MEGAKARYOCYTIC DIFFERENTIATION OF HIMeg-1 CELLS INDUCED BY INTERFERON γ AND TUMOUR NECROSIS FACTOR α BUT NOT BY THROMBOPOIETIN

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248

Article No. ck980345

MEGAKARYOCYTIC DIFFERENTIATION OF HIMeg-1 CELLS INDUCED BY INTERFERON g AND TUMOUR NECROSIS FACTOR a BUT NOT BY THROMBOPOIETIN Jian Li,1 Robert S. Franco,1 Yisheng Wang,2 Hui-Qi Pan,1 Dan Eaton,3 Tao Cheng,4 Kenneth Kaushansky,5 Wei Dai1 Activated macrophage-conditioned medium (M-CM) induces megakaryocytic differentiation of HIMeg-1 cells. The megakaryocytic differentiation activity (MDA) is proteinaceous since it is susceptible to treatments by proteinases, heat, and reducing agents. MDA is not thrombopoietin (TPO) since (1) TPO alone or in conjunction with several other recombinant cytokines fails to induce any degree of HIMeg-1 cell differentiation; and (2) a neutralizing antibody against TPO or an antibody against the extracellular domain of c-mpl is unable to abolish M-CM-induced CD41 expression on HIMeg-1 cells. Reverse transcriptase-mediated polymerase chain reaction shows that HIMeg-1 cells express c-mpl but not TPO. Additional neutralizing antibody studies suggest that MDA is not one of the cytokines known to induce some degree of megakaryopoiesis in vitro or in vivo including interleukin 3 (IL-3), IL-6, IL-11, granulocyte–macrophage colony-stimulating factor, erythropoietin, or stem cell factor. On the other hand, MDA appears to be a combination of interferon g (IFN-g) and tumour necrosis factor a (TNF-a), since neutralizing antibodies against these two cytokines completely abolish MDA-induced CD41 expression. In addition, either recombinant human IFN-g or TNF-a alone is capable of inducing CD41 and CD42 expression on HIMeg-1 cells. In combination, IFN-g and TNF-a induce a maximal level of CD41 and CD42 expression, which is also accompanied by an increase in cell size and DNA ploidy level. Thus, our studies indicate that IFN-g/TNF-a is capable of inducing megakaryocytic differentiation of the HIMeg-1 cell line and that HIMeg-1 is a good system for studying the molecular mechanism mediating megakaryocytic differentiation. 7 1998 Academic Press

Megakaryocytes are derived from pluripotent haematopoietic stem cells through a complex process including proliferation, differentiation, and maturation. This developmental process is regulated by a spectrum of soluble polypeptide hormones commonly called as cytokines or blood cell growth factors.1,2 In the past decade, recombinant technology has enabled molecular cloning of cytokines regulating the proliferation and/or differentiation of haematopoietic progenitors of various lineages. Because of the lack of a quick,

*From the 1Division of Hematology/Oncology, Department of Internal Medicine, University of Cincinnati College of Medicine; 2 Department of Molecular and Cellular Physiology, University of Cincinnati College of Medicine; 3Division of Cardiovascular Research, Genentech, Inc.; 4The MGM Cancer Center, Massachusetts General Hospital; 5Division of Hematology/Oncology, Washington State University Correspondence to: Wei Dai, Division of Hematology/Oncology, Department of Internal Medicine, University of Cincinnati College of Medicine, ML-508, K-Pavilion, 231 Bethesda Avenue, Cincinnati, OH 45267-0508; E-mail: wei.dai.uc.edu Received 10 October 1997; received in revised form 17 December 1997; accepted 29 January 1998 7 1998 Academic Press 1043–4666/98/110880 + 10 $30.00/0 880

sensitive and specific assay, it has been difficult, until recently, to identify the elusive factor(s) regulating the proliferation and/or differentiation of progenitor cells of the megakaryocytic lineage. About 3 years ago, several groups reported the molecular cloning and functional characterization of thrombopoietin (TPO), a cytokine specifically regulating the production of megakaryocytes and platelets.3–5 However, there is evidence in the literature suggesting the presence of thrombopoietic activities in addition to TPO.6–8 In the past, it was widely accepted that the early proliferative response of committed megakaryocytic precursor cells was stimulated by megakaryocyte colony-stimulating activity (Meg-CSA), and the cell maturation processes were stimulated by thrombopoietin-like activity that was different from Meg-CSA6,7. Recently, it has been shown that mice with targeted disruption at the c-mpl (coding for the TPO receptor) and TPO genomic loci contain a low level but functionally normal megakaryocytes and platelets,8,9 suggesting the existence of a complementary pathway for megakaryopoiesis. Although some known cytokines such as interleukin 3 (IL-3), IL-6, IL-11, granulocyte–macrophage colonystimulating factor (GM-CSF), stem cell factor (SCF), CYTOKINE, Vol. 10, No. 11 (November), 1998: pp 880–889

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 Leukaemia cell differentiation induced by cytokines / 881

and erythropoietin (EPO) are implicated in promoting, to some extent, megakaryopoiesis in vivo and/or in vitro,10–14 their physiological relevance in megakaryocyte development remains unclear. HIMeg-1 is a subclone of the HIMeg cell line that was originally established from a patient with chronic myelogenous leukaemia.15,16 HIMeg-1 cells are essentially negative for CD41 (a megakaryocyte/plateletspecific marker) but express very high levels of CD33 (an early myeloid progenitor marker).16 These cells also express a low level of CD11c and glycophorin A. HIMeg-1 is capable of multi-lineage differentiation when exposed to appropriate stimuli, and when treated with phorbol 12-myristate 13-acetate (PMA), HIMeg-1 cells express an enhanced level of CD11c and CD14 (both are myeloid markers).16 On the other hand, treatment with activated lymphocyte-conditioned medium induces some degree of CD41 expression on HIMeg-1 cells.16 We report here that conditioned medium from peripheral blood macrophages treated with Phytolecca americana contains an activity that is capable of inducing megakaryocytic differentiation of HIMeg-1 cells in vitro. This megakaryocytic differentiation activity (MDA) is proteinaceous and is not TPO. Our further studies have revealed that MDA present in macrophage-conditioned medium (M-CM) is a combined activity of interferon g (IFN-g) and tumor necrosis factor a (TNF-a).

