Kinetic analysis of various heparin fractions and heparin substitutes in the thrombin inhibition reaction

Biochimica etBiophysicaActa 838 (1985) 106-113

106

Elsevier BBA21952

Kinetic analysis of various heparin fractions and heparin substitutes in the thrombin inhibition reaction Carol H. Pletcher *, Mark Cunningham and Gary L. Nelsestuen ** Department of Biochemistry, University of Minnesota, 1479 Gortner A re., St. Paul, MN 55108 (U.S.A.) (Received June 5th, 1984)

Key words: Antithrombin; Heparin kinetics; Fluorescence

Kinetic characteristics of several heparin preparations and substitute heparins were determined to help understand the bases for activity differences. Several materials were highly active in factor Xa inhibition and the reaction rate at constant factor Xa concentration appeared to be predicted by the extent of intrinsic antithrombin III fluorescence change induced by the polysaccharide. Heparin fractions of different molecular weight and affinity for antithrombin III showed similar kinetic parameters in catalysis of the thrombin-antithrombin IIl reaction when these parameters were expressed on the basis of antithrombin III-binding heparin. The latter was determined by stoichiometric titration of the antithrombin III fluorescence change by the heparin preparation. However, the various heparin fractions showed very different specific activities per mg of total polysaccharide. This indicated that functional heparin molecules had similar kinetic properties regardless of size or antithrombin III-binding affinity and is possible because the K m for antithrombin | I I is determined by diffusion rather than by binding affinity. Substitute heparins and depolymerized heparin were poor catalysts for thrombin inhibition, due at least partially to their affinity for thrombin. This latter binary interaction inhibits thrombin reaction in the heparin-catalyzed reaction.

Introduction Heparin, a complex polysaccharide, serves as a catalyst in the antithrombin III inactivation of several proteinases of the coagulation-fibrinolytic system. Since heparin is often used therapeutically to treat thrombosis disorders, many studies have focused on understanding heparin's mode of action in this reaction. Lundblad et al. [1] can be consulted for a collection of works in this area. The heparin-dependent antithrombin I I I / proteinase reaction was modelled over a broad range of heparin concentrations as an ordered sequential two-substrate heparin-catalyzed reac* Present address: Cargill Research, P.O. Box Minneapolis, MN 55440, U.S.A. ** To whom correspondence should be addressed.

9300,

0304-4165/85/$03.30 © 1985 Elsevier Science Publishers B.V,

tion [2]. The K m for bovine thrombin was less than 2 riM; the K m for antithrombin III, 160 nM; and the kcat, 0.16 s -1. Subsequently Griffith [3] reported a K m of 7 nM for human thrombin and a kcat of 0.76 s-1. (The published data were a kc~t of 3.8 nM t h r o m b i n / m i n . n g heparin per ml for heparin of M r 12000). A half-maximal reaction velocity at about 160 nM thrombin was reported by Nesheim [4]. These studies were conducted in the presence of a potent active-site inhibitor of thrombin so that the concentration of free thrombin was less than 1.5 nM. In addition, Pletcher and Nelsestuen [5] as well as Griffith [3] observed that factor Xa was a poor proteinase substrate due to a high K m (about 100 nM) rather than to a lower kca t. It is clear that enhanced understanding of heparin's action is emerging from the use of saturation kinetics.

107 A major potential problem in comparing kinetic studies involving heparin is the nature of the heparin preparation used. Commercial heparin consists of polysaccharides that are heterogeneous with respect to size and charge. Heparin fractions that display higher activity on a weight basis as well as tighter binding to antithrombin III have been reported [6-8]. Higher activity could be achieved by changes in g m o r kcat or by removal of inert polysaccharides [6,9]. Another problem concerns the activity of substitute heparins (sulfated polysaccharides) and depolymerized heparins. Often materials are as active as heparin in the antithrombin Ill/factor Xa reaction but are substantially less active in the antithrombin I I I / thrombin reaction [10-14]. In addition, high concentrations of heparin inhibit the thrombin-antithrombin III reaction but not the antithrombin III-Xa reaction [6]. The results of this study indicated that the functional heparin molecules present in commercial heparin displayed a range of binding affinities for antithrombin III but had virtually identical kinetic properties in catalysis of the thrombin-antithrombin III reaction. These observations were consistent with the fact that g m values were diffusionally controlled rather than affinity controlled and that kca t w a s not related to product dissociation [5]. The inhibitory effect of various polysaccharides as well as their failure to function in the thrombin-antithrombin III reaction was due to tight binding of the polysaccharide to thrombin. A feature of native heparin necessary for high activity against thrombin is a relatively lower thrombin-binding affinity. Materials and Methods

