THE JOURNAL t.Cb 1987 by The American Society for Biochemistry and

OF BIOLOG~CAL CHEMISTRY Molecular Biology, Inc

Vol. 262, No. 33, Issue of November 25, PP. 16157-16163,1987 Printed in U.S.A.

Trigramin A LOW MOLECULAR WEIGHT PEPTIDE INHIBITING FIBRINOGEN INTERACTION WITH PLATELET RECEPTORS EXPRESSED ON GLYCOPROTEIN IIb-IIIa COMPLEX*

(Received for publication, February 4,1987)

Tur-Fu HuangS, John C. Holt, Hanna Lukasiewicz, and Stefan Niewiarowskis From the Thrombosis Research Center and Department of Physiology, Temple University School of Medicine, Philadelphia, Pennsylvania 19140

Trigramin, a highly specific inhibitor of fibrinogen binding to platelet receptors, was purified to homoge- neity from Trimeresurusgramineus snake venom. Tri- gramin is a single chain (approximately 9 kDa) cys- teine-rich peptide with the Glu-Ala-Gly-Glu-Asp-Cys- Asp-Cys-Gly-Ser-Pro-Ala NHz-terminal sequence. Chymotryptic fragmentation showed the Arg-Gly-Asp sequence in trigramin. Trigramin inhibited fibrino- gen-induced aggregation of platelets stimulated by ADP (IC6,, = 1.3 X lo-‘ M) and aggregation of chymo- trypsin-treated platelets. It did not affect the platelet secretion. Trigramin was a competitive inhibitor of the Iz5I-fibrinogen binding to ADP-stimulated platelets (Ki = 2 X M). ‘261-Trigramin bound to resting platelets ( ICd = 1.7 X M; n = 16,500), to ADP-stimulated platelets (& = 2.1 X lo-’ M; n = 17,600), and to chymotrypsin-treated platelets (ICd = 8.8 X lo-’ M; n = 13,800) in a saturable manner. The number of lZ6I- trigramin binding sites on thrombasthenic platelets amounted to 2.7-5.4% of control values obtained for normal platelets and correlated with the reduced num- ber of GPIIb-GPIIIa molecules on the platelet surface. EDTA, monoclonal antibodies directed against the GPIIb-GPIIIa complex, and synthetic peptides (Arg- Gly-Asp-Ser and Tyr-Gly-Gln-Gln-His-His-Leu-Gly- Gly-Ala-Lys-Gln-Ala-Gly-Asp-Val) blocked both “‘1- fibrinogen binding and ”‘1-trigramin binding to plate- lets. Fibrinogen binding was more readily inhibited by these compounds than was trigramin binding. Mono- clonal antibodies directed either against GPIIb or GPIIIa molecules did not block the interaction of either ligand with platelets. Reduced, S-pyridylethyl, trigra- min did not inhibit platelet aggregation and fibrinogen binding to platelets, and it did not bind to platelets, suggesting that the secondary structure of this mole- cule is critical for expression of its biological activity.

It is well established that interaction of fibrinogen with specific receptors associated with the glycoprotein IIb-IIIa

* These studies have been supported in part by National Institutes of Health Grants HL 15226, HL 14217, and HL 36579. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement’’ in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

.$ Recipient of Fogarty Fellowship 1 F05 TW 03682-01 from the National Institutes of Health. Permanent address: Dept. of Pharma- cology, College of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China.

§ To whom reprint requests should be addressed.

(GPIIb-GPIIIa)’ complex is essential for platelet aggregation (1). Unstimulated platelets do not bind fibrinogen, and, there- fore, they do not aggregate in the circulation. When platelets are stimulated by agonists such as ADP (2-5), epinephrine (3, 5 ) , thrombin (6, 71, or prostaglandin endoperoxides (8, 9), fibrinogen receptors associated with the GPIIb-GPIIIa com- plex become exposed on the platelet surface, resulting in fibrinogen binding and subsequent platelet aggregation. The common interpretation is that ADP is an essential mediator of fibrinogen receptor exposure under physiological conditions (10). Pretreatment of intact platelets with proteolytic en- zymes such as chymotrypsin, Pronase, or elastase (4, 11-15) can also expose fibrinogen binding sites on the platelet surface even in the absence of ADP, resulting in a spontaneous platelet aggregation upon addition of fibrinogen.

The mechanism of fibrinogen receptor exposure on the platelet surface and molecular events associated with the interaction of fibrinogen with the GPIIb-GPIIIa complex are not fully understood. There is evidence that fibrinogen reacts with an epitope involving both components of the complex (16, 17) and that the exposure of this epitope depends on the conformation of the GPIIb-GPIIIa molecules (18). Calcium is essential for the formation of the GPIIb-GPIIIa complex and for its interaction with fibrinogen (19,20). Studies with mono- clonal antibodies directed against the GPIIb-GPIIIa complex (16, 17, 21-23), GPIIb (24), GPIIIa (25), and with synthetic peptides (24-28) provide reagents that can be utilized to map fibrinogen receptors.

