Isolation and characterization of thrombomodulin from bovine lung

Biochimica et Biophysica Acta 882 (1986) 343-352 Elsevier

343

BBA 22358

Isolation and characterization of t h r o m b o m o d u l i n f r o m bovine lung

K o j i S u z u k i *, H i r o s h i K u s u m o t o a n d S e n i c h i r o H a s h i m o t o Department of Laboratory Medicine, Mie University School of Medicine, Tsu City, Mie 514 (Japan)

(Received January 2nd, 1986)

Key words: Thrombomodulin; Protein C activation; Anticoagulant; (Bovine lung)

Bovine thrombomodulin was isolated from the lung by Triton X extraction, affinity chromatography on diisopropyl phosphate-thrombin-agarose, and gel filtration on Ultrogel AcA-44. The final preparation was purified 6000-fold from the membrane extract with a yield of 21%. It showed apparent M r of 78000 and 105 000, before and after reduction, respectively, on polyacrylamide gel electrophoresis in SDS. The activity of the thrombomodulin was stable under the conditions of 1% SDS, 8 M urea, pH 2 and 10, and heat treatment at 60°C for 30 min, but was unstable against treatment with 2-mercaptoethanol. Activation of protein C by thrombin in the presence of the thrombomodulin depended on Ca2+, and an equimolar complex formation between thrombin and thrombomodulin was required for the maximum rate activation. The rate of protein C activation by thrombin was increased 900-fold by thrombomodulin. Thrombomodulin inhibited the thrombin-induced fibrinogen clotting and platelet activation. However, it did not affect the inhibition of thrombin by antithrombin III with or without heparin, a protein C inhibitor or several synthetic inhibitors. These properties of bovine thrombomodulin were similar to those of rabbit thrombomodulin reported earlier.

Introduction

Vitamin K-dependent protein C is a zymogen of plasma serine proteinase, which plays a role in the regulation of blood coagulation due to inactivation of the coagulation factors Va and Villa [1-6]. Protein C is activated by thrombin, that binds to a receptor on the vascular endothelial surface [7,8]. This receptor protein, called thrombomodulin [9], acts as a cofactor to accelerate the thrombin-catalyzed activation of protein C. Thrombomodulin also has an anticoagulant activ-

* To whom correspondence should be addressed. Abbreviations: DFP, diisopropyl fluorophosphate; SDS, sodium dodecyl sulfate; MD-805, (2R,4R)-4-methyl-l-[N 2(3-methyl-l,2,3,4-tetrahydro-8-quinoline sulfonyl)-L-arginyl]2-piperidinecarboxylic acid monohydrate; 1-2581, N% dansyl( p-guanidino)phenylalanine-piperidine hydrochloride; Boc-, butoxycarbonyl-; -MCA, -amidomethylcoumarin.

ity itself. Once the protein is bound, thrombin loses its ability not only to clot fibrinogen and to activate factor V, but also to trigger platelet aggregation [10,11]. Therefore, thrombomodulin is an important regulator for intravascular coagulation. This protein so far has been isolated from the lung of rabbit [9], rat [12], dog [12] and human placenta [13]. We isolated thrombomodulin from bovine lung and examined its physicochemical properties. Further, the functional properties of this protein were compared with those of the protein isolated from rabbit lung, which has been described previously by Esmon et al. [9]. Materials and Methods Materials

The following chemicals were obtained from commercial sources and were of the highest grade

0304-4165,/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

344

available: Ultrogel AcA-44 and Ampholine-containing polyacrylamide gel plate (LKB Product, Sweden); CNBr-activated Sepharose 4B and Sephadex G-25 (Pharmacia Fine Chemicals, Sweden); diisopropyl fluorophosphate (DFP) (Fluka A.G., F.R.G.); fibrinogen and bovine serum albumin (Sigma Chemical Co., MO, U.S.A.). Triton X-100 and heparin (Wako Pure Chemicals, Osaka); SDS, benzamidine chloride and Lubrol PX (Nakarai Chemicals, Kyoto); chromogenic substrates for thrombin, Boc-Val-Pro-Arg-MCA [14], and activated protein C, Boc-Leu-Ser-Thr-Arg-MCA [15] (Protein Research Foundation, Osaka). Synthetic thrombin inhibitors (2R,4R)-4-methyl-1[ N 2-(3-methyl-l,2,3,4-tetrahydro-8-quinoline sulfonyl)-L-arginyl]-2-piperidinecarboxylic acid monohydrate, (MD-805) [16] and N%dansyl(pguanidino)phenylalanine-piperidine hydrochloride (I-2581) were provided by Mitsubishi-Kasei, Tokyo and Kabi, Sweden, respectively.

