, USA). They were cloned in the pTripΔU3EF1α lentiviral vector upstream an IRESECMV-green fluorescent protein (GFP) cassette. Infectious vector particles were produced in 293T cells by transient co-transfection with the p8.91- and the VSV-G-encoding plasmids. The empty vector devoid of insert was used as control. Lentivirus transduction was performed twice at days 0 and 1. Due to the limited number of purified CD34+ cells, five MDS and three control samples were transduced with WT-Bcl-2 and control vectors only.

Flow cytometry

Transduction efficacy was measured as the percentage of GFP-positive cells (FL1). Bcl-2 expression was followed by intracellular fluorescence of the PE-Bcl-2 monoclonal antibody (FL2). For cell-cycle analysis, cells were fixed in 1% paraformaldehyde, permeabilized with 70% ethanol for 1 h at −20 °C and then labeled with 1 mg/ml PI in the presence of 20 μl RNAse at 10 mg/ml for 30 min at room temperature. Cell-cycle-oriented histogram analysis was performed using the MultiCycle software (De Novo, Los Angeles, CA, USA). Phosphatidylserine exposure was measured by annexin V binding. Cells were incubated with 1 μg/ml PE-coupled annexin V (FL2) and 2 μl per 104 cells viability dye 7-AAD for 15 min at 37 °C in HEPES 50 mM (pH 7.4), NaCl 140 mM, CaCl2 1 mM buffer. AnnexinV+/7-AAD cells were referred as apoptotic cells. Mitochondrial apoptosis was quantified by measuring the variation of mitochondrial transmembrane potential (ΔΨm) that reflects the MMP. Briefly, cells were labeled with 5 nM of the fluorescent dye HIDC that incorporated into intact mitochondrial membrane and 2.5 μg/ml of the viability dye PI, for 15 min at 37 °C and diluted in phosphate-buffered saline (PBS; pH 7.40) before analysis. Increase in MMP is proportional to the decrease in HIDC fluorescence emission (FL4). A positive control was performed for each sample by preincubating cells with the respiratory chain decoupling agent CCmClP. HIDClow/PI cells were referred as cells with MMP. When cells were transduced with a lentivirus, apoptosis markers were always quantified in the GFP-positive population (FL1). Analyses were performed on a FC500 apparatus (Beckman Coulter).


For immunofluorescence studies, erythroid precursors were spread on glass slides by cytocentrifugation. Cytochrome c was detected as followed: cells were fixed in 4% paraformaldehyde and permeabilized in 0.1% saponin containing PBS. Cells were stained with a mouse monoclonal anti-cytochrome c antibody, followed by Alexa 568 coupled goat anti-mouse secondary antibody. Results were expressed as percentage of cells with diffuse cytochrome c labeling among 100 cells for each condition. WT-Bcl-2 or ER-Bcl-2 was detected with the monoclonal anti-FLAG antibody. Immune complexes were visualized using cyanine 5-conjugated goat anti-mouse antibody (Jackson ImmunoResearch, West Grove, PA, USA). Mitochondria were stained with Mitotracker Red. Calreticulin, as a marker of ER was stained with a rabbit polyclonal antibody and revealed by an Alexa 568-conjugated goat anti-rabbit antibody (Molecular Probes). Nuclei were colored with DAPI. Fluorescence images were obtained on a Leica DMB microscope and analyzed using the Metamorph software (Molecular Devicer Corp, Dowington, PA, USA).

Caspase activities

Caspase-3 like activities were determined using a fluorometric assay. Cells were incubated in lysing buffer (150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 0.1% SDS, 1% Nonidet P-40, 0.5% sodium deoxycholate) for 30 min at 4 °C and clarified by centrifugation (10 000 g, 20 min, 4 °C). Proteins (30 μg) were incubated in 100 mM HEPES (pH 7.0), 1 mM ETDA, 0.1% CHAPS, 10% glycerol, 20 mM dithiothreitol, in the presence of 100 μM fluorogenic peptide substrate the Ac-Asp-Glu-Val-Asp (DEVD)-7-amino-4-methylcoumarin (AMC) for caspase-3 and related enzymes (Calbiochem). AMC released from the substrate was excited at 380 nm to measure emission at 460 nm. Fluorescence was monitored continuously at 37 °C for 30 min in a dual luminescence fluorometer MicroTek OS (Bio-Tek Kontron Instruments, Winooski, VT, USA). Caspase activities were determined as initial velocities expressed as relative intensity of fluorescence per minute per milligram.

Spectrofluorimetry for Ca2+ measurement

Endoplasmic reticulum (ER) and cellular Ca2+ concentrations were measured with the use of the fluorescent probe Fura-2/AM. Cells (1 × 106 cells per ml) were loaded with 1 μM Fura-2/AM at 37 °C in the dark for 20 min, washed and resuspended in saline buffer (140 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 11 mM glucose, 10 mM HEPES, pH 7.4). EGTA (10 mM) was added as an extracellular Ca2+ chelator immediately before stimulation with 100 nM thapsigargin. Dual excitation alternating at 340 mm and 380 nm was provided by a Cary Eclipse spectrofluorimeter (Varian, Les Ulis, France), equipped with two excitation monochromators and emission was measured at 510 nm. Fluorescence ratio R=F380/F340 was recorded every 3 s for around 5 min at 37 °C. Results were expressed as the rate of Ca2+ exit that is the initial slope (first derivative) of calcium rise upon thapsigargin treatment.

