The effect of plasma antithrombin concentration on thrombin generation and fibrin gel structure

Thrombosis Research, Vol. 75. No. 2. pp. 203-212. 1994 Copyright 0 1994 Elscvicr Science Ltd Printed in the USA. All righls resaved 0049.3848/94 $6.00 + .oO

Pergamon

0049-3848(94)EOOOSS-1

THE EFFECT OF PLASMA ANTITHROMBIN CONCENTRATION THROMBIN GENERATION AND FIBRIN GEL STRUCTURE

ON

Graciela Elgue, Javier Sanchez, Kamaran Fatah *, Per Olsson and Birger Blomback# Departments of Surgery, Division of Experimental Surgery, and *Laboratory Medicine, Karolinska Hospital and #Department of Blood Coagulation Research, Karolinska Institutct, Stockholm, Sweden.

(Received 2 March 1994 by Editor P Kierulf; revised/accepted

Abstract

29 April 1994)

Congenital dcl‘icicncy cl‘antithrombin (AT) is associated with thrombotic cvcnts and AT consumption occurs in some sevcrc disorders and after treatment with heparin. The aim of this study was to investigate whcthcr variations in the lcvcl of plasma AT modify thrombin generation and the fibrin formation process after the intrinsic coagulation mechanism is triggered. Normal plasma was depleted of AT hy immunoadsorption on CNBr-Sepharosc coupled with the anti-AT-I@ traction of‘antiserum. The AT-depleted plasma was reconstituted with AT (between 0.3 and I.5 AT units per ml). Thromhin generation was measured as the development of thromhin-antithromhin complexes (TAT). The lag phase prcccding lihrin formation dcpcnded on the concentration of AT. The short lag phase was seen in complctcly AT-dcplctcd plasma and the long in plasma with 1.5 AT units per ml. TAT generation, dctcimined in parallel consecutive samples, showed that the rate at which thromhin was generated was inverse to the AT concentration in plasma. The network structure of hydrated fibrin gels in the clotted plasma was studied by measuring the wavclcngth dependence of gel turhidity. The mass/length ratio value, -i.e. the thickness of fiber strands and porosity of the gel increased with increasing AT concentrations. It is concluded that plasma AT regulates the rate of prothromhin-thromhin conversion, the clotting time and the consequently network structure of the fibrin gel.

Antithromhin (AT) is the main inhibitor ol‘the plasma coagulation process. By definition, plasma pooled from a healthy population contains 1.0 AT unit /ml. The notmal range is considered te he between 0.85 and 1.16 AT units per ml (l), and thcrcl‘orc a lcvcl of less than 0.8 units per ml might be regarded abnormal. Puticnts with halt’ ol‘the normal AT concentration suffer thrombotic events (2).In various severe disorders and after continuous intravenous heparin therapy, consumption cl‘ AT occurs tc levels below 0.8 units (3.4) and should therefore lead to impaired inhihition of coagulation.

Key-words: antithrombin, prothromhin-thrombin Corresponding author: G. Elgue, Department Karolinska Hospital, S- 17 1 76 STOCKHOLM,

conversion, ol‘ Surgcry, Sweden.

203

I‘ibrin gel structure. Division of Experimental

Surgery,

204

AT LEVEL

VARIATIONS

IN PLASMA

Vol. 75, No. 2

It has previously been shown that in plasma with approximately 50 per cent of normal AT concentration, more prothrombin is activated than in the same plasma reconstituted with AT (5), findings which suggest that the plasma AT concentration regulates the rate by which thrombin is formed on activation of the intrinsic coagulation system. It has also been demonstrated that thrombin added to fibrinogen solutions or plasma governs the structure of the ultimate fibrin gel i.e., the higher initial thrombin concentration, the thinner fibrin threads and tighter fibrin gel (6,7). Preliminary clinical investigations have shown that a potential to form tight fibrin network structures h vitro are associated with myocardial infarction at young age (8). The aim of the present study was to investigate whether variations in the plasma AT concentration modify the thrombin generation and fibrin formation processes after activation of the intrinsic coagulation system. Serial determinations of thrombin-antithrombin complex (TAT) formation in recalcified plasma with varying levels of AT were taken as a measure of thrombin generation. At the same time fibrin formation was monitored spectrophotometrically and the final fibrin network structure was determined from the wavelength dependence of turbidity.

