- Email: [email protected]
Expression of cyclin-dependent kinase inhibitor p15INK4B during normal and leukemic myeloid differentiation
Experimental Hematology 28 (2000) 519–526
Expression of cyclin-dependent kinase inhibitor during normal and leukemic myeloid differentiation
INK4B
p15
Luciana Teofilia, Roberta Morosettia, Maurizio Martinib, Raffaella Urbanoa, Rossana Putzulua, Sergio Rutellaa, Luca Pierellia, Giuseppe Leonea, and Luigi Maria Laroccab Departments of aHematology and bPathology, Catholic University, Rome, Italy (Received 2 August 1999; revised 10 January 2000; accepted 20 January 2000)
Objective. Expression of the cyclin-dependent kinase inhibitor p15INK4B frequently is altered in myeloid malignancies. We previously demonstrated that p15INK4B is expressed in normal myeloid cells. The aim of this study was to investigate whether p15INK4B expression is restricted to the granulomonocytic lineage and to evaluate its modulation during normal and leukemic myeloid differentiation. Materials and Methods. Normal CD34⫹ cells were cultured in serum-free media to obtain granulomonocytic, erythroid, or megakaryocytic unilineage differentiation. NB4 promyelocytic cell line and fresh leukemic blasts from seven patients with acute promyelocytic leukemia were cultured with all-trans retinoic acid. At different times of culture, cell samples were collected to evaluate p15INK4B by semiquantitative reverse transcriptase polymerase chain reaction. Results. p15INK4B mRNA was found during granulomonocytic and megakaryocytic, but not erythroid, differentiation. In the granulomonocytic lineage, p15INK4B was detectable when the majority of cells were at the promyelocytic stage and increased progressively in more mature elements. In the megakaryocytic lineage, p15INK4B was expressed in the early phase of differentiation, before megakaryoblasts had appeared, and was mantained throughout the time of culture. NB4 cell line and five of seven leukemic samples displayed undetectable or very low level of p15INK4B that rapidly increased during retinoic acid-induced differentiation. Two leukemic samples (both collected from two patients developing all-trans retinoic acid syndrome) showed high basal levels of p15INK4B, which was not modified by retinoic acid treatment. Conclusions. p15INK4B upregulation occurs specifically during normal granulomonocytic and megakaryocytic commitment. In acute promyelocytic leukemic blasts, p15INK4B, which is detectable at a very low level, is promptly increased by retinoic acid. In contrast, two acute promyelocytic leukemia samples obtained from patients who developed all-trans retinoic acid syndrome showed high basal levels of p15INK4B that did not increase further during all-trans retinoic acid-induced differentiation. © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. Keywords: Hematopoietic differentiation—Cyclin-dependent kinase inhibitors—Acute promyelocytic leukemia—All-trans retinoic acid syndrome
Introduction An important function of the cyclin-dependent kinases inhibitors is the induction of cell cycle arrest in response to differentiative stimuli [1]. Cyclin-dependent kinases inhibitors usually are grouped in two major families, INK4 and the Kip/Cip families, according to their structural, functional, and biochemical properties [2]. Recent observations suggest that two cyclinOffprint requests to: Luciana Teofili, M.D., Department of Hematology, Catholic University, Largo A. Gemelli 8, 00168 Rome, Italy; E-mail: [email protected]
dependent kinases inhibitors belonging to different families cooperate to arrest cell cycle progression in response to antiproliferative stimuli [3]. Exponentially growing epithelial cells express high levels of Kip/Cip family inhibitors p27Kip1 (p27) and p21WAF1 (p21), which are bound to cyclin D-CDK-4 and -6 complexes, without inhibiting these kinases. In these cells, transforming growth factor  (TGF-) induces a progressive increase of p15INK4B (p15), which, in turn, displaces p21 and p27, and binds to and inhibits cyclin D-CDK complexes. In the meantime, p21 and p27 bind and inhibit cyclin-E-CDK2 complexes. As a result, cells are blocked in the G1 phase of
0301-472X/00 $–see front matter. Copyright © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(00)0 0 1 3 9 - 9
520
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
the cell cycle [3,4]. In hematopoietic cells, normal differentiation is associated with the progressive withdrawal from cell cycle, to restrict the proliferative potential of differentiating cells. Previous studies clearly demonstrated that p21 gene is upregulated during the granulocytic and monocytic cell differentiation of normal [5] and leukemic cells [6–9]. Moreover, we recently found a high expression of p15 in normal cells belonging to the granulomonocytic lineage [10], suggesting that p15 and p21 cooperate in inducing cell cycle arrest associated with normal myeloid differentiation. The importance of p15 in normal hematopoietic differentiation is suggested by the frequent loss of p15 expression in myeloid malignancies [11]. Actually, p15 gene methylation is a common event in acute myeloid leukemias (AML) and in myelodisplastic syndromes [11–17]. According to this finding, in a previous study investigating the expression of p15 in AML, we found that most of them, including acute promyelocytic leukemias (APL), were p15 negative [18]. To investigate the role of p15 in normal and neoplastic differentiation, we evaluated its expression in normal CD34⫹ progenitors differentiated toward different maturative lineages. We found that p15 is upregulated during granulomonocytic and megakaryocytic commitments, whereas it is completely unexpressed during erythroid differentiation. Considering that APL blasts are homogeneously blocked at the promyelocytic stage and that all-trans retinoic acid (ATRA) induces their growth inhibition and differentiation both in vivo and in vitro [19], we investigated the modulation of p15 in this leukemia. We studied NB4 promyelocytic cell line and fresh APL blasts that were induced to differentiate in vitro by ATRA and we found that p15, which is undetectable or expressed at very low levels in NB4 cells and in most APL samples, is progressively upregulated by ATRA. Interestingly, two APL samples obtained from patients who developed ATRA syndrome showed high basal levels of p15 mRNA, which did not increase further during ATRA-induced differentiation.
Materials and methods Patients Normal CD34⫹ cells were isolated from bone marrow samples of normal marrow donors after obtaining informed consent. Fresh leukemic blasts were obtained at diagnosis from peripheral blood or bone marrow of seven patients with APL. Diagnosis was made by morphologic criteria according to the French-American-British classification [20] and confirmed by detection of the PML/RARa rearrangement, as previously described [21]. All evaluated patients received ATRA (45 mg/m2/day administered orally) plus chemotherapy. Two of the patients developed ATRA syndrome that was diagnosed, as previously reported [22], based on the presence of fever, dyspnea, pleural or pericardial effusions, pulmonary infiltrates on chest radiograph, and unexplained increase of the body weight. Cell samples Normal cells. Aliquots of bone marrow were diluted 1:5 with Iscove’s modified Dulbecco’s medium (IMDM; Hyclone, UK) and
centrifuged over Ficoll-Hypaque (density 1.077; Pharmacia, Uppsala, Sweden). For the CD34⫹ cell selection, CD34⫹ Multisort kit (Miltenyi Biotec Inc., Auburn, CA) was used according to the manufacturer’s instructions. Briefly, mononuclear cells were washed twice in IMDM, suspended in phosphate-buffered saline (PBS; Aerobe, France) containing 1% of bovine serum albumin (Sigma, St. Louis, MO) and incubated with anti-CD34⫹ monoclonal antibody-coated beads for 30 minutes at 4⬚C. Cell suspension was passed through the column (MS⫹ columns, Miltenyi); attached cells were eluted from the column by washings with PBS. Collected cells were passed through a new column. Purity of cells was assessed by flow cytometry as previously described [10], using the phycoerythrin-conjugated anti-CD34 monoclonal antibody 8G12 (Becton-Dickinson, Mountain View, CA). In all the experiments, more than 95% of the collected cells were CD34⫹. Leukemic cells. Bone marrow or peripheral blood samples were diluted 1:5 with IMDM and centrifuged over Ficoll-Hypaque to isolate the mononuclear cell fraction. More than 95% of the obtained cells were blasts at morphologic examination. NB4 promyelocytic cell line (kindly provided from Dr. Clara Nervi, University “La Sapienza” of Rome) had exponentially grown in RPMI (Hyclone) containing 10% fetal bovine serum (FBS; StemCell Technologies, Vancouver, Canada) . Cell cultures Normal cells. For unilineage differentiation, CD34⫹ cells were seeded in liquid cultures at 5 ⫻ 104/mL (erythroid and myeloid cultures) or 10 ⫻ 104/mL (megakaryocytic cultures), in serumfree medium consisting of IMDM containing 20% of serum substitute (BIT 9500, StemCell Technologies), and the following growth factors (all purchased from PeproTech EC Ltd., London UK): granulocyte colony-stimulating factor (10 ng/mL), stem cell factor (100 ng/mL), granulocyte-macrophage colony-stimulating factor (10 ng/mL), interleukin 3 (10 ng/mL) for myeloid differentiation; stem cell factor (10 ng/mL) and erythropoietin (4 U/mL) for erythroid differentiation; and interleukin 6 (20 ng/mL) and thrombopoietin (10 ng/mL) for megakaryocytic differentiation as previously described [23]. Growing cells were maintained at cell concentration lower than 3 ⫻ 105/mL. At days 0, 3, 6, 9, 12, and 15 in myeloid differentiation experiments or at days 0, 6, and 12 in erythroid and megakaryocytic differentiation experiments, aliquots of cells were assayed for cell concentration, morphologic examination, and reverse transcriptase polymerase chain reaction (RT-PCR) analysis. To examine a pure erythroid or megakaryocytic cell population, before performing RT-PCR analysis, cells recovered at day 6 and 12 from these cultures were incubated with anti-glycophorin-A or anti-CD41 fluorescein isothiocyanate (FITC)-conjugated monoclonal antibodies (both purchased from Caltag Laboratories, Burlingame, CA) and then with antiFITC antibody-coated beads (Miltenyi). Cell suspensions were applied on MS⫹ columns as described for CD34⫹ cell selection. Glycophorin-A-positive (Gly⫹) and glycophorin-A-negative (Gly⫺) cells , CD41⫹, and CD41⫺ cells were recovered and assayed for RT-PCR. Cell cycle analysis was performed at different times of culture by propidium iodide staining, as previously described [10]. To evaluate the cell differentiative stage, cytospins were prepared, stained by May-Grunwald-Giemsa solution and independently examined by two different observers.