Proteins Protein C was purified from bovine plasma by the method of Stenflo [1]. Thrombin was prepared by the method of Lundblad et al. [17], and its specific activity was 2300 units/mg protein. Bovine antithrombin III was purified by the method of Kurachi el al. [18]. The protein C inhibitor was purified from human plasma as described previously [19]. Rabbit thrombomodulin was obtained from American Diagnostica Inc. The concentration of each protein was determined spectrophotometrically. The M r and absorption coefficient used for the respective proteins were as follows: protein C, 58000, a280n m l ~ =13.7 [2]; thrombin, 37000, a 1~ 280nm = 18.3 [20]; rabbit thrombomodulin, 74000 [9], a280n m l ~ = 14.0 (tentative value).

Diisopropyl phosphate-thrombin-agarose Thrombin-agarose was prepared with bovine thrombin and CNBr-activated Sepharose 4B according to the manufacturer's instructions. Thrombin was coupled to the agarose to give 2.7 mg thrombin/ml agarose. The thrombin-agarose was then treated with 1 mM DFP in 0.05 M Tris-HC1/0.1 M NaC1 (pH 7.5).

Isolation of thrombomodulin from bovine lung Extraction of cell membrane proteins. Isolation

of thrombomodulin from bovine lung was performed basically according to the method of Esmon et al. for its isolation from rabbit lung [9]. 1 kg of freshly frozen bovine lung, obtained from the local slaughterhouse, was chopped after separation of attached tissues, and washed with stirring at 4°C in 10 liters 0.02 M Tris-HCl/0.1 M NaC1/0.05 mM EDTA/1 mM benzamidine/1 mM DFP (pH 7.5) to remove the blood. The material was then homogenized in 700 ml 0.02 M Tris-HC1/1 M NaC1/0.25 M sucrose/0.05 mM EDTA/1 mM benzamidine/1 mM DFP (pH 7.5) using a Waring blender on ice. The homogenate was centrifuged at 12000 x g for 30 min at 4°C to obtain the pellet. The pellet was suspended in the same buffer and washed twice more. The washed pellet was suspended in 500 ml 0.02 M TrisHC1/0.25 M sucrose/0.5 mM CaC12/1 mM benzamidine/0.02% azide/0.5% Triton X-100 (pH 7.5) and homogenized with the blender. The homogenate was centrifuged at 12 000 x g for 50 min at 4°C to obtain the supernatant containing cell membrane proteins. This extraction from the pellet was repeated twice more. Finally, 1.9 liters of the membrane extract were obtained.

Affinity chromatograph), on diisopropyl phosphate-thrombin-agarose. Membrane extract (1.9 liters) was applied to a column (2.5 x 16 cm) of diisopropyl phosphate-thrombin-agarose equilibrated with 0.02 M Tris-HC1/0.1 M NaC1/0.5 mM CaC12/1 mM benzamidine/0.5% Lubrol PX (pH 7.5) (equilibration buffer) at room temperature. To remove the undissolved materials, a column (2.5 x 10 cm) packed with Sephadex G-25 was connected just before the diisopropyl phosphate-thrombin-agarose column. After application of the extract, the column was washed with 500 ml of the equilibration buffer. Then the thrombomodulin was eluted from the column with the buffer consisting of 0.02 M Tris-HC1/1 M NaC1/0.1 mM EDTA/1 mM benzamidine/0.5% Lubrol PX (pH 7.5). The fractions containing thrombomodulin, indicated by the bar in Fig. 1, were pooled and dialyzed against 0.02 M TrisHC1/0.1 M NaC1/0.5 mM CAC12/0.05% Lubrol PX (pH 7.5) at 4°C overnight.

Rechromatography on diisopropyl phosphatethrombin-agarose. The dialysate (36 ml) was applied to the column (1.5 x 16 cm) of diisopropyl

345

phosphate-thrombin-agarose in the equilibration buffer, and the column was washed with 60 ml 0.02 M Tris-HC1/0.3 M NaC1/0.5 mM CAC12/0.5% Lubrol PX (pH 7.5). Thereafter, thrombomodulin was eluted with 0.02 M TrisHC1/1 M NaC1/0.5 mM EDTA/0.05% Lubrol PX (pH 7.5). The pooled eluate was concentrated using a Millipore Immersible CX-30. Gel filtration on Ultrogel AcA-44. The concentrated sample (4 ml) was applied to a column (1.5 x 97 cm) of Ultrogel AcA-44 equilibrated with 0.02 M Tris-HC1/0.1 M NaC1/0.5 mM CAC12/0.05% Lubrol PX (pH 7.5) and eluted with the same buffer at a flow rate of 13 ml/h. Eluates containing thrombomodulin were pooled and stored at - 80 °C.

