CD34 cells from patients with trisomy 8 myelodysplastic syndrome (MDS) express early apoptotic markers but avoid programmed cell death by up-regulation of antiapoptotic proteins

Myelodysplastic syndromes (MDSs) are heterogeneous disorders showing varying degrees of bone marrow (BM) failure and a tendency to evolve to acute leukemia. To clarify the cellular and molecular mechanisms in MDS, we have focused on syndromes that are defined by cytogenetics. In trisomy 8 MDS, patients suffer predominantly from pancytopenia; indeed, the diagnosis can be confounded with aplastic anemia, and some patients with aplastic anemia can evolve to trisomy 8 MDS.¹  We have observed that trisomy 8 MDS is associated with oligoclonal T-cell expansion, suggestive of an antigen-driven pathophysiology.²  We also observed that trisomy 8 cells, compared with normal cells from individual patients' BM, are Fas-receptor⁺ and annexin⁺,³  findings consistent with a cell population under immune attack and undergoing apoptosis.⁴ 

This model of pathophysiology is compatible with the larger role for apoptosis hypothesized for the MDS. However, that trisomy 8 cells are targeted and undergo apoptosis remains paradoxical because this mechanism would be expected to eradicate the abnormal clone. A possible explanation for the survival of trisomy 8 cells despite an apparently specific immune attack and the induction of apoptosis was suggested by experiments demonstrating that annexin⁺ BM cells from MDS showed good hematopoietic colony formation.⁵  Survival of abnormal CD34 cells in MDS might therefore be explicable based on apoptosis that is incomplete, as, for example, due to failure of signal transduction upstream of caspase 3 activation and the phosphatydyinositol membrane changes that result in annexin binding.^6,7 

Previously we demonstrated on microarray analysis that proteins affecting apoptosis and proliferation, c-myc and CD1, were up-regulated in CD34 cells of patients with trisomy 8.⁸  In this study we examined CD1 and c-myc expression by real-time polymerase chain reaction (PCR) and assessed the clonogenicity of annexin⁺ trisomy 8 cells in culture.

After informed consent, using protocols approved by the Institutional Review Board of the National Heart, Lung, and Blood Institute, BM was obtained from a cohort of patients with MDS described in a previous publication.⁹  All samples were obtained prior to immunosuppressive treatment.

Separated cell preparations were enriched for CD34 cells using the magnetic-activated cell-sorting (MACS) column or by flow cytometric sorting. The purity of CD34 cells was 88% to 94% as determined by flow cytometry (EPICS Altra; Beckman Coulter, Miami, FL).

Peripheral blood mononuclear cells (PBMCs) were stained with annexin and propidium iodide (Tat assay) as previously described;¹⁰  cells which stained with propidium iodide were excluded from analysis. Cells for sorting were only stained with annexin-FITC. An in situ thymidine deoxyneuclease (TdT) assay (Apotag; Oncor, Gaithersburg, MD) also was used to quantitate the number of cells that underwent apoptosis as previously described;¹¹  all assays were performed in duplicate.

Intracellular staining for CD1 expression was performed using the PharMingen Intracellular Staining Kit (San Diego, CA)¹²  staining with FITC-labeled anti-CD1 antibody. Samples were sorted by flow cytometry as previously described.¹³  A number (5000-10 000) of cells were examined for each sample, and in most experiments in duplicate.

Fluorescent in situ hybridization (FISH) was performed as previously described¹⁴  with probes for chromosomes 5q, 7, and 8 (Vysis Inc, Downers Grove, IL). Three different observers, blinded with respect to sample identity, examined 3 different sets of slides. A positive healthy control and a trisomy 8⁺ control were hybridized and scored with each separate run. Samples were run in triplicate and each read by 3 blinded observers. Standard error of the mean is calculated for each individual sample.

DNA analysis of cell-cycle status was performed using flow cytometry. CD34 cells were purified using a MACS column as previously described.¹⁵  In some instances CD34 cells which were sorted into annexin⁺ and annexin⁻ fractions using flow cytometry. Cell-cycle analysis of propidium iodide was performed using MCycle software version 2.5 (Phoenix Flow Systems, Phoenix, AZ). This method allows determination of cell-cycle status and also quantitation of apoptotic cells by measurement of the hypodiploid peak. The positive apoptotic control was prepared by coculturing normal CD34 cells with IFN-γ (1000 U/mL) and Fas agonist, CH11 (Immunotech, Marseilles, France), for 1 day at 37°C.

