Expression of Bcl-2-related genes in normal and AML ... - Nature

Leukemia (1999) 13, 1881–1892  1999 Stockton Press All rights reserved 0887-6924/99 $15.00 http://www.stockton-press.co.uk/leu

Expression of Bcl-2-related genes in normal and AML progenitors: changes induced by chemotherapy and retinoic acid M Andreeff1, S Jiang1, X Zhang1, M Konopleva1, Z Estrov2, VE Snell1, Z Xie1, MF Okcu1, G Sanchez-Williams1, J Dong3, EH Estey4, RC Champlin5, SM Kornblau1, JC Reed6 and S Zhao1 Departments of 1Molecular Hematology and Therapy, 2Bioimmunotherapy, 4Leukemia, 5Blood and Marrow Transplantation, and 3 Biomathematics, The University of Texas MD Anderson Cancer Center, Houston, Texas; and 6Burnham Institute, La Jolla, California, USA

The expression of Bcl-2 family members was examined in normal and leukemic hematopoietic cells. Immature hematopoietic progenitor cells (CD34+/33−/13−) did not express Bcl-2 but BclXL, the majority of CD34 cells expressed Bcl-2, Bcl-XL and BAD, and normal promyelocytes (CD34−/33+) lacked expression of both Bcl-2 and Bcl-XL, while leukemic CD34+ progenitors and promyelocytes expressed these anti-apoptotic proteins. In AML, Bcl-2 expression was higher on CD34+ than on all AML cells, however, expression of Bcl-2 or Bcl-XL did not predict achievement of complete remission. Surprisingly, low Bcl-2 content was associated with poor survival in a group of patients with poor prognosis cytogenetics. The anti-apoptotic BAD protein was found to be expressed in AML, but was phosphorylated in 41/42 samples. Phosphorylation was found at both sites, Ser 112 and Ser 136. During induction chemotherapy, Bcl-2 levels of CD34 cells increased significantly. In the context of evidence for small numbers of leukemic CD34+ cells expressing very high levels of Bcl-2 prior to therapy, this finding is interpreted as a survival advantage of Bcl-2 overexpressing progenitors and rapid elimination of cells with low Bcl-2. Bcl-2 and Bcl-XL were both expressed in minimal residual disease cells. Downregulation of Bcl-2 mRNA and protein was observed by ATRA and the combination of Ara-C, followed by ATRA, resulted in markedly increased cytotoxicity in HL-60 cells, as compared to Ara-C alone or ATRA followed by AraC. Implications of these findings for the development of new therapeutic strategies for AML are discussed. Keywords: apoptosis; AML; Bcl-2; Bcl-XL; BAD; minimal residual disease

Introduction Programmed cell death (apoptosis) is controlled by an intrinsic genetic program remarkably conserved in evolution. All currently available cytotoxic drugs induce tumor death by triggering apoptosis, but many tumors have defects in the regulation of genes that control the programmed cell death process, thus rendering them resistant to chemotherapeutic agents. The members of the Bcl-2 family are critical regulators of the cell death pathway.1 In gene transfection experiments, overexpression of Bcl-2 and its homolog Bcl-XL have been shown to render neoplastic cells resistant to induction of apoptosis by a variety of chemotherapeutic drugs.2–6 Likewise, downregulation of Bcl-2 protein has been shown to reverse chemoresistance in several experimental systems.7–10 Chemosensitization has also been achieved by overexpression of proapoptotic proteins such as Bcl-XS and Bax.11,12 A schematic

Correspondence: M Andreeff, The University of Texas MD Anderson Cancer Center, Department of Molecular Hematology and Therapy, 1515 Holcombe Boulevard, Box 81, Houston, Texas 77030, USA; Fax: (713) 794-4747 Received 12 August 1998; accepted 12 July 1999

diagram of the apoptotic pathway is shown in Figure 1. The assignment of pro- and anti-apoptotic function to particular Bcl-2 family members oversimplifies some of the complex alterations in this pathway: Bcl-2 can be cleaved by caspase 3 and becomes pro-apoptotic, and Bax can mutate and loose its death-promoting activity.13 The mechanism by which Bcl-2 and Bcl-XL promote survival is not yet understood. Current data suggest the role of competitive dimerization between selective pairs of antagonists and agonists. Homodimerization of Bcl-2 and Bcl-XL results in strong anti-apoptotic function, while heterodimerization of the proapoptotic members Bax and BAD with Bcl2 and Bcl-XL counteracts this function and shifts the balance towards apoptosis.14,15 Several recent experiments have indicated that post-translational modification of Bcl-2-related proteins can play an important regulatory role as well. Phosphorylation of Bcl-2 on serine residues appears to inhibit its anti-apoptotic function in some contexts.16 In particular, the ability of chemotherapeutic agents that act on microtubules (eg taxol, vinblastine and vincristine) to induce phosphorylation of Bcl-2 during the G2– M phase has been implicated as a mechanism by which these anti-cancer drugs promote cell death.16,17 Conversely, serine phosphorylation of BAD reportedly disrupts its ability to bind to Bcl-XL, thus negating the proapoptotic function of BAD.18,19 In addition to dimerization with other Bcl-2 family members, Bcl-2 can bind to a p53-binding protein, p53-BP2, conceivably explaining how overexpression of Bcl-2 can interfere with the reported translocation of p53 from the cytosol into the nucleus in some types of cells.20,21 Furthermore, it has been demonstrated that Bcl-2 not only inhibits apoptosis but also restricts cell cycle entry22–24 and that these two functions can be genetically dissociated.25 This effect could provide additional cytoprotection, because proliferating cells are in general more vulnerable to apoptotic stimuli than nonproliferating cells. It was recently reported that the most primitive hematopoietic precursors express Bcl-XL but not Bcl-2.26 Bcl-2 is universally expressed in AML progenitors cells, and a subset of patients may have higher levels of expression than is found in normal CD34+ cells.27,28 High levels of Bcl-2 protein have been reported to be of independent prognostic significance in AML. A high percentage of Bcl-2+ AML blasts correlated with failure to achieve complete remission (CR) and with shorter overall survival in one study.29 Higher levels of Bcl-2 or higher ratios of Bcl-2:Bax have been reported to correlate with failure to achieve CR and with relapse by univariate analysis in other studies of adult AML.30–32 Identifying strategies for overcoming the anti-apoptotic effects of Bcl-2 and Bcl-XL is a challenging goal of translational cancer research. Inhibition of Bcl-2 function might have a more pronounced effect on neoplastic than on normal cells for reasons related both to its anti-apoptotic and anti-cell proliferative functions. For example, since the loss of cell

Expression of Bcl-2-related genes in normal and AML progenitors M Andreeff et al

1882

Figure 1 Pathways controlling apoptosis and necrosis. Activation of death receptors, DNA damage, growth factor loss, radio- or chemotherapy can result in activation of upstream caspases, activation of mitochondria, release of cytochrome c, activation of Apaf-1, subsequent activation of downstream caspases, and finally DNA fragmentation and apoptosis. The central role of anti-apoptotic Bcl-2 family members (Bcl-2, Bcl-XL) and of inhibitors like IAP (inhibitors of apoptosis proteins) is demonstrated. Mitochondrial activation resulting in release of Ca++, generation of free radicals, lipid peroxidation and ATP-depletion leads to necrosis (adapted from John Reed).