Western blot

Cells at days 10–14 were solubilized in HEPES 50 mM (pH 7.4), NaCl 150 mM, Triton X-100 0.5% containing protease inhibitors for 20 min on ice and then lysed in Laemmli buffer. Proteins were analysed by SDS/polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane for immunoblot analysis. Membrane were saturated in 5% low fat milk for 2 h at room temperature and incubated with 1 μg/ml primary monoclonal antibody at 4 °C overnight and with horseradish peroxidase-coupled secondary antibody. Specific labeling was revealed by ECL chemiluminescence kit (General Electric Healthcare, Bucks, UK).

Statistical analysis

Values are expressed as means±s.e.m. Continuous variables were compared using the Student's t-test and P-values <0.05 were considered statistically significant.


Ectopic expression of WT-Bcl-2 and ER-Bcl-2 in normal and MDS erythroid precursors

To determine whether the ER was involved in apoptosis of low-grade MDS erythroid precursors, we transduced MDS or control bone marrow CD34+ cells with a lentiviral vector encoding WT-Bcl-2 or ER-Bcl-2. The corresponding c-DNAs, fused to a FLAG tag, were under the control of EF1α promoter. The vectors also encoded the GFP under the control of an IRES. The percentage of GFP-expressing cells 7 days after transduction ranged from 65 to 91%. An anti-FLAG antibody detected a 28 kDa (WT-Bcl-2) or a 30 kDa (ER-Bcl-2) band in erythroid cells obtained after 10 days of liquid culture of transduced CD34+ cells. Endogenous Bcl-2 remained equally expressed in all transduced cells, whatever the vector used (Figure 1a, left panel). We confirmed the overexpression of Bcl-2 in WT- and ER-Bcl-2-transduced cells by flow cytometry. Transduced cells exhibited a sixfold increase in the ratio of mean fluorescence intensity (RFI) as compared to those transduced with the empty vector (Figure 1a, right). In U937 cells transduced with these constructs, WT-Bcl-2 colocalized with both the mitochondrial fluorescent Mitotracker red dye (Figure 1b, upper panel) and the ER protein calreticulin (Figure 1b, lower panel) whereas ER-Bcl-2 colocalized only with calreticulin (Figure 1b). When expressed in erythroid precursors, both wild-type and ER-targeted proteins prevented apoptosis (annexinV+/7-AAD cells) induced by the ER-targeting agent thapsigargin, an inhibitor of sarcoplasmic ER Ca2+-ATPases (SERCA) family of proteins, whereas apoptosis induced by the mitochondria-targeting agent lonidamine was prevented only by WT-Bcl-2 (Figure 1c). These observations confirmed that ER-targeted Bcl-2 was expressed at the ER level and specifically inhibited apoptotic pathways involving the ER.

Figure 1

Ectopic expression of Bcl-2 in erythroid precursors. (a) CD34+ progenitors were transduced with an empty lentivirus (vector) or a virus encoding wild-type (WT)-Bcl-2 or endoplasmic reticulum (ER)-targeted Bcl-2, then cultured for 10 days to generate erythroid precursors. (Left) Immunoblot analysis with anti-Bcl-2 or anti-FLAG antibodies. Actin: loading control. *: Endogenous Bcl-2. Right. Flow cytometry analysis with a PE-coupled anti-Bcl-2 antibody. RFI: ratio of mean fluorescence intensity of Bcl-2-PE to control immunoglobulin G (IgG)-PE. (b) Immunolocalization of ER- and WT-Bcl-2 in transduced U937 cells. 4,6-diamidino-2-phenyl indole (DAPI) was used to stain the nuclei, an anti-calreticulin antibody and mitotracker to stain the ER and the mitochondria, respectively, and an anti-FLAG to detect transduced Bcl-2. Magnification × 40. (c) Functional specificity of ER- and WT-Bcl-2. Erythroid precursors obtained as in (a), were either left untreated or treated for 24 h with 5 μM thapsigargin or for 2 h with 200 μM lonidamine before measuring apoptosis as annexinV+/7-aminoactinomycin D (7-AAD) cells in green fluorescent protein (GFP)+ cells by flow cytometry. White bars: empty vector, black bars: WT-Bcl-2 and grey bars: ER-Bcl-2. Results: mean±s.e.m. of three independent experiments.

ER-targeted Bcl-2 inhibits spontaneous and Fas-dependent apoptosis of MDS erythroid precursors

Apoptosis markers were analyzed in GFP+ erythroid precursors. For example, at day 14 of liquid culture, the basal level of annexin V+/7-AAD cells in normal samples (n=7) was 9.8±3.5%. Expression of WT- or ER-targeted Bcl-2 did not affect this basal level in normal samples. A spontaneous increase in the percentage of annexin V+/7-AAD cells was observed in MDS samples (n=18, 16.8±6.4%) and was completely suppressed by both constructs (P<0.05; Figure 2a). WT- and ER-targeted Bcl-2 also significantly decreased the Ac-DEVD-AMC cleavage activity, suggesting that caspase-3 activation in MDS erythroid precursors (n=4; P<0.05), which was not observed in normal samples (n=4; Figure 2b) was inhibited.

