- Email: [email protected]
The effect of human thrombomodulin on the inactivation of thrombin by human antithrombin III
THROMBOSIS RESEARCH 57; 117-126, 1990 0049-3848/90 $3.00 t .OO Printed in the USA. Copyright (c) 1990 Pergamon Press plc. All rights reserved.
THE EFFECTOF HUHANTHRCGHWODULIN ONTHE INACTIVATION OF THROWNBY HUNAN ANTITHRUMBIN III Keizo Hirahara”, l4asayoshi Koya&, Tetsuro Hatsuishie and l!unetsugu Kuratab Pharma Research Laboratories, Hoechst Japan Limiteda l-3-2, Binamidai, Kawagoe, Saitama, 350 Japan and Faculty of Pharmaceutical Sciences, Josai Universityb Keyakidai, Sakado, Saitama, 350-02 Japan (Received 12.6.1989; accepted in revised form 20.9.1989 by Editor S. Okamoto) (Received in final form by Executive Editorial Office 30.10.1989)
ABSTRACT The effect of human thrombomodulin (TH) on the inactivation of thrombin by human antithrombin III (AT1111 was evaluated in comparison with that produced from rabbits. HumanTl did not accelerate the thrombin inhibition by AT111 but rabbit TN enhanced the activity of ATIII. Also inclusion of human Tl at increasing concentration suppressed the throabin inhibitory activity of ATIII. The intensity of AT111 activity in the presence of heparin (O.OlU/ml> was also diminished by the human TM. However, this ATIII- heparin cofactor activity recovered with the addition of a IO-fold amount of heparin (O.lU/ml). In SDS-polyacrylamide gel electrophoresis and immunoblotting analysis, we found a complex formation of AT111 with both human and rabbit Tll (and further confirmed their presence with isoelectrofocusing electrophoresisdata not shown). These results indicate that human TM is substantially different from rabbit TN. Our results suggest that human TN show the crucial role on protein C activation system via thrombin.
INTRODUCTION Antithrombin III (AT1111 is one of the plasma protease activity-inhibiting proteins which participate in the control and regulation of blood coagulation cascades(1). AT111 illsediately inactivates the generated-thrombin in vivo by the interaction with heparin like substances which exist on endothelial cell surfaces (2). while thrombomodulin (RI) also exists on the cell surfaces and immediately binds thrombin with high affinity (Kd: 1 n!l> to form a 1:l molar complex (3). The thrombin-TM complex converts protein C (PC) to the anticoagulant enzyme activated PC (APC) which is then able to inhibit two auxiliary proteins of the coagulation mechanism, factors Va and VIIIa by limiting proteolytic cleavage of these moleKey words : antithrombin III, thrombomodulin, protein C, heparin, thrombin, antithrombin III-thrombomodulin complex 117
AT111 INHIBITION BY TM
118
Vol. 57, No. 1
cules(4,5). Therefore we speculated that, if thrombin generated in the coagulation sites which binds to TM is immediately inhibited by ATIII, protein C can not work as an anticoagulant in vivo. In our previous reports, we found that AT111 easily binds to fatty acids (6,7) and tissue thromboplastin (8,9), then loses the acti vity. These results suggest that human AT111 may easily interact with hydrophobic sites of proteins. Preissner et al (10) reported that rabbit TN accelerated the thrombin inhibitory activity of ATIII, and suggested the existence of a new regulation of AT111 by TM. At the same time, Thompson and Salem (11) reported that human thrombin bound to human TN was protected from inactivation by ATIII. These results concluded that TM influences AT111 itself and thus the reactivity between AT111 and thrombin became poor. It seemed that TMregulates AT111 activity, and thereby efficiently promotes protein C activation in the presence of thrombin. Consequently, in this present study, we investigated the effect of human TMon the interaction between AT111 and thrombin in comparison with that of rabbit TM. MATERIALS ANDMETHODS hterials Materials were purchased from the following suppliers: Tris, glycine, sodium chloride, sulphosalicylic acid, trichloro-acetic acid, ethanol, acetic acid (Wako pure chemical,Tokyo), Bovine serum albumin (Calbiochem, USA), Ammoniumpersulphate, Acrylamide, Bisacrylamide, coomassie brilliant blue R-250, nitrocellulose (BioRad,USA), Triton X-100 (Nakarai chemical,Tokyo), Lubrol PX (Sigma,USA), Rabbit thrombomodul in (American Diagnost ica, USA), DEAB-Sepharose CL-GB, CNBracivated Sepharose 4B (Pharmacia-LXB, Sweden), Silver staining kit (Daiichi Chemical,Tokyo),Purified ATIII, anti-human AT111 antibody (rabbit) and BCP-100 (D-phenyl-propyl-arginyl-5-amino-2-ni trobenzoic acid isopropylamide) of chromogenic substrate for thrombin were provided by Behringwerke AG (FRG). The purity of human AT111 was calculated to be above 99% from its SDS polyacrylamide gel electrophoretic (SDS-PAGE)pattern. Protein concentration was assayed by use of BCA protein assay reagent from Pierce (USA). Human thrombin was prepared according to the method of Xawabata et al (12). Purification
of human TM
