Apoptosis Induction by Retinoids in Eosinophilic Leukemia Cells: Implication of Retinoic Acid Receptor- Signaling in All-Trans-Retinoic Acid Hypersensitivity

2006;66:6336-6344. Cancer Res Carine Robert, Laurent Delva, Nicole Balitrand, et al. All-Trans-Retinoic Acid Hypersensitivity

Signaling inαCells: Implication of Retinoic Acid Receptor-Apoptosis Induction by Retinoids in Eosinophilic Leukemia

Updated version

http://cancerres.aacrjournals.org/content/66/12/6336

Access the most recent version of this article at:

Cited Articles

http://cancerres.aacrjournals.org/content/66/12/6336.full.html#ref-list-1

This article cites by 56 articles, 24 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/66/12/6336.full.html#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on June 7, 2013. © 2006 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/content/66/12/6336

http://cancerres.aacrjournals.org/content/66/12/6336.full.html#ref-list-1

http://cancerres.aacrjournals.org/content/66/12/6336.full.html#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

mailto:[email protected]

http://cancerres.aacrjournals.org/

Apoptosis Induction by Retinoids in Eosinophilic Leukemia Cells:

Implication of Retinoic Acid Receptor-A Signaling in

All-Trans-Retinoic Acid Hypersensitivity

Carine Robert,1Laurent Delva,

1Nicole Balitrand,

1Sarolta Nahajevszky,

2

Tamas Masszi,2Christine Chomienne,

1and Bela Papp

1

1Institut National de la Sante et de la Recherche Medicale, UMR-S 718, Institut Universitaire d’Hematologie, University of Paris VII, Paris,France and 2National Medical Center, Department of Haematology and Stem Cell Transplantation, Budapest, Hungary

Abstract

Hypereosinophilic syndrome (HES) has recently been recog-nized as a clonal leukemic lesion, which is due to a specificoncogenic event that generates hyperactive platelet-derivedgrowth factor receptor-A–derived tyrosine kinase fusionproteins. In the present work, the effect of retinoids on theleukemic hypereosinophilia-derived EoL-1 cell line and onprimary HES-derived cells has been investigated. We showthat all-trans-retinoic acid (ATRA) inhibits eosinophil colonyformation of HES-derived bone marrow cells and is a powerfulinducer of apoptosis of the EoL-1 cell line. Apoptosis wasshown in the nanomolar concentration range by phosphati-dylserine externalization, proapoptotic shift of the Bcl-2/Bakratio, drop in mitochondrial membrane potential, activationof caspases, and cellular morphology. Unlike in other ATRA-sensitive myeloid leukemia models, apoptosis was rapid andwas not preceded by terminal cell differentiation. Use ofisoform-selective synthetic retinoids indicated that retinoicacid receptor-A–dependent signaling is sufficient to induceapoptosis of EoL-1 cells. Our work shows that the scopeof ATRA-induced apoptosis of malignancies may be widerwithin the myeloid lineage than thought previously, that theEoL-1 cell line constitutes a new and unique model for thestudy of ATRA-induced cell death, and that ATRA may havepotential for the management of clonal HES. (Cancer Res 2006;66(12): 6336-44)

Introduction

Molecularly targeted therapy of malignancies is a promising newconcept based on the selective pharmacologic modulation of keysignaling/transcription mechanisms specific to a given tumor/leukemia type. All-trans-retinoic acid (ATRA), the first clinicallyrelevant example of such treatment, targets the promyelocyticleukemia (PML)/retinoic acid receptor (RAR)-a fusion oncoproteinin acute promyelocytic leukemia (APL). APL blasts undergoterminal granulocytic differentiation on treatment with ATRAand are thereafter eliminated from the body by apoptosis, leadingto complete remission (1). ATRA treatment in combination withchemotherapy has dramatically improved the prognosis of APL (1).Despite the marked differentiation-inducing effect of the molecule

observed in various other in vitro solid tumor and leukemiamodels, ATRA treatment of malignancies is, however, currentlylimited to PML/RAR-a-positive APL in the clinic.

Recently, the identification of an interstitial deletion, del(4)(q12q12), on chromosome 4 that generates the FIP1-like-1/platelet-derived growth factor receptor-a (FIP1L1/PDGFRa)–transformingtyrosine kinase fusion protein in a significant subset of cases ofhypereosinophilic syndrome (HES) permitted to unequivocallyestablish the clonal, leukemic origin of these lesions (2). Inaddition, as the tyrosine kinase activity of the FIP1L1/PDGFR-afusion protein can be inhibited by molecules such as imatinibmesylate (STI571 or Gleevec), HES that expresses this fusionprotein can be successfully treated with imatinib mesylate (2–6).However, clinical experience has shown that treatment of sensitivetypes of leukemia with targeted therapy, such as imatinib mesylateor ATRA as a single agent, often leads to relapse and emergence ofresistant clones (7, 8), and, indeed, imatinib mesylate resistance inHES has already been described (2, 9).

Although certain sublines of the HL-60 and AML14 myeloidleukemia cell lines present some eosinophilic features (10, 11), theFIP1L1/PDGFR-a-positive (PML/RAR-a and BCR/ABL negative)EoL-1 cell line (12), established from an acute eosinophilic leukemiathat followed a chronic HES (13, 14), has recently emerged as a keymodel system for the investigation of the molecular and cellularbiology of FIP1L1/PDGFR-a-positive malignancies of the eosino-philic lineage (15). The EoL-1 cell line constitutes the only in vitrotool currently available for the study of this type of malignancy andhas recently been used by many laboratories for the investigation ofeosinophil-granulocytic differentiation (13, 16, 17) and the functionof the FIP1L1/PDGFR-a fusion protein (5, 15, 18).

ATRA can exert marked biological responses, including celldifferentiation and apoptosis, in several experimental models ofmalignancy, including myeloid leukemias (19). In addition, ATRAhas also a suppressive effect on normal eosinopoiesis ex vivo (20).However, the effect of ATRA on the behavior of neoplasticeosinophil cells has not been studied thus far. Therefore, in thepresent study, we investigated the effects of retinoids on EoL-1 cellsand on eosinopoiesis of HES-derived bone marrow ex vivo . Weshow that EoL-1 cells are exquisitely sensitive to retinoic acid andundergo extensive programmed cell death at very low doses ofATRA and, importantly, that ATRA also inhibits eosinophil colonyformation of bone marrow cells obtained from FIP1L1/PDGFR-a-positive HES.

Materials and Methods

Cell culture and treatments. The EoL-1 and HL-60 human cell lines

were obtained from the Deutsche Sammlung von Mikroorganismen und

Zellkulturen (Braunschweig, Germany) and American Type Culture

Requests for reprints: Bela Papp, Institut National de la Sante et de la RechercheMedicale, UMR-S 718, Institut Universitaire d’Hematologie, Hopital Saint-Louis, 16 ruede la Grange aux Belles, 75010 Paris, France. Phone: +33-1-53-72-40-10; Fax: +33-1-53-72-40-16; E-mail: [email protected].