RESULTS M-CM contains an MDA HIMeg-1 and its parental cell line HIMeg are capable of megakaryocytic differentiation when they are treated with retinoids, vitamin D3 or a lymphocyteconditioned medium.15,16 This differentiation is characterized by an enhanced expression of CD41 or an increase in cell size and DNA ploidy levels.15 To understand better the mechanism of differentiating agents in regulating megakaryopoiesis, we tested several other conditioned media including M-CM for their megakaryocytic differentiation activities using the HIMeg-1 cell line as a model system. HIMeg-1 cells treated with the vehicle expressed barely detectable levels of CD41 but high levels of CD33 (Fig. 1B). However, when cultured in RPMI-1640 medium supplemented with 20% M-CM for 3 days, HIMeg-1 cells were consistently observed to express an elevated level of CD41, whereas CD33 expression was not affected by M-CM treatment (Fig. 1D). A higher percentage of CD41-positive cells was obtained when more conditioned medium was supplemented to the HIMeg-1 culture (Table 1). These data strongly suggest that M-CM contains an MDA.

MDA is proteinaceous Since small chemical compounds such as vitamin D3 and retinoids are known to induce megakaryocytic differentiation of HIMeg cells,15 we asked whether or not MDA in M-CM was proteinaceous. M-CM as well as the vehicle (IMDM with 5% FBS) were boiled. After centrifugation, the boiled supernatant of each medium was collected for the treatment of HIMeg-1 cells. Figure 2A shows that the boiled M-CM (20% final concentration) induced little, if any, CD41 expression as compared with the vehicle-treated negative control. On the other hand, the unboiled M-CM (20%) induced significant CD41 expression as expected. As an alternative approach to confirm the proteinaceous property of MDA, M-CM was pre-incubated with pronase, or bovine serum albumin (BSA) as a control, that had been immobilized onto Affi-gel 10 beads. After overnight incubation, M-CM was collected after centrifugation, and sterilized by filtration. HIMeg-1 cells were then cultured in the presence of the pronase- or the BSA-treated M-CM. Figure 2B shows that pronase-treated M-CM lost its capability to induce CD41 expression. On the other hand, M-CM incubated with BSA-coupled beads induced CD41 expression with a potency close to that of the untreated M-CM (Fig. 2B). In addition, a SDS-PAGE analysis showed that no discrete protein bands were observed in pronase-treated M-CM samples (data not shown), indicating an efficient digestion by the immobilized pronase. To determine whether one or more disulfide bonds were essential for the activity of MDA, M-CM and the vehicle were treated with the reducing agent dithiothreitol (DTT). The treated media were then thoroughly dialysed against phosphate-buffered saline (PBS) and used for the treatment of HIMeg-1 cells. At the end of culture, HIMeg-1 cells were analysed for CD41 expression. It was shown (Fig. 2C) that compared with the positive control (20% regular M-CM), DTT treatment markedly reduced the ability of M-CM to induce CD41 expression on HIMeg-1 cells, again suggesting that MDA is a protein and that a disulfide bond(s) is important for its activity.

MDA is different from thrombopoietin (TPO) TPO is a lineage-specific cytokine primarily regulating the production of megakaryocytes and platelets.1 We asked whether MDA is TPO. HIMeg-1 cells were treated for three days with rhTPO at final concentrations of 2, 20, and 200 ng/ml, respectively. The treated cells were then analysed for CD41 expression. Figure 3B shows that no CD41 expression was activated by rhTPO as compared with the negative control (Fig. 3A). On the other hand, M-CM (20%)

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 882 / Li et al.

Figure 1.

CYTOKINE, Vol. 10, No. 11 (November, 1998: 880–889)

CD41 expression induced by M-CM.

HIMeg-1 cells cultured in medium supplemented with the vehicle (A,B) or M-CM (C,D) were either double labelled with isotype control IgGs (A,C) or double labelled with PE-conjugated anti-CD33 and FITC-conjugated anti-CD41 monoclonal antibodies (B,D).

stimulated CD41 expression (Fig. 3C). As an alternative approach, HIMeg-1 cells were supplemented with a TPO-neutralizing antibody at a final

TABLE 1. Concentration-dependency of CD41 expression on M-CM Amount of M-CM 0% 5% 10% 20% 50%

% CD41 positive cells

W test

1 2 0.2 4 2 0.5 10 2 0.8 32 2 1.6 55 2 3.3

P Q 0.001 P Q 0.01 P Q 0.01 P Q 0.01

Results were summarized from three individual experiments (mean 2 SEM) as analysed by flow cytometry. Differences between each treatment were analysed using Wilcoxon’s signed-rank test (W test).

concentration of 3.2 mg/ml along with 20% M-CM. After three days’ incubation, the treated HIMeg-1 cells were analysed for CD41 expression. Compared with the positive control (20% M-CM, Fig. 3C), no reduction in CD41 expression was observed when HIMeg-1 cells were incubated with the anti-TPO antibody (Fig. 3D). Similar experiments were performed by supplementation of M-CM-treated HIMeg1 cells with an anti-c-mpl antibody at final concentrations of 2 mg/ml and 20 mg/ml, respectively. No modulation of CD41 expression by the anti-c-mpl antibody was observed either (Table 3). The TPO antibody is effective in blocking murine CFU-Meg (colony-forming units-megakaryocyte) assays (data not shown). Thus, all the above data strongly suggest that MDA and TPO are different.