Antithrombin III was purified to apparent homogeneity from bovine plasma by methods described previously [2]. The extinction coefficient used (E280nm) 1~ w a s 6.1 [15]. Bovine a-thrombin was prepared from pure bovine prothrombin using the activating system outlined previously [2]. Thrombin was separated from other reagents on a heparin-Sepharose affinity column [16]. Heparin-Sepharose and dextran sulfate-Sepharose affinity columns were prepared from heparin (Riker Laboratories, 161 units/mg) or dextran

sulfate (M r, 500000, Sigma) and Sepharose 4B (Pharmacia). The coupling and washing procedures were essentially those of Ref. 16. 1-2 mg of polysaccharide were covalently linked per ml of Sepharose gel as determined from the amount of heparin remaining in the reaction supernatant. Pentosane polysulfate was lot No. 17068 obtained from Biosynth., Ziarich. Thrombin activity was measured in 0.05 M Tris buffer (pH 8.4)/0.1 M NaCI/0.1% poly(ethylene glycol) 6000 (Sigma Chemical Co.) at 25°C. The substrate was 0.1 mM S-2288, H-D-isoleucyl-L-prolyl-L-arginine-p-nitroanilide dihydrochloride (KabiVitrum, Stockholm), which is cleaved to peptide plus p-nitroaniline (E4o5n m = 1.06- 10 -4 M -1 -cm-l). The kcat for thrombin acting on S2288 at 25°C, 100 s -1 [2], was used to estimate the molar concentration of thrombin. This velocity is substantially greater than that reported by the manufacturer. Factor Xa was prepared as described previously [5]. The assay was similar to the thrombin assay except that the substrate was 0.6 mM S-2222, N-benzoyl-L-isoleucyl-L-glutamylglycyl-L-arginineparanitroanilide hydrochloride (KabiVitrum, Stockholm). A kcat of 140 s -1 and the extinction coefficient for p-nitroaniline were used to estimate factor Xa concentrations. Proteinase inactivation rates were determined by either of two methods. One method, described previously [2] consisted of removing samples from a large reaction mixture, and mixing with polybrene (0.1% final concentration) to stop the reaction. These quenched samples were assayed immediately for proteinase. The other method involved samples that were incubated for a defined length of time and the entire sample was quenched with polybrene at one time point and then assayed for remaining proteinase. In this latter case all assays were done in triplicate. The determinations were quite consistent and usually fell within less than a 10% range. The amount of proteinase inhibited during the reaction time interval was used to calculate reaction velocity. Thrombin inhibition in the absence of heparin was determined for each reaction in a separate experiment and subtracted to give the heparin-catalyzed reaction velocity. Heparin was fractionated by chromatography on an antithrombin III affinity column by stan-

