Heparin enhances thrombin-stimulated prostaglandin I2 production by cultured endothelial cells

THROMBOSIS RESEARCH 57;

481-488,1990 0049-3848/90 $3.00t .OO Printed in the USA. Copyright (c) 1990 Pergamon Press plc.

HEPARIN

ENHANCES

Fumiaki

Itoh,

All rights reserved.

THROMBIN-STIMULATED PROSTAGLANDIN BY CULTURED ENDOTHELIAL CELLS

Toshiyuki and

Kaji,', Yumiko Hayakawa, Nobuo Sakuragawa

12 PRODUCTION

Yutaka

Oguma

Department of Clinical Laboratory Medicine, Faculty of Medicine, Toyama Medical and Pharmaceutical IJniversity, 2630 Sugitani, Toyama-shi, Toyama 930-01, Japan

(Received11.lO. 1989; accepted in revised form 19.11.1989 by Editor W.H. Seegers)

ABSTRACT Endothelial cells human from umbilical vein were incubated in vitro with thrombin in the absence or presence of unfractionated heparin low molecular (UFH), weight heparin (LMWH) or dermatan sulfate (DS) to clarify the effect of these glycosaminoglycans on thrombinstimulated prostaglandin I, (PGI,) production. Although DS did not affect the thrombin stimulation, UFH and LMWH each enhanced it about 1.5-fold. Pretreatment of cells with either UFH or LMWH also enhanced the thrombin stimulation. In addition, UFH and LMWH each enhanced PGI, production ionophore A23187, but stimulated by calcium did not affect that stimulated by arachidonic acid. From these results, it was suggested that UFH and LMWH might cause a physical change of cell membrane, which would caused an easier liberation of arachidonic acid by phospholipases.

INTRODUCTION Heparin activity and

is well known other biological

to have activities

antithrombotic a potent (1, 2). The ma,jor mech-

Heparin, Glycosaminoglycans, Endothelium, : Words Key molecular weight heparin, Prostaglandin I,, Thrombin " To whom correspondence should be addressed. 481

Low

482

HEPARIN AND ENDOTHELIAL CELLS

Vol. 57, No. 4

anisms of its antithrombotic action were recognized as antiserine prot.ease actions. It greatly accelerates antithrombin III binding to factor Xa and thrombin, or inhibits serine proteases directly. Recently, it has been well known that heparin and vascular endothelial cell interaction occupies a prominent position in anticoagulant action of heparin. Heparin releases tissue plasminogen activator (t-PA) and glycosaminoglycans (GAGS) from endothelial cells (1). Also, endothelial cells produce the antithrombotic agent, arachidonic prostaglandin Ia (PGIZ). PGI,, a major acid (AA) metabolite of endothelial cells, strongly prevents platelet aggregation, and is one of the potent vasodilators. Talking about the effect of heparin on PGI, production, Buchanan et al. (3) heparin prior to reported that rabbit aorta, infused with stimulated sacrifice, exhibited an enhancement of AA PGI, of coronary disease, heparin In patients production in vitro. concentration in coronary sinus blood (4). On increased the PGI, showed that an intravenous et al. (5) the other hand, Ljungberg of heparin did not affect urinary excretion of PGI, injection metabolite. we studied the effect of heparin cofactor 11 and Recently, production by thrombin-stimulated PGI, antithrombin III on endothelial cells from human umbilical vein (HUVE)(G). In the we observed that unfractionated heparin process of this study, (LMWH) each enhanced low molecular weight heparin (UFH) and In this study, we thrombin-induced PGI,? production by HUVE. (DS) on LMWH and dermatan sulfate describe effect of UFH, and discuss the production by HUVE, thrombin-induced PGI, mechanism and significance of the effect of UFH and LMWH. MATERIALS

AND

METHODS

Materials Thrombin (from human), collagenase calcium (Type I), ionophore A23187 (A23187), AA and endothelial cell growth (ECGS) were purchased from Sigma. supplement UFH (Av.M.W. LMWH (M.W. 4,000 - 6,000) and DS (M.W. 11,000 - 25,000) lZ,OOO), were obtained from Sigma, KabiVitrum AB and Seikagaku Kogyo, respectively. 6-keto-prostaglandin Fla The (6-Keto-PGFla) radioimmunoassay kit was purchased from Amersham Corp. RPM11640 medium and foetal bovine serum were obtained Nissui from Pharmaceutical Co.Ltd. and Filtron, respectively. Gelatincoated petri dishes (35 mm, 60 mm) and culture plates (24 well) were purchased from Corning. Cell from were

culture Endothelial cells were prepared by collagenase treatment human umbilical cord vein as previously described (7) and cultured in gelatin-coated 60 mm petri dishes at 37 "C in a