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 Leukaemia cell differentiation induced by cytokines / 883

It was observed that no single antibody against the above cytokines significantly reduced the ability of M-CM to activate CD41 expression (Table 3), suggesting that these individual cytokines may not be involved in regulating megakaryocytic differentiation of HIMeg-1 cells.

MDA is a combined activity of IFN-g and TNF-a

Figure 2.

MDA is proteinaceous.

M-CM was treated with heat (A), pronase (B), or DTT (C) as described in Materials and Methods. The treated M-CM as well as untreated (or vehicle treated) M-CM were added to HIMeg-1 cell cultures. At the end of the treatments, HIMeg-1 cells were analysed for CD41 expression via flow cytometry. Representative results from two independent experiments were shown.

MDA appears to differ from those cytokines reported to be involved in regulating megakaryopoiesis Before the molecular cloning of TPO, various recombinant human cytokines such as IL-3, IL-6, IL-11, GM-CSF, EPO and SCF were shown to be involved, or play a role, in megakaryopoiesis.10–14 To determine whether MDA is one of these factors, HIMeg-1 cells were treated with the individual cytokines listed in Table 2. In some experiments, three different cytokines (SCF, IL-3 and TPO), or seven cytokines (SCF, IL-3, GM-CSF, IL-6, IL-11, EPO and TPO) were added to HIMeg-1 cell cultures. Flow cytometric analyses showed that neither individual cytokines nor combinations stimulated CD41 expression on HIMeg-1 cells (Table 2). Experiments were also performed in which HIMeg-1 cells were treated with 20% M-CM supplemented with a neutralizing antibody against SCF, IL-3, IL-6, IL-11, GM-CSF, EPO, TPO or c-mpl.

Extensive screening using antibodies against known cytokines revealed that anti-IFN-g or antiTNF-a antibodies significantly reduced the capability of M-CM to induce CD41 expression on HIMeg-1 cells (Table 3). When combined, anti-IFN-g and anti-TNFa antibodies completely abolished MDA in M-CM (Table 3). To confirm that MDA in M-CM is partly or fully due to the presence of IFN-g and/or TNF-a, HIMeg-1 cells were cultured in the presence of various concentrations of IFN-g and/or TNF-a. After 3 days, cells were collected for analysis of CD41 expression. It was observed (Fig. 4A) that either IFN-g or TNF-a alone induced a significantly higher percentage of CD41 positive cells than the vehicle (BSA). For both cytokines the CD41 cell positivity is dose-dependent up to 20 ng/ml. It was shown that 40 ng/ml of TNF-a and 40 ng/ml IFN-g achieved the maximal level of induction, about 40% and 55%, respectively. A combination of these two cytokines further stimulated the percentage of CD41 positive cells (Fig. 4A), suggesting an additive effect of IFN-g and TNF-a. To confirm further that IFN-g and TNF-a are capable of inducing megakaryocytic differentiation, we examined CD42 surface antigen expression on HIMeg-1 cells treated with or without these two cytokines. Flow cytometric analyses showed that CD42 was induced by either IFN-g or TNF-a in a concentration-dependent manner, and a combination of both cytokines induced maximal levels of CD42 expression (Fig. 4B). The HIMeg-1 cell line contains a very small percentage (less than 1%) of spontaneously differentiated cells, which are characterized by low level expression of CD41 and an increase in cell size. We consistently noticed that when supplemented with optimal concentrations of IFN-g and TNF-a for 2 weeks, HIMeg-1 culture contained up to 10% large cells, and flow cytometric analyses of propidium iodide-stained cells confirmed an increase in the DNA ploidy level in large cells (data not shown). When these cells were studied using immunocytochemistry, most of HIMeg-1 cells treated with a combination of IFN-g and TNF-a were positive for CD41 (Fig. 5). The large cells enriched in the cytokine-treated samples were all strongly positive for CD41 antigen (Fig. 5B). On the other hand, almost all vehicle-treated control cells were negative for CD41 staining (Fig. 5A). No positive

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 884 / Li et al.

signals could be detected in isotype control IgG-stained cells (data not shown).

HIMeg-1 cells express c-mpl but not TPO TPO does not appear to mediate HIMeg-1 cell differentiation based on cytokine and neutralizing antibody experiments (Table 3 and Fig. 3). However, since neutralizing antibodies only work to block cytokine loops that involve extra-cellular expression, it is conceivable that TPO, potentially activated by IFN-g and/or TNF-a, may work through an internal cellular loop. In other words, it is possible that HIMeg-1 may express TPO that interacts with c-mpl on the endoplasmic reticulum, thereby transmitting a differentiation signal. To test this possibility, we examined whether HIMeg-1 cells express TPO and/or c-mpl, and

Figure 3.

