THROMBOSIS RESEARCH 63; 609-616,199i 0049-3848/91 $3.00 + .OOPrinted in the USA. Copyright (c) 1991 Pergamon Press pk. All rights reserved.
RADIOIMMUNOASSAY
II.
FOR THE DETECTION OF ACTIVE-SITE INHIBITORS IN BIOLOGICAL FLUIDS
HEPARIN
AFFECTS
THE BINDING
Martin Lackmann and Carolyn. L. Geczy. Camperdown, NSW, Australia.
OF HIRlJDIN
Heart Research Institute,
SPECIFIC
TI-IRO~IBIN
TO a-THROMBIN’
145 Missenden
Road,
(Received 7.3.1991; accepted in revised form 12.6.1991 by Editor H.H. Salem)
A B S’I’RACT A radioimmunoassay for recombinant heparin and other glycosaminoglycans thrombin.
Binding of ‘251-r.hirudin
(r.) hirudin was used to study the effect of (GAG’s) on interactions between hirudin and
to thrombin
in the presence or absence of heparin
(and other GAG’s) was monitored by immunoprecipitation of thrombin-“‘1-r.hirudin complexes with a monoclonal antibody to human a-thrombin. Heparin and dextran sulfate substantially reduced hirudin binding to thrombin in human plasma. Inhibition by heparin was restored by addition of increasing amounts of antithrombin (AT) to samples containing constant amounts of heparin, not neutralized by protamine sulfate.
thrombin
At very low heparin
displacement
concentrations
competitive
and ‘751-r.hirudin
and was
of. ‘2’l-r.hil-udin
was
observed, while in the presence of heparin-AT complexes some ‘151-r.hirudin remained bound to thrombin, suggesting that two or more binding sites for hirudin on the thrombin molecule may be occupied simultaneously.
INTRODUCTION Thrombin is one of the key enzymes of the blood coagulation system. The intrinsic and extrinsic pathways converge at thrombin, a serine protease which cleaves fibrinogen leading to the formation of insoluble fibrin clots (Review, 1). The activity of thrombin in plasma is regulated predominantly by antithrombin (AT), a single chain glycoprotein which inhibits several serine proteases involved in the coagulation cascade (Factors Xa.XIa,XIIa and thrombin). Interaction
1. Key words: hirudin, thrombin,
radioimmunoassay, 609
heparin, glucosaminoglycans,
anticoagulants
610
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
between thrombin and AT involves the active-site serine in a-thrombin and an arginlne-serine bond in the active site of AT (2, 3). Binding results in the formation of a 1: 1 stoichiometric cornpIe\ in which the proteolytic domain of thrombin is physically blocked by AT (4). The rate of complex formation is increased up to 2000 fold by heparin, a widely used anticoagulant which binds with high affinity to both thrombin and AT (6,7), causing an increased activity of the protease inhibitor (2, 8). A variety of biological effects of heparin have recently been described (9) and accumulating evidence of side effects of heparin therapy has increased the interest in alternatiw mtico:tgulant substances. Hirudin is the most potent and specific natural inhibitor of thrombin (10). Interaction with thrombin involves the three regions of the protease thought to determine the substrate specificity (11, 12): the active site, the apolar binding site and the anion binding site specific for binding ;f fibrinopeptides (13- 15). Chemical modification of thrombin at its active site serine reduces the affinity of hirudin for thrombin lo3 to lo6 fold (12, 13) emphasizing the importance of this site in hirudin binding. Others have also reported that the acidic C-terminal region of hirudin (i.e. residues 54-65) inhibits cleavage of fibrinogen without interfering with the amidolytic site of thrombin (16-18) and there is good evidence that this region of hirudin binds to a region of ;Ithrombin including residues 62-73 of the B-chain (14, 15, 19,20). We describe the use of a radioimmunoassay (RIA) for hirudin to study the interaction between hirudin and thrombin in plasma, in the presence of various GAG’s, Active-site specific inhibitors of thrombin were captured from solution with thrombin and an anti-thrombin monoclonal antibody (MAb) and quantitation achieved using ‘251-r.hirudin of high specific activity (21). In plasma samples containing heparin or dextran sulfate, substantially reduced hirudin binding to thrombin was observed. We show that even in the presence of excess heparin-AT-complexes some 11’1. r.hirudin thrombin
remains bound to thrombin, suggesting may be occupied simultaneously. MATERIALS
that two or more binding
sites for hirudin
on
AND METHODS
Reagents In addition to the reagents listed in the accompanying publication (21) human AT (40 inhibitor U/mg, Calbiochem), heparin-sodium salt (David Bull Laboratories, Victoria, Australia) and protamine sulfate, chondroitin sulfate (B), dextran sulfate (M.W.:.500,000) and hyaluronic acid from Sigma were used. Methods The methods for radioiodination accompanying article in this issue.
of r.hirudin
and the hirudin RIA are described
in detail in the
RESULTS Plasma components are required for the inhibition of hirudin-thrombin interactions by heparin. Various GAG’s were tested for their effect on the binding of 1251-r.hirudin to thrombin. Heparin, dextran sulfate, chondroitin sulfate and hyaluronic acid at 5 or 10 mg/ml in buffer (TBS + 0.1% OVA) or in 10% titrated plasma were exposed to a-thrombin 1251-r.hirudin (approx. 3pM).
(40.3 PM) in the presence of
‘251-r.hirudin was substantially reduced by 5 Figure 1 shows the quantity of thrombin-bound mg heparin or dextran sulfate in plasma (33% and 38% inhibition respectively, columns 5,8) whereas hyaluronic acid and chondroitin sulfate had only minor effects. Identical results were obtained when 10 mg/ml GAGS were used (not shown). When plasma was omitted, slightly increased 1251-r.hirudin binding occurred in samples containing heparin (8.0-8.6%, column l), hyaluronic acid (4.8-5.9%, column 2) or dextran sulfate (6.7-8.3%, column 4). Chondroitin sulfate weakly inhibited binding when tested either in buffer or plasma (4-5%, columns 3, 7).
