CPP32 enzyme is increased in “early stage” myelodysplastic syndromes with excessive apoptosis, but caspase inhibition does not enhance colony formation in vitro

Experimental Hematology 28 (2000) 784–791

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 Didier Bouscarya,d, Yan Lian Chenb, Martine Guesnua, Françoise Picarda, Franck Viguierc, Catherine Lacombea,d, François Dreyfusa,d, and Michaela Fontenay-Roupiea,d a Service d’Hématologie, Hôpital Cochin, AP-HP, Université René Descartes, Paris, France; bService de Pharmacologie, Hôpital Cochin, Paris, France; cLaboratoire de Cytogénétique, Hôpital de l’Hôtel Dieu, Paris, France; dUnité INSERM U363 ICGM, Hôpital Cochin, Paris, France

(Received 13 October 1999; revised 8 March 2000; accepted 17 March 2000)

Objective. Excessive apoptosis may have a role in the ineffective hematopoiesis and cytopenias observed in myelodysplastic syndromes. The goals of this study were 1) to quantify apoptosis in patients with “early stage” myelodysplasia [including patients with refractory anemia (RA), RA with ringed sideroblasts (RARS), RA with excess blasts and with less than 10% blasts (RAEB⬍10)], and in patients with “late stage” myelodysplasia [including RAEB with more than 10% blasts (RAEB⬎10), RAEB in transformation (RAEB-t), and acute myeloid leukemia secondary to myelodysplasia (LAM2)]; 2) to study the activation of the caspase-3/CPP32 enzyme, a major “effector” caspase in hematopoiesis, in patients with “early stage” myelodysplasia, and 3) to evaluate the effect of caspase inhibition on the apoptotic phenotype and clonogenicity of hematopoietic progenitors in vitro in these patients. Materials and Methods. Patients: Fifty-four patients with myelodysplastic syndromes, including 30 with “early stage” myelodysplasia and 24 with “late stage” myelodysplasia were studied. Study of apoptosis: TUNEL assay performed on bone marrow smears and/or quantification of annexin V positive bone marrow mononuclear cells by flow cytometric analysis. Caspacse-3/CPP32 activity: Quantitative measurement of caspase-3/CPP32 activity on total bone marrow mononuclear cells using a fluorogenic substrate. Effect of the caspase-inhibitor Z-VAD-FMK: 1) on the apoptotic phenotype of total bone marrow mononuclear cells and 2) on the clonogenicity of hematopoietic progenitor cells. Results. The group of 30 patients with “early stage” myelodysplasia had statistically increased apoptosis compared to the group of 24 patients with “late stage” myelodysplasia (44.1% ⫾ 4.8 vs 21.8% ⫾ 3.6; p ⫽ 0.02) using the TDT-mediated dUTP nick-end labeling (TUNEL) assay. In the group of patients with RAEB, those with MDSRAEB⬍10 had excessive apoptosis compared to those with MDSRAEB⬎10 (44.0% ⫾ 3.5% vs 29.5% ⫾ 3.6%; p ⫽ 0.042) The median caspase-3 activity in 20 “early stage” myelodysplasia patients was 19,000 U (range 3,460–41,000) and significantly increased compared to normal individuals (4,256 U, range 3,200–5,200; p ⫽ 0.032) Bone marrow mononuclear cells from 12 “early stage” MDS patients (including 11 from the 20 studied for caspase-3 activity) were incubated with or without the broad-spectrum caspase inhibitor Z-VAD-FMK. In 4 of 9 evaluable patients (44.4%) with excessive apoptosis, the number of annexin V positive cells decreased in a dose-dependent manner in the presence of Z-VAD-FMK. However, in none of these patients was caspase inhibition with Z-VAD-FMK able to enhance colony formation in vitro. Conclusion. These results confirm that a major characteristic of patients with “early stage” myelodysplasia is increased apoptosis. The results also indicate that excessive apoptosis in these patients is accompanied by increased caspase-3/CPP32 activity. However, caspase inhibition with the broad-spectrum inhibitor Z-VAD-FMK cannot improve hematopoiesis in this group of patients, even when apoptosis is attenuated. © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. Offprint requests to: Didier Bouscary, Service d’Hématologie. Hôpital Cochin, 27 rue du Faubourg St. Jacques, 75563 Paris, France; E-mail: bouscary@ cochin.inserm.fr

0301-472X/00 $–see front matter. Copyright © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(00)0 0 1 7 9 - X

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

785

Keywords: Myelodysplastic syndromes—Apoptosis—Caspases—Peptide inhibitors of caspase activation

