During the course of differentiation, megakaryocytes undergo endomitosis (polyploidization) ( Arriaga et al, 1987 ; Nagasawa et al, 1985 ; Straneva et al, 1986 ). However, little is known about the mechanism of megakaryocytic polyploidization. This is due mainly to the difficulty in obtaining sufficient numbers of megakaryocytes for analysis ( Rabelleno et al, 1979 ; Levine et al, 1982 ; Long et al, 1982 ). Therefore we attempted to establish an alternative model to mimic the normal process of polyploidization of megakaryocytes, with the use of established megakaryoblastic cell lines. Phorbol-12-myristate-13-acetate (TPA) was reported to induce megakaryocytic differentiation in some megakaryocytic cell lines. However, its ability to induce polyploidization was not sufficient to allow analysis of the mechanism of polyploidization ( Ogura et al, 1988 ; Long et al, 1990 ; Greenberg et al, 1988 ; Murata et al, 1991 ). K-252a, an indolocarbasole derivative protein kinase inhibitor isolated from the culture broth of Nocardiosis SP ( Kase et al, 1986 , 1987) causes polyploidization without intervening mitosis in various cultured cell lines ( Usui et al, 1991 ). Longer incubation with K-252a produced highly polyploid cells with a single giant nuclei which were reminiscent of polyploid and mature normal megakaryocytes. It was also reported that K-252a enhanced the 1,25-dihydroxyvitamin D3 [1,25(OH)2D3], all-trans-β-retinoic acid (RA), or dimethyl sulphoxide (DMSO) induced differentiation of the myeloblastic leukaemia cell line HL-60 ( Taoka et al, 1990 ). These observations prompted us to investigate the effect of K-252a on megakaryocytic cells. We found that K-252a induced prominent differentiation of a human megakaryocytic cell line, Meg-J. Using this novel model, we analysed the changes of cdc2 kinase activity in the process of megakaryocytic polyploidization.
Megakaryocytes are unique haemopoietic cells which undergo DNA replication, giving rise to polyploid cells. However, little is known about the mechanism of megakaryocytic polyploidization. To address this issue, we used the human megakaryocytic cell line Meg-J. In the presence of K-252a (an indolocarbasole derivative), Meg-J cells stopped proliferation and exhibited additional megakaryocytic features, including morphological changes, polyploidization, and increases in the levels of surface expression of platelet glycoprotein (GP) IIb/IIIa and GPIb. Thrombopoietin (TPO) promoted the K-252a-induced polyploidization and megakaryocytic differentiation. In the process of K-252a-induced polyploidization, levels of expression of both cdc2 and cyclin B1 were elevated transiently and subsequently decreased. This suggested that the polyploidization process in Meg-J cells was at least in part associated with a transient elevation and subsequent decrease in the expression of cdc2/cyclin B1 complex, a critical kinase involved in G2/M cell cycle transition.
MATERIALS AND METHODS
K-252a was purchased from Funakoshi Co. (Tokyo, Japan), dissolved in dimethyl sulphoxide (DMSO) and stored in the dark at −4°C. Purified recombinant human TPO was a gift from Kirin Brewery Co., Japan. Propidium iodide (PI), RNase and hydroxyurea were purchased from Sigma (St Louis, Mo., U.S.A.). Fluorescein isothiocyanate (FITC) conjugated monoclonal antibodies (MoAbs) against platelet glycoprotein GPIIb/IIIa and GPIb were purchased from Immunotech (Marseille, France). Leupeptin, protease inhibitor, trypsin inhibitor, aprotinin, dl-Dithiothreitol (DTT), Phenylmethylsulphonyl Fluoride (PMSF) and Tween 20 were from Sigma. Protein G-sepharose beads were purchased from Pharmacia (Uppsala, Sweden), and [γ-32P]ATP (111 TBq/mmol) was from Amersham (Little Chalfont, Bucks., U.K.). Antibodies used for Western blotting were anti-cyclin B1 antibody (14541) from Pharmingen (San Diego, Calif., U.S.A.), anti-cdc2 antibody (no. 9110), and anti-phospho-specific cdc2 (Tyr-15) antibody (no. 9110) from New England Biolabs (Beverly, Mass., U.S.A.). Anti-cyclin B1 antibody (14551) from Pharmingen was used for in vitro kinase assay. Peroxidase-conjugated secondary sheep anti-mouse and sheep anti-rabbit antibodies were obtained from Amersham.