Vol. 57, No. 4

HEPARIN AND ENDOTHELIAL CELLS

483

They humidified chamber with an atmosphere of 5% CO2 in air. 1640 medium containing 20% in RPM1 were grown to confluence and 100 foetal bovine serum, 200 JJg/ml ECGS, 100 U/ml penicillin utilized for experiments at Cells were rrg/ml streptomycin. in 35 mm petri dishes or 24-well tissue passages 6 through 9, culture plates coated with gelatin. Cell

treatment 35 mm petri dishes were used, In the time course experiment, The 24-well tissue culture plates were used. and in the others, were washed twice confluent cells (3.88 + 0.27 x lo* cells/cmz) 11 mM HEPES, 137 mM NaCl, 4 in HEPES buffer (pH 7.1), containing 1 mM MgCl, and 11 mM glucose. They were mM KCl, 3 mM CaCl,, incubated with 1 ml (for 35 mm petri dishes) 01' 0.25 ml (for 40 24-well tissue culture plates) of HEFES buffer at 37 "C up to min. Thrombin (1 NIH U/ml and below), I'FH ( 1 IJSP U/ml), LMWH ( 1 U/ml ; anti-factor Xa activity), or DS (0.2 mg/ml) were solved in The activity of HEPES buffer and added to the incubation buffer. UFH and LMWH were expressed as each units because USP unit and anti-factor Xa activity are almost equal. After incubation, the buffer was harvested, and an aliquot was used for the assay of PGI,. A stimulation of the cells by 10 Jl?I A23187 or 25 !JM AA was performed in the same way. Assay

of PGIa The content radioimmunoassay FGI=.

of kit

PGI, for

in incubation 6-keto-PGFlti,

Statistical analysis The significance of was estimated by student's

the difference t-test..

buffer was measllred a stable metabolite

from

the

control

by of

value

RESULTS PGI, production was stimulated by 0.1 and 1 NIH U/ml thrombin, 20 min (Fig. 1). PGI, was and reached plateau within hardly produced in control and 0.01 NIH U/ml thrombin groups. Fig. Cells were incubated with 0.5 NIH II/ml thrombin for 20 min. 2 shows individual UFH, LMWH and DS on the effect of HUVE produced 70.3 f 5.0 thrombin-stimulated FGI, production. UFH and pg/well PGI, in the presence of 0.5 NIH U/ml thrombin. thrombinenhanced the LMWH at 1 U/ml each significantly stimulated PGI, production to 103.4 + 4.6 pg/well and 102.2 _+ 4.3 pg/well, respectively. The enhancement occurred when the (data not concentration of UFH or LMWH was 0.01 U/ml and above thrombin stimulation only shown). DS, enhanced the however, slightly. In the absence of thrombin, all of the tested GAGS did Although we not induce FG12 production (data not shown).

484

HEPARIN AND ENDOTHELIAL CELLS

lb

0

lo

TIME

(min)

Vol. 57, No. 4

3-0

Time course of thrombin-induced PGI= production by HUVE. Fig. 1 Cells were incubated with thrombin at 37 "C for indicated time intervals in 35 mm dishes, and PGI, production was measured by radioimmunoassay for 6-keto-PGFla. ; 0, 0.01 NIH 0, control ??, 1.0 NIH U/ml thrombin. Each point , 0.1 NIH (J/ml ; U/ml ; ? ? is the mean f S.E. of three samples.

0

CONT UFH LMWH DS

PGI, proFig. 2 Effect of UFH, LMWH and DS on thrombin-induced duction. Cells were incubated with 0.5 NIH U/ml thrombin for 20 min at 37 "C in the absence or presence of UFH (1 U/ml), LMWH (1 U/ml) or DS (0.2 mg/ml). Each column represents the mean + S.E. of eight samples. XxF P
HEPARIN AND ENDOTHELIAL CELLS

Vol. 57, No. 4

TABLE 1 with UFH,

Effect of pretreatment induced PGI, production.

LMWH

6-Keto-PGFla CONTROL UFH LMWH DS

9.37 16.96 12.55 5.92

485

and

DS

on

thrombin-

(pg/well) + + + k

2.66 2.00 2.20 2.06

Cells were incubated with UFH (1 U/ml), LMWH (1 U/ml) were washed (0.2 mg/ml) at 37 "C for 1 hour. Monolayers and then incubated with thrombin (0.5 NIH U/ml) at 37 "C f S.E. of three samples. min. Each value is the mean

Effect of production.