In this paper we describe trigramin, a naturally occurring low molecular weight peptide from the venom of Trimeresurus gramineus snake. This peptide specifically and competitively inhibits fibrinogen binding to the receptors associated with the GPIIb-GPIIIa complex and exposed either by ADP or by limited proteolysis.

EXPERIMENTAL PROCEDURES~

RESULTS

Separation of Trigramin from Phospholipase A-Crude preparation of trigramin (previously designated as “platelet aggregation inhibitor”) contained phospholipase A activity

The abbreviations used are: GP, glycoprotein; HPLC, high pres- sure liquid chromatography.

Portions of this paper (including “Experimental Procedures” and Tables 1-111) are presented in miniprint a t the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Doc- ument No. 87M-346, cite the authors, and include a check or money order for $2.80 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

16157

16158 Trigramin Interaction with Glycoprotein IZb-IIIa Complex

which did not affect platelet aggregation as documented by experiments with the use of specific inhibitors of phospholi- pase A (29). Separation of trigramin from phospholipase activity was accomplished by means of reverse phase HPLC (Fig. 1). The first component eluted after a retention time of 37 min had inhibitory activity on ADP-induced platelet ag- gregation (at a concentration of 1.0 pg/ml), but it was devoid of phospholipase A activity. The second component contained phospholipase A activity, but it had no effect on platelet aggregation even at a concentration of 20 pg/ml. This com- ponent also did not alter platelet inhibitory activity of trigra- min as evaluated in the platelet aggregation assay.

Characterization of Purified Trigramin-The purified tri- gramin, further analyzed by sodium dodecyl sulfate-polya- crylamide gel electrophoresis, was insensitive to staining with Coomassie Brilliant Blue (at 20 pg). However, trigramin could be stained by silver (1 pg). It appeared as a single band, and the apparent molecular weight of trigramin was estimated to be 9 kDa approximately, using 20% gels (Fig. 2). The trigra- min appeared to be composed of a single polypeptide chain since it migrated with the same mobility whether reduced or nonreduced. The homogeneity of '"1-trigramin was also con- firmed by a combination of sodium dodecyl sulfate-polyacryl- amide gel electrophoresis (20% gels) and autoradiography (not shown). Edman degradation of trigramin revealed a single NHZ-terminal sequence, Glu-Ala-Gly-Glu-Asp-Cys-Asp-Cys- Gly-Gly-Ser-Pro-Ala (Table I, see Miniprint). This table also shows amino acid yields during NH2-terminal sequencing of S-pyridylethyl trigramin. The identification of phospholipase A was confirmed by the sequence His-Leu-Met determined for the material eluted from the reverse phase column. The amino acid composition of trigramin is shown in Table I1 (see Miniprint). Residues notable by their abundance are half- cystine, aspartate, glutamate, and glycine. Trigramin had approximately 86 amino acid residues with a minimal molec- ular weight of 8820. The protein content of trigramin calcu- lated on the basis of amino acid analysis and determined by the method of Lowry (31) using bovine serum albumin as a standard gave almost identical values. The study of chymo- tryptic fragmentation showed that trigramin contained an Arg-Gly-Asp sequence probably closer to COOH-terminal. Table I11 (see Miniprint) shows amino acid yields during NH2-terminal sequencing of chymotryptic peptide 3 of S- pyridylethyl trigramin.

Effect of Trigramin on Platelet Aggregation in Platelet-rich

g 0.0 I ID 0 N l- a 8 0.06

a m z

LL s: 5 0.02

To - e o PLA I - 60 -

c_

W _"" """' - 4 0 -1 E

20 2 - _"" _ _ "

"" z

u W

L "h\ - O ae I

20 30 40 50 RETENTION TIME, min

FIG. 1. Separation of trigramin (Tg) from phospholipase A (PLA) using reverse phase HPLC. The crude preparation of trigramin ("platelet aggregation inhibitor") was obtained from snake venom of 7'. Eramineus as described previously (29). An aliquot of 150 pg of crude trigramin (in 200 pl of 0.15 M NaCI) was injected into a column (250 X 4.6 mm) of Vydac T P R P equilihrated in 0.1% trifluoroacetic acid a t a flow rate of 1.0 ml/min. After the column was washed for 3 min, fractions were eluted during 50 min with a gradient of 0 4 5 % acetonitrile.