Figure 2

Effects of wild-type (WT)- and endoplasmic reticulum (ER)-Bcl-2 on spontaneous and Fas-induced apoptosis in myelodysplastic syndromes (MDS) erythroid precursors. (a) Normal controls (n=7) or low-grade MDS (n=18) CD34+ cells were transduced with the empty vector (white), the WT-Bcl-2 (black) or the ER-targeted (grey) Bcl-2 vectors and cultured for 14 days. (Left panel) Representative flow cytometry histograms for one control and one patient are shown. Percentages indicate the proportion of green fluorescent protein (GFP)+/annexin V+/7-aminoactinomycin D (7-AAD) cells. Right panel: apoptosis was quantified as the mean percentage (±s.e.m.) of GFP+/annexin V+/7-AAD cells (*P<0.05). (b) Ac-Asp-Glu-Val-Asp (DEVD)-7-amino-4-methylcoumarin (AMC) cleavage activity, suggesting activation of caspase-3, was measured in lysates of erythroid precursors obtained as above (control samples=4; MDS samples=4). Results: relative intensities of fluorescence per minute per milligram representing caspase activities are expressed as means±s.e.m. (c) MDS cells (n=7) transduced with the indicated constructs were stimulated at day 10 of the culture with 500 ng/ml CH11 antibody (hatched bars) or a control immunoglobulin M (IgM) (open bars) for 16 h. Results: mean percentages (±s.e.m.) of GFP+/annexin V+/7-AAD cells (*P<0.05).

Ectopic expression of WT- and ER-Bcl-2 in normal and MDS CD34+ bone marrow cells did not significantly affect cell growth (Supplementary Figure 1A). The quantification of GFP+ cells at days 7, 10 and 14 demonstrated that WT-Bcl-2- or ER-Bcl-2-transduced cells had no growth advantage over nontransduced cells (data not shown). Erythroid progenitors of BFU-E type were quantified at day 5 of the liquid culture as previously described.8 The ratio of BFU-E in cultures transduced with WT-Bcl-2 over empty vector was equivalent in MDS (n=4) and control samples (n=4) (Supplementary Figure 1B). To understand why Bcl-2 expression does not increase cell proliferation although it prevents cell death, we performed cell-cycle analysis of erythroid cells at day 7 of the culture. The percentage of S-phase cells decreased in erythroid cultures transduced with WT-Bcl-2 compared to the control vector, indicating that Bcl-2 delayed cell-cycle entry in erythroid cells, as observed in transgenic mice and other tissue culture models.25, 26 In agreement, WT- and ER-Bcl-2 delayed the differentiation of both normal and MDS erythroid precursors as demonstrated by lower expression of glycophorin A (Supplementary Figure 1D) and increased percentage of proerythroblast and basophilic erythroblast cytological analysis (Supplementary Figure 1E). Altogether, these results demonstrate the pure anti-apoptotic effect of Bcl-2 that increases erythroid cell survival but does not induce cell proliferation.

We have previously shown that MDS erythroid precursors overexpressed Fas receptor and were responsive to Fas-induced cell death, even in the presence of Epo,8, 11 whereas normal erythroid precursors were responsive to Fas-dependent cell death in the absence of Epo.27 Here, we show that ER-targeted Bcl-2 is as efficient as WT-Bcl-2 to prevent apoptosis induced by Fas engagement with an anti-Fas agonistic antibody (CH11) in MDS erythroid precursors cultured with Epo (Figure 2c). Thus, this first series of experiments suggests an important role for the ER in apoptosis of MDS erythroid precursors.

ER-Bcl-2 inhibits the mitochondrial pathway to apoptosis

We then observed that both transgenes significantly prevented the increase in HIDClow/PI MDS erythroid precursors at day 14 of the liquid culture (P<0.05; Figure 3a), indicating that ER-Bcl-2 was as efficient as WT-Bcl-2 in preventing spontaneous MMP. ER- and WT-Bcl-2 also prevented the increase of MMP induced by Fas stimulation in MDS erythroid precursors (Figure 3b). The percentage of cells with cytosolic cytochrome c, studied by indirect immunofluorescence on cytospins, was decreased in erythroid precursors obtained from MDS CD34+ cells transduced with either ER- or WT-Bcl-2 compared to those transduced with the empty vector (P<0.05; Figure 3c). We conclude from these experiments that the ER might play a role in the spontaneous and Fas-involving apoptosis of MDS erythroid precursors, upstream of the mitochondria.

Figure 3

Effects of wild-type (WT)- and endoplasmic reticulum (ER)-Bcl-2 on the mitochondrial membrane permeabilization (MMP) and cytochrome c release. (a) MMP was detected by flow cytometry (green fluorescent protein (GFP)+/hexamethylindodicarbo-cyanine iodide (HIDC)low/propidium iodide (PI) cells) in normal controls (n=7) and myelodysplastic syndromes (MDS, n=18) erythroid precursors transduced as in Figure 2a (empty vector: white; WT-Bcl-2: black; ER-Bcl-2: grey) collected at day 14. (Left panel) Representative flow cytometry histograms for one control and one patient. Percentages indicate the proportion of GFP+/HIDClow/PI cells. Right panel indicates the global results as mean percentages±s.e.m. (b) MDS cells (n=7) transduced with the indicated constructs were stimulated at day 10 of the culture with 500 ng/ml CH11 antibody (hatched bars) or a control immunoglobulin M (IgM, open bars) for 16 h. Results: mean percentages (±s.e.m.) of GFP+/HIDClow/PI cells (*P<0.05). (c) Indirect immunofluorescence analysis of cytochrome c was performed at day 10 of the culture of normal control (n=5) or MDS (n=6) samples transduced as indicated. Results: percentages of cells with a diffuse cytoplasmic cytochrome c staining pattern. Upper panel: one representative immunofluorescence microphotograph from a normal and a MDS samples transduced with the empty vector. Lower panel: Percentages of cells with a diffuse cytoplasmic cytochrome c staining pattern. Each symbol represents one individual sample.