I2006 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-06-0078

Cancer Res 2006; 66: (12). June 15, 2006 6336 www.aacrjournals.org

Research Article



Collection (Manassas, VA), respectively, and NB4 cells were kindly providedby Dr. Michel Lanotte (U685 Institut National de la Sante et de la Reserche

Medicale, Paris, France). Cells were grown in RPMI 1640 supplemented with

glutamax-1, 2 mmol/L glutamine plus 15% ( for HL-60) or 10% ( for NB4 and

EoL-1 cells) FCS (Life Technologies, Inc., Grand Island, NY) that supportedgrowth of EoL-1 cells as suspension of single cells without significant cell

clumping.

At the beginning of experiments, exponentially growing cells were

resuspended in complete medium at a density of 2 � 103/mL, and drugswere added from concentrated stock solutions made in DMSO. Final

concentration of DMSO did not exceed 0.1% and did not interfere with

the assays. ATRA and AM580 (4-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-

naphtamido)benzoic acid) were purchased from Sigma-Aldrich (Saint-QuentinFallavier, France), Ro-41 5253 (p-[(E)-2-[3¶4¶-dihydro-4¶4¶-dimethyl-7¶-(hepty-

loxy)-2¶H-1-benzothiopyran-6¶-yl]propenyl]benzoic acid 1¶,1¶dioxyde) was a

gift from Drs. Eva-Maria Gutknecht and Pierre Weber (Hoffmann-La RocheSA, Basel, Switzerland), and imatinib mesylate was provided by Novartis,

Inc. (Basel, Switzerland).

Cell counts and viability were determined by trypan blue exclusion using

a hemocytometer. Morphologic assessment of cells was made on cytospinpreparations stained with May-Grunwald-Giemsa (MGG).

Early apoptosis was detected on a FACScalibur flow cytometer (BD

Biosciences, Le Pont de Claix, France) and by Annexin V and propidium

iodide staining according to the instructions of the manufacturer (AnnexinV-FITC kit, Beckman Coulter, Marseille, France; ref. 21).

Mitochondrial membrane potential (DCm) was measured on a

FACScalibur flow cytometer by staining cells with 20 nmol/L 3,3¶-dihexyloxacarbocyanine iodide DiOC6(3) (Uptima Interchim, Montlucon,

France) for 30 minutes at 37jC in serum-free RPMI 1640 in the dark (22).

Surface marker expression was detected on a FACScalibur flow

cytometer by incubating 1 � 106 cells for 30 minutes at 4jC withmonoclonal antibodies directed against CD11b, CD11c, CD14, CD49d,

CDw125, or isotype-matched control antibodies (BD Biosciences). Detection

of intracellular interleukin (IL)-5Ra was done after fixation and permeabi-

lization with 0.1% Triton X-100 in PBS.Transfections. Thirty million EoL-1 cells in 250 AL Opti-MEM (Life

Technologies) were electroporated at 280 V, 950 AF with the hRAR-b2luciferase reporter gene (8 Ag; ref. 23), and, as control, the h-galactosidase

vector (pCH110; 1 Ag) using an IBI Gene Zapper 450/2500 apparatus

(Eastman-Kodak, New Haven, CT). Cells were then resuspended in complete

medium, and, 3 hours later, drugs were added. Assays for luciferase

(Luciferase Assay System kit, Promega, Charbonnieres, France) and h-

galactosidase (h-Gal Reporter Gene Assay kit, Roche, Fontenay-sous-bois,

France) were done on harvested cells after an overnight treatment with

drugs according to the protocols of the manufacturers. Results are

expressed as fold induction based on the relative luciferase activity

(arbitrarily set at 1) displayed in the absence of any drug. Experiments

were normalized to h-galactosidase (24).Immunoblotting and antibodies. Cells were quickly washed twice with

ice-cold NaCl (150 mmol/L), precipitated with 5% trichloroacetic acid

overnight at 4jC, and then centrifuged (25). Protein pellet was dissolved in

sample buffer (25), and equal amounts of lysates (50 Ag protein/well) wererun on SDS-PAGE and electroblotted onto nitrocellulose (Hybond enhanced

chemiluminescence (ECL), Amersham, Freiburg, Germany).

Blocking of nonspecific sites was done by incubating the membranes

for 1 hour with TBS supplemented with 0.1% Tween 20 (TBST) and 5%

nonfat dry milk (TBST-milk). Immunostaining was done by incubating the

membranes under agitation overnight at 4jC with primary antibodies

diluted in TBST-milk for caspase-3 (2,000-fold), caspase-8, caspase-9, and

Mcl-1 (1,000-fold; Cell Signalling, Beverly, MA); Bcl-2 (2,000-fold; DAKO,

Trappes, France); Bak (500-fold) and Bcl-x (2,000-fold; PharMingen, San

Diego, CA); Bax (2,000-fold), PDGFR-a (C-20), and P-Y754-PDGFR-a (1,000-

fold; Santa Cruz Biotechnology, Santa Cruz, CA); h-actin (4,000-fold) and

poly(ADP-ribose) polymerase (PARP; 2,000-fold; Sigma-Aldrich); and plasma

membrane calcium ATPase (PMCA; 1,000-fold; Ozyme, Saint-Quentin en

Yvelines, France). The 9a9A6 anti-RAR-a antibody, kindly provided by

Prof. Pierre Chambon and Dr. Cecile Rochette-Egly (Institut de Genetique

et de Biologie Moleculaire et Cellulaire, Illkirch, France) was diluted 500-fold

in TBST. After three washes in TBST-milk, membranes were incubated with

appropriate peroxidase-conjugated secondary antibodies for 30 minutes in

TBST-milk and washed. Signal was detected using the ECL system and

quantified by scanning nonsaturated luminograms using the ScionImage

software (version 4.0.2, Scion Corp.).3

Collagen gel culture of human bone marrow. Bone marrow samples

were obtained from residual biopsy material from healthy donors and from

a HES patient after informed consent. The patient, a 29-year-old man, wastreated during 4 months with 500 to 1500 mg/d hydroxyurea, and the

symptoms persisted. After 2 weeks of hydroxyurea withdrawal, a bone

marrow biopsy was taken for the detection of the FIP1L1/PDGFR-atranscript by reverse transcription-PCR and for cell culture. At this point, hepresented marked hepatosplenomegaly by ultrasonography, WBC:11.7 G/l,

neutrophil:4.76 G/l, basophil:0.02 G/l, monocyte:1.37 G/l, lymphocyte:1.63

G/l, and eosinophil:3.88 G/l. After detection of the FIP1L1/PDGFR-atranscript, treatment with 100 mg/d imatinib mesylate was implemented,resulting in disappearance of hepatosplenomegaly and other symptoms

within 1 week (WBC:2.2 G/l, Hgb:126 g/l, PLT:156 G/l, and eosinophil:0.00 G/l).