Mr

k D a

66-

43-

3 1-

22-

14-

9- - *-

R NR FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel electro-

phoresis of trigramin (1 pg) in 20% gels in both nonreduced (NR) and in reduced ( R ) systems. The gels were st.ained by silver after fixation by glutaraldehyde. The molecular weight standards are indicated by the urrows. * indicates the dye front.

I

I I

025 0.5 I .o 2.0 5 0

[TRIGRAMIN].pq/ml

FIG. 3. Concentration-dependent inhibitory effect of tri- gramin on the extent of fibrinogen (200 pg/ml)-induced plate- let aggregation of ADP (10 pM)-stimulated platelets (A-A) or chymotrypsin (CT)-treated platelets (W). Various doses of trigramin (0.25-5.0 pg/ml) were added to platelet suspensions and incuhated for 1 min prior to the addition of ADP and fibrinogen (in the case of ADP-stimulated platelets) or fibrinogen (in the case of platelets pretreated with chymotrypsin). The extent of the aggrega- tion of chymotrypsin-treated platelets was one-third of that of the ADP-stimulated platelets. Each point represents mean of a t least five experiments.

Plasma and in Isolated Platelet Suspension-In platelet-rich plasma, trigramin inhibited platelet aggregation induced by ADP (10 p ~ ) , epinephrine (50 p ~ ) , U46619 (2.5 p ~ ) , and sodium arachidonate (200 p M ) with ICsll = 2-4 X lo-' M. Fig. 3 shows dose-dependent inhibition of fibrinogen (200 pg/ml)- induced aggregation of ADP (10 pM)-stimulated platelets and of a-chymotrypsin-treated platelets. The IC,,, values for tri- gramin effects on the aggregation of ADP-stimulated platelets and on the aggregation of chymotrypsin-treated platelets were 1.3 X lo-' and 2.8 X lo-" M, respectively. It is known that chymotrypsin-treated platelets interact with fibrinogen di- rectly since they have fibrinogen receptors exposed on the surface (12). Therefore, the inhibitory effect of trigramin on fibrinogen-induced aggregation of chymotr?rpsin-treated

Trigramin Interaction with Glycoprotein Ilb-IIIa Complex 16159

platelets indicates interaction of trigramin with fibrinogen receptors on the platelet membranes. A significantly lower IC,,, value for trigramin effect on chymotrypsin-treated plate- lets can be explained assuming that chymotrypsin is a much less potent activator of platelet aggregation than ADP. Chy- motrypsin only exposed one-third of the total fibrinogen binding sites exposed by ADP (Table 111). For this reason lower concentrations of trigramin would be required to inhibit fibrinogen-induced aggregation of chymotrypsin-treated platelets than to inhibit fibrinogen-induced aggregation of ADP-stimulated platelets. Trigramin also inhibited in a con- centration-dependent manner platelet aggregation induced by U46619 (2.5 WM) and thrombin (0.5 unit/ml) whereas ["C] serotonin release was not inhibited (Fig. 4). Reduced trigra- min (2 x M, as S-pyridylethyl trigramin) did not inhibit ADP (10 FM)- or thrombin (0.5 unit/ml)-induced platelet aggregation. Also, reduced trigramin did not affect the inhib- itory effect of intact trigramin (2 X lo-? M ) or Arg-Gly-Asp- Ser (100 PM) or tyrosylpentadecapeptide (250 PM) on throm- bin (0.5 unit/ml)-induced platelet aggregation.

Effect of Trigramin on 12'I-Fibrinogen Binding to Platelets- As shown in Fig. 5, trigramin inhibited '251-fibrinogen binding to ADP (10 FM)-stimulated platelets in a concentration-de- pendent manner with an IC, of 2.8-5.6 X lo-' M. The double- reciprocal plot of the data is consistent with a competitive inhibitory mechanism of trigramin action ( K , = 2 X lo-' M). Trigramin also inhibited '251-fibrinogen binding to cu-chymo- trypsin-treated platelets with an IC,, of 1.1 X lo-' M indicat- ing its direct effect on the exposed fibrinogen receptors (data not shown). Reduced trigramin (2 X M) did not inhibit l2'1-fibrinogen binding of ADP-stimulated platelets.