cycle control mechanisms in neoplasia drives cells into the cycle in spite of drug-induced damage,33 tumor cells which lack the G1 cell cycle checkpoint regulated by p53, might proceed into cycle in the absence of Bcl-2, thus putting themselves at risk for programmed cell death. It has been reported that retinoids downregulate the expression of Bcl-2. We were therefore interested how ATRA would affect the chemosensitivity of AML cells, since this retinoid is commonly used in the treatment of acute promyelocytic leukemia. Antisense oligonucleotides (ODN) targeted against Bcl-2 have been employed to induce apoptosis of malignant cells or sensitize them to conventional chemotherapeutic drugs,7–9 implying that retinoid-induced decreases in Bcl-2 ought to be similarly effective. The first clinical study of Bcl-2 antisense therapy in relapsed patients with relapsed and refractory non-Hodgkin lymphoma has recently been reported with encouraging results.34 The activity of Bcl-2 was also selectively abrogated and drug-mediated cytotoxicity was enhanced by an intracellular anti-Bcl-2 single-chain antibody.35 However, this approach necessitates delivery of antibody expression vectors into the target cells, making clinical applications for AML unlikely. Further elucidation of the molecular mechanisms of Bcl-2 function may therefore create the basis for the therapeutic correction of this disease-mediating dysfunction of cell death control. We report here our results on studies of the expression of Bcl-2 family members in normal and leukemic hematopoietic progenitor cells, during induction chemotherapy, and in the minimal residual disease state. We also investigated the effects of retinoids on Bcl-2 expression and explored the cytotoxicity of retinoids in combination with Ara-C in AML cells in vitro. Materials and methods

Cells Bone marrow or peripheral blood cells for in vitro studies were obtained from patients newly diagnosed with AML or

relapsed patients after obtaining informed consent according to institutional policy. The mononuclear cells were separated by Ficoll–Hypaque (Sigma Chemical, St Louis, MO, USA) density gradient centrifugation. The clinical features are listed in Table 1. Morphological subtypes were determined according to FAB classification. Normal bone marrow and G-CSF-mobilized progenitor cells from pheresis samples were separated by magnetic cell sorting (MACS). The pheresis samples were resuspended to 50 ml using cold RPMI medium and two ‘soft-spins’ were performed (centrifuge at 200 g for 10 min) to remove platelets. Bone marrow samples were Ficolled and washed twice. Cells were mixed with magnetic beads coated with anti-CD34 antibody (No. 467–01; Miltenyi Biotec, Auburn, CA, USA) according to the manufacturer’s instructions, except that one half the recommended volumes of reagents were used. A MACS buffer consisting of Hank’s buffered saline solution (HBSS) without calcium and magnesium (Cellgro Mediatech, Herndon, VA, USA), plus 0.6% ACD-A (Baxter, Deerfield, IL, USA), 0.5% BSA (Sigma) (pH 6.5) was used throughout staining and separation to prevent clumping of cells while maintaining optimal progenitor viability. Cells were separated on VS-positive selection columns (No. 413–06; Miltenyi Biotec) using a VarioMACS (Miltenyi Biotec) according to the manufacturer’s instructions and as previously described.36 The exception to these methods was that after washing, cells were backflushed to the top of the column outside of the magnet to increase purity rather than running the positive fraction over a second column. Cell purity was assessed by flow cytometry using CD34-PE and CD45-FITC as described by Sutherland et al37 and was found to be ⬎98%.

Flow cytometric analysis and cell sorting Different hematopoietic cell subpopulations (see Results) were FACS sorted using a FACS Vantage (Becton Dickinson, San Jose´, CA, USA) equipped with an argon ion laser (Coherent


Table 1

Pt. No.

FAB

A. Patients achieving complete 1 M1 2 RAEBT 3 M1 4 5 6 7 8 9 10 11

1883

Bcl-2 expression of myeloid leukemic cells and their CD34+ progenitors

M2 M7 M7 M5 M4 M4 M5 M1

Cytogenetics remission (CR) (20)46,xx (18)46,xy,t(8;21) (2)46,xy (2)49,xx,+8,t(9;22), +17,+21 (3)50,xx,+3,+8,t(9;22),+17,+21 (7)50,x,−x,+3,+8,t(9;22),+17,+21,+22q−(8)PH+ w/random chgs. (19)48,xy,+4,+8(1)46,xy (13)47,xy,+11 (7)46,xy (29)46,xx,t(9;22)(q34;q11) (29)46,xy (4)46,xy,del(16)(q21) (26)46,xy (20)46,xx (20)45,xy,t(4;12),−7 (21)45,x,−y, t(8;21)(q22;q22)(2)46,x,−y,t(8;21),+8

B. patients failing to achieve CR 1 M4 (20)43–44,xx,t(2;3),−7,+19,−22,+mar 2 M4 (19)45,xy,−7 3 M7 (15)43–45,xx,−2,−5,−7,−12,−13,−19,−22,+6–10mar. 4 M4 (2)43,xx,t(2;3)(q37;q21),−7,add(16)(q24), add(17)(p13),−18,−20 (7)41–46,x,−x w/common abn:t(2;3)(q37;q21),−7,+9,−10,−11,− 3add(16)(q24),add(17)(p13),−18,−20,add(21)(p11),+r,+1−2mar 5 M2 (3)46,xx 6 M2 (20)46,xx,inv(9)(p11q12) 7 M4 (20)46,xy,t(9;11)(p22;q23) 8 M4 (13)46,xy(6)49,xy,+9,+13,+15(1)47,xy,+9 9 M2 (19)45,xy,inv(3)(q25;q27)del(5)(q23;q34),−7(1)46,xy 10 CMML (20)46,xx 11 M2 (20)46,xy 12 M4 (19)47,xy,+8 (1)46,xy 13 M4 11/30/94:(9)46,xy,add(11)(q25),der(12;18)(q10;q10) 14 M1 (13)46,xy,del(7)(q22) (2)46,xy,del(7)(q22),del(9)(q22)(5)46,xy 15 M4 (20)46,xy 16 M4 (17)46,xx,i(17)(q10) (2)48,xx,+13,i(17)(q10),+19

ABC total

ABC CD34

133809 3278 25246

146006 5277 25918

6678 6858 16409 25156 69153 40055 64043 33218

7500 14055 19099 27401 71598 44358 84772 35055

30752 39636 6142 26771

32577 52357 4194 26861

27131 113680 9988 38475 2107 719 59161 37270 9453 47752

42670 121861 12363 44171 2434 2507 75180 39822 12827 49551

15164 35009

20721 41734

Bcl-2 expression was determined by quantitative flow cytometry and expressed as ABC = antibody binding capacity/per cell. ABC total = ABC for the entire population, ABC CD34 = ABC of gated CD34 cells. Bcl-2 expression of CD34 AML cells is significantly higher than that of the entire population (for details, see text).

Enterprise, Palo Alto, CA, USA) operated at 488 nm and 300 mW. Fluorescent signals were detected by 530/30 nm, 575/26 nm bandpass and 640 nm long pass filters. Cells were kept on ice during the sorting procedure and used subsequently for RT-PCR or fixed for FISH analysis.