108 dard procedures. The affinity column contained about 1.6 m g / m l of antithrombin III linked by cyanogen bromide procedure to Sepharose 4B. 10 mg of commercial heparin was applied to a 2 × 6 cm antithrombin III affinity column. The flowthrough effluent was collected and the column was washed with 4 vol. of 0.05 M Tris buffer (pH 8.4)/0.1 M NaC1. Bound heparin was eluted with 3 M sodium chloride. The eluted heparin was pooled and dialyzed against 0.1 M a m m o n i u m bicarbonate in low M r cut-off dialysis tubing (3500 M r cutoff) and finally lyophilized. This was referred to as heparin I and contained 1.4 mg polysaccharide. The flow-through effluent from the first column was reapplied to the regenerated affinity column and the procedure for isolating bound heparin was repeated. The resulting heparin was referred to as heparin II. This process was repeated two more times to generate heparin III and IV. The final run-through effluent was referred to as heparin V and was dialyzed and lyophilized as above. The weight concentration of heparin present in various samples was estimated from Azure A dye-binding assays [9] or from turbidity measured at 600 nm in the presence of 0.1% polybrene. Commercial heparin was used to produce a standard curve and the two methods agreed within about 10%. The molar concentration of antithrombin IIIactive polysaccharide was estimated by standard fluorescence titration methods described previously [2,16]. Briefly, a known concentration of antithrombin III was placed in the fluorometer and aliquots of polysaccharide were added until the fluorescence emission (340 nm, excitation at 280 nm) reached a maximum. If the protein concentration was far above the K d for polysaccharide-antithrombin III binding, there would be virtually no free heparin and the titration would be stoichiometric. The molar concentration of antithrombin Ill-active polysaccharide was then obtained from the antithrombin III concentration and the volume of polysaccharide needed to produce the maximum change. Technically, the same procedure was used to determine the K j for heparin-antithrombin III binding. In the latter case the antithrombin III concentration was close to or below the K d so there are two heparin populations, free and antithrombin Ill-bound. The K~ is de-

fined as the free heparin concentration where the antithrombin III fluorescence change is half maximal. To substantiate that an equilibrium was titrated, this value was determined at three or four antithrombin III concentrations (typically 12 100 nM). Heparin was depolymerized with nitrous acid by the method in Ref. 17. The neutralized product was dialyzed against 0.1 M ammonium bicarbonate and lyophilized as outlined above. Alternatively, when small amounts of heparin were needed and the ionic strength would not be altered by added heparin, the neutralized product was used directly. Controls were run to ensure that the added reagent did not affect the reaction velocity. Molecular weights of antithrombin III and heparin-antithrombin III complexes were determined by light scattering using the methods and apparatus described by Pletcher et al. [18]. In order to eliminate dust, the proteins were prepared by the gel filtration method described. The initial molecular weights of various antithrombin III preparations used were 55 000 to 59 000, which agreed well with the molecular weight of monomeric antithrombin III [15]. Excess heparin was added and the resulting molecular weights of the protein were measured. The estimates were reproducible with a range of 2% for the same protein and polysaccharide preparations. The greatest variability was therefore the monodispersity of the individual protein preparation (see above). Alternatively, antithrombin III was chromatographed on a gel filtration column (1.1 × 60 cm of Sephadex G-100) containing buffer equilibrated with heparin (0.5 m g / m l ) . The molecular weight of the eluted antithrombin III was measured. Results

Characterization of fractionated heparins Heparin was fractionated according to its antithrombin III binding affinity as detailed in Materials and Methods. In summary, heparin I was the first material to bind to the antithrombin III affinity matrix and it showed the lowest K d (Table I). Heparin III was the material binding to the affinity gel on the third pass of the same heparin through the column. The K d for heparin III was similar to the K~ of unfractionated heparin.

109 TABLE I CHARACTERISTICS OF VARIOUS HEPARINS K d values were determined by fluorescence titration. K m and kea t values are data from plots drawn in Fig. 3. Vmax units are nmol antithrombin/ng polysaccharide per s. A M r values are molecular weight changes for antithrombin at saturating heparin. Heparin

Kd (nM)

K m for antithrombin (nM)

k¢, t ( s - 1)

limax ( z 10 s )

AM r

Heparin I Heparin 11I Heparin V Commercial

12 25 40 25

250 250 250 160 a

0.40 0.50 0.50 0.28 a

4.9 4.1 0.56 1.47 a

30000 9000

a Data not shown.