Real-time PCR was performed as previously described¹⁶  on column-purified CD34 cells using primer pairs for c-myc,¹⁷  CD1,¹⁸  and survivin.¹⁹  Relative expression levels and differences in CD34 cells of individual patients compared with controls were validated using the TaqMan 5′ nuclease real-time reverse transcriptase–PCR (RT-PCR) assay. Samples from 8 patients (4 with monosomy 7, and 4 with trisomy 8) were subjected to real-time RT-PCR using appropriate primers for 10 genes of interest and β-actin as an internal control. Primers were designed using Clonal Manager Suite software except for TRAIL (R&D Systems, Minneapolis, MN). Using RT, cDNA was synthesized from 500 ng total RNA extracted from CD34 cells; aliquots of cDNA were used as templates for real-time PCR reactions containing primers of either target genes or β-actin. Each amplification was performed in duplicate for each sample in a 25-μL reaction mixture using the components of the TaqMan PCR Core Reagent Kit (Perkin Elmer, Les Ulis, France) and 300 nM of forward and reverse primers. The PCR reactions, TaqMan analyses, and subsequent calculations were performed with the ABI Prism 7700 sequence detection system (Perkin Elmer Biosystems, Foster City, CA) and SYBR Green reagents. The threshold cycle (C_T) value was defined as the fraction cycle number and set at 10 times the standard deviation above the mean baseline fluorescence calculated from cycles 3 to 15. The fold changes in the target genes normalized to β-actin and relative expression of healthy controls were calculated for each sample using the 2^−CT method, where −C_T = (C_{T, Target} − C_{T, Actin}) _{Patient's sample} −(C_{T, Target} − C_{T, Actin}) _{Normal control}.

BM mononuclear cells (BMMNCs) were used for the preparation of the extracts as previously described.²⁰  The protein content of the extracts was determined using the Micro BCA Protein Assay kit (Pierce, Rockford, IL).

Proteins (10 μg/lane) were resolved in 12% Tris-glycine SDS gels (Invitrogen, Carlsbad, CA) electrophoresis at 125 V. Resolved proteins from the gel were transferred to polyvinylidene difluoride (PVDF) membranes (Invitrogen). The membrane was blocked for 2 hours with 5% BSA in PBS and 0.05% Tween-20 followed by incubation in primary antibody (Ab). Subsequently, membranes were incubated in horseradish peroxidase (HRP)–conjucated secondary Ab, and detection of the bands of interest was performed using the electrochemiluminescence (ECL) plus system (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom). Membranes were stripped after the first blotting in ImmunoPure IgG elution buffer (Pierce), reblocked, and reblotted with another Ab. To evaluate equal loading of the lanes, membranes were reblotted with antiactin polyclonal Ab. Densitometry analysis of the bands of interest was performed using the ImageQuant analysis software from Amersham (Amersham Biosciences, Piscataway, NJ). Anti–c-myc Ab was purchased from Calbiochem (San Diego, CA) and antisurvivin Ab was from Zymed Laboratories (San Francisco, CA). The antiactin Ab and the HRP-conjugated Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

Aliquots of cells were cultured in 6-well plates in MethoCult GF H4535 (Stem Cell Technologies, Vancouver, BC, Canada). A c-myc inhibitor, Int-H1-S6A, F8A (Marligen Biosciences, Ijamsville, MD), was added to final concentrations of 15 μM and 20 μM. Int-H1-S6A, F8A interferes with specific c-Myc DNA binding through the cooperative binding of H1 peptides with tetrameric c-Myc-92.²¹  Cultures were incubated at 37°C, 5% CO₂ for 14 days; cells were then harvested for FISH and Western analysis.

BMMNCs (7 × 10⁵cells/well) were plated into a 12-well plate and cultured for 24 hours at 37°C with 5%CO₂. The transfections was performed following the Cell Signaling protocol (Dancers, MA) and were incubated for 36 hours at 37°C with 5%CO₂. SignalSilence survivin siRNA and goat anti–rabbit IgG-HRP were purchased from Cell Signaling Technology (Danvers, MA). The transfection reagent, TransIT-TKO, supplied with the SignalSilence Survivin siRNA Kit, was produced by Mirus Bio (Madison, WI). Transfected and control BMMNCs were cultured for 36 hours in hematopoietic growth factors before cell count, viability, and FISH were performed.

We performed real-time PCR on flow-cytometrically sorted CD34 cells from 8 patients with trisomy 8 MDS, 10 healthy controls, and 6 patients with monosomy 7 MDS. For each subject, the procedure was performed by 3 blinded observers each with 3 repeats, so that a total of 9 readings per subject were obtained. Within each reading, c-myc and cyclin D1 (CD1) expressions were evaluated based on a sample of 10 000 cells. Through a series of analysis of variance (ANOVA) tests, no significant interobserver and intraobserver differences for c-myc and CD1 expressions were detected. Consequently, the means and the corresponding standard errors of the c-myc and CD1 expressions were computed for each subject by pooling his/her data from all 9 readings, and mean and standard error for the mean c-myc and cyclin D1 expressions were computed for the 10 healthy controls. For each of the patients with trisomy 8 and monosomy 7 MDS, fold increases for c-myc and CD1 were computed by the ratios of the corresponding mean c-myc and CD1 expressions of the patient over the means of the 10 healthy controls. Sample means and their standard errors of the fold increases were computed for the 8 patients with trisomy 8 MDS and the 6 patients with monosomy 7 MDS, respectively. Linear regression models were used to analyze the relationships between the fold increases (c-myc and CD1) and the percent of trisomy 8 cells for the patients with trisomy 8 MDS.

