von willebrand factor

402

Biochimica et Biophysica Acta, 623 (1980) 402--411 © Elsevier/North-Holland Biomedical Press

BBA 38421

STUDIES ON FACTOR VIII-RELATED PROTEIN IV. INTERACTION OF GALACTOSE-SPECIFIC LECTINS WITH HUMAN FACTOR VIII/VON WILLEBRAND FACTOR

MIHA FURLAN, BEAT A. PERRET and EUGENE A. BECK

Central Hematology Laboratory, Inselspital and University of Berne, School of Medicine, 3010 Berne (Switzerland) (Received September 26th, 1979)

Key words: Factor VIII; Von Willebrand activity; Galactose-specific lectin

Summary Factor VIII of human cryoprecipitate was purified and fractionated on Sepharose CL-2B into three fractions of progressively decreasing multimer size and ristocetin cofactor activity. Following complete disulfide reduction, the resulting subunits were electrophoresed on 3% polyacrylamide gels and subsequently stained with two galactose-specific, fluorescein-labelled lectins from Ricinus c o m m u n i s (RCAI and RCAII). Measurements of fluorescence indicated that the reduced chains, derived from the largest factor VIII multimers, have a stronger binding affinity for both lectins than those obtained after reduction of smaller factor VIII species. Ristocetin cofactor activity of purified factor VIII was competitively inhibited by both Ricinus lectins and by concanavalin A. RCAi-lectin was found to be a considerably more efficient inhibitor than RCAII or concanavalin A. Following removal of sialic acid from factor VIII, the inhibiting effect of RCAii-lectin was markedly potentiated, probably by exposing additional galacrose residues, some of which must be located close to the ristocetin cofactor 'active site' of factor VIII. Ristocetin cofactor activity was also strongly inhibited by specific rabbit antibodies to human factor VIII. Such antibodies competed with RCAH-lectin for binding sites which are located on the surface of factor VIII multimers. Our results suggest that RCA~-lectin, which contains two galactose-specific binding sites per molecule, and anti-factor VIII antibodies inhibit ristocetin cofactor activity by crosslinking and aggregation of factor VIII multimers.

403 Introduction Human factor VIII is a glycoprotein with two distinct biologic activities: procoagulant activity ( V I I I : C ) corrects the blood coagulation defect in patients with hemophilia A, while von Willebrand factor activity (VIIIR : WF) participates in platelet adhesion at the site of vascular injury. The latter activity of human factor VIII is assayed by measuring the rate of aggregation of washed normal human platelets in the presence of an antibiotic, ristocetin, and has therefore also been named ristocetin cofactor. Factor VIII has been shown to be a family of multimers which vary in size, but contain the same repeating subunits [1--3 ]. The mechanism of interaction of factor VIII with human platelets is still unknown. Several reports have suggested that carbohydrate side-chains of factor VIII are involved in this interaction [4,5]. In our previous work, we presented evidence that neither the size
Purification o f factor VIII Human factor VIII was prepared from cryoprecipitate of fresh frozen normal plasma by gel filtration on Sepharose CL-2B as previously described [9], using 0.13 M NaC1/0.01 M Tris-HC1/0.01 M citrate buffer (pH 7.4) for elution at room temperature. SDS-electrophoresis in agarose and polyacrylamide gels Chromatographic fractions from the Sepharose column were characterized by electrophoresis on 1% agarose (Serva, Heidelberg) gels in 0.075 M sodium barbital/0.01 M EDTA buffer containing 6 M urea and 0.2% SDS [10], and stained for protein with Coomassie brilliant blue. Complete reduction of disulfide bonds was performed by heating factor VIII at 95°C for 10 min in 1% 2-mercaptoethanol/l% SDS/0.8 M urea. Reduced subunits of factor VIII were electrophoresed on 3% polyacrylamide gels (containing 3% by weight bisacrylamide : acrylamide) as reported elsewhere [9]. Polyacrylamide gels were fixed for 20 h in methanol/acetic acid/water (10 : 3 : 27 v/v/v) and stained for protein and galactose with Coomassie brilliant blue and fluorescein isothiocyanate (FITC) labelled Ricinus lectins, respectively. Staining with fluorescent lectins FITC-RCAz (FITC-Ricinus communis agglutinin-120) and FITC-RCAn (FITCR. communis agglutinin-60), both from Mfles-Yeda, Rehovot, was performed

404 essentially as recently reported [11 ]. Fixed gels were washed in 5% acetic acid and in four changes of 0.1 M NaC1/0.05 M Tris-HC1 (pH 7.0), and subsequently incubated for 24 h at room temperature with FITC-lectin solution. Staining solutions contained 0.4 mg/ml of RCA I (molar ratio FITC/protein = 7.4) or RCAII (molar ratio FITC/protein = 2.1) in 0.1 M NaC1/0.05 M Tris-HC1 (pH 7.0), respectively. Following destaining for 2 days with 0.1 M NaC1/0.05 M Tris-HC1 buffer, the gels were photographed under long-wave ultraviolet light using a Polacolor type 108 film [6]. Photographs of the gels were then scanned in a Joyce-Loebl reflectance densitometer through an orange filter (JoyceLoebl No. 5-045) and the resulting peaks converted into arbitrary units of FITC concentration using a calibration curve as previously described [6,11]. Polyacrylamide gels, which had been stained for protein with Coomassie brilliant blue, were photographed under white light illumination on Polacolor type 108 film, and the protein concentrations quantitated following scanning of the photographs through a 620 nm optical filter. Ratios of fluorescence and Coomassie staining intensities, representing arbitrary units of galactose concentration, were calculated from data of ten gels for each staining procedure. Treatment with neuraminidase and N-acetylgalactosaminidase Asialo factor VIII was prepared essentially as described by Gralnick [7]. Factor VIII containing 4 units VIIIR : WF/ml was incubated for 5 h at room temperature with 0.05 U/ml of neuraminidase from Clostridium perfringens, type V (Sigma) in 0.13 M NaC1/0.01 M Tris-HC1/0.01 M citrate, pH 7.0. To prevent proteolytic degradation during incubation, 100 U/ml of Trasylol (Bayer, Leverkusen) and 0.06 M benzamidine (Merck, Darmstadt) were incorporated in the digestion mixture. Incubation with 0.04 U/ml of a-N-acetylgalactosaminidase (Miles, Slough) was performed for 24 h at 37°C in 0.13 M NaC1/0.01 M CaC12/0.01 M Tris-HC1 (pH 7.4). Assay o f factor VIII-related activities Ristocetin cofactor was assayed using washed formalin-fixed human platelets [12] at a final ristocetin (Lundbeck, Copenhagen) concentration of 1.5 mg/ml. Fixed platelets, suspended in 0.14 M Tris-HC1/5 mM CaC12/1 mM MgC12/5 mM KC1 (pH 7.4) buffer, were added to a factor VIII-free supernatant of partially dehydrated plasma [ 13 ]. The resulting suspension had a protein concentration comparable to that of normal plasma. Factor VIII-related antigen (VIIIR : AG) was determined by quantitative electroimmunoassay (Laurell) in agarose gels containing a monospecific antiserum raised in rabbits with purified human factor VIII [14]. Inhibition o f VIIIR : WF by lectins, rabbit anti-factor VIII antibodies and sugars For inhibition experiments, gel chromatographic fractions between 70 ml and 120 ml of eluate were pooled. Aliquots of 0.1 ml were incubated for 15 min at room temperature with 10 pl of inhibitor solution. Concanavalin A (Calbiochem, San Diego), FITC-labelled lectins from R. communis RCAI and RCAn and a whole globulin fraction of rabbit antiserum to factor VIII, obtained by four precipitations with 50% saturated (NH4)2SO4, were used as

405 inhibitors. The crude globulin fraction from antiserum contained 60% IgG as determined b y electrophoresis on cellulose acetate and SDS-polyacrylamide. Incubation mixtures were then diluted with 0.14 M Tris-HC1 buffer (pH 7.4) and immediately assayed for VIIIR : WF. To examine the effect of lectins on washed platelets alone, platelet suspensions in factor VIII-deficient plasma were stirred for 30 s with lectins and left to stand for 15 min at r o o m temperature, prior to the assay of VIIIR : WF with unmodified factor VIII. Reversal of lectin-induced inhibition of VIIIR : WF was studied b y adding 1 M D-galactose (Serva, Heidelberg) or 1 M ~-methyl-D-mannoside (Fluka, Buchs) to the incubation mixture of factor VIII and lectin. Following incubation for 5 min at room temperature, the samples were diluted with 0.1 M Tris-HC1 buffer (pH 7.4) and assayed for VIIIR : WF. Competitive inhibition of VIIIR : WF b y lectins and antibodies was also examined. Factor VIII was first incubated for 15 min with the lectin RCAII at a final concentration of 3.2 #M. Following partial inhibition with this lectin, the incubation was continued for a further 15 min in the presence of the lectin RCA I or of the globulin fraction from rabbit antiserum. Effects o f the following sugars on factor VIII-induced platelet aggregation, were investigated: D-galactose, a-methyl-D-mannoside, sucrose (Merck, Darmstadt), N-acetyl-D-galactosamine (Fluka, Buchs) and N-acetyl-D-glucosamine (Fluka, Buchs). Individual sugars were dissolved in the platelet suspension to give the desired concentration in the final VIIIR : WF assay mixture. Unmodified factor VIII was added and the suspension was stirred for 1 min at 37°C prior to addition of ristocetin. Results

A representative elution diagram of human cryoprecipitate' on Sepharose CL2B is shown in Fig. 1. It is evident that the early factor VIII fractions carry

F i g . 1. F r a c t i o n a t i o n of factor V I I I b y S e p h s . r o s e C L - 2 B gel filtration. • --, r i s t o c e t i n cofactor activity; • •, factor VIII-related antigen; • A, a b s o r b a n c e a t 2 8 0 n m . S h a d e d a r e a s represent three fractions of f a c t o r V I I I w h i c h w e r e used for analysis; E = e a r l y f r a c t i o n , M = m i d d l e f r a c t i o n , L = l a t e f r a c t i o n . S D S - a g a r o s e electrophoretic gels of these fractions, stained with Coomassie b r i l l i a n t b l u e , a r e s h o w n a b o v e the respective elution v o l u m e s .

406 more ristocetin cofactor activity, relative to the antigen expression, than those of the later elution fractions. Superimposed SDS-electrophoretic gels in 1% agarose demonstrate a shift of the size distribution of factor VIII multimers with increasing elution volume. Three representative fractions, i.e. early (E), middle (M) and late (L), were chosen for sugar staining with galactose-specific lectins. All stained bands in these three fractions reacted with monospecific antiserum to human factor VIII, as determined by the second direction electrophoresis against factor VIII antibodies [10,15]. Agarose gels consist of galactose polymers and react strongly with galactose-specific lectins. Since polyacrylamide gels are not suitable for fractionation of large factor VIII multimers, we electrophoresed and stained factor VIII subunits in 3% polyacrylamide gels, following complete reduction of disulfide bridges. The resulting single polypeptide band represented the average subunit of the corresponding factor VIII population. As an example, gels of the reduced late fraction (L) from the gel filtration column are shown in Fig. 2, following staining with Coomassie brilliant blue and FITC-labelled lectin RCAI. Ten gels of each fraction (E, M, and L) were stained either with Coomassie brilliant blue, FITC-RCAI or FITC-RCAII, respectively. The ratio of VIIIR : AG/Coomassie does not seem to be size-dependent (Table I). On the other hand, the specific activity of ristocetin cofactor (VIIIR :WF/Coomassie) and binding affinity of the reduced subunit for both galactose-specific lectins decreases progressively with decreasing size of factor VIII oligomers. Sigmoidal inhibition curves, demonstrating competitive inhibition of platelet aggregating activity by lectins, are shown in Fig. 3. Concentrations of lectins that produced 50% inhibition of the aggregation response are given in Table II. It is evident that blocking of the binding sites on factor VIII for platelet receptors depends both on lectin and factor VIII concentration. 1 unit VIIIR : WF/ ml was inhibited 50% with 0.18 pM RCAI, whereas 50% inhibition of 0.1 unit activity was achieved at 0.03 pM RCAL Assuming that 1 ml of normal human plasma contains 10 #g factor VIII [16], corresponding to 1 unit VIIIR : WF, and that a single factor VIII protein subunit has a molecular weight of 250 000 [3], molar ratios of lectin/factor VIII subunit were calculated (Table II). Thus, 4.5 and 8.0 mol RCAx per mol subunit were required for 50% inhibition of 1 U/ml and 0.1 U/ml V I I I R : W F , respectively. This difference can be explained by the slow rate of reaction between factor VIII and lectin at room

Coomassie brilliant blue FITC-Ricinus lectin RCAI Fig. 2. Samples of the late fraction (L) w e r e r e d u c e d with 2-mercaptoethanol, electrophoresed in SDSp o l y a c r y l a m i d e gels a n d s t a i n e d f o r p r o t e i n (Coomassie brilliant blue) and galactose (FITC-Ricinus lectin RCAI).

407 TABLE I RISTOCETIN COFACTOR ACTIVITY, FACTOR VIII-RELATED ANTIGEN AND SUBUNIT STAINING I N T E N S I T I E S IN P O L Y A C R Y L A M I D E GELS OF E A R L Y , M I D D L E A N D L A T E S E P H A R O S E FRACTIONS

VIIIR :AG/Coomassie * VIIIR :WF/Coomassie * F I T C - R C A I / C o o m a s s i e ** F I T C - R C A i i / C o o m a s s i e **

Early

Middle

Late

0.99 2.48 0.99 ± 0.10 0.97 ± 0 . 1 1

1.12 1.72 0.81 ± 0 . 0 7 0.72 ± 0.10

1.04 0.76 0.66 ± 0.10 0.65 ± 0.11

• M e a n o f t h r e e assays. • * Mean ± S.D. o f t e n gels.

100

80

6O

~e

~

20

o

0.01

I

I

0.1 Lectin,

I

1.0

10.0

~M

Fig. 3. I n h i b i t i o n o f I u h i t / m l r i s t o c e t i n c o f a c t o r b y galactose-specific tectins R C A I (e e) and RCAII (a A) a n d b y g l u c o s e / m a n n o s e - s p e c i f i c c o n e a n a v a l i n A (m m ) . F a c t o r V I I I was i n c u b a t e d w i t h i n d i c a t e d c o n c e n t r a t i o n s o f l e c t i n s f o r 15 m i n a t r o o m t e m p e r a t u r e , p r i o r t o t h e r i s t o c e t i n c o f a c t o r assaY.

T A B L E II CONCENTRATIONS ACTIVITY

OF LECTINS PRODUCING

50% I N H I B I T I O N O F R I S T O C E T I N C O F A C T O R

T h e a s s u m e d m o l e c u l a r w e i g h t s a r e : R C A I , 1 2 0 0 0 0 ; R C A I I , 6 0 0 0 0 ; c o n c a n a v a l i n A, 1 1 0 0 0 0 ; f a c t o r V I I I subunit, 250 000.

1 unit VIIIR:WF/ml 0.1 u n i t V I I I R : W F / m l 0.1 u n i t V I I I R : W F / m l + 3 . 2 / ~ M R C A I I 1 u n i t asialo V I I I R : W F / m l

L e e t i n c o n c e n t r a t i o n (~uM)

Mol l e c t i n / m o ! f a c t o r V I I I subunit

RCA I

RCAII

Concanavalin A

RCA I

RCAII

Concanavalin A

0.13 0.03 0.27 0.13

6.8 4.7 -0.29

13.0 5.6 -12.5

4.5 8.0 67.5 3.3

170 1170 -7.2

325 1400 -313

408

temperature. RCAII and concanavalin A produced 50% inhibition of 1 U VIIIR : WF at 6.8 pM and 13.0 #M, respectively; the effect of these t w o lectins was considerably less dependent on factor VIII concentration than that of RCAI. To examine the competition between both galactose-specific lectins for the binding site on factor VIII molecule, 0.1 U VIIIR : WF/ml was first incubated with 3.2/aM RCAn and then with increasing concentrations of RCAI. Our results (50% inhibition at 0.27 #M RCAI) suggests that the RCAII-pretreated factor VIII is significantly less susceptible to inhibition b y RCAI than untreated material. Inhibition of platelet aggregation b y lectins can be fully ascribed to the masking of binding sites for platelet receptors on factor VIII, and not to a direct effect of lectins on platelets, since the aggregation assays were performed within less than 2 min with diluted factor VIII-lectin incubation mixtures. Moreover, more than 80% response to factor VIII was recovered when platelet suspensions were incubated for 15 min with 4 pM RCAI, 12 #M RCAII or 6 pM concanavalin A, prior to using them for the assay of ristocetin cofactor activity. More than 90% of VIIIR : WF was found in three purified factor VIII preparations following neuraminidase treatment as described in Materials and Methods. The resulting asialo-factor VIII (1 U/ml) was readily inhibited by much lower concentrations of RCAn, than unmodified factor VIII (Fig. 4 and Table II). Inhibition of VIIIR : WF b y RCAI or concanavalin A was hardly affected b y desialylation. Treatment of factor VIII with 0.04 U/ml a-N-acetylgalactosaminidase for 24 h at 37°C had no significant effect on ristocetin cofactor activity. All three lectins bind reversibly to factor VIII-related protein. A complete recovery of initial VIIIR : WF activity was observed when incubation mixtures of factor VIII with Ricinus lectins or concanavalin A were further incubated for 5 min with 0.1 M D-galactose or 0.1 M a-methyl-D-mannoside, respectively. Prior to activity assay, samples were diluted with buffer, and the resulting sugar concentrations in the platelet suspension were less than 0.01 M. Sugars alone

100

8O

60

4o iv= 20

o 0.01

I

I

I

0.1

1.0

10.0

Lectin, ~M Fig. 4. I n h i b i t i o n o f 1 u n i t / m l r i s t o c e t i n c o f a c t o r a c t i v i t y o f asialo f a c t o r V I I I b y l e c t i n s R C A I (e

RCAII (4

A) and concanavalin A (m

m).

e),

409 T A B L E III INHIBITION OF FACTOR VIII-INDUCED AGGREGATION L E T S BY S U G A R S IN T H E P R E S E N C E O F R I S T O C E T I N Sugar

D-Galactose ~-Methyl-D-mannoside Sucrose N-Ace tyl- D-galact osamine N-Acetyl-D-glucosamine

VnIR:WF

OF WASHED

FIXED HUMAN PLATE-

(% o f c o n t r o l )

0.15 M

0.30 M

0.60 M

102 100 97 101 90

88 93 75 83 66

82 88 57 64 43

exhibited an inhibiting effect on platelet aggregation by ristocetin cofactor only at considerably higher concentrations (Table III). Our results demonstrate that sucrose and both acetylhexosamines are more effective inhibitors than galactose and methylmannoside. Heterologous antibodies to factor VIII inhibit ristocetin cofactor activity [7 ]. We performed inhibition experiments with rabbit antibodies to factor VIII under the same conditions as those described with lectins. Again, the rate of VIIIR : WF inhibition was slow, especially at low antiserum concentrations, and the activity assay was done well before an equilibrium was achieved. We achieved 50% inhibition at 16 #g whole globulin/ml, containing about 60% IgG. If 7% of the total rabbit IgG is assumed to be anti-factor VIII [18], 50% inhibition of 0.1 U VIIIR : WF/ml is brought about by 0.67 pg anti-factor VIII per ml. This corresponds to approx. 1 mol antibody per mol factor VIII subunit. It is conceivable that not all rabbit antibodies are directed towards determinants located closely to the binding sites for the platelet receptors, and that the reaction was far from being completed within 15 min of incubation. Therefore, the inhibiting effect of antibodies can hardly be explained by blockade of factor VIII binding sites alone; the observed loss of activity appears to be rather a consequence of crosslinking and aggregation of factor VIII oligomers by antibodies. Pretreatment of factor VIII with 3.2 pM RCAH, prior to addition of antibodies, resulted in a shift of the antibody inhibition curve; for 50% inhibition of ristocetin cofactor activity five times as much globulin is required as with unmodified factor VIII, suggesting that binding of RCAn impairs the attachment of antibodies to the surface of factor VIII multimers.

Discussion Our present study demonstrates that reduced chains of factor VIII, electrophoresed in polyacrylamide gels, bind two fluorescein labelled lectins from R. c o m m u n i s (RCAI and RCAn). Both lectins bind to oligosaccharides containing D-galactose [19]. Reduced chains, derived from the large molecular weight forms of factor VIII, had a higher binding affinity for both galactosespecific lectins than those from the smaller species of factor VIII. The ratios of VIIIR : AG/Coomassie do not appear to be dependent on size of factor

410

VIII multimers, thus indicating that large forms of factor VIII have truly higher specific activity of ristocetin cofactor than smaller aggregates. It thus appears that factor VIII multimers are composed of qualitatively different subunits: the subunits of large aggregates may either contain more galactose, or have receptor sites better accessible to galactose-specific lectins than is the case with subunits of smaller multimers. Ristocetin cofactor activity of purified human factor VIII was strongly inhibited by the lectin RCAI; 50% inhibition of 0.1 unit VIIIR : WF/ml was observed following 15 min incubation with 0.03 pM RCAI. A comparable effect was achieved with 4.7 pM RCAII or 5.6 pM concanavalin A, suggesting that either receptors on factor VIII multimers for these two lectins are less accessible, or that reactive carbohydrate chains are located far from the binding site for platelets. Since both Ricinus lectins bind to galactose residues, we examined whether they compete for the same binding sites on factor VIII. We found that preincubation of factor VIII with RCAn impairs the inhibiting effect of RCAI, indicating that both lectins bind to the same carbohydrate receptors. The stronger inhibition of ristocetin cofactor activity by RCAI lectin, in comparison with RCAII, might be ascribed to differences in their molecular size. Moreover, RCAI has two sugar-binding sites per molecule, whilst RCAn has only one. A bivalent lectin can, in accordance with the Law of Mass Action, first react with the carbohydrate side chain of one factor VIII molecule. In a later stage, the lectin bound to one factor VIII molecule is able to attach to the carbohydrate chain of another factor VIII multimer. This crosslinking of factor VIII molecules leads to formation of large aggregates with a progressively decreasing ratio of surface to volume and consequent loss of binding affinity for platelet membrane receptors. Following desialylation, the inhibiting effect of RCAII was markedly potentiated. Obviously, additional reactive galactose residues were exposed by removal of sialic acid, some of.which must be located close to the binding sites for platelet receptors. These results are in agreement with observations of Sodetz et al. [8] that asialo factor VIII had a higher binding affinity for galactose~pecific hepatic lectin than native factor VIII. Increased ristocetin cofactor inhibition in asialo factor VIII by RCAn appears to reflect blocking of the penultimate galactose residue which seems to be involved in platelet aggregation [ 7,8 ]. Since the lectin RCAII also reacts with N-acetyl-D-galactosamine, we were interested whether this terminal amino sugar might participate in aggregation of platelets by factor VIII. Exhaustive treatment of factor VIII-related protein with ~-N-acetylgalactosaminidase resulted in complete recovery of ristocetin cofactor activity, suggesting that N-acetylgalactosamine residues are not involved in aggregation of platelets by factor VIII. Receptors for factor VIII on human platelets were shown to be membrane glycoproteins [20]. In order to examine whether these platelet receptors bind factor VIII by reacting with specific carbohydrate (galactose) residues, involved in platelet aggregation, we studied the effect of sugars on lectin-induced inhibition of ristocetin cofactor activity. This activity was completely recovered from inactive complexes between factor VIII and Ricinus lectins or concanavalin A,

411

following short incubation with 0.1 M D-galactose or 0.1 M a-methyl-D-mannoside, respectively. These two sugars, at the same concentration, did not influence aggregation of washed, fixed platelets by factor VIII in the presence of ristocetin; a 6-fold increase in sugar concentration produced only 20% inhibition of VIIIR : WF activity. These observations indicate that binding of human factor VIII to platelet membrane receptors does not correspond to a sugarlectin reaction. Sucrose, N-acetyl-D-galactosamine and N-acetyl-D-glucosamine appear to be more efficient inhibitors of ristocetin cofactor activity than D-galactose or a-methyl-D-mannoside. Kuwahara and Malik [21] also described stronger inhibition of factor VIII-induced platelet aggregation by sucrose and N-acetyl-glucosamine than by D-galactose; their platelets were, however, sensitive to considerably lower sugar concentration, than ours. This discrepancy might be ascribed to the fact that Kuwahara and Malik [21] measured the interaction of metabolically active guinea-pig platelets with bovine factor VIII. In platelet-rich plasma, primary aggregation of viable platelets is followed by a release reaction and secondary aggregation, whereas washed, formalintreated platelets, employed in our experiments, are metabolically inactive; their aggregation by factor VIII, in the ristocetin cofactor assay, reflects direct interactions between factor VIII and platelet membrane receptors [22 ]. Acknowledgements This work was supported by a grant from the Central Laboratory of the Swiss Red Cross Blood Transfusion Service and grant No. 3.936-078 from the Swiss National Science Foundation. We gratefully acknowledge the skillful technical assistance of Miss M. Stalder. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Fass, D.N., Knutson, G.J. a n d B o w i e , E.J.W. (1978) J. Lab. Clln. Med. 91,307--320 Counts, R.B., Paskell, S.L. and Elgee, S.K. (1978) J. Clin. Invest. 62, 702--709 Perret, B.A., FvLrlan, M. and Beck, E.A. (1979) B i o c h i m . B i o p h y s . A c t a 578, 164--174 Gralnick, H.R., Coller, B.S. and Sultan, Y. (1976) Science 192, 56--59 Peake, I.R. and B l o o m , A.L. (1977) T h r o m b o s . H a e m o s t a s . 37, 361--362 Fttrlan, M., Perret, B.A. and Beck, E.A. (1979) Biochim. B i o p h y s . A c t a 5 7 9 , 3 2 5 - - 3 3 3 Gralnick, H.R. (1978) J. Clin. Invest. 6 2 , 4 9 6 - - 4 9 9 Sodetz, J.M., Paulson, J.C., Pizzo, S;V. and McKee, P.A. (1978) J. Biol. C h e m . 253, 7202--7206 Fuzlan, M. and Beck, E.A. (1977) Thromb. Res. 10, 153--158 Beck, E.A., Tranqul-Poult, L., Chapel, A., Perret, B.A., Furlan, M., Hudry-Clergeon, G. and Suscillon, M. (1979) B i o e h i m . B i o p h y s . A c t a 5 7 8 , 1 5 5 - - 1 6 3 Furlan, M., Perret, B.A. and Beck, E.A. (1979) Anal, Biochem. 9 6 , 2 0 8 - - 2 1 4 Cooper, H.A. Reisner, F.F., Hall, M. and Wagner, R.H. (1975) J. Clln. Invest. 56, 751--760 Owen, W.G. and Wagner, R.H. (1972) Thrombos. Res. 1, 71--88 Meyer, D., Dreyfus, M.D. and L a ~ i e u , M.J. (1973) Path. Biol. 21 (Suppl.), 66--71 Converse, C.A. and Papermaster. D.S. (1975) Science 189, 469--472 Zimmerman , T.S., Ratnoff, O.D. a n d P o w e l l , A.E. (1971) J. Clln. Invest. 50, 244--254 Meyer, D., Jenkins, C.S.P., Dreyfus, M.D., Fressinaud, E. and Larrieu, M.J. (1974) Br. J. H a e m a t o l . 28, 579--599 Hoyer, L.W. (1972) Blood 3 9 , 4 8 1 - - 4 8 9 Nicolson, G.L., Blaustein, J. a n d Etzler, M.E. (1974) Biochemistry 13, 196--204 Cooper, H.A., Clemetson, K.J. and Liischer, E.F. (1979) Proc. Natl. A c a d . Sci. U.S.A. 76, 1069--1073 Kuwahara, S.S. and Malik, M. (1977) Am. J. Haematol. 2, 173--181 ~l|Airl, J.p., Cooper, H.A., Wagner, R.H. and B r i n k h o u s , K.M. (1975) J. Lab. Clln. Med. 85, 318--328