ER Ca2+ stores are increased in MDS erythroid precursors

To further understand the importance of the ER in the spontaneous apoptosis of MDS erythroid precursors, we first explored the Ca2+ storage and homeostasis in these cells. Normal and MDS erythroid precursors at day 10 of the culture were loaded with the Ca2+ dye Fura-2/AM and stimulated with the inhibitor of SERCA pumps, thapsigargin. Experiments were performed in the presence of EGTA as a chelator of extracellular Ca2+. The rise of Ca2+ (F380/F340) induced by thapsigargin reflected ER Ca2+ pools (Figure 4a). Treatment of Fura-2/AM-loaded cells with thapsigargin indicated that ER Ca2+ stores were increased in MDS compared to normal samples (Figure 4a, left panel). Ca2+ exit from the ER, represented by the initial slope (first derivative) of the Ca2+ rise upon thapsigargin exposure, occurred much faster in MDS (n=5) than in normal (n=4) erythroid precursors (P<0.05; Figure 4a, right panel).

Figure 4

Endoplasmic reticulum (ER) Ca2+ stores in control and myelodysplastic syndromes (MDS) erythroid precursors. (a) Normal control (gray line) and MDS (black line) erythroid progenitors (106 cells) at day 10 of in vitro culture were loaded with Fura-2/acetomethyl ester (AM) for 20 min and the fluorescence ratio R=F380/F340 was measured at 510 nm every 3 s for indicated times. ER Ca2+ stores were assessed by measuring the fluorescence ratio after addition of 100 nM thapsigargin in the presence of 10 mM EGTA (arrow). A representative experiment is shown (left panel). The mean rate of Ca2+ exit (±s.e.m.; normal control samples=4; MDS samples=5), corresponding to the initial slope (primitive derivate) of the thapsigargin-induced Ca2+ change is plotted as a bar graph (*P<0.05, right panel). (b) ER Ca2+ stores were quantified in green fluorescent protein (GFP)+ normal control (n=4) and MDS (n=4) erythroid precursors sorted at day 10 of the culture. Cells had been transduced as in Figure 2a (empty vector: white bars; wild-type (WT)-Bcl-2: black bars; ER-Bcl-2: grey bars). Results: mean rates±s.e.m. of ER Ca2+ exit (*P<0.05).

Bcl-2 was proposed to reduce the amount of Ca2+ releasable from the ER by increasing ER Ca2+ leak, thus preventing mitochondrial Ca2+ overload and cell death in response to various stimuli.15, 19, 20 A homogeneous population of GFP+ erythroid precursors was selected by cell sorting, loaded with Fura-2/AM and stimulated with thapsigargin in the presence of EGTA. As expected, expression of WT- and ER-Bcl-2 in normal erythroid precursors significantly decreased the ER Ca2+ stores, as demonstrated by exposure to thapsigargin (n=4; Figure 4b; P<0.05). Surprisingly enough, such an effect was not observed in MDS erythroid cells, that is neither WT- nor ER-Bcl-2 modified ER Ca2+ stores measured by the efflux provoked by exposure to thapsigargin (n=4; Figure 4b). Altogether, these experiments demonstrate an increase in ER Ca2+ stores in MDS compared to normal erythroid precursors. They also indicate that the protective effect of ER-targeted Bcl-2 toward spontaneous apoptosis of MDS erythroid precursors cannot be explained by a Bcl-2-induced leak of ER Ca2+ stores.

BAP31 protein is cleaved in MDS erythroid precursors undergoing apoptosis

To further explore the importance of the ER in MDS erythroid precursor apoptosis, we examined the expression of BAP31, a protein whose cleavage was specifically associated with ER-mediated apoptosis.13, 17 Immunoblot experiments detected BAP31 in U937 cells as well as in MDS (n=10) and control (n=5) erythroid precursors. Exposure of U937 cells to thapsigargin induced the appearance of a p20 BAP31 fragment (Figure 5a). A similar p20 BAP31 fragment was identified in all the tested MDS samples, suggesting cleavage of the protein in a fraction of the cells (Figure 5a). Both WT- and ER-targeted Bcl-2 decreased apoptosis of MDS erythroid precursors and the associated cleavage of BAP31 into a p20 fragment (Figure 5b). These experiments indicated that cleavage of the ER-associated BAP31 protein might be related to apoptosis occurring spontaneously in MDS erythroid precursors.

Figure 5

: Endoplasmic reticulum (ER)-Bcl-2 inhibits the spontaneous cleavage of BAP31. Cell lysates from normal control and myelodysplastic syndromes (MDS) erythroid cells were obtained at day 10 for immunoblot analysis of BAP31. (a) U937 cells were stimulated or not with thapsigargin (5 μM for 16 h). Hsc70: loading control. One representative (a unique gel) of three separated experiments is shown. Apoptosis was quantified by flow cytometry (percentage of annexin V+/7-aminoactinomycin D (7-AAD) cells). (b) Normal control and MDS erythroid precursors were collected 10 days after transduction of CD34+ cells as in Figure 2a for immunoblot analysis of BAP31. Apoptosis was measured as above in sorted green fluorescent protein (GFP)+ cells (upper panel). Ratio of p20 to Hsc70 expression measured by densitometry (lower panel).

To further demonstrate the link between apoptosis and BAP31 cleavage in these cells, we inhibited caspase-8 activity with increasing concentrations of z-IETD-fmk (Figure 6a). Z-IETD-fmk inhibited both BAP31 cleavage into a p20 fragment and MDS cell apoptosis. To investigate whether Fas engagement could trigger BAP31 cleavage in MDS erythroid precursors, we exposed these cells to 500 ng/ml Fas agonistic antibody CH11 for 16 h (Figure 6b). CH11 enhanced the generation of the BAP31 cleavage fragment p20 in MDS erythroid precursors, which was prevented by z-IETD-fmk (Figure 6b). At the end, we checked in four MDS samples that all the steps of the described Fas-caspase-8-BAP31-mitochondria pathway could be identified and modulated by ER-targeted Bcl-2 in CD34+-derived erythroid cells.