Fluorescence in situ hybridization (FISH) FISH analysis of FACS sorted cells was performed as described previously.38

RNA isolation and reverse transcription RNA was isolated according to the single-step acid guanidinium thiocyanate-phenol-chloroform method. A reversetranscription kit was used to synthesize cDNA, according to the manufacturer’s instructions (Boehringer-Mannheim, Indianapolis, IN, USA). One microgram of the total RNA template was used per 10 ␮l of reverse transcriptase reaction. Primers for Bcl-2, Bcl-X, Bax and ␤2-microglobulin were synthesized on an Applied Biosystems oligonucleotide synthesizer (Applied Biosystems, Foster City, CA, USA; Model 392). One-

tenth of the RT reaction was used for each PCR reaction. The total volume of PCR is 25 ␮l. Bcl-2, Bcl-XL, Bax and ␤2-M cDNAs were amplified simultaneously in separate 0.2 ml MicroAmp tubes (Perkin Elmer, Foster City, CA, USA). The reaction conditions were as follows: 1.0 ␮M primers (each), 0.2 mM dNTP, 2 ␮Ci 32P-␣-dCTP, 1 × PCR reaction buffer (Perkin Elmer) and 1.5 U AmpliTaq DNA polymerase (Perkin Elmer). The thermal cycling was performed in a GeneAmp PCR system 9600 (Perkin Elmer) using the following parameters: 29 cycles of 94°C for 30 s, 63°C for 30 s, 72°C for 45 s, followed by a 10 min post-extension step at 72°C. Following thermal cycling, 3.0 ␮l of the PCR produced was analyzed on an 8% polyacrylamide gel. Electrophoresis was performed for 3 h at 300 volts. The radioactive products were detected and quantitated on the Betascope-603 (Betagen, Waltham, MA, USA) for 30 min, followed by exposure on Kodak (Rochester, NY, USA) film at −70°C for 4–8 h without an intensifying screen. All experiments were performed in duplicate.

Immunoblotting Cells were lysed in protein lysis buffer. Equal amounts of protein lysate were placed on 8% SDS-PAGE for 2 h at 100 volts


1884

followed by transfer of the protein on the Nytran membrane (S&S, Heween, NH, USA). Immunoblotting was performed at room temperature for 2 h with 5% milk, incubated with the first antibody in 1:1000 dilution for another 2 h, followed by three washes in PBS. The procedure was repeated for the secondary antibody followed by soaking the blot in ECL buffer for 1 min and then exposing it to the ECL films. The antibodies to BCL-XL and Bax were used at 1:1000 dilution.39,40 Immunoblotting for BAD was carried out as follows: the protein from the lysed cell was loaded on 12% SDS-PAGE gel. Then, the protein was transferred to the nitrocellulose membrane. The blot was blocked in 5% milk overnight, and then incubated with specific BAD and BAD phosphorylation (S-112 and S-136) antibodies (BioLabs, Beverly, MA, USA). The secondary antibody was used for 1 h, followed by three washes with TBS buffer. ECL plus was used for signal detection.

instead of TdT enzyme into the reaction buffer. A positive control slide was prepared from peripheral blood lymphocytes treated with 1 ␮M dexamethasone for 4 h.

Statistical analysis All results are reported, where applicable, as the mean ± standard error of mean (s.e.m.). Statistical analysis between paired data was performed using the paired Student’s t-test. Generalized linear model was used for analyzing the repeated measurements of Bcl-2. Results

Expression of Bcl-2 in myeloid leukemic cells and their progenitors Detection of Bcl-2 protein by flow cytometry AML blast cells were first stained with PE or PerCP-conjugated anti-CD34 monoclonal antibody (Becton Dickinson) and used at 1/10 dilution. Staining procedure was performed as described elsewhere. Cells were fixed in 1% formaldehyde (Sigma) and permeabilized with 10% Triton X in PBA (1% albumin, 0.1% NaN3) for 10 min at 4°C. The cells were washed twice in PBS. Appropriate PE or PerCP-conjugated lgG1-stained cells were incubated for 15 min at 4°C with 10 ␮l of FITC-conjugated mouse IgG1 control antibody (DAKO, Carpinteria, CA, USA). All assays were performed with a single lot of FITC-conjugated antibody, therefore samples were analyzed with the same coupling ratio of FITC/antibody. Cells stained with CD34-PE or PerCP antibody were incubated with 10 ␮l of FITC-conjugated monoclonal mouse anti-human bcl2 protein (DAKO bcl-2) lgG1. The cells were washed in PBS and analyzed immediately using the FACSCAN flow cytometer equipped with an argon (488 nm) laser and Lysys software (Becton Dickinson). Appropriate single-color stained cells were used for compensation control. for each sample, at least 10 000 cells were analyzed; cells were live-gated according to CD34- and/or Bcl-2-positivity, and the mean channel number was used for comparison of Bcl-2 expression by AML blasts (Figure 2). The relative channel number (RCN) was measured from the upper limit of the negative control. The relative channel number of a series of calibrated FITC microbeads having levels of fluorescence intensity ranging from 8.9 × 103 to 206.1 × 103 antibody-binding capacity (ABC) of equivalent soluble fluorochrome per bead was calculated (FCSC Quantum; Becton Dickinson) and a standard curve constructed. The RCN and the number of ABC per cell for the test samples were calculated using this standard curve.

Apoptosis Apoptotic cells were detected by a DNA fragmentation assay using ApoTag Plus detection kit (Oncor, Gaithersburg, MD, USA) as recommended by the manufacturer. Briefly, sorted cells were fixed in 1% paraformaldehyde. After washing and applying equilibration buffer, the cells were incubated with TdT in a humidified chamber at 37°C for 1 h. After washing, cells were stained with anti-digoxigenin-fluorescein for 30 min at room temperature. Thereafter, slides were washed, counterstained with DAPI and viewed by Nikon UFX-II fluorescence microscope. For negative control, water was mixed

As shown in Table 1, 27 samples were analyzed by quantitative flow cytometry for Bcl-2 expression within all cells (ABC total) and within CD34+ cells (ABC CD34). The correlations between Bcl-2 mRNA detected by RT-PCR, and protein determination by either immunoblot or quantitative flow cytometry (FCM) are shown in Figure 3. In AML, the FCM Bcl-2 values ranged from 2107 to 146 006 ABC/cell. The lowest value recorded was 719 ABC using cells from a patient with CMML. In all cases (except patient No. 3 in Table 1), CD34 cells expressed higher levels of Bcl-2 as compared to the total population (P = 0.000061). This was true for patients who achieved CR (P = 0.019) and who failed to achieve CR (P = 0.002). In Table 2, results are analyzed with regard to achievement of CR vs failure to achieve CR. There was no significant difference in Bcl-2 expression either when measured in all cells (P = 0.57) or in CD34 cells between the CR and non-CR groups (P = 0.60), suggesting that Bcl-2 expression prior to treatment did not predict for response to induction chemotherapy. No apparent correlation was found between Bcl-2 expression and FAB classification or cytogenetic group, within the limitations of this sample size.