Heparin V was the material that did not bind after four passes through the column; the K d for heparin V was the highest. The amount of bound heparin decreased with each pass through the column and it appeared that the heparin in fraction III failed to saturate the affinity gel even though more heparin (fraction IV) was bound during the next application to the regenerated column. This property appeared similar to the results of Danielson and Bjork [7] who reported that a considerable amount of low molecular weight, but active heparin failed to adhere to an antithrombin III affinity gel under similar conditions. Affinity chromatography also appeared to fractionate the heparin by size. Chromatography of

<

heparins I and V on a gel filtration column indicated that heparin I had a higher molecular weight (Fig. 1). Unfractionated heparin showed a major peak eluting just ahead of heparin V (data not shown). The molecular weights were quantified using relative light scattering (Table I) to estimate the molecular weight of the heparin-antithrombin III complex. Heparin I caused a 30000 increase in molecular weight of antithrombin III, while the unfractionated heparin caused a 9000 increase. Affinity chromatography therefore appeared to automatically fractionate heparin by size with the highest affinity polysaccharide showing a much higher apparent molecular weight. These results are consistent with several previous studies on size-fractionated heparins [7,8]. As indicated below, the results do not distinguish the possibility that the heparin I molecules may self-associate and thereby only appear to have high molecular weight.

Kinetic properties of fractionated heparins

0,,) t~

rr

20

40

Fraction Number

Fig. 1. Chromatography of heparin 1 (e) and heparin 5 (zx) on TSK-2000 gel filtration column. Aliquots (50 ~1) of the two preparations (1.1 and 3.4 m g / m l , respectively) were loaded onto a 0,5 × 20 cm column and eluted with 25 mM Tris-HC1 (pH 7.5)/0.15 M NaCI. The buffer was delivered by a Varian model 5000 high-performance liquid chromatograph. Fractions (0.3 ml) were collected and assayed for heparin activity. Relative activity is shown.

The reaction mechanism of the fractionated heparins in the antithrombin III/thrombin reaction appeared identical to that of unfractionated heparin [2]. The initial reaction velocity was independent of thrombin down to at least 5 nM (Fig. 2). The K m for thrombin was therefore less than 5 nM for both heparins I and V. In addition, studies with the fractionated heparins showed no detected differences in the K m for antithrombin III (Fig. 3). All the reactions showed normal saturation kinetics, with the kinetically determined K m for antithrombin III higher than the K s (Table I). The K m for all the heparin preparations appeared

110 .8 ~

_

I

A

1.6 F-

~

~

- '

r

T

/

.

j

14 nM

1.4

1.2 .4 ~_

x

1.3

~: 0.8

.2

~

~ 1.2

E

~o 1.1 i 0

2

4

6

Polysaccharide

10

Concentration

~>20

1.0

(14g/ml)

Fig. 4. Comparison of factor Xa inhibition and antithromhin

.2 ~ 0

8

0

i

210

40

610

Time (sec) Fig. 2. Velocity of thrombin inhibition as a function of thrombin concentration for heparin I (part A ) and heparin V (Part

B). The concentrations of heparin were 1.3 n M in part A and 0.9 nM in part B. The antithrombin concentration was 25 nM in part A and 58 nM in part B. The thrombin concentrations are shown. The reactions were sampled at the times shown and thrombin concentration is proportional to the absorbance change at 405 nM.

to be a diffusionally controlled parameter. The k~,t, expressed per tool of antithrombin III-active polysaccharide, was similar for the various heparin preparations.

Ill fluorescence change induced by various heparins. The antithrombin llI concentration was 32 nM in all experiments. Factor Xa was 16 nM in the activity measurements. F/Fo is the ratio of antithrombin Ill fluorescence in the presence of given polysaccharide concentration ( F ) to that in its absence (k])). Results for heparin (A, activity; A F/Fo)' depolymerized heparin (11, activity; O, F/Fo) and for pentosane polysulfate (O, activity; ©, F/Fo) are shown.

When expressed on the basis of polysaccharide mass, however, the activity varied considerably for the different preparations (Table I). Affinity purified heparin showed over 3-times the specific activity of commercial heparin which is similar to the results of other laboratories [3,6]. The low affinity fraction was only about 10% as active as the high-affinity material. These differences appeared 100 u

o

,,.

/

//

> 50

2~

i

o ~ - ~ 0.1

01

,

0

.008

J

.016

I

.024

,

0.2

- - ~ 2.0

Polysaccharide

° 20

Concentration

200 (lag/ml)

.032

[Antithrombin I ~ (nM ~)

Fig. 3. Double-reciprocal plots of heparin-catalyzed thrombinantithrombin III reaction. The heparin concentration was adjusted to 2.5 nM, thrombin was 10 nM and antithrombin III was varied. Each point represents the average of three determinations. Results for heparin fractions I (O), III (©) and V (zx) are shown. The temperature was 26°C. The procedure and other conditions are as given in Materials and Methods.