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Leukemic cells. To exclude that p15 upregulation could be induced by the presence of FBS, we performed preliminary experiments with NB4 cells in serum-free and serum-containing medium. No differences were found between these culture conditions, so all the experiments were carried out in the presence of FBS. Fresh leukemic promyelocytes and exponentially growing NB4 cells were seeded at 106/mL or 3 ⫻ 105/mL, respectively, in RPMI containing 10% FBS, in the presence of ATRA (Sigma) 10⫺7 M. At hours 0, 6, 12, 24, 48, and 72, cells were assayed for RT-PCR. ATRA was dissolved in 100% ethanol at 10⫺2 mol/L, stored at ⫺20⬚C, and protected from light. Control cultures were performed by adding the same concentration of ethanol as that present in ATRA-containing cultures.
521
Leukemic cell differentiation was assessed by flow cytometric evaluation of CD11b expression before and after ATRA exposure, as previously described [24]. RT-PCR All the samples were analyzed for p15 expression by semiquantitative RT-PCR. Total RNA was extracted and RT-PCR performed as previously described [24]. Bone marrow mononuclear cells were used as positive control. PCR was performed using the following primers and conditions: 5⬘-AAG-GTC-CCA-TGG-ACG-CGTGT-3⬘, 5⬘-TCG-AGG-TCA-TGA-TGG-GC-3⬘; annealing temperature 55⬚C, and MgCl2 1.5 mM for 30 cycles. To control for genomic DNA contamination, an RT reaction was performed without
Figure 1. p15 upregulation during granulomonocytic differentiation of bone marrow CD34⫹ cells. At day 0 and at 3-day intervals, cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (May-Grunwald-Giemsa staining) is shown as percentage of different cell types. pml ⫽ promyelocytes; myel ⫽ myelocytes; bands ⫽ band cells; segm ⫽ segmented granulocytes. (B) RT-PCR analysis of p15 and -actin. Each PCR reaction included primers specific for p15 and primers for -actin as internal control. PCR products were hybridized by radiolabeled cDNA probes and autoradiography was performed. Bone marrow mononuclear cells were used as positive control. On the bottom of the figure, densitometric analysis is shown as relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. The p15 signal becomes detectable at day 6 and peaks at day 12, when the majority of cells are myelocytes and metamyelocytes.
522
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
the addition of the enzyme; one reaction using this sample was included for each set of primers. An intron is present between this two primers so that any contaminating genomic DNA would have given a product of different size. A set of primers for -actin [25] was used in each PCR reaction as positive control. PCR products were transferred onto nylon membranes (Amersham, Arlington Heights, IL) after agarose gel electrophoresis, and standard hybridization was performed using cDNA probes, as previously described [25]. Densitometric analysis was performed for all the samples using the Molecular Imager (Biorad, Hercules, CA), and ratios of p15 vs -actin were obtained for quantitative analysis.
Figure 2. Flow cytometric evaluation of DNA content in CD34⫹ cells and in cells differentiating toward the myeloid lineage (propidium iodide staining). The percentage of cells in S⫹G2M phases of the cell cycle is shown. DNA distribution histograms were generated and analyzed with ModFIT software.
Results Normal CD34⫹ cell differentiation Myeloid differentiation. p15 was undetectable by RT-PCR in normal bone marrow CD34⫹ cells. As expected on the basis of our previous observations, we found a progressive increase of p15 during the granulomonocytic commitment [10], starting from day 6 and peaking on day 12 (Fig. 1 ). A signal was detectable when the majority of cells displayed
Figure 3. p15 evaluation in cultures of CD 34⫹ cells differentiating toward the erythroid lineage. At day 0, 6 and 12 cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (MayGrunwald staining). myel ⫽ contaminant cells belonging to the granulomonocytic lineage, at different stage of differentiation; proeryth ⫽ proerythroblasts; bas ⫽ basophilic erythroblasts; poly ⫽ polychromatic erythroblasts; orth ⫽ orthochromatic erythroblasts. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. Lane 1 ⫽ day 0 CD34⫹ cells; lane 2 ⫽ day 6 glycophorin-A⫺ cells; lane 3 ⫽ day 6 glycophorin-A⫹ cells; lane 4 ⫽ day 12 glycophorin-A⫺ cells; lane 5 ⫽ day 12 glycophorin-A⫹ cells. (C) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to the -actin signals. P15 is not expressed in erythroid cells, either at day 6 or day 12. The p15 signal detectable at day 12 in the Gly⫺ cell fraction is due to the granulomonocytic cell contaminant.
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
morphologic features of promyelocytes; nevertheless, p15 mRNA expression was maximal when cell population predominantly consisted of myelocytes and metamyelocytes (day 12, Fig. 1A). The percentage of cycling cells increased progressively from day 0 (15% to 30% of CD34⫹ cells in S phase) to day 6 (up to 46 % of S phase), and it rapidly declined afterward (8% to 12 % of S phase on day 15) (Fig. 2). Therefore, the cycling activity of differentiating cells was inversely related to the modulation of p15 expression.
523
ter ATRA exposure. Interestingly, in the other patient we observed a downregulation of gene expression during the first 24 hours of culture with ATRA (Fig. 7). In all cases, when fresh blasts were cultured without ATRA, no modifications of p15 message were detectable. In all APL samples, ATRA-induced differentiation was confirmed by the increase of the CD11b expression at immunocytometry, without significant differences between patients with or without ATRA syndrome.