Mononuclear cells of normal and patient bone marrow were isolated bydensity gradient centrifugation on lymphocyte separation medium (density,

1,077; Eurobio, Courtaboeuf, France). Cells were grown in collagen gel

cultures using a modified protocol from the Megacult-C kit (Stem Cell

Technologies, Inc., Vancouver, British Columbia, Canada). Cells were platedfor the patient and normal donors at 1 � 105 and 2.5 � 104 per chamber of

double chamber slides, respectively, in RPMI 1640 supplemented with

Megacult-C medium, agar human leukocyte-conditioned medium with 10%

fetal bovine serum, and collagen in the presence or absence of ATRA asshown in the figures. Human recombinant IL-3, recombinant granulocyte

macrophage colony-stimulating factor (GM-CSF), and recombinant IL-5

(R&D Systems, Lille, France) were added at a final concentration of 10 ng/mL.Bone marrow cultures were made in duplicate, and colony-forming units

were enumerated on day 14 of culture after MGG staining or immuno-

histochemical staining (DAKO LSAB+ System, DAKO) with an antibody

directed against major basic protein (MBP; BMK13, Chemicon, Hampshire,United Kingdom) according to the instructions of the manufacturer.

The results presented in this work are representative of three

independent experiments.

Results

ATRA induces the death of EoL-1 cells. EoL-1 cells werecultured in the presence or absence of various concentrations ofATRA in the 1 nmol/L to 10 Amol/L range, and cell viability wasdetermined by trypan blue exclusion daily for 15 days. ATRA inducedcomplete cell death in the low nanomolar concentration range.Although untreated cells, as well as cells treated with 1 nmol/LATRA, continued to proliferate and maintain f100% viability, celldeath could be detected >10 nmol/L ATRA (Fig. 1A). The timecourse of cell death was dependent on the concentration of ATRA,with faster kinetics at higher doses of the drug. ATRA-induced celldeath proceeded somewhat more slowly than that induced byimatinib at maximally efficient concentrations (1 nmol/L).

To investigate further the exquisite sensitivity of EoL-1 cells toATRA, the dose-dependence of ATRA-induced EoL-1 cell death wasexplored in greater detail in the 1 to 10 nmol/L concentrationrange at day 12. Reduction of cell viability was clearly detectablealready at a concentration of 2 to 4 nmol/L of exogenously addedATRA, with only 20% viable cells remaining at day 12 in thepresence of 4 nmol/L of the drug (Fig. 1B).

ATRA treatment of EoL-1 cells induced cell clumping. UntreatedEoL-1 cells grew in suspension, predominantly as single cells, with

3 http://www.scioncorp.com.

Apoptosis in Hypereosinophilia

www.aacrjournals.org 6337 Cancer Res 2006; 66: (12). June 15, 2006



occasional small cell aggregates consisting of only a few (<10) cells.Treatment with various concentrations of ATRA induced theformation of large cell aggregates from day 1, which could reachthe size of several hundred cells (Fig. 5A, 1 and 2) and which couldbe dissociated by gentle pipetting. Aggregation was accompaniedby an almost complete loss of viable single cells in the culture, anddeath of cells proceeded within the aggregates and could bedetected by trypan blue staining (Fig. 5A, 3 and 4).

On ATRA treatment, the promyelocytic NB4 and myeloblasticHL-60 leukemia cell lines, which constitute the most widely usedand best characterized in vitro models of ATRA-induced cell differ-entiation within the myeloid lineage, undergo terminal neutrophilgranulocytic differentiation in 4 and 6 days, respectively (26, 27).Moreover, induction of neutrophil-granulocytic differentiationfollowed by apoptosis of promyelocytic blasts constitutes the basisof ATRA treatment of APL. To compare ATRA sensitivity ofEoL-1 cells to that of HL-60 and NB4, cells were incubated with100 nmol/L ATRA for 4 days, and viability was determined bytrypan blue exclusion. A faster and more marked death response ofEoL-1 cells could be observed when compared with HL-60 or NB4cells (Fig. 1C). The viability of HL-60 and NB4 cells remainedessentially unaltered, whereas the viability of EoL-1 cells decreasedto f50%, indicating that ATRA-induced cell death proceeds fasterthan that of HL-60 or NB4 cells.

To study morphologic changes that take place during ATRA-treatment of EoL-1 cells, cytospin preparations were made fromcells treated with 100 nmol/L drug for up to 7 days and stainedwith MGG. Untreated cells displayed a blastic morphology withlarge, round nuclei and high nuclear/cytoplasmic ratio (Fig. 5B, 1).During treatment, some cells showed a discrete modification in cell

size, diminution of nuclear/cytoplasmic ratio, and occasionalappearance of nuclear invaginations and vacuoles (Fig. 5B, 2-6).It is worth mentioning that remaining viable cells that persisted inculture in late stages of treatments (such as day 7; Fig. 5B, 6) didnot display signs of more advanced myeloid differentiation, such asindented or multilobed nuclei or specific cytosolic granules oreosinophilic staining patterns, and nitroblue tetrazolium cyto-chemical staining remained negative as well (data not shown).

To document more precisely the expression of proteins involvedin eosinophil differentiation, analysis of the modulation of theimmunophenotype of EoL-1 cells during ATRA treatment was doneby flow cytometry using CD11b (integrin aM subunit), CD11c(integrin aX subunit), CD49d (integrin a4 chain), CD14 (lipopoly-saccharide receptor), and CDw125 (IL-5Ra) specific antibodies andappropriate isotype-matched negative controls. ATRA treatmentfor 3 days led to a marked induction of CD11b and CD11c expres-sion in the plasma membrane, and this was accompanied by aparallel decrease of the CD49d level (Fig. 1D). Our results agreewith previous studies, showing that EoL-1 cells under normalculture conditions express a high level of CD49d and very littleCD11b and that the expression of CD49d decreases and CD11bincreases under differentiating stimuli (16, 28). On the other hand,ATRA had no effect on CD14 and intracellular or extracellular IL-5Ra expression. This agrees with previous observations, indicatingthat a small population of EoL-1 cells express a low level of CD14(17) and that IL-5, the major cytokine of the eosinophil lineage, isnot able to enhance the growth or induce the differentiation ofEoL-1 cells (13) and is also in accordance with our observationsthat IL-5 added to cultures did not rescue EoL-1 cells from ATRA-induced cell death (data not shown). Importantly, during ATRA

Figure 1. ATRA induces death of EoL-1cells. The percentage of viable EoL-1 cellstreated or not treated with the indicatedconcentrations of ATRA (continuous lines )or 1 nmol/L imatinib (dotted line ) wasassessed by trypan blue staining every dayduring 15 days (A ) and after 12 days oftreatment (B). C, the percentage of viableHL-60, NB4, and EoL-1 cells treated or nottreated by ATRA (10�7 mol/L) wasassessed by trypan blue staining after4 days of treatment. D, EoL-1 cells weretreated (white bars ) or not treated (blackbars ) with ATRA (10�7 mol/L). Cells wereharvested 3 days after treatment, and thespecific expression of CD11b, CD11c,CD14, CD49d, and CDw125 (IL-5Ra) wasassessed by direct staining using specificantibodies. Cell surface (e ) or intracellular(i) staining was determined using aFACScalibur flow cytometer. Columns,mean of three independent experiments;bars, SE.

