Prostaglandin D Synthase in Human Megakaryoblastic Cells*

The cytosol fraction of human platelets did not convert prostaglandin (PG) H2 to PGD2. However, a homogenate of human megakaryoblastic CMK cells (precursor cells of platelets) produced PGD2 from PGH2. The PGD synthase activity was localized in the cytosol of CMK cells, and absolutely required glutathione. The catalytic properties and Western and Northern blottings indicated that the enzyme was PGD synthase of the hematopoietic type rather than the lipocalin type. When CMK cells were differentiated to megakaryocytes with phorbol ester along with induction of cyclooxygenase-1, the PGD synthase activity increased about 2-fold for 2 days and then decreased. In another human megakaryoblastic cell line, Dami, the PGD synthase increased about 10-fold by the addition of phorbol ester. Thus, the PGD synthase, which was undetectable in platelets, appeared during differentiation of megakaryoblasts to megakaryocytes.

Prostaglandin (PG) 1 D 2 is an anti-aggregatory, vasodilating and bronchoconstrictive eicosanoid (1). Furthermore, the role of PGD 2 in sleep induction has been extensively investigated (2). The PGD 2 is formed from arachidonic acid by successive enzyme reactions: the oxygenation of arachidonic acid to PGH 2 via PGG 2 by PG endoperoxide synthase and the isomerization of PGH 2 to PGD 2 by PGD synthase. There are two different types of PGD synthase: the hematopoietic type and the lipocalin type (3). The former was purified from rat spleen (4,5), identified as a -type glutathione S-transferase (6,7), and immunohistochemically detected in mast cells of various organs (8) and antigen-presenting cells, histiocytes, dendritic cells, Kupffer cells, and Langerhans cells (9). The latter was purified from rat brain (10), and later found to be identical with ␤-trace, a major constituent in cerebrospinal fluid (11,12).
A human megakaryoblastic cell line CMK was established from the peripheral blood of an acute megakaryoblastic leukemia patient with Down's syndrome (13,14). The cells differentiate to mature megakaryocyte-like cells by treatment with phorbol ester (TPA) or dimethyl sulfoxide (13,14). Recently we showed that cyclooxygenase-1, but not cyclooxygenase-2, was induced in CMK cells during the differentiation (15). The cyclooxygenase-1 is well known as a constitutive enzyme present in platelets (16,17), and is responsible for the production of proaggregatory thromboxane A 2 essentially as a sole cyclooxygenase metabolite. However, we found that the TPA-treated CMK cells produced PGD 2 as a major product in addition to thromboxane A 2 (15). Since PGD 2 was reported not to be produced in platelets (18), we were interested in the megakaryoblastic production of PGD 2 and attempted to characterize the responsible enzyme.

EXPERIMENTAL PROCEDURES
Materials-[1-14 C]Arachidonic acid (2.1 GBq/mmol) was purchased from Amersham International (Amersham, UK); arachidonic acid from Nu-Chek-Prep (Elysian, MN); lyophilized powder of sheep seminal vesicle microsomes from Novamed (Jerusalem, Israel); glutathione (reduced form), dithiothreitol, cysteine, and 1-chloro-2,4-dinitrobenzene from Wako Pure Chemical Industries (Osaka, Japan); 2-mercaptoethanol from Nacalai Tesque (Kyoto, Japan); precoated Silica Gel 60 F254 glass plates for TLC (20 cm ϫ 20 cm, 0.25-mm thickness) from Merck (Darmstadt, Germany); fetal calf serum from Nippon Bio-Supplies Center (Tokyo, Japan); RPMI 1640 medium and Iscove's modified Dulbecco's medium from Life Technologies, Inc. (Gaithersburg, MD); TPA, esculetin, streptomycin, and penicillin G from Sigma; and glutathione-Sepharose 4B from Pharmacia Biotech Inc. (Uppsala, Sweden). Other reagents used for Northern (15) and Western (19) blots were purchased as described previously. CMK cells were kindly provided by Dr. Eiji Shimizu of the Third Department of Internal Medicine, Tokushima University School of Medicine, with the consent of Dr. Takeyuki Sato of Chiba University, and Dami cells by Dr. Hiroshi Miyazaki of Kirin Brewery Company (Takasaki, Japan). Standard PGs and thromboxane B 2 were gifts from Ono Pharmaceutical Co. (Osaka, Japan). The antiserum was raised in a rabbit against the recombinant human hematopoietic PGD synthase which was expressed in Escherichia coli and purified by glutathione-affinity chromatography (7). The rabbit antiserum against human lipocalin type of PGD synthase was prepared as described previously (20). cDNA for human hematopoietic PGD synthase was isolated by the 3Ј-rapid amplification of cDNA end method (7).
