Interaction of heparin with fibrinogen using surface plasmon resonance technology: Investigation of heparin binding site on fibrinogen

Thrombosis

Pergamon

Research, Vol. 81. No. 4. DD. 503-509. 1996 Copyright 0 1996 Elsevier Scieke Ltd Printed in the USA. All rights reserved OtM9-3848/96$12.00 + .OO

PII SOO49-3848(96)00024-2

Brief Communication

INTERACTION OF HEPARIN WITH FIBRINOGEN USING SURFACE PLASMON RESONANCE TECHNOLOGY: INVESTIGATION OF HEPARIN BINDING SITE ON FIBRINOGEN Sanjew Raut and Patrick J. Gaffney Division of Haematology, National Institute for Biological Standards and Control Blanche Lane, South Mimms, Potters Bar, Hertfordshire, EN6 3QG, UK

(Received 12 December 1995 by Editor M.J. Seghatchian; accepted 8 January 1996)

Heparin is widely used as an antithrombotic drug, having excellent anticoagulant properties. However in certain clinical situations heparin’s efficacy seems to be somewhat limited. Despite the administration of heparin there is a high incidence of reocclusion of coronary arteries following thrombolytic therapy (1,2), and it has been observed that a significant number of patients receiving heparin treatment still exhibit thrombus extension (3). Although it is well established that the in vitro and in viva anticoagulant activities of heparin is mediated via the potentiation of the major coagulation inhibitor, antithrombin III (ATIII), some in vivo antithrombotic mechanisms are not fully understood. There is poor correlation between the anticoagulant activity of heparin as measured by in vitro assays and their in vivo antithrombotic efficacy. This may be due to heparin being targeted to many blood constituents whose resultant activities on the coagulation system have not been measured as yet. The antithrombotic activity of heparin as well as the pathogenesis of bleeding complications during heparin treatment cannot be completely explained by the inhibition of blood coagulation factors. Platelet dysfunction and acceleration of fibrinolytic process (4) have been implicated as additional factors involved. Recently a number of reports (5,6,7) have suggested that the inhibition of the antithrombotic activity of heparin in these clinical situations may be due to the interaction of heparin with other plasma proteins specifically with fibrin(ogen) present in the thrombus. Despite the possible pathophysiological significance of heparin-fibrin(ogen) interaction, little is known about the physicochemical aspects of this reaction. In this study an attempt was made to locate where heparin binds to fibrin(ogen), Key words: Fibrinogen, heparin, surface plasmon resonance, perfusion chromatography. Corresponding author: Dr. S. Raut, Division of Haematology, National Institute for Biological Standards and Control, Blanche Lane, South Mimms, Potters Bar, Her&-&hire, ~pg ~QG, UK. 503

504

HEPARIN BINDING SITE ON FBG

Vol. 81, No. 4

using various isolated structural domains from the plasmin-mediated digests of fibrinogen and the individual chains of fibrinogen. The BIALITETM system (Pharmacia Biosensor AB, Uppsala, Sweden) was employed for such a study. This utilises the Surface Plasmon Resonance (SPR) phenomenon (8,9) and allows a direct quantitative analysis of the label-free molecular interaction, in real-time, from which association and dissociation rate constants can readily be obtained.