Cancer Research




treatment, morphologic changes characteristic for apoptosis couldbe observed, including cell shrinkage with maintained plasmamembrane integrity, condensation, and fragmentation of nuclei.Such apoptotic bodies became preponderant by day 7 of ATRAtreatment, and this was followed on longer incubations by acomplete loss of the viability of the culture (Fig. 5B, 4-6).

Taken together, our results show that ATRA regulates theexpression of integrins specifically expressed during eosinophilicdifferentiation, has no effect on IL-5Ra expression (Fig. 1D), anddoes not induce mature eosinophil morphology (Fig. 5B ),indicating that the cell death effect of ATRA is not linked to theinduction of terminal eosinophilic differentiation.

Characterization of ATRA-induced EoL-1 cell death. Tocharacterize the molecular events taking place during ATRA-induced death of EoL-1 cells, general apoptosis markerscorresponding to various stages of the apoptotic process wereinvestigated. Early apoptosis was detected by phosphatidylserineexternalization as measured by FITC-labeled Annexin V bindingand propidium iodide staining. Treatment of cells with ATRAinduced, in a time-dependent manner, the externalization of phos-phatidylserine in EoL-1 cells (Fig. 2A and B). Although untreatedcells were predominantly unlabeled by Annexin V or propidiumiodide, treatment with 100 nmol/L ATRA for 3 days resulted in amarked increase of the number of Annexin V-positive, propidiumiodide–negative cells, indicative of apoptosis. On longer treatment,this was followed by the appearance of Annexin V and propidiumiodide double-positive cells, indicative of more advanced stages ofapoptosis and cell destruction.

Two major pathways lead to the induction of apoptosis.Although the extrinsic apoptotic pathway, triggered by the acti-vation of the death receptors and leading to caspase-8 activation,is a well-established mechanism of the apoptosis of matureeosinophils (29, 30), data are currently not available about the

intrinsic pathway in this lineage. The intrinsic pathway involvesproteins of the Bcl-2 family and the mitochondria to which signalsconverge (31). This process results in the loss of mitochondrialmembrane potential and release of proapoptotic proteins, such ascytochrome c , leading to the formation of the apoptosome, whichin turn cleaves and activates caspase-9 (31).

As ATRA-induced apoptosis in leukemic neutrophil-granulocyticdifferentiation implicates mitochondria and proteins of the Bcl-2family, the mitochondrial apoptosis pathway was further investi-gated by measuring DCm in control and ATRA-treated EoL-1 cellsby flow cytometry using DiOC6(3) (22), a fluorescent carbocyaninedye. This molecule distributes into mitochondria depending on themembrane potential of the organelle and therefore a decrease ofDiOC6(3) fluorescence indicates diminished DCm. Treatment ofEoL-1 cells with 100 nmol/L ATRA induced a significant drop inDiOC6(3) fluorescence when compared with untreated cells,indicating that ATRA treatment at days 1 to 4 leads to a gradualdecrease of DCm (Fig. 2C).

We investigated also the expression of proteins of the Bcl-2family by Western immunoblotting in EoL-1 cells treated withATRA for 3 days. ATRA treatment decreased Bcl-2 and increasedBak expression, whereas the expression levels of other Bcl-2 familymembers, such as Bax, Bcl-xL, and Mcl-1, did not changesignificantly (Fig. 3A). Taken together, these data indicate thatATRA treatment induced an f3.5-fold proapoptotic shift in theBcl-2/Bak ratio (Fig. 3B).

Extrinsic and intrinsic pathways lead to activation of initiatorcaspases, such as caspase-8 and caspase-9, which can activate theeffector caspases, such as caspase-3 and caspase-6, which cleave,degrade, or activate other cellular proteins (32). Activation ofprocaspase-8, procaspase-9, and procaspase-3 was monitored dailyin EoL-1 cells treated with 100 nmol/L ATRA for 4 days bydetection of active caspase fragments by Western blot. Formation

Figure 2. ATRA induces phosphatidylserine externalization and DCm decrease. EoL-1 cells were treated or not treated with ATRA (10�7 mol/L). A, apoptoticcells were detected by Annexin V binding and propidium iodide labeling after the indicated times of treatment. Apoptotic cells in the lower right quadrant (LR ), necroticcells in the upper right quadrant (UR ), and viable cells in the lower left quadrant (LL ) were detected. B, percentage of apoptotic and necrotic cells. C, DCm wasstudied by flow cytometry in untreated (dotted lines) or treated (continuous lines) EoL-1 cells with 10�7 mol/L ATRA after the indicated time of treatment. DecreasedDCm was detected as a leftward shift in the curve shape of the DiOC6(3) signal.





of proteolitically active caspase-3 (17 kDa), as well as caspase-8(36 and 40 kDa) and caspase-9 (34 kDa) could be detected inATRA-treated EoL-1 cells from day 2 (Fig. 3C). Cleavage of caspasesubstrates was investigated by the detection of proteolyticfragments of PARP, a well-established caspase substrate (33) andof PMCA, a recently identified target of caspase action (34).Treatment of EoL-1 cells with 100 nmol/L ATRA induced PARPcleavage with the appearance of an 85-kDa proteolytic PARPfragment (Fig. 3D). Cleavage followed a time course compatiblewith that of caspase-3 activation and could be completely inhibitedby the pan-caspase inhibitor z-VAD-(O-Me)-fmk (data not shown).Proteolytic fragmentation of PMCA was also observed in ATRA-treated EoL-1 cells, and this led to the formation of a 125-kDaPMCA fragment, in accordance with the literature (Fig. 3D ; ref. 34).Taken together, these data indicate that ATRA-induced apoptosisin EoL-1 cells is caspase dependent.

Analysis of ATRA-activated signal transduction in EoL-1cells. The biological effect of retinoids is mediated by ligand-inducible transcription factors that belong to the RAR and RXRgene families. Both gene families are composed of 3 members(RAR-a, RAR-h, and RAR-g and RXR-a, RXR-h, and RXR-g, respect-ively), which may perform distinct functions involved in theinduction of differentiation, growth inhibition, or apoptosis (35).

To identify the retinoid receptor pathway involved in apoptosisinduction in EoL-1 cells, experiments with retinoid receptorisoform–specific ligands were carried out. AM580, a selectiveRAR-a-specific agonist induced cell death of EoL-1 cells (Figs. 4Aand 5C, 1-3). AM580 (10 nmol/L) was f10-fold more potent in

apoptosis induction when compared with ATRA (100 nmol/L). Thisagrees with the relative potencies of the two molecules observed inother experimental settings (36). Similar to other cell types (36),ATRA could transactivate the RAR-a-sensitive RAR-h2 promoteralso in EoL-1 cells (Fig. 4B). ATRA-induced death of EoL-1 cellscould be inhibited by the RAR-a-specific antagonist Ro-41 5253(Figs. 4C and 5C, 4-6 ; ref. 37). Endogenous RAR-a protein could bedetected in EoL-1 cells, and ATRA treatment induced RAR-aprotein down-regulation (Fig. 4D). This is in accordance with theATRA-induced proteolytic degradation of RAR-a observed in othermyeloid cell types (38). Altogether, these observations indicate thatRAR-a is expressed and functional in EoL-1 cells and is sufficient toinitiate ATRA-induced apoptosis. On the other hand, differentconcentrations of ATRA had no discernible effect on the expressionlevel or the state of tyrosine phosphorylation (P-Y754-PDGFR-a) ofthe FIP1L1/PDGFR-a fusion protein, whereas imatinib totallyabrogated the autophosphorylation of the tyrosine kinase fusionprotein at Y754 of the PDGFR-a moiety (Fig. 4E).