Expression of Bcl-2 family members in AML Bcl-2 family members with pro-apoptotic (Bax, Bcl-XS, Bak, Bad) and anti-apoptotic (Bcl-2, Bcl-XL, MCL-1) function were analyzed in newly diagnosed AML (n = 48). Figure 4 (upper panel) shows the comparison between RT-PCR for Bax, Bcl2, Bcl-XL, Bcl-XS (left) and immunoblot analysis (right) of the same samples. In general, there was good correlation, although the anti-Bcl-X antibody used did not detect Bcl-XS, which was expressed at very low levels as determined by RTPCR, and some discrepancies in the mRNA and protein levels were apparent (eg lanes 1 and 7). This suggests that the BclXS protein may be unstable or its levels regulated by posttranslational mechanisms. Figure 4 (lower panel) shows Bcl2, Bcl-XL, and Bax levels, as determined by immunoblotting in all 48 patients studied. Bcl-2 was expressed in essentially all samples, although at different levels, which was consistent with the results obtained by quantitative flow cytometry. BclXL was likewise expressed in essentially all samples. Bax was also expressed at different levels in all samples. The control cell line, MO7e, had low levels of Bcl-2 but high Bcl-XL expression. In this series, no correlation between the expression of these Bcl-2 family members and FAB classi-


1885

Figure 2 Correlated analysis of Bcl-2 (FITC) and CD34 (PE). The AML sample shown here has 1.5% CD34+ and 98.5% CD34− cells. To obtain more information on the CD34+ population, this region was life-gated (upper right). Bcl-2 histograms are shown in the lower panel and the respective ABCs, as described in Materials and methods were calculated (CD34− cells: 3300 ABC; CD34+ cells: 16 900 ABC).

Figure 3 Comparison of methods to determine Bcl-2 expression in four samples (1–4) of newly diagnosed AML. RT-PCR (left), immunoblot and determination of antibody binding capacity by quantitative flow cytometry all demonstrate pronounced heterogeneity among samples.

fication, sex, cytogenetic group, age, or achievement of CR was found for patients receiving Ara-C/anthracycline-based therapy. Univariate and Cox regression analysis identified high Bcl2 expression in association with better overall survival (P = 0.04), after cytogenetics and achievement of CR. These results were corroborated in a separate group of patients with unfavorable cytogenetics (n = 87): in this series, low Bcl-2 levels were associated with shorter survival (P = 0.003).41

This unexpected result raised the question whether Bcl-2 in AML is inactivated by phosphorylation. In a series of 20 AML samples, no evidence of Bcl-2 phosphorylation was detected by immunoblot analysis, where slower migration of Bcl-2 protein is the expected result (C Croce, M Andreeff, personal communication). Expression of BAD was found in 84% of AML (Figure 5a). In 41/42 samples analyzed, the BAD protein migrated as a doublet, indicating the presence of phosphorylated BAD.


1886

Phosphorylation of BAD is known to inactivate this proapoptotic gene and this finding suggests widespread phosphorylation of BAD protein in AML, although to varying degrees. We then examined the phosphorylation sites of BAD, using antibodies specific for phosphorylated serine 112 and serine 136 (Figure 5b). In 10/11 AML, SER 136 and in 11/11 SER 112 was phosphorylated, confirming the results shown in Figure 5a.

a lesser extent in CD34+/13+ cells, but was absent in cells differentiated beyond the CD34 compartment (CD34−/33+). BclXS was detectable only in the most immature compartment CD34+/33−/13−, which also showed mRNA for Bax. Interestingly, Bcl-2 was not detected in CD34+/33−/13− cells nor in CD34−/33+ cells. Expression was strong however in CD34+/13+ cells. A tentative schema of Bcl-2, Bcl-XL and Bax expression in normal and leukemic cells is shown in Table 3. The interesting finding here is that both anti-apoptotic genes, Bcl-2 and Bcl-XL are expressed in the majority of CD34 cells, but that the most immature compartment lacks expression of Bcl-2, while Bcl-XL is highly expressed. When cells leave the CD34 compartment and develop into promyelocytes (CD34−/33+), they loose both Bcl-2, Bcl-XL and Bax. Leukemic and normal promyelocytes are compared in Table 4 which shows the absence of Bcl-2, Bcl-XL and Bax expression in normal, as compared to the high expression of Bcl-2 and Bcl-XL in leukemic promyelocytes. As shown in Figure 5a, BAD is expressed in normal CD34 cells, in both phosphorylated and unphosphorylated forms; its expression in normal promyelocytes has not been studied.

Expression of Bcl-2, Bcl-XL, Bcl-XS and Bax in normal hematopoietic progenitor cells

Changes of Bcl-2 expression of AML progenitor cells during induction therapy

Different subpopulations of progenitor cells were FACS-sorted following enrichment by an immunomagnetic procedure into CD34+/33−/13−, CD34+/13+ and CD34−/33+ populations of ⬎95% purity. As shown in Figure 6, Bcl-XL was highly expressed in the earliest compartment CD34+/33−/13− and to

After having failed to demonstrate a prognostic importance of Bcl-2 for achievement of CR in bulk AML cells and in CD34+ AML progenitors, we examined changes of Bcl-2 levels during induction chemotherapy of patients with AML. Twenty-five patients were investigated at baseline, and on days 1, 2, 3, 4

Table 2

Bcl-2 expression and response to therapy

CR All cells CD34

Fail

38.5 ± 1.2 31.2 ± 7.0 43.7 ± 12.7 36.4 ± 7.7 ×103 ABC/cell

P 0.57 0.60

No significant difference in Bcl-2 expression of all and of CD34+ gated AML cells was found between patients who achieved complete remission and those who failed (summary of data shown in Table 1).

Figure 4 Comparison of Bcl-2 family genes in AML (RT-PCR and immunoblot). The upper left panel shows results from RT-PCR analysis of Bax, Bcl-2, Bcl-XL, Bcl-XS, and ␤-2 microglobulin as control. Bcl-XS is notably expressed at low levels, while expression levels for Bax, Bcl-2 and Bcl-XL varied. The upper right panel shows immunoblot analysis of the same samples. Again, expression is variable; Bcl-XS is not detected by the antibody used. There are discrepancies between RT-PCR and immunoblot analysis (lanes 1 and 7). Lower panel shows the expression of Bcl-2, Bcl-XL and Bax in 48 samples from patients with AML. Bcl-2, Bcl-XL and Bax are expressed in almost all samples, at different levels.


Table 3 Tentative schema of Bcl-2 Bcl-XL, and Bax expression in normal and leukemic cells

Normal

Leukemia

Figure 5 (a) Expression of BAD in AML, CML and normal CD34+ cells. The lower band seen by immunoblot analysis represents unphosphorylated, the upper band phosphorylated BAD (pBAD). Most samples display both BAD and pBAD. (b) Phosphorylation of BAD. Phosphorylation site-specific antibodies reveal phosphorylation on both, serine 112 and 136 in primary AML cells.

Bcl-2

Bcl-XL

Bax

CD34+33−13− CD34+13+ (very rare) CD34+33+13+ CD34−33+ (Promyel.)