MATERlALS

AND METHODS

Materials

CNBr-clctivcrted Sephnrosr 4B. was purchased from Pharmacia,

Stockholm,

Sweden,

Antithrombil? antibody, rabbit anti-human

antithrombin was purchased from Dako, Glostrup, The purified IgG fraction of antiserum had a protein concentration of 7.0 g/l.

Denmark.

Antithrombin (AT) for clinical use, obpained from Pharmacia, protein. The lyophilized powder was dissolved -70°C in 0.3 ml aliquots.

had a specific activity of 6.1 IU/mg in a water stock solution of 50 IU/ml and stored at

Factor Xo (FXo), was purchased obtained

from Chromogenix, Miilndal, Sweden. The lyophilized from bovine plasma, was diluted in water to 0.67 nkat/ml (pH 7.4).

powder,

Synthetic pepticle chromogenic arbstrotrs ,for FXu (S-2765), kallikrein (S-2302) and plasmin (S22.51), also purchased from Chromogenix, were dissolved in water to a final concentration of 0.9 mmol/l, 2.5 mmol/l and 3 mmol/l, respectively,

r-Hirudin, antithrombin

supplied by Pentapharm, Base], Switzerland, had a specific activity >lO,OOO units per mg pro&. The lyophilizcd powder was diluted in water to 4,000 U/ml.

Calcium chloride (Cr/C/2) was obtained water was prepared. Thr~mbin-antithrombin Behringwerke,

and stored in the dark at 4%.

Marburg,

from Merck, Dalmstadt,

Gelmany.

cornplrx (TAT) usstry kit, Enzygnost-TAT,

A 1.O mol/l solution in

was

purchased

from

Germany.

Plastic cuvettes, semi-micro disposable polystyrene cuvettes, They were supplied by Brand, Wcrtheim/Main, Germany.

with 1 cm light path, were used.

Methods Plcmicr wmpling Venous blood was drawn from 15 normal healthy volunteers. Nine ml of blood was mixed with 1 ml trisodium citrate 0. I3 mol/l and ccntrifugcd at room temperature for 20 min at 2,500 x g to obtain platelet-poor plasma (PPP). Pooled PPP was dialysed with TRIS-HCl 50 mmol/l buffer containing NaCl 0. I mol/l and Na2EDTA I .O mmol/l, pH 7.4, for three hours at 4°C to eliminate excess of &sodium citrate. The plasma was portioned in 5 ml aliquots and kept at -70%

Vol. 75, No. 2

AT LEVEL

VARIATIONS

IN PLASMA

205

Deplrtion ofAT in plrrsm was performed as follows: the dialysed PPP was depleted by passing 40 ml portions, at a rate of 20 ml per hour, at 4”C, through a column with 17 ml CNBr-activated Sepharose 4B coupled with 10 mg AT-IgG per ml gel, as recommended by the manufacturers I, equilibrated in NaCl 0.15 mol/l, with pH ad.justed to 7.4. Each 5 ml fraction collected was assayed for AT, spontaneous proteolytic activity and protein concentration. The spontaneous proteolytic activity was assayed as described hcfore (9), using the chromogenic substrates S-2302 (kallikrein) and S-2251 (plasmin). The protein concentration in the fractions was compared to that of the dialysed PPP. All samples were diluted l(X)-fold in NaCl 0.15 mol/l and the absorbance at 280 nm was measured in the Shimadzu UV-240 spectrophotomcter with an attached OPI-4. The fractions containing less than 0.005 AT units per ml, with an ahsorhance close to that of the original plasma and with no measurable spontaneous proteolytic activities, were pooled in 5 ml aliquots and stored at -70°C. The dilution factor of the AT-depleted plasma, according to the spectrophotometric determinations, was 1.25 times that of the non-depleted plasma. The AT-depleted plasma was assayed for coagulation factors XII, XI, X, IX, VIII, V, prothrombin, as well as for a2macroglobulin and at-antitrypsin hy routine clinical methods, and were found to he within the normal range. even when ignoring the dilution factor. The TAT concentration in normal plasma was 0.033.02 O.OlfO.01 pmol per ml.