We modified the method of Salem et al (13). Three kilograms of frozen placentas were shredded and mixed with 10 volumes of 20 mMTris buffer pH 7.4 containing 50 mMNaCl. The material was then homogenized using a commercial Waring blender and centrifuged at 12,000 g for 45 min. The pellets were suspended in 20 mMTris buffer pH 7.4 containing 50 mNNaCl and 0.5 % Triton X-100 and the suspension was incubated overnight at 4” C with continuous stirring. At the end of this period, the cofactor containing supernatant was collected after centrifugation at 12,000 g for 45 min. The supernatant was applied to a column of DEAE-Sepharose CL6B equilibrated with 20 mMTris buffer pH 7.4 containing 150 mMNaCl and 0.5 % Lubrol PX (starting buffer). The column was washed with the starting buffer and then eluted with NaCl gradient from 150 mMto 1.0 H in 20 mMTris buffer pH 7.4 containing 0.5 % Lubrol PX. Fractions were collected and assayed for protein content and TMactivity. TMcontaining fractions were pooled. The pooled fraction was dialyzed for 8 h against 20mMTris buffer pH 7.4 containing 0.1 M NaCl and 0.5 % Lubrol PX. The material was applied to a DIP-thrombin Sepharose 4B column equilibrated with 20 mNTris buffer pH 7.4 containing 0.1 H NaCl and 0.5 % Lubrol PX. The column was washed with this starting buffer and eluted using NaCl gradient from 0.1 M to 1.0 M in 20 mMTris buffer pH 7.4 containing 0.5 % Lubrol PX. Frac-
Vol. 57, No. 1
AT111 INHIBITIONBY TM
119
tions were collected and tested for TMactivity and protein content. Active fractions were pooled, then rechromatographed using a DIP-thronbin Sepharose 48 column. Finally fractions containing TN activity were dialyzed against 20 mMTris buffer, pH 7.4 containing 150 ml4NaCl and 0.1 % Lubrol PX. Determination
of human TMactivity
TMactivity was assayed according to the method described by Kurosawa and Aoki (14). Both protein C and thrombin used here were of human origin. Our chromogenic substrate for protein C was ChromozymePCs (Behringer Manheim,PRG). Detection
of the complexes between AT111 and TM
A mixture of equal volumes of AT111 (12 pl) and either human TN (2OO~gg/ml, 42 U/ml> or rabbit TM (30fig/ml, 30U/ml) were subjected to SDS-PAGEaccording to the method of Laemmli (15). Four to twenty % gradient gel was used for electrophoresis. After the electrophoresis,the gel was analysed according to the immunoblotting method of Render et al (16) except that Horize Blot (ATTGCo, Japan) was used as the blotting apparatus and that the electrophoretic run was done at a fixed voltage of 11 V for 2 h. Interference
of TMwith the inhibition
of thrombin by AT111
The effect of TMon the inactivation of thrombin by AT111 followed the method of Preissner et al (10). Various concentrations of AT111 were preincubated for 1 min at 37” C in the absence or presence of TM in a total volume of 140~ 1 containing 20 mMTris buffer, pH 7.4, 150 mMNaCl and 0.1 % Lubrol PX. The inhibition reaction was started by the addition of 10,~ 1 of thrombin (28nH). At selected time periods, 100~ 1 portions were removed from the reaction mixture and transferred into a cuvette with 0.25 mMcompound BCP-100 dissolved in 400~ 1 of 50 mM Tris buffer pH 8.3 containing 150 mMNaCl and 0.2 % human serum albumin and kept at 37” C. Residual thrombin amidolytic activity was recorded at 405 nm in a Hitachi spectrophotometer A-220. RESULTS Purification
of human TM
The human TMwas extracted from placentas with Triton X-100 and purified with three column chromatography as described in Materials and Methods.The product obtained by affinity chromatography on DIP-thrombin Sepharose 4B was not homogenous and necessitated a further purification step using the same DIP-thrombin column. Table 1 shows the summary of the human TMpurification procedure.The over all increase in TMactivity was about 2,755 at a recovery of 5 %. Loss of activity was marked in the DEAK-Sepharose chromatography step. After SDS-PAGE,purified TN migrated as a single band of apparent NW.75Kd (Fig.l,lane 1) under non-reducing conditions and MN.100 Kd (Fig. 1 lane 2) under reducing conditions. Effect
of TMon the inhibition
of thrombin by AT111
The effect of increasing human TMconcentration on the activation of thrombin with AT111 is shown in Fig. 2. As we increased the human TMconcentration, a progressive reduction of AT111 activity was observed. We next compared the effect of human TMwith that of rabbit TMon the inactivation of thrombin by ATIII. As
120
AT111 INHIBITION BY TM
Vol. 57, No. 1
TABLEI -_ Summary of Thrombomodulin Purification
from Human Placenta ~-
Step
Protein (mg>
Activity Specific (units>
Triton extract ion
105,000
8,000
0.076
100
1
DEAE-
2,610
928
0.356
11.6
4.7
DIP-thrombin Sepharose
20
448
22.4
5.6
294
DIP-thrombin Sepharose
1.92
403
209.9
5.0
2,755
activity
Yield -Fold (%I
Sepharose
One unit corresponds to the amount of thrombomodulin which produces 1 n!l activated protein Clmin.
MW 200
Kd
97.4 Kd
68 Kd 43 Kd
25.7 Kd I 8.4 Kd
Fig. 1 SDS-polyacrylamide gel electrophoretic pattern of purified human thrombomodulin lane 1: unreduced human thrombomodulin, lane 2: reduced human thromboaodulin, lane 3: molecular size markers. Gel was visialized by silver staining.
o.j
j
121
AT111 INHIBITIONBY TM
Vol. 57, No. 1
.