ATRA inhibits eosinophil colony formation. To extend ourresults and to explore the clinical relevance of our observations, weinvestigated eosinophil colony formation on normal and HES-derived bone marrow mononuclear cells in the presence andabsence of ATRA.

Bone marrow–derived mononuclear cells of healthy donorsand of a FIP1L1/PDGFR-a-positive HES patient were grown incollagen containing semisolid cultures as described in Materialsand Methods. The number of eosinophil and noneosinophilcolonies was determined by cellular morphology on MGG-stained

Figure 3. ATRA induces diminution of theBcl-2/Bak ratio and caspase activation.EoL-1 cells were treated or not treated withATRA (10�7 mol/L). A, Western blotanalysis was done with specific antibodiesagainst Bak (23 kDa), Bax (23 kDa), Bcl-2(26 kDa), Bcl-xL (26 kDa), Mcl-1 (37 kDa),and h-actin (42 kDa). B, protein expressionlevel of Bcl-2 and Bak in cells treated withATRA for 3 days was compared withuntreated cells. Result is fold induction ofrelative protein expression compared withuntreated cells. Columns, mean of threeindependent experiments; bars, SE.Western blot analysis was done withspecific antibodies against caspase-3,caspase-8, caspase-9, (C ) PARP (116kDa), PMCA (140 kDa), and h-actin (D ).*, p85 and p120 fragments resulting fromthe cleavage of PARP and PMCA bycaspases, respectively.

Cancer Research




preparations and by immunocytochemical staining for MBP, aneosinophil granule protein that constitutes a specific marker of thislineage (39). Figure 5D, 1 and 2 shows morphologic features ofeosinophilic cells (bilobed nuclei and crystalloid granules) and cellsof noneosinophilic colonies (predominantly of neutrophil-granulo-cytic morphology), respectively. Figure 5D, 3 and 4 shows positiveMBP immunostaining (red immunocytochemical staining) ineosinophilic and lack of staining in noneosinophilic colonies,respectively. Under ATRA treatment, a specific decrease of theabsolute number of eosinophil colonies (Fig. 6C) as well as adecrease of the eosinophil/noneosinophil colony ratio (Fig. 6Aand B) could be observed by morphologic as well as immunocy-tochemical criteria on MGG- and MBP-stained samples of HESmyeloid progenitors, respectively. As shown in Fig. 6D, 5 and 6 ,when HES mononuclear cells were grown in collagen cultures inthe presence of 1 Amol/L ATRA, residual eosinophilic coloniesdisplayed a marked apoptotic morphology with nuclear conden-sation and numerous apoptotic bodies.

These observations show for the first time that ATRA exerts atclinically attainable concentrations a selective antiproliferativeeffect on eosinopoiesis of HES-derived bone marrow mononuclearcells.

Discussion

Retinoids are natural and synthetic vitamin A derivatives andanalogues that are involved in many important biologicalprocesses, including vision, morphogenesis, differentiation, growth,metabolism, and cellular homeostasis. ATRA regulates biologicalprocesses by binding to specific nuclear receptors, RARs and RXRs(35). The antiproliferative activity observed for ATRA is thoughtto be mostly due to direct growth inhibition and induction of

differentiation. ATRA can induce the differentiation of many tumorcell lines in vitro (40–42), but, currently, its efficacy in vivo isrestricted to APL, where ATRA targets the PML/RAR-a-chimericoncoprotein, inducing the progressive differentiation of theleukemic cells and their elimination from the body (1).

Whereas some synthetic retinoids are potent inducers ofapoptosis (43), although apoptosis induction has been reported forATRA in some cell lines, this is usually weak and is induced atrelatively high concentrations (26, 43–45). The mechanisms by whichATRA induces apoptosis are not well understood. The moleculewas shown to down-regulate the expression of the antiapoptoticprotein Bcl-2 (45–47) and to induce the decrease of mitochondrialrespiration (48) and calcium sequestration (49) causing cell deathby apoptosis (50). In addition to the activation of the intrinsic apop-totic pathway, ATRA has also been reported to induce apoptosis inAPL blasts through the induction of the expression of the tumornecrosis factor–related apoptosis-inducing ligand (51).

Recently, imatinib mesylate that is in clinical use for thetreatment of chronic myeloid leukemia, has been shown to induceremission in some cases of HES (4). This led to the identification ofthe imatinib mesylate–inhibitable FIP1L1/PDGFR-a-oncogenicfusion protein (12). The discovery of FIP1L1/PDGFR-a, observedin a majority of cases of HES, permitted to establish the clonal,leukemic nature of this disorder. HES can be life threatening due toinfiltration of vital organs by excessive amounts of eosinophils,which can release toxic substances normally involved in antipar-asitic host defense. In addition, HES can transform to acuteleukemia (18). Thus far, the only currently available in vitro cellularmodel for the study of FIP1L1/PDGFR-a-positive HES is the EoL-1cell line, established from an acute eosinophilic leukemia that arosefollowing chronic HES (13, 14). EoL-1 cells express the FIP1L1/PDGFR-a fusion oncoprotein (15). In vitro , imatinib mesylate

Figure 4. Effect of RAR-a analogues onEoL-1 cell death. A, EoL-1 cells weretreated or not treated with ATRA(10�7 mol/L) or AM580 (10�8 mol/L) for4 days, and the percentage of viable cellswas measured by trypan blue staining.B, EoL-1 cells were electroporated with theRAR-b2 luciferase reporter gene andtreated or not treated with ATRA(10�7 mol/L) or AM580 (10�8 mol/L) for18 hours, and extracts were tested forluciferase activity. Results are foldinduction of luciferase activity relative tothat displayed in the absence of ATRA.All experiments were normalized toh-galactosidase. C, EoL-1 cells weretreated or not treated with ATRA(10�8 mol/L), Ro-41 5253 (Ro ; 10�6 mol/L),or ATRA + Ro-41 5253 (ATRA + Ro ) for6 days, and the percentage of viable cellswas measured by trypan blue staining.D, Western blot analysis was done withspecific antibodies against RAR-a(50 kDa) and h-actin (42 kDa). E, EoL-1cells were treated for 3 days with ATRA orimatinib, and Western blot analysis wasdone with antibodies against PDGFR-a,P-Y754-PDGFR-a, and h-actin. *, FIP1L1/PDGFR-a fusion protein (110 kDa)specifically expressed by EoL-1 cells. As anegative control, NB4 cells were used.Columns, mean of three independentexperiments; bars, SE.





induces the apoptosis of EoL-1 cells (12). However, the exactmolecular mechanisms of this effect remain to be established.