MATERIALS

AND METHODS

Reagents Human fibrinogen (plasminogen free) was purchased from Imco (Stockholm, Sweden) and stored in the lyophilised form. Porcine intestinal mucosal heparin (UFH) (Mr(,-,) - 13kDa) was from Diosynth (Oss, Netherlands); Purified antithrombin (AT III) [ the 1” International Standard for AT III] was from the National Institute for Biological Standards and Control (NIBSC) (code W548). Streptavidin sensor chip SA5, surfactant P20 and amine coupling kit containing Nhydroxysuccinimide (NHS), (N-ethyl-N’-(3-diethylaminopropyl)~carbodiimide (EDC) and 1M ethanolamine-hydrochloride pH 8.5 were obtained from Pharmacia Biosensor AB (Uppsala, Sweden). Preparation and purification of fragments D and E, and the chains of human fibrinogen Human fibrinogen was digested by plasmin to the terminal and antigenically distinct fragments D and E as described elsewhere (10,ll) using 0.14 international units (iu) of human plasmin per mg of fibrinogen. From this digest fragments D and E were isolated in high purity by DEAE ion-exchange chromatography essentially as described by Nussenzweig et al, 1961 (12). These fragments were established at the NIBSC as internal research standards (code numbers being 73/503 and 73/504 for fragments D and E respectively) following freeze-drying in ampoules. The three individual polypeptide chains of the dimeric human fibrinogen were prepared and separated as described by Raut et al, 1994 (13). Briefly fibrinogen was reduced, Scarboxymethylated and the individual chains were isolated via perfusion chromatography. These chains were shown to be of high purity and they were character&d using chain specific monoclonal antibodies (14). Biotinylation of Heparin via Oxidized Cis-Dial Groups The procedure of O’Shannessy, 1990 (15) was used where heparin (4.5mg/ml) was first oxidized at room temperature with 3.3mM sodium metaperiodate (Sigma) in O.lM sodium acetate, pH 5.5. After l-h incubation the oxidization was stopped by adding an excess of sodium sulfite. The oxidised heparin was then incubated with a six-fold molar excess of biotinx-hydrazide (Calbiochem) for l-h. The excess biotin was then removed using a PD-10 column. Plasmon Resonance Analysis (Preparation of sensor surlfaces) All SPR measurements were carried out on a BIALITETM analytical system from Pharmacia Biosensor AB (Uppsala, Sweden) and at a constant temperature of 25.1“C. The cis-diol biotinylated heparin was immobilised to a covalently bound streptavidin sensor chip (SA5) using the high affinity biotin-streptavidin principle recommended by the immobilisation procedures described elsewhere (16). The amount of streptavidin on the SA5 chip quoted by the manufacturers was 65 rig/mm*. 35~1 of biotinylated-heparin (lOOpg/ml in 1OmM acetate buffer

Vol. 81, No. 4

HEPARIN BINDING SITE ON FBG

505

pH 5.5) was injected over the SA5 surface at a flow rate of SFl/min. In order to avoid severe limitation by masstransport of analytesbeing testedfrom bulk solutionsand achieve a suitable balance between mass transport and intrinsic reaction rate, an optimal concentration of approximately 500RU of surface bound heparin was obtained. Aflnity

and Kinetics determination

Analytes of interest were diluted in the running buffer (PBS buffer, pH 7.2, containing 5mM Na,SO,, 1mM EDTA, and 0.05% Tween 20) to various concentrationsin the interval of 0.1-5 PM and each in turn were injected over the heparin surface. The injected volume of each analyte was 511 at a constant eluent flow rate of lO@min. The heparinized surface was regeneratedwith 1.3M NaCl in PBS(54). Data points were collected continuously during the binding and dissociation processesand evaluated as described elsewhere (9). The kinetic analysis of sensogramsfrom the interaction of various analyteswith the immobilised heparin was based on the following rate equation (8,9): dWdt=k,,,,C%,-&,,C+k,&+ where R is the plasmon-resonanceresponse(RU) due to analyte interaction with immobilised heparin at time t seconds,k, and kA are the association @I%‘) and dissociation (s-l) rate constants,respectively, C is the concentrationof the injected analyte w’) and R- is the RU responseat saturation of heparin binding sites. The rate constantswere calculated from the analysis of sensogramsusing BIA Evaluation software at five different analyte concentrations, and the dissociationequilibrium constant(Ku) was calculatedas the ratio of &,,/b.

FIG. 1 Overlay plots of sensogramsshowing the dose responsebinding of fibrinogen fragment-D to immobilised heparin over a concentration range of 5.32 - 0.17 PM.

HEPARIN BINDING

Vol. 81, No. 4

SITE ON FBG

RESULTS In confirmation of data reported elsewhere (6), plasminogen free fibrinogen was shown to bind to the immobilised heparin in a dose response fashion with medium affinity, (K,, = 5.6 x lo-’ M). Furthermore, the individual k, and hx values of 9.1 x 103 M-k’ and 1.79 x lc3 S’ respectively, were obtained (Table 1). In order to determine the binding site of heparin on fibrinogen, the two major plasmin resistant structural domains of fibrinogen (D and E) were used as probes in the above SPR-type analysis in which it was shown that the isolated fragmentD bound to the immobilised heparin in a dose-response manner, FIG. 1. Fragment-D was injected at point A and the injection was terminated at point B by the influx of running buffer over the sensor surface. The interaction cycle was completed with an injection of 1.3M NaCl which dissociated the bound analyte and regenerated the heparinized surface for the next run. A KD value of 7.79 x 10m8M was obtained for the binding of fragment-D to heparin. Individual k, and b values of 1.86 x 10s M-k’ and 2.43 x 10” s”, respectively, were observed for this interaction. There was no binding observed for fragment-E to the immobilised heparin. Furthermore, the isolated individual polypeptide chains of fibrinogen (Aa, BB, y) showed no binding affinity for the immobilised heparin. As a positive control, SPR analysis confirmed that AT-III bound to heparin in a dose response fashion with a K,, value of 24.1 x10%4 which is similar to the values obtained elsewhere (17). Individual k, and b values of 9.74 x 10s M”s” and 7.89 x 1O-3s-’ respectively for ATIII-heparin interaction were also obtained. As a negative control, albumin at a concentration of 5OOpg/ml showed no binding affinity for heparin. TABLE Evaluation of the Interaction Between Immobilised E, and the Isolated Chains of Fibrinogen.