0
100
200
Human
Thrombomodulin(
ps/ml
)
Fig. 2 Influence of humanthrombomodulinwith the inhibition of thrombin by antithrombin III. One hundred CL1 of a mixtuer of antithrombin III(120 nH>, thrombin(28 nM) and humanthrombomodulin were added to 400 p 1 of chromogenic substrate solution to measurethe residual thrombinactivity. 1.0 -
I
Human
Antithrombin
Ill
1nM
)
Fig. 3 Comparisonbetween humanand rabbit thrombomodulin on the inhibition of antithrombin III. One hundred p 1 of a mixture of humanantithrombin III (30 - 600 I@, and thrombin (28 190 with either humanor rabbit thrombomodulin were addedto 400 CL1 of chromogenic substrate solution to measure the inactivated thrombinactivity. A: rabbit thrombomodulin,~:human thrombomodulin,0: free of thrombomodulin
122
Vol. 57, No. 1
AT111 INHIBITION BY TM
(A>
O1
1
3 Incubation
Time
5 (min
j
(9)
O-
1
5
3 Incubation
Time
( min )
Fig. 4 Effect of human thrombomodulin on the inhibition of thrombin activity by antithrombin III in the presence or absence of heparin. a: thrombin + antithrombin III t thrombomodulin, A: thrombin + antithrombin III + thrombomcdulin + heparin, ??: thrombin + antithrombin III, 0: thrombin + antithrombin III + heparin. One hundred P 1 of a mixture of human thrombomodulin, heparin( A, 0.01 U/ml, B, 0.1 U/ml), sntithrombin III(120 III) and thrombin (28 nM) was added to 400 ~11 of chromogenic substrate to measure the inactivated thrombin.
AT111 INHIBITION BY TM
Vol. 57, No. 1
123
Fig.3 shows, the inhibition of thrombin by AT111 was substantially accelerated by rabbit TM. However, human TMhad no such an activity. The thrombin inhibitory activity of AT111 (1OOn.M)was significantly suppressed by the presence of human TMas compared with those of AT111 alone and rabbit TM (t-test,p
of
complexes between TMand AT111
Whether AT111 and TMhad formed complexes or not was examined by SDS-PAGE and immunobloting using an anti-human AT111 antibody. As Fig. 5 shows, several bands appeared in the presence of either human or rabbit TM. The molecular size of these complexes was not significant between human and rabbit TM (lane 1,3). These complexes were also comfirmed by means of isoelectrofocusing electrophoresis and immunobloting (data not shown). These results indicate that TMs bind to ATIII.
1234
5
Fig. 5 Detection of complexes between thrombomodulin and antithrombin III by using SDS polyacrylamide gel electrophoresis and inununoblotting. lane 1 : human thrombomodulin + antithrombin III, lane 2 : antithrombin III alone,lane 3 : rabbit thrombomodulin + antithrombin III lane 4 : rabbit thrombomodulin, lane 5 : human thrombomodulin
DISCUSSION TM from rabbit lung and human placenta has been previouly shown to inhibit the procoagulant action of thrombin on fibrinogen and platelets (17,18). Also,
124
AT111 INHIBITION BY TM
Vol. 57, No. 1
Preissner et al (10) reported that rabbit TWsubstantially accelerated the inhibition of thrombin by ATIII. However, this function of human TMwas not shown in their report. Therefore,in this present study, we investigated the effect of human TMon the inactivation of thrombin by AT111 in comparison with rabbit TM. Our purified human TM activity was one-sixth that of rabbit TH on the protein C activation (data not shown). This result was corresponded to published report (1s). Thus we used human Tl which was about 6-fold higher than concentration of rabbit TM when comparing the complexes formation with ATIII. As shown in Fig. 1, human TM used here appeared nearly as a single band in SDS-PAGEand silver staining, and the molecular weight corresponded to the result of other reports (13,14). Our data indicates that human Tl4 acts as a weak competitive inhibitor of thrombin inactivation by AT111 but rabbit TH accelerates thorombin inactivation (Fig. 2,3). These results suggest a substantial difference between human and rabbit TM. Therefore, we beleive this difference by in regulation of the coagulation cascade via protein C activation on each animals. Preissner et aI reported that accelerative effect of rabbit TM in the thrombin-AT111 was not mediated by heparin like substances but included an increased sensitivity of the enzyme for inactivation by AT111 ; he suggested the existence of heparin independent mechanism for thrombin inactivation by ATIII. Since, however human TMdid not show such an activity (Pig. 2,3), while AT111 activity is extremely enhanced by the presence of heparin, we examined the combined effect on the inactivation of thrombin by ATIII. As Fig.4 shows, human TH suppressed the ATIII-heparin cofactor activity in the presence of heparin (0.01 U/ml). However the heparin-neutralizing effect of human m was diminished by the presence of high amount of heparin (0.1 U/ml). Next, we looked for the complex between AT111 and either human or rabbit Tl by using SDS-PAGEand immunoblotting finding the results shown in Fig 5. Preissner et al (10) reported that rabbit TH did not bind to ATIII-Se&arose. However, our result clearly indicated the formation of complexes. We beleive this binding of TH to AT111 observed here is an interaction between the hydrophobic moiety of AT111 and TMsince fatty acids (4,5), steroid hormones (20) and tissue thromboplastin (6) combine with ATIII. Since heparin binds to phospholipids (21) and lipoprotein (22,23), it may also bind to human TM. Therefore human TB may react competitively with thrombin, heparin and ATIII. On the other hand, Suzuki et al (24) reported that bovine Tl did not affect the inhibition of thrombin by AT111 with or without heparin. These results suggest that the interaction between human TM, heparin and AT111 is substantially differnt from the cases of rabbit and bovine TM. Consequently, we have to consider this species difference of Tl when we discuss the protein C activation system via TMand thrombin. In conclusion, the trapping or inactivation of AT111 by human TMmay show a crucial role in sufficient formation of thrombinTMcomplexes, since protein C is acivated by thrombin-Td complex. In this report, we argued the interaction between AT111 and human TN. REFFERENCES
1) DAV1E.E.W.and HANAHAN, D.J. ” The plasma proteins Academic Press, New York, 1977, p 421
” ed. by PUTNAH,F. W. Vol Ill
2) HARCUW, J.A., FRITZ&L., GALLI,S. J., KARP,G and ROSENBKRG,R. D. Microvascular heparin like species with anticoagulant activity. Am.J. Physiol. ,%.725-733(1983) 3) ESHON, C.T. and thrombin-catalyzed (1981)
0WKN.W.G. Identification activation of protein C.