Assay of thrombomodufin activity Activity of thrombomodulin was assayed as cofactor activity towards thrombin-catalyzed activation of protein C. A typical assay was as follows. To 100/~10.05 M Tris-HC1/0.1 M NaC1/2 mM CaC12/0.1% bovine serum albumin (pH 8.5) were added 10 ~tl protein C (280 /tg/ml), 10 /zl thrombin (5 /~g/ml) and 10 #1 of the sample containing thrombomodulin. The mixture was incubated at 37°C for 30 min, and the reaction was terminated by the addition of 20 ~1 antithrombin III (150 /~g/ml) and 5 t~l heparin (0.1 mg/ml). The amount of activated protein C formed was determined as amidolytic activity for the fluorogenic substrate, Boc-Leu-Ser-Thr-ArgMCA, as described previously. One unit of thrombomodulin activity was defined as I nmol activated protein C formed/ml incubation mixture per min. To study the effect of Ca2+ on the protein C activation, buffers were rendered Ca2+-free by treatment with Chelex (Bio-Rad) prior to the experiment.

Fibrinogen clotting assay Fibrinogen clotting time was measured by a KC-10 coagulometer (Amelung, F.R.G.). To 80/~1 0.05 M Tris-HC1/0.1 M NaC1/0.1% bovine serum albumin (pH 7.5) were added 10 btl thrombin (6.25 /~g/ml) and 10 /~1 thrombomodulin solution, and the mixture was incubated at 37°C for 2 min. Then 100 /~1 bovine fibrinogen (4 mg/ml) in the

Tris-buffered saline was added to the mixture and the clotting time was counted simultaneously.

Platelet aggregation assay Platelet aggregation was measured turbidimetrically in a platelet aggregometer (Sienco) at 37°C with continuous stirring at 1000 rpm.

SDS-polyacrylamide slab gel electrophoresis Slab gel electrophoresis in SDS was performed according to the method of Blobel and Dobberstein [21] by using 5-15% gradient gels. 2 ~g of the sample protein were treated beforehand with 2% SDS in the presence or absence of 2-mercaptoethanol at 80°C for 10 min. After the electrophoresis, the gel was stained with a silver stain, using Ag-Stain 'Daiichi' (Daiichi Pure Chemicals, Tokyo) according to the manufacturer's instructions.

A mino-acid analysis Thrombomodulin (0.5 nmol) dialyzed against 0.1% SDS was hydrolyzed in 6 M HC1 at 110°C for 24 h and 72 h under vacuum. Analyses were carried out by a Hitachi amino-acid analyzer, Model 835, as recommended by the manufacturer.

Antiserum Antiserum for thrombomodulin was obtained from the rabbits immunized by administration of the purified thrombomodulin (50 /~g per animal each time) in Freund's complete adjuvant biweekly five times. Results

Purification of thrombomodulin The bovine thrombomodulin was purified from the Triton X extract of the lung with three column chromatographies as described in Materials and Methods. Fig. 1A, B, C shows the elution profile of each chromatography. Like rabbit thrombomodulin, bovine thrombomodulin strongly bound to diisopropyl phosphate-thrombin-agarose in the presence of low concentrations of Ca2+ and dissociated from the agarose in the presence of EDTA and a high concentration of NaC1. On the second diisopropyl phosphate-thrombin-agarose chromatography, almost all of the proteins eluted with

~

0.4

~

¢;:

:

c

-.

"Oo~,~...o.o.o.o.~

1

~"O"O"

i •

i

...... O " ' - - O i

0.6

0.8

0.2

0.3

.o <~ 0.1

=

c

o=

3

i

25

0

0

Elution

=

50

-~

Elution

.O " ' ' ' O ' ' ' ' 0

O.3M Nm(~

""0""0""0

O-...O'" -~

(B)

2000

i

0

•

volume

75

i

:;

120

i

(ml)

100

(ml)

1M NaC~

volume

60

i

i

125

...

180

.....

O.