pmol per ml and, in AT-depleted

plasma,

AT was assayed hy measurement of its FXa inhihitory capacity. Samples were diluted 15 with heparin buffer (50 mmol/l TRIS, containing 7.5 mmol/l Na2EDTA, 3 IU/ml heparin and 1% PEG, pH X.4, ionic strength (I) 0.2). Aliquots of 200 ~1 of thcsc dilutions were mixed with 200 ~1 01 FXa solution. Al‘tcr 5 min. at 37°C the residual FXa activity was determined by incubation with 200 ~1 of the chromogenic substrate S-2765 for 5 min. at 37%. The substrate hydrolysis was terminated by addition of 200 l_~lof 20% citric acid and the absorbance at 405 nm was read. If no residual FXa activity indicative ol‘cxccss AT in relation to FXa was dctcctcd, samples wcrc further diluted and rc-analyscd. A standard curve for AT was prepared by diluting PPP standardized against the WHO standard (National Institute for Biological Standards and Controls, Hcrtl‘ordshire, U.K.), with heparin bufcr from l:So() to 1:2X00. The resulting AT conccntralions ranged from 5.6 to I .O pmol/ml. The assay procedure was identical with that outlined above. TAT was assayed by a modified solid phase immunoassay before (IO). Fourteen pmol TAT - i.e., 14 pmol thromhin,

(ELISA) Enzygnost-TAT, as described cquuls to I .O unit thromhin.

Gel formation after addition of CaCl2 to a final concentration of 20 mmol/l, in AT-depleted and reconstituted plasma was monitored in polystyrcnc cuvcttcs at 600 nm, at 37%, in six cells of the Hitachi Model 150-20 spectrophotomctcr, as previously described (1 1). In each cuvette fibrin formation was monitored hy the turbidity dcvclopcd at onc-minute intervals. Clotting time was obtained from the turbidity profile by drawing a tangent to the first turbidity change in the turbidity curve. Its intcrscction with the time axis is dcfincd as clotting time (6). After 24 hours the gels were scanned from 550 to 790 nm wavclcngth and the mass/length ratio of the fibrin fihers wcrc determined as described hy Carr ( 12.13).

AT, 11 to 54 yl of a 50 IU/ml solution, was uddcd to I.8 ml AT-depleted plasma in cuvcttes to obtain final concentrations hctwccn 0.3 and 1.S AT units per ml. The AT-reconstituted plasma aliquots wcrc prc-warmed to 37°C and rccalcificd with 36 ~1 CaC12. Immediately after recalcilication, 0.7 ml was translcn’cd to ;L second cuvettc, which was placed in the Hitachi

1 Affinity Chromatography: Principles and Methods. Phannucia. Division. Chapter 3, pp IS- IX (19X6).

Laboratory

Separation

206

AT LEVEL VARIATIONS

IN PLASMA

Vol. 75, No. 2

spectrophotometer. From the remaining 1.1 ml in the original cuvette, 0.1 ml samples were taken four minutes after recalcification and then at every minute or every second minute until clotting occurred in the cuvette placed in the spectrophotometer. They were then transferred to tubes containing 10 ~1 trisodium citrate and 2.5 1.11 hirudin. Non-parametric comparisons were made between groups (Mann-Whitney, *P&.05, **PcO.Ol).