UFH,

LMWH

and

TABLE DS on

2 A23187-

6-Keto-PGFla

113.0 133.1 138.9 114.2

+ + f +

AA-stimulated

PGIz

(pg/well)

A23187 CONTROL UFH LMWH DS

or

or DS twice, for 20

AA 8.4 7.8 8.0' 4.8

378.4 393.5 402.7 416.9

+ f + +

20.2 17.4 13.1 35.3

Cells were incubated with A23187 (10 MM) or AA (25 PM) in the or DS (0.2 absence or presence of UFH (1 (J/ml), LMWH (1 U/ml) mg/ml) at 37 "C for 20 min. Each value is the mean f S.E. of six samples. a:P(O.05, compared with control.

showed the data using heparin extracted from bovine lung, either porcine hepatic or intestinal mucosal heparin had an equal effect (data not shown). In investigate which is more important, an order to interaction of thrombin or that of HUVE with heparin for the enhancement of PGI, production, cells were preincubated with UFH, LMWH or DS, and then stimulated by thrombin. PGIz content The PGI, production in released from HUVE is shown in Table 1. this experiment was less than those in the other experiments as The thrombin stimulation of cells were preincubated for 1 hour. was enhanced by the pretreatment of either UFH or PGI, production This result indicates that heparin acted on cells but did LMWH.

486

HEPARIN AND ENDOTHELIAL CELLS

Vol. 57, No. 4

not

act on thrombin. In the process of PGI, production, it is possible that heparin facilitated the thrombin-receptor binding, increased an elevation of intracellular Ca?+ concentration, activated phospholipases or changed AA metabolism. To clarify these possibilities, we examined the effect of UFH, LMWH or DS on PG12 production in the presence of nonreceptor-mediated stimulants, A23187 or AA. As shown in .Table 2, UFH and LMWH each enhanced PGI, production stimulated by A23187. In contrast, they did not change the production stimulated by AA. These data suggest that UFH and LMWH enhanced liberation of AA by phospholipases. DISCUSSION Recently, role of endothelial cells in the blood the has received much attention. Thrombin coagulation system stimulates the release of GAGS (8), t-PA (9) and PGI, (6, 10) cells. In addition, thrombin-thrombomodulin from endothelial of factor complex activates protein C (ll), leading to inhibition endothelial cells convert thrombin Thus, V and VIII activation. action from thrombotic to anti-throm‘ stic or fibrinolytic. In that small amoul_;s of thrombin were infused fact, it was reported treating thrombosis (12). intravenously in the interest of Although heparin usually inhibits coagulative thrombin action, in that UFH and LMWH each enhanced this study, we demonstrated thrombin stimulation of PGI, production by HUVE (Fig. 2). As UFH and LMWH may enhance prevents platelet aggregation, PGI, This converting activity of endothelial cells against thrombin. high degree of efficiency for effect is a enhancement anticoagulant action of heparin to prevent an excess of platelet since aggregation reaction is usually initiated by aggregation, vessel wall, the adhesion to damaged parts of the near that interaction of endothelial cells. It is possible the thrombin with UFH and LMWH will occur in vivo because the blood concentration of heparin reaches about 1 U/ml in cases where heparin is used clinically. DS is produced by fibroblasts and smooth muscle cells in subendothelial or extravascular tissues (13). DS did not enhance PGI, production stimulated by thrombin (Fig. 2). It is profitable for the coagulant system, since DS exists on the injured vascular wall, where the coagulation of the blood is needed. Thrombin binding to endothelial cells reaches a plateau within 1 min and the thrombin-induced elevation of (I4), intracellular Ca"+ levels peaks within 30 set (15), resulting in an activation of phospholipases. On the other hand, it takes a few hours for heparin-endothelial cell binding to reach a steady-state Vannucchi et al. reported that (14). (16) E3Hlheparin bound on the cell surface was internalized for 15

Vol. 57, No. 4

HEPARIN AND ENDOTHELIAL CELLS

487

Thus, it is unlikely that the enhancement of thrombin min. intracellular was caused by PGI, production stimulation of would not contribute heparin within 20 min, and protein synthesis to this reaction. by the physical Phospholipase A, activity is influenced reported that substrate. Barlow et al. state of the (17) of phospholipids played a specific packing-induced conformations substantial role in the interfacial activations of phospholipase that the enzyme activity Roberts et al. (18) also reported A bound collision between the enzyme, random wzs dependent on We speculate that and substrate in mixed micelles. phospholipid, induced a physical changes of cell binding of heparin or LMWH Such a change would provide an easy hydrolysis of membrane. and might induce an increase in phospholipids by phospholipases, Although UFH and LMWH did not have a stimulatory efCa influx. pretreatment with either of individually, fect on PGI, production them could induce an enhancement of the thrombin stimulation (Table 1). In addition, UFH and LMWH each enhanced PG12 production stimulated by Ca ionophore (Table 2). These results suggest that UFH and LMWH each induced an easier hydrolysis of phospholipids by thrombin-stimulated phospholipases rather than activated phospholipases by an increase in Ca influx. As a result, UFH and LMWH would enhance the thrombin stimulation of PGI, production by HUVE. In conclusion, it is that heparin possibly suggested enhances endothelial cell-mediated anticoagulant activity against the coagulant action of thrombin via the PGI, production. This enhancement the important one of roles in the may play anticoagulant action of heparin in vivo. REFERENCES 1. 2. 3.