Fibrinogen Fragment-D Fragment-E Aar Chain B/3 Chain y Chain

I Heparin and Fibrinogen,

Binding to Heparin

Association Rate Constant k,, (M”s-‘1

Dissociation Rate Constant k,, (s-‘1

Dissociation Constant K, 0

+ + -

9.10x103 1.86~10’ -

1.79x10” 2.43~10~

560x10-’ 7.79x10a

+ Indicates Positive Binding to Immobilised Heparin .

Fragments D and

-

Heparin; - Indicates No Binding to Immobilised

Vol. 81, No. 4

507

HEPARIN BINDING SITE ON FBG

DISCUSSION The binding of heparin to a variety of analytes such as fibrinogen, isolated plasmin mediated fibrinogen fragments and isolated fibrinogen chains was studied using Surface Plasmon Resonance (SPR) technology. Biotinylated heparin was immobilised to streptavidin covalently attached to a dextran matrix, and this form of heparin was used to test the binding of various of the above mentioned proteins. Specific one to one interaction was measured in resonance units (RU) which is a direct measurement of mass interacting on the heparinized surface. This technology allowed us to observe unlabelled protein-heparin interactions and to study their binding kinetics, in real-time. Due to the limitation on the amount of each analyte, it was not possible to obtain the equilibrium response for each analyte-heparin interaction and the individual rate constants were obtained from the individual association and dissociation phases in the sensograms as described by Karlsson ef al, 1991(9). Five different concentrations were used for each analyte, and the dissociation equilibrium constants (K,) were calculated from the ratio of k.,,, and kff values. While this study confirmed the binding of fibrinogen to heparin (K, = 5.6 x 10m7M) as reported elsewhere (6) further experiments have demonstrated the domain in fibrinogen where heparin binds. Fragments D was found to avidly bind to heparin in a dose response fashion over a wide concentration range, but no binding with fragment-E was observed. This contradicted an earlier implication that the E domain of fibrinogen was involved (5,7). To delineate the primary sequence of the heparin binding site in the D domain of fibrin(ogen), individual Scarboxymethylated chains of fibrinogen, isolated by perfusion chromatography (13), were used to assess their binding to immobilised heparin. The finding that none of the individual chains of fibrinogen bound to the immobilised heparin suggested that the heparin site on fragment D was unlikely to be composed of linear sequences on a single chain. This seems to contradict the report (19) suggesting that heparin binds to fibrinogen via linear sequences in the I3 and y chains. Heparin therefore appears to bind to a region which is conformational amalgam of linear sequences in two or three of the polypeptide chains of fragment D, or a conformational binding site on one of the three subunit chains. This may explain at least one mechanism by which fibrin(ogen) exerts its anti-heparin activity. The fact that the D domain is involved may be significant in that this domain is present in each fibrin subunit as a dimeric structure. It is hoped that these results may aid in the design of novel heparins that have reduced affinity for fibrin(ogen) and hence display improved antithrombotic activity in common clinical situations where standard heparin may be compromised. Acknowledgements This work was supported by a grant from The Katholic University,

Leuven, Belgium.

REFERENCES 1. GOLD, H.K., LEINBACH, R.C., GARABEDIAN, H.D., YASUDA, T., JOHNS, J.A., GROSSBARD, E.B., PALACIOS, J. and COLLEN, D. Acute coronary reocclusion after thrombolysis with recombinant human tissue-type plasminogen activator: prevention by a maintenance infusion. Circulation 73: 347-352, 1986. 2. VERSTRAETE, M., ARNOLD, A.E.R., BROWER, R.W., COLLEN, D., de BONO, D.P., De ZWAAN, C., ERBEL, R., HILLIS, W.S., LENNANE, R.L., LUBSEN, J., MATHEY, D., REID, D-S., RUTSCH, W., SCHARTL, M., SCHOFER, J., SERRUYS,