Binding of '2'I-Trigramin to Platelets-The binding of '*'I- trigramin to ADP-stimulated platelets was time-dependent at 22 "C. After 1 min of incubation the amount of Iz5I-trigramin bound to platelets corresponded to 50% of the maximal amount of trigramin bound under optimal conditions. The binding of trigramin approached maximal values after 5 min of incubation; further incubation up to 30 min did not affect the amount of "'I-trigramin radioactivity recovered in the platelet pellet. Addition of 6 mM EDTA or 20 pg of unlabeled

THROMBIN U46619 FIG. 4. Effect of trigramin on thrombin (0.5 unit/ml)- and

U46619 (2.5 aM)-induced aggregation and ['4C]serotonin re- lease. Trigramin (4 pg/ml) was incubated with ['4C]serotonin-labeled platelet suspension for 1 min prior to the addition of thrombin or U46619. Following addition of the agonists, platelets were stirred for 5 min in a Payton aggregometer. The reaction was terminated by the addition of 0.1% glutaraldehyde, and the platelet suspension was centrifuged immediately at 15,000 x g in an Eppendorf centrifuge. Thrombin and U46619 released 76 f 5% and 30 f 3%, respectively, of the total ["CCJserotonin radioactivity. The extent of platelet aggre- gation induced by U46619 and by thrombin was similar. The bars represent percentages of control of serotonin release (Ed) and platelet aggregation (0) by trigramin observed in three experiments (mean values and standard deviations).

-0.01 0.01 0.05 0.b 0.20

FIG. 5. Double-reciprocal plot of lZ6I-fibrinogen binding to human platelet stimulated by 10 PM ADP in the absence ( O " 4 ) or presence of trigramin (0.5 rg/ml, a"0; 1.0 fig/ ml, A-A). Trigramin was added 3 min prior to the addition of ADP. The amount of 1251-fibrinogen bound to platelet pellet was measured following 10 min of incubation at 22 "C. Nonspecific bind- ing of fibrinogen was measured in the presence of 6 mM EDTA. The data are consistent with competitive inhibition of fibrinogen binding by trigramin with K, = 2 X lo-' M. This experiment is a representative one of three similar experiments.

[FlERlNOGEN]~ l p ~ l m l ~

TRIGRAYIN BOUND. M (a10.b

FIG. 6. Scatchard plot of the binding of '251-trigramin to unstimulated platelets (A-A) and to ADP-stimulated plate- lets (-). Various concentrations of '251-trigramin were added to platelet suspension (5 X 10' platelets/ml) 3 min prior to the addition of ADP (10 /IM) or an equivalent volume of Tyrode's buffer. The incubation mixture was incubated at 22 "C for 10 min before centrifugation through silicone oil. The Kd of trigramin was estimated to be 1.7 X lob7 M (unstimulated platelets) and 2.1 X M (ADP- stimulated platelets), respectively. This experiment is a representa- tive one of three similar experiments.

trigramin prior to ADP resulted in a displacement of about 90% of '2sII-trigramin radioactivity from the platelet pellet. However, if EDTA or unlabeled trigramin were added follow- ing 5 min after addition of ADP, the values of 12'I-trigramin radioactivity displaced from the pellets were 20 -+ 16% or 26 1- 6% of the total radioactivity bound to platelet pellets in control samples (n = 3). These experiments suggested that binding was in equilibrium and reversible only during a short incubation period; during longer incubation time trigramin became irreversibly bound to ADP-stimulated platelets. The binding of I2'I-trigramin to resting and to ADP-stimulated platelets approached saturation around 1 Fg/ml. Similar data was obtained with chymotrypsin-treated platelets (not shown). Scatchard analysis (Fig. 6) demonstrated that ADP increased binding affinity of trigramin to platelets, but it did not affect the number of binding sites. Table IV compares binding of Iz5I-trigramin and '2sI-fibrinogen to resting, ADP- stimulated, and chymotrypsin-treated platelets. As shown in the table, the number of binding sites for fibrinogen exposed by chymotrypsin is about one-third of the total number of

16160 Trigramin Interaction with Glycoprotein IIb-IIIa Complex

TABLE IV Binding of '"I-fibrinogen and of '"I-trigramin to unstimulated, ADP-stimulated, and

chymotrypsin-treated platelets The values represent means and standard deviations; number of experiments is given in parentheses; n, number

of binding sites/platelet. ND, not determined. The low values of "'I-fibrinogen binding to unstimulated platelets probably represent nonapecific binding.

'251-Fibrinogen binding Platelets

'"I-Trigramin binding

n K d X 10-I M n K d X lo-' M

Not stimulated ND ND 16,500 f 3,420 (7) 17 f 12 ADP-stimulated 34,200 f 13,800 (16) 5.4 f 2.4 17,600 f 3,000 (11) 2.1 f 1.4 Chymotrypsin-treated 10,800 f 6,400 (12) 4.2 -C 1.5 13,800 f 6,200 (9) 8.8 f 3.8

binding sites exposed by ADP. The binding affinities of fi- brinogen to ADP-stimulated and to chymotrypsin-treated platelets were similar. The number of trigramin binding sites on resting, chymotrypsin-treated, and ADP-stimulated was similar when the binding approached saturation. When com- pared to normal platelets, trigramin binding affinities to chymotrypsin-treated platelets and to ADP-stimulated plate- lets were increased by 2- and by 8-fold, respectively. Fibrino- gen (400 pg/ml) did not alter the number of trigramin binding sites on platelets (data not shown). The reduced '"1-trigramin bound to ADP-stimulated platelets with Kd = 2 X 1O" j M, and it occupied only 3% of the total binding trigramin sites at the condition approaching saturation.