Introduction Increased apoptosis of bone marrow (BM) cells has been described in myelodysplastic syndromes (MDS) [1–10]. The high proportion of BM cells undergoing programmed cell death (PCD) may contribute to the cytopenias that occur in MDS [3,11]. Some mechanisms for apoptosis in MDS have been proposed. The Fas/Fas-ligand (FasL) pathway probably plays a major role in apoptosis of MDS. The expression of Fas, which is a cell-surface receptor that belongs to the tumor necrosis factor receptor (TNF-R) family, and of FasL, which induces apoptosis by oligomerization of Fas molecules, is increased in myeloid, erythroid, and mononuclear cells from patients with MDS [2,12–14]. A correlation between increased expression of TNF-␣ and interferon ␥ (IFN-␥) and the degree of apoptosis has been described in MDS [11]. TNF-␣ may induce apoptosis either directly or indirectly, by induction of a functional Fas protein on CD34⫹ cells in synergy with IFN-␥ [15,16]. IFN-␥ has been reported to induce Fas expression on normal erythroid progenitor cells and to stimulate apoptosis by Fas/FasL interaction [17]. The Fas and FasL proteins are functional in hematopoietic progenitor cells and may account for autocrine and/or paracrine apoptosis, suggesting that pharmacologic blockade of this pathway may be of clinical value in these disorders [12–14]. However, the susceptibility of hematopoietic progenitor cells to apoptosis in MDS may also be explained by 1) an abnormally high ratio of pro vs antiapoptotic proteins in CD34⫹ hematopoietic cells [5,6,10], 2) defects in anti-apoptotic signal transduction pathways generated in response to hematopoietic growth factors such as erythropoietin [18,19], 3) a defective BM stroma unable to sustain normal hematopoiesis [7,20], and/or 4) an acquired genetic defect intrinsic to the hematopoietic stem cell at the origin of myelodysplasia [5]. Apoptotic signal transduction pathways can be induced by different stimuli [21]. Most of these pathways lead to activation of a novel family of cytosolic proteases termed caspases [22–24]. These proteins are expressed as inactive zymogens (the pro-caspases) that require proteolytic processing to generate an active enzyme. They all cleave peptide substrates after aspartic acid residues [25]. The processing sites for pro-caspases are themselves caspase consensus sequences. Caspases can be divided into initiators and effectors, based on their structure and order in the cell death cascade [23]. “Initiator” caspases (caspase-8 and caspase-10) activate downstream “effector” caspases (caspase-3, caspase6, caspase-7) Then, “effector” caspases such as caspase-3/ CPP32 cleave and inactivate proteins crucial to determine the biochemical and morphologic changes associated with apoptosis [26]. It has been reported that, in addition to in-

creasing the expression of Fas, IFN-␥ induces apoptosis in normal human erythroid progenitor cells by upregulation and activation of caspase-1, caspase-3, and caspase-8 [27]. Fas ligation leads to formation of the death-inducing signaling complex (DISC), which includes Fas, FasL, the Fasassociated protein with death domain (FADD), the recently described FLASH protein [28], and the pro–caspase-8 (also called FLICE, for FADD-like ICE) [29]. Activation of the DISC results in proteolytic activation of the pro–caspase-8, which in turn initiates a cascade of caspase activation leading to the final activation of caspase-3. However, the death receptor Fas also can activate a mitochondrial pathway of apoptosis [30,31]. There is some evidence that reducing apoptosis in MDS can improve the cytopenias observed in patients with MDS [3,10,11]. With the aim of developing new therapies, it is necessary to study the abnormalities of the apoptotic signal transduction pathways implicated in these disorders. In this report, we studied apoptosis in the BM of 54 patients with newly diagnosed MDS using the TUNEL assay. In accordance with previous observations, we found that excessive apoptosis is mainly observed in “early stage” myelodysplasia, that is, patients with refractory anemia with ringed sideroblasts (RARS), refractory anemia (RA), or RA with excess blasts (RAEB) and less than 10% blasts. We studied the activation state of the caspase-3/CPP32 enzyme, a major “effector” caspase highly expressed in cells of hematopoietic origin [27]. The levels of proteolytically active caspase-3 protein were increased in the majority of these “early stage” MDS patients compared to normal individuals. Finally, we investigated the ability to reverse the apoptotic phenotype and to improve hematopoiesis in vitro using the broad-spectrum irreversible caspase inhibitor Z-VADFMK. We demonstrate that this inhibitor can reverse the apoptotic phenotype in some patients but that it never improves the clonogenicity of erythroid and myeloid progenitor cells. These studies indicate that caspase-3 is involved in the mechanisms of increased apoptosis in MDS. However, they also demonstrate that caspase inhibition is insufficient to enhance the in vitro growth of mature erythroid and myeloid progenitors, suggesting that peptide inhibitors act too late in the apoptotic pathway.

Materials and methods Patients Fifty-four patients with newly diagnosed MDS and 12 normal BM donors were included in this study. They were studied for apoptosis using the TUNEL assay performed on BM smears. Eleven of

786

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

these 54 also were evaluated for apoptosis using the annexin V-fluorescein isothiocyanate (FITC) assay and 20 were evaluated for caspase-3 activity, whereas a group of 11 patients among these 20 were tested for the effect of the peptide inhibitor of caspases, Z-VAD-FMK. BM samples were collected before the initiation of therapy, after obtaining informed consent for study. The diagnosis of MDS was confirmed according to the French-American-British (FAB) committee criteria. MDS patients were subdivided into “early stage” MDS [including RA (n ⫽ 6); RARS (n ⫽ 9); and RAEB and less than 10% blasts (RAEB ⬍10, n ⫽ 15)] and “late stage” MDS [including RAEB with more than 10% blasts (RAEB ⬎10, n ⫽ 14), RAEB in transformation (RAEB-t, n ⫽ 6); and acute myelogenous leukemia secondary to MDS (LAM2, n ⫽ 4)]. Patients were also classified according to the International Prognosis Scoring System (IPSS) [32]. The clinical and laboratory parameters of these patients are given in Table 1. BM cell preparation BM was harvested by BM aspirates. BM mononuclear cells (BMMCs) were isolated by density gradient centrifugation using a Ficoll lymphocyte separation medium. Apoptosis detection In situ cell death was detected on BM smears using the TUNEL technique with the in situ cell death detection kit POD (Boehringer, Mannheim, Germany) as previously described [2,18]. Quantification of apoptosis was performed on total BMMCs using the annexin V-FITC binding assay (Boerhinger). Phosphatidylserine (PS) are anionic phospholipids that normally localize on the internal surface of cellular plasma membranes but are externalized when cells undergo apoptosis. PS externalization is thought to be a downstream event of caspase enzyme family activation during apoptosis [33]. Annexin V, a Ca2⫹–dependent phospholipid-binding protein, binds PS with high affinity.