of an endothelial cell cofactor for Proc. Nat 1. Acad. Sci. USA,71.2249-2252
Vol. 57, No. 1
AT111 INHIBITION BY TM
125
4) SUZUKI,E., STENFLO, J., DAHLBACK, B. and TKODORSSON, G. Inactivation of human coagulation factor V by activated protein C. J. Biol. Cher ,E, 1914-1920(1983) 5) VBHAR.G.A.and DAV1E.E.W. Preparation and properties (antihaemophil ic factor). Biochemistry, l9,401-4090980)
of bovine factor
VIII
6) H1RAHARA.K.. AND0H.N..HATSUISH1.T.. SUZUK1.N.TERAO,T.and KURATA,H.Interaction between oleic acid and human antithrombin III. Yakugaku Zasshi,lO8,860-866 (1988) N., TER0A.T. and KURATA.H.Effect 7) HIRAHARA, K. , HATSUISHI, T., SUZUKI, kinds of fatty acids on the human antithrombin III activity. -108, 867-871(1988)
of various
8) HIRAHARA, II., HATSUISHI, T., SUZUKI, N. and KURATA,H.Interact ion between human tissue thromboplastin and human antithrombin III. Chem.Pharm.Bull.,(in press) 9) HIRAHARA,K.,ETOH,Y., NATSUISHI,T., SUZUK1.N.and KURATA,H.Synergistic effect of antithrombin III, activated protein C and heparin on the inhibition of the tissue-thromboplastin mediated coagulation. Chem.Pharm. Bull. .37,692-696(1989) 10) PREISSNER. K.T., DELVOS, U. and HULLER-BERGHAUS. G. Binding of thrombin to thrombomodulin accelerates inhibition of the enzyme by antithrombin III. Evidence for a heparin-independent mechanism. Biochemistry,%, 2521-2528(1987) 11) THOHPSON, E. A. and SALEB.H.H. The effect of human thrombomodulin on the Thromb.Hemostas. .g. 806-810 inactivation of thrombin by its serum inhibitors. (1987) 12) KAWABATA,S., HOR1TA.T.. 1WANAGA.S. and properties between thrombin-staphylocoagulase 97,1073-1078(1985)
1GARASHI.H. Difference in enzymatic complex and free thrombin. J.Biochem.,
P. W. Isolation and character13) SALKN,H.H., HARUYANA, I., ISHII, H. and NAJKRUS, ization of thrombomodulin from human placenta. J. Biol. Chem.,259,12246-12251(1984) 14) WROSAWA, S. and AOKI,N. Preparation Thromb.Res. ,37,353-3640985)
of thrombomodul in from human placenta.
15) LAKNNLI,U. K. Cleavage of structual proteins bacteriophage T4. Nature, 2x.680-685 (1970)
during the assembly of the head of
16) RKNART, J., RKISNKR, J. and STARK, G.R. Transfer of proteins from gels to diazobenzyloxymethyl-paper and detection with antisera: A method for studying antibody specificity and antigen structure. Proc. Natl. Acad. Sci. USA,73 3116-3120 (1979) 17) ESHON, C.T., ESN0N.N.L. and HARR1S.K.W.Complex formation between thrombin and thrombomodulin inhibits both thrombin-catalyzing fibrin formation and factor V J. Biol. Chem,,x.7944-7947(1982) activation. _ 18) ESNON, N.L. , CARROLL, R. C. and ESNON, C.T. Thrombomodulin blocks the abi 1i ty of thrombin to activate platelets. J. Biol. Chem.,x12238-12242(1983)
126
AT111 INHIBITION BY TM
Vol. 57, No. 1
19) MARUYAWA I., SALEMH.H., ISHII H. and WAJKRUS P. W. Humanthrombomodulin is not an efficient inhibitor of the procoagulant activity of thrombin. J. Clin. Invest. ,75,987-9910985) 20) NAGASAWA, H., KIM,B. K., STEINER,M. and BAIDIN1.k G. Inhibition of thrombin neutralyzing activity of antithrombin III by steroid hormones. Thromb.Hemostas . , 3 157-161(1982) 21) WLKKR, P. J. and ESI4ON.C.T.Interaction between heparin and factor Ka. Inhibition of prothrombin activation. B&hem. Biophys. Acta,5& 405- 4150979) 22) VIJAYOGOPAL, P., SPINIVASAN, S. R., RADHAKRISH-NAHURTHY, B. and BKRENSKN, G. S. Interaction of serum lipoprotein and a proteoglycan from bovine aorta. J.Biol. Chem.,256,8234-8241(1981) I_23) 1VKRIUS.P.H. The interaction between human plasma lipoproteins tissue glycosaminoglycans. J. Biol. Chem.,x2607-26130972)
and connective
24) SUZUKI,K., KUSUMOTO,H. and HASHIMGT0.S. Isolation and characterization thrombomodulin from bovine lung. Biochem. Biophys. Acta,882,343-3520986)
of
THE EFFECTOF HUHANTHRCGHWODULIN ONTHE INACTIVATION OF THROWNBY HUNAN ANTITHRUMBIN III Keizo Hirahara”, l4asayoshi Koya&, Tetsuro Hatsuishie and l!unetsugu Kuratab Pharma Research Laboratories, Hoechst Japan Limiteda l-3-2, Binamidai, Kawagoe, Saitama, 350 Japan and Faculty of Pharmaceutical Sciences, Josai Universityb Keyakidai, Sakado, Saitama, 350-02 Japan (Received 12.6.1989; accepted in revised form 20.9.1989 by Editor S. Okamoto) (Received in final form by Executive Editorial Office 30.10.1989)