Vol. 63, No. 6
HEPARIN-HIRUDIN THROMBIN COMPLEX
611
Protamine sulfate (1.2520 pg/ml) did not abolish the inhibition of hirudin-thrombin complex fomration by heparin even though it neutralized the effect of protamine sulfate on the anticoagulant activity of heparin in plasma when tested in a clotting assay.
I
1
2 samples
5
3
6 samples
in buffer
7
8
in plasma
Figure 1. Effect of glucosaminoglycans on the hirudin-thrombin interaction. Heparin, hyaluronic acid, chondroitin sulfate B and dextran sulfate (Mwt 500.000) were adjusted to 5 mg/ml in buffer (TBS/O.l% OVA/O.l% TritonXlOO) or in 10% plasma. Aliquots of samples (0.1 ml) were incubated in a total volume of 0.4 ml with 1251-r.hirudin (5000 cpm/tube), cc-thrombin (40.3 pM) and the anti-thrombin MAb for 22 h at 4OC. Bound hirudin was determined using the hirudin RIA as outlined in Materials & Methods of the accompanying manuscript. Inhibition of total t251r.hirudin binding of triplicate determinations for each sample is shown: columns l-4 represent samples
in buffer,
pg/ml hyaluronic
columns
5-8 samples
in 10% plasma;
acid, 3 and 7,5 pg/ml chondroitin
1 and 5, 5 pg/rnl heparin,
2 and 6, 5
sulfate B, 4 and 8,5 l_tg/ml dextran sulfate.
Antithrombin restores the inhibitory effect of heparin on the hirudin-thrombin buffer. Increasing amounts of AT (5.4 pM-86.2 nM) incubated with thrombin approx. 3 pM ‘251-r.hirudin in buffer resulted in a proportionally r.hirudin from thrombin (Fig. 2).
increased
interaction in (40.3 pM) and
displacement
of 12?-
Heparin alone and together with low concentrations of AT (5.4-21.6 pM, columns enhanced hirudin-thrombin complex formation by 6.7-2.9%, whereas increasing
l-5) AT
concentrations (43.1 and 862.0 pM) inhibited 12sI-r.hirudin binding from 7.0 to 45.2% (Fig. 2, columns 8-11). At concentrations of AT from 0.862-86.2 nM displacement of hirudin (452 45.9%) reached the levels in plasma samples containing the same amount of heparin (column 11). A twofold increase of heparin (12.5 l_tg/ml) only slightly enhanced displacement of ‘251-r.hirudin (48.1%, column 12). In comparison, displacement of tracer with 5 rig/ml hirudin (0.7 1 nM) in plasma was 56.3% (column 13).
612
HEPARIN-HIRUDIN
THROMBIN
COMPLEX
Vol. 63. No. 6
60 50 40 30 20
6% 10 .e
-5
0
“”
-10
I
1
2
3
4
5
6
samples
7
8
assayed
10
9
11
12
Ii
in RIA
Figure 2. Displacement of thrombin-bound 1251-r.hirudin by heparin-AT complexes. Using a solution of heparin (25 mg/ml) in buffer/O.l% OVA, AT was diluted to concentrations of: 21.6pM (3), 43.lpM (4), 86.2pM (5) 0.172nM (6), 0.345nM (7), 3.448nM (8) 34.483nM (9) and 344.828nM (10). Aliquots of 0.1 ml were assayed in a total of 0.4 ml in the hirudin RIA (see legend to Fig. 1 and Methods)
and compared
to samples of 25 (11) and 50 pg/ml (12) heparin in
10% plasma or in buffer/O.l% OVA (2,l) without AT. The percent inhibition of 1251-r.hirudin binding of triplicate samples (mean and I S.D.) is shown. The competitive displacement with 20 ng/ml r.hirudin in plasma is given for comparison (13).
Excess formation. r.hirudin
heparin
in plasma
Serial dilutions
does not completely
of heparin in plasma
(approx. 3 PM) and thrombin
to values obtained
by displacing
inhibit
1251-r.hirudin-thrombin
were incubated
(40.3 PM). Thrombin-bound
1251-r.hirudin with increasing
with constant
amounts
‘251-r.hirudin
doses of r.hirudin
complex of 1251-
was compared (Fig.3).
In the range of 0.65-X ng/ml heparin displacement of radiolabelled hirudin followed a similar dose-response curve to that observed using r.hirudin in the range 0.16520 ng/ml. Displacement only reached 64.5% (35.5% B/Be) with 2.5 ng/ml heparin and did not increase with lo3 fold higher heparin concentrations. Displacement with 20 rig/ml hirudin was 80.1% (19.9% B/Be). Non-specific binding of tz51-r.hirudin accounted for 8.6% B/Be.
in the RIA (no anti-thrombin
MAb present in the sample)
613
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
1
1
10
100
anticoagulant
1000
10000
100000
added (ng/miTni)
Figure 3. Limiting dilution of heparin in plasma. Serial ng/ml) were prepared in 10% plasma. Aliquots of samples for lT51-r.hirudin binding to thrombin in the hirudin RIA. as percent B/l30 (-+-) and compared to the response (standard curve).
dilutions of heparin (25 pg/ml - 0.156 (0.1 mll) were tested in a total of 0.4 ml The response in the assay is expressed of serial dilutions of r.hirudin (-O-)
DISCUSSION The inhibition of thrombin by antithrombin/heparin or by hirudin involves complex interactions of the participating molecules and the exact mechanism of thrombin inactivation by either anticoagulant remains to be clarified. The possible therapeutic use of hirudin and its synthetic C-terminal fragments as anticoagulants or as adjuncts to heparin therapy have led to increased interest in the mode of action of this thrombin inhibitor. We have used an RIA for hirudin to study some interactions and thrombin.