Table 1. Clinical and biologic characteristics of 54 myelodysplastic syndrome (MDS) patients Patient characteristics Number Median age Sex (female:male) IPSS Low Intermediate-1 Intermediate-2 High Cytogenetics Good risk Intermediate risk Poor risk TUNEL Annexin V-FITC

“Early stage” MDS*

“Late stage” MDS†

30 62.5 (34–85) 18:12

24 63.5 (29–88) 14:10

18 (60%) 11 (36.6%) 1 (3.3%) 0

0 0 10 (41.6%) 14 (58.4%)

22 (73.4%) 8 (26.6%) 0 44.1 ⫾ 4.8% (N ⫽ 30) 36 ⫾ 9% (N ⫽ 11)

4 (16.6%) 9 (37.6%) 11 (45.8%) 21.8 ⫾ 3.6% (N ⫽ 24) ND

*“Early stage” MDS includes MDS patients with RA, RARS, and RAEB with less than 10% blasts (RAEB⬍10). † “Late stage” MDS includes patients with RAEB with more than 10% blasts (RAEB⬎10), RAEB-t, and LAM2. IPSS ⫽ International Prognosis Scoring System [32]; ND ⫽ not done.

Briefly, 5 ⫻ 105 BMMCs were used by assay. They were washed once in binding buffer containing 10 mM HEPES/NaOH (pH 7.4), 140 mM NaCl, and 5 mM CaCl2, then incubated with annexin V-FITC and propidium iodide (PI) (0.5 ␮g/mL, Interchim) in 100 ␮L of buffer. The suspension was incubated 30 minutes at room temperature in the dark and cells were resuspended in 400 of ␮L of binding buffer before flow cytometric analysis. The samples were analyzed using an XL cytometer (Coulter) Only cells positive for annexin V-FITC and negative for PI were considered to be truly apoptotic. Double-positive cells for annexin V-FITC and PI were considered to be necrotic and excluded. Excessive apoptosis was considered if the percentage of annexin V positive cells was above the mean ⫹2 SD of normal controls. Protease inhibitor The peptide inhibitor of caspases used in this study was the benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (Z-VAD-FMK) (Bachem, Voisins le Bretonneaux, France), which is a broad specificity caspase inhibitor. The dose used ranged from 50 to 500 ␮M according to the assay (see Results). Hematopoietic cell cultures Clonogenic progenitor assays. Isolated BMMCs were plated in methylcellulose StemGEM 1a containing interleukin 3 (IL-3), IL-6, stem cell factor, erythropoietin, and granulocyte-macrophage colony-stimulating factor (StemGEM 1a, Villejuif, France), in the presence or absence of protease inhibitors. Total BMMCs were plated at a density of 105 cells for MDS and 5 ⫻ 104 cells for normal BM in 1 mL of medium in 35-mm tissue culture dishes (Falcon, Becton-Dickinson) All experiments were performed in triplicate. Colony-forming unit erythroid (CFU-E) was enumerated at day 7 of culture, and colony-forming unit granulocyte-monocytes (CFU-GM) and burst-forming unit erythroid (BFU-E) at day 14. Liquid cultures of BMMCs. Total BMMCs were washed twice in Hank’s balanced salt solution (HBSS, Gibco, Eragny, France) and resuspended at 106/mL in Iscove’s modified Dulbecco’s medium (IMDM) supplemented with 4 mg/mL bovine serum albumin (BSA, Sigma, St. Louis, MO), 2 mg/mL iron-saturated transferrin (Sigma), 0.3 mg/mL L-glutamine, 100 U/mL penicillin G, and 0.1 mg/mL steptomycin. Cells were incubated for 12 hours with or without the protease inhibitor Z-VAD-FMK at a concentration of 100 or 500 ␮M. Evaluation of caspase-3 activity Evaluation of caspase-3 activation in total BMMCs was performed using the CaspACETM assay kit (Promega), according to the manufacturer’s instructions. Briefly, lysates from 107 cells were incubated with the fluorogenic substrate for CPP32/caspase-3, AcDEVD-AMC (where AMC is the fluorochrome 7-amino-4-methyl coumarin), for 60 minutes at 30⬚C in a buffer containing 312.5 mM HEPES (pH 7.5), 31.2% sucrose, and 0.3% CHAPS (3-[(3cholamidopropyl)-dimethylammonio]-1 propane-sulfonate) All the reactions were performed in duplicate. The control of specificity for caspase-3 activation was performed in each case by incubation of cell lysates with the CPP32/caspase-3 inhibitor Ac-DEVDCHO, before reaction with the Ac-DEVD-AMC substrate. Briefly, the caspase-3 substrate used in this assay is labeled with the fluorochrome 7-amino-4-methyl coumarin AMC that produces a blue fluorescence that can be detected by exposure to ultraviolet light at 360 nM. AMC is released from these substrates on cleavage by

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

CPP32 enzyme, and free AMC produces a yellow-green fluorescence that is monitored by a fluorometer at 460 nM. The amount of yellow-green fluorescence produced on cleavage is proportional to the amount of proteolytically active caspase-3 enzyme. Results are given as fluorescence intensity unit for 107 cells. Caspase-3 activity was considered to be increased if it was above the mean ⫹2 SD of normal controls. Statistical analysis The Student’s t-test was used to compare quantitative data expressed as means ⫾ SEM. Correlation between paired variables was studied using a Spearman test.