To our knowledge, no published work has addressed thus far theeffects of retinoids on apoptosis of malignant cells belonging to theeosinophilic lineage (52), although ATRA has been reported toexert a suppressive effect on normal eosinopoiesis ex vivo (20).

In the present work, we show that ATRA induces the apoptosisof EoL-1 cells and inhibits eosinopoiesis of FIP1L1/PDGFR-a-positiveHES-derived bone marrow mononuclear cells. Apoptosis proceedssomewhat more slowly than that induced by imatinib and could beshown by phosphatidylserine externalization, activation of initiatingand executor caspases, proteolytic cleavage of caspase substrates,proapoptotic shift of the ratio of proapoptotic versus antiapoptotic

Bcl-2 family proteins, decreased DCm, and cell morphology.Interestingly, apoptosis could be induced in EoL-1 cells at extremelylow ATRA concentrations, with clearly detectable cell death in theslightly supraphysiologic range (4 to 10 nmol/L) of ATRA. Unlike inthe myeloblastic HL-60 and promyelocytic NB4 cells, in EoL-1 cells,although integrin expression varied, death was not preceded byterminal differentiation as shown by cell morphology or IL-5Raexpression. These characteristics taken together suggest that EoL-1cells constitute a new experimental model with unique features forthe study of the apoptosis-inducing effect of retinoids.

The understanding of the exact molecular mechanisms set inmotion by ATRA in EoL-1 cells requires further research. However,the capacity of an RAR-a-specific agonist to induce and of an

Figure 5. Morphologic features of ATRA-treated EoL-1cells. A, photographs of EoL-1 cells were taken 4 daysafter 10�7 mol/L ATRA treatment before (1 and 2)or after (3 and 4) trypan blue staining. B, cells weretreated or not treated with ATRA (10�7 mol/L) during7 days, and MGG staining was done at different periods.1, control; 2, ATRA day 1; 3, day 2; 4, day 3; 5, day 4;6, day 7. Current magnification, �75. Black arrows,apoptotic morphology. C, MGG staining was done onEoL-1 cells 4 or 6 days after treatment. 1, control;2, ATRA 10�7 mol/L; 3, AM580 10�8 mol/L;4, Ro 10�6 mol/L; 5, ATRA 10�8 mol/L; 6, ATRA +Ro-41 5253. Current magnification, �75.D, HES-derived bone marrow culture shows themorphologic characteristics of eosinophilic andnoneosinophilic cells after MGG staining (1 and 2 ,respectively; current magnification, �75) and positiveMBP immunostaining in eosinophilic cells versusnegative immunostaining in noneosinophilic cells(3 and 4 , respectively; current magnification, �40).5 and 6, apoptotic morphology of eosinophil coloniesgrown in collagen culture in the presence of 1 Amol/LATRA. Current magnification, �40 and �75,respectively.

Cancer Research




RAR-a-specific antagonist to prevent ATRA-induced cell death, incombination with the demonstration of functional RAR-a in EoL-1cells, strongly suggests that RAR-a-dependent signaling is sufficientto initiate apoptosis in these cells. In addition, the observation thatATRA-induced cell death was independent from initial cell densityin the 1 � 103 to 2 � 105/mL range and the inability of exogenouslyadded IL-5 to rescue ATRA-treated cells suggest that cell death isnot initiated by autocrine/paracrine mechanisms or by theacquisition by the cells of an IL-5-dependent phenotype.

In this work, activation of initiating (caspase-8 and caspase-9)and effector caspases (caspase-3) has been observed in ATRA-treated EoL-1 cells. Activation of caspase-9, which is induced in themitochondrial apoptotic pathway, has been observed in parallelwith a decrease of DCm. In addition, caspase-8 activation was alsoobserved as well as cleavage of caspase substrates (PARP andPMCA). Because of the considerable complexity of apoptoticpathways and, in particular, due to the existence of various positivefeedback mechanisms (53, 54), the precise identification of theprimary initiating event of ATRA-induced apoptosis in EoL-1 cellsrequires further studies. Our observations indicate, however, thatATRA-induced killing of EoL-1 cells proceeds via RAR-a-inducedcaspase-dependent apoptosis, without influencing the activity ofthe fusion protein as measured by the state of its autophosphor-ylation or its level of expression. This suggests that ATRA-inducedapoptosis is independent from the expression or activity of FIP1L1/PDGFR-a tyrosine kinase fusion protein.

Rather than being uniform, the mechanisms of the control andthe induction and execution of programmed cell death display cell-type and stimulus-dependent characteristics. Apoptosis of normalmature eosinophils can be induced in the presence of prosurvivalcytokines, such as IL-3, IL-5, or GM-CSF, by glucocorticoids (55),transforming growth factor-h (29), or by activation of the Fasreceptor in vitro , and Fas-dependent signal transduction (30) andreactive oxygen species (56) are involved in this process. Thesignaling pathways that control apoptosis constitute an intricatenetwork, in which cross-talk between various proapoptotic andantiapoptotic mechanisms operates in a highly dynamic and inter-connected manner, and this has not been investigated in eosi-nophils in great detail thus far. The marked sensitivity and rapidityof ATRA-induced apoptosis of EoL-1 cells clearly distinguish thesecells from other known models of ATRA-induced cell death(26, 43–45). This model therefore may prove useful for the in-depthanalysis of the molecular mechanisms of ATRA-induced cell death.

Our results may have some relevance in the context of the clinicalmanagement of HES. We show that the formation of eosinophilcolonies from the bone marrow mononuclear cell fraction of aFIP1L1/PDGFR-a-positive HES patient can be inhibited in vitro byclinically attainable concentrations of ATRA (1 Amol/L; refs. 1, 57), asdetected by early (MBP expression) as well as late eosinophildifferentiation markers (i.e., formation of colonies that displaymature eosinophil-granulocytic morphology). This effect wasspecific in that ATRA treatment had no significant effect oneosinopoiesis from normal bone marrow samples under the con-ditions used in our experiments. Morphologic analysis of residualeosinophilic colonies in ATRA-treated HES collagen cultures re-vealed the presence of nuclear condensation and numerous apop-totic bodies compatible with apoptosis. Due to the sensitivity ofthe tyrosine kinase activity of the FIP1L1/PDGFR-a fusion proteinto inhibition by imatinib mesylate, neoplastic HES is currentlytreated by imatinib mesylate (2–6). However, spontaneous as well asacquired resistance to this drug has been observed in a significantsubset of patients (2, 7, 9). Therefore, our observations support theidea that ATRA, which is used in clinical practice for APL, may beof some use also in non-APL-type malignancies, such as HES.

Acknowledgments

Received 1/7/2006; revised 1/9/2006; accepted 3/31/2006.Grant support: Institut National de la Sante et de la Recherche Medicale,

Fondation de France, Association pour la Recherche sur le Cancer, and AssociationLaurette Fugain (C. Robert and B. Papp).