Cell Culture-CMK cells (1 ϫ 10 7 cells/dish) and Dami cells (0.5 ϫ 10 7 cells/dish) were plated in a 150-mm plastic dish with 30 ml of RPMI 1640 medium containing 10% fetal calf serum, 23.8 mM sodium bicarbonate, and 34.3 M streptomycin sulfate or with 30 ml of Iscove's modified Dulbecco's medium containing 10% horse serum, 35.7 mM sodium bicarbonate, 69 M streptomycin sulfate, and 0.28 mM penicillin G, respectively. The Petri dishes were placed in a humidified 7% CO 2 , 93% air incubator at 37°C. The cells were subcultured every 3 days. For cell differentiation, TPA was dissolved in 30 l of ethanol, and added to the cells in 30 ml of the medium. After incubation for various periods, the cells were scraped from the dishes, washed with phosphate-buffered * This work was supported in part by grants-in-aid for scientific research from the Ministry of Education, Science, Sports and Culture of Japan, the Japanese Foundation of Metabolism and Disease, Ono Medical Research Foundation, the Japan Foundation for Applied Enzymology, Ono Pharmaceutical Co., Kissei Pharmaceutical Co., Sankyo Co., Japan Tobacco Co., and Takeda Pharmaceutical Industry (to S. Y. and N. U.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ʈ Supported by the grants-in-aid for Scientific Research Program of the Ministry of Education, Science, Sports and Culture of Japan (07558108 and 07457033) and grants from Sankyo Foundation of Life Science, Japan Foundation for Applied Enzymology, and the Cell Science Research Foundation. 1 The abbreviations used are: PG, prostaglandin; TPA, 12-O-tetradecanoylphorbol 13-acetate. saline at pH 7.4, and resuspended in 0.5 ml of phosphate-buffered saline per 1 ϫ 10 7 cells. Platelets were prepared from whole blood as described previously (21).
Enzyme Preparation-The cells were sonicated three times for 3 s at 30 kHz with an interval of 1 min after each sonication. The homogenate was centrifuged at 187,000 ϫ g at 4°C for 20 min, and the resultant supernatant was used as the cytosol. Protein concentration was determined by the method of Bradford (22) using bovine serum albumin as standard. The cytosol (2 ml containing 10 mg of protein) of CMK cells was loaded onto a glutathione-Sepharose 4B column (a bed volume of 10 ml) equilibrated with phosphate-buffered saline at pH 7.4 including 1 mM dithiothreitol. Proteins were eluted from the column with the same buffer, and 1.2-ml fractions were collected. The column was regenerated by washing with 50 mM Tris-HCl (pH 9) including 5 mM glutathione. The purified enzyme was stabilized by adding ␥-globulin at a final concentration of 1 mg/ml.
Western Blot-The cytosol (15 g of protein) was subjected to 12% polyacrylamide gel electrophoresis in the presence of 0.1% SDS. Immunostaining was performed as described for cyclooxygenase (15) by the use of rabbit antiserum against recombinant human hematopoietic PGD synthase (1:2000 dilution). Immunostained bands were visualized by the enhanced chemiluminescence method.
Northern Blot-Total RNA was extracted with an ISOGEN solution (Nippon Gene, Tokyo, Japan) (23). Electrophoresis and transfer of RNA onto the membrane were carried out as described for cyclooxygenase (15). The blots were hybridized with 32 P-labeled cDNA probes for human hematopoietic PGD synthase or glyceraldehyde-3-phosphate dehydrogenase in Rapid-hyb buffer (Amersham International). Distribution of the radioactivity on the membrane was detected by a BAS2000 imaging analyzer.