CYTOKINE, Vol. 10, No. 11 (November, 1998: 880–889)

whether TPO production is activated by IFN-g and TNF-a via the RT-PCR approach. Figure 6 shows that HIMeg-1 cells express a detectable level of c-mpl but not TPO. In addition, IFN-g and TNF-a treatment did not activate TPO gene expression in HIMeg-1 cells (Fig. 6, lane 6).

DISCUSSION In an attempt to identify factors other than TPO that play a role in regulating megakaryocyte development, we found that the conditioned-medium from plant lectin-stimulated macrophages contained an activity stimulating megakaryocytic differentiation of the HIMeg-1 cell line. Our studies revealed that the

MDA is not TPO.

HIMeg-1 cells treated with the vehicle (A), 200 ng/ml TPO (B), 20% M-CM (C) or 20% M-CM plus anti-TPO antibody (D) were analysed for CD41 and CD33 expression by flow cytometry. Representative results from three independent experiments are shown.

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 Leukaemia cell differentiation induced by cytokines / 885

TABLE 2.

CD41 expression induced by various cytokines

Cytokine EPO GM-CSF IL-3 IL-6 IL-11 TPO SCF IFN-g TNF-a IFN-g + TNF-a TPO + SCF + IL-3 TPO + SCF + IL-3 + GM-CSF + IL-6 + IL-11 + EPO

% CD41 positive cells

W test

1.0 2 0.1 1.1 2 0.2 1.1 2 0.1 1.0 2 0.2 1.0 2 0.2 1.2 2 0.2 1.1 2 0.1 45 2 2.5 38 2 2.2 80 2 3.5 1.2 2 0.2 1.2 2 0.2

P q 0.05 P q 0.05 P q 0.05 P q 0.05 P q 0.05 P q 0.05 P q 0.05 P Q 0.05 P Q 0.01 P Q 0.01 P q 0.05 P q 0.05

Results were summarized from three individual experiments (mean 2 SEM) using flow cytometry. Differences between vehicle (1.1 2 0.1) and each cytokine were analysed using Wilcoxon’s signed-rank test (W test).

megakaryocytic differentiation activity present in M-CM was due to the combined action of IFN-g and TNF-a. IFN-g and TNF-a have been previously implicated in regulating proliferation and differentiation of megakaryocytes both in vivo or in vitro.17–23 For example, TNF-a enhances colony formation by a megakaryoblastic leukaemia cell line CMK.22 IFN-g is shown to be a potentiator for murine megakaryocyte development,19 which may at least partly explain why this cytokine accelerates platelet recovery in mice with marrow aplasia induced by 5-FU treatment.18 IFN-g receptor is expressed on the surface of megakaryoblastic leukaemia cells.21 In addition, it has been observed that the immunomodulator AS101 protects mice from lethal and sublethal effects such as myelosuppression, thrombocytopenia and anaemia from radiotherapy and chemotherapy, and this protection effect is most likely due to the ability of AS101 to induce mouse and human haematopoietic cells to produce several cytokines including IFN-g and TNF-a.18 It has been proposed TABLE 3. Effects of various neutralizing antibodies on M-CM induced CD41 expression Antibody Anti-EPO Anti-GM-CSF Anti-IL-3 Anti-IL-6 Anti-IL-11 Anti-TPO Anti-SCF Anti-c-mpl Anti-IFN-g Anti-TNF-a Anti-IFN-g + Anti-TNF-a

% Inhibition of CD41 expression Q10 Q10 Q10 Q10 Q10 Q10 Q10 Q10 60 2 15 50 2 10 90 2 12

Results were summarized from flow cytometry analyses for three independent experiments. Twenty percent M-CM was used to induce CD41 expression. The percentage of inhibition was obtained by comparing antibody/M-CM-treated cells with the cells treated with M-CM alone.

that AS101-induced cytokines such as IFN-g and TNF-a may play an important role in inducing proliferation and differentiation of megakaryocytic as well as other lineage progenitors.19 Furthermore, it has been shown that megakaryocytic cell lines such as CHRF-288-11 and CMK constitutively produces TNF-a as well as several haematopoietic growth factors,20,23 suggesting a possible autocrine loop during megakaryocyte growth and differentiation. IFN-g and TNF-a have multiple functions such as anti-inflammation and immunomodulation. In addition, these two cytokines are potent inhibitors of cell proliferation. In fact, we have observed that IFN-gand TNF-a-treated HIMeg-1 cells have a greatly reduced proliferation rate as compared with that of the vehicle-treated control (data not shown). It is known that cell differentiation is normally coupled with a decrease in cell proliferation. It is likely that a slow-down in proliferation caused by IFN-g and TNF-a may facilitate turning on the differentiation programme. On the other hand, recent gene knock-out studies indicate that normal platelets and megakaryocytes are produced, albeit at a low level, in mice lacking functional thrombopoietin and its receptor c-mpl.9 This suggests that certain non-TPO cytokines can compensate for the terminal maturation of megakaryocytes. Recent studies have shown that TPO is a lineage-specific cytokine regulating megakaryopoiesis.1,3,5 TPO-induced megakaryocytic proliferation and differentiation of haematopoietic cells is mediated through its cell surface receptor, c-mpl.1,3,5 The binding of TPO to c-mpl results in activation of the receptor and protein tyrosine kinases such as JAK2 and TYK2,24–25 which in turn activate STAT3, and/or STAT5.24 The activated STAT transcription factors, along with other transcription factors, are thought to modulate megakaryocyte-specific gene expression. It is known that IFN-g also activates JAK2 on targeted cells,26 which may explain at least in part the

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 886 / Li et al.