Erythroid differentiation. According to our previous observations [10], p15 was undetectable throughout erythroid differentiation (Fig. 3). p15 mRNA was detectable only in the Gly⫺ cell population recovered at day 12, whereas Gly⫹ cells recovered at days 6 and 12 were p15⫺. The Gly⫺ cells recovered at day 6 also were p15⫺. Morphologic examination showed that most cells were undifferentiated blasts or erythroid differentiating cells, and no myeloid cells were detectable (Fig. 3A). Megakaryocytic differentiation. p15 was strongly expressed during the megakaryocytic differentiation: both CD41⫹ and CD41⫺ cell populations showed high levels of p15 mRNA at days 6 and 12 (Fig. 4).This finding was quite surprising, considering that in a previous immunohistochemical study of bone marrow biopsies we did not find p15 expression in megakaryocytes [10]. In contrast, in this unilineage culture system, p15 was clearly detectable at day 6, when most cells were undifferentiated blasts or megakaryoblasts (Fig. 4A). Considering that no differentiated myeloid elements were detectable by this time in these culture conditions, the finding of p15⫹ cells in CD41⫺ population suggests that p15 is sharply upregulated in the early phase of the megakaryocytic commitment, before the appearance of the CD41 antigen; thereafter, its expression was maintained at high levels throughout the time of culture. According to the high expression of p15, the percentage of cycling cells was low at days 6 and 12 (9% to 13 % and 8% to 11% of cells in S phase, respectively; data not shown). Expression of p15 in NB4 cell line and APLs. In the NB4 cell line we detected low basal levels of p15 mRNA. However, when this cell line was induced to differentiate with ATRA, a rapid upregulation of p15 was observed, followed by a slight decrease and a late peak at hour 72 (Fig. 5). NB4 cells cultured in the absence of ATRA did not show any significant variation in p15 mRNA expression (Fig. 5). Similar results were obtained in five of seven APL patients. The expression of p15 was dramatically upregulated during the first 6 hours of ATRA exposure. Only a slight increase was observed in the following hours of culture, with a late peak at 72 hours (Fig. 6). In two APL patients who developed ATRA syndrome, we found a completely different modulation of p15 mRNA expression. Blasts from these patients showed a high basal expression of p15 mRNA. In one of these patients, p15 mRNA levels did not increase further af-
Figure 4. p15 evaluation in cultures of CD 34⫹ cells differentiating toward the megakaryocytic lineage. At day 0, 6 and 12 cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (May-Grunwald staining). myel ⫽ contaminant cells belonging to the granulomonocytic lineage, at different stage of differentiation; megak bl ⫽ megakaryoblasts; early meg ⫽ early megakaryocytes; mat meg ⫽ mature megakaryocytes. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. Lane 1 ⫽ day 0 CD34⫹ cells; lane 2 ⫽ day 6 CD41⫺ cells; lane 3 ⫽ day 6 CD41⫹ cells; lane 4 ⫽ day 12 CD41⫺ cells; lane 5 ⫽ day 12 CD41⫹ cells. (C) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to the -actin signals. p15 is expressed throughout the entire culture period.
524
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Discussion We had demonstrated by immunohistochemistry in a previous study that p15 is largely expressed in normal cells belonging to the granulocytic and monocytic lineages [10]. In this study, we confirmed that p15 expression is upregulated when hematopoietic progenitors undergo myeloid differentiation. p15 mRNA, undetectable in CD34⫹ progenitors, becomes detectable when the majority of cells differentiating toward the granulomonocytic lineage are promyelocytes, i.e., the most differentiated myeloid elements able to proliferate; thereafter, as more mature nonproliferating myeloid cells predominate, p15 mRNA levels progressively increase. In a previous study, Steinman et al. [5] showed that p21 is upregulated during myeloid differentiation of CD34⫹ cells isolated from cord blood. The authors found that p21 expression depends on a new regulatory activity that binds a 44-bp fragment of the p21 promoter; this binding activity is highly expressed in normal differentiated cells and in HL60 cells treated with ATRA. That both p15 and p21 genes are upregulated in myeloid cells agrees with the recent observation showing that two cyclin-dependent kinases inhibitors, belonging to different families of cyclin-dependent kinases
Figure 6. p15 evaluation in fresh blasts obtained from a patient with acute promyelocytic leukemia at diagnosis. Cells were cultured in the presence of all-trans retinoic acid (ATRA) 10⫺7M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. For NB4 cells, p15 message is promptly induced by ATRA (hour 6); thereafter, only a slight increase peaking at 72 hours is found.
Figure 5. p15 evaluation in NB4 cell line cultured with and without alltrans retinoic acid (ATRA) 10⫺7 M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. p15 message, expressed at a low level in basal samples, is sharply upregulated in cells treated with ATRA.
inhibitors, cooperate to induce cell cycle arrest [3,4]. In addition, p15 and p21 could have different roles, specific for each of them, in the early or late phase of myeloid differentiation, respectively, when predominantly p21 [26] or p15 is expressed. It was reported recently that both p21 and p27 are expressed in megakaryocytic colonies derived from CD34⫹ progenitors [27]. In this study, we also show that p15 is expressed during megakaryocytic differentiation and that its upregulation occurs before the appearance of CD41⫹ megakaryoblastic cells. In a previous study, we investigated the presence of p15 in megakaryocytes using dual-color immunofluorescence analysis of bone marrow cells with antiCD61 and anti-p15 antibodies and immunohistochemical study of bone marrow biopsies. We did not found p15 expression in megakaryocytes [10]. This discrepancy can be due to the fact that by immunohistologic and immunocytologic analysis we examined megakaryocytes at different stages of differentiation, whereas CD34⫹ cells give rise a homogeneous megakaryocytic cell population. Moreover, the expression of p15 in the megakaryocytic lineage could
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Figure 7. p15 evaluation in fresh blasts obtained from a patient with acute promyelocytic leukemia at diagnosis, who developed all-trans retinoic acid (ATRA) syndrome during the treatment. Cells were cultured in the presence of ATRA 10⫺7 M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. The p15 mRNA level is high in basal samples and is downregulated by ATRA during the culture.
be related to megakaryocyte production of TGF-, an important inhibitor of megakaryocyte growth and endomitosis [28]. It was demonstrated recently that high levels of TGF-, as observed in patients with autoimmune thrombocytopenia, are able to regulate in vivo megakaryocytopoiesis by inducing marrow stromal cells to produce thrombopoietin; thrombopoietin, in turn, upregulates the expression of TGF- receptors on the megakaryocyte surface [29]. We can hypothesize that, in the unilineage culture system, high levels of thrombopoietin induce p15 by sensitizing megakaryocytes to the effect of TGF- produced by themselves. Finally, the finding that p15 upregulation is selectively induced in myeloid and megakaryocytic, but not erythroid, lineage suggests that its gene could be regulated by transcriptional activators specific to these lineages. Liu et al. [30] demonstrated that in U937 myelomonocytic cell line, which was differentiated by retinoic acid, the gene encoding p21 is transcriptionally activated by the retinoic acid receptor. As described for myeloid differentiation, retinoids play
525
an important role in the regulation of normal megakaryocytopoiesis [31]. Interestingly, patients receiving ATRA as differentiating therapy often show an increase of platelet count above normal values [32]. Whereas the p15 gene deletion has been documented mainly in T-lymphoid neoplasms [11], altered p15 expression due to p15 gene methylation is very frequently observed in AMLs [12,13,15] and myelodysplastic syndromes [14,16]. In this study, we show that, according to p15 modulation, leukemic promyelocytes can be distinguished into two groups. In one group, including the majority of APL samples and the NB4 cell line, p15 mRNA is undetectable or detected at very low levels by RT-PCR; nevertheless, ATRA exposure is able to induce rapidly p15 expression in these cells, showing that p15 gene is not hypermethylated in such leukemias. In contrast, in blasts obtained from patients developing ATRA syndrome, p15 is highly expressed in untreated cells and ATRA treatment downmodulates p15 mRNA. NB4 cell line was derived from APL blasts of the patient who was treated with ATRA who did not develop ATRA syndrome [33]; therefore, it is not surprising that NB4 cells show low basal levels of p15. ATRA is commonly used as a differentiating agent in the treatment of APL, and its association with anthracyclines has greatly improved the outcome of APL patients [19]. The most important side effect of ATRA administration is the development of ATRA syndrome, which is well defined on the basis of the occurrence of clinical signs and symptoms [22]. ATRA syndrome is a life-threatening complication, but unfortunately no predictive criteria are available, so that, in the presence of clinical suspicion, it is mandatory to immediately stop ATRA administration and to start dexamethasone administration [34]. The etiology of the ATRA syndrome is unknown; all the suggested pathogenetic mechanisms indicate that APL blasts from patients who develop ATRA syndrome release greater amounts of cytokines and growth factors in comparison to blasts from APL patients without ATRA syndrome [35–37]. If so, the high expression of p15 mRNA in cells from these patients could reflect a more advanced level of differentiation in comparison to patients without ATRA syndrome. This hypothesis is supported further by recent observations showing high basal levels of C/EBP⑀ in blasts from patients with ATRA syndrome [38]. In conclusion, this study definitely establishes the association between p15 upregulation and normal and leukemic myeloid differentiation. The observation of a different p15 modulation by ATRA in patients who develop ATRA syndrome offers new insights into better understanding the pathogenesis of this syndrome.