In a study on two patients we demonstrated that the binding of "'I-trigramin to thrombasthenic platelets is greatly reduced (Fig. 7). The total number of trigramin binding sites on ADP- activated thrombasthenic platelets as analyzed by Scatchard plot amounted to 2.7 and 5.4% of the normal values. In one patient we found 7% of the trigramin binding sites on resting thrombasthenic platelets as compared to the total number of binding sites on normal platelets. This experiment suggested that the GPIIb-GPIIIa complex that is deficient in patients with Glanzmann's thrombasthenia is a requirement for trigramin binding to platelets. To confirm this observa- tion, we studied the effect of various monoclonal antibodies on the binding of '*'I-trigramin to normal and ADP-stimu- lated platelets. Table V shows that antibodies (A2A6, 7E3,

[TRIGRAMIN]. pq/ml

FIG. 7. Binding isotherms of laaI-trigramin to ADP-stimu- lated platelets obtained from a normal subject (o"--o) and from a patient with Glanzmann's thrombasthenia (A-A). Both platelet suspensions were activated with 10 p~ ADP. For other experimental details, see Fig. 6. Scatchard plot of the data showed that the number of trigramin binding sites on normal and on throm- basthenic platelets was 16,700 and 900 molecules/platelet, respec- tively; the corresponding Kd values were 1.2 X lo-' and 0.36 X lo-' M.

TABLE V Effect of monoclonal antibodies directed against the GPIIb-GPIIIa

complex, GPIIb, and GPIIIa on the binding of lZ5I-trigramin to platelets stimulated with ADP

Binding of 'zSI-trigramin to platelets was studied in parallel in two aliquots of the suspension prepared from the same donor. Ten pl of antibody (at the final concentration indicated in parentheses) or 10 pl of Tyrode's buffer were added to 1.0 ml of platelet suspension (5 X 10' platelets) and preincubated for 5 min. Subsequently, 1 pg of '"I- trigramin (specific radioactivity, 50,000 cpm) was added, and the suspension was incubated for another 3 min. This was followed by the addition of 10 p~ ADP, incubation for 10 min, and centrifugation through silicone oil. The radioactivity was determined in parallel in pellets prepared of control platelets or of platelets preincubated with the antibodies. In control experiments the amount of trigramin bound to platelets was 20-30 ng/108 platelets. The concentrations of anti- GPIIb-GPIIIa antibodies used were sufficient to inhibit completely binding of 100 pg '251-fibrinogen to ADP-stimulated platelets. The anti-GPIIb and anti-GPIIIa antibodies did not interfere with the fibrinogen binding to ADP-stimulated platelets.

Antibody Number of lZ6I-Trigrarnin

(per ml) specificity experiments binding, % of Antigenic

control

A2A6 (20 pg) GPIIb-GPIIIa 4 A2A6 (10 pg) GPIIb-GPIIIa 4

22 -C 5" 40 f 7

7E3 (10 p g ) GPIIb-GPIIIa 5 38 f 10 AP2 (10 pg) GPIIb-GPIIIa 3 TAB (10 pg) GPIIb 3

47 f 9 106 f 7

AP3 (10 pg) GPIIIa 3 130 f 6 SSA6 (20 pg) GPIIIa 2 110 f 8

a Complete inhibition of '*'I-trigramin binding to platelets was obtained at the same concentration of antibody and at 0.5 pg of '"I- trigramin.

and AP, at 10 pg/ml) directed against the GPIIb-GPIIIa complex consistently inhibited lZ5I-trigramin binding to plate- lets (by 60%), whereas antibodies directed against either GPIIIa (AP, and SSA6 at 10 pg/ml) or GPIIb (Tab, 10 pg/ ml) had no effect. A2A6 (20 pg/ml) inhibited about 80% of "'I-trigramin binding.