. ....

i

150

0

240

0.5

1.0

1.5

0

0.2

c

NICI EDTA

.=0.1

=-

1M

0.1ram

1.0

0.4

o•

0

(A)

0.2

=~

o=

o~ 0 . 3

o

o

0.5

<

E

E

E

A

E

E

c E

v

0.05

0.10

0

3t/

i

50

i

60

i

5

i

80

I

90

Elution v o l u m e

70

1 234

i

(ml)

100

i

110

120

i

Mr

85,000

94,000

-~- 1 0 5 , 0 0 0

0.5

1.0

1.5

2.0

Fig. 1. (A) Affinity chromatography of the Triton X extract on diisopropyl phosphate-thrombin-agarose chromatography. The bar indicates the eluate pooled for rechromatography on diisopropyl phosphate-thrombin-agarose. (B) Rechromatography of the crude thrombomodulin on diisopropyl phosphate-thrombin-agarose. The bar indicates the eluate pooled for gel filtration. (C) Gel filtration on Ultrogel AcA-44 column. Eluates indicated as a bar were pooled and stored at - 80 o C. SDS-polyacrylamide slab gel electrophoretic patterns of the reduced thrombomodulin are inserted in the figure. Gels 1 - 5 correspond to the respective fractions indicated by the arrows. M r of the respective proteins were estimated from the mobility of the standard proteins. APC, activated protein C.

<

O o3

.Q

O E

O 00 O4

E t-

b

0.15

(c)

n

o

E O

"10

E

E -5 E

B

E

347

TABLE I P U R I F I C A T I O N OF BOVINE T H R O M B O M O D U L I N 1 unit = 1 nmol activated protein C f o r m e d / m i n per ml. DIP, diisopropyl phosphate. Step

Volume (ml)

Total protein (mg)

Total activity (units)

Triton X extract DIP-thrombinagarose (I) DIP-thrombinagarose (II) Ultrogel AcA-44

1900

28 900

9 250

Specific activity (units/mg)

Yield (%)

0.3

Purification (fold)

100

1

36

7.9

3 050

390

33

1 300

39 19

2.4 1.1

2 440 1970

1030 1790

26 21

3 430 6000

EDTA buffer consisted of the intact thrombomodulin and its degraded proteins, as described later. On the Ultrogel AcA:44 gel filtration, the intact thrombomodulin and the lower-M r proteins were separated (Fig. 1C). The SDS gel electrophoretic pattern of the proteins that were reduced is also shown in Fig. 1C. The main peak protein showed an M r of 78 000 and 105 000, before and after reduction, respectively. To elucidate whether

or not the lower-M r proteins, which still had cofactor activity, were derivatives of the thrombomodulin, Western blotting analysis [22] was performed by using antiserum that had been prepared against the main peak protein with an M, of 105 000 after reduction. The immunoglobulin reacted with all of the bands showing an M r from 85 000 to 105 000 after reduction (data not shown). This suggests that the lower-M r proteints bound to diisopropyl phosphate-thrombin-agarose are the degradation products of the intact thrombomodulin.

T A B L E II C O M P A R I S O N OF A M I N O - A C I D C O M P O S I T I O N OF BOVINE, H U M A N A N D RABBIT T H R O M B O M O D U L I N A m i n o acid "

Bovine

Human b

Rabbit c

(%)

(%)

(%)

1.5

c

Asp Thr Ser Glu Pro Gly Ala Val Met Ile Leu Tyr Phe His Lys Arg

7.9 5.1 5.9 13.1 8.6 11.8 10.5 6.1 1.3 3.9 9.7 1.3 3.5 2.7 3.5 5.1

9.1 4.4 7.4 12.5 5.9 14.1 8.8 6.3 1.6 3.4 8.8 2.8 2.9 2.2 5.0 4.7

9.0 4.3 6.8 10.0 10.0 12.5 10.9 7.0 1.5 3.1 9.5 2.5 3.4 2.3 2.6 4.7

a Results expressed as percent of residue in protein. b Salem et al. [13]. c Esmon et al. [9].

E "6

1.0

E E 2

/

0

['~]

/

0.5 1'o

5

0

5

10

[,c2"+,] 0 1

2

3

4

Ca 2~ ( m M )

Fig. 2. Effect of Ca 2 + on protein C activation by thrombin and thrombomodulin. Protein C was activated at 3 7 ° C for 30 min with 0.72 ~tM protein C, 1.9 n M thrombomodulin and 5.6 nM thrombin in 100/tl 0.05 M Tris-HCl/0.1 M NaC1/0.1% bovine serum albumin (pH 8.5), and CaCI 2. For this experiment, Ca 2÷ was removed from the buffers by Chelex before use. APC, activated protein C.

