UMR CNRS 6101, Service d Hématologie Clinique, Service de Pédiatrie, CHU Dupuytren, Limoges, France

Transcription

1 Stem Cells Original Article Membrane and Intracellular Platelet-Activating Factor Receptor Expression in Leukemic Blasts of Patients with Acute Myeloid and Lymphoid Leukemia M. DONNARD, L. GUGLIELMI, P. TURLURE, C. PIGUET, M.J. COURAUD, D. BORDESSOULE, Y. DENIZOT UMR CNRS 6101, Service d Hématologie Clinique, Service de Pédiatrie, CHU Dupuytren, Limoges, France Key Words. AML ALL PAF receptor Blast cells CD34 + cells ABSTRACT Platelet-activating factor (PAF), a phospholipid mediator with a wide range of actions on mature leukocytes, acts through PAF-receptors (PAF-Rs) on the membranes of responsive cells. No results are available concerning the putative presence of PAF-Rs on leukemic blasts. Using multiparameter flow cytometry, we assessed intracellular and membrane PAF-Rs on blast cells of acute myeloid leukemic (AML) and acute lymphoid leukemic (ALL) patients. Membrane PAF-Rs were documented in 7/15 cases of ALL and 0/28 cases of AML. Putative intracellular PAF-Rs were found in blasts of 8/8 ALL and 13/13 AML patients. Vitamin D 3 and dimethyl sulfoxide that induced the expression of PAF-Rs on the membrane of the human promyelocytic leukemia cell line, HL60, failed to induce their expression on the membranes of CD34 + AML blasts. The lack of membrane PAF-Rs on the membranes of AML blasts confirms that these receptors represent a marker of mature cells and that their membrane induction is a consequence of cell maturation and differentiation. Stem Cells 2002;20: INTRODUCTION Platelet-activating factor (PAF) is a phospholipid mediator that sparks off a wide range of immunoregulatory actions on various blood cell types, such as monocytes and B lymphocytes. Indeed, PAF is well known in enhancing the immunoglobulin secretion by B lymphocytes [1] and in stimulating the production of various cytokines by circulating monocytes [2]. The PAF-receptor (PAF-R), a 7-transmembrane molecule that belongs to the G-protein-coupled family, is present on the membranes of responsive cells [3]. Indeed, functional PAF-Rs are detected on mature monocytes [4] and B tonsillar lymphocytes [5]. While the HL-60 promyelocytic leukemia cell line did not express membrane PAF-Rs, their differentiation toward the monocyte/ macrophage and granulocyte phenotype was associated with induction of PAF-R gene expression [6, 7], thus suggesting that this receptor might represent a marker of mature cells, and its induction might be a consequence of cell maturation and differentiation. So far, few results are available concerning the presence of PAF-Rs on immature leukocytes. In view of the potentially important role of PAF in processes of leukocyte maturation and function [1, 8], we Correspondence: Y. Denizot, Ph.D., UMR CNRS 6101, Faculté de Médecine, 2 rue Dr. Marcland, Limoges, France. Telephone: Fax: ; yves.denizot@unilim.fr Received February 8, 2002; accepted for publication June 6, AlphaMed Press /2002/$5.00/0 STEM CELLS 2002;20:

2 395 PAF Receptors in AML and ALL Blasts investigated, using multiparameter flow cytometry, the presence of intracellular and/or membrane PAF-Rs on blast cells of acute myeloid leukemic (AML) and acute lymphoid leukemic (ALL) patients. PATIENTS AND METHODS Patients Blood samples were obtained from patients at diagnosis over a period of 18 months. Samples were first used for routine laboratory investigations. The remaining blood was then used for PAF-R investigation. Blood recovered on EDTA was obtained from 28 AML patients (11 men and 17 women, mean age 69 years) and 15 ALL patients (11 men and four women, mean age 41 years) according to the Helsinki recommendations. The AML population (graded according to the French-American-British classification) consisted of 15 AML0, five AML1, and eight AML2 patients. The ALL population consisted of five pre-b1 ALL, seven pre-b2 ALL, and three pre-b3 ALL patients. Flow Cytometry Analysis Fresh cells were used, and membrane PAF-Rs were assessed as previously described [9]. Briefly, for membrane PAF-Rs 100 µl of blood were incubated for 10 minutes in the presence of 10% serum anitbody (AB) (ETS; Nantes, France) at 37 C (serum AB was used to reduce the unspecific background). After washing, cells were incubated with antihuman PAF-R mouse IgG antibodies (Spi Bio; Massy, France; or irrelevant mouse IgG antibodies, as an isotype control (Immunotech; Marseille, France) for 30 minutes at 4 C. After washing, cells were incubated with fluorescein isothiocyanate (FITC)-labeled goat anti-mouse antibodies ([GAM] Dakopatts; Glostrup, Denmark; For AML patients, cells were then labeled with phycoerythrin (PE)-labeled Table 1. Expression of membrane and intracellular PAF receptors (PAF-Rs) on blast cells of 28 AML patients Diagnosis Blast phenotype Membrane PAF-R Intracellular PAF-R AML0 CD33 CD34 + No nd AML0 CD33 CD34 + No nd AML0 CD33 CD34 + No Yes AML0 CD33 CD34 + No Yes AML0 CD33 CD34 + No Yes AML0 CD33 + CD34 + No nd AML0 CD33 + CD34 + No nd AML0 CD33 + CD34 No nd AML0 CD33 + CD34 No Yes AML1 CD33 + CD34 + No Yes AML1 CD33 + CD34 + No Yes AML1 CD33 + CD34 No nd AML1 CD33 + CD34 No nd AML1 CD33 + CD34 No nd AML2 CD33 CD34 + No Yes AML2 CD33 CD34 + No nd AML2 CD33 + CD34 + No nd AML2 CD33 + CD34 + No Yes AML2 CD33 + CD34 + No Yes AML2 CD33 + CD34 + No Yes AML2 CD33 + CD34 + No Yes AML2 CD33 + CD34 No nd Membrane and intracellular PAF-Rs were investigated in 15 AML0, five AML1, and eight AML2 patients. For some patients (nd), the amount of available blood allowed only the assessment of membrane PAF-R. Abbreviation: nd = not done.

3 Donnard, Guglielmi, Turlure et al. 396 anti-cd33 or SpectralRed (PC5)-labeled anti-cd34 antibodies (Immunotech). For ALL patients, cells were then labeled with PE-labeled anti-cd10 or PC5-labeled anti-cd34 antibodies. Cell suspensions were then submitted to flow cytometric analysis (XL II; Coulter; Margency, France; For intracellular PAF-R assessment, 50 µl of blood were fixated and permeabilized with the IntraPrep permeabilization reagent (Immunotech) according to the manufacturer s recommendations prior to incubation with anti-human PAF-R mouse antibodies or irrelevant mouse antibodies. In Vitro Expression of PAF-Rs from Blast Cells In a first set of experiments, blood cells from four AML patients (one AML0, one AML1, and two AML2) were harvested by centrifugation, washed in Hanks balanced salts A B PAF-R-FITC C Counts GFP CD33-PE CD34-PC5 PAF-R-FITC Figure 1. Membrane and intracellular PAF-R of AML blasts. A) A representative result from CD33 + CD34 + blasts from one AML0 patient. B) A representative result from CD33 + CD34 + blasts from one AML1 patient. C) A representative result from CD33 CD34 + blasts from one AML2 patient. Left panel: Blasts gated on physical parameters were analyzed for the presence of membrane PAF-Rs (FITC labeling), CD33 (PE labeling), and CD34 (PC5 labeling). Right panel: Blasts were fixed, permeabilized, and stained with a control antibody (unshaded area) or specific anti-paf-r antibody followed by an FITC-conjugated secondary antibody (shaded area).