Since Arg-Gly-Asp-Ser inhibits binding of fibrinogen to GPIIb-GPIIIa complex on the platelet surface (26, 27) and the Arg-Gly-Asp sequence is present in trigramin (Table 111), the effect of this peptide on the binding of I2'I-trigramin to ADP-stimulated platelets was also investigated. Arg-Gly-Asp- Ser in a concentration-dependent manner (50-500 p M ) inhib- ited binding of "51-trigramin. On a double-reciprocal plot (Fig. 8) the Arg-Gly-Asp-Ser effect was consistent with a compet- itive inhibition of trigramin to platelets. In 11 experiments, the IC,, Arg-Gly-Asp-Ser effect on trigramin binding to plate- lets was about 125-200 p ~ . The maximal inhibition was obtained with 500 p~ Arg-Gly-Asp-Ser (62 f 9%, n = 4). The inhibitory effect of Arg-Gly-Asp-Ser on Iz5I-fibrinogen bind- ing to platelets was significantly stronger (ICs", around 50 p ~ ) . Lys-Gly-Asp-Phe-Ser-Ser (220 p ~ ) in three experiments had no significant effect on '*'I-trigramin binding and '''1- fibrinogen binding to platelets. Analysis of data presented in

Trigramin Interaction with Glycoprotein IIb-IIIa Complex 16161

900 /A

/ /

[TRIGRAMINj'. (pp/mli'

FIG. 8. Double-reciprocal plot of '"1-trigramin binding to ADP-stimulated platelets. Binding of control samples of lZ51-tri- gramin (U); binding of '=I-trigramin in the presence of 125 pM (A-A) or 250 IM (A-A) Arg-Gly-Asp-Ser. Synthetic peptides were added to platelet suspensions 3 min prior to ADP and '"1- trigramin. For another explanation, see Fig. 6. It shows that Arg-Gly-

9.0 x M). Asp-Ser inhibited trigramin binding in a competitive manner ( K , =

Fig. 8 by the method of Scatchard suggested that Arg-Gly- Asp-Ser did not alter the number of '2sI-trigramin binding sites on platelets but reduced binding affinity of this peptide. In six experiments the Kd values for binding of lZSI-trigramin to ADP-stimulated platelets in the presence of 250 pM Arg- Gly-Asp-Ser and in the absence of Arg-Gly-Asp-Ser were 1.8 X and 4.7 X lo-' M, respectively.

In five experiments we observed that tyrosylpentadecapep- tide (Tyr-Gly-Gln-Gln-His-His-Leu-Gly-Gly-Ala-Lys-Gln- Ala-Gly-Asp-Val) corresponding to the C-terminal end of the fibrinogen y chain blocked both '"I-fibrinogen binding to ADP-stimulated platelets (ICso = 120 @M) and "'I-trigramin binding to platelets stimulated by ADP (ICso = 240 PM) (data not shown). The maximal inhibition of this peptide on lZ5I- trigramin binding was observed at 800 p~ (60 k 4%, n = 4).

DISCUSSION

Several lines of evidence indicate that trigramin, a low molecular weight peptide purified from T. grumineus snake venom, binds specifically to the GPIIb-GPIIIa complex and that it blocks competitively fibrinogen binding to the recep- tors associated with this complex. Competitive inhibition of lZ5I-fibrinogen binding by trigramin was observed using ADP- stimulated platelets. Accordingly, trigramin inhibited the ag- gregation of intact platelets stimulated by thrombin and pros- taglandin endoperoxide analogues without inhibiting the platelet release reaction. Trigramin appeared to be a stronger inhibitor of fibrinogen-induced aggregation of chymotrypsin- treated platelets. There were several similarities between the binding of '251-trigramin and the binding of '251-fibrinogen to human platelets. The binding of both ligands was inhibited by EDTA, by monoclonal antibodies interacting with the GPIIb-GPIIIa complex (21-23), and by synthetic peptides representing putative platelet binding sites on the fibrinogen molecule, Arg-Gly-Asp-Ser (26, 27), and tyrosyl pentadeca- peptide of the C-terminal portion of the y chain, Gly-Gln- Gln-His-His-Leu-Gly-Gly-Ala-Lys-Gln-Ala-Gly-Asp-Val (28). Moreover, neither trigramin nor fibrinogen bound sig- nificantly to platelets of patients with Glanzmann's thromb- asthenia, deficient in the GPIIb-GPIIIa complex. After pro-

longed incubation of radiolabeled fibrinogen with activated platelets (44) and after prolonged incubation of trigramin with platelets (this study) binding becomes irreversible. How- ever, significant differences between the binding of trigramin and of fibrinogen to human platelets were observed. Fibrino- gen does not bind to resting platelets, and the stimulation of platelets by ADP or treatment with proteolytic enzymes is a requirement for the exposure of fibrinogen binding sites. On the other hand, the number of trigramin binding sites on resting platelets, on ADP-stimulated platelets, and on chy- motrypsin-treated platelets was similar and amounted to 50% of the total number of fibrinogen binding sites exposed by ADP.