348 TABLE Ill STABILITY OF BOVINE T H R O M B O M O D U L I N Conditions 1-5: thrombomodulin (3.0 /~g/ml) was incubated in 0.02 M Tris-HCl/0.1 M NaC1/0.05% Lubrol PX (pH 7.5) (reaction buffer) containing the reagent indicated in the table, at room temperature for 3 h. Then the sample was diluted 10-fold with 0.05 M Tris-HC1/0.1 M NaCI/0.1% bovine serum albumin/0.05% Lubrol PX (pH 8.5) (assay buffer) and the cofactor activity was assayed. Conditions 6 and 7: thrombomodulin (3.0 /~g/ml) was incubated at room temperature in 0.02 M glycine-HC1/0.1 M NaC1/0.05% Lubrol PX (pH 2.0) containing 5 m M CaC12 or 5 m M EDTA, or 0.02 M glycineN a O H / 0 . 1 M NaC1/0.05% Lubrol PX (pH 10.0), containing 5 m M CaCI 2 or 5 m M EDTA. Then the sample was diluted 10-fold with the assay buffer, and the cofactor activity was assayed. Condition 8: thrombomodulin (3.0 /~g/ml) in the reaction buffer containing 0.5 m M CaCI 2 was treated at 6 0 ° C for 30 min. The sample was then cooled on ice, and the cofactor activity was assayed. Condition 9: thrombomodulin (3.0 /xg/ml) in the reaction buffer containing 0.5 m M CaCI 2 was boiled for up to 120 min. At intervals, the samples were cooled on ice and the cofactor activity was assayed. A half-life (/'1/2) was estimated from the rate of the terminal activity in relation to the original activity. Cofactor activity of thrombomodulin after treatment is indicated as percent of the original activity.

obtained by using the protein with an M r of 105000 after reduction. Table II shows the amino-acid composition of the purified protein, compared with the rabbit and human thrombomodulin. Although bovine thrombomodulin appeared to contain a relatively larger amount of glutamic acid and a smaller amount of proline than those occurring in rabbit thrombomodulin, the overall composition of the bovine protein was similar to that of the rabbit and human proteins. Bovine thrombomodulin displayed heterogeneity on isoelectric focusing carried out on a polyacrylamide gel plate, and its isoelectric point was 4.0-4.3. The absorption coefficient (a280nm) 1~ in 0.02 M Tris-HC1/0.1 M NaC1/0.5 mM CAC12/0.05% Lubrol PX (pH 7.5) was 13.8 as determined from the protein concentration by the dye-binding method [23]. The stability of the bovine thrombomodulin

200

Conditions

Percent of original activity

1.2-mercaptoethanol 0.1% 1% 2.1% SDS 3.8 M urea 4. 6 M guanidine 5.3 M KSCN 6. p H 2 : 5 m M C a z÷ 5 m M EDTA 7. pH 10:5 m M Ca 2+ 5 m M EDTA 8 . 6 0 ° C , 30 min 9. 1 0 0 ° C

2 0 110 100 40 40 70 80 100 90 100

7"1/2 =

175

150

125

.E_

-

1oo

75

20 min 50

Table I summarizes the procedure for the purification of the thrombomodulin from bovine lung Triton X extract, where the specific activity and the yield are shown as the means of several preparations. From 1 kg of lungs, 1.1 mg thrombomodulin was obtained at 6000-fold purification with a yield of 21%.

Physicochemical characterization of thrombomodulin The following results on characterization were

25

0

1

2

3

Thrombomodulin / Thrombin

4 ( mol / mol )

Fig. 3. Effect of thrombomodulin on thrombin-induced fibrinogen clotting. 100 #1 of a mixture of thrombin (17 nM) and thrombomodulin were added to 100/~1 fibrinogen (12 # M ) to measure the fibrinogen clotting time at 37°C. ©, bovine thrombomodulin; L rabbit thrombomodulin.

349

was analyzed by measuring the cofactor activity of the protein after treatment under the conditions described in Table III. It was stable in SDS and urea, and also resistant to heat treatment, but was readily damaged by reduction with 2-mercaptoethanol.

Functional characterization of thrombomodulin Activation of protein C depended on both thrombomodulin and thrombin. The kinetic studies showed that both proteins form a 1 : 1 complex to activate protein C. The activation rate of protein C by thrombin was increased 880-fold by an excess of thrombomodulin compared with the rate in the absence of thrombomodulin (data not shown). All these properties of bovine thrombomodulin were similar to those of rabbit [9] and human [13] protein. Fig. 2 shows that the activation of protein C by the thrombin-thrombomodulin complex is dependent on Ca 2+. The concentration of Ca 2+ necessary to attain half-maximum velocity was calculated to be 0.2 mM. In the presence of 1 mM EDTA, bovine thrombomodulin did not act as a cofactor.