Figure 6

Inhibition of Bcl-2-associated protein (BAP) 31 cleavage by caspase-8 inhibitor and erythropoietin (Epo). (a) z-IETD-fmk was added at indicated concentrations for 16 h to myelodysplastic syndromes (MDS) erythroid precursors at day 10 of the culture before immunoblot analysis of BAP31 (Hsc70: loading control) and apoptosis measurement (percentage of annexin V+/7-aminoactinomycin D (7-AAD) cells) (b) MDS erythroid precursors at day 10 of the culture were stimulated with 500 ng/ml CH11 for 16 h with or without 20 μM z-IETD-fmk before immunoblot analysis of BAP31 (Hsc70: loading control) and apoptosis measurement (percentage of annexin V+/7-AAD cells). In both panels, one representative of three independent experiments is shown. (c) MDS erythroid precursors at day 10 were washed, and cultured for 16 h with fresh medium containing 0, 1 or 10 UI/ml Epo, and stimulated or not with 500 ng/ml CH11 before immunoblot analysis of BAP31 (Hsc70: loading control) and apoptosis measurement (percentage of annexin V+/7-AAD cells) was performed.

To test the protective effect of rHuEpo against spontaneous and Fas-induced apoptosis in MDS, we added increasing amounts of Epo to the culture of MDS precursors during the stimulation by CH11 for 16 h (Figure 6c). Epo addition partially inhibited spontaneous and CH11-induced BAP31 cleavage, as well as cell apoptosis. Altogether, these experiments indicate that both spontaneous and Fas-stimulated MDS erythroid cell apoptosis are associated with a caspase-8-mediated BAP31 cleavage at the ER level, and can be inhibited by increasing doses of Epo in the culture.


Apoptosis of erythroid precursors is one of the pathophysiological events that account for anemia in low-grade MDS. We showed previously that the deregulated expression of the death receptor Fas and its ligand at the cell surface was at least in part responsible for apoptosis of dysplastic erythroid lineage in RA or RAEB-1.8, 10 Whether disease-associated changes in the apoptotic pathways exist downstream of Fas receptor in MDS erythroid cells remained unknown. By specifically expressing Bcl-2 at the ER level, we show here that the ER is a component of the pathway leading from Fas engagement to mitochondrial changes and apoptosis of erythroid precursors in low-grade MDS.

We also show that Ca2+ stores are spontaneously increased in the ER as compared to normal cells. Nevertheless, ER-targeted Bcl-2 does not induce ER Ca2+ leak as observed in normal cells, which suggests alterations of ER Ca2+ homeostasis in these cells. The ER-associated protein BAP31 is shown to be cleaved in MDS erythroid precursors, downstream of caspase-8 activation. These results indicate a central role for ER in MDS erythroid cell apoptosis, downstream of Fas and upstream of the mitochondria.

Several nonlethal functions of caspases have been identified, including a role in red cell differentiation.28, 29, 30 Caspase activation is increased in MDS compared to normal erythroid precursors undergoing differentiation.3 Death of low-grade MDS erythroid precursors cannot be viewed as an amplification of the physiological caspase-3 activation that contributes to normal erythroid differentiation because this physiological pathway does not involve Fas or caspase-8.29, 30 It remains to be demonstrated whether apoptosis of MDS erythroid precursors could be the consequence of the deregulation of the negative loop that controls normal erythropoiesis through Fas and caspase-8.27

Spontaneous apoptosis of MDS erythroid precursors was associated with an increase in MMP and a cytosolic release of cytochrome c.12 The connection between Fas and the mitochondria remained to be established. The usual connection involves the caspase-8-mediated cleavage of the BH3-only protein Bid.11 We failed to reproducibly identify a cleavage of this protein in MDS erythroid cell precursors undergoing apoptosis (data not shown). Although less well defined, another connection between death receptors and the mitochondria involves the ER, which propagates the death induced by other stress signals such as Ca2+, ceramide and protein unfolding.15, 16, 31 Several proteins of the Bcl-2 family are expressed at the ER level and modulate ER-mediated cell death.15, 19, 20 By overexpressing ER-targeted Bcl-2, we inhibited apoptosis, MMP and cytochrome c release as efficiently as when WT-Bcl-2 or Bcl-xL (data not shown) were overexpressed in MDS erythroid precursors, which ascertains the implication of the ER in the control of the mitochondrial pathway of apoptosis in these cells. Whether caspase-independent death pathway exist in MDS erythroid cells as in megakaryocytes4 and could be blocked in ER-targeted Bcl-2 remains a matter of speculation.