ABSTRACT The effect of human thrombomodulin (TH) on the inactivation of thrombin by human antithrombin III (AT1111 was evaluated in comparison with that produced from rabbits. HumanTl did not accelerate the thrombin inhibition by AT111 but rabbit TN enhanced the activity of ATIII. Also inclusion of human Tl at increasing concentration suppressed the throabin inhibitory activity of ATIII. The intensity of AT111 activity in the presence of heparin (O.OlU/ml> was also diminished by the human TM. However, this ATIII- heparin cofactor activity recovered with the addition of a IO-fold amount of heparin (O.lU/ml). In SDS-polyacrylamide gel electrophoresis and immunoblotting analysis, we found a complex formation of AT111 with both human and rabbit Tll (and further confirmed their presence with isoelectrofocusing electrophoresisdata not shown). These results indicate that human TM is substantially different from rabbit TN. Our results suggest that human TN show the crucial role on protein C activation system via thrombin.
INTRODUCTION Antithrombin III (AT1111 is one of the plasma protease activity-inhibiting proteins which participate in the control and regulation of blood coagulation cascades(1). AT111 illsediately inactivates the generated-thrombin in vivo by the interaction with heparin like substances which exist on endothelial cell surfaces (2). while thrombomodulin (RI) also exists on the cell surfaces and immediately binds thrombin with high affinity (Kd: 1 n!l> to form a 1:l molar complex (3). The thrombin-TM complex converts protein C (PC) to the anticoagulant enzyme activated PC (APC) which is then able to inhibit two auxiliary proteins of the coagulation mechanism, factors Va and VIIIa by limiting proteolytic cleavage of these moleKey words : antithrombin III, thrombomodulin, protein C, heparin, thrombin, antithrombin III-thrombomodulin complex 117
AT111 INHIBITION BY TM
118
Vol. 57, No. 1
cules(4,5). Therefore we speculated that, if thrombin generated in the coagulation sites which binds to TM is immediately inhibited by ATIII, protein C can not work as an anticoagulant in vivo. In our previous reports, we found that AT111 easily binds to fatty acids (6,7) and tissue thromboplastin (8,9), then loses the acti vity. These results suggest that human AT111 may easily interact with hydrophobic sites of proteins. Preissner et al (10) reported that rabbit TN accelerated the thrombin inhibitory activity of ATIII, and suggested the existence of a new regulation of AT111 by TM. At the same time, Thompson and Salem (11) reported that human thrombin bound to human TN was protected from inactivation by ATIII. These results concluded that TM influences AT111 itself and thus the reactivity between AT111 and thrombin became poor. It seemed that TMregulates AT111 activity, and thereby efficiently promotes protein C activation in the presence of thrombin. Consequently, in this present study, we investigated the effect of human TMon the interaction between AT111 and thrombin in comparison with that of rabbit TM. MATERIALS ANDMETHODS hterials Materials were purchased from the following suppliers: Tris, glycine, sodium chloride, sulphosalicylic acid, trichloro-acetic acid, ethanol, acetic acid (Wako pure chemical,Tokyo), Bovine serum albumin (Calbiochem, USA), Ammoniumpersulphate, Acrylamide, Bisacrylamide, coomassie brilliant blue R-250, nitrocellulose (BioRad,USA), Triton X-100 (Nakarai chemical,Tokyo), Lubrol PX (Sigma,USA), Rabbit thrombomodul in (American Diagnost ica, USA), DEAB-Sepharose CL-GB, CNBracivated Sepharose 4B (Pharmacia-LXB, Sweden), Silver staining kit (Daiichi Chemical,Tokyo),Purified ATIII, anti-human AT111 antibody (rabbit) and BCP-100 (D-phenyl-propyl-arginyl-5-amino-2-ni trobenzoic acid isopropylamide) of chromogenic substrate for thrombin were provided by Behringwerke AG (FRG). The purity of human AT111 was calculated to be above 99% from its SDS polyacrylamide gel electrophoretic (SDS-PAGE)pattern. Protein concentration was assayed by use of BCA protein assay reagent from Pierce (USA). Human thrombin was prepared according to the method of Xawabata et al (12). Purification
of human TM
We modified the method of Salem et al (13). Three kilograms of frozen placentas were shredded and mixed with 10 volumes of 20 mMTris buffer pH 7.4 containing 50 mMNaCl. The material was then homogenized using a commercial Waring blender and centrifuged at 12,000 g for 45 min. The pellets were suspended in 20 mMTris buffer pH 7.4 containing 50 mNNaCl and 0.5 % Triton X-100 and the suspension was incubated overnight at 4” C with continuous stirring. At the end of this period, the cofactor containing supernatant was collected after centrifugation at 12,000 g for 45 min. The supernatant was applied to a column of DEAE-Sepharose CL6B equilibrated with 20 mMTris buffer pH 7.4 containing 150 mMNaCl