− ++

++ +

+ +

++ −

++ −

+ −

CD34+13+ (MRD) CD34−33+ (APL)

+++

++

+

++

++

+

Different normal and leukemic cell populations were FACS-sorted based on their antigen expression and subsequently analyzed by RT-PCR. Of note, the most immature normal progenitor cells (CD34+33−13−) expressed high levels of Bcl-XL but no detectable Bcl-2. More mature CD34 cells (CD34+33+13+) express both, Bcl-2 and Bcl-XL, while normal promyelocytes have no detectable Bcl-2 and Bcl-XL mRNA. Leukemic CD34+13+ cells were FACS sorted from patients in CR with minimal residual disease (MRD) and exhibited high levels of both, Bcl-2 and Bcl-XL mRNA. CD34−33+ leukemic promyelocytes, express both anti-apoptotic genes. Results were obtained by RT-PCR as shown in Figure 6. Table 4 Bcl-XL

Leukemic promyelocytes express high levels of Bcl-2 and

Normal CD34−CD33+13− Leukemic CD34−33+

Bcl-2

Bcl-XL

Bcl-XS

− ++

− +++

+ +

Bcl-2 and Bcl-XL mRNAs were not detectable by RT-PCR in normal promyelocytes.

Figure 6 Expression of Bcl-2, Bcl-XL, Bax and Bcl-XS in normal hematopoietic progenitors. Normal bone marrow CD34+ cells were purified by MACS, and subsequently FACS-sorted into CD34+/33−/13−, CD34+/13+ and CD34−/33+ cell populations. RT-PCR analysis of Bax (right), Bcl-2 (middle) and Bcl-XL and XS (left lanes) demonstrates the absence of Bcl-2 in immature CD34+/33−/13− cells and in promyelocytes (CD34−/33+), expression of Bcl-XL in immature cells, but not in promyelocytes, and expression of Bcl-2 and Bcl-XL in CD34+/13+ cells.

of Ara-C and idarubicin-based induction chemotherapy. Seventeen patients (65%) achieved complete remission, four were resistant, and four failed to achieve CR for other reasons (Table 5). Their CD34 Bcl-2 baseline levels varied but did not predict for achievement of CR (P = 0.28). However, increasing Bcl-2 levels in CD34 cells were observed during subsequent days of chemotherapy. The increase can be described by the equation: y = 8.74 + 0.74 x day − 1.8F x F − 1.45 x R, where y is the square root of Bcl-2 mean channel fluorescence, x are days of therapy, and F = 1, CR = 0, R = 1. Figure 7 shows the

Bcl-2 increase over time for the three groups of patients. This effect was highly significant (P = 0.004) and did not differ between patients who subsequently achieved CR, failed or were classified as resistant to chemotherapy. This finding allows two interpretations: either chemotherapy increases Bcl-2 expression levels in AML cells, or chemotherapy eliminates progenitors with low Bcl-2 expression suggesting that increased Bcl-2 levels confer survival advantage. Sequential flow cytometric studies of CD34/Bcl-2 suggest that the latter explanation is correct, without eliminating the first one: as shown in Figure 8, the pretreatment Bcl-2 content of CD34 cells was intermediate (day 0). Only 210/10 000 cells expressed high levels of Bcl-2 (right upper quadrant). On day 2 of induction chemotherapy, when the circulating WBC had decreased from 80 × 106 cells/ml to 0.8 × 106/ml, a population of CD34+ cells expressing high levels of Bcl-2 was detected. The relative frequency of these cells had increased from 2% to 25%. ‘Lifegating’ of these cells in the pretreatment sample clearly identified them as preexisting, at a frequency of 1527/100 000. Induction chemotherapy therefore eliminated cells with lower Bcl-2 content and spared CD34 cells with very high Bcl-2 content.

Bcl-2 and Bcl-XL are overexpressed in minimal residual disease in AML If Bcl-2 and Bcl-XL confer survival advantage during induction chemotherapy, it might be possible that residual leukemic

1887


1888

Table 5

Sequential analysis of Bcl-2 expression of CD34 cells during induction chemotherapy of AML

No.

Response

Status

EFS

Survival

D0

D1

D2

D3

D4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

CR Fail CR Fail Fail Fail CR CR Resistant CR CR CR Resistant Resistant CR Resistant CR CR CR CR CR CR CR CR CR

D D D D D D D A D D D D D D A D D D A D D D D A D

37 8 11 13 2 6 98 126 29 100 71 112 41 48 164 49 24 49 134 126 35 35 22 78 86

41 8 11 13 2 6 97 133 29 100 71 112 41 48 164 49 24 49 134 126 36 35 22 185 85

54.7 67 47 8.1 14.3 69.6 113.6 41.6 36.8 48.6 257.1 14.9 49.3 100.5 11.7 126 68.2 95.9 159.1 127.7 34.9 193.6 67 30.3 133.1

39 88.9 60.7 36.7 76.2 76.7 235.4 61.6 35.6 34.8 247.8 4.5 53.9 122 16.7 113.5

195.5 124.9 50.1

304.2 69 51.4

334.6 83.7 52.2

67 62.6 250.7

90.3 88.6 195.5

130.9 226

171.2 48.5 204.1

52.4 52.9

61.3 166 20.4 122.1 43.3 138.5

54.4 59.7

122.1 245.6 149.1 79.1 149.6 93.5 49.6 144.6

200.7 118.8 36.2 130.3

204.5 149.9 200.6 54.6 35.3 28.2 43.8 127.2

148

74.8 62.4 129.6 657.9 215.6 96.6 112.9 7.4

Twenty-five patients with AML were followed before (day 0) and on days 1, 2, 3 and 4 of induction chemotherapy. CR, complete remission; resistant, resistant AML; fail, failure to achieve CR. Status: D, dead; A, alive. EFS, event-free survival (months); survival, survival from diagnosis (months). D0, pretreatment expression of Bcl-2 on CD34 cells (×103 ABC/cell), D1, 2, 3, 4, days 1–4 of chemotherapy.

cells, in patients in CR, share these molecular characteristics. We therefore FACS-sorted cells expressing the abnormal CD34+/13+ phenotype in four patients in CR with known monosomies of chromosome 7. The normal compartment CD34+ 13− was also FACS-sorted and samples were subjected to RT-PCR and FISH analysis. FISH demonstrated the presence of over 50% of cells with monosomy 7 in the CD34+ 13+ but not in the CD34+ 13− population.38 RT-PCR showed high levels of Bcl-2 (not shown) and Bcl-XL mRNA (Figure 9) in minimal residual disease, but not in the normal cells. Of note are also the altered Bcl-XL/XS ratios in samples at diagnosis and at relapse: the anti-apoptotic Bcl-XL/XS ratio was increased in minimal residual leukemic cells compared to newly diagnosed AML. Flow cytometric analysis of CD34, CD13, and Bcl-2 identified these cells in one patient: a population of CD34+/CD13+ leukemic cells with very high (two log) Bcl-2 content was identified, while CD34+/13− cells contained only intermediate levels of Bcl-2. The difference in Bcl-2 expression was approximately 10-fold between residual leukemic and normal cells.