The high affinity of hirudin for thrombin
between
various anticoagulants
(Kd=2x10Y’3 M) (20) the availability
of
lz51-r.hirudin of high specific radioactivity (21) and the use of a MAb to thrombin (recognising an epitope removed from the active site of the molecule) which does not interfere with the thrombinAT complex (22) enabled us to investigate the interaction between hirudin, thrombin and AT in the presence and absence of various GAGS. Results presented in the accompanying paper indicated inhibitors which bind to the catalytic site of thrombin.
the specificity
of the assay for those
Stone and Hofsteenge (23) suggested that complex formation between hirudin and thrombin proceeds via a two step reaction in which the first rate limiting step is dependent on ionic strength and involves the acidic C-terminus of hirudin and a secondary binding site of thrombin, and is independent of substrate binding to the catalytic site. In agreement with the “docking” function of this initial interaction (20) C-terminal hirudin peptides which block the cleavage of fibrinogen (16, 17, 24,25) by binding to the cationic secondary binding site of thrombin (19, 2S), had little effect
HEPARIN-HIRUDIN
614
THROMBIN
COMPLEX
~YJI.63, No ;:
In this study we show that thrombin-hirudin binding in plasma was substantlall~ reduced ‘j> heparin and dextran sulfate. Although the exact mechanism of hep3rin-enhan~~ti Inhibition 01’ thrombin by AT remains to be clarified, accumulating evidence from kinetic studies suggests rha: by acting as a template, heparin enhances the ability of both thrombin and AT to form complesex (5, 6, 26). Our results suggest that dextran sulfate may act in a similar mannc‘I- to heparin ;I\ previously reported (7, -38) and would indicate that highly sulfated GAGS may ser\e this function, since hyaluronic acid and chondroitin sulfate only weakly inhibited hirudin binding. The direct effect of dextran sulfate and other GAGS with varying molecular weights and sulfate content on the interactions between hirudin, AT and thrombin warrant further investigation. When plasma wah omitted from the test system, heparin, dextran sulfate and hyaluronic acid consistently incrcasetl thrombin bound lZ51-r.hirudin by about 6-X% (Fig. 1). Increased reactivity of thrombin anti hirudin may occur in the presence of these polyanions since ionic interactlorls may play ;III important role in the formation of the hirudin-thrombin complex (20. 23). Figure 2 suggests that decreased hirudin-thrombin complex formation in heparinised plasma was primarily due to AT binding to the active site of thrombin. AT restored the inhibitory effect of heparin in the plasma-free buffer system; increasing AT concentrations (5.4 pM-86.2 nM) resulted in proportionally increased ‘251-r.hirudin displacement to levels achieved in plasma (columtl 1 1, Fig. 2). At least equimolar AT concentrations (in relation to the thrombin concentration of 40.3 pM) were required to produce a “competitive effect” with hirudin (column 6. Fig. 3). Protamine sulfate reduced the thrombin time of heparinised plasma but did not abolish the effect of heparin on the hirudin-thrombin interaction confirming results of Rosenber, (1 and Damus who demonstrated that protamine did not release thrombin from the thrombin-AT-heparin complex (2). Titration of heparin in the presence of constant concentrations of AT, thrombin and “‘Ir.hirudin produced a sigmoidal dose-response curve similar to the standard curve obtained from competitive displacement of tracer with r.hirudin, as shown in the same figure (Fig. 3). This finding is in apparent contradiction to the observation that only catalytic amounts of heparin are necessary to accelerate the formation of non-dissociable stoichiometric complexes between AT and thrombin (2) in which case a steep, non-competitive dose-response curve should result. However, both hirudin and varying amounts of heparin-AT complexes were present in the plasma sample when thrombin was added to the assay (see legend, Fig. 3) and both compete for the same binding site on the effector molecule. Due to its very low dissociation constant (2~10Y’~ M), lZ51-r.hirudin will remain bound to thrombin at the ratio defined by the initial concentrations of hirudin, heparinbound AT and thrombin. Since the concentrations of hirudin (limiting), thrombin (limiting) and AT (saturating) were constant, this ratio was determined by the amount of AT complexed to heparin at a given concentration of this GAG at the beginning of the incubation. The plateau of the dose-response curve observed at very low heparin concentrations (i.e. 0.16-0.64 ng/ml) may indicate reduced rates of heparin-AT complex formation (5, 26), resulting in preferred binding of lZ51-r.hirudin
under these conditions.
In the range 0.64-25 ng/ml heparin,
increasing
numbers
of
AT-heparin complexes may compete with 12”1-r.hirudin for the binding site on thrombin to yield the approximately linear portion of the dose-response curve. Heparin concentrations above 25 ng/ml (up to 25 pg/ml)
had no further effect on hirudin
binding.
Because
approximately
27%
lZ51-r.hirudin remained bound to thrombin under conditions when heparin and AT were both present in excess, we suggest that hirudin binds to thrombin even when the active site of the enzyme is blocked by AT. Recent reports on the crystallographic structure of a r.hirudin-thrombin complex demonstrate, that apart from strong interactions of the three N-terminal amino acids of hirudin with the catalytic site of thrombin, numerous electrostatic and hydrophobic interactions of the COOH-terminal domain of hirudin with the anion binding site of thrombin are evident (14, 15). These results confirm earlier models based on kinetic studies of the hirudin-thrombin interactions (12, 19, 20,23,27) and may explain our observations.
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
The RIA described here, and designed to quantitate hirudin, interactions between thrombin and its acive-site specific inhibitors. Acknowledgements: Lum for invaluable
615
may also be used to study
We would wish to thank Ms. Brigid Phelan , Mrs. Olga Lutak and Nls. Tina secretarial assistance in the preparation of this manuscript. REFERENCES and NEMERSON,
JACKSON, CM., 765-811, 1980.
3 -.
ROSENBERG, R.D., and DAMUS, P.S. The. purification and mechanism of action of human antithrombin-heparin cofactor. J. Bid. Chem, 248, 6490-6505, 1973.
3.
BJORK, I., JACKSON, CM., JORNVALL, H., LAVINE, K.K., NORDLING, K., and SALSGIVER, W.J. The active site of antithrombin. Release of the same proteolytically cleared form of the inhibitor from complexes with factor IXa, factor Xa, and thrombin. .I. Biol. Cl~ern, 257, 2406-2311, 1982.
4.
TRAVIS, Biochem,
5.