and 0.5 % Lubrol PX (starting buffer). The column was washed with the starting buffer and then eluted with NaCl gradient from 150 mMto 1.0 H in 20 mMTris buffer pH 7.4 containing 0.5 % Lubrol PX. Fractions were collected and assayed for protein content and TMactivity. TMcontaining fractions were pooled. The pooled fraction was dialyzed for 8 h against 20mMTris buffer pH 7.4 containing 0.1 M NaCl and 0.5 % Lubrol PX. The material was applied to a DIP-thrombin Sepharose 4B column equilibrated with 20 mNTris buffer pH 7.4 containing 0.1 H NaCl and 0.5 % Lubrol PX. The column was washed with this starting buffer and eluted using NaCl gradient from 0.1 M to 1.0 M in 20 mMTris buffer pH 7.4 containing 0.5 % Lubrol PX. Frac-
Vol. 57, No. 1
AT111 INHIBITIONBY TM
119
tions were collected and tested for TMactivity and protein content. Active fractions were pooled, then rechromatographed using a DIP-thronbin Sepharose 48 column. Finally fractions containing TN activity were dialyzed against 20 mMTris buffer, pH 7.4 containing 150 ml4NaCl and 0.1 % Lubrol PX. Determination
of human TMactivity
TMactivity was assayed according to the method described by Kurosawa and Aoki (14). Both protein C and thrombin used here were of human origin. Our chromogenic substrate for protein C was ChromozymePCs (Behringer Manheim,PRG). Detection
of the complexes between AT111 and TM
A mixture of equal volumes of AT111 (12 pl) and either human TN (2OO~gg/ml, 42 U/ml> or rabbit TM (30fig/ml, 30U/ml) were subjected to SDS-PAGEaccording to the method of Laemmli (15). Four to twenty % gradient gel was used for electrophoresis. After the electrophoresis,the gel was analysed according to the immunoblotting method of Render et al (16) except that Horize Blot (ATTGCo, Japan) was used as the blotting apparatus and that the electrophoretic run was done at a fixed voltage of 11 V for 2 h. Interference
of TMwith the inhibition
of thrombin by AT111
The effect of TMon the inactivation of thrombin by AT111 followed the method of Preissner et al (10). Various concentrations of AT111 were preincubated for 1 min at 37” C in the absence or presence of TM in a total volume of 140~ 1 containing 20 mMTris buffer, pH 7.4, 150 mMNaCl and 0.1 % Lubrol PX. The inhibition reaction was started by the addition of 10,~ 1 of thrombin (28nH). At selected time periods, 100~ 1 portions were removed from the reaction mixture and transferred into a cuvette with 0.25 mMcompound BCP-100 dissolved in 400~ 1 of 50 mM Tris buffer pH 8.3 containing 150 mMNaCl and 0.2 % human serum albumin and kept at 37” C. Residual thrombin amidolytic activity was recorded at 405 nm in a Hitachi spectrophotometer A-220. RESULTS Purification
of human TM
The human TMwas extracted from placentas with Triton X-100 and purified with three column chromatography as described in Materials and Methods.The product obtained by affinity chromatography on DIP-thrombin Sepharose 4B was not homogenous and necessitated a further purification step using the same DIP-thrombin column. Table 1 shows the summary of the human TMpurification procedure.The over all increase in TMactivity was about 2,755 at a recovery of 5 %. Loss of activity was marked in the DEAK-Sepharose chromatography step. After SDS-PAGE,purified TN migrated as a single band of apparent NW.75Kd (Fig.l,lane 1) under non-reducing conditions and MN.100 Kd (Fig. 1 lane 2) under reducing conditions. Effect
of TMon the inhibition
of thrombin by AT111
The effect of increasing human TMconcentration on the activation of thrombin with AT111 is shown in Fig. 2. As we increased the human TMconcentration, a progressive reduction of AT111 activity was observed. We next compared the effect of human TMwith that of rabbit TMon the inactivation of thrombin by ATIII. As
120
AT111 INHIBITION BY TM
Vol. 57, No. 1
TABLEI -_ Summary of Thrombomodulin Purification
from Human Placenta ~-
Step
Protein (mg>
Activity Specific (units>
Triton extract ion
105,000
8,000
0.076
100
1
DEAE-
2,610
928
0.356
11.6
4.7
DIP-thrombin Sepharose
20
448
22.4
5.6
294
DIP-thrombin Sepharose
1.92
403
209.9
5.0
2,755
activity
Yield -Fold (%I
Sepharose
One unit corresponds to the amount of thrombomodulin which produces 1 n!l activated protein Clmin.
MW 200
Kd
97.4 Kd
68 Kd 43 Kd
25.7 Kd I 8.4 Kd
Fig. 1 SDS-polyacrylamide gel electrophoretic pattern of purified human thrombomodulin lane 1: unreduced human thrombomodulin, lane 2: reduced human thromboaodulin, lane 3: molecular size markers. Gel was visialized by silver staining.
o.j
j
121
AT111 INHIBITIONBY TM
Vol. 57, No. 1
.