4 397 PAF Receptors in AML and ALL Blasts Table 2. Expression of membrane and intracellular PAF receptors (PAF-Rs) on blast cells of 15 ALL patients Diagnosis Blast phenotype Membrane PAF-R Intracellular PAF-R pre-b1 ALL CD10 CD34 + No Yes pre-b1 ALL CD10 + CD34 + No Yes pre-b1 ALL CD10 + CD34 + Yes nd pre-b1 ALL CD10 + CD34 No Yes pre-b1 ALL CD10 + CD34 Yes Yes pre-b2 ALL CD10 CD34 + No nd pre-b2 ALL CD10 + CD34 + No Yes pre-b2 ALL CD10 + CD34 + Yes nd pre-b2 ALL CD10 + CD34 + Yes Yes pre-b2 ALL CD10 + CD34 No nd pre-b2 ALL CD10 + CD34 Yes nd pre-b2 ALL CD10 + CD34 Yes nd pre-b3 ALL CD10 + CD34 + No Yes pre-b3 ALL CD10 + CD34 + Yes Yes pre-b3 ALL CD10 + CD34 No nd Membrane and intracellular PAF-Rs were investigated in five pre-b1, seven pre-b2, and three pre-b3 ALL patients. For some patients (nd), the amount of available blood allowed only the assessment of membrane PAF-R. Abbreviation: nd = not done. solution and incubated up to 3 hours at 37 C in a PAF-free medium (RPMI 1640). The presence of membrane PAF-Rs on CD34 + AML blasts was then assessed as described above. In a second set of experiments, PAF-Rs were investigated on CD34 + blast cells from two AML2 patients before and after 4 days of growth in liquid medium (RPMI 1640 with 10% fetal calf serum, 10% 5637-conditioned medium, as source of colony-stimulating factor, and antibiotics) with 0.1 µm vitamin D 3 (Sigma; Saint Quentin Fallavier, France; 0.5% dimethyl sulfoxide (DMSO), or 50 U/ml of macrophage-colony-stimulating factor (M-CSF) at 37 C in 5% C0 2 in air [10]. RESULTS Results indicated no membrane PAF-Rs on blast cells of 28 different AML patients. Thus, CD34 + CD33 blasts (five AML0 and two AML2), CD34 + CD33 + blasts (eight AML0, two AML1, and five AML2), and CD34 CD33 + blasts (two AML0, three AML1, and one AML2) did not express membrane PAF-Rs (Table 1, Figure 1A). In contrast, blasts from 13/13 AML patients (six AML0, two AML1, and five AML2) contained intracellular PAF-Rs (Table 1, Figure 1B). For these experiments, the means (± standard deviation [SD]) of fluorescence obtained with PAF-R antibodies (2.5 ± 0.3, n = 13) were higher (p < 0.01, t test for paired data) than those obtained with control antibodies (0.7 ± 0.1, n = 13). Results indicated that membrane PAF-Rs were present on blast cells of seven ALL patients but absent on blast cells of eight other ALL patients (Table 2). Flow cytometry analysis showed heterogeneity in the distribution of PAF-Rs on the ALL blasts obtained from a single patient (Figure 2A). Intracellular PAF-Rs were found in blast cells of 8/8 ALL patients (Table 2, Figure 2B). For these experiments, the means (± SD) of fluorescence obtained with PAF-R antibodies (5.1 ± 1.7, n = 8) were higher (p < 0.01, t test for paired data) than those obtained with control antibodies (0.6 ± 0.1, n = 8). Incubation of washed CD34 + AML blasts up to 3 hours in a PAF-free medium did not induce the expression of PAF-Rs on CD34 + cell membranes (Figure 3A). As shown in Figure 3B, 4 days of treatment with vitamin D 3, M-CSF, and DMSO (data not shown) did not induce the expression of PAF-Rs on the membranes of CD34 + AML2 blasts. DISCUSSION Previous results reported that human blood CD34 + cells isolated by the magnetic cell sorting procedure expressed PAF-R mrna but did not express membrane PAF-Rs [10]. The high sensitivity of the reverse-transcriptase polymerase chain reaction (RT-PCR) procedure may reveal levels of mrna expression that could be physiologically irrelevant and that do not lead to the production of PAF-Rs. The in vitro growth of normal CD34 + cells for 3 days in the presence of stimulating cytokines induced the coexpression on CD34 + progenitors of PAF-Rs with the CD33 differentiation marker [10]. These results suggest that PAF-Rs were expressed only by CD34 + cells undergoing several changes in maturational stage, and thus, in antigen expression. Results of this short clinical study indicate that CD34 + AML blasts (M0, M1, and M2 types) did not express

5 Donnard, Guglielmi, Turlure et al. 398 A B PAF-R-FITC C PAF recept-fit GFP Counts PAF recept-fit CD10-PE CD34-PC5 PAF-R-FITC Figure 2. Membrane and intracellular PAF-Rs of ALL blasts. A) A representative result from CD10 CD34 + blasts from one pre-b1 ALL patient. B) A representative result from CD10 + CD34 + blasts from one pre-b2 ALL patient. C) A representative result from CD10 CD34 + blasts from one pre-b3 ALL patient. Left panel: Blasts gated on physical parameters were analyzed for the presence of membrane PAF-Rs (FITC labeling), CD10 (PE labeling), and CD34 (PC5 labeling). Right panel: Blasts were fixed, permeabilized, and stained with a control antibody (unshaded area) or specific anti-paf-r antibody followed by an FITC-conjugated secondary antibody (shaded area). membrane PAF-Rs, even in the presence of the maturation marker CD33. Exposure of cells to PAF led to rapid sequestration of PAF-Rs into their intracellular compartments. This effect is mostly a temporary event, with reexpression of PAF-Rs within 1 hour [11-13]. High amounts of PAF are present in human marrow [14]. To test the hypothesis that bone-marrow-derived PAF might be responsible for the observed PAF-R sequestration inside cells, blasts were washed and incubated up to 3 hours in a PAF-free medium. Under these experimental conditions, no PAF-Rs were