One of the mechanisms by which Bcl-2 and related proteins modulate apoptosis at the ER level is the regulation of ER Ca2+ concentration. Bcl-2 interacts with inositol 1,4,5-triphosphate receptors (InsP3R) to provoke ER Ca2+ leak. The subsequent decrease in ER Ca2+ concentration prevents the release of large amounts of Ca2+ and subsequent apoptosis upon ER stress.32, 33 Here, we show that ER-targeted Bcl-2 decreases ER Ca2+ stores in normal erythroid precursors and protect these cells from apoptosis. Surprisingly, this effect was not identified in MDS erythroid precursors. In these latter cells, Ca2+ homeostasis appears to be impaired with enhanced Ca2+ stores in the ER and Bcl-2 does not induce any leakage of ER Ca2+. Mechanisms that could account for this observation include either altered InsP3R expression, or deregulated interaction of InsP3R with post-translationally modified Bcl-2 or with activated Bax and Bak.34, 35 In preliminary experiments using real-time reverse transcription PCR, we did not detect any change in InsP3RI, II and III transcripts in MDS erythroid precursors (data not shown), which does not rule out functional defect in the InsP3R proteins. Further study is currently done to understand which changes at the ER level account for abnormal Ca2+ storage and regulation in MDS erythroid precursors. Ca2+ disturbances may stimulate adaptative responses to protect the cells from death, for example, by triggering autophagy that ensures the turnover of macromolecules and ATP production in injured cells and can lead to cell death. This process is inhibited by ER-targeted, not by mitochondrial Bcl-2.36

At least in some cell types, ER stress leading to apoptosis can implicate the cleavage of caspase-4, and downregulation of caspase-4 expression by small-interfering RNA renders these cells resistant to ER stress-induced apoptosis.18, 37, 38 Caspase-4 activation was shown to account for spontaneous death of human plasma cells.22 Although we detected a cleavage of caspase-4 into a p35 fragment in erythroid precursors from most studied patients with MDS (Supplementary Figure 2), this cleavage was inconstantly modulated by ER-targeted Bcl-2 (data not shown), suggesting that caspase-4 cleavage might not play a central role in the spontaneous death of these cells. Consistent with our data, a previous study demonstrated that ER stress did not induce the cleavage of caspase-4 in a MDS cell line.39 Nevertheless, together with the enhanced ER Ca2+ stores, this observation argues for a deregulation of several ER functions in MDS erythroid precursors.

Another ER-associated protein linked to ER stress-induced cell death is the Bcl-2-associated protein, BAP31. We here report that the ability of ER-targeted Bcl-2 to prevent spontaneous apoptosis of MDS erythroid precursors correlates with its ability to prevent the cleavage of BAP31. Overexpression of WT-Bcl-2 was previously shown to prevent BAP31 cleavage in epithelial cells undergoing apoptosis.21 BAP31, which is part of a large ER-anchored complex that regulates the export of membrane and secreted proteins, can be cleaved by activated caspase-8.17, 21, 37 The p20 fragment generated by this cleavage promotes mitochondrial events leading to apoptosis such as cytochrome c release, and BAP31 cleavage connects Fas engagement to mitochondrial events in epithelial and breast cancer cells.13, 17 Our data demonstrate that BAP31 cleavage in MDS erythroid precursors undergoing spontaneous apoptosis is prevented by a caspase-8 inhibitor and can be amplified by exposure of these cells to an anti-Fas agonistic antibody, which correlates with enhanced apoptosis. In addition to the pro-apoptotic role of the 20 kDa fragment of BAP31, the cleavage of this protein could affect its functions at the ER level. These functions include the export of several tetraspanins that contribute to the maintenance of functional β1 integrins at the cell surface,40 thus promote erythroid cell survival.41 Interestingly, integrin adhesion to fibronectin has shown to be defective in MDS bone marrow cells.42

Erythropoietin is used in clinical practice for treatment of low-grade MDS patients, as it allows correction of anemia and transfusion independence in anemic patients (review in Hellström-Lindberg and Malcovati43), and increased survival.44 In this study, we observe a protective effect of increasing doses of Epo on the Fas-induced apoptosis of MDS precursors, as well as a decrease of the cleavage of the ER protein BAP31. This is consistent with the known anti-apoptotic effect of Epo on erythroid cells.

The spontaneous apoptosis of erythroid precursors that characterizes low-grade MDS and leads to anemia, involves the deregulated expression of death receptors and their ligands. Our data demonstrate a role for the ER, downstream of Fas, in the pathway leading to death, possibly through caspase-8-mediated cleavage of BAP31. They also demonstrate several alterations in the functions of the ER, such as deregulated Ca2+ storage and spontaneous cleavage of caspase-4. Thus, together with excessive death receptor-dependent apoptosis and genetic instability, changes in ER functions could play a role in MDS dyserythropoiesis.


  1. 1

    Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, Sanz G et al. International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood 1997; 89: 2079–2088.

  2. 2

    Raza A, Gezer S, Mundle S, Gao XZ, Alvi S, Borok R et al. Apoptosis in bone marrow biopsy samples involving stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes. Blood 1995; 86: 268–276.

  3. 3

    Bouscary D, Chen YL, Guesnu M, Picard F, Viguie F, Lacombe C et al. Activity of the caspase-3/CPP32 enzyme is increased in ‘early stage’ myelodysplastic syndromes with excessive apoptosis, but caspase inhibition does not enhance colony formation in vitro. Exp Hematol 2000; 28: 784–791.

  4. 4

    Braun T, Carvalho G, Grosjean J, Ades L, Fabre C, Boehrer S et al. Differentiating megakaryotyctes in myleodysplastic syndromes succumb to mitochondrial derangement without caspase activation. Apoptosis 2007; 12: 1101–1104.

  5. 5

    Fabre C, Carvalho G, Tasdemir E, Braun T, Ades L, Grosjean J et al. NF-kappaB inhibition sensitizes to starvation-induced cell death in high-risk myelodysplastic syndromes and acute myeloid leukemia. Oncogene 2007; 26: 4071–4083.

  6. 6

    Bouscary D, De Vos J, Guesnu M, Jondeau K, Viguier F, Melle J et al. Fas/Apo-1 (CD95) expression and apoptosis in patients with myelodysplastic syndromes. Leukemia 1997; 11: 839–845.