Figure 6. ATRA inhibits eosinophil colony formation in HES. Bone marrowmononuclear cells obtained from normal donors (Normal ) and from a case ofFIP1L1/PDGFR-a-positive HES (Patient ) were grown as described in Materialsand Methods in the presence or absence of 10�6 mol/L ATRA. Eosinophil (E)and noneosinophil (NE ) colonies were identified morphologically onMGG-stained preparations (A ) and independently by MBP immunostaining (B).Results are percentage of eosinophil (white bars ) versus noneosinophil colonies(black bars ). C, absolute number of eosinophil and noneosinophil colonies.Columns, mean of three independent experiments; bars, SE.





The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Drs. Maurizio Giannı (Istituto di Ricerche Farmacologiche MarioNegri, Milan, Italy), Michel Lanotte, Jerome Larghero (Unite de Therapie Cellulaire,Hopital Saint-Louis, Paris, France), Dr. Patrice Richard and Marie-Laurence Menot,Fabien Zassadowski (UMR-S 718 Institut National de la Sante et de la Recherche

Medicale), Dr. Agota Apati (National Medical Center, Budapest, Hungary), and Prof.Lionel Prin (Faculte de Medecine, Lille, France) for support and stimulatingdiscussions during this work; Prof. Pierre Chambon and Dr. Cecile Rochette-Eglyfor the RAR-a antibody; Drs. Eva-Maria Gutknecht and Pierre Weber for syntheticretinoids; Drs. Patrice Berthaud and Frederique Oustry (Novartis) for imatinibmesylate; and Robin Nancel (Service d’Infographie, Institut Universitaire d’Hema-tologie, Hopital Saint-Louis) for excellent artwork.

Cancer Research


References

1. Fenaux P, Chomienne C, Degos L. Acute promyelocyticleukemia: biology and treatment. Semin Oncol 1997;24:92–102.

2. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinasecreated by fusion of the PDGFRa and FIP1L1 genes as atherapeutic target of imatinib in idiopathic hyper-eosinophilic syndrome. N Engl J Med 2003;348:1201–14.

3. Coutre SE, Gotlib J. Targeted treatment of hyper-eosinophilic syndromes and chronic eosinophilic leuke-mias with imatinib mesylate. Semin Cancer Biol 2004;14:307–15.

4. Gleich GJ, Leiferman KM, Pardanani A, Tefferi A,Butterfield JH. Treatment of hypereosinophilic syn-drome with imatinib mesilate. Lancet 2002;359:1577–8.

5. Gotlib J, Cools J, Malone JM III, et al. The FIP1L1-PDGFRa fusion tyrosine kinase in hypereosinophilicsyndrome and chronic eosinophilic leukemia: implica-tions for diagnosis, classification, and management.Blood 2004;103:2879–91.

6. Pardanani A, Brockman SR, Paternoster SF, et al.FIP1L1-PDGFRa fusion: prevalence and clinicopatho-logic correlates in 89 consecutive patients with moder-ate to severe eosinophilia. Blood 2004;104:3038–45.

7. Tauchi T, Ohyashiki K. Molecular mechanisms ofresistance of leukemia to imatinib mesylate. Leuk Res2004;28 Suppl 1:S39–45.

8. Gallagher RE. Retinoic acid resistance in acutepromyelocytic leukemia. Leukemia 2002;16:1940–58.

9. Cools J, Stover EH, Boulton CL, et al. PKC412overcomes resistance to imatinib in a murine modelof FIP1L1-PDGFRa-induced myeloproliferative disease.Cancer Cell 2003;3:459–69.

10. Tomonaga M, Gasson JC, Quan SG, Golde DW.Establishment of eosinophilic sublines from humanpromyelocytic leukemia (HL-60) cells: demonstration ofmultipotentiality and single-lineage commitment of HL-60 stem cells. Blood 1986;67:1433–41.

11. Paul CC, Tolbert M, Mahrer S, et al. Cooperativeeffects of interleukin-3 (IL-3), IL-5, and granulocyte-macrophage colony-stimulating factor: a new myeloidcell line inducible to eosinophils. Blood 1993;81:1193–9.

12. Griffin JH, Leung J, Bruner RJ, Caligiuri MA,Briesewitz R. Discovery of a fusion kinase in EoL-1 cellsand idiopathic hypereosinophilic syndrome. Proc NatlAcad Sci U S A 2003;100:7830–5.

13. Mayumi M. EoL-1, a human eosinophilic cell line.Leuk Lymphoma 1992;7:243–50.

14. Saito H, Bourinbaiar A, Ginsburg M, et al. Establish-ment and characterization of a new human eosinophilicleukemia cell line. Blood 1985;66:1233–40.

15. Cools J, Quentmeier H, Huntly BJ, et al. The EoL-1 cellline as an in vitro model for the study of FIP1L1-PDGFRa-positive chronic eosinophilic leukemia. Blood2004;103:2802–5.

16. Wong CK, Ho CY, Lam CW, Zhang JP, Hjelm NM.Differentiation of a human eosinophilic leukemic cellline, EoL-1: characterization by the expression ofcytokine receptors, adhesion molecules, CD95, andeosinophilic cationic protein (ECP). Immunol Lett1999;68:317–23.

17. Shintaku N, Ohshima Y, Jung EY, et al. Induction ofeosinophilic granules, nonspecific esterase activity andCD14 expression in the human eosinophilic leukemiacell line, EoL-1. Hematol Oncol 1994;12:129–39.

18. Gilliland G, Cools J, Stover EH, Wlodarska I, MarynenP. FIP1L1-PDGFRa in hypereosinophilic syndrome andmastocytosis. Hematol J 2004;5:S133–7.

19. Sun SY, Lotan R. Retinoids and their receptors incancer development and chemoprevention. Crit RevOncol Hematol 2002;41:41–55.

20. Upham JW, Sehmi R, Hayes LM, et al. Retinoic acidmodulates IL-5 receptor expression and selectivelyinhibits eosinophil-basophil differentiation of hemopoi-etic progenitor cells. J Allergy Clin Immunol 2002;109:307–13.

21. Vermes I , Haanen C, Stef fens-Nakken H,Reutelingsperger C. A novel assay for apoptosis. Flowcytometric detection of phosphatidylserine expressionon early apoptotic cells using fluorescein labelledAnnexin V. J Immunol Methods 1995;184:39–51.

22. Korchak HM, Rich AM, Wilkenfeld C, Rutherford LE,Weissmann G. A carbocyanine dye, DiOC6(3), acts as amitochondrial probe in human neutrophils. BiochemBiophys Res Commun 1982;108:1495–501.

23. Umesono K, Giguere V, Glass CK, Rosenfeld MG,Evans RM. Retinoic acid and thyroid hormone inducegene expression through a common responsive element.Nature 1988;336:262–5.

24. Delva L, Bastie JN, Rochette-Egly C, et al. Physicaland functional interactions between cellular retinoicacid binding protein II and the retinoic acid-dependentnuclear complex. Mol Cell Biol 1999;19:7158–67.

25. Gelebart P, Kovacs T, Brouland JP, et al. Expression ofendomembrane calcium pumps in colon and gastriccancer cells. Induction of SERCA3 expression duringdifferentiation. J Biol Chem 2002;277:26310–20.