GRIFFITH, M.J. The heparin enhanced antithrombin IIVthrombin reaction is saturable with respect to both thrombin and antithrombin III. .I.Biol. Chem, 257, 13899-13902, 1982.
6
OLSON, S.T. Transient kinetics of heparin-catalyzed protease inactivation by antithrombin III. Linkage of protease-inhibitor-heparin interactions in the reaction with thrombin. J. Biol. Chem, 263, 1698-1708. 1988.
7.
DAWES, J. Measurement of the affinities of heparin, naturally occurring glucosaminoglycons and other sulfated polymers for antithrombin III and thrombin. .4tlal. Biochem, 174, 177-186, 1988.
x.
BJORK, I., and LINDAHL, U. Mechanism Cell. Biocllem, 48, 161-l 82, 1982.
9.
JAQUES, 1980.
10.
MARKWARDT, F., and WALSMANN, P. Reindarstellung und Analyse des Thrombin inhibitors Hirudin. Hope-Seyler’s Z. Physiol. Chem, 338. 1381-1386, 1967.
11.
MARK WARDT, F. Pharmacology of hirudin: one hundred years after the first report of the anticoagulant agent in medicinal leeches. Biomcd., Bioc/lir?l. Actn, 44, 1007-1013, 1983.
12.
STONE, thrombin
1Ii.
FENTON,
14.
RYDEL, T.J., RAVICHANDRAN, K.G., TULINSKY, A., BODE, W., HUBER, R., ROITSCH, C., and FENTON, J.W. II. The structure of II complex of recombinant hirudin and human alpha strombin. Science, 249, 277-280, 1990.
15.
GRUTTER, M.G., PRIESTEL, J.P., REHUEL, J., GROSSENBACHER, H., BODE, W., HOFSTEENGE, J., and STONE, S.R. Crystal structure of the thrombin-hirudin comples: a novel mode of serine protease inhibition. EMBO J, 9,2361-2365, 1990.
J., and SALVESEN, 52, 655-709. 1983.
L.B.
Heparin-anionic
G.S.
Y.
Blood Coagulation.
Ann. Rev. Biochern,
1.
Human
plasma
proteinase
of the anticoagulant
polyelectrolyte
drugs.
inhibitors.
.4nnir. Rev.
action of heparin.
Pharmacol.
49,
Mol.
Rev, 31. 99- 166,
S.R., BRAUN, P.J., and HOFSTEENGE, J. Identification of regions of CIinvolved in its interaction with hirudin. Biochemistry, 26, 4617-4624. 1987. J.W. II. Thrombin
specificity.
Ann. N.Y. Acod. Sci, 370, 468-495.
1981.
616
HEPARIN-HIRUDIN
THROMBIN
Vol. 63, No 6
COMPLEX
IMAO, S.J.‘f., YATES, M.l‘.. OWEN, ‘I .J.. and KRSTENAYSKY, .I.[.. intcractiol.1 I)I hirudin with thrombin indentification of a minimal binding domain of hirudir rhat inhibit\ clotting activitv. Wii,c,/itr,rzi.\t,_\., 27. X 170 ‘i! 7 1, IYXX. 17.
MARAGANORE. J.M.. CIlAO. B,, RAMACIHANDRAN, K.L. Anticoagulant Cllc/n, 264, 11729-l 173.5, 19X9.
18
JAKUBOWSKY, J.A., and MARAGANORE, thrombin-induced platelet activities by a synthetic terminus of hirudin. Blood, 75. 399-406, 1990.
19.
NOE, G., HOFSTEENGE, J., ROVELL,I, specific antibodies to identify a secondary 11729-l 171.5, 1988.
20.
STONE,
JOSEPIH, M.I. JABLONSKY J. .in(l activity of synthetic hirudin peptitles. 1 /3i~~~
J.M. Inhibition of coagulation and dodecupeptide modeled on the carbosy-
G., and STONE, S.R. The use of sequencebinding site in thrombin. .I. Biol. Chem, 263,
S.R., and HOFSTEENGE, J. Kinetics 2.5, 4622-4628, 1986.
of the inhibition
of thrombin
by IHirudin.
Biochemistry, 21.
LACKMANN, M., HOAD, R., KAKAKIOS, A., and GECZY, C.L. Radioimmunoassay for the detection of assay characteristics and quantitation of recombinant hirudin. this issue, 1991.
33
STONE, S.R., DENNIS, S., and HOFSTEENGE, contribution of ionic interactions to the formation Biochemistry. 28, 6X57-6861, 1989.
23.
DAWES,
J., JAMES,
Monoclonnl antibodies III complex. Thromh. 23
‘5.
K., MICKLEM,
L.R.,
J. Quantitative evaluation of the of the thrombin-hirudin complex.
PEPPER,
directed against human a-thrombin Res, 36, 397-409, 1984.
D.S.,
and
PROWSE,
and the thrombin-anti
C.V.
thrombin
KRSTENANSKY, J.L., OWEN, T.A., YATES, M.T., and MAO, S.J.T. Anticoagulant peptides: nature of the interaction of the c-terminal region of hirudin with a noncatalytic binding sites on thrombin. .I. Med. Chern, 30, 16X8-1691, 1987. FENTON,
J.W. II, OWEN,
exosite of human cn-thrombin 1988.
T.J., ZABINSKI,
M.P., and WII,NER,
and fibrin (ogen) recognition.
G.D.
Anion-binding 27, 7106-71 12.
Bioc/wni.stry,
26.
NESHEIM, M.E. A simple rate law that describes the kinetics of the heparin-catalysed reaction between Antithrombin III and thrombin. ./. Biol. C/WUL, 264, 1470X- 147 17. 19X?.
27.
CHANG, J.Y. Binding of heparin to human 2ntithrombin III activates modification at lysine 236. ./. Biol. Chern, 264, 311 l-31 15, 19X9.
28.
PRATT, C.W., WHINNA, H.C., MEADE, J.B., TREANOR, CHURCH, T.C. Physicochemical aspects of heparin cofactor Acad. Sci, 556, 104-115, 1989.
selective
R.E., and 11. Am. N.1
chemical
RADIOIMMUNOASSAY
II.