RESULTS The course of clotting in the normal dialysed recalcified PPP as well as in the corresponding ATdepleted plasma, reconstituted to 1.0 unit per ml with puritied AT, is demonstrated in Figure 1. On recalcification in the cuvette, the fibrin gelation took place after a lag phase of about 15 minutes. The graph shows no obvious difference in turbidity development between the values obtained with normal plasma and those obtained with the AT-dcplcted and AT-reconstituted plasma. Figure 2 panel A and Figure 3, illustrate the typical outcome in one experiment in which the fibrin gelation and TAT formation were simultaneously monitored at different plasma AT levels. Concerning the fibrin gelation (Figure 2 panel A) the shortest lag phase was seen in the completely AT-depleted plasma and the longest in plasma with 1.5 AT units per ml. The clotting time was found to be positively correlated to plasma AT concentration (Figure 2, panel B). This correlation did not essentially change when clotting time was defined as the tangent to the time axis intersection of the steep part of the turbidity CUI-VC . serial determinations of TAT formation until the commencement of fibrin gelation revealed that the mode of thromhin generation also depended on the AT concentration. Figure 3 shows that TAT formation was measurable at all AT concentrations l’ivc minutes after recalcification. However, with increasing AT, the initial rate of T.4T formation decreased with increasing AT concentrations. At zero AT concentration, no TAT could hc measured before or after fibrin gelation since the prerequisite for TAT formation was lacking. Comparison of Figure 3 with Figure 2A show that the curves depicting initial thromhin gcncration at all AT concentration precede gelation, -i.e. the formation of fibrin gels.

The

FIG. I Relation between turbidity aticr rcculcification of dialysed (0) PPP or (0) AT-depleted PPP reconstituted with AT to 1.O units/ml. The control PPP was diluted in 0.15 NaCl to the same protein concentration as the AT-depleted PPP, measured as absorbance at 280 nm.

Vol. 75, No. 2

AT LEVEL VARIATIONS IN PLASMA

207

B 13 12

I

11 n=5

10

n=5 n=5

Y 8

??

n=5

57

7 G3 6 5

uuuuuu 0

0.3

Time (min)

0.5

AT

0.8

1.0

(units/ml)

FIG. 2 (A) Turbidity dcvelopcd in one experiment alter recalcification of AT-depleted PPP reconstituted with different amounts of AT: (a) 0, (h) 0.3, (c) 0.5, (d) 0.8, (e) 1.0 and (f) 1.5 units/ml. (B) Box plot of clotting times after recalcification of AT-depleted plasma reconstituted with din‘crcnt AT concentrations. The circles represent the extremes and the horizontal lines I‘rom the bottom of the figure correspond to the 10th percentile, lower quartile, median, upper quartile and 90th percentile. The asterisks refer to tests between zero and the higher AT concentrations. Regression analysis AT concentration vs. clotting time: Y = 628.5+ +2 10.2 X, r?=O.96.

24

16

8

4

8

12

16

Time (min)

FIG. 3 TAT generated in the cxpcriment illustrated in Figure 2A, after recalcification of dialysed AT-depleted plasma rcconstitutcd with dillcrcnt amount of AT: (a) 0, (h) 0.3, (c) 0.5, (d) 0.X (e) 1.0 and (f) 1.5 units/ml.

1.5

208

AT LEVEL VARIATIONS

0.0

0.5

IN PLASMA

1.0

Vol. 75, No. 2

1.5

AT (units/ml) FIG. 4

TAT formed (M) eight. (+) nine, (A) ten and (0) 11 minutes after recalcification in ATdepleted plasma reconstituted with various amounts of AT. The TAT value obtained at 0.3 AT units /ml was arbitrarily regarded as 100 per cent, and the other values calculated as a per cent of it. Each point represents the mean value of four experiments. The solid line (0)shows the mean curve of all values; 16 determinations at each point. The data from the four experiments showed that TAT formation did not occur during the first four minutes after recalcification. At five minutes, values significantly above the zero value were recorded in two experiments. At seven minutes, the TAT level was at all AT concentrations increased to between 0.3 to 0.7 picomol/ml (corresponding to 0.02 to 0.05 thrombin units/ml) and then in each case it increased with time, in a similar manner as shown in Figure 3. At corresponding time-points after the recalcification, the highest values were always found in plasma with 0.3 U/ml AT. In Figure 4 the TAT values obtained at the higher AT concentrations were related to those measured in plasma with 0.3 AT units per ml. The features at eight, nine, ten and eleven minutes are illustrated, TAT formation always decreased gradually with increasing concentrations of AT. The results clearly demonstrate that plasma AT regulates the initial rate at which thrombin is generated when the intrinsic coagulation system is activated. The influence of the plasma AT concentration on the fibrin network structure was studied in five separate experiments but with the same PPP hatch as used for the above thrombin generation studies. From the box plot diagram in Figure 5 it appears that the mass/length ratio of the fibrin strands was positively correlated to the AT level - i.e., the higher AT, the thicker fibrin threads. The achieved median values of the thickness of the fibrin threads were correlated to the median values of the clotting time measured as described above. Figure 6 shows that the gel structure becomes a positive function of the clotting time. The plasma AT concentration thus, apart from regulating the rate by which thrombin also governs the structure of the being tibrin gel t’ormed.