787

Table 2. Percentage of apoptotic cells in the BM smears of MDS patients using the TUNEL assay FAB RARS RA RAEB ⬍10% blasts ⬎10% blasts RAEB-t/LAM2 Normal BM

No. of patients

Apoptotic cells by TUNEL (%)

9 6 29 15 14 10 12

47% ⫾ 6 40% ⫾ 3 37% ⫾ 4.8 44% ⫾ 3.5 29.5% ⫾ 3.6 11% ⫾ 2.4 12% ⫾ 2.5

patients. Only 11 among these 20 belonged to the 54 previously described patients studied by TUNEL. Total BMMCs obtained from these 20 patients were incubated in the presence of the caspase-3 substrate DEVD-AMC with or without previous incubation with the caspase-3 inhibitor Ac-DEVD-CHO to assess the specificity of the reaction. Apoptosis was quantified in the same cellular population using annexin V-FITC binding. In five normal individuals, the mean caspase-3 activity was 4,256 U (range 3,200–5,200). The median percentage of apoptotic cells using annexin V-FITC fixation to externalized PS in these controls was 5% (range 3–6%). When normal BMMCs were treated overnight with 50 ␮g/mL of etoposide, which is an inhibitor of topoisomerase II with prominent antitumor activity, apoptosis increased (mean percentage of annexin V-FITC positive cells 47%) as well as the level of caspase-3 activity (mean 25,000 U, range 18,000–37,000) (Fig. 1). The 20 “early stage” MDS patients studied for the caspase-3 activity included 4 MDSRA, 4 MDSRARS, and 12 MDSRAEB⬍10. As shown in Figure 1, their median caspase-3 activity was 19,000 U (range 3,460–41,000) and signifi-

RAEB⬍10

Results Apoptosis of BM cells is increased in patients with “early” myelodysplasia We studied apoptosis on BM smears using the TUNEL assay as previously reported [2,18] in a group of 54 patients with newly diagnosed MDS. The mean percentage of apoptotic cells was 12.0% ⫾ 2.5% (range 6–18%) in 12 normal individuals. Apoptosis of BM cells in the 54 patients with MDS was statistically increased compared to controls (mean 34.1% ⫾ 4.3%; range 1–92%; p ⫽ 0.03). Table 2 gives the percentage of BM cells engaged in the process of programmed cell death according to the FAB group. The patients with MDSRA/RARS and MDSRAEB had significantly increased apoptosis compared to normal individuals (44.2% ⫾ 4.5%, p ⫽ 0.012; and 37.0% ⫾ 4.8%; p ⫽ 0.035 respectively). Fifteen of the 29 patients (52%) with RAEB had less than 10% blasts in the BM at the time of the study (MDSRAEB⬍10). Patients with MDSRAEB⬍10 had increased apoptosis compared to MDSRAEB⬎10 patients (44.0% ⫾ 3.5% vs 29.5% ⫾ 3.6%; p ⫽ 0.042). The group of patients with “early stage” MDS (MDSRA/RARS/RAEB⬍10) who completely overlapped with the “Low” and “INT-1” subgroups of the IPSS had increased apoptosis compared to the group of patients with “late stage” MDS (MDSRAEB⬎10/RAEB-t/LAM2) corresponding to the “INT-2” and “High” subgroups of the IPSS (44.1% ⫾ 4.8% vs 21.8% ⫾ 3.6%; p ⫽ 0.02). The percentage of apoptotic cells using the annexin V-FITC assay was 36% ⫾ 9% in 11 patients tested among the 30 with “early stage” MDS (Table 1), whereas the TUNEL assay performed in the same 11 patients gave a mean percentage of 41% ⫾ 7.5% of apoptotic cells. High rates of apoptosis correlated with low blast numbers (Spearman test, p ⫽ 0.012). These results confirm that a major biologic characteristic of patients with “early” MDS at diagnosis is excessive apoptosis of BM hematopoietic progenitors. Proteolytic activity of the caspase-3/CPP32 enzyme is increased in BM of patients with “early” myelodysplasia As apoptosis is increased in “early stage” MDS, we studied the level of caspase-3 activity in the BM of 20 MDSRA/RARS/

Figure 1. Increased caspase-3/CPP32 activity in MDS. Lysates from 107 BMMCs were incubated with the fluorogenic caspase-3/CPP32 substrate DEVD-AMC alone (Ac⫺DEVD⫺, light shading) or after previous incubation with the caspase-inhibitor Ac-DEVD-CHO (Ac-DEVD⫹; dark shading) to control the specificity of the reaction. Results are given as the mean fluorescence intensity for 107 BMMCs obtained for the 5 normal individuals and the 20 patients with “early” MDS. These included 4 RA, 4 RARS, and 12 RAEB⬍10.

788

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

cantly increased compared to normal individuals (p ⫽ 0.032) Seventeen of these 20 patients (85%) had increased caspase-3 activity defined as a value above the mean ⫹2 SD of normal controls, whereas 3 patients (patients 1, 2, and 6 reported in Table 3) had normal values. Using the annexin V-FITC assay, the percentage of apoptotic cells was 34% ⫾ 12% (range 6–93%) in these 20 patients and significantly increased. The TUNEL assay gave a mean percentage of 41% of apoptotic cells (range 15–54%) in the 11 patients studied. Peptide inhibitor of caspases Z-VAD-FMK can reverse the apoptotic phenotype of BMMCs in vitro Because the proteolytic activity of the CPP32/caspase-3 enzyme is increased in the majority of “early stage” MDS patients, we sought to determine if caspase inhibition could rescue the apoptotic phenotype of BMMCs in these disorders, by decreasing the percentage of annexin V-FITC positive BMMCs. We first determined the concentrations of Z-VAD-FMK necessary to inhibit etoposide-induced apoptosis of normal BMMCs (exposition to 50 ␮g/mL of etoposide for 12 hours). Figure 2 shows that Z-VAD-FMK blocked morphologic apoptosis in a dose-dependent manner. The protective effect of Z-VAD-FMK began at 100 ␮M and was complete at 500 ␮M. We used concentrations of 100 and 500 ␮M for further experiments in MDS. We studied 12 patients with “early stage” MDS, including 11 from the 20 previously reported, for the caspase-3 activity. The FAB classification was as follows: RARS 3, RA