Acknowledgments This work was supported by grants from Associazione Italiana Ricerca sul Cancro and from M.U.R.S.T.
526
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
References 1. Hirama T, Koeffler HP (1995) Role of the cyclin-dependent kinase inhibitors in the development of cancer. Blood 86: 841 2. Hall M, Bates S, Peters G (1995) Evidence for different modes of action of cyclin-dependent kinase inhibitors: p15 and p16 bind to kinases, p21 and p27 bind to cyclins. Oncogene 11:1581 3. Reynisdottir I, Polyak C, Iavarone A, Massagué J (1995) Kip/Cip and Ink4 Cdk inhibitors cooperate to induce cell cycle arrest in response to TGF-. Genes Dev 9:1831 4. Reynisdottir I, Massagué J (1997) The subcellular localizations of the p15Ink4b and p27Kip1 coordinate their inhibitory interactions with cdk4 and cdk2. Genes Dev 11:492 5. Steinman RA, Huang J, Yaroslavskiy B, Goff JP, Ball ED, Nguyen A (1998) regulation of p21(WAF1) expression during normal myeloid differentiation. Blood 91:4531 6. Steinman RA, Hoffman B, Iro A, Guillouf C, Liebermann DA, ElHouseini ME (1994) Induction of p21(WAF1/CIP1) during differentiation. Oncogene 9:3389 7. Jiang H, Lin J, Su ZZ, Collart FR, Huberman E, Fisher PB (1994) Induction of differentiation in human promyelocytic HL-60 leukemia cells activates p21WAF1/CIP1 expression in the absence of p53. Oncogene 9:3397 8. Zhang W, Grasso L, McClain CD, et al. (1995) p53-independent induction of WAF1/CIP1 in human leukemia cells is correlated with growth arrest accompanying monocyte/macrophage differentiation. Cancer Res 55:668 9. Liu M, Lee MH, Cohen M, Bommakanti M, Freedman LP (1996) Transcriptional activation of the cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev 10:142 10. Teofili L, Rutella S, Chiusolo P, et al. (1998) Expression of p15INK4B in normal hematopoiesis. Exp Hematol 26:1133 11. Drexler HG (1998) Review of alterations of the cyclin-dependent kinase inhibitor INK4 family genes p15, p16, p18 and p19 in human leukemia-lymphoma cells. Leukemia 12:845 12. Herman GJ, Jen J, Merlo A Baylin SB (1996) Hypermethylation-associated inactivation indicates a tumor suppressor role for p15INK4B1. Cancer Res 56:722 13. Herman JG, Civin CI, Issa JPJ, Collector MI, Sharkis SJ, Baylin SB (1997) Distinct pattern of inactivation of p15INK4B and p16INK4A characterize the major types of hematological malignancies. Cancer Res 57:837 14. Uchida T, Kinoshita T, Nagai H, et al. (1997) Hypermethylation of the p15INK4B gene in myelodysplastic syndromes. Blood 90:1403 15. Aggerholm A, Guldberg P, Hokland M, Hokland P (1997) Extensive intra- and interindividual heterogeneity of p15INK4B methylation in acute myeloid leukemia. Cancer Res 59:436 16. Cameron EE, Baylin SB, Herman JG (1999) p15INK4B CpG island methylation in primary acute leukemia is heterogeneous and suggests density as a critical factor for transcriptional silencing. Blood 94:2445 17. Quesnel B, Guillerm G, Vereecque R, et al. (1998) Methylation of the p15INK4B gene in myelodysplastic syndromes is frequent and acquired during disease progression. Blood 91:2985 18. Larocca LM, Teofili L, Salutari P, et al. (1996) Expression of p15INK4B in primary leukemic blasts and its modulation by quercetin. Exp Hematol 24:1122 (abstr) 19. Grignani F, Fagioli M, Alcalay M, et al. (1994) Acute promyelocytic leukemia: from genetics to treatment. Blood 83:10 20. Bennet JM, Catowsky D, Daniel MT, et al. (1976) Proposal for the classification of the acute leukemias. Br J Haematol 33:451
21. Diverio D, Riccioni R, Mandelli F, Lo Coco F (1995). The PML/ RARa fusion gene in the diagnosis and monitoring of acute promyelocytic leukemia. Hematologica 80:155 22. Frankel SR, Eardley A, Lauwers G, Weiss M, Warrel R (1992) The “retinoic acid syndrome” in acute promyelocytic leukemia. Ann Intern Med 117:292 23. Pierelli L, Scambia G, Fattorossi A, et al. (1998) Functional, phenotypic and molecular characterization of cytokine low-responding circulating CD34⫹ haematopoietic progenitors. Br J Haematol 102: 1139 24. De Stefano V, Teofili L, Sica S, et al. (1995) Effect of all trans retinoic acid on procoagulant and fibrinolytic activities of cultured blast cells from patients with acute promyelocytic leukemia. Blood 86:3535 25. Morosetti R, Park DJ, Chumakov AM, et al. (1997) A novel myeloid transcription factor, C/EBP⑀, is up-regulated during granulocytic but not monocytic differentiation. Blood 90:2591. 26. Yaroslavskji B, Watkins S, Donnenberg AD, Patton TJ, Steinman RA (1999) Subcellular and cell-cycle expression profiles of CDK-inhibitors in normal differentiating myeloid cells. Blood 93:2907 27. Taniguchi T, Endo H, Chikatsu N, et al. (1999) Expression of p21Cip1/ Waf1/Sdi1 and p27Kip1 cycling dependent kinases inhibitors during human hematopoiesis. Blood 93:4167 28. Kuter DJ, Gminski DM, Rosemberg RD (1992) Transforming growth factor- inhibits megakaryocyte growth and endomitosis. Blood 79:619 29. Sakamaki S, Hirayama Y, Matsunaga T, et al. (1999) Transforming growth facto-1 (TGF-1) induces thrombopoietin from bone marrow stromal cells, which stimulates the expression of TGF- receptor on megakaryocytes and, in turn, renders them susceptible to suppression by TGF- itself with high specificity. Blood 94:1961. 30. Liu M, Iavarone A, Freedman LP (1996) Transcriptional activation of the human p21WAF1/CIP1 gene by retinoic acid receptor. J Biol Chem 271:31723 31. Visani G, Ottaviani E, Zauli G, et al. (1999) All-trans retinoic acid at low concentration directly stimulates normal adult megakaryocytopoiesis in the presence of thrombopoietin or combined cytokines. Eur J Hematol 63:149 32. Kinjo K, Kizaki M, Takayama N, et al. (1999) Serum thrombopoietin and erythropoietin levels in patients with acute promyelocytic leukemia during all-trans retinoic acid treatment. Br J Haematol 105:382 33. Lanotte M, Martin-Thouvenin V, Najman S, Balerini P, Valensi F, Berger R (1991) NB4, a maturation inducible cell line with t(15;17) marker isolated from a human acute promyelocytic leukemia (M3). Blood 77:1080 34. De Bottom S, Dombret H, Sanz S, et al. (1998) Incidence, clinical feature and outcome of all trans-retinoic acid syndrome in 413 cases of newly diagnosed acute promyelocytic leukemia. Blood 92:2712 35. Larson R, Brown D, Sklar L (1997) Retinoic acid induces aggregation of the acute promyelocytic leukemia cell line NB4 by utilization of LFA-1 and ICAM-2. Blood 90:2747 36. Grande A, Manfredini R, Tagliafico E, et al. (1995) All trans retinoic acid induces simultaneously granulocytic differentiation and expression of inflammatory cytokines in HL-60 cells. Exp Hematol 23:117 37. Dubois C, Schlageter MH, De Gentile A, et al. (1994) Modulation of IL-8 and IL-1b and G-CSF secretion by all trans retinoic acid in acute promyelocytic leukemia. Leukemia 8:1750 38. Morosetti R, Martini M, Teofili L, et al. (1999) ATRA up-regulates the expression of C/EBP⑀ in blasts from APL patients in a secondary AML, but not in blasts from APL patients with ATRA syndrome. Cancer Res 40:4069 (abstr)
Expression of cyclin-dependent kinase inhibitor during normal and leukemic myeloid differentiation
INK4B
p15
Luciana Teofilia, Roberta Morosettia, Maurizio Martinib, Raffaella Urbanoa, Rossana Putzulua, Sergio Rutellaa, Luca Pierellia, Giuseppe Leonea, and Luigi Maria Laroccab Departments of aHematology and bPathology, Catholic University, Rome, Italy (Received 2 August 1999; revised 10 January 2000; accepted 20 January 2000)
Objective. Expression of the cyclin-dependent kinase inhibitor p15INK4B frequently is altered in myeloid malignancies. We previously demonstrated that p15INK4B is expressed in normal myeloid cells. The aim of this study was to investigate whether p15INK4B expression is restricted to the granulomonocytic lineage and to evaluate its modulation during normal and leukemic myeloid differentiation. Materials and Methods. Normal CD34⫹ cells were cultured in serum-free media to obtain granulomonocytic, erythroid, or megakaryocytic unilineage differentiation. NB4 promyelocytic cell line and fresh leukemic blasts from seven patients with acute promyelocytic leukemia were cultured with all-trans retinoic acid. At different times of culture, cell samples were collected to evaluate p15INK4B by semiquantitative reverse transcriptase polymerase chain reaction. Results. p15INK4B mRNA was found during granulomonocytic and megakaryocytic, but not erythroid, differentiation. In the granulomonocytic lineage, p15INK4B was detectable when the majority of cells were at the promyelocytic stage and increased progressively in more mature elements. In the megakaryocytic lineage, p15INK4B was expressed in the early phase of differentiation, before megakaryoblasts had appeared, and was mantained throughout the time of culture. NB4 cell line and five of seven leukemic samples displayed undetectable or very low level of p15INK4B that rapidly increased during retinoic acid-induced differentiation. Two leukemic samples (both collected from two patients developing all-trans retinoic acid syndrome) showed high basal levels of p15INK4B, which was not modified by retinoic acid treatment. Conclusions. p15INK4B upregulation occurs specifically during normal granulomonocytic and megakaryocytic commitment. In acute promyelocytic leukemic blasts, p15INK4B, which is detectable at a very low level, is promptly increased by retinoic acid. In contrast, two acute promyelocytic leukemia samples obtained from patients who developed all-trans retinoic acid syndrome showed high basal levels of p15INK4B that did not increase further during all-trans retinoic acid-induced differentiation. © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. Keywords: Hematopoietic differentiation—Cyclin-dependent kinase inhibitors—Acute promyelocytic leukemia—All-trans retinoic acid syndrome