CYTOKINE, Vol. 10, No. 11 (November, 1998: 880–889)

100

% CD41 positive cells

80

60

40

20

0

–20

0

40

80 120 160 Concentration (ng/ml)

200

240

0

40

80 120 160 Concentration (ng/ml)

200

240

100

% CD42 positive cells

80

60

TPO treatment may be due to a structural alteration in c-mpl, resulting in a non-functional receptor. On the other hand, activation of c-mpl alone in non-compatible cell types is insufficient for megakaryocytic differentiation since factor-dependent cell lines such as MO7e and Ba/F3-c-mpl express functional c-mpl and TPO treatment of these cell lines does not stimulate CD41 expression and megakaryocytic differentiation.27 Phorbol esters are capable of inducing megakaryocytic differentiation of erythroleukaemia cell lines (K562 & HEL)30–32 or megakaryoblastic leukaemia cell lines (Dami, CMK).33–34 Phorbol ester-induced megakaryocytic differentiation of these cell lines is characterized by downregulation of erythroid phenotype and upregulation of megakaryocytic phenotype. Although useful in providing some insights into the mechanism of haematopoietic cell differentiation phorbol esters are non-physiological regulators. IFN-g and TNF-a are known to induce myeloid differentiation of leukaemic cell lines,35–37 but few, if any, known cytokines induce dramatic megakaryocytic differentiation of established cell lines. Our current studies have provided unequivocal evidence that

40

20

0

–20

Figure 4. CD41 and CD42 expression induced by IFN-g and TNF-a. HIMeg-1 cells were cultured in the presence of various concentration of BSA (vehicle) (q), IFN-g (Q), TNF-a (w) or TNF-a + IFN-g (W). At the end of the treatments, HIMeg-1 cells were analysed for CD41 and CD42 expression by flow cytometry. Representative results from at least three independent experiments are shown.

differentiation activity of IFN-g on HIMeg-1 cells. On the other hand, the JAK/STAT pathway does not appear to be a part of the TNF-a signal transduction pathway,29 and therefore, TNF-a-induced CD41 expression on HIMeg-1 cells may involve a different mechanism. Both HIMeg-1 and its parental cell line HIMeg are capable of megakaryocytic differentiation when exposed to appropriate stimuli such as retinoids, vitamin D3 or certain conditioned medium.15,16 It is interesting that TPO is unable to induce megakaryocytic differentiation of HIMeg-1 cells (Fig. 3). Our RT-PCR analysis indicates that there is no endogenously produced TPO in HIMeg-1 cells either in the presence or absence of IFN-g and TNF-a (Fig. 6). One possibility exists that HIMeg-1’s unresponsiveness to

Figure 5.

Immunocytochemistry study of HIMeg-1.

HIMeg-1 cells were treated with vehicle (A) or IFN-g (40 ng/ml) plus TNF-a (40 ng/ml) (B). The treated cells were processed for CD41 expression via immunocytochemistry as described in Materials and Methods. CD41-expressing cells are stained with the brown colour. Original maganification ×600, reproduced at 65%.

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 Leukaemia cell differentiation induced by cytokines / 887

were collected, centrifuged (12 500 × g × 10 min) to remove cell debris, and stored in aliquots for subsequent analyses.

Cell culture

Figure 6.

HIMeg-1 cells express c-mpl but not TPO.

Total RNAs from HIMeg-1 were analysed for TPO (lanes 4–6), c-mpl (lanes 7–8), or b-actin (lane 10) expression via RT-PCR. As controls, TPO cDNA (lane 2) or total RNAs from K562 (lane 3) or MO7e (lane 9) cells were also used for analyses. Lane 1, 100-bp ladders; lane 2, PCR product amplified from TPO cDNA using TPO primers; lane 3, RT-PCR products amplified from K562 RNAs using TPO primers; lane 4, RT-PCR products amplified from HIMeg-1 RNAs using TPO primers; lane 5, RT-PCR products amplified from RNAs of HIMeg-1 cells treated with IFN-g and TNF-a using TPO primers; lane 6, PCR products amplified from HIMeg-1 RNAs without reverse transcription using TPO primers; lane 7, PCR products amplified from HIMeg-1 RNAs without reverse transcription using c-mpl primers; lane 8, RT-PCR products amplified from HIMeg-1 RNAs using c-mpl primers; lane 9, RT-PCR products amplified from MO7e RNAs using c-mpl primers; lane 10, RT-PCR products amplified from HIMeg-1 RNAs using b-actin primers.

defined cytokines IFN-g and TNF-a secreted by macrophages induce megakaryocytic differentiation of HIMeg-1 cells. Although significant progress has been made in the past decade in understanding the molecules mediating proliferative signals of activated haematopoietic receptors little is known regarding molecular pathways regulating differentiation and maturation of haematopoietic progenitor cells. Thus, the HIMeg-1 cell line and IFN-g/TNF-a do provide an excellent model system to study cytoplasmic and nuclear events mediating haematopoietic cell differentiation in general and megakaryocytic cell differentiation in particular.