Rabbit thrombomodulin has been reported to inhibit thrombin-induced fibrin formation and also platelet activation [10,11]. We therefore tested whether bovine thrombomodulin has the same activity on thrombin. As shown in Fig. 3, bovine thrombomodulin prolonged the clotting time in proportion to the increase of the ratio of thrombomodulin to thrombin, as was observed with rabbit thrombomodulin. Fig. 4 shows the effect of bovine thrombomodulin on the thrombin-induced platelet aggregation of human platelet-rich plasma. In the absence of thrombomodulin, thrombin first stimulated the platelet aggregation and then induced plasma clotting. After addition of a 10-times molar

100

0 8o

~ 60

~

2

40

< 20 ¢n (n

1

E (n c 0

I..£: o) ..J

TM

T

i

a

0

0.01

i

0.02 Heparin

1 rain

Fig. 4. Effect of thrombomodulin on thrombin-induced platelet aggregation. To 250/~1 of platelet-rich plasma (4.1-104 cells//~l) were added 150/~1 thrombomodulin (230 nM), and the mixture was incubated at 37°C for 1 min with stirring at 1000 rpm. Then 10 ~l of thrombin (330 nM) were added to initiate platelet aggregation. T, thrombin; TM, thrombomodulin. In the control experiment (in the absence of thrombomodulin, left pattern), 1 and 2 indicate the changes of light transmission corresponding to platelet aggregation and plasma clotting, respectively.

i

0.03

i

0.04

0.05

( U/ml )

Fig. 5. Inhibition of thrombin-thrombomodulin-induced protein C activation by antithrombin III and heparin. To 50/xl of buffer consisting of 0.05 M Tris-HCl/0.1 M NaC1/2 mM CaCI2/0.1% bovine serum albumin (pH 8.5) 10 /.tl protein C (5.3 #M), 10 VI antithrombin III (12.5, 25 and 50 nM), or the buffer and 10 #1 heparin (at varying concentrations) were added. To this mixture 10 p.1 thrombin (25 nM) and 10 #1 thrombomodulin (50 nM) or the buffer were added, and the mixture was incubated at 37°C for 10 min. Protein C activation was stopped by the addition of 25 /~1 of a mixture of antithrombin II1 (1.5 vM) and heparin (16 U/ml). APC, activated protein C. O, without antithrombin III; • antithrombin III (final concentration = 1.25 nM); zx antithrombin III (2.5 n M ) ; • antithrombin III (5.0 nM).

350

100

A

-------0

O

0

80

v

.m > .g 4-,

60

t~ tc~

E

40

o tl--

20

I

i

i

i

i

|

0

0.1

0.2

0.3

0.4

0.5

Heparin

(U/ml)

Fig. 6. Inhibition of amidolytic activity of thrombin-thrombomodulin by antithrombin III and heparin. To 60 /~1 of the buffer noted in Fig. 5, 10 ~1 thrombin (25 nM) and l0 /~1 thrombomodulin (50 nM) or the buffer were added and the mixture was incubated at 37°C for 5 min. Then, 10 #1 of various concentrations of heparin and 10 #1 of antithrombin I11 (12.5, 25, 50 nM) or the buffer were added to the mixture, which was further incubated at 37°C. After 5 rain, 2 ml of the buffer containing 100 /~M Boc-Val-Pro-Arg-MCA was added to determine the residual thrombin amidolytic activity by the initial-rate assay method. (3, without antithrombin III; e, antithrombin III (final concentration =1.25 nM); z~, antithrombin III (2.5 nM); A, antithrombin III (5.0 nM).

excess of thrombomodulin, thrombin no longer induced platelet aggregation or clotted fibrinogen. These results indicate that bovine thrombomodulin inhibits the procoagulant action of thrombin as does rabbit thrombomodulin. The effects of antithrombin III and heparin on the protein C activation by thrombin-thrombomodulin was then studied. As shown in Fig. 5, in the absence of heparin the inhibitory activity of antithrombin III was weak, but in proportion to the increase of heparin the inhibition by antithrombin III was augmented remarkably. This stimulatory effect of heparin was also observed on the inhibition of the amidolytic activity of thrombin-thrombomodulin by antithrombin III (Fig. 6). In addition to these, the effects of the protein C inhibitor and a few synthetic inhibitors on the catalytic and amidolytic acitivity of thrombin in the absence and presence of thrombomodulin were studied (Table IV). The concentration of each inhibitor, except for the protein C inhibitor, that was necessary to obtain a half-maximum velocity (V1/2) was basically the same, regardless of the presence or absence of thrombomodulin. However, the concentration of protein C inhibitor that was required to obtain a half-maximum velocity for protein C activation by thrombin in the presence of thrombomodulin was observed to be smaller than that in the absence of the cofactor protein.