Introduction An important function of the cyclin-dependent kinases inhibitors is the induction of cell cycle arrest in response to differentiative stimuli [1]. Cyclin-dependent kinases inhibitors usually are grouped in two major families, INK4 and the Kip/Cip families, according to their structural, functional, and biochemical properties [2]. Recent observations suggest that two cyclinOffprint requests to: Luciana Teofili, M.D., Department of Hematology, Catholic University, Largo A. Gemelli 8, 00168 Rome, Italy; E-mail: [email protected]
dependent kinases inhibitors belonging to different families cooperate to arrest cell cycle progression in response to antiproliferative stimuli [3]. Exponentially growing epithelial cells express high levels of Kip/Cip family inhibitors p27Kip1 (p27) and p21WAF1 (p21), which are bound to cyclin D-CDK-4 and -6 complexes, without inhibiting these kinases. In these cells, transforming growth factor  (TGF-) induces a progressive increase of p15INK4B (p15), which, in turn, displaces p21 and p27, and binds to and inhibits cyclin D-CDK complexes. In the meantime, p21 and p27 bind and inhibit cyclin-E-CDK2 complexes. As a result, cells are blocked in the G1 phase of
0301-472X/00 $–see front matter. Copyright © 2000 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(00)0 0 1 3 9 - 9
520
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
the cell cycle [3,4]. In hematopoietic cells, normal differentiation is associated with the progressive withdrawal from cell cycle, to restrict the proliferative potential of differentiating cells. Previous studies clearly demonstrated that p21 gene is upregulated during the granulocytic and monocytic cell differentiation of normal [5] and leukemic cells [6–9]. Moreover, we recently found a high expression of p15 in normal cells belonging to the granulomonocytic lineage [10], suggesting that p15 and p21 cooperate in inducing cell cycle arrest associated with normal myeloid differentiation. The importance of p15 in normal hematopoietic differentiation is suggested by the frequent loss of p15 expression in myeloid malignancies [11]. Actually, p15 gene methylation is a common event in acute myeloid leukemias (AML) and in myelodisplastic syndromes [11–17]. According to this finding, in a previous study investigating the expression of p15 in AML, we found that most of them, including acute promyelocytic leukemias (APL), were p15 negative [18]. To investigate the role of p15 in normal and neoplastic differentiation, we evaluated its expression in normal CD34⫹ progenitors differentiated toward different maturative lineages. We found that p15 is upregulated during granulomonocytic and megakaryocytic commitments, whereas it is completely unexpressed during erythroid differentiation. Considering that APL blasts are homogeneously blocked at the promyelocytic stage and that all-trans retinoic acid (ATRA) induces their growth inhibition and differentiation both in vivo and in vitro [19], we investigated the modulation of p15 in this leukemia. We studied NB4 promyelocytic cell line and fresh APL blasts that were induced to differentiate in vitro by ATRA and we found that p15, which is undetectable or expressed at very low levels in NB4 cells and in most APL samples, is progressively upregulated by ATRA. Interestingly, two APL samples obtained from patients who developed ATRA syndrome showed high basal levels of p15 mRNA, which did not increase further during ATRA-induced differentiation.
Materials and methods Patients Normal CD34⫹ cells were isolated from bone marrow samples of normal marrow donors after obtaining informed consent. Fresh leukemic blasts were obtained at diagnosis from peripheral blood or bone marrow of seven patients with APL. Diagnosis was made by morphologic criteria according to the French-American-British classification [20] and confirmed by detection of the PML/RARa rearrangement, as previously described [21]. All evaluated patients received ATRA (45 mg/m2/day administered orally) plus chemotherapy. Two of the patients developed ATRA syndrome that was diagnosed, as previously reported [22], based on the presence of fever, dyspnea, pleural or pericardial effusions, pulmonary infiltrates on chest radiograph, and unexplained increase of the body weight. Cell samples Normal cells. Aliquots of bone marrow were diluted 1:5 with Iscove’s modified Dulbecco’s medium (IMDM; Hyclone, UK) and
centrifuged over Ficoll-Hypaque (density 1.077; Pharmacia, Uppsala, Sweden). For the CD34⫹ cell selection, CD34⫹ Multisort kit (Miltenyi Biotec Inc., Auburn, CA) was used according to the manufacturer’s instructions. Briefly, mononuclear cells were washed twice in IMDM, suspended in phosphate-buffered saline (PBS; Aerobe, France) containing 1% of bovine serum albumin (Sigma, St. Louis, MO) and incubated with anti-CD34⫹ monoclonal antibody-coated beads for 30 minutes at 4⬚C. Cell suspension was passed through the column (MS⫹ columns, Miltenyi); attached cells were eluted from the column by washings with PBS. Collected cells were passed through a new column. Purity of cells was assessed by flow cytometry as previously described [10], using the phycoerythrin-conjugated anti-CD34 monoclonal antibody 8G12 (Becton-Dickinson, Mountain View, CA). In all the experiments, more than 95% of the collected cells were CD34⫹. Leukemic cells. Bone marrow or peripheral blood samples were diluted 1:5 with IMDM and centrifuged over Ficoll-Hypaque to isolate the mononuclear cell fraction. More than 95% of the obtained cells were blasts at morphologic examination. NB4 promyelocytic cell line (kindly provided from Dr. Clara Nervi, University “La Sapienza” of Rome) had exponentially grown in RPMI (Hyclone) containing 10% fetal bovine serum (FBS; StemCell Technologies, Vancouver, Canada) . Cell cultures Normal cells. For unilineage differentiation, CD34⫹ cells were seeded in liquid cultures at 5 ⫻ 104/mL (erythroid and myeloid cultures) or 10 ⫻ 104/mL (megakaryocytic cultures), in serumfree medium consisting of IMDM containing 20% of serum substitute (BIT 9500, StemCell Technologies), and the following growth factors (all purchased from PeproTech EC Ltd., London UK): granulocyte colony-stimulating factor (10 ng/mL), stem cell factor (100 ng/mL), granulocyte-macrophage colony-stimulating factor (10 ng/mL), interleukin 3 (10 ng/mL) for myeloid differentiation; stem cell factor (10 ng/mL) and erythropoietin (4 U/mL) for erythroid differentiation; and interleukin 6 (20 ng/mL) and thrombopoietin (10 ng/mL) for megakaryocytic differentiation as previously described [23]. Growing cells were maintained at cell concentration lower than 3 ⫻ 105/mL. At days 0, 3, 6, 9, 12, and 15 in myeloid differentiation experiments or at days 0, 6, and 12 in erythroid and megakaryocytic differentiation experiments, aliquots of cells were assayed for cell concentration, morphologic examination, and reverse transcriptase polymerase chain reaction (RT-PCR) analysis. To examine a pure erythroid or megakaryocytic cell population, before performing RT-PCR analysis, cells recovered at day 6 and 12 from these cultures were incubated with anti-glycophorin-A or anti-CD41 fluorescein isothiocyanate (FITC)-conjugated monoclonal antibodies (both purchased from Caltag Laboratories, Burlingame, CA) and then with antiFITC antibody-coated beads (Miltenyi). Cell suspensions were applied on MS⫹ columns as described for CD34⫹ cell selection. Glycophorin-A-positive (Gly⫹) and glycophorin-A-negative (Gly⫺) cells , CD41⫹, and CD41⫺ cells were recovered and assayed for RT-PCR. Cell cycle analysis was performed at different times of culture by propidium iodide staining, as previously described [10]. To evaluate the cell differentiative stage, cytospins were prepared, stained by May-Grunwald-Giemsa solution and independently examined by two different observers.