MATERIALS AND METHODS Conditioned medium To obtain M-CM, mononuclear cells collected from the peripheral blood of individual healthy donors were first incubated in RPMI-1640 medium (GIBCO/BRL, Grand Island, NY) containing 10% (V/V) fetal bovine serum (FBS) in Falcon tissue culture plates or flasks for four hours. Non-adherent cells were discarded, and adherent cells (enriched macrophages) were cultured in Iscove’s modified Dulbecco’s medium (IMDM) containing 5% FBS, antibiotics (penicillin, 100 U/ml and streptomycin sulfate 100 mg/ml, GIBCO/BRL), and Phytolecca americana (10 ng/ml, Sigma, St Louis, MO) for three days. At the end of culture, media

HIMeg-1 cells were grown in suspension in RPMI-1640 supplemented with 10% FBS and antibiotics as above. Cultured cells were passed twice each week, seeding at a density of about 2 × 105 cells/ml. When ready to use, cells (1 × 105 cells/well) were seeded into 6-well tissue culture plates (Corning, NY) with 5 ml/well culture medium. M-CM was added to HIMeg-1 cell cultures at various final concentrations (5, 10, 20 or 50%). After incubation for three days, HIMeg-1 cells were collected and analysed for their surface antigen expression by flow cytometry. To determine the effect of known cytokines on HIMeg-1 differentiation, a panel of individual cytokines [EPO (2 U/ml), GM-CSF (20 ng/ml), SCF (10 ng/ml), IL-3 (20 ng/ml), IL-6 (20 ng/ml), IL-11 (20 ng/ml), IFN-g (2–200 ng/ml), TNF-a (2–200 ng/ml) and TPO (2, 20 or 200 ng/ml)], as well as some of their combinations were added to the HIMeg-1 cell cultures. The concentrations of various cytokines (EPO, GM-CSF, SCF, IL-3, IL-6 and IL-11) were based on literatures.17,18 In some experiments, neutralizing antibodies against various cytokines [anti-TPO and anti-c-mpl antibodies (Genentech, San Francisco, CA), anti-GM-CSF (10 mg/ml), anti-SCF (5 mg/ ml), anti-IL-3 (10 mg/ml), anti-IL-6 (2 mg/ml), anti-IL-11 (15 mg/ml), anti-IFN-g (5–20 mg/ml), anti-TNF-a (0.1–1 mg/ ml) antibodies] were also included in the HIMeg-1 cell cultures that were treated with M-CM. The concentrations of neutralizing antibodies used in each experiment were as recommended by the suppliers. Cytokines such as EPO, GM-CSF, IL-3, IL-6, IL-11 and SCF, and antibodies against GM-CSF, IL-3, IL-6, IL-11, SCF, IFN-g and TNF-a were purchased from R & D System (Minneapolis, MN), and IFN-g and TNF-a were purchased from Genzyme (Cambridge, MA).

Analyses of proteinaceous property of MDA (1) Heat treatment: M-CM was boiled for 10 min. After a brief spin to remove any precipitates, the supernatant was collected for treatment of HIMeg-1 cells. (2) Pronase treatment: pronase (20 mg/ml, Sigma) as well as bovine serum albumin (BSA) (20 mg/ml, Sigma) were immobilized onto a solid matrix by incubating with Affi-gel 10 Sepharose beads (50% slurry, Bio-Rad, Hercules, CA) at 4°C for 4 h. After coupling, the beads were washed sequentially with 1 × phosphate-buffered saline (PBS), PBS with 1 M NaCl, and PBS to remove the uncoupled proteins. The pronase- or BSA-coupled beads (200 ml of 50% slurry) were then added to the vehicle (IMDM with 5% FBS) or M-CM, and the mixtures were incubated at room temperature for 18 h. At the end of incubation, the beads were removed by centrifugation. Supernatants from the vehicle or M-CM were sterilized by filtration and the filtered media were tested for their ability to induce CD41 expression. (3) Reducing agent treatment: M-CM and the vehicle were treated with the reducing agent dithiothreitol (DTT, Sigma) at a final concentration of 0.1 M for 4 h at 4°C. The DTT-treated media were then extensively dialyzed against 1 × PBS at 4°C. The dialysed media were

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 888 / Li et al.

sterilized by filtration and used for treatment of HIMeg-1 cells.

Flow cytometry HIMeg-1 cells (2 × 105) were analysed for CD41, CD42 and CD33 surface antigen expression as described.16 Briefly, cells were resuspended in PBS with 0.05% bovine serum albumin (BSA) and incubated with an anti-CD33 monoclonal antibody conjugated with phycoerythrin (PE) (Becton Dickinson, Mountain View, CA) and an anti-CD41 or anti-CD42 monoclonal antibody conjugated with fluorescein isothiocyanate (FITC) (Immunotech, Marseille, France) at 4°C for 30 min. Isotype IgGs conjugated with PE or FITC (Becton Dickinson) were used as negative controls. After three washes with PBS containing 0.05% BSA, the cells were resuspended in the same buffer. The intensity of the staining by each antibody was analysed by flow cytometry (Coulter XL-MCL).

Immunocytochemistry study HIMeg-1 cells with various treatments were collected, washed with 1 × PBS and centrifuged onto slides by cytospin (Shanndon Cytocentrifuge, Pittsburgh, PA). A DAKO LSAB2 kit (DAKO, Carpinteria, CA) was used for cell staining according to the protocol provided by the manufacturer. Endogenous peroxidase activities were first quenched by incubating the specimen for 5 min with 3% hydrogen peroxidase. The specimen was then incubated with an anti-CD41 monoclonal antibody (Immunotech) for 30 min followed by sequential incubations (10 min each) with a biotinylated second antibody and peroxidase-labelled streptavidin. Staining was completed by the addition of a freshly prepared substrate-chromogen solution and counterstaining with haematoxylin.