FOR THE DETECTION OF ACTIVE-SITE INHIBITORS IN BIOLOGICAL FLUIDS
HEPARIN
AFFECTS
THE BINDING
Martin Lackmann and Carolyn. L. Geczy. Camperdown, NSW, Australia.
OF HIRlJDIN
Heart Research Institute,
SPECIFIC
TI-IRO~IBIN
TO a-THROMBIN’
145 Missenden
Road,
(Received 7.3.1991; accepted in revised form 12.6.1991 by Editor H.H. Salem)
A B S’I’RACT A radioimmunoassay for recombinant heparin and other glycosaminoglycans thrombin.
Binding of ‘251-r.hirudin
(r.) hirudin was used to study the effect of (GAG’s) on interactions between hirudin and
to thrombin
in the presence or absence of heparin
(and other GAG’s) was monitored by immunoprecipitation of thrombin-“‘1-r.hirudin complexes with a monoclonal antibody to human a-thrombin. Heparin and dextran sulfate substantially reduced hirudin binding to thrombin in human plasma. Inhibition by heparin was restored by addition of increasing amounts of antithrombin (AT) to samples containing constant amounts of heparin, not neutralized by protamine sulfate.
thrombin
At very low heparin
displacement
concentrations
competitive
and ‘751-r.hirudin
and was
of. ‘2’l-r.hil-udin
was
observed, while in the presence of heparin-AT complexes some ‘151-r.hirudin remained bound to thrombin, suggesting that two or more binding sites for hirudin on the thrombin molecule may be occupied simultaneously.
INTRODUCTION Thrombin is one of the key enzymes of the blood coagulation system. The intrinsic and extrinsic pathways converge at thrombin, a serine protease which cleaves fibrinogen leading to the formation of insoluble fibrin clots (Review, 1). The activity of thrombin in plasma is regulated predominantly by antithrombin (AT), a single chain glycoprotein which inhibits several serine proteases involved in the coagulation cascade (Factors Xa.XIa,XIIa and thrombin). Interaction
1. Key words: hirudin, thrombin,
radioimmunoassay, 609
heparin, glucosaminoglycans,
anticoagulants
610
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
between thrombin and AT involves the active-site serine in a-thrombin and an arginlne-serine bond in the active site of AT (2, 3). Binding results in the formation of a 1: 1 stoichiometric cornpIe\ in which the proteolytic domain of thrombin is physically blocked by AT (4). The rate of complex formation is increased up to 2000 fold by heparin, a widely used anticoagulant which binds with high affinity to both thrombin and AT (6,7), causing an increased activity of the protease inhibitor (2, 8). A variety of biological effects of heparin have recently been described (9) and accumulating evidence of side effects of heparin therapy has increased the interest in alternatiw mtico:tgulant substances. Hirudin is the most potent and specific natural inhibitor of thrombin (10). Interaction with thrombin involves the three regions of the protease thought to determine the substrate specificity (11, 12): the active site, the apolar binding site and the anion binding site specific for binding ;f fibrinopeptides (13- 15). Chemical modification of thrombin at its active site serine reduces the affinity of hirudin for thrombin lo3 to lo6 fold (12, 13) emphasizing the importance of this site in hirudin binding. Others have also reported that the acidic C-terminal region of hirudin (i.e. residues 54-65) inhibits cleavage of fibrinogen without interfering with the amidolytic site of thrombin (16-18) and there is good evidence that this region of hirudin binds to a region of ;Ithrombin including residues 62-73 of the B-chain (14, 15, 19,20). We describe the use of a radioimmunoassay (RIA) for hirudin to study the interaction between hirudin and thrombin in plasma, in the presence of various GAG’s, Active-site specific inhibitors of thrombin were captured from solution with thrombin and an anti-thrombin monoclonal antibody (MAb) and quantitation achieved using ‘251-r.hirudin of high specific activity (21). In plasma samples containing heparin or dextran sulfate, substantially reduced hirudin binding to thrombin was observed. We show that even in the presence of excess heparin-AT-complexes some 11’1. r.hirudin thrombin
remains bound to thrombin, suggesting may be occupied simultaneously. MATERIALS
that two or more binding
sites for hirudin
on
AND METHODS
Reagents In addition to the reagents listed in the accompanying publication (21) human AT (40 inhibitor U/mg, Calbiochem), heparin-sodium salt (David Bull Laboratories, Victoria, Australia) and protamine sulfate, chondroitin sulfate (B), dextran sulfate (M.W.:.500,000) and hyaluronic acid from Sigma were used. Methods The methods for radioiodination accompanying article in this issue.
of r.hirudin
and the hirudin RIA are described
in detail in the
RESULTS Plasma components are required for the inhibition of hirudin-thrombin interactions by heparin. Various GAG’s were tested for their effect on the binding of 1251-r.hirudin to thrombin. Heparin, dextran sulfate, chondroitin sulfate and hyaluronic acid at 5 or 10 mg/ml in buffer (TBS + 0.1% OVA) or in 10% titrated plasma were exposed to a-thrombin 1251-r.hirudin (approx. 3pM).
(40.3 PM) in the presence of
‘251-r.hirudin was substantially reduced by 5 Figure 1 shows the quantity of thrombin-bound mg heparin or dextran sulfate in plasma (33% and 38% inhibition respectively, columns 5,8) whereas hyaluronic acid and chondroitin sulfate had only minor effects. Identical results were obtained when 10 mg/ml GAGS were used (not shown). When plasma was omitted, slightly increased 1251-r.hirudin binding occurred in samples containing heparin (8.0-8.6%, column l), hyaluronic acid (4.8-5.9%, column 2) or dextran sulfate (6.7-8.3%, column 4). Chondroitin sulfate weakly inhibited binding when tested either in buffer or plasma (4-5%, columns 3, 7).
Vol. 63, No. 6
HEPARIN-HIRUDIN THROMBIN COMPLEX
611
Protamine sulfate (1.2520 pg/ml) did not abolish the inhibition of hirudin-thrombin complex fomration by heparin even though it neutralized the effect of protamine sulfate on the anticoagulant activity of heparin in plasma when tested in a clotting assay.