0
100
200
Human
Thrombomodulin(
ps/ml
)
Fig. 2 Influence of humanthrombomodulinwith the inhibition of thrombin by antithrombin III. One hundred CL1 of a mixtuer of antithrombin III(120 nH>, thrombin(28 nM) and humanthrombomodulin were added to 400 p 1 of chromogenic substrate solution to measurethe residual thrombinactivity. 1.0 -
I
Human
Antithrombin
Ill
1nM
)
Fig. 3 Comparisonbetween humanand rabbit thrombomodulin on the inhibition of antithrombin III. One hundred p 1 of a mixture of humanantithrombin III (30 - 600 I@, and thrombin (28 190 with either humanor rabbit thrombomodulin were addedto 400 CL1 of chromogenic substrate solution to measure the inactivated thrombinactivity. A: rabbit thrombomodulin,~:human thrombomodulin,0: free of thrombomodulin
122
Vol. 57, No. 1
AT111 INHIBITION BY TM
(A>
O1
1
3 Incubation
Time
5 (min
j
(9)
O-
1
5
3 Incubation
Time
( min )
Fig. 4 Effect of human thrombomodulin on the inhibition of thrombin activity by antithrombin III in the presence or absence of heparin. a: thrombin + antithrombin III t thrombomodulin, A: thrombin + antithrombin III + thrombomcdulin + heparin, ??: thrombin + antithrombin III, 0: thrombin + antithrombin III + heparin. One hundred P 1 of a mixture of human thrombomodulin, heparin( A, 0.01 U/ml, B, 0.1 U/ml), sntithrombin III(120 III) and thrombin (28 nM) was added to 400 ~11 of chromogenic substrate to measure the inactivated thrombin.
AT111 INHIBITION BY TM
Vol. 57, No. 1
123
Fig.3 shows, the inhibition of thrombin by AT111 was substantially accelerated by rabbit TM. However, human TMhad no such an activity. The thrombin inhibitory activity of AT111 (1OOn.M)was significantly suppressed by the presence of human TMas compared with those of AT111 alone and rabbit TM (t-test,p
of
complexes between TMand AT111
Whether AT111 and TMhad formed complexes or not was examined by SDS-PAGE and immunobloting using an anti-human AT111 antibody. As Fig. 5 shows, several bands appeared in the presence of either human or rabbit TM. The molecular size of these complexes was not significant between human and rabbit TM (lane 1,3). These complexes were also comfirmed by means of isoelectrofocusing electrophoresis and immunobloting (data not shown). These results indicate that TMs bind to ATIII.
1234
5
Fig. 5 Detection of complexes between thrombomodulin and antithrombin III by using SDS polyacrylamide gel electrophoresis and inununoblotting. lane 1 : human thrombomodulin + antithrombin III, lane 2 : antithrombin III alone,lane 3 : rabbit thrombomodulin + antithrombin III lane 4 : rabbit thrombomodulin, lane 5 : human thrombomodulin
DISCUSSION TM from rabbit lung and human placenta has been previouly shown to inhibit the procoagulant action of thrombin on fibrinogen and platelets (17,18). Also,
124
AT111 INHIBITION BY TM
Vol. 57, No. 1
Preissner et al (10) reported that rabbit TWsubstantially accelerated the inhibition of thrombin by ATIII. However, this function of human TMwas not shown in their report. Therefore,in this present study, we investigated the effect of human TMon the inactivation of thrombin by AT111 in comparison with rabbit TM. Our purified human TM activity was one-sixth that of rabbit TH on the protein C activation (data not shown). This result was corresponded to published report (1s). Thus we used human Tl which was about 6-fold higher than concentration of rabbit TM when comparing the complexes formation with ATIII. As shown in Fig. 1, human TM used here appeared nearly as a single band in SDS-PAGEand silver staining, and the molecular weight corresponded to the result of other reports (13,14). Our data indicates that human Tl4 acts as a weak competitive inhibitor of thrombin inactivation by AT111 but rabbit TH accelerates thorombin inactivation (Fig. 2,3). These results suggest a substantial difference between human and rabbit TM. Therefore, we beleive this difference by in regulation of the coagulation cascade via protein C activation on each animals. Preissner et aI reported that accelerative effect of rabbit TM in the thrombin-AT111 was not mediated by heparin like substances but included an increased sensitivity of the enzyme for inactivation by AT111 ; he suggested the existence of heparin independent mechanism for thrombin inactivation by ATIII. Since, however human TMdid not show such an activity (Pig. 2,3), while AT111 activity is extremely enhanced by the presence of heparin, we examined the combined effect on the inactivation of thrombin by ATIII. As Fig.4 shows, human TH suppressed the ATIII-heparin cofactor activity in the presence of heparin (0.01 U/ml). However the heparin-neutralizing effect of human m was diminished by the presence of high amount of heparin (0.1 U/ml). Next, we looked for the complex between AT111 and either human or rabbit Tl by using SDS-PAGEand immunoblotting finding the results shown in Fig 5. Preissner et al (10) reported that rabbit TH did not bind to ATIII-Se&arose. However, our result clearly indicated the formation of complexes. We beleive this binding of TH to AT111 observed here is an interaction between the hydrophobic moiety of AT111 and TMsince fatty acids (4,5), steroid hormones (20) and tissue thromboplastin (6) combine with ATIII. Since heparin binds to phospholipids (21) and lipoprotein (22,23), it may also bind to human TM. Therefore human TB may react competitively with thrombin, heparin and ATIII. On the other hand, Suzuki et al (24) reported that bovine Tl did not affect the inhibition of thrombin by AT111 with or without heparin. These results suggest that the interaction between human TM, heparin and AT111 is substantially differnt from the cases of rabbit and bovine TM. Consequently, we have to consider this species difference of Tl when we discuss the protein C activation system via TMand thrombin. In conclusion, the trapping or inactivation of AT111 by human TMmay show a crucial role in sufficient formation of thrombinTMcomplexes, since protein C is acivated by thrombin-Td complex. In this report, we argued the interaction between AT111 and human TN. REFFERENCES
1) DAV1E.E.W.and HANAHAN, D.J. ” The plasma proteins Academic Press, New York, 1977, p 421
” ed. by PUTNAH,F. W. Vol Ill
2) HARCUW, J.A., FRITZ&L., GALLI,S. J., KARP,G and ROSENBKRG,R. D. Microvascular heparin like species with anticoagulant activity. Am.J. Physiol. ,%.725-733(1983) 3) ESHON, C.T. and thrombin-catalyzed (1981)
0WKN.W.G. Identification activation of protein C.