Table 3. Effect of the caspase inhibitor Z-VAD-FMK on apoptosis of BMMCs in patients with MDS†

Patient

Caspase-3 activity*

Percent annexin V ⫹ Z-VAD-FMK 0/500 ␮M

1. RARS 2. RARS 3. RARS 4. RA 5. RA 6. RA 7. RAEB ⬍10 8. RAEB ⬍10 9. RAEB ⬍10 10. RAEB ⬍10 11. RAEB ⬍10 12. RAEB ⬍10 Control (n ⫽ 5)

3,460 U 5,500 U 9,200 U ND 39,400 U 6,800 U 12,000 U 9,500 U 22,000 U 11,000 U 9,200 U 24,900 U 4,256 U

29/32% ND/16% 32/29% 7/7% 22/8% 17/16% 22/7% 6/10% 65/22% 9/8% 32/10% 16/18% 5/5%

*The proteolytic activity of the caspase-3 enzyme was quantified at T0 as defined in the Materials and methods section. † BMMCs were cultured for 12 hours in IMDM without or with Z-VAD-FMK at a concentration of 100 or 500 ␮M. Only the results for 500 ␮M are reported. Apoptosis was quantified at 12 hours by the binding of annexin V-FITC. Results of patients 5, 7, 9, and 11 with reduced apoptosis due to Z-VAD are given in bold. ND ⫽ not done.

Figure 2. Effect of the caspase-inhibitor Z-DEVD-FMK on etoposideinduced apoptosis of normal BMMCs. BMMCs were incubated with etoposide in the absence or presence of the Z-VAD-FMK inhibitor. Apoptosis was quantified at 12 hours by binding of annexin V-FITC to externalized PS.

3, and RAEB 6. BMMCs obtained from these patients were incubated for 12 hours in IMDM without added fetal calf serum or cytokines, with or without Z-VAD-FMK at 100 or 500 ␮M. Nine of these 12 patients had increased apoptosis using the annexin V-FITC assay. Results are shown in Table 3. We observed a dose-dependent decrease of apoptotic cells due to caspase inhibition by the Z-VAD-FMK peptide in 4 of these 9 patients (patients 5, 7, 9, and 11 reported in bold in Table 3) These patients had increased caspase-3 activity of 39,400, 12,000, 22,000, and 9,200 U, respectively. The mean percentage of annexin V-FITC positive cells in these 12 patients was 22,8% ⫾ 16 %(range 6–65%). Peptide inhibitor of caspases Z-VAD-FMK does not enhance colony formation in vitro Because the caspase-inhibitor Z-VAD-FMK can attenuate apoptosis in some MDS patients, we sought to determine if this compound could restore hematopoiesis in vitro. We first determined the toxicity of Z-VAD-FMK used at different concentrations (50, 100, 200, 500, or 1,000 ␮M) on the clonogenicity of normal hematopoietic progenitor cells. From four different experiments, there was no toxicity of Z-VAD-FMK used at concentrations of 50, 100 or 500 ␮M on the growth of erythroid and myeloid colonies. However, when Z-VAD-FMK was used at 1,000 ␮M, a mean 32% decrease of both BFU-E and CFU-E colony number was observed (data not shown). BMMCs from the MDS patients reported in Table 3 were plated in methylcellulose with Z-VAD-FMK at 100 and 500 ␮M. Table 4 shows the mean number of CFU-E, BFU-E, and CFU-GM–derived colonies obtained with or without the in-

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

hibitor. Caspase inhibition did not improve the proliferation and differentiation of BM hematopoietic progenitors in any of these patients. To ensure that the half-life of the Z-VADFMK inhibitor was not too short to protect MDS colony formation, this compound was re-added at 100 ␮M in two patients (patients 7 and 10 reported in Table 3) every 2 days up to day 8 of culture (cumulative dose of 500 ␮M). In these conditions, hematopoiesis was not improved (data not shown). Moreover, at 500 ␮M, the Z-VAD-FMK inhibitor could only partially protect hematopoiesis of normal BMMNCs exposed to 50 ␮g/mL of etoposide for 12 hours (Table 4). Discussion Ineffective intramedullary hematopoiesis is a characteristic feature of MDS, especially observed in the erythroid lineage. Among the biologic mechanisms that can lead to impaired erythropoiesis and myelopoiesis in MDS, excessive apoptosis may play a role. Some characteristics of apoptosis in MDS have been described. 1) Most studies reported increased apoptosis in FAB subtypes RA, RARS, and RAEB [1,2,4–6,10], although Raza et al. [7] found excessive apoptosis even in advanced stages of the disease. 2) Apoptotic cells are observed in the erythroid and myeloid lineages at all stages of differentiation [2]. 3) Some cells of stromal origin may be engaged in PCD [7]. 4) Hematopoietic stem cells expressing the CD34 antigen are highly apoptotic [5,6,10]. 5) Altered expression of proteins with pro-apoptotic function (such as c-myc or Bad and Bax) vs proteins with anti-apoptotic function (such as Bcl-2 and Bcl-xl) is observed in CD34⫹ cells from patients with “early” stage MDS compared to normal CD34⫹ cells and may render them more sensitive to apoptosis [5,6,10]. 6) A role of the Fas-FasL system has been emphasized. The death receptor Table 4. Effect of the caspase-inhibitor Z-VAD-FMK on hematopoietic progenitor growth in normal BM and MDS