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Leukemic cells. To exclude that p15 upregulation could be induced by the presence of FBS, we performed preliminary experiments with NB4 cells in serum-free and serum-containing medium. No differences were found between these culture conditions, so all the experiments were carried out in the presence of FBS. Fresh leukemic promyelocytes and exponentially growing NB4 cells were seeded at 106/mL or 3 ⫻ 105/mL, respectively, in RPMI containing 10% FBS, in the presence of ATRA (Sigma) 10⫺7 M. At hours 0, 6, 12, 24, 48, and 72, cells were assayed for RT-PCR. ATRA was dissolved in 100% ethanol at 10⫺2 mol/L, stored at ⫺20⬚C, and protected from light. Control cultures were performed by adding the same concentration of ethanol as that present in ATRA-containing cultures.
521
Leukemic cell differentiation was assessed by flow cytometric evaluation of CD11b expression before and after ATRA exposure, as previously described [24]. RT-PCR All the samples were analyzed for p15 expression by semiquantitative RT-PCR. Total RNA was extracted and RT-PCR performed as previously described [24]. Bone marrow mononuclear cells were used as positive control. PCR was performed using the following primers and conditions: 5⬘-AAG-GTC-CCA-TGG-ACG-CGTGT-3⬘, 5⬘-TCG-AGG-TCA-TGA-TGG-GC-3⬘; annealing temperature 55⬚C, and MgCl2 1.5 mM for 30 cycles. To control for genomic DNA contamination, an RT reaction was performed without
Figure 1. p15 upregulation during granulomonocytic differentiation of bone marrow CD34⫹ cells. At day 0 and at 3-day intervals, cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (May-Grunwald-Giemsa staining) is shown as percentage of different cell types. pml ⫽ promyelocytes; myel ⫽ myelocytes; bands ⫽ band cells; segm ⫽ segmented granulocytes. (B) RT-PCR analysis of p15 and -actin. Each PCR reaction included primers specific for p15 and primers for -actin as internal control. PCR products were hybridized by radiolabeled cDNA probes and autoradiography was performed. Bone marrow mononuclear cells were used as positive control. On the bottom of the figure, densitometric analysis is shown as relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. The p15 signal becomes detectable at day 6 and peaks at day 12, when the majority of cells are myelocytes and metamyelocytes.
522
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
the addition of the enzyme; one reaction using this sample was included for each set of primers. An intron is present between this two primers so that any contaminating genomic DNA would have given a product of different size. A set of primers for -actin [25] was used in each PCR reaction as positive control. PCR products were transferred onto nylon membranes (Amersham, Arlington Heights, IL) after agarose gel electrophoresis, and standard hybridization was performed using cDNA probes, as previously described [25]. Densitometric analysis was performed for all the samples using the Molecular Imager (Biorad, Hercules, CA), and ratios of p15 vs -actin were obtained for quantitative analysis.
Figure 2. Flow cytometric evaluation of DNA content in CD34⫹ cells and in cells differentiating toward the myeloid lineage (propidium iodide staining). The percentage of cells in S⫹G2M phases of the cell cycle is shown. DNA distribution histograms were generated and analyzed with ModFIT software.
Results Normal CD34⫹ cell differentiation Myeloid differentiation. p15 was undetectable by RT-PCR in normal bone marrow CD34⫹ cells. As expected on the basis of our previous observations, we found a progressive increase of p15 during the granulomonocytic commitment [10], starting from day 6 and peaking on day 12 (Fig. 1 ). A signal was detectable when the majority of cells displayed
Figure 3. p15 evaluation in cultures of CD 34⫹ cells differentiating toward the erythroid lineage. At day 0, 6 and 12 cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (MayGrunwald staining). myel ⫽ contaminant cells belonging to the granulomonocytic lineage, at different stage of differentiation; proeryth ⫽ proerythroblasts; bas ⫽ basophilic erythroblasts; poly ⫽ polychromatic erythroblasts; orth ⫽ orthochromatic erythroblasts. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. Lane 1 ⫽ day 0 CD34⫹ cells; lane 2 ⫽ day 6 glycophorin-A⫺ cells; lane 3 ⫽ day 6 glycophorin-A⫹ cells; lane 4 ⫽ day 12 glycophorin-A⫺ cells; lane 5 ⫽ day 12 glycophorin-A⫹ cells. (C) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to the -actin signals. P15 is not expressed in erythroid cells, either at day 6 or day 12. The p15 signal detectable at day 12 in the Gly⫺ cell fraction is due to the granulomonocytic cell contaminant.
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
morphologic features of promyelocytes; nevertheless, p15 mRNA expression was maximal when cell population predominantly consisted of myelocytes and metamyelocytes (day 12, Fig. 1A). The percentage of cycling cells increased progressively from day 0 (15% to 30% of CD34⫹ cells in S phase) to day 6 (up to 46 % of S phase), and it rapidly declined afterward (8% to 12 % of S phase on day 15) (Fig. 2). Therefore, the cycling activity of differentiating cells was inversely related to the modulation of p15 expression.
523
ter ATRA exposure. Interestingly, in the other patient we observed a downregulation of gene expression during the first 24 hours of culture with ATRA (Fig. 7). In all cases, when fresh blasts were cultured without ATRA, no modifications of p15 message were detectable. In all APL samples, ATRA-induced differentiation was confirmed by the increase of the CD11b expression at immunocytometry, without significant differences between patients with or without ATRA syndrome.