PCR analysis Total RNA was isolated from HIMeg-1, MO7e or K562 cells using a kit from Tel-Test Inc (Friendwood, TX) according to the protocol provided by the supplier. Reverse transcriptase-mediated polymerase chain reaction (PCR) was performed using a SuperScriptTM pre-amplification system (GIBCO/BRL). A pair of TPO primers has the following sequences: 5'AGC CCG GCT CCT CCT GCT TGT3' (forward) and 5'GTA GTC GGC AGT GTC TG3' (reverse). A pair of c-mpl primers has the following sequences: 5'GAC TGG AAG GTG CTG GAG3' (forward) and 5'TTA GAG TGT AAG GAG CCG3' (reverse). The b-action primer pair has the following sequence: 5'GTG ACG AGG CCC AGA GCA AG3' (forward) and 5'AGG GGC CGG ACT CAT CGT AC3' (reverse). PCR products were analysed on 1.5% agarose gels.

Statistical analysis The significance of means was determined using Wilcoxon’s signed-rank test.

Acknowledgements This work was supported in part by a grant from Genentech Inc.

CYTOKINE, Vol. 10, No. 11 (November, 1998: 880–889)

REFERENCES 1. Kaushansky K (1995) Thrombopoietin: the primary regulator of platelet production. Blood 86:419–431. 2. Hoffman R (1989) Regulation of megakaryopoiesis. Blood 74:1196–1212. 3. de Sauvage WJ, Hass P, Spencer SD, Malloy BE, Gurney AL, Spencer SA, Barbonne WC, Henzel WJ, Wong SC, Kuang WJ, Hultgren B, Solberg LA, Goeddel DV, Eaton DL (1994) Stimulation of megkaryocytopoiesis and thrombopoiesis by the c-mpl ligand. Nature 369:533–538. 4. Kaushansky K, Lok S, Holly RD, Broudy VC, Lin N, Balley MC, Forstrom JW, Buddle MM, Oort P, Hagen FS, Roth GJ, Papayannopoulou T, Foster DC (1994) Promotion of megakaryocyte progenitor expression and differentiation by the c-mpl ligand thrombopoietin. Nature 369:568–571. 5. Bartley TD, Bogenberger J, Hunt P, Li Y-S, Lu HS, Martin F, Chang M-S, Samal B, Nichol JL, Swift S, Johnson J, Hsu R-Y, Parker VP, Suggs S, Skrine JD, Merewether LA, Clogston C, Hsu E, Hokom MM, Hornkohl A, Chol E, Pangelinan M, Sun Y, Mar V, McNinch J, Simonet L, Jacobsen F, Xie C, Shutter J, Chute H, Basu R, Selander L, Trollinger D, Sieu L, Padilla D, Trail G, Elliott G, Izumi R, Covey T, Crouse J, Garcia A, Xu W, Del Castillo J, Biron J, Cole S, Hu MCT, Pacifici R, Ponting I, Saris C, Wen D, Yung YP, Lin H, Bosselman RA (1994) Identification and cloning of a megakaryocyte growth and development factor that is a ligand for the cytokine receptor mpl. Cell 77:1117–1124. 6. Williams N, Eger RR, Jackson HM, Nelson DT (1982) Two factor requirement for murine megakaryocyte colony-formation. J Cell Physiol 110:101–104. 7. Williams N, Jackson H, Walker F, Uon SH (1989) Multiple levels of regulation of megakaryocytopoiesis. Blood Cells 15:123– 133. 8. Gurney AL, Carver-Moore K, de Sauvage F-J, Moore MW (1994) Thrombocytopenia in c-mpl-dependent mice. Science 265:1445–1457. 9. Bunting S, Widmer R, Lipari T, Rangell L, Steimetz H, Carver-Moore K, Moore MK, Keller GA, de Sauvage FJ (1997) Normal platelets and megakaryocytes are produced in vitro in the absence of thrombopoietin. Blood 90:3423–3429. 10. Debili N, Masse JM, Katz A, Guichard J, Breton-Gorius J, Vainchenker W (1993) Effects of the recombinant hematopoietic growth factors interleukin-3, interleukin-6, stem cell factor, and leukemia inhibitory factor on the megakaryocytic differentiation of CD34+ cells. Blood 82:84–95. 11. Avraham H, Vannier E, Cowley S, Jiang S, Chi S, Dinorello CA, Zsebo KM, Groopman JE (1992) Effects of the stem cell factor, c-kit ligand on human megakaryocytic cells. Blood 79:365–371. 12. Yonemura Y, Kawakita M, Masuda T, Fujimato K, Kato K, Takatuski K (1992) Synergistic effects of interleukin 3 and interleukin 11 on murine megakaryopoiesis in serum-free culture. Exp Haematol 20:1011–1016. 13. Withy RM, Rafield LF, Beck AK, Hoppe H, Williams N, McPherson JM (1992) Growth factors produced by human embryonic kidney cells that influence megakaryopoiesis include erythropoietin, interleukin 6, and transforming growth factor-beta. J Cell Physiol 153:362–372. 14. Frisch J, Ganser A, Hoelzer D, Brugger W, Kanz L, Mertelsmann R, Schulz G (1992) Interleukin 3 and granulocytemacrophage colony-stimulating factor in combination: clinical implication. Med Pediatric Oncol (Suppl.) 2:34–37. 15. Song LN, Cheng T (1992) Effects of 1,25-dihydroxyvitamin D3 analogue 1,24(OH)2-22-ene-24-cyclopropyl D3 on proliferation and differentiation of a human megakaryoblastic leukemia cell line. Biochem Pharmacol 43:2292–2295. 16. Cheng T, Erickson-Miller CL, Li C, Cardier J, Wang Y, Dempsey J, Mogle M, Barbera E, Murphy MJ Jr, Dai W (1995) HIMeg-1, a cell line derived from a CML patient, is capable of monocytic and megakaryocytic differentiation. Leukemia 9:1257– 1263.