6 399 PAF Receptors in AML and ALL Blasts A Control 1 hour 3 hours PAF-R-FITC CD34-PC5 B PAF-R-FITC Control Vitamin D M-CSF CD34-PC5 Figure 3. Analysis of membrane PAF-Rs on CD34 + AML blasts after washing (A) and after 4 days of culture (B). A) AML blasts were labeled with PC5-conjugated anti-cd34 antibodies and anti-paf-r antibodies associated with FITC-conjugated GAM as a secondary fluorochrome before and after washing, followed by 1 or 3 hours in PAF-free medium. One representative experiment out of four is shown. B) AML2 blasts were labeled with PC5-conjugated anti-cd34 antibodies and anti-paf-r antibodies associated with FITC-conjugated GAM as a secondary fluorochrome after 4 days of growth with and without 0.1 µm of vitamin D 3 or 50 U/ml of M-CSF. One representative experiment out of two is shown. detected on the membranes of CD34 + AML blasts, suggesting that their lack of membrane PAF-Rs is not related to PAF-R internalization in response to marrow-derived PAF. Flow cytometry experiments suggest the presence of PAF- Rs inside AML blast cells. Several authors have already reported intracellular PAF binding sites in various cells, such as neutrophils [15, 16] and platelets [17]. Normal blood CD34 + cells did not express membrane PAF-Rs [10]. However, their treatment with carbamyl-paf (a nonmetabolizable PAF agonist) reduced the number of erythroid (BFU-E) and myeloid (colony-forming units-granulocytemacrophage) colonies in semisolid cultures [18] suggesting the existence of functional intracellular PAF-Rs in normal CD34 + cells. At the present time, we have no direct experimental evidence that intracellular PAF-Rs from CD34 + AML blasts are functional. Indeed, intracellular PAF-Rs are not always functional. For example, when Cys 90 or Cys1 73 is mutated, the resulting PAF-Rs are only found intracellularly and are defective in ligand-binding activity [19]. Clearly, further studies are required to confirm the functionality of intracellular PAF-R from CD34 + AML blasts. Experiments with cell lines reported that the expression of membrane PAF-Rs was upregulated by various agents, including those responsible for cell differentiation [6, 7, 20]. Thus, HL-60 cells (an M2/M3-type cell line) are a classical model of myeloid maturation, because they differentiate to granulocytes or monocytes when treated with DMSO or vitamin D 3, respectively [6, 7]. Spontaneous M0/M1 blast cells are not responsive to vitamin D 3 [20]. Contrary studies have reported that M2-type AML blasts