ATRA downregulates Bcl-2 mRNA and protein in leukemic cells and enhances the cytotoxic effect of Ara-C We then examined the effect of all-trans retinoic acid (ATRA) on HL-60 leukemic cells, alone and in combination with AraC (Table 6). The ATRA concentration used was 10−6 M, as was the concentration of Ara-C. In these experiments, we also altered the sequence of ATRA and Ara-C, giving ATRA first followed by Ara-C and vice versa. Ara-C or ATRA or saline was added to exponentially growing HL60 cells at time 0. After

48 h, the drugs were washed out and cells were subsequently exposed to ATRA, Ara-C or RPMI medium. Bcl-2 mRNA levels were determined at 1, 2, 4, 8, 24 and 48 h after Ara-C or ATRA (right lower panel), Bcl-2 protein levels were measured by quantitative flow cytometry at 24 and 48 h following AraC or ATRA, and again at 24 and 48 h following washout and ATRA, RPMI, or Ara-C (right upper panel). The number of apoptotic cells was determined at these timepoints (left lower panel) as was the number of clonogenic cells, calculated as the number of colonies/6 × 104 cells (left upper panel). All experiments were carried out in duplicates or triplicates. Results are shown as means. As shown in Table 6, control conditions at 48 or 96 h had no effect on the number of apoptotic (3.3–4.8%) or clonogenic (565–515 colonies) cells. Bcl-2 mRNA levels decreased at 2 and 4 h after Ara-C to 70.7%±, but recovered to baseline levels at 24 and 48 h. ATRA induced downregulation of Bcl2 mRNA within 2 h to 苲45% of control, with sustained decreases at 24 and 48 h. Bcl-2 protein levels after ATRA treatment decreased to 76.3% at 48 h and to 55.4% at 72 h, with subsequent recovery at 96 h. ATRA followed by Ara-C had no additional effect on Bcl-2 protein levels. However, Ara-C followed by ATRA resulted in a sustained ⬎50% decrease in Bcl-2. The number of apoptotic cells was likewise highest for Ara-C followed by ATRA (25.6%), more than twice as high as those found when ATRA was followed by Ara-C (10.7%). Moreover, clonogenic assays also demonstrated that Ara-C followed by ATRA reduced the number of colonyforming units to 16% (91 colonies) of control at 96 h while ATRA followed by Ara-C reduced colony counts only to 40% (221 colonies) of controls. These results suggest that the sequence of Ara-C followed by ATRA exerts a maximum effect on Bcl-2 protein levels, reduction in clonogenic survival, and induction of apoptosis, while ATRA followed by ARA-C was less than half as effective.


Figure 7 Change in Bcl-2 expression of CD34 cells during induction chemotherapy of AML. Y axis: square root of Bcl-2 mean channel fluorescence, X axis: days. Day 0, pretreatment; days 1–4, days on chemotherapy. Results are shown for patients who achieved CR, failed to achieve CR or were classified as resistant. There was a significant increase in Bcl-2 expression of CD34+ cells from pretreatment values to day 4 (P = 0.004).

Discussion Proteins structurally related to Bcl-2 form a complex network that regulates important steps leading to or preventing apoptosis. Dysregulation may contribute to the accumulation of malignant cells, but also to the resistance of malignancies to therapeutic interventions. Hence, we attempted to better understand the role of these molecules in myeloid leukemia. When we determined Bcl-2 expression in newly diagnosed AML cells, a very large variation in expression was observed. Selective determination of Bcl-2 expression in CD34+ cells revealed significantly higher expression in leukemic CD34+

cells than in the entire leukemic population. This finding follows the paradigm of normal hematopoiesis, where Bcl-2 expression is maximal in CD34 cells, followed by downregulation as cells differentiate. We did not find gross evidence of phosphorylation of Bcl-2 in newly diagnosed AML. Surprisingly, no difference in Bcl-2 expression was found between patients who achieved CR and those who failed to do so. This finding suggests that other factors may be better determinants of chemosensitivity or resistance, but does not preclude a role for Bcl-2 in drug resistance, given the overall poor outcome of AML therapy. The development of multiparametric flow cytometric assays to determine Bcl-2 on selected progenitor cell populations opened opportunities for studying very small stem cell populations, which are not amenable to Western blot analysis. Our findings confirm those by Park et al,26 and show that normal CD34+/33− progenitors do not express Bcl-2 but have high Bcl-XL expression. This finding is of potential importance with respect to attempts to employ Bcl-2 antisense oligonucleotides therapeutically and suggests that the most primitive hematopoietic stem cells would be spared. Once cells leave the CD34+ compartment and differentiate to promyelocytes (CD34−/33+), Bcl2 and Bcl-XL are both downregulated in normal, but not in leukemic cells, thus providing a potential differential target. The finding that high Bcl-2 expression was associated with good survival was surprising. However, this was confirmed in a second series by Kornblau et al41 and appears related to the cytogenetic subgroup investigated: in patients with ‘good prognosis’ cytogenetics, low Bcl-2 levels were associated with good prognosis, while in poor prognosis cytogenetics, the reverse was found. Our data do not elucidate this finding. Possibly, other regulators of apoptosis are selectively expressed in patients with ‘poor prognosis’ cytogenetics and disrupt the canonical apoptotic pathway that is operational in patients with ‘good prognosis’ cytogenetics. The human CED4 homologue Apaf-1,42 which participates in cytochrome cdependent activation of caspase 3 (Figure 1), was found by us to be expressed in 58% of AML43 and will be examined in this context. An alternative explanation for the ‘paradoxical’ role of Bcl-2 in poor prognosis AML could be the antiproliferative effect of Bcl-2: leukemias with high Bcl-2 levels would have slower repopulation kinetics, which would result in longer survival. This hypothesis also remains to be tested. Results presented here for BAD show expression in almost all samples studied. Importantly, BAD is phosphorylated on Ser 112 and Ser 136 in essentially all samples studied. BlumeJensen et al44 demonstrated that P13-kinase and Akt-mediated phosphorylation of BAD on Ser 136 by cytokine receptor pro-

Figure 8 Survival/Selection of Bcl-2high/CD34+ leukemic cells during induction chemotherapy of AML. Serial flow cytometric studies of CD34/Bcl-2 before (d 0) and 2 days after chemotherapy demonstrate survival of pre-existing population of CD34+/Bcl-2 high cells which increase from 2.1% to 25.85%. Life-gating reveals that these cells exist already on day 0 (middle panel, for details, see text).

1889


1890

Figure 9 Bcl-XL is expressed in residual leukemic cells in CR. Four AML samples at diagnosis (dx) (left panel) express both, Bcl-XL and XS. When minimal residual leukemic cells (CD34+/13+) were FACS-sorted in two patients and compared to normal (CD34+13−), or pre-treatment cells, Bcl-XL expression and Bcl-XL/XS ratios increased (right panel). The clonality of CD34+13+ cells was determined by FISH, which showed monosomy of chromosome 7 in the CD34+13+ population.