I
1
2 samples
5
3
6 samples
in buffer
7
8
in plasma
Figure 1. Effect of glucosaminoglycans on the hirudin-thrombin interaction. Heparin, hyaluronic acid, chondroitin sulfate B and dextran sulfate (Mwt 500.000) were adjusted to 5 mg/ml in buffer (TBS/O.l% OVA/O.l% TritonXlOO) or in 10% plasma. Aliquots of samples (0.1 ml) were incubated in a total volume of 0.4 ml with 1251-r.hirudin (5000 cpm/tube), cc-thrombin (40.3 pM) and the anti-thrombin MAb for 22 h at 4OC. Bound hirudin was determined using the hirudin RIA as outlined in Materials & Methods of the accompanying manuscript. Inhibition of total t251r.hirudin binding of triplicate determinations for each sample is shown: columns l-4 represent samples
in buffer,
pg/ml hyaluronic
columns
5-8 samples
in 10% plasma;
acid, 3 and 7,5 pg/ml chondroitin
1 and 5, 5 pg/rnl heparin,
2 and 6, 5
sulfate B, 4 and 8,5 l_tg/ml dextran sulfate.
Antithrombin restores the inhibitory effect of heparin on the hirudin-thrombin buffer. Increasing amounts of AT (5.4 pM-86.2 nM) incubated with thrombin approx. 3 pM ‘251-r.hirudin in buffer resulted in a proportionally r.hirudin from thrombin (Fig. 2).
increased
interaction in (40.3 pM) and
displacement
of 12?-
Heparin alone and together with low concentrations of AT (5.4-21.6 pM, columns enhanced hirudin-thrombin complex formation by 6.7-2.9%, whereas increasing
l-5) AT
concentrations (43.1 and 862.0 pM) inhibited 12sI-r.hirudin binding from 7.0 to 45.2% (Fig. 2, columns 8-11). At concentrations of AT from 0.862-86.2 nM displacement of hirudin (452 45.9%) reached the levels in plasma samples containing the same amount of heparin (column 11). A twofold increase of heparin (12.5 l_tg/ml) only slightly enhanced displacement of ‘251-r.hirudin (48.1%, column 12). In comparison, displacement of tracer with 5 rig/ml hirudin (0.7 1 nM) in plasma was 56.3% (column 13).
612
HEPARIN-HIRUDIN
THROMBIN
COMPLEX
Vol. 63. No. 6
60 50 40 30 20
6% 10 .e
-5
0
“”
-10
I
1
2
3
4
5
6
samples
7
8
assayed
10
9
11
12
Ii
in RIA
Figure 2. Displacement of thrombin-bound 1251-r.hirudin by heparin-AT complexes. Using a solution of heparin (25 mg/ml) in buffer/O.l% OVA, AT was diluted to concentrations of: 21.6pM (3), 43.lpM (4), 86.2pM (5) 0.172nM (6), 0.345nM (7), 3.448nM (8) 34.483nM (9) and 344.828nM (10). Aliquots of 0.1 ml were assayed in a total of 0.4 ml in the hirudin RIA (see legend to Fig. 1 and Methods)
and compared
to samples of 25 (11) and 50 pg/ml (12) heparin in
10% plasma or in buffer/O.l% OVA (2,l) without AT. The percent inhibition of 1251-r.hirudin binding of triplicate samples (mean and I S.D.) is shown. The competitive displacement with 20 ng/ml r.hirudin in plasma is given for comparison (13).
Excess formation. r.hirudin
heparin
in plasma
Serial dilutions
does not completely
of heparin in plasma
(approx. 3 PM) and thrombin
to values obtained
by displacing
inhibit
1251-r.hirudin-thrombin
were incubated
(40.3 PM). Thrombin-bound
1251-r.hirudin with increasing
with constant
amounts
‘251-r.hirudin
doses of r.hirudin
complex of 1251-
was compared (Fig.3).
In the range of 0.65-X ng/ml heparin displacement of radiolabelled hirudin followed a similar dose-response curve to that observed using r.hirudin in the range 0.16520 ng/ml. Displacement only reached 64.5% (35.5% B/Be) with 2.5 ng/ml heparin and did not increase with lo3 fold higher heparin concentrations. Displacement with 20 rig/ml hirudin was 80.1% (19.9% B/Be). Non-specific binding of tz51-r.hirudin accounted for 8.6% B/Be.
in the RIA (no anti-thrombin
MAb present in the sample)
613
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
1
1
10
100
anticoagulant
1000
10000
100000
added (ng/miTni)
Figure 3. Limiting dilution of heparin in plasma. Serial ng/ml) were prepared in 10% plasma. Aliquots of samples for lT51-r.hirudin binding to thrombin in the hirudin RIA. as percent B/l30 (-+-) and compared to the response (standard curve).
dilutions of heparin (25 pg/ml - 0.156 (0.1 mll) were tested in a total of 0.4 ml The response in the assay is expressed of serial dilutions of r.hirudin (-O-)
DISCUSSION The inhibition of thrombin by antithrombin/heparin or by hirudin involves complex interactions of the participating molecules and the exact mechanism of thrombin inactivation by either anticoagulant remains to be clarified. The possible therapeutic use of hirudin and its synthetic C-terminal fragments as anticoagulants or as adjuncts to heparin therapy have led to increased interest in the mode of action of this thrombin inhibitor. We have used an RIA for hirudin to study some interactions and thrombin.