RESULTS
When the homogenate of human megakaryoblastic CMK cells was incubated with [ 14 C]PGH 2 , the major product was PGD 2 (lane 2 of Fig. 1A). Under this assay condition only a small amount of unstable PGH 2 was spontaneously converted to PGD 2 and PGE 2 in a protein-free buffer used as a control (lane 1). The activity to generate PGD 2 in the homogenate was mostly recovered in the cytosol fraction (lane 4), but was hardly detectable in the particulate fraction (lane 3). The enzyme activity required glutathione and the PGD 2 production was very low in the absence of glutathione (lanes 5 and 6 of Fig. 1B). Furthermore, when pretreated at 90°C for 5 min, the cytosol totally lost the activity (lane 7). These results showed the presence of a cytosolic glutathione-dependent PGD synthase in CMK cells.
It should be noted that the cytosol fraction of human platelets did not significantly convert PGH 2 to PGD 2 either in the presence (Fig. 1, lane 8) or absence (lane 9) of glutathione. In addition, the cytosol fractions of porcine and rat platelets did not show any significant activity to produce PGD 2 (data not shown).
The PGD synthase activity was dependent on the concentration of PGH 2 ( Fig. 2A), and a high K m value for PGH 2 (about 200 M) was obtained by Lineweaver-Burk plots. The specific enzyme activity was about 150 nmol/min/mg with a subsaturating concentration (40 M) of PGH 2 . The enzyme activity was also dependent on the concentration of glutathione (Fig. 2B), and the K m for glutathione was about 1 mM. However, the enzyme was inactive when glutathione was replaced by other sulfhydryl compounds such as cysteine, dithiothreitol, and 2mercaptoethanol at the same concentration. These sulfhydryl compounds activated the lipocalin type PGD synthase but not the hematopoietic type enzyme which required selectively glutathione (3). Furthermore, 1-chloro-2,4-dinitrobenzene at 1 mM, which is known to inhibit the lipocalin-type enzyme but not the hematopoietic-type enzyme (5), hardly inhibited the PGD synthase activity of CMK cells. These properties suggest that the PGD synthase of the CMK cells is of the hematopoietic type rather than the lipocalin type.
In view of the glutathione requirement, the cytosol of CMK cells was applied to glutathione-affinity chromatography (Fig.  3A). A bulk of proteins passed through the column, and the PGD synthase activity was eluted with a significant delay in the absence of glutathione, suggesting a weak interaction between the enzyme and glutathione. The enzyme was highly purified by the glutathione-affinity chromatography, and the specific activity of three active fractions pooled was about 300 mol/min/mg protein. Active fractions showed a 26-kDa protein band stained with silver nitrate (Fig. 3B). This protein band was immunostained with rabbit antiserum against human hematopoietic PGD synthase (Fig. 3C), but not with antiserum against human lipocalin type PGD synthase.
The PGD synthase activity was followed during differentiation of CMK cells. When the cells were cultured in the presence of 0.1 M TPA for different days, the cyclooxygenase activity in the cell homogenate was almost undetectable at first, but increased markedly in a time-dependent manner (Fig. 4A) as we reported previously (15). In contrast, the PGD synthase activity was considerably high at day 0, and increased only about 2-fold during the incubation with TPA. The maximum PGD synthase activity was observed at day 2, and thereafter the enzyme activity decreased. As examined by Western blotting with anti-human hematopoietic PGD synthase antibody, the intensity of the immunopositive band changed in agreement with the enzyme activity (Fig. 4B). When the amount of PGD synthase mRNA was followed by Northern blotting, two radioactive bands were detected at positions around 1.8 and 1.0 kilobases through the 5 days, and the most intense bands were seen at day 1 (Fig. 4C).