Surprisingly, binding affinity of '*'I-trigramin to ADP- stimulated platelets, as judged on the basis of &, was approx- imately 25-fold greater than binding affinity of '"I-fibrinogen to ADP-stimulated platelets. This may explain the observa- tion that an excess of fibrinogen did not interfere with the binding of trigramin to platelets, that trigramin was efficient in platelet-rich plasma, and that both monoclonal antibodies and synthetic peptides inhibited fibrinogen binding to plate- lets more efficiently than trigramin binding to ADP-stimu- lated platelets. The binding affinity of trigramin to ADP- stimulated platelets resembled that of some monoclonal an- tibodies such as A2A6 (23) and PAC-1 (18); it was several orders of magnitude higher than that of synthetic peptides (26-28). I t is interesting that in the case of monoclonal antibody PAC-1 and in the case of trigramin, ADP was essential for the optimal binding to platelets and that the number of trigramin binding sites and PAC-1 binding sites/ platelet were similar. It is possible that each ligand may block two fibrinogen receptors on the platelet surface. Trigramin may bind to the same epitope as fibrinogen on the GPIIb- GPIIIa complex, or it may bind in close vicinity to the fibrinogen receptor epitope. The observation that trigramin, in contrast to fibrinogen, binds to resting platelets can be explained in agreement with Coller (45) who suggested that activation of platelets with ADP or proteolytic enzymes alters the microenvironment of platelet membranes, thus providing "access" for a large size molecule such as fibrinogen to the receptors. Increased binding affinity of trigramin to ADP- stimulated platelets is also compatible with the hypothesis that the conformational change of the GPIIb-GPIIIa complex induced by ADP is critical for the exposure of fibrinogen receptors (18). Most recently an attempt has been made to synthesize peptides modeled on both sequences present in the A 01 and y chain of the fibrinogen molecule and representing putative platelet binding sites. A basic peptide consisting of 13 amino acids has been synthesized (46). This peptide bound to thrombin-stimulated platelets with a Kd of 3.8 X M and it blocked, at the micromolar concentration, binding of radiolabeled fibrinogen to platelets. Still, the trigramin ap- peared to be a much stronger inhibitor of fibrinogen binding to platelets than this basic peptide.

Monoclonal antibodies against the GPIIb-GPIIIa complex and peptide Arg-Gly-Asp-Ser inhibited fibrinogen binding to platelets completely whereas in most experiments only partial inhibition of trigramin binding was obtained. The most likely explanation of this finding is that trigramin binds to platelets with a much higher affinity than that of fibrinogen (Table IV). However, it also does not exclude the possibility that the binding sites of trigramin and fibrinogen on the GPIIb-GPIIIa complex may not be entirely identical. Trigramin may be a useful probe for elucidation of the interaction of fibrinogen with its receptors on the platelet surface. Of particular interest is the relationship between the structure of trigramin and its

16162 Trigramin Interaction with Glycoprotein IIb-IIIa Complex

ability to bind to the GPIIb-GPIIIa complex. It is conceivable that the Arg-Gly-Asp sequence present in trigramin may play a role in the interaction of this molecule with the complex. Accordingly, we found that Arg-Gly-Asp-Ser inhibited com- petitively the binding of trigramin to the platelets (Fig. 8). However, the high content of half-cystine in trigramin and the loss of its binding ability and biological activity after reduction suggest an importance of the conformation of the molecule in the expression of its activity. I t is conceivable that another sequence of trigramin which is situated in close vicinity to Arg-Gly-Asp in the folded molecule, but is distant from Arg-Gly-Asp in the unfolded (reduced) molecule, may be important for trigramin interaction with the GPIIb-GPIIIa complex. In fact, a similar spatial arrangement may be critical for fibrinogen binding to platelets since the Ki for Arg-Gly- Asp-Ser or for His-His-Leu-Gly-Gly-Ala-Lys-Gln-Ala-Gly- Asp-Val are a few orders of magnitude higher than Kd for fibrinogen binding to ADP-stimulated platelets.

Recent evidence indicates that monoclonal antibodies blocking fibrinogen binding to the GPIIb-GPIIIa complex prevent formation of platelet microthrombi in experimental animals (47, 48). Further studies are required to explore the antithrombotic potential of trigramin or related molecules.

Acknowledgments-We wish to thank Drs. B. Rucinski and A. Schmaier for helpful discussion and Annette Eckardt for technical assistance in a few experiments and Patricia Joseph for typing this manuscript.

Note Added in Proof-Preliminary amino acid sequencing of tri- gramin suggests the chain is composed of 72 residues, significantly fewer than the estimate of 86 made from the amino acid composition.

1. 2.

3.

4.

5.

6.

7. 8.

9.

10. 11.

12.

13.

14.