Table 6

Effect of ATRA and Ara-C on HL-60 cells

Clonogenic assays, induction of apoptosis, and effects on Bcl-2 protein and Bcl-2 mRNA are shown. The sequence: Ara-C followed by ATRA results in the most significant downregulation of Bcl-2 protein, reduction of clonogenic cells and induction of apoptosis (for details, see text). Results shown are from experiments conducted in duplicates or triplicates and are expressed as means.

moted cell survival. Hence, BAD phosphorylation could result in inactivation of the pro-apoptotic function of BAD. Mechanistic studies to evaluate this hypothesis and to identify the relevant kinases are underway. When patients were monitored during induction therapy, a significant increase in Bcl-2 expression of leukemic CD34 cells was noted. This change was consistent, regardless of the patient’s response to therapy. This could either indicate induction of Bcl-2 by chemotherapy, or selection of Bcl-2-overexpressing cells during therapy. Serial studies in a small number of patients indicate that cells with very high Bcl-2 content are pre-existing, and have a survival advantage during the first

days of chemotherapy. Therefore, downregulation of Bcl-2 would be potentially advantageous. The putative cytoprotective function of Bcl-2 and Bcl-XL is further substantiated by the finding of expression of Bcl-XL and Bcl-2 in minimal residual leukemic cells in patients in remission. Again, these antiapoptotic genes could be potential therapeutic targets in AML in remission and antisense oligonucleotides directed against the translocation initiation site of Bcl-2 have been used by us to induce apoptosis in AML progenitor cells.45 ATRA was found to downregulate Bcl-2 mRNA in HL-60 cells, with subsequent declines in Bcl-2 protein levels and apoptosis.46,47 Therefore, we tested the hypothesis that ATRA


in combination with Ara-C would increase the efficacy of AraC therapy in vitro. Interestingly, there was no enhancement of Ara-C toxicity when ATRA preceded Ara-C. However, significant reductions in Bcl-2 protein levels, decreases in clonogenic survival and increases in induction of apoptosis were seen in vitro, when ATRA was given after Ara-C. The combination of Ara-C and ATRA (with and without G-CSF) has been tested in a large clinical trial and initial results indicated that ATRA + G-CSF increased survival of patients with AML.48 In that trial, ATRA was given before, during and after chemotherapy consisting of fludarabine, Ara-C and idarubicin. However, the final analysis of 215 patients demonstrated no beneficial effect of ATRA (+G-CSF) on CR rate, survival or event-free survival in poor prognosis AML or high-risk MDS.49 This could be due to the non-canonical role of Bcl-2 in poor prognosis AML, ie high levels of Bcl-2 conferring relatively good prognosis and downregulation by ATRA may be detrimental, or to the suboptimal sequence of ATRA preceding chemotherapy. The results presented here, if confirmed in samples from primary AML, suggest that clinical trials should be conducted in which ATRA follows chemotherapy in patients who could potentially benefit from the transcriptional down-regulation of this anti-apoptotic protein.

Acknowledgements This work was supported in part by grants from the National Institutes of Health (PO1 CA55164, PO1 CA49639, and CA16672) and from the Stringer Professorship for Cancer Treatment and Research (MA).

References 1 Reed JC. Bcl-2 family proteins: regulators of apoptosis and chemoresistance in hematologic malignancies. Semin Hematol 1997; 34: 9–19. 2 Miyashita T, Reed JC. Bcl-2 gene transfer increases relative resistance of S49.1 and WEH17.2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res 1992; 52: 5407–5411. 3 Miyashita T, Reed JC. Bcl-2 oncoprotein blocks chemotherapyinduced apoptosis in a human leukemia cell line. Blood 1993; 81: 151–157. 4 Walton WI, Whysong D, O’Connor PM, Hockenberry D, Korsmyer SJ, Kohn KW. Constitutive expression of human bcl-2 modulates nitrogen mustard and camptothecin induced apoptosis. Cancer Res 1993; 53: 151–157. 5 Kamesaki S, Kamesaki H, Jorgensen TJ, Tanizawa A, Pommier Y, Cossman J. Bcl-2 protein inhibits etoposide-induced apoptosis through its effects on events subsequent to topoisomerase IIinduced DNA strand breaks and their repair. Cancer Res 1993; 53: 4251–4256. 6 Tang C, Willingham MC, Reed JC, Miyashita T, Ray S, Ponnathpur V, Huang Y, Mahoney ME, Bullock G, Bhalla K. High levels of p26BCL-2 oncoprotein retard taxol-induced apoptosis in human pre-B leukemia cells. Leukemia 1994; 8: 1960–1969. 7 Kitada S, Takayama S, DeRiel K, Tanaka S, Reed JC. Reversal of chemoresistance of lymphoma cells by antisense-mediated reduction of bcl-2 gene expression. Antisense Res Dev 1994; 4: 71–79. 8 Campos L, Sabido O, Rouault J-P, Guyotat D. Effects of Bcl-2 antisense oligodeoxynucleotides on in vitro proliferation and survival of normal marrow progenitors and leukemic cells. Blood 1994; 84: 595–600. 9 Berchem GJ, Bosseler M, Sugars LY, Voeller HJ, Zeitlin S, Gelmann EP. Androgens induce resistance to Bcl-2-mediated apoptosis in LNCaP prostate cancer cells. Cancer Res 1995; 55: 735–738. 10 Teixeira C, Reed JC, Pratt MAC. Estrogen promotes chemothera-