  7. 7

    Schmidt-Mende J, Tehranchi R, Forsblom AM, Joseph B, Christensson B, Fadeel B et al. Granulocyte colony-stimulating factor inhibits Fas-triggered apoptosis in bone marrow cells isolated from patients with refractory anemia with ringed sideroblasts. Leukemia 2001; 15: 742–751.

  8. 8

    Claessens YE, Bouscary D, Dupont J-M, Picard F, Melle J, Gisselbrecht S et al. In vitro proliferation and differentiation of erythroid progenitors from patients with myelodysplastic syndrome: evidence for Fas-dependent apoptosis. Blood 2002; 99: 1594–1601.

  9. 9

    Gersuk GM, Lee JW, Beckham CA, Anderson J, Deeg HJ . Fas (CD95) receptor and Fas-ligand expression in bone marrow cells from patients with myelodysplastic syndrome. Blood 1996; 88: 1122–1123.

  10. 10

    Claessens YE, Park S, Dubart-Kupperschmitt A, Mariot V, Garrido C, Chretien S et al. Rescue of early-stage myelodysplastic syndrome-deriving erythroid precursors by the ectopic expression of a dominant-negative form of FADD. Blood 2005; 105: 4035–4042.

  11. 11

    Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ et al. Two CD95 (APO-1/Fas) signaling pathways. EMBO J 1998; 17: 1675–1687.

  12. 12

    Tehranchi R, Fadeel B, Forsblom A-M, Christensson B, Samuelsson J, Zhivotovsky B et al. Granulocyte colony-stimulating factor inhibits spontaneous cytochrome c release and mitochondria-dependent apoptosis of myelodysplastic syndrome hematopoietic progenitors. Blood 2003; 101: 1080–1086.

  13. 13

    Nguyen M, Breckenridge DG, Ducret A, Shore GC . Caspase-resistant BAP31 inhibits fas-mediated apoptotic membrane fragmentation and release of cytochrome c from mitochondria. Mol Cell Biol 2000; 20: 6731–6740.

  14. 14

    Orrenius S, Zhivotovsky B, Nicotera P . Regulation of cell death: the calcium-apoptosis link. Nat Rev Mol Cell Biol 2003; 4: 552–565.

  15. 15

    Pinton P, Ferrari D, Rapizzi E, DiVirgilio FD, Pozzan T, Rizzuto R . The Ca(2+) concentration of the endoplasmic reticulum is a key determinant of ceramide-inducd apoptosis: significance of the molecular mechanism of Bcl-2 action. EMBO J 2001; 20: 2690–2701.

  16. 16

    Szalai G, Krisnamurthy R, Hajnoczky G . Apoptosis driven by IP(3)-linked mitochondrial calcium signals. EMBO J 1999; 18: 6349–6361.

  17. 17

    Breckenridge DG, Stojanovic M, Marcellus RC, Shore GC . Caspase cleavage product of BAP31 induces mitochondrial fission through endoplasmic reticulum calcium signals, enhancing cytochrome c release to the cytosol. J Cell Biol 2003; 160: 1115–1127.

  18. 18

    Hitomi J, Katayama T, Eguchi Y, Kudo T, Taniguchi M, Koyama Y et al. Involvement of caspase-4 in endoplasmic reticulum stress-induced apoptosis and Aβ-induced cell death. J Cell Biol 2004; 165: 347–356.

  19. 19

    Foyouzi-Youssefi R, Artnaudeau S, Borner C, Kelley WL, Tschopp J, Lew DP et al. Bcl2 decreases the free Ca2+ concentration within the endoplasmic reticulum. Proc Natl Acad Sci USA 2000; 97: 5723–5728.

  20. 20

    Vanden Abeele F, Skryma R, Shuba Y, VanCoppenolle F, Slomianny C, Roudbaraki M et al. Bcl2-dependent modulation of Ca(2+) homeostasis and store-operated channels in prostate cancer cells. Cancer Cell 2002; 1: 169–179.

  21. 21

    Ng FW, Nguyen M, Kwan T, Branton PE, Nicholson DW, Cromlish JA et al. P28BAP31, a Bcl-2/Bcl-xL, and procaspase-8-associated protein in the endoplasmic reticulum. J Cell Biol 1997; 139: 327–338.

  22. 22

    Pelletier N, Casamayor-Palleja M, De Luca K, Mondiere P, Saltel F, Jurdic P et al. The endoplasmic reticulum is a key component of the plasma cell death pathway. J Immunol 2006; 176: 1340–1347.

  23. 23

    Gregory T, Yu C, Ma A, Orkin SH, Blobel GA, Weiss MJ . GATA-1 and erythropoietin cooperate to promote erythroid cell survival by regulating bcl-xL expression. Blood 1999; 94: 87–96.

  24. 24

    Wang NS, Unkila MT, Reineks EZ, Distelhorst CW . Transient expression of wild-type or mitochondrially targeted Bcl-2 induces apoptosis, whereas transient expression of endoplasmic reticulum-targeted Bcl-2 is protective against Bax-induced cell death. J Biol Chem 2001; 276: 44117–44128.

  25. 25

    Linette GP, Roth K, Korsmeyer SJ . Cross talk between cell death and cell cycle progression: Bcl-2 regulates NFAT-mediated activation. Proc Natl Acad Sci USA 1996; 93: 9545–9552.

  26. 26

    O’Reilly LA, Huang DC, Strasser A . The cell death inhibitor Bcl-2 and its homologues influence control of cell cycle entry. EMBO J 1996; 15: 6979–6990.