26. Martin SJ, Bradley JG, Cotter TG. HL-60 cells inducedto differentiate towards neutrophils subsequently dievia apoptosis. Clin Exp Immunol 1990;79:448–53.

27. Taimi M, Breitman TR. Growth, differentiation, anddeath of retinoic acid-treated human acute promyelo-cytic leukemia NB4 cells. Exp Cell Res 1997;230:69–75.

28. Jung EY, Ohshima Y, Shintaku N, et al. Effects ofcyclic AMP on expression of LFA-1, Mac-1, and VLA-4and eosinophilic differentiation of a human leukemiacell line, EoL-1. Eur J Haematol 1994;53:156–62.

29. Alam R, Forsythe P, Stafford S, Fukuda Y. Trans-forming growth factor h abrogates the effects ofhematopoietins on eosinophils and induces theirapoptosis. J Exp Med 1994;179:1041–5.

30. Druilhe A, Cai Z, Haile S, Chouaib S, Pretolani M. Fas-mediated apoptosis in cultured human eosinophils.Blood 1996;87:2822–30.

31. Cory S, Adams JM. The Bcl-2 family: regulators of thecellular life-or-death switch. Nat Rev Cancer 2002;2:647–56.

32. Degterev A, Boyce M, Yuan J. A decade of caspases.Oncogene 2003;22:8543–67.

33. Tewari M, Quan LT, O’Rourke K, et al. Yama/CPP32 h,a mammalian homologue of CED-3, is a CrmA-inhibitable protease that cleaves the death substratepoly(ADP-ribose) polymerase. Cell 1995;81:801–9.

34. Paszty K, Verma AK, Padanyi R, et al. Plasmamembrane Ca2+ATPase isoform 4b is cleaved andactivated by caspase-3 during the early phase ofapoptosis. J Biol Chem 2002;277:6822–9.

35. Minucci S, Pelicci PG. Retinoid receptors in healthand disease: co-regulators and the chromatin connec-tion. Semin Cell Dev Biol 1999;10:215–25.

36. Giannı M, Li Calzi M, Terao M, et al. AM580, a stablebenzoic derivative of retinoic acid, has powerful andselective cyto-differentiating effects on acute promyelo-cytic leukemia cells. Blood 1996;87:1520–31.

37. Apfel C, Bauer F, Crettaz M, et al. A retinoic acidreceptor a antagonist selectively counteracts retinoicacid effects. Proc Natl Acad Sci U S A 1992;89:7129–33.

38. Kopf E, Plassat JL, Vivat V, et al. Dimerization withretinoid X receptors and phosphorylation modulate theretinoic acid-induced degradation of retinoic acidreceptors a and g through the ubiquitin-proteasomepathway. J Biol Chem 2000;275:33280–8.

39. Al-Rabia MW, Blaylock MG, Sexton DW, Thomson L,Walsh GM. Granule protein changes and membrane

receptor phenotype in maturing human eosinophilscultured from CD34+ progenitors. Clin Exp Allergy 2003;33:640–8.

40. Breitman TR, Selonick SE, Collins SJ. Induction ofdifferentiation of the human promyelocytic leukemiacell line (HL-60) by retinoic acid. Proc Natl Acad SciU S A 1980;77:2936–40.

41. Lanotte M, Martin-Thouvenin V, Najman S, et al. NB4,a maturation inducible cell line with t(15;17) markerisolated from a human acute promyelocytic leukemia(M3). Blood 1991;77:1080–6.

42. Doyle LA, Giangiulo D, Hussain A, et al. Differenti-ation of human variant small cell lung cancer cell linesto a classic morphology by retinoic acid. Cancer Res1989;49:6745–51.

43. Pfahl M, Piedrafita FJ. Retinoid targets for apoptosisinduction. Oncogene 2003;22:9058–62.

44. Liu Y, Lee MO, Wang HG, et al. Retinoic acid receptorh mediates the growth-inhibitory effect of retinoic acidby promoting apoptosis in human breast cancer cells.Mol Cell Biol 1996;16:1138–49.

45. Pettersson F, Dalgleish AG, Bissonnette RP, ColstonKW. Retinoids cause apoptosis in pancreatic cancercells via activation of RAR-g and altered expression ofBcl-2/Bax. Br J Cancer 2002;87:555–61.

46. Park JR, Robertson K, Hickstein DD, et al. Dysregu-lated bcl-2 expression inhibits apoptosis but notdifferentiation of retinoic acid-induced HL-60 granulo-cytes. Blood 1994;84:440–5.

47. Calabresse C, Barbey S, Venturini L, et al. In vitrotreatment with retinoids or the topoisomerase inhibitor,VP-16, evidences different functional apoptotic path-ways in acute promyelocytic leukemic cells. Leukemia1995;9:2049–57.

48. Notario B, Zamora M, Vinas O, Mampel T. All-trans -retinoic acid binds to and inhibits adenine nucleotidetranslocase and induces mitochondrial permeabilitytransition. Mol Pharmacol 2003;63:224–31.

49. Schmidt-Mende J, Gogvadze V, Hellstrom-Lindberg E,Zhivotovsky B. Early mitochondrial alterations in ATRA-induced cell death. Cell Death Differ 2006;13:119–28.

50. Orrenius S, Zhivotovsky B, Nicotera P. Regulation ofcell death: the calciumapoptosis link. Nat Rev Mol CellBiol 2003;4:552–65.

51. Altucci L, Rossin A, Raffelsberger W, et al. Retinoicacid-induced apoptosis in leukemia cells is mediated byparacrine action of tumor-selective death ligand TRAIL.Nat Med 2001;7:680–6.

52. Tai G, Eun-Young J, Yuji H, et al. Different effects ofcyclic AMP and butyrate on eosinophilic differentiation,apoptosis and bcl-2 expression of a human eosinophilicleukemia cell line, EoL-1. Hematol Oncol 1996;14:181–92.

53. Cowling V, Downward J. Caspase-6 is the directactivator of caspase-8 in the cytochrome c -inducedapoptosis pathway: absolute requirement for removal ofcaspase-6 prodomain. Cell Death Differ 2002;9:1046–56.

54. Van de Craen M, Declercq W, Van den brande I, FiersW, Vandenabeele P. The proteolytic procaspase activa-tion network: an in vitro analysis. Cell Death Differ 1999;6:1117–24.

55. Wallen N, Kita H, Weiler D, Gleich GJ. Glucocorti-coids inhibit cytokine-mediated eosinophil survival.J Immunol 1991;147:3490–5.

56. Wedi B, Straede J, Wieland B, Kapp A. Eosinophilapoptosis is mediated by stimulators of cellularoxidative metabolisms and inhibited by antioxidants:involvement of a thiol-sensitive redox regulation ineosinophil cell death. Blood 1999;94:2365–73.

57. Cambier N, Menot ML, Schlageter MH, et al. All transretinoic acid abrogates spontaneous monocytic growthin juvenile chronic myelomonocytic leukaemia. HematolJ 2001;2:97–102.



Documents

Apoptosis Induction by Retinoids in Eosinophilic Leukemia Cells: Implication of Retinoic Acid Receptor- Signaling in All-Trans-Retinoic Acid Hypersensitivity