is generated

AT LEVEL VARIATIONS

Vol. 75, No. 2

209

IN PLASMA

140 -

” b rl

t*

X

f

$

P I

0.3

I

I

I

I

I

1.0

0.8

0.5

1.5

AT (units/ml) FIG. 5

Box plot of mass/length ratio (p) in AT dcpletcd plasma reconstituted to different concentrations with AT. The asterisks rcfcr to tats bctwccn xro and the higher AT concentrations.

60

I

7.5

9.5

I

I

I I.5

13.5

I 15.5

Clotting time (set x lo*) FIG. 6 Linear rcgrcssion analysis hetwecn median of muss/length ratio (p) and median clotting time obtained at different AT concentrations. (Y=3.94 + 0.086 X, r2=0.89)

of

DISCUSSION A prerequisite for inducing

for answering the question prcscntly at issue is, of course:, whether the procedure AT deplction in plasma dots not alter the inherent procoagulant properties. The

210

AT LEVEL

VARIATIONS

IN PLASMA

Vol. 75, No. 2

described method for depleting PPP seems to fulfill the requirements, which include minor dilution, absence of signs of contact activation and maintenance of coagulation factors at normal levels. Furthermore, the clotting potential of the AT-depleted plasma as compared to normal plasma, was fully restored by the reconstitution with AT. Boyer et al (5) studied thrombin generation at different plasma AT levels, determined by thrombin generation test according to Biggs, prothrombin and AT consumption as well as by formation of TAT measured with a semiquantitative assay. The methods did not petmit accurate determination of small amounts of thrombin generated. The determination of TAT, in the present ELISA assay system, is undoubtedly an accurate method for measuring minute amounts of thrombin. In a bimolecular system, the TAT formation is strictly equimolar to the amount of thrombin added to the inhibitor present in molar excess (10). In the experiments performed, the AT concentration varied between 840 and 4200 pmol per ml. A lower rate of binding between AT and the initially formed thrombin may be expected to occur at lower than at higher levels of AT. Thus, since hirudin was added to the TAT assay, increasing amounts of free thrombin may have escaped detection in the TAT determination, especially at low AT concentrations in plasma. Even so, under the present conditions this would result in an undercstimatc of thrombin generation at low AT concentrations. The spectrophotometric monitoring of the development of fibrin offers an opportunity for determining the lag phase prior to gclation. It is evident that this lag phase was a function of the AT content. AT, as the main inhibitor of the coagulation process, would be expected to modulate the period from the triggering of the coagulation system until the formation of a fibrin clot. In the presence of heparin, AT is thought to exercise its inhibitory cfl’ect mainly on thrombin (14,15). In accordance with the findings of Boyer ct al (5), the present experiments demonstrate that prothrombin-thrombin transformation is invcrscly dependent on plasma AT concentrations. The obtained data do not disclose whether the lag phase until the appearance of the first thrombin molecules is regulated by AT. However, it is shown that the rate by which the subsequent thrombin generation takes place is a function of plasma AT -i.e., the higher AT concentration, the lower is the rate of thrombin generation. It has been suggested that the heparin-acccleratcd AT, by inhibiting thrombin, prevents the thrombin-induced activation of the coagulation factor VIII (FVIII) - the so-called positive feed-back loop in the intrinsic system, which is necessary to accelerate the clotting process (14,15). Provided such a mechanism is valid under the kinetic conditions prevailing in the absence of heparin, increasing AT concentrations could imply a mom efficient inhibition of the initially generated thrombin and therel’orc a lower degree of FVIII activation, leading to a decrease in prothrombinase activity. However, AT is also an inhibitor of the activated coagulation factors XII (FXIIa), XI (XIa), IX (FIXa) and X (FXa) (16). Although the bimolecular rate constant for AT-induced inhibition of FXa is about 50% of that for thrombin inhibition (17), it is still probable that variations in plasma AT concentrations will inlluencc FXa inhibition and thereby influence the degree of prothrombin cleavage (18). The inhibition rate constant for FIXa is one tenth that for FXa and the inhibition of FIXa and FXIIa is even lower (17). Indeed the coagulation enzyme cascade is an amplification system (IO). Hence, the molar excess of AT in relation to enzymes in the plasma rises upwards in the cascade. This means that a partial inhibition of FXIIa, FXIa and FIXa may occur and reduce the enzymatic impact on FX. Recent investigations have dealt with the structure of the fibrin networks upon addition of thrombin to fibrinogen solutions and plasma (6,7,11). It was shown that the gel structure was kinetically determined, -i.e. by the rate of activation of fibrinogen and that the initial network formed the scafold into which the activated fibrinogen molecules were deposited. The more thrombin added the more rigid and tight clots were formed. The biophysical environment influences the fibrin network structure (6,20,21) e.g. plasma proteins promote formation of more porous gel structures. Since AT constitutes such a minute fraction of the protein content in plasma, it is hardly probable that even the complete AT depletion of plasma significantly could change this environment and cause the fibrin structure changes seen here. If AT prr se exerted an effect on fibrin gelation, it would be expected to dccrcase the mass/length ratio of the fibrin strands according to a recent