11

12 13

14 15

16 17

18 19 20

21 22 23

24 25 26 27 28

29

30

31

peutic drug resistance by a mechanism involving Bcl-2 protooncogene expression in human breast cancer cells. Cancer Res 1995; 55: 3902–3907. Bargou RC, Wagener C, Bommert K, Mapara MY, Daniel PT, Arnold W, Dietel M, Guski H, Feller A, Royer HD, Dorken B. Overexpression of the death-promoting gene bax-␣ which is downregulated in breast cancer restores sensitivity to different apoptotic stimuli and reduces tumor growth in SCID mice. J Clin Invest 1996; 97: 2651–2659. Sumantran VN, Ealogeva MW, Nunez G, Clarke MF, Wicha MS. Overexpression of Bcl-XS sensitizes MCF-7 cells to chemotherapyinduced apoptosis. Cancer Res 1995; 55: 2507–2510. Meijerink JPP, Mensink EJBM, Wang K, Sedlak TW, Sloetjes AW, de Witte T, Waksman G, Korsmeyer SJ. Hematopoietic malignancies demonstrate loss-of-function mutations of bax. Blood 1998; 91: 2991–2997. Sato T, Irie S, Krajewski S, Reed JC. Cloning and sequencing of cDNA encoding rat bcl-2 protein. Gene 1994; 140: 291–292. Sedlak TW, Oltvai ZN, Yang E, Wang K, Boise LH, Thompson CB, Korsmeyer SJ. Multiple Bcl-2 family members demonstrate selective dimerizations with Bax. Proc Natl Acad Sci USA 1995; 92: 7834–7838. Haldar S, Chintapalli J, Croce CM. Taxol induces bcl-2 phosphorylation and death of prostate cancer cells. Cancer Res 1996; 56: 1851–1854. Blagosklonny M, Schulte TW, Nguyen P, Mimnaugh EG, Trepel J, Neckers L. Taxol-induced apoptosis and phosphorylation of Bcl2 protein involves c-Raf-1 and represents a novel c-Raf-1 transduction pathway. Cancer Res 1995; 55: 4623–4626. Zha J, Harada H, Yang E, Jockel J, Korsmeyer SJ. Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-XL. Cell 1996; 87: 619–628. Wang H-G, Takayama S, Rapp UR, Reed JC. Bcl-2 interacting protein, BAG-1, binds to and activates the kinase Raf-1. Proc Natl Acad Sci USA 1996; 93: 7063–7068. Ryan JJ, Prochownik E, Gottlieb CA, Apel IJ, Merino R, Nunez G, Clarke MF. c-myc and bcl-2 modulate p53 function by altering p53 subcellular trafficking during the cell cycle. Proc Natl Acad Sci USA 1994; 91: 5878–5882. Naumovski L, Cleary ML. The p53-binding protein 53BP2 also interacts with Bcl-2 and impedes cell cycle progression at G2/M. Mol Cell Biol 1996; 16: 3884–3892. Borner C. Diminished cell proliferation associated with the deathprotective activity of Bcl-2. J Biol Chem 1996; 271: 12695–12698. Mazel S, Burtrum D, Petrie HT. Regulation of cell division cycle progression by bcl-2 expression: a potential mechanism for inhibition of programmed cell death. J Exp Med 1996; 183: 2219– 2226. Linette GP, Li Y, 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. Huang D, O’Reilly L, Strasser A, Cory S. The anti-apoptosis function of Bcl-2 can be genetically separated from its inhibitory effect on cell cycle entry. EMBO J 1997; 16: 4628–4638. Park JR, Bernstein ID, Hockenberry DM. Primitive human hematopoietic precursors express Bcl-x but not Bcl-2. Blood 1995; 86: 868–876. Bradbury DA, Russel NH. Comparative quantitative expression of bcl-2 by normal and leukaemic myeloid cells. Br J Haematol 1995; 91: 374–379. Porwit MacDonald A, Ivory K, Wilkinson S, Wheayley K, Wong L, Janossy G. Bcl-2 protein expression in normal human bone marrow precursors and in acute myelogenous leukemia. Leukemia 1995; 9: 1191–1196. Campos L, Rouault JP, Sabido O, Oriol P, Roubi N, Vasselon C, Archimbaud E, Magaud JP, Guyotat D. High expression of bcl-2 protein in acute myeloid leukemia cells is associated with poor response to chemotherapy. Blood 1993; 81: 3091–3096. Maung ZT, MacLean FR, Reid MM, Pearson AD, Proctor SJ, Hamilton PJ, Hall AG. The relationship between bcl-2 expression and response to chemotherapy in acute leukaemia. Br J Haematol 1994; 88: 105–109. Stoetzer OJ, Nussler V, Darsow M, Gullis E, Pelka-Fleischer R, Scheel U, Wilmanns W. Association of bcl-2, bax, bcl-xL and interleukin-1 beta-converting enzyme expression with initial

1891


1892 32 33 34 35

36 37

38

39 40

41

response to chemotherapy in acute myeloid leukemia. Leukemia 1996; 10 (Suppl. 3): S18–S22. Karakas T, Maurer U, Weidmann E, Miething CC. High expression of bcl-2 mRNA as a determinant of poor prognosis in acute myeloid leukemia. Ann Oncol 1998; 9: 159–165. Elledge SJ. Cell cycle checkpoints: preventing an identity crisis. Science 1996; 274: 1664–1667. Webb A, Cunningham D, Cotter F, Clarke PA, Stefano PA, Ross P, Corbo M, Dziewanowska Z. Bcl-2 antisense therapy in patients with non-Hodgkin lymphoma. Lancet 1997; 349: 1137–1141. Piche A, Grim J, Rancourt C, Gomez-Navarro J, Reed JC, Curiel DT. Modulation of bcl-2 protein levels by an intracellular antibcl-2 single-chain antibody increases drug-induced cytotoxicity in the breast cancer cell line MCF-7. Cancer Res 1998; 58: 2134– 2140. Miltenyi S, Muller W, Weichel W, Radbruch A. High gradient magnetic separation with MACS. Cytometry 1990; 11: 231–238. Sutherland DR, Keating A, Nayar R, Anania S, Stewart AK. Sensitive detection and enumeration of CD34+ cells in peripheral and cord blood by flow cytometry. Exp Hematol 1994; 22: 1003– 1010. Engel H, Goodacre A, Keyhani A, Jiang S, Van NT, Kimmel M, Sanchez-Williams G, Andreeff M. Minimal residual disease in acute myelogenous leukemia and myelodysplastic syndromes: a follow-up of patients in clinical remission. Br J Haematol 1997; 99: 64–75. Krajewski S, Krajewska M, Shabaik A, Wang HG, Irie S, Fong L, Reed JC. Immunohistochemical analysis of in vivo patterns of BclX expression. Cancer Res 1994; 54: 5501–5507. Krajewski S, Blomqvist C, Franssila K, Krajewska M, Wasenius VM, Niskanen E, Nordling S, Reed JC. Reduced expression of proapoptotic gene BAX is associated with poor response rates to combination chemotherapy and shorter survival in women with metastatic breast adenocarcinoma. Cancer Res 1995; 55: 4471–4478. Kornblau SM, Thall P, Estrov Z, Kantarjian H, Estey E, Andreeff M.

42 43

44

45

46

47 48

49

Proliferation and apoptosis regulating proteins and cytogenetics predict prognosis in acute myelogenous leukemia. Blood 1997; 90 (Suppl. 1): 308, 71a. Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 1997; 90: 405. Andreeff M, Zhao S, Xie Z, Konopleva M, Estey E, Munsell M, Sanchez-Williams G, Kornblau S, Reed J. Control of apoptosis in AML by Bcl-2 and Bcl-XL: critic impact of cytogenetics. Blood 1998; 92 (Suppl. 1): 3887. Blume-Jensen P, Janknecht R, Hunter T. The kit receptor promotes cell survival via activation of PI 3-kinase and subsequent Aktmediated phosphorylation of Bad on Ser136. Curr Biol 1998; 8: 779–782. Konopleva M, Tari A, Estrov Z, Harris D, Xie Z, Zhao S, Lo´pezBerestein G, Andreeff M. Liposomal Bcl-2 antisense oligonucleotides enhance proliferation, sensitize AML to cytosine-arabinoside, and induce apoptosis independent of other anti-apoptotic proteins. Blood (submitted). Konopleva M, Zhao S, Xie Z, Segall H, Younes A, Claxton DF, Estrov Z, Kornblau SM, Andreeff M. Apoptosis: molecules and mechanisms. In: Veerman AJP, Pieters R, Kaspers GJL (eds). Drug Resistance in Leukemia and Lymphoma III. Plenum Press: Amsterdam (in press). Konopleva M, Andreeff M. Regulatory pathways in programmed cell death. Cancer Mol Biol 1999; 6: 1229–1260. Estey E, Beran M, Pierce S, Kantarjian H, Keating M. All-trans retinoic acid (ATRA) may improve results of chemotherapy in poor prognosis non-APL AML and MDS: a randomized study. Blood 1997; 90 (Suppl. 1): 1851, 416a. Estey EH, Thall PF, Pierce S, Cortes J, Beran M, Kantarjian H, Keating MJ, Andreeff M, Freireich E. Randomized phase II study of fludarabine + cytosine arabinoside + idarubicin ± all-trans retionic acid ± granulocyte-colony stimulating factor in poor prognosis newly-diagnosed non-APL AML and MDS. Blood 1999; 93: 2478–2484.