Colonies*

Controls (n ⫽ 4) [normal BM]

CFU-E

152 ⫾ 40

BFU-E

110 ⫾ 35

CFU-GM

316 ⫾ 62

Control (n ⫽ 1) [a/ control b/ ⫹Etoposide c/ ⫹Etoposide ⫹ Z-VAD 500 ␮M]

MDS ⫾ Z-VAD (n ⫽ 12) [a/ ⫺Z-VAD b/ Z-VAD 100 ␮M c/ Z-VAD 500 ␮M]

a/ 158 b/ 45 c/ 68 a/ 125 b/ 35 c/ 55 a/ 328 b/ 75 c/ 155

a/ 5 (1–18) b/ 4 (12–17) c/ 6 (13–16) a/ 10 (4–22) b/ 11 (6–18) c/ 8 (4–14) a/ 80 (55–150) b/ 85 (60–160) c/ 82 (62–162)

*Numbers are given as the mean numbers of colonies for 105 BMMCs plated in triplicate, for the 12 patients with MDS. Z-VAD-FMK was used at 100 or 500 ␮M. Normal BMMCs were incubated for 12 hours with 50 ␮g/mL etoposide with or without Z-VAD-FMK at 500 ␮M.

789

Fas/CD95 is highly expressed on CD34⫹, CD33⫹, CD14⫹, and glycophorin⫹ cells from MDS patients, and Fas triggering by an agonistic anti-Fas antibody induces apoptosis, impairs proliferation. and decreases the in vitro growth of erythroid and myeloid progenitors in myelodysplasia [2,4,14,34]. We recently reported a correlation between Fas expression on glycophorin⫹ cells and impaired clonogenicity of mature erythroid progenitors [18]. Moreover, FasL itself is highly expressed in BM of MDS patients [12–14]. These results indicate that hematopoietic progenitors can express simultaneously a functional Fas protein and its ligand, strongly suggesting a major role for this system in apoptosis of myelodysplasia. However, the exact role assigned to excessive apoptosis in the pathogenesis of MDS is still unclear. It could be a TNF-␣ or IFN-␥–mediated process raised in reaction to excessive proliferation observed in patients with MDS. Alternatively, it could arise from an intrinsic defect within the stem cell at the origin of myelodysplasia. Whatever the hypothesis, it is essential to have a clearer understanding of the molecular mechanisms underlying apoptosis in these disorders. We studied apoptosis using the TUNEL assay in 54 newly diagnosed patients with MDS. As previously reported, apoptosis was significantly increased in patients with RA, RARS, and RAEB compared with apoptosis observed in normal individuals and in the 10 patients with RAEB-t or LAM secondary to myelodysplasia. We also confirmed in a larger series of patients that increased apoptosis is inversely correlated with the number of blasts [2]. Moreover, RAEB patients with less than 10% blasts had higher apoptosis rates than RAEB patients with more than 10% blasts (44.0% ⫾ 3.5% vs 29.5% ⫾ 3.6%, respectively, p ⫽ 0.042), suggesting that decrease of apoptosis in patients with RAEB may be a bad prognosis factor if associated with progressive disease. Overall, our results confirmed that increased apoptosis is mainly restricted to the group of patients with “early” MDS (including RA, RARS, and RAEB with less than 10% blasts). The apoptotic signaling pathways activated in this group of patients with “early” MDS are essentially unknown. Because caspases are the main proteins implicated in apoptosis described in response to many different apoptotic stimuli, we studied the proteolytic activity of the central caspase-3/ CPP32 enzyme in the BM of these patients. Mundle et al. [35] reported increased activity of the caspase-1/ICE protein in MDS patients. However, gene targeting of caspase-1 revealed that it is probably dispensable for most apoptotic pathways [36]. More recently, the same group reported a correlation between TNF-␣ expression and high caspase-3 activity in BMMCs from patients with MDS [37]. The proteolytic activation of caspase-3 is mediated by upstream caspases from different apoptotic pathways [23]. In particular, ligation of Fas or the TNF-␣–R1 results in direct activation of caspases, including caspase-3. To assess the activation state of caspase-3 in the BM of our patients, we used a highly specific and quantitative assay. A narrow range of

790

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791

caspase-3 activity was detected in normal BM used as controls, indicating that a basal level of caspase-3 activity is present in normal BM progenitor cells. This phenomenon was reported recently in purified human erythroid progenitors [27]. It may reflect the dynamic process of proliferation and apoptosis during normal hematopoiesis. Seventeen of the 20 MDS patients (85%) with “early stage” MDS we studied had increased caspase-3 activity. Only 3 of 20 (15%) had levels comparable to those of normal individuals (patients 1, 2, and 6 reported in Table 3), whereas they had increased apoptosis by annexin V-FITC assays (reported in Table 3), suggesting that caspases other than caspase-3 or caspase-independent pathways of apoptosis are involved in programmed cell death in these patients. The 12 MDSRAEB⬍10 patients studied had increased caspase-3 activity. However, we did not find any statistic correlation between apoptosis and caspase-3 levels. The best explanation for this is the important cellular heterogeneity of samples analysed, with the possibility of variable expression levels of the caspase-3 pro-enzyme according to the cell type. Interestingly, there was no correlation between Fas expression, analyzed by flow cytometry on total BMMCs, on CD34⫹, CD33⫹, or glycophorin⫹ cells as previously reported [2] and the caspase-3 activity in the 20 “early” MDS patients we studied (data not shown). In this study, we used the irreversible broad-spectrum protease inhibitor Z-VAD-FMK. In 4 of 9 MDS patients with increased apoptosis studied, we observed a dose-dependent decrease in the percentage of annexin V-FITC positive BMMCs using Z-VAD-FMK (Table 3). In the same conditions, Z-VAD-FMK was able to inhibit etoposide-induced apoptosis of normal BMMCs (Fig. 2). The lack of effect observed in some MDS patients cannot be explained by a limiting concentration of Z-VAD-FMK, because a concentration of 1,000 ␮M of the inhibitor was tested in the first four patients without any effect. These results indicate that caspase inhibition can modify the apoptotic phenotype of BMMCs in some MDS patients. The other goal of this work was to examine the possibility to enhance proliferation and differentiation of hematopoietic progenitor cells in vitro by protease inhibitors. It was justified by two recently published works, using different techniques, demonstrating that apoptosis is mainly increased in the CD34⫹ cells of “early stage” MDS patients [5,6,10]. Experiments performed in 12 MDS patients with various rates of apoptosis and caspase-3 activity indicated that the inhibitor never rescued hematopoiesis in vitro (Table 4). We were unable to prevent or repair the apoptotic cell damage even in a small fraction of the cells using Z-VADFMK. These findings are in agreement with a recent report demonstrating that protease inhibitors can inhibit the apoptotic phenotype of normal human erythroid progenitors exposed to erythropoietin deprivation, but do not modify their fonctionality in term of rescued clonogenicity [38]. Moreover, the same inhibitor was unable to efficiently protect colony formation after etoposide-induced apoptosis of normal