We first sought to confirm our previous observation by microarray analysis that CD1 and c-myc mRNA were up-regulated in trisomy 8 MDS. We performed real-time PCR and FISH on flow-cytometrically sorted CD34 cells from 8 patients with trisomy 8 MDS and 10 healthy controls using probes described in “Materials and methods.” Both c-myc and CD1 expression were substantially increased in CD34 cells from patients with trisomy 8 compared with healthy individuals (fold increase is seen in Figure 1A). Expression of c-myc and CD1 was proportional to the number of trisomy 8 cells in the sample (Figure 1B-C).

We examined CD1 and c-myc protein expression using flow cytometry and immunoblot. Unseparated BM was stained with CD34-PE and CD1-FITC, and CD34⁺ cells were gated and examined by flow cytometry (Figure 2A) and compared with a group of 10 healthy controls. The amount of protein was significantly increased compared with controls (P < .01) and correlated positively with the number of trisomy 8 CD34⁺ cells in the sample (Figure 2B). When BMMNCs were sorted into CD34⁺, CD1⁺, and CD1⁻ fractions and FISH performed, there was a predominance of trisomy 8 cells in the CD1⁺ fraction on FISH analysis (Figure 2C). c-myc was not reliably detectable with flow cytometric staining but could be detected in excess amounts when nuclear protein was extracted from lymphocyte-depleted BMMNCs and analyzed by Western blot (immunoblots are seen in Figure 2D).

Up-regulation of cyclin D in mantle cell lymphomas is associated with increases in survivin.²²  In addition, survivin is one of the few inhibitors of apoptosis that will block apoptosis upstream of caspase 8 activation (already previously shown to be present in trisomy 8 cells²³ ). We examined survivin mRNA and protein expression. We performed real-time PCR on column-purified CD34 cells from 6 patients with trisomy 8 MDS. Survivin mRNA expression was compared with a group of 10 healthy individuals and to a group of patients with MDS monosomy 7 who were equally cytopenic. Survivin mRNA was significantly increased in patients with trisomy 8 (P < .05; fold difference is seen in Figure 3A); the relationship between survivin expression and numbers of trisomy 8 cells are seen in Figure 3B. Increases in survivin protein in purified CD34 cells were demonstrated by flow cytometry (Figure 4A-B) and by immunoblot (Figure 4C).

We previously demonstrated that trisomy 8 cells express annexin and activated caspase 8.²⁴  Here, we determined whether these signs of early apoptosis were associated with DNA fragmentation. For this purpose, we performed both cell-cycle analysis and the TdT-mediated dUTP nick-end labeling (TUNEL) assay. Cell-cycle analysis on purified CD34 cells from 5 patients with trisomy 8 showed no significant DNA degradation in the annexin⁺ fraction enriched with trisomy 8 cells (Table 1; Figure 5A). Annexin⁺ and annexin⁻ CD34⁺ cells from healthy donors served as controls. As expected, significant DNA degradation was seen in a positive control. When the TUNEL assay was performed on CD34 cells from an additional 4 patients with trisomy 8, there were many fewer TUNEL⁺ cells compared with annexin⁺ cells (Table 1; P = .04, paired t test; example seen in Figure 5B).

In order to assess the function of primarily annexin⁺ CD34 cells from patients with trisomy 8, colony culture assays for hematopoietic progenitors were performed using sorted annexin⁺ and annexin⁻ cells. Unseparated BM from 6 patients with trisomy 8, 12 healthy controls, and 4 patients with monosomy 7 was stained with annexin-FITC and CD34-PE and sorted into annexin⁺ and annexin⁻ CD34 cell fractions. The annexin⁺ CD34 cells were primarily trisomy 8 (Table 1). The annexin⁺ BM from patients with trisomy 8 showed comparable colony growth in methylcellulose culture to the annexin⁻ marrow (Figure 6). Control annexin⁺ BM from healthy donors and patients with monosomy 7 (data not shown) did not give rise to any colonies.