4.

3.

FAREED,J. Heparin, its fractions, fragments and derivatives. Some newer perspectives. Semin.Thromb.Hemost. 11, 1-9, 1985. ENGELBERG,H. Heparin and the aterosclerotic process. Pharmacological Reviews, 36, 91-110, 1984. BUCHANAN, M.R., DEJANA, E., CAZENAVE, J.P., MUSTARD,J.F. and by whole vessel wall HIRSH,J. Uncontrolled PGIa production in vivo and its prevensegments due to thrombin generation tion by heparin. Thromb.Res. 16, 551-555, 1979. GOLDBERG, WALLIS,J., MOSES,J.W., BORER,J.S., WEKSLER,B.B., TALK-GOLDMAN,K., CARTER,J. and H.L., FISHER,J., KASSE,M. , Coronary blood flow in coronary artery disease : CALLE,S. Heparin-induced potentiation caused by prostacyclin release. Circulation 9 : suppl. II, 263, 1982. JOHNSSON, H. and BEVING, H., EGBERG, N., LJUNGBERG, B., Immediate effect of heparin and LMW heparin on VESTERQVIST,O. Thromb. Res. and endothelial derived factors. some platelet 1988. _$_J, 209-217,

488

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

16.

17.

18.

HEPARIN AND ENDOTHELIAL CELLS

Vol. 57, No. 4

KAJI, T., ITOH, F., HAYAKAWA, Y., OGUMA,Y. and SAKURAGAWA,N. Interaction of thrombin with heparin cofactor Jl and antithrombin Ill on prostacyclin production by cultured endothelial cells. Thromb.Res. (in press), 1989. JAFFE, E.A., NACHMAN, R.L., BECKER,C.G. and MINICK,C.R. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic immunologic criteria. and J.Clin.Invest. 52, 27125-2756, 1973. of heparan sulfate proteoSHTMADA, K. and OZAWA, T. Release glycans from cultured aortic endothelial cells by thrombin. Thromb.Res. 2, 387-397, 1985. LEVIN, E.G., MARZEC,U., ANDERSON,J. and HARKER,L.A. Thrombin stimulates tissue plasminogen activator release from cultured human endothelial cells. J.Clin.lnvest. 72, 1988-1995, 1984. and thrombin on prostacyclin HONG, S.L. Effect of bradykinin in endothelial cells from calf and pig aorta and synthesis human umbilical cord vein. Thromb.Res. l8, 787-795, 1980. ESMON,C.T., ESMON,N.L. and HARRIS,K.W. Complex formation bethrombomodulin inhibits both thrombinthrombin and tween activation. fibrin formation and factor V catalyzed J.Biol.Chem. 257, 7944-7947, 1982. SEEGERS, W.H. and OUYANG, C. Snak; venoms and blood coagulation. ,Handb.Exp.Pharmacol. 52, 684-750, 1979. Activation of heparin and TOLLEFSEN,D.M. coMcGUIRE,E.A., by fibroblasts and vascular smooth muscle cells. factor II 1987. J.Biol.Chem. 262, 169-175, MACHOVICH, R., ARANYI, P., BUKI,K.G., CSONKA,E. BAUER, P.I., Mechanism of thrombin binding to endothelial and HORVSTH,I. 1983. cells. Blood 62, 368-372, JAFFE,E.X., GRULICH,J., WEKSLER,B.B., HAMPEL,G.and WATANABE, Correlation between thrombin-induced prostacyclin producK. tion and inositol trisphosphate and cytosolic free calcium levels in cultured human endothelial cells. J.Biol.Chem. G, 8557-8565, 1987. VANNUCCHI, S., PASQUALI, F., CHIARUGI, V. and RUGGIERO, M. Internalization and metabolism of endogenous heparin by culcells. Biochem.Biophys.Res.Commun. tured endothelial 140, 294-301, 1986. BARLOW, P.N., LISTER,M.D., SIGLER,P.B. and DENNIS,E.A. Probing the role of substrate conformation in phospholipase Aa action on aggregated phospholipids using constrained phosJ.Biol.Chem. 12954-12958, phatidylcholine analogues. 263, 1988. ROBERTS,M.F., DEEMS,R.A. and DENNIS,E.A. Dual role of interfacial phospholipid in phospholipase A, catalysis. Proc. Natl.Acad.Sci.USA E, 1950-1954, 1977.