The high affinity of hirudin for thrombin
between
various anticoagulants
(Kd=2x10Y’3 M) (20) the availability
of
lz51-r.hirudin of high specific radioactivity (21) and the use of a MAb to thrombin (recognising an epitope removed from the active site of the molecule) which does not interfere with the thrombinAT complex (22) enabled us to investigate the interaction between hirudin, thrombin and AT in the presence and absence of various GAGS. Results presented in the accompanying paper indicated inhibitors which bind to the catalytic site of thrombin.
the specificity
of the assay for those
Stone and Hofsteenge (23) suggested that complex formation between hirudin and thrombin proceeds via a two step reaction in which the first rate limiting step is dependent on ionic strength and involves the acidic C-terminus of hirudin and a secondary binding site of thrombin, and is independent of substrate binding to the catalytic site. In agreement with the “docking” function of this initial interaction (20) C-terminal hirudin peptides which block the cleavage of fibrinogen (16, 17, 24,25) by binding to the cationic secondary binding site of thrombin (19, 2S), had little effect
HEPARIN-HIRUDIN
614
THROMBIN
COMPLEX
~YJI.63, No ;:
In this study we show that thrombin-hirudin binding in plasma was substantlall~ reduced ‘j> heparin and dextran sulfate. Although the exact mechanism of hep3rin-enhan~~ti Inhibition 01’ thrombin by AT remains to be clarified, accumulating evidence from kinetic studies suggests rha: by acting as a template, heparin enhances the ability of both thrombin and AT to form complesex (5, 6, 26). Our results suggest that dextran sulfate may act in a similar mannc‘I- to heparin ;I\ previously reported (7, -38) and would indicate that highly sulfated GAGS may ser\e this function, since hyaluronic acid and chondroitin sulfate only weakly inhibited hirudin binding. The direct effect of dextran sulfate and other GAGS with varying molecular weights and sulfate content on the interactions between hirudin, AT and thrombin warrant further investigation. When plasma wah omitted from the test system, heparin, dextran sulfate and hyaluronic acid consistently incrcasetl thrombin bound lZ51-r.hirudin by about 6-X% (Fig. 1). Increased reactivity of thrombin anti hirudin may occur in the presence of these polyanions since ionic interactlorls may play ;III important role in the formation of the hirudin-thrombin complex (20. 23). Figure 2 suggests that decreased hirudin-thrombin complex formation in heparinised plasma was primarily due to AT binding to the active site of thrombin. AT restored the inhibitory effect of heparin in the plasma-free buffer system; increasing AT concentrations (5.4 pM-86.2 nM) resulted in proportionally increased ‘251-r.hirudin displacement to levels achieved in plasma (columtl 1 1, Fig. 2). At least equimolar AT concentrations (in relation to the thrombin concentration of 40.3 pM) were required to produce a “competitive effect” with hirudin (column 6. Fig. 3). Protamine sulfate reduced the thrombin time of heparinised plasma but did not abolish the effect of heparin on the hirudin-thrombin interaction confirming results of Rosenber, (1 and Damus who demonstrated that protamine did not release thrombin from the thrombin-AT-heparin complex (2). Titration of heparin in the presence of constant concentrations of AT, thrombin and “‘Ir.hirudin produced a sigmoidal dose-response curve similar to the standard curve obtained from competitive displacement of tracer with r.hirudin, as shown in the same figure (Fig. 3). This finding is in apparent contradiction to the observation that only catalytic amounts of heparin are necessary to accelerate the formation of non-dissociable stoichiometric complexes between AT and thrombin (2) in which case a steep, non-competitive dose-response curve should result. However, both hirudin and varying amounts of heparin-AT complexes were present in the plasma sample when thrombin was added to the assay (see legend, Fig. 3) and both compete for the same binding site on the effector molecule. Due to its very low dissociation constant (2~10Y’~ M), lZ51-r.hirudin will remain bound to thrombin at the ratio defined by the initial concentrations of hirudin, heparinbound AT and thrombin. Since the concentrations of hirudin (limiting), thrombin (limiting) and AT (saturating) were constant, this ratio was determined by the amount of AT complexed to heparin at a given concentration of this GAG at the beginning of the incubation. The plateau of the dose-response curve observed at very low heparin concentrations (i.e. 0.16-0.64 ng/ml) may indicate reduced rates of heparin-AT complex formation (5, 26), resulting in preferred binding of lZ51-r.hirudin
under these conditions.
In the range 0.64-25 ng/ml heparin,
increasing
numbers
of
AT-heparin complexes may compete with 12”1-r.hirudin for the binding site on thrombin to yield the approximately linear portion of the dose-response curve. Heparin concentrations above 25 ng/ml (up to 25 pg/ml)
had no further effect on hirudin
binding.
Because
approximately
27%
lZ51-r.hirudin remained bound to thrombin under conditions when heparin and AT were both present in excess, we suggest that hirudin binds to thrombin even when the active site of the enzyme is blocked by AT. Recent reports on the crystallographic structure of a r.hirudin-thrombin complex demonstrate, that apart from strong interactions of the three N-terminal amino acids of hirudin with the catalytic site of thrombin, numerous electrostatic and hydrophobic interactions of the COOH-terminal domain of hirudin with the anion binding site of thrombin are evident (14, 15). These results confirm earlier models based on kinetic studies of the hirudin-thrombin interactions (12, 19, 20,23,27) and may explain our observations.
HEPARIN-HIRUDIN THROMBIN COMPLEX
Vol. 63, No. 6
The RIA described here, and designed to quantitate hirudin, interactions between thrombin and its acive-site specific inhibitors. Acknowledgements: Lum for invaluable
615
may also be used to study
We would wish to thank Ms. Brigid Phelan , Mrs. Olga Lutak and Nls. Tina secretarial assistance in the preparation of this manuscript. REFERENCES and NEMERSON,
JACKSON, CM., 765-811, 1980.
3 -.
ROSENBERG, R.D., and DAMUS, P.S. The. purification and mechanism of action of human antithrombin-heparin cofactor. J. Bid. Chem, 248, 6490-6505, 1973.
3.
BJORK, I., JACKSON, CM., JORNVALL, H., LAVINE, K.K., NORDLING, K., and SALSGIVER, W.J. The active site of antithrombin. Release of the same proteolytically cleared form of the inhibitor from complexes with factor IXa, factor Xa, and thrombin. .I. Biol. Cl~ern, 257, 2406-2311, 1982.
4.
TRAVIS, Biochem,
5.