Fig. 5. Inhibition of the heparin-catalyzed thrombin-antithrombin 11I reaction by various polysaccharides. All reactions contained 60 n M antithrombin III, 10 nM thrombin and 5 /~g/ml heparin. The velocity of the reaction in the presence of various additional polysaccharides are shown for dextran sulphate ( M r 500000) (©), pentosane polysulfate (O), depolymerized heparin (D) and heparin (zx) are shown. In the latter case, total heparin is plotted. The temperature was 17°C and full velocity was 1.05.10 -9 M thrombin per s.

111

i

i

lOO 80

A /

/

iiit II/

60

~.

0

I

30

I

1O0

,

LL

1000

,

100 B 80 ~

40 00

, =

10,000

,

//T =

•

Discussion

L / 250 PolysaccharideConcentration(IJg/ml) 40

widely different abilities to inhibit the heparincatalyzed antithrombin III/thrombin reaction (Fig. 5). The inhibitory capability and polysaccharide concentration at 50% inhibition were: dextran sulfate (0.4 /Lg/ml)> pentosane polysulfate (1.4 # g / m l ) > depolymerized heparin (4 #g/ml) > heparin itself (40/xg/ml). The ability of these polysaccharides to inhibt the antithrombin I I I / thrombin reaction coincided with the ability of the polysaccharides to elute thrombin from a polydextran sulfate-Sepharose affinity gel (Fig. 6A; dextran sulfate > pentosane polysulfate > depolymerized heparin > heparin). Moreover, the pentosane polysulfate appeared to remove thrombin from a heparin-Sepharose affinity gel more efficiently than heparin and in an essentially stoichiometric manner (50/xg pentosane polysulfate eluted 180/~g thrombin; Fig. 6B).

so

120

160

Fig. 6. Ability of various polysaccharides to elute thrombin from dextran sulfate-Sepharose (panel A) and heparin-Sepharose (panel B) affinity columns. The mixture contained 0.2 ml packed affinity gel and 0.36 mg thrombin in a total volume of 0.6 ml. The proportion of free thrombin is plotted as a function of added polysaccharide. The polysaccharides were dextran sulfate ( M r 500000, O), pentosane polysulfate (O), depolymerized heparin (D) and heparin (zx).

to be due to the presence of inactive materials in the unfractionated or low-affinity heparins.

Studies with sulfated polysaccharides Attempts to mimic heparin's mode of action with other sulfated polysaccharides have proved unsuccessful in the antithrombin III/thrombin reaction, even though they display high activity in the antithrombin Ill/factor Xa reaction [10-14]. At constant factor Xa concentration, the reaction velocity appeared to be directly proportional to the fluorescence change induced in antithrombin III (Fig. 4). High heparin concentrations are known to inhibit the thrombin-antithrombin III reaction. The various sulfated polysaccharides used here showed

Our studies on the heparin-stimulated reaction of thrombin or other proteinases with antithrombin III (Refs. 2 and 5 and this study) can be interpreted by the following series of reactions:

k3

(1)

T+H~TH

kl A~: HAT kcat --* H A k] k2

H+A~H

T # H+A-

T

H is heparin or a heparin substitute, A is antithrombin III, T is thrombin or other proteinase and the symbol - indicates the irreversible association of antithrombin III and proteinase. The resuits indicated that direct association of thrombin with H (K3, Eqn. 1) decreased reaction velocity. Previous results indicated that the TH species still reacted with HA but with a much higher K m [4,5]. The major productive pathway in Eqn. 1 corresponds to an ordered sequential two-substrate reaction with heparin as the catalyst and antithrombin III plus proteinase as the two substrates (Ref. 5 and references therein). A deviation from this reaction sequence has been proposed [3,4,19] in which either thrombin or antithrombin III can bind first to the heparin. Evidence that was inconsistent with random-

112 ordered addition or template mechanism for heparin function has been presented [2]. In addition, the activity of the polysaccharides was inversely proportional to its thrombin-binding affinity (see above), suggesting that an initial thrombin-polysaccharide binding was inhibitory.