Vol. 75, No. 2

211

AT LEVEL VARIATIONS IN PLASMA

rapport (22). The prcscnt data, thus indeed indicate that the rate of the initially generated thrombin after triggering OTthe intrinsic coagulation system dctcnnines the final structure of the clot. Networks with thin fiber strands and, therefore, rigid and space-l‘illing structures have heen implied in thrombotic disease processes (8). Important in this regard is the finding that such structures are less accessihlc I‘or fibrinolytic enzymes than networks with thicker fibril strands and larger pores (23). It is conceivable that the present i/7 \liho characterization of the clot structure reflects the structure of the clots formed in the vasculature. Thor degree of “malignancy” of a thrombus could thus vary. The present results confirm that even a moderate decrease in plasma AT concentr;ltion incrcascs the rate hy which thromhin is initially generated. Thereby the clotting time is shortened and the I‘inal clot changed to ;Lmore thromhogcnic structure. The implication OT the findings would he that AT consumed in states associated with increased intravascular coagulation or acquired by temporary hcparin administration should be suhstitutcd in order to maintain the paticnL’s antithromhotic mechanisms intact.

Acknowledgments This work was supported by the Swedish Foundation and grants I‘rom the Karolinska Swedish Heart and Lung Foundation.

Board of Technical Development, Tornspiran Institute. Kamuran Fatah was supported by the

REFERENCES

1. ODECARD, OR., LIE, M. and ABILGAARD U. Hcparin cofactor activity measured amidolytic method. Thromb Res 6, 2X7-294, 1975. 2. EGEBERG, 0. Inherited antithromhin Hacmonh 13. 5 16-530, 1965.

dcl‘icicncy

3. BLOMBACK, B., BLOMBACK, M., LAGERGREN, factor activity in plasma and its relation to the anticoagulant Acta Physiol Stand 58, 306-3 1X, 1963. 4. MARCINIAK, E. and GOCKERMAN, antithrombin-III. Lancct II, 58 I -5X4. 1977.

J.P.

causing

thromhophilia.

Thromh

with an

Diath

H. and OLSSON, P. The hcparin cocl‘fcct ol‘intravenously irjectcd hcparin.

Hcparin-induced

decrcasc

5. BOYER, C., WOLF, M., LAVARGNE, J.M. and LARRIEU, M.J. Thromhin formation of thromhin-antithromhiii III complcxcs in congenital antithromhin Thromh Rcs 20, 207-2 1X, 19X0.

in circulating

generation and III deficiency.