The megakaryoblastic leukaemia cell line Meg-J was established in our laboratory from a patient with blastic crisis of chronic myelogenous leukaemia (CML). Meg-J cells were positive for platelet GPIIb/IIIa, GPIb and platelet peroxidase, and were tetraploid. Cells were maintained in Falcon 3031 plastic flasks (Becton Dickinson, Oxford, Calif., U.S.A.) in RPMI 1640 medium (GIBCO, Grand Island, N.Y., U.S.A.) supplemented with 10% heat-inactivated fetal calf serum (FCS) at 37°C in a humidified atmosphere of 5% CO2.
Meg-J cells (1 × 105 cells) were plated in 1 ml of RPMI 1640 medium containing 0.3% agar and 10% FCS. K-252a (0.3 μM) and TPO (10 ng/ml) were added alone or in combination. The plates were incubated at 37°C in a humidified atmosphere of 5% CO2. After 5 d in culture, cell morphology was investigated with an Olympus inverted microscope. Plates were dried using filter papers stacked on the agar, stained with May-Grünwald-Giemsa (MGG), and morphological changes were further studied under an Olympus microscope. To synchronize Meg-J cells to G1 (strictly early S) phase of the cell cycle, exponentially growing cells (3 × 105 cells/ml) were cultured in 10 ml of RPMI 1640 medium containing 10% FCS and 20 m M hydroxyurea (Sigma) for 15 h. After washing by centrifugation, cells (1.5 × 105 cells/ml) were incubated containing 10 ml of culture medium supplemented with 10% FCS containing K-252a (0.3 μM) alone or in combination with TPO (10 ng/ml). After 5 d in culture, morphological changes were investigated using an Olympus inverted microscope.
The cell viability was estimated by the erythrosine exclusion test. Non-synchronized cells (1.5 × 105 cells/ml) were incubated in 10 ml of culture medium containing 10% FCS, K-252a (0.3 μM) and TPO (10 ng/ml) and the cell viability was estimated on day 5.
Meg-J cells (5 × 103 cells/well) were cultured for 48 h in 150 μl of RPMI 1640 medium supplemented with 2% FCS in 96-well plates (Flow Laboratories, Irvine, U.K.) in the presence of TPO at 37°C in a fully humidified incubator at 5% CO2. Then, samples were labelled with 37 kBq of [3H]thymidine (Amersham) during the last 4 h of incubation and harvested onto glass-fibre filters. The radioactivity was measured using a MATRIXTM96, Direct Beta Counter (Hewlett Packard, Palo Alto, Calif., U.S.A.).
Flow cytometric analysis of ploidy and surface expression of GPIIb/IIIa and GPIb
Meg-J cells (1.5 × 105 cells/ml) synchronized to G1 (strictly early S) phase of cell cycle were cultured with K-252a (0.3 μM) alone or in combination with TPO (10 ng/ml) containing 10 ml of culture medium and 10% FCS for 5 d. To analyse the ploidy distribution of cells, cells were harvested and stained with 0.1% sodium citrate solution containing 25 mg/ml propidium iodide (Sigma), 0.1% Triton X (Sigma) and 1 mg/ml ribonuclease A (Sigma) at room temperature for 60 min. Then, DNA content was measured by flow cytometry (Coulter, Miami, Fla., U.S.A., EPICS Profile II). To investigate the surface expression of GPIIb/IIIa and GPIb, cells were washed once in phosphate-buffered saline (PBS) containing 1% bovine serum albumin (Sigma), incubated for 30 min at 4°C with fluorescein isothiocyanate (FITC)-conjugated MoAb against GPIIb/IIIa and GPIb or control mouse IgG1 antibody (Coulter), and analysed by flow cytometry.