GRIFFITH, M.J. The heparin enhanced antithrombin IIVthrombin reaction is saturable with respect to both thrombin and antithrombin III. .I.Biol. Chem, 257, 13899-13902, 1982.
6
OLSON, S.T. Transient kinetics of heparin-catalyzed protease inactivation by antithrombin III. Linkage of protease-inhibitor-heparin interactions in the reaction with thrombin. J. Biol. Chem, 263, 1698-1708. 1988.
7.
DAWES, J. Measurement of the affinities of heparin, naturally occurring glucosaminoglycons and other sulfated polymers for antithrombin III and thrombin. .4tlal. Biochem, 174, 177-186, 1988.
x.
BJORK, I., and LINDAHL, U. Mechanism Cell. Biocllem, 48, 161-l 82, 1982.
9.
JAQUES, 1980.
10.
MARKWARDT, F., and WALSMANN, P. Reindarstellung und Analyse des Thrombin inhibitors Hirudin. Hope-Seyler’s Z. Physiol. Chem, 338. 1381-1386, 1967.
11.
MARK WARDT, F. Pharmacology of hirudin: one hundred years after the first report of the anticoagulant agent in medicinal leeches. Biomcd., Bioc/lir?l. Actn, 44, 1007-1013, 1983.
12.
STONE, thrombin
1Ii.
FENTON,
14.
RYDEL, T.J., RAVICHANDRAN, K.G., TULINSKY, A., BODE, W., HUBER, R., ROITSCH, C., and FENTON, J.W. II. The structure of II complex of recombinant hirudin and human alpha strombin. Science, 249, 277-280, 1990.
15.
GRUTTER, M.G., PRIESTEL, J.P., REHUEL, J., GROSSENBACHER, H., BODE, W., HOFSTEENGE, J., and STONE, S.R. Crystal structure of the thrombin-hirudin comples: a novel mode of serine protease inhibition. EMBO J, 9,2361-2365, 1990.
J., and SALVESEN, 52, 655-709. 1983.
L.B.
Heparin-anionic
G.S.
Y.
Blood Coagulation.
Ann. Rev. Biochern,
1.
Human
plasma
proteinase
of the anticoagulant
polyelectrolyte
drugs.
inhibitors.
.4nnir. Rev.
action of heparin.
Pharmacol.
49,
Mol.
Rev, 31. 99- 166,
S.R., BRAUN, P.J., and HOFSTEENGE, J. Identification of regions of CIinvolved in its interaction with hirudin. Biochemistry, 26, 4617-4624. 1987. J.W. II. Thrombin
specificity.
Ann. N.Y. Acod. Sci, 370, 468-495.
1981.
616
HEPARIN-HIRUDIN
THROMBIN
Vol. 63, No 6
COMPLEX
IMAO, S.J.‘f., YATES, M.l‘.. OWEN, ‘I .J.. and KRSTENAYSKY, .I.[.. intcractiol.1 I)I hirudin with thrombin indentification of a minimal binding domain of hirudir rhat inhibit\ clotting activitv. Wii,c,/itr,rzi.\t,_\., 27. X 170 ‘i! 7 1, IYXX. 17.
MARAGANORE. J.M.. CIlAO. B,, RAMACIHANDRAN, K.L. Anticoagulant Cllc/n, 264, 11729-l 173.5, 19X9.
18
JAKUBOWSKY, J.A., and MARAGANORE, thrombin-induced platelet activities by a synthetic terminus of hirudin. Blood, 75. 399-406, 1990.
19.
NOE, G., HOFSTEENGE, J., ROVELL,I, specific antibodies to identify a secondary 11729-l 171.5, 1988.
20.
STONE,
JOSEPIH, M.I. JABLONSKY J. .in(l activity of synthetic hirudin peptitles. 1 /3i~~~
J.M. Inhibition of coagulation and dodecupeptide modeled on the carbosy-
G., and STONE, S.R. The use of sequencebinding site in thrombin. .I. Biol. Chem, 263,
S.R., and HOFSTEENGE, J. Kinetics 2.5, 4622-4628, 1986.
of the inhibition
of thrombin
by IHirudin.
Biochemistry, 21.
LACKMANN, M., HOAD, R., KAKAKIOS, A., and GECZY, C.L. Radioimmunoassay for the detection of assay characteristics and quantitation of recombinant hirudin. this issue, 1991.
33
STONE, S.R., DENNIS, S., and HOFSTEENGE, contribution of ionic interactions to the formation Biochemistry. 28, 6X57-6861, 1989.
23.
DAWES,
J., JAMES,
Monoclonnl antibodies III complex. Thromh. 23
‘5.
K., MICKLEM,
L.R.,
J. Quantitative evaluation of the of the thrombin-hirudin complex.
PEPPER,
directed against human a-thrombin Res, 36, 397-409, 1984.
D.S.,
and
PROWSE,
and the thrombin-anti
C.V.
thrombin
KRSTENANSKY, J.L., OWEN, T.A., YATES, M.T., and MAO, S.J.T. Anticoagulant peptides: nature of the interaction of the c-terminal region of hirudin with a noncatalytic binding sites on thrombin. .I. Med. Chern, 30, 16X8-1691, 1987. FENTON,
J.W. II, OWEN,
exosite of human cn-thrombin 1988.
T.J., ZABINSKI,
M.P., and WII,NER,
and fibrin (ogen) recognition.
G.D.
Anion-binding 27, 7106-71 12.
Bioc/wni.stry,
26.
NESHEIM, M.E. A simple rate law that describes the kinetics of the heparin-catalysed reaction between Antithrombin III and thrombin. ./. Biol. C/WUL, 264, 1470X- 147 17. 19X?.
27.
CHANG, J.Y. Binding of heparin to human 2ntithrombin III activates modification at lysine 236. ./. Biol. Chern, 264, 311 l-31 15, 19X9.
28.
PRATT, C.W., WHINNA, H.C., MEADE, J.B., TREANOR, CHURCH, T.C. Physicochemical aspects of heparin cofactor Acad. Sci, 556, 104-115, 1989.
selective
R.E., and 11. Am. N.1
chemical