Kinetics of commercial and affinity-fractionated heparins Polyanionic materials bind to both thrombin ( K 3, Eqn. 1) and to antithrombin III ( K 1, Eqn. 1). Jordan et al. [6] used fluorescent labelled heparins and estimated dissociation constants of about 100 nM for heparin binding to either thrombin or antithrombin III. The conditions used here included lower ionic strength and the observed antithrombin III-heparin binding was somewhat tighter (Table I), which corresponded closely with other reports [8,20]. The results indicated that the range of fractionated heparins obtained in this study included the range obtained in most studies [8,21]. Olson et al. [22] provided evidence that reaction 1 (Eqn. 1) could be separated into a larger number of distinct processes. Fractionation of commercial heparin by affinity chromatography invariably increases the maximum specific velocity of the reaction. Typical increases are 3-5-fold (Refs. 3 and 6; Table I). This could be due to isolation of a subpopulation of heparins with a higher k~at or it could be due to removal of inert polysaccharides. Since the K m for antithrombin III is dependent on k~, t (i.e., kc~t > k 1, Eqn. 1), a heparin subpopulation with a higher k~at should display an increased K,1 for antithrombin III except in the unlikely case where kl changed in an exactly compensating manner. That fact that the K m values for antithrombin III were essentially unchanged for high-affinity heparin and commercial heparin (Ref. 3 and Table I above) therefore provides circumstantial evidence that the higher velocity of high-affinity heparin was primarily due to removal of inert polysaccharides rather than to isolation of heparins with higher kc~t values. Direct evidence for this conclusion was obtained when reaction velocity was expressed per mol of antithrombin III-binding polysaccharide (kc,t, Table I), which gave similar values for all heparin fractions studied. This included commer-

cial heparin, high-molecular-weight high-affinity heparin, and low-molecular-weight heparin from which most of the antithrombin Ill-binding material had been removed to give a very low activity per mg of total polysaccharide (Table I). Given the several steps needed to determine the molar concentration of antithrombin Ill-binding polysaccharide, the small variations reported appeared minimal and most functional heparin molecules in the commercial heparin appeared to have similar in vitro kinetic parameters.

Kinetics of substitute and depolymerized heparins The results indicated that the anti-Xa activity was closely related to the extent of the intrinsic antithrombin lII fluorescence change induced by the polysaccharide (Fig. 4) and the fluorescence change may allow prediction of the k~,. These molecules had very low activity in the antithrombin III-thrombin reaction and actually inhibited the heparin-catalyzed reaction in a manner that correlated with their highly different abilities to bind to and elute thrombin from affinity gel matrixes. In contrast, antithrombin Ill appears to bind more tightly to heparin than to the other polysaccharides such as pentosane [14]. These results indicated that polysaccharide-thrombin interactions were inhibitory. These interactions may function by increasing the K m for thrombin [5]. A partial reason for low thrombin activity of depolymerized heparin in the thrombin-antithrombin III reaction may be a higher thrombin-binding affinity. Heparin may assume a specific secondary structure in solution that is not optimal for thrombin binding. Depolymerization may disrupt this in some manner to give an anionic polymer which better fits the polysaccharide-binding site on thrombin. This structural feature appears to be in addition to the intensely studied (e.g., Refs. 23-26) heparin structures responsible for optimal interaction with antithrombin III. After preparation of this manuscript, Scully and Kakkar [27] reported a study on the inhibitory action of pentosane polysulfate on heparin-catalyzed thrombin-antithrombin III reaction. They also concluded that the inhibition was due to pentosane-thrombin interaction. Lane et al. [28] have presented results showing that an octadecasaccharide is the minimum structure capable of

113 significant

stimulation

of

the

thrombin-anti-

t h r o m b i n I I I r e a c t i o n . T h i s size m a y b e n e c e s s a r y for s i m u l t a n e o u s b i n d i n g to b o t h p r o t e i n s [28] a n d for m a i n t a i n i n g the c o r r e c t s t r u c t u r e w i t h l o w a f f i n i t y for t h r o m b i n .