  27. 27

    De Maria R, Testa U, Luchetti L, Zeuner A, Stassi G, Pelosi E et al. Apoptotic role of Fas/Fas ligand system in the regulation of erythropoiesis. Blood 1999; 93: 796–803.

  28. 28

    Launay S, Hermine O, Fontenay M, Solary E, Garrido C . Vital functions for lethal caspases. Oncogene 2005; 24: 5137–5148.

  29. 29

    Zermati Y, Garrido C, Amsellem S, Fishelson S, Bouscary D, Valensi F et al. Caspase activation is required for terminal erythroid differentiation. J Exp Med 2001; 193: 247–254.

  30. 30

    Ribeil J-A, Zermati Y, Kersual J, Dussiot M, Coulon S, Moura IC et al. Hsp70 regulates erythropoiesis by preventing caspase-3-mediated cleavage of GATA-1. Nature 2007; 445: 102–105.

  31. 31

    Xu C, Bailly-Maitre B, Reed JC . Endoplasmic reticulum stress. Cell life and death decisions. J Clin Invest 2005; 115: 2656–2664.

  32. 32

    Chen R, Valencia I, Zhong F, McColl KS, Roderick HL, Bootman MD et al. Bcl-2 functionally interacts with inositol 1,4,5-trisphosphate receptors to regulate calcium release from the ER in response to inositol 1,4,5-trisphosphate. J Cell Biol 2004; 166: 193–203.

  33. 33

    White C, Li C, Yang J, Petrenko NB, Madesh M, Thompson CB et al. The endoplasmic reticulum gateway to apoptosis by Bcl-xL modulation of the InsP3R. Nat Cell Biol 2005; 7: 1021–1028.

  34. 34

    Bassik MC, Scorrano L, Oakes SA, Pozzan T, Korsmeyer SJ . Phosphorylation of Bcl-2 regulates ER Ca2+ homeostasis and apoptosis. EMBO J 2004; 23: 1207–1216.

  35. 35

    Scorrano L, Oakes SA, Opferman JT, Cheng EH, Sorcinelli MD, Pozzan T et al. Bax and Bak regulation of endoplasmic reticulum Ca2+: a control point for apoptosis. Science 2003; 300: 135–139.

  36. 36

    Hover-Hansen M, Bastholm L, Szynlarowski P, Campanella M, Szabadkai G, Farkas T et al. Control of macroautophagy by calcium, calmodulin-dependent kinase kinase-beta and Bcl-2. Mol Cell 2007; 25: 193–205.

  37. 37

    Chandra D, Choy G, Deng X, Bhatia B, Daniel P, Tang DG . Association of active caspase 8 with the mitochondrial membrane during apoptosis: potential roles in cleaving BAP31 and caspase 3 and mediating mitochondrion-endoplasmic reticulum cross talk in etoposide-induced cell death. Mol Cell Biol 2004; 24: 6592–6607.

  38. 38

    Sanges D, Comitato A, Tammaro R, Marigo V . Apoptosis in retinal degeneration involves cross-talk between apoptosis-inducing factor (AIF) and caspase-12 and is blocked by calpain inhibitors. Proc Natl Acad Sci USA 2006; 103: 17366–17371.

  39. 39

    Jie H, Donghua H, Xinghui X, Liang G, Wenjun W, Xiaoyan H et al. Homoharringtonine-induced apoptosis of MDS cell line MUTZ-1 cells is mediated by the endoplasmic reticulum stress pathway. Leuk Lymphoma 2007; 48: 964–977.

  40. 40

    Stojanovic M, Germain M, Nguyen M, Shore GC . BAP31 and its cleavage product regulate cell surface expression of tetraspanins and integrin-mediated cell survival. J Biol Chem 2005; 280: 30018–30024.

  41. 41

    Eshghi S, Vogelezang MG, Hynes RO, Griffith LG, Lodish HF . Alpha4beta1 integrin and erythropoietin mediate temporally distinct steps in erythropoiesis: integrins in red cell development. J Cell Biol 2007; 177: 871–880.

  42. 42

    Delforge M, Raets V, Van Duppen V, Vandenberghe P, Boogaerts M . CD34+ marrow progenitors from MDS patients with high levels of intramedullary apoptosis have reduced expression of α4β1 and α5β1 integrins. Leukemia 2005; 19: 57–63.

  43. 43

    Hellström-Lindberg E, Malcovati L . Supportive care and use of hematopoietic growth factors in myelodysplastic syndromes. Semin Hematol 2008; 45: 14–22.

  44. 44

    Park S, Grabar S, Kelaidi C, Beyne-Rauzy O, Picard F, Bardet V et al. Predictive factors of response and survival in myelodysplastic syndrome treated with erythropoietin and G-CSF: the GFM experience. Blood 2008; 111: 574–582.

Download references


We thank Dr GC Shore for providing tools, Dr G Szabadkai and Dr B Papp for helpful discussions. This study was supported by grants from the Direction regionale de la Recherche Clinique, AP-HP (MUL03009), and from the Canceropole Ile-de-France. EG was the recipient of a fellowship from Inserm (poste d’accueil) and from the Fondation de France—Fondation contre la Leucémie. PM and ES groups were supported by the Ligue Nationale Contre le Cancer. EG, EF, J-CD and CP-E performed research study and analyzed data; OB-R, FD and ES recorded patients, CR, AD-K and CG reviewed the paper and contributed analytic tools; CL and PM analyzed data; ES and MF analyzed data and wrote the paper.

Author information

Correspondence to M Fontenay.

Additional information

Supplementary Information accompanies the paper on the Leukemia website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article


  • myelodysplastic syndromes
  • apoptosis
  • death receptors

Further reading