(p.b.e.) — CYTO — 10/11 — ISSUE — ( 862814 ) —MS 0248 Leukaemia cell differentiation induced by cytokines / 889 17. Erickson-Miller CL, Ji H, Murphy MJ Jr (1993) Megakaryocytopoiesis and platelet production: does stem cell factor play a role? Stem Cells 11 (Suppl. 2) 163–169. 18. Quertmeier H, Zaborski M, Graf G, Ludwig WD, Drexler HG (1996) Expression of the receptor MPL and proliferative effects of its ligand thrombopoietin on human. Leukemia 10:297–310. 19. Tsuj I, Takayama KT, Takata H, Harushima A, Nishida Y, Izumi N, fukuda S, Ohta T, Kurimoto M (1996) Interferon-g enhances megakaryocyte colony-stimulating activity in murine bone marrow cells. J Interferon Cytokine Res 16:701. 20. Tsuj I, Takayama K, Harashima A, Tahata H, Izumi N, Nishida Y, Namba M, Okuda Y, Ohta T, Kurimoto M (1996) IFN-g in combination with IL-3 accelerates platelet recovery of mice with 5-fluorouracil-mediated marrow aplasia. J Interferon Cytokine Res 16:447–451. 21. Kalechman Y, Rushkim G, Nerubay J, Albeck M, Sredni B (1995) Effect of the immuno-modulator AS101 on chemotherapyinduced multilineage myelosuppression, thrombocytopenia, and anemia in mice. Exp Hematol 23:1358–1366. 22. Sandrock B, Hudson KM, Williams DE, Lieberman MA (1996) Cytokine production by a megakaryocytic cell line. In Vitro Cell Dev Biol 32:233. 23. Monte D, Wietzerbin J, Pancre V, Merlin G, Greenberg SM, Kusnierz JP, Capron A, Auriault C (1991) Identification and characterization of a functional receptor for interferon-g on a megakaryocytic cell line. Blood 8:2062–2069. 24. Akiyama Y, Yamaguchi K, Sato T, Abe K (1992) Tumor necrosis factor-a stimulates colony formation by a megakaryoblastic leukemia cell line CMK. Jpn J Cancer Res 83:989–994. 25. Avraham H, Vannier E, Chi SY, Dinarello CA, Groopman JE. Cytokine gene expression and synthesis by human megakaryocytic cells. Int J Cell Cloning 10:70–79. 26. Tortolani J, Johnson JA, Bacon CM, McVicar DW, Shimosaka A, Linnekim D, Longo DL, O’Shea JJ (1995) Thrombopoietin induces tyrosine phosphorylation and activation of the Janus kinases, JAK2. Blood 85:3444–3451.

27. Sattler M, Durtin MA, Frank DA, Okuda K, Kaushansky K, Salgin R, Griffin JD (1995) The thrombopoietin receptor c-mpl activates JAK2 and Tyk2 tyrosine kianses. Exp Hematol 23:1040–1048. 28. Darnell JE Jr, Kerr IM, Stark GR (1994) Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science 264:1415–1421. 29. Obeid LM, Hannun YA (1995) Ceramide: a stress signal and mediator of growth suppressor and apoptosis. J Cell Biochem 58:191–198. 30. Cheng T, Wang Y, Dai W (1994) Transcription factor egr-1 is involved in phobol 12-myristate 13-acetate-induced megakaryocytic differentiation of K562 cells. J Biol Chem 269:30848–30853. 31. Lumelsky N, Forget B (1991) Negative regulation of globin gene expression during megakaryocytic differentiation of a human erythroleukemia cell line. Mol Cell Biol 11:3528–3536. 32. Dai W, Murphy MJ (1993) Down-regulation of GATA-1 expression during phorbol myristate acetate-induced megakaryocytic differentiation of human erythroleukemia cells. Blood 81:1214–1221. 33. Greenberg SM, Rosenthal DS, Greeley TA, Tantravahi R, Hardin RI (1988) Characterization of a new megakaryocytic cell line: the Dami cell. Blood 72:1968–1977. 34. Adachi M, Ryo R, Yoshida A, Sugano W, Yasunaga M, Saigo K, Yamaguchi N, Sato T, Sano K, Kaibuchi K (1992) Induction of smg p21/rap1A and p21/krev-1 gene expression during phorbol ester-induced differentiation of a human megakaryocytic leukemia cell line. Oncogene 7:323–329. 35. Spear GT, Paulnock DM, Helgeson DO, Borden EL (1998) Requirement of differentiative signals of both interferon-gamma and 1,25-dihydroxyvitamin D3 for induction and secretion of interleukin1 by HL-60 cells. Cancer Res 48:1740–1744. 36. Pinter C, Magnani I, Paini C, Clivio A (1995) Reversible macrophage differentiation induced in a new human myeloid cell line by gamma interferon. Cell Biol Int 19:9–15. 37. Kim MY, Linardic C, Obeid L, Hannun Y (1991) Identification of sphingomyelin urnover as an effector mechanism for the action of tumor necrosis factor-a and g-interferon: specific role in cell differentiation. J Biol Chem 266:484–489.