Erythroid differentiation. According to our previous observations [10], p15 was undetectable throughout erythroid differentiation (Fig. 3). p15 mRNA was detectable only in the Gly⫺ cell population recovered at day 12, whereas Gly⫹ cells recovered at days 6 and 12 were p15⫺. The Gly⫺ cells recovered at day 6 also were p15⫺. Morphologic examination showed that most cells were undifferentiated blasts or erythroid differentiating cells, and no myeloid cells were detectable (Fig. 3A). Megakaryocytic differentiation. p15 was strongly expressed during the megakaryocytic differentiation: both CD41⫹ and CD41⫺ cell populations showed high levels of p15 mRNA at days 6 and 12 (Fig. 4).This finding was quite surprising, considering that in a previous immunohistochemical study of bone marrow biopsies we did not find p15 expression in megakaryocytes [10]. In contrast, in this unilineage culture system, p15 was clearly detectable at day 6, when most cells were undifferentiated blasts or megakaryoblasts (Fig. 4A). Considering that no differentiated myeloid elements were detectable by this time in these culture conditions, the finding of p15⫹ cells in CD41⫺ population suggests that p15 is sharply upregulated in the early phase of the megakaryocytic commitment, before the appearance of the CD41 antigen; thereafter, its expression was maintained at high levels throughout the time of culture. According to the high expression of p15, the percentage of cycling cells was low at days 6 and 12 (9% to 13 % and 8% to 11% of cells in S phase, respectively; data not shown). Expression of p15 in NB4 cell line and APLs. In the NB4 cell line we detected low basal levels of p15 mRNA. However, when this cell line was induced to differentiate with ATRA, a rapid upregulation of p15 was observed, followed by a slight decrease and a late peak at hour 72 (Fig. 5). NB4 cells cultured in the absence of ATRA did not show any significant variation in p15 mRNA expression (Fig. 5). Similar results were obtained in five of seven APL patients. The expression of p15 was dramatically upregulated during the first 6 hours of ATRA exposure. Only a slight increase was observed in the following hours of culture, with a late peak at 72 hours (Fig. 6). In two APL patients who developed ATRA syndrome, we found a completely different modulation of p15 mRNA expression. Blasts from these patients showed a high basal expression of p15 mRNA. In one of these patients, p15 mRNA levels did not increase further af-
Figure 4. p15 evaluation in cultures of CD 34⫹ cells differentiating toward the megakaryocytic lineage. At day 0, 6 and 12 cells were harvested for morphologic analysis and RNA isolation. (A) Morphologic analysis (May-Grunwald staining). myel ⫽ contaminant cells belonging to the granulomonocytic lineage, at different stage of differentiation; megak bl ⫽ megakaryoblasts; early meg ⫽ early megakaryocytes; mat meg ⫽ mature megakaryocytes. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. Lane 1 ⫽ day 0 CD34⫹ cells; lane 2 ⫽ day 6 CD41⫺ cells; lane 3 ⫽ day 6 CD41⫹ cells; lane 4 ⫽ day 12 CD41⫺ cells; lane 5 ⫽ day 12 CD41⫹ cells. (C) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to the -actin signals. p15 is expressed throughout the entire culture period.
524
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Discussion We had demonstrated by immunohistochemistry in a previous study that p15 is largely expressed in normal cells belonging to the granulocytic and monocytic lineages [10]. In this study, we confirmed that p15 expression is upregulated when hematopoietic progenitors undergo myeloid differentiation. p15 mRNA, undetectable in CD34⫹ progenitors, becomes detectable when the majority of cells differentiating toward the granulomonocytic lineage are promyelocytes, i.e., the most differentiated myeloid elements able to proliferate; thereafter, as more mature nonproliferating myeloid cells predominate, p15 mRNA levels progressively increase. In a previous study, Steinman et al. [5] showed that p21 is upregulated during myeloid differentiation of CD34⫹ cells isolated from cord blood. The authors found that p21 expression depends on a new regulatory activity that binds a 44-bp fragment of the p21 promoter; this binding activity is highly expressed in normal differentiated cells and in HL60 cells treated with ATRA. That both p15 and p21 genes are upregulated in myeloid cells agrees with the recent observation showing that two cyclin-dependent kinases inhibitors, belonging to different families of cyclin-dependent kinases
Figure 6. p15 evaluation in fresh blasts obtained from a patient with acute promyelocytic leukemia at diagnosis. Cells were cultured in the presence of all-trans retinoic acid (ATRA) 10⫺7M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. For NB4 cells, p15 message is promptly induced by ATRA (hour 6); thereafter, only a slight increase peaking at 72 hours is found.
Figure 5. p15 evaluation in NB4 cell line cultured with and without alltrans retinoic acid (ATRA) 10⫺7 M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals. (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. p15 message, expressed at a low level in basal samples, is sharply upregulated in cells treated with ATRA.
inhibitors, cooperate to induce cell cycle arrest [3,4]. In addition, p15 and p21 could have different roles, specific for each of them, in the early or late phase of myeloid differentiation, respectively, when predominantly p21 [26] or p15 is expressed. It was reported recently that both p21 and p27 are expressed in megakaryocytic colonies derived from CD34⫹ progenitors [27]. In this study, we also show that p15 is expressed during megakaryocytic differentiation and that its upregulation occurs before the appearance of CD41⫹ megakaryoblastic cells. In a previous study, we investigated the presence of p15 in megakaryocytes using dual-color immunofluorescence analysis of bone marrow cells with antiCD61 and anti-p15 antibodies and immunohistochemical study of bone marrow biopsies. We did not found p15 expression in megakaryocytes [10]. This discrepancy can be due to the fact that by immunohistologic and immunocytologic analysis we examined megakaryocytes at different stages of differentiation, whereas CD34⫹ cells give rise a homogeneous megakaryocytic cell population. Moreover, the expression of p15 in the megakaryocytic lineage could
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
Figure 7. p15 evaluation in fresh blasts obtained from a patient with acute promyelocytic leukemia at diagnosis, who developed all-trans retinoic acid (ATRA) syndrome during the treatment. Cells were cultured in the presence of ATRA 10⫺7 M. (A) Relative intensity of p15 signals, deduced on the basis of densitometric scanning relative to -actin signals (B) RT-PCR analysis of p15 and -actin, performed as described in Figure 1. The p15 mRNA level is high in basal samples and is downregulated by ATRA during the culture.
be related to megakaryocyte production of TGF-, an important inhibitor of megakaryocyte growth and endomitosis [28]. It was demonstrated recently that high levels of TGF-, as observed in patients with autoimmune thrombocytopenia, are able to regulate in vivo megakaryocytopoiesis by inducing marrow stromal cells to produce thrombopoietin; thrombopoietin, in turn, upregulates the expression of TGF- receptors on the megakaryocyte surface [29]. We can hypothesize that, in the unilineage culture system, high levels of thrombopoietin induce p15 by sensitizing megakaryocytes to the effect of TGF- produced by themselves. Finally, the finding that p15 upregulation is selectively induced in myeloid and megakaryocytic, but not erythroid, lineage suggests that its gene could be regulated by transcriptional activators specific to these lineages. Liu et al. [30] demonstrated that in U937 myelomonocytic cell line, which was differentiated by retinoic acid, the gene encoding p21 is transcriptionally activated by the retinoic acid receptor. As described for myeloid differentiation, retinoids play
525
an important role in the regulation of normal megakaryocytopoiesis [31]. Interestingly, patients receiving ATRA as differentiating therapy often show an increase of platelet count above normal values [32]. Whereas the p15 gene deletion has been documented mainly in T-lymphoid neoplasms [11], altered p15 expression due to p15 gene methylation is very frequently observed in AMLs [12,13,15] and myelodysplastic syndromes [14,16]. In this study, we show that, according to p15 modulation, leukemic promyelocytes can be distinguished into two groups. In one group, including the majority of APL samples and the NB4 cell line, p15 mRNA is undetectable or detected at very low levels by RT-PCR; nevertheless, ATRA exposure is able to induce rapidly p15 expression in these cells, showing that p15 gene is not hypermethylated in such leukemias. In contrast, in blasts obtained from patients developing ATRA syndrome, p15 is highly expressed in untreated cells and ATRA treatment downmodulates p15 mRNA. NB4 cell line was derived from APL blasts of the patient who was treated with ATRA who did not develop ATRA syndrome [33]; therefore, it is not surprising that NB4 cells show low basal levels of p15. ATRA is commonly used as a differentiating agent in the treatment of APL, and its association with anthracyclines has greatly improved the outcome of APL patients [19]. The most important side effect of ATRA administration is the development of ATRA syndrome, which is well defined on the basis of the occurrence of clinical signs and symptoms [22]. ATRA syndrome is a life-threatening complication, but unfortunately no predictive criteria are available, so that, in the presence of clinical suspicion, it is mandatory to immediately stop ATRA administration and to start dexamethasone administration [34]. The etiology of the ATRA syndrome is unknown; all the suggested pathogenetic mechanisms indicate that APL blasts from patients who develop ATRA syndrome release greater amounts of cytokines and growth factors in comparison to blasts from APL patients without ATRA syndrome [35–37]. If so, the high expression of p15 mRNA in cells from these patients could reflect a more advanced level of differentiation in comparison to patients without ATRA syndrome. This hypothesis is supported further by recent observations showing high basal levels of C/EBP⑀ in blasts from patients with ATRA syndrome [38]. In conclusion, this study definitely establishes the association between p15 upregulation and normal and leukemic myeloid differentiation. The observation of a different p15 modulation by ATRA in patients who develop ATRA syndrome offers new insights into better understanding the pathogenesis of this syndrome.