Acknowledgement T h i s w o r k was s u p p o r t e d in p a r t b y g r a n t s H L 15728 a n d H L 26989 f r o m the N a t i o n a l I n s t i t u t e s of H e a l t h .

References 1 Lundblad, R.L., Brown, W.V., Mann, K.G. and Roberts, H.R., (eds.) (1981) Chemistry and Biology of Heparin, Elsevier/North-Holland, New York 2 Pletcher, C.H. and Nelsestuen, G.L. (1982) J. Biol. Chem. 257, 5342-5345 3 Griffith, M.J. (1983) Proc. Natl. Acad. Sci. USA 80, 5460-5464 4 Nesheim, M.E. (1983) J. Biol. Chem. 258, 14708-14717 5 Pletcher, C.H. and Nelsestuen, G.L. (1983) J. Biol. Chem. 258, 1086-1091 6 Jordan, R.E., Oosta, G.M., Gardner, W.T. and Rosenberg, R.D. (1980) J. Biol. Chem. 255, 10081-10090 7 Danielsson, A. and Bjork, I. (1981) Biochem. J. 193, 427-433 8 Radoff, S. and Danishefsky, I. (1982) Arch. Biochem. Biophys. 215, 163-170 9 Lain, L.H., Silbert, J.E. and Rosenberg, R.D. (1976) Biochem. Biophys. Res. Commun. 69, 570-577 10 Linhardt, R.J., Grant, A., Cooney, C.L. and Langer, R. (1982) J. Biol. Chem. 257, 7310-7313 11 Holmer, E., Kurachi, K. and Soderstrom, G. (1981) Biochem. J. 193, 395-400

12 Fischer, A.M., Barrowcliffe, T.W. and Thomas, D.P. (1982) Thromb. Haemostasis 47, 104-108 13 Aiach, M., Michaud, A., Balian, J.-L., Lefebure, M., Woler, M. and Fourtillan, J. (1983) Thromb. Res. 31,611-621 14 Scully, M.F., Weerasinghe, K.M., Ellis, V., Djazaeri, B. and Kakkar, V.V. (1983) Throm. Res. 31, 87-97 15 Kurachi, K., Schmer, G., Hermodson, M.A., Teller, D.C. and Davie, E.W. (1976) Biochemistry 15, 368-373 16 Nordenman, B. and Bjork, I. (1978) Biochemistry 17, 3339-3344 17 Shively, J.E. and Conrad, H.E. (1976) Biochemistry 15, 3932-3942 18 Pletcher, C.H., Resnick, R.M., Wei, G.J., Bloomfield, V.A. and Nelsestuen, G.L. (1980) J. Biol. Chem. 255, 7433-7438 19 Griffith, M.J. (1982) J. Biol. Chem. 257, 13899-13902 20 Nordenman, B. and Bjork, I. (1981) Biochim. Biophys. Acta 672, 227 21 Laurent, T.C., Tengblad, A., Thunberg, L., Hook, M. and Lindahl, U. (1978) Biochem. J. 175, 691-701 22 Olson, S.T., Srinivasan, K.R., Bjork, I. and Shore, J.D. (1981) J. Biol. Chem. 256, 11073-11079 23 Casu, B., Oreste, P., Torri, G., Zoppetti, G., Choay, J., Lormeau, J.-C., Petitou, M. and Sinay, P. (1981) Biochem. J. 197, 599-609 24 Thunberg, L., Backstrom, G., Grundberg, H., Riesenfeld, J. and Lindahl, U. (1980) FEBS Lett. 117, 203-206 25 Riesenfeld, J., Thurnberg, L., Hook, M. and Lindahl, U. (1981) J. Biol. Chem. 256, 2389-2394 26 Lindahl, U., Backstrom, G., Hook, M., Thunberg, L., Fransson, L.-A. and Linker, A. (1979) Proc. Natl. Acad. Sci. USA 76, 3198-3202 27 Scully, M.F. and Kakkar, V.V. (1984) Biochem. J. 218, 657-665 28 Lane, D.A., Denton, J., Flynn, A.M., Thunberg, L. and Lindahl, U. (1984) Biochem. J. 218, 725-732