of an endothelial cell cofactor for Proc. Nat 1. Acad. Sci. USA,71.2249-2252
Vol. 57, No. 1
AT111 INHIBITION BY TM
125
4) SUZUKI,E., STENFLO, J., DAHLBACK, B. and TKODORSSON, G. Inactivation of human coagulation factor V by activated protein C. J. Biol. Cher ,E, 1914-1920(1983) 5) VBHAR.G.A.and DAV1E.E.W. Preparation and properties (antihaemophil ic factor). Biochemistry, l9,401-4090980)
of bovine factor
VIII
6) H1RAHARA.K.. AND0H.N..HATSUISH1.T.. SUZUK1.N.TERAO,T.and KURATA,H.Interaction between oleic acid and human antithrombin III. Yakugaku Zasshi,lO8,860-866 (1988) N., TER0A.T. and KURATA.H.Effect 7) HIRAHARA, K. , HATSUISHI, T., SUZUKI, kinds of fatty acids on the human antithrombin III activity. -108, 867-871(1988)
of various
8) HIRAHARA, II., HATSUISHI, T., SUZUKI, N. and KURATA,H.Interact ion between human tissue thromboplastin and human antithrombin III. Chem.Pharm.Bull.,(in press) 9) HIRAHARA,K.,ETOH,Y., NATSUISHI,T., SUZUK1.N.and KURATA,H.Synergistic effect of antithrombin III, activated protein C and heparin on the inhibition of the tissue-thromboplastin mediated coagulation. Chem.Pharm. Bull. .37,692-696(1989) 10) PREISSNER. K.T., DELVOS, U. and HULLER-BERGHAUS. G. Binding of thrombin to thrombomodulin accelerates inhibition of the enzyme by antithrombin III. Evidence for a heparin-independent mechanism. Biochemistry,%, 2521-2528(1987) 11) THOHPSON, E. A. and SALEB.H.H. The effect of human thrombomodulin on the Thromb.Hemostas. .g. 806-810 inactivation of thrombin by its serum inhibitors. (1987) 12) KAWABATA,S., HOR1TA.T.. 1WANAGA.S. and properties between thrombin-staphylocoagulase 97,1073-1078(1985)
1GARASHI.H. Difference in enzymatic complex and free thrombin. J.Biochem.,
P. W. Isolation and character13) SALKN,H.H., HARUYANA, I., ISHII, H. and NAJKRUS, ization of thrombomodulin from human placenta. J. Biol. Chem.,259,12246-12251(1984) 14) WROSAWA, S. and AOKI,N. Preparation Thromb.Res. ,37,353-3640985)
of thrombomodul in from human placenta.
15) LAKNNLI,U. K. Cleavage of structual proteins bacteriophage T4. Nature, 2x.680-685 (1970)
during the assembly of the head of
16) RKNART, J., RKISNKR, J. and STARK, G.R. Transfer of proteins from gels to diazobenzyloxymethyl-paper and detection with antisera: A method for studying antibody specificity and antigen structure. Proc. Natl. Acad. Sci. USA,73 3116-3120 (1979) 17) ESHON, C.T., ESN0N.N.L. and HARR1S.K.W.Complex formation between thrombin and thrombomodulin inhibits both thrombin-catalyzing fibrin formation and factor V J. Biol. Chem,,x.7944-7947(1982) activation. _ 18) ESNON, N.L. , CARROLL, R. C. and ESNON, C.T. Thrombomodulin blocks the abi 1i ty of thrombin to activate platelets. J. Biol. Chem.,x12238-12242(1983)
126
AT111 INHIBITION BY TM
Vol. 57, No. 1
19) MARUYAWA I., SALEMH.H., ISHII H. and WAJKRUS P. W. Humanthrombomodulin is not an efficient inhibitor of the procoagulant activity of thrombin. J. Clin. Invest. ,75,987-9910985) 20) NAGASAWA, H., KIM,B. K., STEINER,M. and BAIDIN1.k G. Inhibition of thrombin neutralyzing activity of antithrombin III by steroid hormones. Thromb.Hemostas . , 3 157-161(1982) 21) WLKKR, P. J. and ESI4ON.C.T.Interaction between heparin and factor Ka. Inhibition of prothrombin activation. B&hem. Biophys. Acta,5& 405- 4150979) 22) VIJAYOGOPAL, P., SPINIVASAN, S. R., RADHAKRISH-NAHURTHY, B. and BKRENSKN, G. S. Interaction of serum lipoprotein and a proteoglycan from bovine aorta. J.Biol. Chem.,256,8234-8241(1981) I_23) 1VKRIUS.P.H. The interaction between human plasma lipoproteins tissue glycosaminoglycans. J. Biol. Chem.,x2607-26130972)
and connective
24) SUZUKI,K., KUSUMOTO,H. and HASHIMGT0.S. Isolation and characterization thrombomodulin from bovine lung. Biochem. Biophys. Acta,882,343-3520986)
of