human BMMCs (see Results section and Table 4). These results suggest the existence of a point of “no return” along the apoptotic signaling cascade, after which the cell is irreparably committed to die by a still unknown biochemical mechanism. The events that commit a cell to irreversibly engage the execution machinery are not well defined, but some data obtained from hematopoietic cell lines suggest that cells with reduced mitochondrial membrane potential and increased cytosolic cytochrome c can survive and maintain clonogenicity, if appropriate survival signals are given to suppress apoptosis [39]. Mitochondrial damages was observed in the majority of apoptotic systems, and these changes may precede activation of caspases, including caspase-3, according to the apoptotic stimuli. The biochemical characteristics of mitochondria in MDS progenitor cells of MDS are actually unknown. Experiments are currently under way to determine if there is loss of electrical potential across the inner mitochondrial membrane in these disorders. We suggest that protease inhibitors used alone in MDS are insufficient in vitro to improve hematopoiesis because they act too late in the apoptotic signaling pathway, after the point of “no return,” even if adequate anti-apoptotic concentrations of cytokines are added. This hypothesis explains why the same caspase inhibitor, used at the same concentrations but administered at the initiation of the apoptotic stimulus, is efficient in protecting normal BMMCs from etoposide-induced apoptosis. To develop new therapeutic tools in the future targeting the apoptotic pathways in pathologic conditions such as MDS, the phases of initiation, commitment, and execution of apoptosis need to be defined. We demonstrated in this study that caspase activation is involved in the pathogenesis of apoptosis in MDS. However, caspase inhibition with a very efficient protease inhibitor cannot rescue hematopoiesis, at least in in vitro short-term assays. However, recent data indicate that inhibition of Fas-mediated signaling using a Fas-immunoglobulin (Fas-Fc) fusion protein or inhibition of TNF-␣ signaling with TNF-R-Fc protein result in enhanced in vitro colony formation in myelodysplasia [12]. These encouraging results suggest that blockade of apoptosis at the initiation phase of the apoptotic stimuli may be a more effective approach. They also suggest that protease inhibitors should be tested in vitro in association with drugs or molecules that target different phases of the apoptotic program and may prove to be efficient to improve hematopoiesis in MDS in the future. Acknowledgments This work was supported by a grant from the DRC (Direction pour la Recherche Clinique) The authors are grateful to Sylvie Gisselbrecht for helpful discussion.

References 1. Bogdanovic AD, Tripnac DP, Jankovic GM, et al. (1997) Incidence and role of apoptosis in myelodysplastic syndrome: morphological and ultrastructural assessment. Leukemia 11:656