508

HEPARIN BINDING SITE ON FBG

Vol. 81, No. 4

P.W., SIMMONS, ML., UEBIS, R., VAHANIAN, A., VERHEUGT, F.W.A. and von ESSEN, R. Acute coronary thrombolysis with recombinant human tissue-type plasminogen activator: Initial patency and influence of maintained infusion on reocclusion rate. Am J C&i01 60: 231-237, 1987. 3. MARDER, V.J.,

SOULEN, R.L., ATICHARTAKARN, V., BUDZYNSKI, A.Z., PARULEKAR, S., KIM, J.R., EDWARD, N., ZAHAVI, J. and ALGAZY, M. Quantitative venographic assessmmt of deep vein thrombosis in the evaluation of streptokinase and heparin therapy. J Lab Clin Med 89, 1018-1029, 1977. 4. VAIREL, E.G., BOUTY-BOYE, H., TOULEMONDE, F., DOUTREMEPUICH, C., MARSH, N.A., and GAFFNEY, P.J. Heparin and low molecular weight fraction enhances thrombolysis and by this pathway exercises a protective effect against thrombosis. Thrombosis Research 30, 219-224, 1983. 5. HOTCHKISS, K.A., CHESTERMAN, C.N. and HOGG, P.J. Inhibition of heparin activity in plasma by soluble fibrin: evidence for ternary thrombin-fibrin-heparin complex formation. Blood 84, 498-503, 1994. 6. MOHRI, H. and OHKUBO, T. Fibrinogen binds to heparin: The relationship of the binding of other adhesive proteins to heparin. Arch Biochem and Biophys 303, 27-31, 1993. 7. HOGG, P.J., and JACKSON, M. Fibrin monomer protects thrombin from inactivation by heparin-AT III: Implications for heparin efficacy. Proc Natl Acad Sci 86, 3619-3623, 1989. 8. JONSSON, U., FAGERSTAM, L., IVARSON, B., JOHNSSON, B., KARLSSON, R., LUNDH, K., LOFAS, S., PERSSON, B., ROOS, H., RONNBERG, I., SJOLANDER, S., STENBERG, E., STAHLBERG, R., URBANICZKY, C., OSTLIN, H. and MALMQUIST, M. Real-time biospecific interaction analysis using surface plasmon resonance and sensor chip technology. Bio Techniques II, 620-627, 1991. 9. KARLSSON, R., MICHAEISSON, A. AND MATTSSON, L. Kinetic analysis of monoclonal antibody-antigen interactions with a new biosensorbased analytical system. J. Immunol. Methods 245, 229-240, 1991. 10. GAFFNEY, P.J. and DOBOS, P. A structural aspect of human fibrinogen suggested by its plasmin degradation. FEBS Lett 25, 13-16, 1971. 11. JAMIESON, G.A. and GAFFNEY, P.J. Nature of the high molecular weight fraction of fibrinolytic digests of human fibrinogen. Biochimica Et Biophysics Acta 154, 96-109, 1968. 12. NUSSENSWEIG, V., SELIGMAN, M., PELMART, J. and GRABAR, P. Les produits de degradation du fibrinogene humaine par la plasmine, I: Separation et prprietes Annales de 1’Institut Pasteur 100, 377-387, 1961. physicochimiques. 13. RAUT, S., CORRAN, P.H. and GAFFNEY, P.J. Ultra-rapid preparation of milligram quantities of the purified polypeptide chains of human fibrinogen. J Chromatogr 660, 390-394, 1994. 14. RAUT, S., CORRAN, P.H. and GAFFNEY, P.J. Characterisation of the the chains of human fibrinogen isolated by perfusion chromatography using fibrin specific monoclonal antibodies. Thrombosis Research 79, 405413, 1995. 15, O’SHANNESSY, D.J. Antibodies biotinylated by sugar moities. Methods Enzymol 184, 162-166, 1990. 16. PHARMACIA BIOSENSOR AB. Ligand immobilisation chemistry. BIA application handbook 4.1-4.33, 1994. 17. WA’ITON, J., LONGSTAFF, C., LANE, D.A., and BARROWCLIFFE, T.W. Heparin binding affinity of normal and genetically modified antithrombin III measured using a monoclonal antibody to the heparin binding site of antithrombin III. Biochemistry 32, 72867292, 1993. 18. MOHRI, H., IWAMATSU,

A. and OHKUBO, T.

Heparin binding sites are located in

Vol. 81, No. 4

HEPARIN BINDING SITE ON FBG

a 40-kD gamma chain and a 36-kD beta chain fragment isolated from human fibrinogen. Thrombosis and Thrombolysis 1, 49-54, 1994.

509

J