After releasing the 15 h hydroxyurea (20 m M) block, K-252a (0.5 μM)-treated or untreated Meg-J cells were pelleted and total cellular proteins were extracted in lysis buffer: 50 m M HEPES/NaOH pH 7.2; 5 m M EDTA; 250 m M NaCl; 0.5% NP-40; 0.3 mg/ml benzamidine; 10.8 mg/ml β-glycerophosphate; 0.5 m M DTT; 50 μg/ml PMSF; 10 μg/ml leupeptin; 10 μg/ml aprotinin; 10 μg/ml trypsin inhibitor; 50 m M sodium fluoride; 200 μM sodium orthovanadate. Protein extracts (200 μg) were subject to SDS-PAGE (12% polyacrylamide gels for detection of cdc2, phosphorylation status of cdc2 at tyrosine 15, and cyclin B1). The gels were blotted to Immobilon membranes (Millipore, Bedford, Mass., U.S.A.) for 1 h at 14 V on a semi-dry transfer apparatus (Owl Scientific Inc., Woburn, Mass., U.S.A.). Filters were blocked for 1 h at room temperature in Tris-buffered saline containing 0.1% Tween 20 (TBST), and 5% non-fat dried milk, and incubated overnight at 4°C with the appropriate primary antibody diluted in TBST. Working dilutions were: 1/1000 for anti-cdc2, anti-phosphorylation status of cdc2 at tyrosine 15, and anti-cyclin B1. The filters were once washed for 15 min, then 3 × 5 min in TBST and then incubated for 1 h at room temperature with sheep anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase (Amersham). After extensive washing in TBST, signals were detected using the enhanced chemiluminescence detection system (Du Pont, NEN, Boston, Mass., U.S.A.).
Immunoprecipitation and in vitro kinase assays
After releasing the hydroxyurea block, K-252a (0.5 μM) treated Meg-J cells (1 × 107) were washed twice with ice-cold PBS and lysed for 30 min on ice in lysis buffer. Insoluble materials were removed by centrifugation at 12 000 g for 10 min at 4°C. Resultant supernatants were incubated for 1 h at 4°C in the presence of appropriate anti-cyclin B1 antibody and 40 μl of 50% v/v protein G-sepharose beads (Pharmacia) in a rotating wheel. Immunoprecipitates were washed three times with 0.5 ml of ice-cold lysis buffer and once with 0.5 ml of kinase buffer (50 m M Tris-HCl, pH 7.4, 10 m M magnesium chloride, 1 m M dithiothreitol).
For kinase assays, pellets were resuspended in 25 μl of kinase buffer with 50 μg/ml histone H1 (Boehringer Mannheim). Reactions were initiated by the addition of 50 μM ATP and 370 kBq of γ-[32P]ATP (11.1 Tbq/mmol), incubated at 30°C for 30 min, stopped by the addition of 7 μl of 6 × Laemmli sample buffer, and boiled for 4 min. Aliquots of 40 μl of each reaction were analysed by SDS-PAGE on a 12% polyacrylamide gel, and bands were detected by autoradiography.
Effects of K-252a and TPO on cell proliferation and morphology
Meg-J cells treated with 0.3 μM K-252a in liquid culture ceased proliferation (Fig 1A) and showed an increase in the size of individual cells (Fig 2C). In contrast, TPO induced a concentration-dependent stimulatory effect on the proliferation of Meg-J cells with a maximum effect at 10 ng/ml TPO (Fig 1B), without altering either size or morphology (Fig 2B). When TPO was added to Meg-J cells in combination with K-252a, the anti-proliferative effect of K-252a was partially inhibited (Fig 1A). This suggested that TPO rescued K-252a-treated Meg-J cells. Furthermore, cells treated with K-252a plus TPO became larger than those cultured with K-252a alone, and some of the large cells exhibited pseudopodia-like structures (Fig 2D). The morphological changes became more prominent when Meg-J cells were plated in semisolid agar culture medium. Pseudopodia-like structures were observed in some cells treated with K-252a and a few ramifications were detected (Fig 2G). Meg-J cells treated with K-252a plus TPO exhibited marked ramification of pseudopodia and proplatelet-like formation ( Figs 2H and 2I).
Effects of K-252a with or without TPO on the ploidy and surface expression of platelet glycoproteins (GPIIb/IIIa and GPIb) of Meg-J cells
Flow cytometric analysis demonstrated that Meg-J cells were tetraploid (Fig 3A). TPO alone had no effect on the ploidy distribution of Meg-J cells at any concentration examined (Fig 3B). K-252a induced polyploidization in Meg-J cells at the concentration of 0.25–0.6 μM (data not shown). In Meg-J cells treated with K-252a (0.3 μM) for 5 d, four distinct peaks corresponding to 4N, 8N, 16N and 32N were observed in the ploidy distribution, and 16N was the largest peak (Fig 3C). In cells treated with K-252a (0.3 μM) plus TPO (10 ng/ml), five sharp peaks corresponding to 4N, 8N, 16N, 32N and 64N were observed in the ploidy distribution and 32N was the largest peak (Fig 3D).