7 Donnard, Guglielmi, Turlure et al. 400 are efficiently responsive [21] or weakly responsive [22, 23] to this molecule. Blast maturity and CD34 expression were reported to determine the effect of a vitamin D analogue [24]. In this study, vitamin D 3 (but also DMSO and M-CSF) did not induce AML2 blast differentiation (data not shown). Moreover, these molecules did not induce the expression of membrane PAF-Rs on CD34 + AML2 blasts, strengthening the hypothesis that this receptor might represent a marker of mature cells, and its induction might be a consequence of cell maturation and differentiation. A previous study reported PAF-R mrna transcripts in CD10 + tonsillar B lymphocytes [5]. Stimulation of CD10 + tonsillar B lymphocytes with PAF increased their IL-4 mrna and IL-13 mrna levels only in a small part of the entire population, suggesting that PAF-Rs were not present on all CD10 + B cells. Our results, using ALL blasts, confirm, first, the presence of membrane PAF-Rs on CD34 + CD10 + and CD34 CD10 + B cells, and second, that membrane PAF-Rs were not present on all CD10 + B cells from the same patient. As for AML patients, blasts from ALL patients expressed intracellular PAF-R. As for AML blasts, the functionality of intracellular PAF-Rs in ALL blasts remains an open question that requires further investigation. In conclusion, blasts from AML0, AML1, and AML2 patients did not express membrane PAF-Rs, suggesting that this receptor represents a marker of mature cells and its membrane induction is a consequence of cell maturation and/or differentiation. Results of this clinical study do not argue in favor of a role for PAF on early lymphopoiesis and myelopoiesis. ACKNOWLEDGMENT We are grateful to the Ligue Nationale Contre le Cancer (Comité de la Corrèze) for funding our project. REFERENCES 1 Mazer BD, Sawami H, Franklin R et al. B-cell activation and regulation of immunoglobulin synthesis by platelet activating factor. Neth J Med 1991;39: Poubelle PE, Gingras D, Demers C et al. Platelet-activating factor (PAF-acether) enhances the concomitant production of tumour necrosis factor-alpha and interleukin-1 by subsets of human monocytes. Immunology 1991;72: Ishii S, Shimizu T. Platelet-activating factor (PAF) receptor and genetically engineered PAF receptor mutant mice. Prog Lipid Res 2000;39: Ouellet S, Müller E, Rola-Pleszczynski M. INF-gamma upregulates platelet-activating factor receptor gene expression in human monocytes. J Immunol 1994;152: Bastien Y, Toledano BJ, Mehio N et al. Detection of functional platelet-activating factor receptors on human tonsillar B lymphocytes. J Immunol 1999;162: Müller E, Dupuis G, Turcotte S et al. Human PAF receptor gene expression: induction during HL-60 cell differentiation. Biochem Biophys Res Commun 1991;181: Vallari DS, Austinhirst R, Snyder F. Development of specific functionally active receptors for platelet-activating factor in HL-60 cells following granulocytic differentiation. J Biol Chem 1990;265: Saito H, Hayakawa T, Mita H et al. PAF-induced eosinophilic and basophilic differentiation in human hematopoietic precursor cells. J Lipid Mediat 1992;5: Trimoreau F, Guglielmi L, Touati M et al. Platelet-activating factor receptors on B cells of chronic lymphocytic leukaemia patients [letter]. Br J Haematol 2001;115: Desplat V, Besse A, Faucher JL et al. Expression of plateletactivating factor transcript-1 but not transcript-2 by human bone marrow cells. STEM CELLS 1999;17: Le Gouill C, Parent JL, Rola-Pleszczynski M et al. Structural and functional requirements for agonist-induced internalization of the human platelet-activating factor receptor. J Biol Chem 1997;272: Ishii I, Saito E, Izumi T et al. Agonist-induced sequestration, recycling, and resensitization of platelet-activating factor receptor. Role of cytoplasmic tail phosphorylation in each process. J Biol Chem 1998;273: Flaherty JT, Jacobson DP, Redman JF. Regulation of platelet-activating factor receptors and the desensitization response in polymorphonuclear neutrophils. Biochem J 1992;288: Denizot Y, Trimoreau F, Dupuis F et al. PAF and haematopoiesis: III. Presence and metabolism of platelet-activating factor in human bone marrow. Biochim Biophys Acta 1995;1265: Koike H, Imanishi N, Natsume Y et al. Effects of platelet activating factor receptor antagonists on intracellular platelet activating factor function in neutrophils. Eur J Pharmacol 1994;269: Müller S, Nigam S. Enhancement by staurosporine of platelet-activating factor formation in N-formyl peptide-challenged human neutrophils is mediated by intracellular platelet-activating factor binding sites. Biochem Biophys Res Commun 1992;189: Hwang SB, Wang S. Wheat germ agglutinin potentiates specific binding of platelet-activating factor to human platelet membranes and induces platelet-activating factor synthesis in intact platelets. Mol Pharmacol 1991;39: Dupuis F, Gachard N, Allegraud A et al. Effect of platelet-activating factor on the growth of human erythroid and myeloid CD34 + progenitors. Mediators Inflamm 1998;7: Le Gouill C, Parent JL, Rola-Pleszczynski M et al. Role of the Cys 90, Cys 95, and Cys 173 residues in the structure and function

8 401 PAF Receptors in AML and ALL Blasts of the human platelet-activating factor receptor. FEBS Lett 1997;402: Kishimoto S, Shimazu W, Izumi T et al. Enhanced expression of platelet-activating factor receptor on human eosinophils by interleukin-3, interleukin-5 and granulocyte-macrophage colony-stimulating factor. Int Arch Allergy Immunol 1996;111(suppl 1): Manfredini R, Trevisan F, Grande A et al. Induction of a functional vitamin D receptor in all-trans-retinoic acidinduced monocytic differentiation of M2-type leukemic blast cells. Cancer Res 1999;59: Howell AL, Stukel TA, Bloomfield CD et al. Induction of differentiation in blast cells and leukemia colony-forming cells from patients with acute myeloid leukemia. Blood 1990;75: Yamada K, Honma Y, Asahi KI et al. Differentiation of human acute myeloid leukaemia cells in primary culture in response to cotylenin A, a plant growth regulator. Br J Haematol 2001;114: Pallis M, Turzanski J, Russell NH. Blast maturity and CD34 expression determine the effects of the differentiating agents KH1060 and 9-cis-retinoic acid on the differentiation and clonogenicity of non-m3 acute myeloid leukaemia cells. Leukemia 1998;12: