THROMBOSIS RESEARCH 61; 301-310,199l 0049-3848/91 $3.00 + .OO Printed in the USA. Copyright (c) 1991 Pergamon Press pk. All rights reserved.
INCREASED EXPRESSION OF THROMBOMODULIN ON THE CULTURED HUMAN UMBILICAL VEIN ENDOTHELIAL CELLS AND MOUSE HEMANGIOMA CELLS BY CYCLIC AMP
Ikuro Maruyama*, Yasuko Soejima*, Mitsuhiro Osame*, 1to**, Kohei Ogawa***, Shuji Yamamoto***, William A. and Hidehiko Saito**
Takahiko Dittman+,
*The Third Department of Internal Medicine, Kagoshima University School of Medicine, Kagoshima 890, Japan; **The First Department Nagoya University School of Medicine, of Internal Medicine, Laboratory Asahi Chemical Nagoya 466, Japan; ***Bio-Science +Division of Shizuoka 416, Japan; and Industry, Fuji, Durham VAMC and Duke Unviersity School of Hematology/Oncology, Medicine, Durham NC, USA
(Received 19.9.1990; accepted in revised form 21 .ll .1990 by Editor H. Yamazaki)
ABSTRACT We previously reported that the expression of thrombomodulin on the MEG-01, a cell line from human megakaryoblastic leukemia, was increased by agents that increase intracellular CAMP. In this paper we examine the effect of these agents on cultured human umbilical vein endothelial cells (HUVEC) and mouse hemangioma Incubation of the cells with 3 mM dibutyryl CAMP cells. (dbcAMP) increased functionally active thrombomodulin by about 2-fold on HUVEC and 4-fold on hemangioma cells. This effect was observed from 1 hour after the incubation and continued up to 24 hours. Dot hybridization of mRNA demonstrated a dose dependent increase in thrombomodulin mRNA in response to dbcAMP. Treatment of HUVEC with 20 UM forskolin or 100 uM isobutylmethylxanthine (IBMX) also increased cell-surface thrombomodulin on HUVEC. These agents prevented the interleukin I(IL-I) or tumor necrosis factor (TNF)-induced decrease in thrombomodulin on HUVEC. These data suggest that the expression of thrombomodulin on HUVEC and mouse hemangioma cells may be regulated by intracellular CAMP level. Key Words:
Thrombomodulin, cell, Cytokine
Protein C, Cyclic AMP, Endothelial
301
302
INCREASED EXPRESSION...
Vol. 61, No. 3
INTRODUCTION Thrombomodulin on the cultured human umbilical vein endothelial cells (HUVEC) is down-regulated by tumor necrosis factor (TNF) (I), interleukin 1 (IL-l)(2) and endotoxin(3); however upregulation of thrombomodulin on HWEC has never been described. Previously we reported that MEG-01, a cell line established from human megakaryoblastic leukemia, expressed thrombomodulin(4), and this expression was enhanced by incubation of the cells with dibutyryl CAMP (dbcAMP)(5). enhanced expression of The thrombomodulin on the MEG-01 was accompanied by the increase in mRNA for thrombomodulin, suggesting that CAMP regulates the transcription of thrombomodulin gene in these cells. Here we examine the effect of dbcAMP, and other agents that increase intracellular CAMP, on the expression of thrombomodulin on the HUVEC and mouse hemangioma cells. We also investigate the effect of dbcAMP on the down regulation of thrombomodulin induced by IL-I or TNF. MATERIALS AND METHODS Materials: Materials were obtained from the following source: Dibutyryl CAMP and isobutylmethylxanthine (IBMX) from Sigma co. (St. Louis, MO).; IL-I and TNF from Boehringer Mannheim Yamanouchi (Tokyo, Japan).; Medium 199 (M-199), fetal calf serum and endothelial cell growth supplement (ECGS) from Flow Laboratories (Mclean, VA).; Penicillin, streptom%3&;]_asnoddit; glutamine from Gibco Laboratories; carrier-free [ iodide from Amersham Corp.: tissue culture flasks and centrifuge tubes from Corning Glass Works (Science Products Division, Corning, NY), and Falcon Labware (Division of Be&on, Dickinson and Co., Oxnard, CA): cluster 24, 48 dish wells from Coster (Cambridge, MA). All other chemicals were reagent grade products of Sigma (St. Louis, MO) or Protein C(6), and antithrombin III(G) were from human origin and isolated as indicated. The proteins were sodium dodecyl sulfate homogeneous as judged by The monoclonal anti-human polyacrylamide gel electrophoresis. thrombomodulin IgG (designated as l&f-TM) was prepared as described previously and lab~$~~_;~;",,', I] iodine by a modified chloramine-T procedure(7). 7,000 to 150,000 cpm/ng, G-25 column. was separated from free iodide on A Sephadex Protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad laboratories, Richmond, CA) with bovine serum albumin as standard. HUVEC were prepared by the method of Jaffe et al.(s) and Cells: cultured in M-199 supplemented with 10% fetal calf serum and 30 Confluent mg/L of ECGS in a 5% Co2 atmosphere at 37'C. monolayers of primary HUVEC were subcultured for experiments. Mouse hemangioma cells were provided by Dr. J.D. Hoak and Before the incubation with cultured as previously described(6). various agents, monolayers were washed three.times with M-199 and resuspended in the complete serum-free medium, COSMEDIUM-001 (Cosmo Bio, Tokyo); the cells could be maintained up to 36 hours in this medium. For isolation of RNA, the confluent monolayers of HUVEC cultured in IO-cm dishes were washed three times with M-199, scraped by rubber policeman and collected by centrifugation.
INCREASED EXPRESSION...
Vol. 61, No. 3
Assav of thrombomodulin
303
Activitv
Conditioned medium was removed from confluent monolayers of Then the cells and the cells were washed three times with M-199. monolayers in 6-mm culture wells were incubated with serum-free To examine the effect of dbcAMP, we medium containing dbcAMP. used increasing amounts of dbcAMP and followed a time course. The effect of other agents that increase intracellular CAMP was examined in the same manner. After incubation of the monolayers, the medium was removed, washed three times with M-199, and thrombomodulin activity on the cells was assayed. Thrombomodulin activity on the cells was measured as previously described with a The amount of activated protein C was slight modification(6). assayed by measuring the rate of hydrolysis of 0.2 mM D-Phepipecolyl-Arg-p-nitroanilide (S2238, Kabi Diagnostica, Sweden). Bindinq Assav of Cell-Surface
thrombomodulin
usinq F12511-IqG-TM
Thrombomodulin molecules on HUVEC were measured as previously Confluent monolayers of human umbilical vein described(7). endothelial cells in lo-mm culture wells were washed three times with 0.14 mol/L of NaCl, 0.004 mol/L of KCl, 0.001 mol/L of phosphate buffer, pH 7.4, 0.011 mol/L glucose containing 0.2% BSA Binding studies were (henceforth referred to as wash buffer). carried out in reaction mixtures containing M-199 supplemented with 5 mg/mL of BSA and 50 g/mL of normal human IgG 'n a total After incubation of cells with [125iI]-labeled volume of 0.3 mL. protein, the cells were washed rapidly five times with cold (O'C) wash buffer. The washed monolayers were solubilized in 0.5 mL of buffer containing 15% glycerol, 2% sodium dodecyl sulfate (SDS), 75 mmol/L of Tris-HCl (pH 6.9), and 2 mmol/L of EDTA. The solubilized cell solution was assayed for radioactivity in a scintillation counter (Beckman Instruments, Biogamma II Fullerton, CA). Specific binding was determined by measuring the difference in cell-bound radioactivity with and without loo-fold excess unlabeled IgG-TM. The cell-bound radioactivity with an excess amount of unlabeled IgG-TM was considered nonspecific and was subtracted from the total binding. Nonspecific binding was lo-15% of total binding. Dot Hybridization
of mRNA
Total RNA was prepared using guanidine isothiocyanate-sarcosyl solution followed by phenol extraction and ethanol precipitation as described previously(9). Dot hybridization was carried out by spotting a 10 pg and 5 pg of formamide denatured RNA onto dry nitrocellulose, which was then dried, hybridized with random labeled Bgl II - Hint II fragment (approximately 0.9 kb) of thrombomodulin cDNA(10) using the klenow fragment in a commercial kit (Amersham) and exposed to X-ray film. /?-actin (Oncor) hybridization was used as a reference for relative mRNA loaded(g).
304
INCREASED EXPRESSION...
Vol. 61, No. 3
RESULTS Treatment of HUVEC or mouse hemangioma cells with dbcAMP for 12 hours resulted in an increase in thrombomodulin activity and antigen. The increase in thrombomodulin by dbcAMP was concentration-dependent up to 5 mM of dbcAMP. Higher concentrations of dbcAMP decreased cell viability with reduced expression of thrombomodulin. The maximum effect was observed at 3 mM of dbcAMP (Fig. l-a). The increase in thrombomodulin at 3 mM of dbcAMP was approximately 200% of Control in HUVEC and 400% in mouse hemangioma cells. Thrombomodulin activity gradually increased during the incubation up to 12 hours of incubation. Longer incubation decreased cell viability with a decrease in thrombomodulin (Fig. l-b).
C
0
1
2
3
a
5mM
1
6
12
dbcAMP
Fig. l-a
Fig. l-b
Fis. l-a
Enhancement of the expression of thrombomodulin activity on HUVEC (0-e) and mouse hemangioma cells (0-0) by increasing amount of dbcAMP. Monolayers of HUVEC or mouse hemangioma cells were incubated for 12 hours with serum-free medium‘containing increasing amount of dbcAMP as indicated and the thrombomodulin activity was measured as described in the METHODS.
Fia. l-b
Time course of the d&-induced enhancement of thrombomodulin activity on HUVEC ‘(0-o) and mouse Monolayer of the cells were hemangioma cells (O-O). incubated with 3 mM of dbcAMP for various times up to activity measured. 24 hours and the thrombomodulin three different from (mean * S.D.) The results experiments are shown by % of control.
"--24h
305
INCREASED EXPRESSION...
Vol. 61, No. 3
We examined the effects of other agents that increase intracellular CAMP. Forskolin also increased expression of thrombomodulin activity in a concentration-dependent manner. An approximately 1.4-fold increase was observed with 30 ,uM of forskolin (Fig. 2). Isobutylmethylxanthine (IBMX, 100 PM), a phosphodiesterase inhibitor, also enhanced the expression of thrombomodulin by about 1.8-fold. The combination of forskolin (30 PM) and IBMX (100 PM) increased thrombomodulin more than either alone (Fig. 2). The effect of these agents was confirmed by radio-labeled monoclonal anti-thrombomodulin IgG binding assay (Fig. 2). 0
Formed PCa(pmofelmfh) 50
‘25f-lgG ~OUND(molecules~cellxl~~
control
dbcAMP forskolin IEMX Forskolm+lBMX
Fiu. 2
the various agents that increase in of intracellular CAMP on the expression of thrombomodulin Monolayers of HUVEC were cultured in 10 mm in HUVEC. wells, incubated with the agents for 12 hours and then evaluated for thrombomodulin activity by functional for thrombomodulin assay and antigen by labeled monoclonal anti-TM IgG binding assay. The results are of three different mean f S.D. experiments. The concentrations of the agents were as follows; dbcAMP: 3 I&?, forskolin: 30 PM,' IBMX: 100 UM Effect
100
0
I
‘1
2
3 mM
CAMP Fis. 3
Effect of dbcAMP on mRNA of thrombomodulin in HUVEC. The monolayers of HUVEC were incubated with increasing amounts of dbcAMP for 12 hours. At the indicated time, total cellular RNAs were extracted, and 5 or 10 fig of aliquots were subjected to dot hybridization as described in METHODS. The hybridized samples were subjected to autoradiography and the intensity of hybridization quantified by densitometry and analysis on a Molecular Dynamics (Sannywale, CA) laser densitometer. Relative mRNA levels of thrombomodulin were calculated by using P-actin as an internal control.
306
INCREASED EXPRESSION...
Vol. 61, No. 3
To determine whether the enhanced thrombomodulin expression by dbcAMP is accompanied by increased thrombomodulin mRNA, we evaluated the mRNA level for thrombomodulin by dot hybridization with labeled fragments of a thrombomodulin cDNA. Levels of thrombomodulin mRNA by dbcAMP were increased in a dose dependent manner and the mRNA increase was observed from 0.5 m+l to 3 mM of dbcAMP (Fig. 3). The increase in thrombomodulin mRNA was detectable within 2 hours of incubation with 3 mM dbcAMP, and peaked at 12 hours of the incubation.
Ficr. 4
Effect of dbcAMP on the IL-I-induced decrease in thrombomodulin on HWEC. The monolayers of HUVEC in wells (6 mm) were incubated with COSMEDIUM supplemented with 10 U/ml of IL-I for three hours (shaded), the medium removed and then incubated with (0-o) or without (0-C) 3 mM dbcAMP in fresh COSMEDIUM. At the indicated the thrombomodulin activity on each of the time, monolayers were measured. The results (% of control) are the mean f S.D. of three different experiments.
OO Fia. 5
6
12
18h
TNF-induced decrease in Effect of dbcAMP on the The monolayers of HUVEC were thrombomodulin on HUVEC. incubated with TNF (10 U/ml) in the presence (0-O) or At the indicated time absence (C-C) of 3 mM dbcAMP. the medium was removed and measured the thrombomodulin The results (% of control) are activity on the cells. mean k S.D. of three different experiments.
Vol. 61, No. 3
INCREASED EXPRESSION...
307
These results suggest that intracellular CAMP levels may regulate transcription of the thrombomodulin gene in HUVEC. We examined the effect of CAMP in the presence of IL-I (10 U/ml) Both IL-I and TNF reduced the thrombomodulin and TNF (10 U/ml). After the preincubation of HUVEC monolayers activity on HUVEC. with IL-I for 3 hours, the medium was removed and incubated the monolayers with serum free medium in the presence or absence of 3 mM of dbcAMP. On the dbcAMP, the monolayers without thrombomodulin activity fell to 50% of control by 9 hours. However, in the presence of dbcAMP, the rate of decrease was inhibited. This was associated with more rapid recover of thrombomodulin levels compared to the monolayers without dbcAMP (Fig. 5). Treatment of HUVEC monolayers with TNF also reduced thrombomodulin expression to approximately 40% by 18 hours. This decline was also partially prevented by treatment with dbcAMP (Fig. 6). All these experim.ents were carried out in complete serum-free medium to exclude any effects of serum. However, serum had no additional effects in the other experiments (data not shown).
DISCUSSION Endothelial cells lining blood vessels possess antithrombogenic functions mainly through production of prostacyclin(ll), secretion of tissue plasminogen activator(ll), cell surface heparin-Iike molecule that stimulates Certain COXyllatiOll factor inhibitors(l3), and thrombomodulin, an endothelial cell surface Thrombomodulin has very high affinity for thrombin glycoprotein. In this complex thrombin and forms a 1:l stoichiometric complex. has less effect on its procoagulant substrates such as whereas the ability to factor V and platelets, fibrinogen, Activated activate protein C is enhanced more than lOOO-fold. protein c functions as a potent natural anticoagulant by Thus, Factors Va and VIIIa. coagulation inactivating thrombomodulin converts thrombin from a procoagulant protease to Previously we measured thrombomodulin ~~t~gnetniso~~l~~~~(13,14) * I]IgG-TM binding assay and reported that HUVEC 30,000-50,000 approximately molecules of contains We have also reported that exposure of thrombomodulin/cells(7). HUVEC to thrombin decreased thrombomodulin on the cell-surface occurrence of endocytosis of suggesting the thrombinthrombomodulin complex(7). Phorbol myristate acetate (PMA), a potent activator of protein kinase C, also induces endocytosis of thrombomodulin(l6). Cytokines also affect the expression of thrombomodulin. TNF rapidly reduces thrombomodulin activity on the cell surface(l). The reduction in activity is associated with endocytosis and degradation of thrombomodulin(l6). TNF also inhibits transcription of thrombomodulin mBNA, and may result the subsequent decrease in thrombomodulin protein synthesis(l). Other cytokines, IL-I(2) and endotoxin(3) also cause decreased thrombomodulin activity on endothelial cell surfaces. Thus thrombomodulin expression appears to be regulated by a variety of mechanisms. However all of these factors decrease the expression of thrombomodulin; factors which increase the expression of thrombomodulin are not well characterized.
308
INCREASED EXPRESSION...
Vol. 61, No. 3
In this paper we have investigated the affect of the agents that increase intracellular CAMP and demonstrated that thrombomodulin on HWEC was up-regulated by CAMP. We previously described enhanced expression of thrombomodulin on MEG-01, a cell line from human megakaryoblas+ic leukemia, by agents that increase in intracellular CAMP(S). The agents include dbcAMP, forskolin, prostaglandin E and IBMX. Thrombomodulin expression was enhanced more than 26-fold by 3 mM dbcAMP during 48 hours and accompanied by the increase in thrombomodulin mRNA, suggesting that CAMP increased transcription of thrombomodulin message. In this paper we expanded our investigation to endothelial cells and mouse hemangioma cells. We demonstrated that incubation of the cultured endothelial cells prepared from human umbilical vein or mouse hemangioma cells with 1-5 mM dbcAMP increased the thrombomodulin activity. The increase in thrombomodulin activity was also confirmed by binding assay using radiolabeled monoclonal anti-TM IgG in HUVEC. The effect on thrombomodulin by dbcAMP was observed within 3 hours of the incubation and continued up 24 hours. Addition of fetal calf serum showed no additive effect (data not shown). Thrombomodulin increase by dbcAMP was up to 2fold in HUVEC, and 4-fold in the mouse hemangioma cells. This is much less than in MEG-07 cells. The difference between HUVEC or the mouse hemangioma cells and MEG-01 may be due to in part, different CAMP metabolism in MEG-01. In preliminary s(udies, the level of CAMP in the cells induced by dbcAMP, forskolin, or IBMX was higher in MEG-01 than in HUVEC (data not shown). As was seen in MEG-01, the increase in cell surface thrombomodulin in HUVEC induced by dbcAMP was associated with the increase in mRNA of thrombomodulin. This suggests that dbcAMP enhanced transcription of mRNA of thrombomodulin, and with subsequent increase of thrombomodulin on the cell surface. Up regulation of thrombomodulin expression by increased intracellular CAMP was supported by Imada et al. They have described CAMP enhancement of the expression of a surface marker glycoprotein, designated fetomodulin in mouse embryos(l8). Recently a cDNA clone for fetromodulin was found to be indentical to thrombomodulin(l9): The cytokine-induce reduction of thrombomodulin may have a role in the pathogenesis of thrombosis in a certain pathologic states. It may be that CAMP in endothelial cells alters thrombotic Hence we examined the effect of dbcAMP on IL-I and tendencies. Incubation of TNF induced decrease in thrombomodulin on HUVEC. HUVEC with IL-I for 3 hours reduced cell surface thrombomodulin This reduction of thrombomodulin activity to 50% of control. persisted for more than 18 hours. However the addition of dbcAMP reduced the decrease and resulted in more rapid recovery to near Similar effect was observed in control level within 18 hours. The decline in TNF-induced reduction of thrombomodulin. thrombomodulin expression in response to TNF or IL-I may have some role in the pathophysiology of prothrombotic state of The findings that dbcAMP prevented inflammation or malignancy. the cytokine-induced reduction of thrombomodulin to some extent suggests a potential role in therapeutic strategies for prothrombotic states.
Vol. 61, No. 3
INCREASED EXPRESSION...
309
ACKNOWLEDGEMENTS Part of this work was presented at the Second Saratoga Conference held on October 1-13, 1989, in Towada, Aomori, Japan This work was supported in and will be published as an abstract. part by grant-in-aid for Scientific Research from the Ministry of grants from the Ministry of Science and Culture, Education, Health and Welfare in Japan.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
Tumor necrosis factor R.D. CONWAY, E.M., and ROSENBERG, suppresses transcription of the thrombomodulin gene in endothelial cells. Mol Cell Biol, 8, 5588-5592, 1988 NAWROTH, P.P., HANDLEY, D.A., ESMON, C.T., and STERN, D.M. Interleukin I induces endothelial cell procoagulant while Proc Nat1 activity. cell-surface anticoagulant supressing Acad Sci USA, 83, 3460-3464, 1986 MOORE, K.L., ANDREOLI, S.P., ESMON, N.L., ESMON, C.T., and Endotoxin enhances tissue factor and suppresses BANG, N.U. thrombomodulin expression of human vascular endothelium in vitro. J Clin Invest, 79, 124-130, 1987 OGURA, M., ITO, T., MARUYAMA, I., TAKAMATSU, J., OGAWA, K., Localization and YAMAMOTO, S., NAGURA, H. and SAITO, H. synthesis of functional thrombomodulin in human megakaryocytes and human megakaryoblastic cell line(MEG-01). Thromb Haemostas, in press ITO, T., OGURA, M., MORISHITA, Y., TAKAMATSU, J., MARUYAMA, S., OGAWA, K., and SAITO, H. Enhanced I ., YAMAMOTO, expression of thrombomodulin by intracellular cyclic AMPincreasing agents in two human megakaryoblastic cell lines. Thromb Res, 58, 615-623, 1990 MARUYAMA, I., SALEM, H.H., and MAJERUS, P.W. Coagulation factor Va binds to human umbilical vein endothelial cells and accelerates protein C activation. J Clin Invest, 73, 224-230, 1984 MARUYAMA, I., and MAJERUS, P.W. The turnover of thrombinthrombomodulin complex in cultured human umbilical vein endothelial cells and A549 lung cancer cells. Endocytosis and degradation of thrombin. J Biol Chem, 260, 15432-15438, 1985
8. 9. 10.
JAFFE, E.A., NACHMAN, R.L., BECKER, C.G., and MINICK, C.R. Culture of human endothelial cells derived from umbilical vein. J Clin Invest, 52, 2745-2756, 1973 SAMBOOK, J., TRITSCH, E.F., and MANIATIS, T. Molecular cloning, a laborqtory manual. Second edition. Cold Spring Harbor laboratory, Cold,Spring Harbor, NY, 1989 SUZUKI, K., KUSUMOTO, H., DEYASHIKI, Y., NISHIOKA, J., MARUYAMA, I., ZUSHI, M., KAWAHARA, S., HONDA, G., YAMAMOTO, S and HORIGUCHI S. Structure and expression of human thiombomodulin, a thrombin receptor on endothelium acting as a cofactor for protein C activation. EMBO J, 8, 1891-1897, 1987
310
11.
12. 13.
14. 15. 16.
17.
18.
19.
20.
INCREASED
EXPRESSION...
Vol. 61, No. 3
WEKSLER, B.B., LEY, C.W., and JAFFE, E.A. Stimulation of by thrombin, endothelial cell prostacyclin production trypsin, and the ionophore A23187. J Clin Invest, 62, 923930, 1987 LOSKUTOFF, D.J., and EDINGTON, T.S. Synthesis of a fibrinolytic activator and inhibitor by endothelial cells. Proc Nat1 Acad Sci USA, 74, 3903-3907, 1977 of BAUER, K.A., and ROSENBERG, R.D. The pathophysiology the prethrombotic state in humans: Insight gained from studies using markers of hemostatic system activation. Blood, 70, 343-350, 1987 ESMON, C.T. The roles of protein C and thrombomodulin in the regulation of blood coagulation. J Biol Chem, 264, 47434746, 1989 ESMCN, C.T. The regulation of natural anticoagulant pathways. Science, 235, 1348-1352, 1987 DITTMAN, W.A., KUMADA, T., SADLER, J.E., and MAJERUS, P.W. The structure and function of mouse thrombomodulin: Phorbol myristate acetate stimulates degradation and synthesis of thrombomodulin without affecting mRNA levels in hemangioma cells. J Biol Chem, 263, 15815-15822, 1988 Tumor necrosis MOORE, K.L., ESMON, C.T., and ESMON, N.L. factor leads to the internalization and degradation of thrombomodulin from the surface of bovine aortic endothelial Blood, 73, 159-165, 1989 cells in culture. IMADA, M., IMADA, S., IWASAKI, H., KUME, A., YAMAGUCHI, H., Marker surface protein of and MOORE, E.E. Fetomodulin: fetal development which is modulatable by CAMP. Dev Biol, 122, 483-491, 1987 IMADA, S., YAMAGUCHI, H., NAGUMO, M., KATAYANAGI, S., IWASAKI, H., and IMADA, M. Identification of fetomodulin, a development, as marker protein of fetal surface Dev thrombomodulin by gene cloning and functional assay. _ 140, 113-122, 1990 Biol, MONTMINY, M.R., SEVARINO, K.A., WAGNER, J.A., HANDEL, G., Identification of a cyclic-AMP-responsive and GODMAN, R.H. element within the rat somatostatin gene. Proc Nat1 Acad Sci USA, 83, 6682-6686, 1986
INCREASED EXPRESSION OF THROMBOMODULIN ON THE CULTURED HUMAN UMBILICAL VEIN ENDOTHELIAL CELLS AND MOUSE HEMANGIOMA CELLS BY CYCLIC AMP
Ikuro Maruyama*, Yasuko Soejima*, Mitsuhiro Osame*, 1to**, Kohei Ogawa***, Shuji Yamamoto***, William A. and Hidehiko Saito**
Takahiko Dittman+,
*The Third Department of Internal Medicine, Kagoshima University School of Medicine, Kagoshima 890, Japan; **The First Department Nagoya University School of Medicine, of Internal Medicine, Laboratory Asahi Chemical Nagoya 466, Japan; ***Bio-Science +Division of Shizuoka 416, Japan; and Industry, Fuji, Durham VAMC and Duke Unviersity School of Hematology/Oncology, Medicine, Durham NC, USA
(Received 19.9.1990; accepted in revised form 21 .ll .1990 by Editor H. Yamazaki)
ABSTRACT We previously reported that the expression of thrombomodulin on the MEG-01, a cell line from human megakaryoblastic leukemia, was increased by agents that increase intracellular CAMP. In this paper we examine the effect of these agents on cultured human umbilical vein endothelial cells (HUVEC) and mouse hemangioma Incubation of the cells with 3 mM dibutyryl CAMP cells. (dbcAMP) increased functionally active thrombomodulin by about 2-fold on HUVEC and 4-fold on hemangioma cells. This effect was observed from 1 hour after the incubation and continued up to 24 hours. Dot hybridization of mRNA demonstrated a dose dependent increase in thrombomodulin mRNA in response to dbcAMP. Treatment of HUVEC with 20 UM forskolin or 100 uM isobutylmethylxanthine (IBMX) also increased cell-surface thrombomodulin on HUVEC. These agents prevented the interleukin I(IL-I) or tumor necrosis factor (TNF)-induced decrease in thrombomodulin on HUVEC. These data suggest that the expression of thrombomodulin on HUVEC and mouse hemangioma cells may be regulated by intracellular CAMP level. Key Words:
Thrombomodulin, cell, Cytokine
Protein C, Cyclic AMP, Endothelial
301
302
INCREASED EXPRESSION...
Vol. 61, No. 3
INTRODUCTION Thrombomodulin on the cultured human umbilical vein endothelial cells (HUVEC) is down-regulated by tumor necrosis factor (TNF) (I), interleukin 1 (IL-l)(2) and endotoxin(3); however upregulation of thrombomodulin on HWEC has never been described. Previously we reported that MEG-01, a cell line established from human megakaryoblastic leukemia, expressed thrombomodulin(4), and this expression was enhanced by incubation of the cells with dibutyryl CAMP (dbcAMP)(5). enhanced expression of The thrombomodulin on the MEG-01 was accompanied by the increase in mRNA for thrombomodulin, suggesting that CAMP regulates the transcription of thrombomodulin gene in these cells. Here we examine the effect of dbcAMP, and other agents that increase intracellular CAMP, on the expression of thrombomodulin on the HUVEC and mouse hemangioma cells. We also investigate the effect of dbcAMP on the down regulation of thrombomodulin induced by IL-I or TNF. MATERIALS AND METHODS Materials: Materials were obtained from the following source: Dibutyryl CAMP and isobutylmethylxanthine (IBMX) from Sigma co. (St. Louis, MO).; IL-I and TNF from Boehringer Mannheim Yamanouchi (Tokyo, Japan).; Medium 199 (M-199), fetal calf serum and endothelial cell growth supplement (ECGS) from Flow Laboratories (Mclean, VA).; Penicillin, streptom%3&;]_asnoddit; glutamine from Gibco Laboratories; carrier-free [ iodide from Amersham Corp.: tissue culture flasks and centrifuge tubes from Corning Glass Works (Science Products Division, Corning, NY), and Falcon Labware (Division of Be&on, Dickinson and Co., Oxnard, CA): cluster 24, 48 dish wells from Coster (Cambridge, MA). All other chemicals were reagent grade products of Sigma (St. Louis, MO) or Protein C(6), and antithrombin III(G) were from human origin and isolated as indicated. The proteins were sodium dodecyl sulfate homogeneous as judged by The monoclonal anti-human polyacrylamide gel electrophoresis. thrombomodulin IgG (designated as l&f-TM) was prepared as described previously and lab~$~~_;~;",,', I] iodine by a modified chloramine-T procedure(7). 7,000 to 150,000 cpm/ng, G-25 column. was separated from free iodide on A Sephadex Protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad laboratories, Richmond, CA) with bovine serum albumin as standard. HUVEC were prepared by the method of Jaffe et al.(s) and Cells: cultured in M-199 supplemented with 10% fetal calf serum and 30 Confluent mg/L of ECGS in a 5% Co2 atmosphere at 37'C. monolayers of primary HUVEC were subcultured for experiments. Mouse hemangioma cells were provided by Dr. J.D. Hoak and Before the incubation with cultured as previously described(6). various agents, monolayers were washed three.times with M-199 and resuspended in the complete serum-free medium, COSMEDIUM-001 (Cosmo Bio, Tokyo); the cells could be maintained up to 36 hours in this medium. For isolation of RNA, the confluent monolayers of HUVEC cultured in IO-cm dishes were washed three times with M-199, scraped by rubber policeman and collected by centrifugation.
INCREASED EXPRESSION...
Vol. 61, No. 3
Assav of thrombomodulin
303
Activitv
Conditioned medium was removed from confluent monolayers of Then the cells and the cells were washed three times with M-199. monolayers in 6-mm culture wells were incubated with serum-free To examine the effect of dbcAMP, we medium containing dbcAMP. used increasing amounts of dbcAMP and followed a time course. The effect of other agents that increase intracellular CAMP was examined in the same manner. After incubation of the monolayers, the medium was removed, washed three times with M-199, and thrombomodulin activity on the cells was assayed. Thrombomodulin activity on the cells was measured as previously described with a The amount of activated protein C was slight modification(6). assayed by measuring the rate of hydrolysis of 0.2 mM D-Phepipecolyl-Arg-p-nitroanilide (S2238, Kabi Diagnostica, Sweden). Bindinq Assav of Cell-Surface
thrombomodulin
usinq F12511-IqG-TM
Thrombomodulin molecules on HUVEC were measured as previously Confluent monolayers of human umbilical vein described(7). endothelial cells in lo-mm culture wells were washed three times with 0.14 mol/L of NaCl, 0.004 mol/L of KCl, 0.001 mol/L of phosphate buffer, pH 7.4, 0.011 mol/L glucose containing 0.2% BSA Binding studies were (henceforth referred to as wash buffer). carried out in reaction mixtures containing M-199 supplemented with 5 mg/mL of BSA and 50 g/mL of normal human IgG 'n a total After incubation of cells with [125iI]-labeled volume of 0.3 mL. protein, the cells were washed rapidly five times with cold (O'C) wash buffer. The washed monolayers were solubilized in 0.5 mL of buffer containing 15% glycerol, 2% sodium dodecyl sulfate (SDS), 75 mmol/L of Tris-HCl (pH 6.9), and 2 mmol/L of EDTA. The solubilized cell solution was assayed for radioactivity in a scintillation counter (Beckman Instruments, Biogamma II Fullerton, CA). Specific binding was determined by measuring the difference in cell-bound radioactivity with and without loo-fold excess unlabeled IgG-TM. The cell-bound radioactivity with an excess amount of unlabeled IgG-TM was considered nonspecific and was subtracted from the total binding. Nonspecific binding was lo-15% of total binding. Dot Hybridization
of mRNA
Total RNA was prepared using guanidine isothiocyanate-sarcosyl solution followed by phenol extraction and ethanol precipitation as described previously(9). Dot hybridization was carried out by spotting a 10 pg and 5 pg of formamide denatured RNA onto dry nitrocellulose, which was then dried, hybridized with random labeled Bgl II - Hint II fragment (approximately 0.9 kb) of thrombomodulin cDNA(10) using the klenow fragment in a commercial kit (Amersham) and exposed to X-ray film. /?-actin (Oncor) hybridization was used as a reference for relative mRNA loaded(g).
304
INCREASED EXPRESSION...
Vol. 61, No. 3
RESULTS Treatment of HUVEC or mouse hemangioma cells with dbcAMP for 12 hours resulted in an increase in thrombomodulin activity and antigen. The increase in thrombomodulin by dbcAMP was concentration-dependent up to 5 mM of dbcAMP. Higher concentrations of dbcAMP decreased cell viability with reduced expression of thrombomodulin. The maximum effect was observed at 3 mM of dbcAMP (Fig. l-a). The increase in thrombomodulin at 3 mM of dbcAMP was approximately 200% of Control in HUVEC and 400% in mouse hemangioma cells. Thrombomodulin activity gradually increased during the incubation up to 12 hours of incubation. Longer incubation decreased cell viability with a decrease in thrombomodulin (Fig. l-b).
C
0
1
2
3
a
5mM
1
6
12
dbcAMP
Fig. l-a
Fig. l-b
Fis. l-a
Enhancement of the expression of thrombomodulin activity on HUVEC (0-e) and mouse hemangioma cells (0-0) by increasing amount of dbcAMP. Monolayers of HUVEC or mouse hemangioma cells were incubated for 12 hours with serum-free medium‘containing increasing amount of dbcAMP as indicated and the thrombomodulin activity was measured as described in the METHODS.
Fia. l-b
Time course of the d&-induced enhancement of thrombomodulin activity on HUVEC ‘(0-o) and mouse Monolayer of the cells were hemangioma cells (O-O). incubated with 3 mM of dbcAMP for various times up to activity measured. 24 hours and the thrombomodulin three different from (mean * S.D.) The results experiments are shown by % of control.
"--24h
305
INCREASED EXPRESSION...
Vol. 61, No. 3
We examined the effects of other agents that increase intracellular CAMP. Forskolin also increased expression of thrombomodulin activity in a concentration-dependent manner. An approximately 1.4-fold increase was observed with 30 ,uM of forskolin (Fig. 2). Isobutylmethylxanthine (IBMX, 100 PM), a phosphodiesterase inhibitor, also enhanced the expression of thrombomodulin by about 1.8-fold. The combination of forskolin (30 PM) and IBMX (100 PM) increased thrombomodulin more than either alone (Fig. 2). The effect of these agents was confirmed by radio-labeled monoclonal anti-thrombomodulin IgG binding assay (Fig. 2). 0
Formed PCa(pmofelmfh) 50
‘25f-lgG ~OUND(molecules~cellxl~~
control
dbcAMP forskolin IEMX Forskolm+lBMX
Fiu. 2
the various agents that increase in of intracellular CAMP on the expression of thrombomodulin Monolayers of HUVEC were cultured in 10 mm in HUVEC. wells, incubated with the agents for 12 hours and then evaluated for thrombomodulin activity by functional for thrombomodulin assay and antigen by labeled monoclonal anti-TM IgG binding assay. The results are of three different mean f S.D. experiments. The concentrations of the agents were as follows; dbcAMP: 3 I&?, forskolin: 30 PM,' IBMX: 100 UM Effect
100
0
I
‘1
2
3 mM
CAMP Fis. 3
Effect of dbcAMP on mRNA of thrombomodulin in HUVEC. The monolayers of HUVEC were incubated with increasing amounts of dbcAMP for 12 hours. At the indicated time, total cellular RNAs were extracted, and 5 or 10 fig of aliquots were subjected to dot hybridization as described in METHODS. The hybridized samples were subjected to autoradiography and the intensity of hybridization quantified by densitometry and analysis on a Molecular Dynamics (Sannywale, CA) laser densitometer. Relative mRNA levels of thrombomodulin were calculated by using P-actin as an internal control.
306
INCREASED EXPRESSION...
Vol. 61, No. 3
To determine whether the enhanced thrombomodulin expression by dbcAMP is accompanied by increased thrombomodulin mRNA, we evaluated the mRNA level for thrombomodulin by dot hybridization with labeled fragments of a thrombomodulin cDNA. Levels of thrombomodulin mRNA by dbcAMP were increased in a dose dependent manner and the mRNA increase was observed from 0.5 m+l to 3 mM of dbcAMP (Fig. 3). The increase in thrombomodulin mRNA was detectable within 2 hours of incubation with 3 mM dbcAMP, and peaked at 12 hours of the incubation.
Ficr. 4
Effect of dbcAMP on the IL-I-induced decrease in thrombomodulin on HWEC. The monolayers of HUVEC in wells (6 mm) were incubated with COSMEDIUM supplemented with 10 U/ml of IL-I for three hours (shaded), the medium removed and then incubated with (0-o) or without (0-C) 3 mM dbcAMP in fresh COSMEDIUM. At the indicated the thrombomodulin activity on each of the time, monolayers were measured. The results (% of control) are the mean f S.D. of three different experiments.
OO Fia. 5
6
12
18h
TNF-induced decrease in Effect of dbcAMP on the The monolayers of HUVEC were thrombomodulin on HUVEC. incubated with TNF (10 U/ml) in the presence (0-O) or At the indicated time absence (C-C) of 3 mM dbcAMP. the medium was removed and measured the thrombomodulin The results (% of control) are activity on the cells. mean k S.D. of three different experiments.
Vol. 61, No. 3
INCREASED EXPRESSION...
307
These results suggest that intracellular CAMP levels may regulate transcription of the thrombomodulin gene in HUVEC. We examined the effect of CAMP in the presence of IL-I (10 U/ml) Both IL-I and TNF reduced the thrombomodulin and TNF (10 U/ml). After the preincubation of HUVEC monolayers activity on HUVEC. with IL-I for 3 hours, the medium was removed and incubated the monolayers with serum free medium in the presence or absence of 3 mM of dbcAMP. On the dbcAMP, the monolayers without thrombomodulin activity fell to 50% of control by 9 hours. However, in the presence of dbcAMP, the rate of decrease was inhibited. This was associated with more rapid recover of thrombomodulin levels compared to the monolayers without dbcAMP (Fig. 5). Treatment of HUVEC monolayers with TNF also reduced thrombomodulin expression to approximately 40% by 18 hours. This decline was also partially prevented by treatment with dbcAMP (Fig. 6). All these experim.ents were carried out in complete serum-free medium to exclude any effects of serum. However, serum had no additional effects in the other experiments (data not shown).
DISCUSSION Endothelial cells lining blood vessels possess antithrombogenic functions mainly through production of prostacyclin(ll), secretion of tissue plasminogen activator(ll), cell surface heparin-Iike molecule that stimulates Certain COXyllatiOll factor inhibitors(l3), and thrombomodulin, an endothelial cell surface Thrombomodulin has very high affinity for thrombin glycoprotein. In this complex thrombin and forms a 1:l stoichiometric complex. has less effect on its procoagulant substrates such as whereas the ability to factor V and platelets, fibrinogen, Activated activate protein C is enhanced more than lOOO-fold. protein c functions as a potent natural anticoagulant by Thus, Factors Va and VIIIa. coagulation inactivating thrombomodulin converts thrombin from a procoagulant protease to Previously we measured thrombomodulin ~~t~gnetniso~~l~~~~(13,14) * I]IgG-TM binding assay and reported that HUVEC 30,000-50,000 approximately molecules of contains We have also reported that exposure of thrombomodulin/cells(7). HUVEC to thrombin decreased thrombomodulin on the cell-surface occurrence of endocytosis of suggesting the thrombinthrombomodulin complex(7). Phorbol myristate acetate (PMA), a potent activator of protein kinase C, also induces endocytosis of thrombomodulin(l6). Cytokines also affect the expression of thrombomodulin. TNF rapidly reduces thrombomodulin activity on the cell surface(l). The reduction in activity is associated with endocytosis and degradation of thrombomodulin(l6). TNF also inhibits transcription of thrombomodulin mBNA, and may result the subsequent decrease in thrombomodulin protein synthesis(l). Other cytokines, IL-I(2) and endotoxin(3) also cause decreased thrombomodulin activity on endothelial cell surfaces. Thus thrombomodulin expression appears to be regulated by a variety of mechanisms. However all of these factors decrease the expression of thrombomodulin; factors which increase the expression of thrombomodulin are not well characterized.
308
INCREASED EXPRESSION...
Vol. 61, No. 3
In this paper we have investigated the affect of the agents that increase intracellular CAMP and demonstrated that thrombomodulin on HWEC was up-regulated by CAMP. We previously described enhanced expression of thrombomodulin on MEG-01, a cell line from human megakaryoblas+ic leukemia, by agents that increase in intracellular CAMP(S). The agents include dbcAMP, forskolin, prostaglandin E and IBMX. Thrombomodulin expression was enhanced more than 26-fold by 3 mM dbcAMP during 48 hours and accompanied by the increase in thrombomodulin mRNA, suggesting that CAMP increased transcription of thrombomodulin message. In this paper we expanded our investigation to endothelial cells and mouse hemangioma cells. We demonstrated that incubation of the cultured endothelial cells prepared from human umbilical vein or mouse hemangioma cells with 1-5 mM dbcAMP increased the thrombomodulin activity. The increase in thrombomodulin activity was also confirmed by binding assay using radiolabeled monoclonal anti-TM IgG in HUVEC. The effect on thrombomodulin by dbcAMP was observed within 3 hours of the incubation and continued up 24 hours. Addition of fetal calf serum showed no additive effect (data not shown). Thrombomodulin increase by dbcAMP was up to 2fold in HUVEC, and 4-fold in the mouse hemangioma cells. This is much less than in MEG-07 cells. The difference between HUVEC or the mouse hemangioma cells and MEG-01 may be due to in part, different CAMP metabolism in MEG-01. In preliminary s(udies, the level of CAMP in the cells induced by dbcAMP, forskolin, or IBMX was higher in MEG-01 than in HUVEC (data not shown). As was seen in MEG-01, the increase in cell surface thrombomodulin in HUVEC induced by dbcAMP was associated with the increase in mRNA of thrombomodulin. This suggests that dbcAMP enhanced transcription of mRNA of thrombomodulin, and with subsequent increase of thrombomodulin on the cell surface. Up regulation of thrombomodulin expression by increased intracellular CAMP was supported by Imada et al. They have described CAMP enhancement of the expression of a surface marker glycoprotein, designated fetomodulin in mouse embryos(l8). Recently a cDNA clone for fetromodulin was found to be indentical to thrombomodulin(l9): The cytokine-induce reduction of thrombomodulin may have a role in the pathogenesis of thrombosis in a certain pathologic states. It may be that CAMP in endothelial cells alters thrombotic Hence we examined the effect of dbcAMP on IL-I and tendencies. Incubation of TNF induced decrease in thrombomodulin on HUVEC. HUVEC with IL-I for 3 hours reduced cell surface thrombomodulin This reduction of thrombomodulin activity to 50% of control. persisted for more than 18 hours. However the addition of dbcAMP reduced the decrease and resulted in more rapid recovery to near Similar effect was observed in control level within 18 hours. The decline in TNF-induced reduction of thrombomodulin. thrombomodulin expression in response to TNF or IL-I may have some role in the pathophysiology of prothrombotic state of The findings that dbcAMP prevented inflammation or malignancy. the cytokine-induced reduction of thrombomodulin to some extent suggests a potential role in therapeutic strategies for prothrombotic states.
Vol. 61, No. 3
INCREASED EXPRESSION...
309
ACKNOWLEDGEMENTS Part of this work was presented at the Second Saratoga Conference held on October 1-13, 1989, in Towada, Aomori, Japan This work was supported in and will be published as an abstract. part by grant-in-aid for Scientific Research from the Ministry of grants from the Ministry of Science and Culture, Education, Health and Welfare in Japan.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
Tumor necrosis factor R.D. CONWAY, E.M., and ROSENBERG, suppresses transcription of the thrombomodulin gene in endothelial cells. Mol Cell Biol, 8, 5588-5592, 1988 NAWROTH, P.P., HANDLEY, D.A., ESMON, C.T., and STERN, D.M. Interleukin I induces endothelial cell procoagulant while Proc Nat1 activity. cell-surface anticoagulant supressing Acad Sci USA, 83, 3460-3464, 1986 MOORE, K.L., ANDREOLI, S.P., ESMON, N.L., ESMON, C.T., and Endotoxin enhances tissue factor and suppresses BANG, N.U. thrombomodulin expression of human vascular endothelium in vitro. J Clin Invest, 79, 124-130, 1987 OGURA, M., ITO, T., MARUYAMA, I., TAKAMATSU, J., OGAWA, K., Localization and YAMAMOTO, S., NAGURA, H. and SAITO, H. synthesis of functional thrombomodulin in human megakaryocytes and human megakaryoblastic cell line(MEG-01). Thromb Haemostas, in press ITO, T., OGURA, M., MORISHITA, Y., TAKAMATSU, J., MARUYAMA, S., OGAWA, K., and SAITO, H. Enhanced I ., YAMAMOTO, expression of thrombomodulin by intracellular cyclic AMPincreasing agents in two human megakaryoblastic cell lines. Thromb Res, 58, 615-623, 1990 MARUYAMA, I., SALEM, H.H., and MAJERUS, P.W. Coagulation factor Va binds to human umbilical vein endothelial cells and accelerates protein C activation. J Clin Invest, 73, 224-230, 1984 MARUYAMA, I., and MAJERUS, P.W. The turnover of thrombinthrombomodulin complex in cultured human umbilical vein endothelial cells and A549 lung cancer cells. Endocytosis and degradation of thrombin. J Biol Chem, 260, 15432-15438, 1985
8. 9. 10.
JAFFE, E.A., NACHMAN, R.L., BECKER, C.G., and MINICK, C.R. Culture of human endothelial cells derived from umbilical vein. J Clin Invest, 52, 2745-2756, 1973 SAMBOOK, J., TRITSCH, E.F., and MANIATIS, T. Molecular cloning, a laborqtory manual. Second edition. Cold Spring Harbor laboratory, Cold,Spring Harbor, NY, 1989 SUZUKI, K., KUSUMOTO, H., DEYASHIKI, Y., NISHIOKA, J., MARUYAMA, I., ZUSHI, M., KAWAHARA, S., HONDA, G., YAMAMOTO, S and HORIGUCHI S. Structure and expression of human thiombomodulin, a thrombin receptor on endothelium acting as a cofactor for protein C activation. EMBO J, 8, 1891-1897, 1987
310
11.
12. 13.
14. 15. 16.
17.
18.
19.
20.
INCREASED
EXPRESSION...
Vol. 61, No. 3
WEKSLER, B.B., LEY, C.W., and JAFFE, E.A. Stimulation of by thrombin, endothelial cell prostacyclin production trypsin, and the ionophore A23187. J Clin Invest, 62, 923930, 1987 LOSKUTOFF, D.J., and EDINGTON, T.S. Synthesis of a fibrinolytic activator and inhibitor by endothelial cells. Proc Nat1 Acad Sci USA, 74, 3903-3907, 1977 of BAUER, K.A., and ROSENBERG, R.D. The pathophysiology the prethrombotic state in humans: Insight gained from studies using markers of hemostatic system activation. Blood, 70, 343-350, 1987 ESMON, C.T. The roles of protein C and thrombomodulin in the regulation of blood coagulation. J Biol Chem, 264, 47434746, 1989 ESMCN, C.T. The regulation of natural anticoagulant pathways. Science, 235, 1348-1352, 1987 DITTMAN, W.A., KUMADA, T., SADLER, J.E., and MAJERUS, P.W. The structure and function of mouse thrombomodulin: Phorbol myristate acetate stimulates degradation and synthesis of thrombomodulin without affecting mRNA levels in hemangioma cells. J Biol Chem, 263, 15815-15822, 1988 Tumor necrosis MOORE, K.L., ESMON, C.T., and ESMON, N.L. factor leads to the internalization and degradation of thrombomodulin from the surface of bovine aortic endothelial Blood, 73, 159-165, 1989 cells in culture. IMADA, M., IMADA, S., IWASAKI, H., KUME, A., YAMAGUCHI, H., Marker surface protein of and MOORE, E.E. Fetomodulin: fetal development which is modulatable by CAMP. Dev Biol, 122, 483-491, 1987 IMADA, S., YAMAGUCHI, H., NAGUMO, M., KATAYANAGI, S., IWASAKI, H., and IMADA, M. Identification of fetomodulin, a development, as marker protein of fetal surface Dev thrombomodulin by gene cloning and functional assay. _ 140, 113-122, 1990 Biol, MONTMINY, M.R., SEVARINO, K.A., WAGNER, J.A., HANDEL, G., Identification of a cyclic-AMP-responsive and GODMAN, R.H. element within the rat somatostatin gene. Proc Nat1 Acad Sci USA, 83, 6682-6686, 1986