TABLE IV INHIBITION OF THROMBIN ACTIVITY BY SEVERAL INHIBITORS IN THE PRESENCE OR ABSENCE OF THROMBOMODULIN Thrombin amidolytic activity in the presence or absence of thrombomodulin a n d / o r inhibitor was assayed at 37 ° C using 50 and 100 # M Boc-Val-Pro-Arg-MCA in 0.05 M Tris-HCI/0.1 M NaC1/2 mM CaCI2/0.1% bovine serum albumin (pH 8.5). Protein C activation by thrombin in the presence or absence of thrombomodulin a n d / o r inhibitor was assayed at 37°C using 100 #M Boc-Leu-Ser-Thr-Arg-MCA in 0.05 M Tris-HC1/0.1 M NaC1/2 mM CaCI2/0.1% bovine serum albumin (pH 8.5). The activation period by thrombin in the presence and absence of thrombomodulin was 30 min and 2 h, respectively. The concentration of an inhibitor required to obtain a half-maximum velocity for amidolytic activity or protein C activation by thrombin (1:1/2) was determined from Dixon Plot analysis [26] of 1/v vs. the molar concentration of the inhibitor. TM, thrombomodulin. Inhibitors

Antithrombin lit Protein C inhibitor 1-2581 MD-805 Benzamidine chloride

VI/2

for

amidolytic activity (10 -8 M)

V1/2 for protein

C activation (10 -8 M)

- TM

+ TM

- TM

+ TM

0.55 4.6 14 26 3.104

0.55 4.6 17 20 2-104

1.1 3.3 26 27 5.104

1.2 0.59 20 18 2.104

351

Discussion The present study demonstrated that thrombomodulin, a cofactor protein which accelerates the thrombin-catalyzed protein C activation, is also present in bovine lung tissues. By our isolation procedure, thrombomodulin was purified nearly 6000-fold from the detergent extracts of the lung tissue, probably derived from the vascular endothelial cell membrane. Though our procedure is essentially the same as that of Esmon et al. for purification from rabbit lung [9], we added gel filtration as a final step, by which the intact thrombomodulin was separated from the degraded proteins that are probably produced during the isolation procedure. The ratio of the degradation products to the intact protein seemed to be different in each preparation. To protect against degradation, we used several proteinase inhibitors such as p-chloromercuribenzoic acid, aprotinin, pepstatin and ~-amino caproic acid during the extraction of membrane proteins, but without success. Probably, bovine thrombomodulin is more sensitive to tissue proteinases than rabbit or human thrombomodulin. Bovine thrombomodulin was similar to the rabbit thrombomodulin; bovine thrombomodulin can stimulate the thrombin-catalyzed activation of protein C approx. 900-fold. The protein showed a high affinity for thrombin. The apparent M r of the bovine protein was estimated to be 78000 before reduction and 105000 after reduction, as analyzed by SDS-polyacrylamide gel electrophoresis. This mobility of bovine thrombomodulin on electrophoresis is similar to that of rabbit and human protein reported earlier [9,13]. The overall amino-acid composition of bovine thrombomodulin was also similar to that of rabbit and human thrombomodulin. The bovine protein is unusually stable, and in this respect it is similar to rabbit protein, though the bovine protein was rather less stable than the rabbit protein in the presence of 6 M guanidine hydrochloride. The activation of protein C depended on the concentration of both thrombin and thrombomodulin. The concentration of Ca 2÷, required for the activation of protein C, to attain a half-maximum rate for the activation was approx. 0.2 mM. Esmon et al. first showed that thrombomodulin

isolated from rabbit lung inhibited the procoagulant action of thrombin on fibrinogen, factor V and platelets [10,11]. This effect of thrombomodulin on thrombin suggests the importance of thrombomodulin as an intravascular anticoagulant agent as well as a cofactor for protein C activation. Recently, however, Maruyama et al. [24] reported that human thrombomodulin isolated from the placenta and also lung was much less efficient in blocking the procoagulant actions of thrombin than rabbit thrombomodulin. Therefore we examined whether bovine thrombomodulin could inhibit the procoagulant activity of thrombin. The results showed that the bovine protein efficiently blocks the thrombin action on fibrin formation and platelet activation. Bovine thrombomodulin may act as an antiprocoagulant on intravascular wall, like rabbit thrombomodulin. Esmon and Esmon [25] found that rabbit thrombomodulin did not affect the antithrombin III inhibition of thrombin, but they did not show whether the inhibition of thrombin-thrombomodulin complex by antithrombin III is accelerated by heparin. We examined the effects of antithrombin III, heparin and other thrombin inhibitors on the action of thrombin in the presence and absence of thrombomodulin. Bovine thrombomodulin did not affect the antithrombin III inhibition of thrombin, as was observed with rabbit thrombomodulin. We found that heparin facilitates the inhibition of the catalytic activity by antithrombin III as well as the amidolytic activity of thrombin, regardless of the presence or absence of thrombomodulin. Thrombomodulin essentially did not affect the inhibition of thrombin by protein C inhibitor and the synthetic inhibitors. Although the inhibitory effect of the protein C inhibitor towards thrombin-catalyzed activation of protein C appeared to increase in the presence of thrombomodulin, this would be due to the direct inhibition of activated protein C, generated by the thrombin-thrombomodulin complex, by the protein C inhibitor. From these studies, bovine thrombomodulin is concluded to be similar to rabbit rather than to human thrombomodulin.