Meg-J cells were positive for GPIIb/IIIa and GPIb. Following treatment of cells with K-252a for 5 d, the levels of expression of GPIIb/IIIa and GPIb were markedly increased (Fig 4). Further increases in the levels of GPIIb/IIIa and GPIb were observed in cells treated with TPO in addition to K-252a, and these increases by TPO were observed when cells with 4N, 8N or 16N ploidy were analysed separately (data not shown).
Expression of cyclin B1 and cdc2, and cdc2 kinase activity in K-252a-treated Meg-J cells
It has been reported that polyploidy is induced by a decrease in cdc2 kinase activity ( Usui et al, 1991 ). We therefore examined the levels of cyclin B1 and cdc2 in Meg-J cells. Fig 5 shows the results of Western blot analysis of cell extracts from Meg-J cells with or without K-252a treatment after release from the hydroxyurea block. Levels of cyclin B1 were transiently increased 12 h after releasing the hydroxyurea block in both cell groups, but after this point cyclin B1 expression decreased gradually in K-252a-treated Meg-J cells as compared to untreated Meg-J cells (Fig 5B(a)). On anti-cdc2 immunoblots (Fig 5B(b)), three forms of cdc2 were detected; the slowest migrating form was most phosphorylated cdc2 (phosphorylated on Thr-14 and Tyr-15), the fastest migrating form was most dephosphorylated cdc2 (dephosphorylated on Thr-14 and Tyr-15), and the intermediate form was suggested to be phosphorylated on either Thr-14 or Tyr-15, as previously reported ( Borgne & Meijer, 1996). This intermediate form reacted with anti-phosphorylated Tyr-15 of cdc2 (Fig 5B(c)). Therefore K-252a treatment appeared to generate the Thr-14 dephosphorylated and Tyr-15 phosphorylated forms of cdc2 that could be detected with anti-cdc2 antibody as an intermediate form. Total cdc2 levels were also transiently increased 12 h after releasing hydroxyurea block in both cell groups (Fig 5B(b)). Furthermore, at 12 h after the release the intermediate form of cdc2 was predominant and the levels of the slowest migrating form of cdc2 was extremely low in K-252a-treated Meg-J cells as compared to untreated cells, which expressed similar levels of all three forms of cdc2. At 18 h or later, cdc2 levels were markedly decreased in K-252a-treated Meg-J cells relative to those of untreated cells. Especially, Tyr-15 phosphorylated forms of cdc2 (slowest and intermediate forms of cdc2) were nearly undetectable at 24 h after release of the hydroxyurea block (Fig 5B(b)). Consistent with this observation, in K-252a-treated Meg-J cells, cdc2 kinase activity was detected at 12 h after releasing the hydroxyurea block when cdc2 and cyclin B1 levels were transiently elevated, but the activity was decreased significantly thereafter (Fig 5C). Ploidy distribution of K-252a-treated cells revealed that cells with 8N ploidy at 12–18 h after releasing the hydroxyurea block did not undergo cell division in contrast to untreated cells despite clearly detecting the cdc2 kinase activity at 12 h.
We examined the effects of K-252a on polyploidization and differentiation of a human megakaryocytic cell line, Meg-J. Although K-252a is known to induce polyploidization in some non-haematological cell lines ( Usui et al, 1991 ), this is the first report concerning its effect on megakaryocytic cells. K-252a induced Meg-J cells to form polyploid cells, which were accompanied by the morphological differentiation and increases in the levels of GPIIb/IIIa and GPIb expression. It has been suggested that DNA synthesis is a prerequisite not only for polyploidization but also for the expression of megakaryocytic markers such as GPs in TPA-treated megakaryocytic differentiation ( Murata et al, 1991 , 1993; Yoshino et al, 1996 ). In our system, such DNA synthesis without cell division may also have induced increased levels of megakaryocytic markers.