,$. BLOMBb;CK. B., CARLSSON, K., HESSEL, B., LILJEBORG, A., PROCYK, R. and ASLUND, N. Native fibrin gel networks ohscrvcd hy 3D microscopy, permeation 2nd turbidity. Biochem Biophys Acta OY7, 96 I IO, 1989. 7. BLOMBACK, B., CARLSSON, K., FATAH, K. and HESSEL, B. Fibrin in human plasma: gel architccturcs governed hy rate and nature of fibrinogen activation. Tromh Rcs Submitted. 8. FATAH, K., HAMSTEN, A., BLOMBACK, B. and BLOMB.&CK, M. Fibrin gel network characteristics and coronary heart disease: relations to plasma lihrinogen concentration, acute phase protein, serum lipoproteins and coronary atherosclerosis. Thromh Haemostas 68, 130- 1.35, 1992.

212

AT LEVEL VARIATIONS

IN PLASMA

Vol. 75, No. 2

9. BLOMBACK, M., CHMIELEWSKA, J., NETRE, C. and AKERBLOM 0. Activation of blood coagulation, fibrinolytic and kallikrcin systems during storage of plasma. VOXSang 47, 335 342. 1984.

10. ELGUE, G., PASCHE, B.. BLOMBACK, M. and OLSSON, P. The use of a commercial ELISA for assay of thrombin-antithrombin complexes in purified systems. Thromb Haemostasis 63, 435438, 1990. 11. BLOMBACK, B. and OKADA, M. Fibrin gel structure and clotting time. Thromb Res 2.5,5170, 1982. 12. CARR, ME. and GABRIEL DA. The effect of dextran 70 on the structure of plasma derived fibrin gels. J Lab Clin Med 96, 985993, 1980. 13. CARR ME. and GABRIEL DA. Dextmn-induced changes in fibrin fiber size and density based on wavelength dependence of gel turbidity. Macromolecules 23, 1473- 1477, 1980. 14. BEGUIN, S., LINDHOUT, T and HEMKER, H.C. The mode of action of heparin in plasma. Thromb Haemostas 60,457-462, 1988. 15. OFOSU, F.A., HIRSH. J., ESMON, C.T., MODI, G.J.. SMITH, L.M., ANVARI, N., BUCHANAN, M.R. FENTON II, J.W. and BLAJCHMAN. M.A. Unfractionated heparin inhibits thrombin-catalyzed amplification reaction more efficiently than those catalyzed by factor Xa. Biochem J 257, 143-150, 1989. 16. BARROWCLIFFE, T.W., JOHNSON, E.A. and THOMAS, D. Antithrombin III and heparin. Br Med Bull 34, 143-150; 1978. 17. BJGRK, I., OLSON, S.T. and SHORE, J.D. Molecular mechanisms of the accelerating effect of heparin on the reactions between antithrombin and clotting proteases. In: Heparin-Chemical and Biological Propertirs and Clinicnl Applict~tions. Lane D and Lindahl U (eds). p. 231, Edwards Arnold, London-Melbourne-Auckland (1989). 18. JESTY, J. Measurement of the kinetics of inhibition of activated coagulation factor X in human plasma: The effect of plasma and inhibitor concentration. Analytical Biochem 52,402-411, 1986. 19. BLOOM, A.L. Physiology of blood coagulation. Haemostasis 20, 14-29, 1990. 20. CARR, M.E. Fibrin formed in plasma is composed of fibers more massive than those formed with purified fibrinogen. Thromb Haemost 59. 535-539, 1988. 21. BLOMBACK, Biotechnology

B. Fibrinogen

and fibrin formation

and its role in fibrinolysis.

In:

of blood.J. Goldstein (ed), pp. 225-279, Buttetworth-Heineman, Boston, 1991.

22. NAIR, C.H. and DHALL, D.P. Studies on fibrin network structure: effect of some plasma proteins. Thromb Res 61, 315-325, 1991. 23. GABRIEL, D.A., MUGA, K. and BOOTHROYD, E.M. The effect of fibrin structure on fibrinolysis. J Biol Chcm 267, 24,259-24,263, 1992.