Acknowledgements This study was supported in part by grants (58570963 and 60571104) from the Scientific Re-

352 s e a r c h F u n d of the M i n i s t r y of E d u c a t i o n , S c i e n c e a n d C u l t u r e , a n d g r a n t s f r o m the R e s e a r c h F o u n d a t i o n for C a n c e r a n d C a r d i o v a s c u l a r D i s ease, T h e R y o i c h i N a i t o F o u n d a t i o n for M e d i c a l Research and The Naito Foundation, Japan.

References 1 Stenflo, J. (1976) J. Biol. Chem. 251, 355-363 2 Kisiel, W., Ericsson, L.H. and Davie, E.W. (1976) Biochemistry 15, 4893-4900 3 Murano, G., Seegers, W.H. and Zolton, R.P. (1974) Thromb. Diath. Haemorrh. 57, 305-313 4 Marlar, R., Kleiss, A. and Griffin J. (1982) Blood 59, 1067-1072 5 Suzuki, K., Stenflo, J., Dahlback, B. and Teodorssen, B. (1983) J. Biol. Chem. 258, 1914-1920 6 Vehar, G.A. and Davie, E.W. (1980) Biochemistry 19, 401-410 7 0 w e n , W.G. and Esmon, C.T. (1981) J. Biol. Chem. 256, 5532-5535 8 Esmon, C.T. and Owen, W.G. (1981) Proc. Natl. Acad. Sci. USA 78, 2249-2252 9 Esmon, N.L., Owen, W.G. and Esmon, C.T. (1982) J. Biol. Chem. 257, 859-864 10 Esmon, C.T., Esmon, N.L. and Harris, K.W. (1982 J. Biol. Chem. 257, 7944-7947 11 Esmon, N.L. Carroll, R.C. and Esmon, C.T. (1983) J. Biol. Chem. 258, 12238-12242 12 Esmon, C.T., Esmon, N.L. and Sangstad, J. (1982) in Pathobiology of the Endothelial Cell (Nossel, H.L. and Vogel, H.J. eds.), pp. 121-136, Academic Press, New York

13 Salem, H.H., Maruyama, I., Ishii, H. and Majerus, P.W. (1984) J. Biol. Chem. 259, 12246-12251 14 Morita, T., Kato, H., Iwanaga, S., Takada, K., Kimura, T. and Sakakibara, S. (1977) J. Biochem. 82, 1495-1498 15 Ohno, Y., Kato, H., Morita, T., Iwanaga, S., Takada, K., Sakakibara, S. and Stenflo, J. (1981) J. Biochem. 90, 1387-1395 16 Okamoto, S., Hijikata, A., Kikumoto, R., Tonomura, S., Hara, H., Ninomiya, K., Maruyama, A., Sugano, M. and Tamo, Y. (1981) Biochem. Biophys. Res. Commun. 101, 440-445 17 Lundblad, R.L., Uhteg, L.C., Vogel, C.N., Kingdon, H.S. and Mann, K.G. (1975) Biochem. Biophys. Res. Commun. 66, 482-489 18 Kurachi, K., Schmer, G., Hermodson, M.A., Teller, D.C. and Davie, E.W. (1976) Biochemistry 15, 368-373 19 Suzuki, K., Nishioka, J. and Hashimoto, S. (1983) J. Biol. Chem. 258, 163-168 20 Fenton, J.W., II, Fasco, K.J., Stackrow, A.B., Aronson, D.L., Young, A.M. and Finlayson, J.S. (1977) J. Biol. Chem. 252, 3587-3598 21 Blobel, G.B. and Dobberstein, B. (1975) J. Cell Biol. 17, 835 851 22 Burnette, W.N. (1981) Anal. Biochem. 112, 195-203 23 Bradford, M. (1976) Anal. Biochem. 72, 248-254 24 Maruyama, I., Salem, H.H., Ishii, H. and Majerus, P.W. (1985) J. Clin. Invest. 75,987-991 25 Esmon, C.T. and Esmon, N.L. (1984) Semin. Thromb. Hemostasis 10, 122-130 26 Dixon, M. (1953) Biochem. J. 55, 170-171