Acknowledgments This work was supported by grants from Associazione Italiana Ricerca sul Cancro and from M.U.R.S.T.
526
L. Teofili et al./Experimental Hematology 28 (2000) 519–526
References 1. Hirama T, Koeffler HP (1995) Role of the cyclin-dependent kinase inhibitors in the development of cancer. Blood 86: 841 2. Hall M, Bates S, Peters G (1995) Evidence for different modes of action of cyclin-dependent kinase inhibitors: p15 and p16 bind to kinases, p21 and p27 bind to cyclins. Oncogene 11:1581 3. Reynisdottir I, Polyak C, Iavarone A, Massagué J (1995) Kip/Cip and Ink4 Cdk inhibitors cooperate to induce cell cycle arrest in response to TGF-. Genes Dev 9:1831 4. Reynisdottir I, Massagué J (1997) The subcellular localizations of the p15Ink4b and p27Kip1 coordinate their inhibitory interactions with cdk4 and cdk2. Genes Dev 11:492 5. Steinman RA, Huang J, Yaroslavskiy B, Goff JP, Ball ED, Nguyen A (1998) regulation of p21(WAF1) expression during normal myeloid differentiation. Blood 91:4531 6. Steinman RA, Hoffman B, Iro A, Guillouf C, Liebermann DA, ElHouseini ME (1994) Induction of p21(WAF1/CIP1) during differentiation. Oncogene 9:3389 7. Jiang H, Lin J, Su ZZ, Collart FR, Huberman E, Fisher PB (1994) Induction of differentiation in human promyelocytic HL-60 leukemia cells activates p21WAF1/CIP1 expression in the absence of p53. Oncogene 9:3397 8. Zhang W, Grasso L, McClain CD, et al. (1995) p53-independent induction of WAF1/CIP1 in human leukemia cells is correlated with growth arrest accompanying monocyte/macrophage differentiation. Cancer Res 55:668 9. Liu M, Lee MH, Cohen M, Bommakanti M, Freedman LP (1996) Transcriptional activation of the cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev 10:142 10. Teofili L, Rutella S, Chiusolo P, et al. (1998) Expression of p15INK4B in normal hematopoiesis. Exp Hematol 26:1133 11. Drexler HG (1998) Review of alterations of the cyclin-dependent kinase inhibitor INK4 family genes p15, p16, p18 and p19 in human leukemia-lymphoma cells. Leukemia 12:845 12. Herman GJ, Jen J, Merlo A Baylin SB (1996) Hypermethylation-associated inactivation indicates a tumor suppressor role for p15INK4B1. Cancer Res 56:722 13. Herman JG, Civin CI, Issa JPJ, Collector MI, Sharkis SJ, Baylin SB (1997) Distinct pattern of inactivation of p15INK4B and p16INK4A characterize the major types of hematological malignancies. Cancer Res 57:837 14. Uchida T, Kinoshita T, Nagai H, et al. (1997) Hypermethylation of the p15INK4B gene in myelodysplastic syndromes. Blood 90:1403 15. Aggerholm A, Guldberg P, Hokland M, Hokland P (1997) Extensive intra- and interindividual heterogeneity of p15INK4B methylation in acute myeloid leukemia. Cancer Res 59:436 16. Cameron EE, Baylin SB, Herman JG (1999) p15INK4B CpG island methylation in primary acute leukemia is heterogeneous and suggests density as a critical factor for transcriptional silencing. Blood 94:2445 17. Quesnel B, Guillerm G, Vereecque R, et al. (1998) Methylation of the p15INK4B gene in myelodysplastic syndromes is frequent and acquired during disease progression. Blood 91:2985 18. Larocca LM, Teofili L, Salutari P, et al. (1996) Expression of p15INK4B in primary leukemic blasts and its modulation by quercetin. Exp Hematol 24:1122 (abstr) 19. Grignani F, Fagioli M, Alcalay M, et al. (1994) Acute promyelocytic leukemia: from genetics to treatment. Blood 83:10 20. Bennet JM, Catowsky D, Daniel MT, et al. (1976) Proposal for the classification of the acute leukemias. Br J Haematol 33:451
21. Diverio D, Riccioni R, Mandelli F, Lo Coco F (1995). The PML/ RARa fusion gene in the diagnosis and monitoring of acute promyelocytic leukemia. Hematologica 80:155 22. Frankel SR, Eardley A, Lauwers G, Weiss M, Warrel R (1992) The “retinoic acid syndrome” in acute promyelocytic leukemia. Ann Intern Med 117:292 23. Pierelli L, Scambia G, Fattorossi A, et al. (1998) Functional, phenotypic and molecular characterization of cytokine low-responding circulating CD34⫹ haematopoietic progenitors. Br J Haematol 102: 1139 24. De Stefano V, Teofili L, Sica S, et al. (1995) Effect of all trans retinoic acid on procoagulant and fibrinolytic activities of cultured blast cells from patients with acute promyelocytic leukemia. Blood 86:3535 25. Morosetti R, Park DJ, Chumakov AM, et al. (1997) A novel myeloid transcription factor, C/EBP⑀, is up-regulated during granulocytic but not monocytic differentiation. Blood 90:2591. 26. Yaroslavskji B, Watkins S, Donnenberg AD, Patton TJ, Steinman RA (1999) Subcellular and cell-cycle expression profiles of CDK-inhibitors in normal differentiating myeloid cells. Blood 93:2907 27. Taniguchi T, Endo H, Chikatsu N, et al. (1999) Expression of p21Cip1/ Waf1/Sdi1 and p27Kip1 cycling dependent kinases inhibitors during human hematopoiesis. Blood 93:4167 28. Kuter DJ, Gminski DM, Rosemberg RD (1992) Transforming growth factor- inhibits megakaryocyte growth and endomitosis. Blood 79:619 29. Sakamaki S, Hirayama Y, Matsunaga T, et al. (1999) Transforming growth facto-1 (TGF-1) induces thrombopoietin from bone marrow stromal cells, which stimulates the expression of TGF- receptor on megakaryocytes and, in turn, renders them susceptible to suppression by TGF- itself with high specificity. Blood 94:1961. 30. Liu M, Iavarone A, Freedman LP (1996) Transcriptional activation of the human p21WAF1/CIP1 gene by retinoic acid receptor. J Biol Chem 271:31723 31. Visani G, Ottaviani E, Zauli G, et al. (1999) All-trans retinoic acid at low concentration directly stimulates normal adult megakaryocytopoiesis in the presence of thrombopoietin or combined cytokines. Eur J Hematol 63:149 32. Kinjo K, Kizaki M, Takayama N, et al. (1999) Serum thrombopoietin and erythropoietin levels in patients with acute promyelocytic leukemia during all-trans retinoic acid treatment. Br J Haematol 105:382 33. Lanotte M, Martin-Thouvenin V, Najman S, Balerini P, Valensi F, Berger R (1991) NB4, a maturation inducible cell line with t(15;17) marker isolated from a human acute promyelocytic leukemia (M3). Blood 77:1080 34. De Bottom S, Dombret H, Sanz S, et al. (1998) Incidence, clinical feature and outcome of all trans-retinoic acid syndrome in 413 cases of newly diagnosed acute promyelocytic leukemia. Blood 92:2712 35. Larson R, Brown D, Sklar L (1997) Retinoic acid induces aggregation of the acute promyelocytic leukemia cell line NB4 by utilization of LFA-1 and ICAM-2. Blood 90:2747 36. Grande A, Manfredini R, Tagliafico E, et al. (1995) All trans retinoic acid induces simultaneously granulocytic differentiation and expression of inflammatory cytokines in HL-60 cells. Exp Hematol 23:117 37. Dubois C, Schlageter MH, De Gentile A, et al. (1994) Modulation of IL-8 and IL-1b and G-CSF secretion by all trans retinoic acid in acute promyelocytic leukemia. Leukemia 8:1750 38. Morosetti R, Martini M, Teofili L, et al. (1999) ATRA up-regulates the expression of C/EBP⑀ in blasts from APL patients in a secondary AML, but not in blasts from APL patients with ATRA syndrome. Cancer Res 40:4069 (abstr)