D. Bouscary et al./Experimental Hematology 28 (2000) 784–791 2. Bouscary D, De Vos J, Guesnu M, et al. (1997) FAS/Apo-1/CD95 expression and apoptosis in patients with myelodysplastic syndromes. Leukemia 11:839 3. Hellstrom-Lindberg E, Kanter-Lewensohn L, Ost A (1997) Morphological changes and apoptosis in BM from patients with myelodysplastic syndromes treated with granulocyte-CSF and erythropoietin. Leukemia Reasearch 21:415-4257 4. Lepelley P, Campergue L, Gradel N, et al. (1996) Is apoptosis a massive process in myelodysplastic syndromes? Br J Haematol 95:368 5. Parker JE, Mufti GJ (1998) Ineffective haematopoiesis and apoptosis in myelodysplastic syndromes. Br J Haematol 101:220 6. Parker JE, Fishlock KL, Mijovic A, et al. (1998) Low risk myelodysplastic syndromes is associated with excessive apoptosis and an increased ratio of pro-versus anti-apoptotic Bcl-2-related proteins. Br J Haematol 103:1075 7. Raza A, Gezer S, Mundle S, et al. (1995) Apoptosis in bone marrow biopsy samples involving stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes. Blood 86:268 8. Raza A, Gregory SA, Preisler HD (1996) The myelodysplastic syndromes in 1996: complex stem cell disorders confounded by dual actions of cytokines. Leuk Res 20:881 9. Raza A, Alvi S, Borok RZ, et al. (1997) Excessive proliferation matched by excessive apoptosis in myelodysplastic syndromes: the cause-effect relationship. Leuk Lymphoma 27:111 10. Rajapaksa R, Ginzton N, Rott LS, et al. (1996) Altered oncoprotein expression and apoptosis in myelodysplastic syndrome marrow cells. Blood 88:4275 11. Schetty V, Mundle S, Alvi S, et al. (1996) Measurement of apoptosis, proliferation and three cytokines in 46 patients with myelodysplastic syndromes. Leuk Res 20:891 12. Gersuk G, Beckham C, Loken MR, et al. (1998) A role for tumor necrosis factor-␣ Fas and Fas-ligand in marrow failure associated with myelodysplastic syndromes. Br J Med 103:176 13. Gupta P, Niehans GA, LeRoy SC, et al. (1999) FasL expression in the bone marrow in myelodysplastic syndromes correlates with FAB subtype and anemia, and predict survival. Leukemia 13:44 14. Kitagawa M, Yamaguchi S, Takahashi M, et al. (1998) Localization of Fas and Fas ligand in bone marrow cells demonstrating myelodysplasia. Leukemia 12:486 15. Maciejewski J, Selleri C, Anderson S, et al. (1995) Fas antigen expression on CD34⫹ human marrow cells is induced by interferon gamma and tumor necrosis factor alpha and potentiates cytokine-mediated hematopoietic suppression in vitro. Blood 85:3086 16. Nagafuji K, Shibuya T, Harada M, et al. (1995) Functional expression of Fas antigen (CD95) on hematopoietic progenitor cells. Blood 86:883 17. Dai C, Price J, Brunner T, et al. (1998) Fas ligand is present in human erythroid colony-forming cells and interacts with Fas induced by Interferon ␥ to produce erythroid cell apoptosis. Blood 91:1235 18. Fontenay-Roupie M, Bouscary D, Guesnu M, et al. (1999) Ineffective erythropoiesis in myelodysplastic syndromes: correlation with Fas expression but not with lack of erythropoietin receptor signal transduction. Br J Haematol 106:464 19. Hoefsloot LH, Van Amelsvoort MP, Broeders LC, et al. (1997) Erythropoietin-induced activation of STAT-5 is impaired in the myelodysplastic syndrome. Blood 89:1690 20. Aizawa S, Nakano M, Iwase O, et al. (1999) Bone marrow stroma

21. 22. 23.

24. 25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

36. 37.

38.

39.

791

from refractory anemia of myelodysplastic syndrome is defective in its ability to support normal CD34 positive cell proliferation and differentiation in vitro. Leuk Re 23:239 Raff MC (1992) Social controls on cell survival and cell death. Science 356:397 Cohen GM (1997) Caspases: the executioners of apoptosis. Biochem J 326:1 Los M, Wesselborg S, Schulze-Osthoff K (1999) The role of caspases in development, immunity, and apoptotic signal transduction: lessons from knockout mice. Immunity 10:629 Nagata S (1997) Apoptosis by death factors. Cell 88:355 Garcia-Calvo M, Peterson EP, Leiting B, et al. (1998) Inhibition of human caspases by peptide-based and macromolecular inhibitors. J Biol Chem 273:32608 Liu XS, Zou H, Slaughter C, Wang XD (1997) DFF, a heterodimeric protein that functions downstream of caspase-3 to trigger DNA fragmentation during apoptosis. Cell 89:175 Dai C, Krantz SB (1999) Interferon ␥ induces upregulation and activation of caspases 1, 3 and 8 to produce apoptosis in human erytroid progenitor cells. Blood 93:3309 Imai Y, Kimura T, Murakami A, et al. (1999) The CED-4-homologous protein FLASH is involved in Fas-mediated activation of caspase-8 during apoptosis. Nature 398:777 Boldin MP, Goncharov TM, Goltsev YV, et al. (1996) Involvement of MACH, a novel MORT1/FADD-interacting protease, in Fas/APO1 and TNF receptor-induced cell death. Cell 85:803 Li P, Allen H, Banerjee S, Franklin S, et al. (1995) Mice deficient in IL-1␤-converting enzyme are defective in production of mature IL-1␤ and resistant to endotoxic shock. Cell 80:401 Luo X, Budihardjo I, Zou H, et al. (1998) Bid, a Bcl-2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptor. Cell 94:481 Greenberg P, Cox C, Lebeau MM, et al. (1997) International Scoring System for evaluating prognosis in myelodysplastic syndromes. Blood 89:2079 Naito M, Nagashima K, Mashima T, et al. (1997) Phosphatidylserine externalization is a downstream event of interleukin-1␤-converting enzyme family protease activation during apoptosis. Blood 99:2060 Lepelley P, Gardel N, Erny O, et al. (1998) Fas/APO-1 (CD95) expression in myelodysplastic syndromes. Leuk Lymphoma 30:307 Mundle SD, Vernugopal P, Cartlidge JD, et al. (1996) Indication of an involvement of Interleukin-1␤ converting enzyme-like protease in intramedullary apoptotic cell death in the bone marrow of patients with myelodysplastic syndromes. Blood 88:2640 Li H, Zhu H, Xu CJ, et al. (1998) Cleavage of BID by caspase-8 mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell 94:491 Mundle SD, Reza S, Ali A, et al. (1999) Correlation of tumor necrosis alpha (TNF-␣) with high caspase-3-like activity in myelodysplastic syndromes. Cancer Lett 140:201 Gregoli PA, Bondurant MC (1999) Function of caspases in regulating apoptosis caused by erythropoietin deprivation in erythroid progenitors. J Cell Physiol 178:133 Chen Q, Takeyama N, Brady G, et al. (1998) Blood cells with reduced mitochondrial membrane potential and cytosolic cytochrome C can survive and maintain clonogenicity given appropriate signals to suppress apoptosis. Blood 92:4545