Another megakaryoblastic cell line, Dami (24), was also treated with TPA, and the cyclooxygenase and hematopoietic PGD synthase activities were followed. The cyclooxygenase activity increased in a similar time course as in the case of CMK cells but to a lower specific enzyme activity. In contrast to CMK cells, Dami cells showed a low PGD synthase activity at time 0, and the addition of TPA increased the enzyme activity by 10-fold at day 3 (Fig. 5A). Thereafter the enzyme activity decreased. Western blot analysis also indicated a marked in-crease in PGD synthase protein at days 2 and 3 (Fig. 5B). As examined by Northern blotting, the content of the PGD synthase mRNA also increased and then decreased with the highest content at days 1 and 2.  (27). B, the cytosol of Dami cells (20 g of protein) was applied to Western blot analysis using anti-human hematopoietic PGD synthase antibody. C, total RNA (10 g) was isolated from Dami cells, and applied to Northern blot analyses using cDNAs of human hematopoietic PGD synthase (hPGDS) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). 18 S indicates the position of 18 S rRNA. of megakaryoblasts or megakaryocytes of bone marrow, and are useful tools for biochemical studies on megakaryoblast differentiation (13,24).
In the present study we revealed potent PGD 2 production by the cytosol fraction of CMK cells with a specific activity as high as about 150 nmol/min/mg protein. This value is comparable to the PGD synthase activity in the cytosol of rat peritoneal mast cells (105 nmol/min/mg protein under the same assay conditions) which was reported to be the highest among various rat tissues (8). Serum albumin was previously shown to convert PGH 2 to PGD 2 in a glutathione-independent manner (18), but the PGD synthesis in CMK cells was absolutely dependent on glutathione. Therefore, it is unlikely that the PGD 2 formation which we observed is attributed to albumin derived from fetal calf serum in the cell culture medium. Several lines of evidence including catalytic properties and Western and Northern blot analyses demonstrated that the enzyme activity was due to the hematopoietic type of PGD synthase rather than the lipocalin type. Dami cells also contained the hematopoietic type of the enzyme. The hematopoietic PGD synthase has so far been found in mast cells of various tissues (8) and antigen-presenting cells (9). To the best of our knowledge, this is the first report for the presence of PGD synthase in megakaryoblastic cell lines, precursor cells of platelets which have no PGD synthase.
The cyclooxygenase-1 activity was almost undetectable in untreated CMK cells, but increased for 5 days up to 10 nmol/2 min/mg protein by the addition of TPA. In contrast, the PGD synthase activity was detected in the untreated cells with a specific activity of about 40 nmol/min/mg protein of the homogenate, and increased only by 2-fold in the presence of TPA. These observations indicate that the cyclooxygenase reaction rather than the PGD synthase reaction is a rate-limiting step in the biosynthetic pathway of PGD 2 starting with arachidonic acid released from phospholipid. Interestingly, the original level of PGD synthase activity in Dami cells was low, but the enzyme activity was increased markedly by the addition of TPA. The induced cyclooxygenase was also lower than the increased PGD synthase in Dami cells.
Previously it was a debatable subject whether or not platelets generated PGD 2 . Later, it was clearly demonstrated that the capacity to generate PGD 2 in platelet-rich plasma was attributable to serum albumin and that neither washed platelets nor platelet lysate formed PGD 2 enzymatically (18). Indeed we could not detect the PGD synthase activity in the cytosol of peripheral platelets from different animal species. Unlike cyclooxygenase-1 and thromboxane A synthase, the induced PGD synthase is not transferred to platelets which are derived from megakaryocytes. Since CMK and Dami cells are leukemia cells, a high expression of PGD synthase may not be a physiological event. Therefore, it is important to examine its expression level in native megakaryoblasts and megakaryocytes, especially by immunohistochemical study, which may lead to a finding of unknown physiological function of PGD 2 . Furthermore, PGD 2 may be further transformed to PGJ 2 and its derivatives which are known to have anti-proliferative activity and may regulate cell proliferation (25). Recently, 15-deoxy-⌬ 12,14 -PGJ 2 , a further metabolite of PGJ 2 , was reported to be a ligand for peroxisome proliferator-activated receptor ␥ which determined differentiation of fibroblasts to adipocytes (26).