The relative resistance of trisomy 8 cells and normal diploid cells to apoptosis under stress conditions was examined using gamma radiation. Lymphocyte-depleted BMMNCs from 2 patients with trisomy 8 were split into 2 aliquots. The cells were either nonirradiated (0 Gy) or irradiated with γ rays from a ¹³⁷Cs source at a dose of 25 Gy. Cells were subsequently placed in short-term methylcellulose culture for 2 weeks as previously described,²⁵  or in liquid culture for 1 hour. The TUNEL assay was performed on samples in liquid culture, which were subsequently sorted based on TUNEL staining and then subjected to FISH. FISH was also performed on the colonies obtained from the irradiated and nonirradiated cells in methylcellulose culture. Samples from the TUNEL⁻ fractions consistently demonstrated more trisomy 8 cells than the positive fractions indicating resistance to apoptosis in the trisomy 8 population (Figure 7A). In short-term methylcellulose culture, the proportion of trisomy 8 cells in the irradiated aliquot was substantially increased compared with the nonirradiated aliquot (4% and 10% vs 29% and 40%, respectively; Figure 7B).

In order to confirm the importance of survivin in blockade of apoptosis in trisomy 8 cells, siRNA for survivin was used to “knock down” survivin expression, and immunoblot was performed to confirm decreased survivin expression. Control and siRNA-transfected BMMNCs from 2 patients with trisomy 8 were cultured for 36 hours in hematopoietic growth factors. Cell viability and count were obtained, and FISH for chromosome 8 was performed. Culture of BMMNCs with c-myc inhibitor Int-H1-S6A, F8A resulted in specific loss of trisomy 8 cell viability (Figure 7C). Similarly, knockdown of survivin resulted in decreased trisomy 8 viability without affecting diploid cells (Figure 7D).

In previous studies, we demonstrated that trisomy 8 cells showed a proliferative advantage over cells of normal karyotype from the same patient when cultures were depleted of lymphocytes or when Fas antagonist was added.²  We also found that caspase 8 was activated preferentially in trisomy 8 cells,²  and demonstrated that patients had cytotoxic T cells that appeared to be specific for the trisomy 8 clone. In the current work, we show that trisomy 8 cells have up-regulation of c-myc and CD1 by real-time PCR and further demonstrate that survivin, an inhibitor of apoptosis (IAP), is up-regulated. Knockdown of survivin and inhibition of c-myc resulted in selective loss of viability of the trisomy 8 cells. We go on to show that these cells fail to undergo DNA degradation and cell death, despite evidence of apoptosis. Paradoxically, the annexin⁺ cells had a comparable capacity for colony formation to annexin⁻ cells. Trisomy 8 cells appeared to be less susceptible to the apoptosis resulting from ionizing radiation than diploid cells and were also less susceptible to the effects of growth factor withdrawal (data not shown). Knockdown of survivin, an IAP, eliminated the survival advantage of trisomy 8 cells in culture.

Under normal conditions, apoptosis has been considered inevitable once the process is initiated and annexin is bound; however, some leukemia cells and cardiomyopathic cardiac myocytes show evidence of uncoupling of caspase 3 activation from the late stages of apoptosis of DNA disintegration and cell death.^23,24  Uncoupling of caspase 8 activation and changes in mitochondrial membrane potential and cell death has also been ascribed to activation of IAPs such as survivin in some leukemia cells.²⁵  In trisomy 8 cells, increases in c-myc might be expected to result in up-regulation of survivin²⁶  as well as cyclin D protein synthesis rates.²⁷  C-myc is located on chromosome 8, and its overexpression in trisomy 8 is likely a consequence of a “dosage effect” also described previously in studies of trisomy 8 leukemias.²⁸  Overexpression of c-myc, and subsequently survivin^29,30  and CD1,^31,32  would be expected to act to increase the proliferation and survival of trisomy 8 cells and could explain the apparent survival advantage of trisomy 8 cells in a bone marrow under immune attack. Levels are all increased in our trisomy 8 patients while they were not increased in patients with MDS who had other cytogenetic abnormalities. There are a number of reasons why survivin expression may not have completely correlated with the up-regulation of c-myc and CD1. Other factors controlling surviving, including the cell's commitment to erythroid versus myeloid development,³³  and the exposure of cells to cytokines (survivin expression is up-regulated by hematopoietic cytokines thrombopoietin, Flt3 ligand, and stem cell factor).³⁴  Expression of survivin is also cell-cycle related, the highest being during G₂/M.

That trisomy 8 is one of the most common cytogenetic abnormalities evolving in aplastic anemia may be due to the relative proliferative advantage and resistance to apoptosis of trisomy 8 cells in BM under immune attack.

Conflict-of-interest statement: The authors declare no competing financial interests.

Correspondence: Elaine Sloand, National Heart, Lung and Blood Institute, National Institutes of Health, 9000 Rockville Pike, Bldg 10, Rm 7C108, Bethesda, MD 20892; e-mail: sloande@nhlbi.nih.gov.