K-252a also induced polyploidization and increased the levels of surface expression of GPIIb/IIIa and GPIb on CMK, MEG-01, UT-7, HEL and K562, all of which are known to have megakaryocytic features or the capacity to differentiate into cells of the megakaryocytic lineage (data not shown). However, K-252a did not induce obvious polyploidization of non-megakaryocytic cell lines such as MOLT-4 (a human lymphoid leukaemia cell line) and HL-60 (a human myeloid leukaemia cell line) (data not shown). These results suggest that, even in haemopoietic cells, polyploidization-inducing activity of K-252a is limited to cells with potential megakaryocytic features. This K-252a-induced polyploidization and differentiation of Meg-J cells was enhanced by simultaneous addition of TPO. This might have been due to the increasing survival of Meg-J cells by TPO.
Given the potential importance of G2/M regulation in polyploidization, we performed detailed analyses of expression and function of cyclin B1 and cdc2 which constitute critical kinase involved in G2/M transition in K-252a-treated Meg-J cells. It has been demonstrated that dephosphorylation of Thr-14 and Tyr-15 of cdc2 during the activation at the prophase/metaphase transition occurs in two steps; first, Thr-14 is dephosphorylated, and then Tyr-15 is dephosphorylated. The Thr-14 dephosphorylated and Tyr-15 phosphorylated form of cdc2 still retains significant kinase activity ( Borgne & Meijer, 1996). In K-252a-treated Meg-J cells, total levels of cdc2 were transiently increased at 12 h after release from the hydroxyurea block. The intermediate migrating form (Thr-14 dephosphorylated and Tyr-15 phosphorylated form) and the fastest migrating form (Thr-14 and Tyr-15 dephosphorylated form), both of which were active as cdc2 kinase, were clearly detected. Furthermore, cdc2 kinase activity was also transiently increased at 12 h after release from hydroxyurea block. These findings indicated that K-252a did not block the dephosphorylation of Thr-14 and Tyr-15 (and also phosphorylation of Thr-161), and did not interfere with the transient increase of cdc2 kinase activity at 12 h after releasing the hydroxyurea block. On the other hand, K-252a reduced the expression of total cdc2 thereafter, especially the phosphorylated Tyr-15 form, indicating that cdc2 was not newly synthesized. In addition, expression of cyclin B1 started to reduce after 12 h. These changes in cdc2 and cyclin B1 expression resulted in decreased cdc2 kinase activity. Polyploidization began to occur more than 48 h after release from the hydroxyurea block in K-252a-treated cells ( Figs 3C and 5A), when the levels of cdc2 and cyclin B1 became markedly decreased. Hence, the elevations and subsequent decreases in cdc2 and cyclin B1 levels (as well as cdc2 kinase activity) preceded K-252a-induced polyploidization. In some megakaryocytic cell lines such as MEG 01, HEL and U937, treatment of cells with TPA induced inhibition of cdc2 kinase activity, followed by polyploidization. In this case, levels of cdc2 and cyclin B1 did not change, and lack of cdc2 kinase activity was due to the down-regulation of cdc25C phosphatase ( Garcia & Cales, 1996). It has also been reported that lack of cdc13 (cyclin B) induces entry of Sch. pombe cells into an endoreplication cycle ( Hayles et al, 1994 ). Furthermore, overexpression of rum1, which appears to be a cdc2 inhibitor, causes Sch. pombe cells to enter an endoreplication cycle ( Moreno & Nurse, 1994). These findings indicate that inhibition of cdc2 kinase activity appears to be a key event for endoreplication (or polyploidization). Intriguingly, our results obtained in megakaryocytic cells indicates that transient elevation of cdc2 kinase occurred before progressive decrease of the kinase activity. Recently, it was reported that normal polyploid megakaryocytes enter mitosis and step into anaphase. In such cells, accumulation of both cyclin B1 and cdc2 was clearly detected during early mitosis ( Nagata et al, 1997 ). Based on the generally accepted idea that activation of cdc2 kinase is an essential prerequisite to initiate mitosis, these findings indicate that, in normal megakaryocytes, elevation of cdc2 kinase activity occurs in the process of polyploidization, as has been demonstrated in K-252a-treated Meg-J cells. Accordingly, our system shown here may be an useful model to further elucidate mechanisms underlying normal megakaryocytic polyploidization.
We are grateful to Kirin-Brewery Company Ltd for supplying TPO. This work was supported in part by a grant-in-aid from the Ministry of Education, Science and Culture, Japan, by a research grant from Uehara Memorial Foundation, and by a grant from the Sankyo Foundation of Life Science.