REFERENCES

Peerscke, E. I. B. (1985) Semin. Hematol. 2 2 , 241-259 Marguerie, G. A., Plow, E. F., and Edgington, T. S. (1979) J. Biol.

Bennett, J. S., and Vilaire, G. J. (1979) J. Clin. Inuest. 6 4 , 1393-

Kornecki, E., Niewiarowski, S., Morinelli, T. A., and Kloczewiak,

Plow, E. F., and Marguerie, G. A. (1980) J. Biol. Chem. 2 5 5 ,

Hawiger, J., Parkinson, S., and Timmons, S. (1980) Nature 2 8 3 ,

Plow, E. F., and Marguerie, G. A. (1980) Blood 56,553-555 Bennett, J. S., Vilaire, G., and Burch, J. W. (1981) J. Clin. Inuest.

Morinelli, T. A., Niewiarowski, S., Kornecki, E., Figures, W . R., Wachtfogel, Y., and Colman, R. W. (1983) Blood 61,41-49

Colman, R. W. (1986) Semin. Hematol. 2 3 , 119-128 Greenherg, J., Or, J. L., Packham, M. A., Guccione, M. A.,

Harfenist, E. J., Kinlough-Rathbone, R. L., Perry, D. W., and Mustard, J . F. (1979) Blood 5 4 , 753-765

Niewiarowski, S., Budzynski, A. Z., Morinelli, T. A., Brudzynski, T. M., and Stewart, G. J. (1981) J. Biol. Chem. 256,917-925

Kornecki, E., Ehrlich, Y. H., DeMars, D. D., and Lenox, R. H. (1986) J . Clin. Inuest. 7 7 , 750-756

Peerschke, E. I. B., and Coller, B. S. (1984) Blood 64,59-63

Chem. 254,5357-5363

1401

M. (1981) J. Biol. Chern. 2 5 6 , 5696-5701

10971-10977

195-197

68,981-987

15. McGregor, J. L., Clezardin, P., James, E., McGregor, L., Dechav- anne, M., and Clemetson, K. J. (1985) Eur. J . Biochem. 148 ,

16. Coller, B. S., Peerschke, E. I. B., Scudder, L. E., and Sullivan, C.

17. McEver, R. P., Bennett, E. M., and Martin, M. N. (1983) J. Biol.

18. Shattil, S. J., Hoxie, J. A., Cunningham, M., and Brass, L. F.

19. Jennings, L. K., and Phillips, D. R. (1982) J. Biol. Chem. 2 5 7 ,

20. Brass, L. F., Shattil, S. J., Kunicki, T. J., and Bennett, J. S. (1985) J. Biol. Chem. 260 , 7875-7881

21. Pidard, D., Montgomery, R. R., Bennett, J . S., and Kunicki, T. J. (1983) J. Biol. Chem. 2 5 8 , 12582-12586

22. Coller, B. S. (1985) Clin. Inuest. 76 , 101-108 23. Bennett, J . S., Hoxie, J. A., Leitman, S. F., Vilaire, G., and Cines,

D. B. (1983) Proc. Natl. Acad. Sci. U. S. A . 80, 2417-2421 24. McEver, R. P., Baenziger, N. L., and Majerus, P. W . (1980) J.

Clin. Inuest. 6 6 , 1311-1318 25. Newman, P., Allen, R. W., Kahn, R. A., and Kunicki, T. J. (1985)

26. Gartner, T. K., and Bennett, J. S. (1985) J. Biol. Chem. 2 6 0 ,

27. Plow, E. F., Pierschbacher, M. D., Ruoslahti, E., Marguerie, G. A., and Ginsberg, M. H. (1985) Proc. Natl. Acad. Sci. U. S. A .

28. Kloczewiak, M., Timmons, S., Lukas, T. J., and Hawiger, J .

29. Ouvane. C.. and Huane. T.-F. (1983) Biochim. BioDhvs. Acta

97-106

A. (1983) J. Clin. Inuest. 72 , 325-338

Chem. 258,5269-5275

(1985) J. Biol. Chem. 2 6 0 , 11107-11114

10458-10466

Blood 6 5 , 227-232

11891-11894

82,8057-8061

(1984) Biochemistry 23 , 1767-1774

757y3321341 I. . , . "

30. Huane. T.-F.. and Ouvane. C. (1984) Thromb. Res. 32 . 125-138 31. Lowri,' 0. H.; Rosebrough; N. J., Farr, A. L., and Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275 32. Tarr, G. E. (1986) in Microcharacterization of Polypeptides: A

Practical Manual (Shively, J . E., ed) pp. 155-194, Humana Press, Inc., Clifton, NJ

33. Lipinska, I., Lipinski, B., and Gurevich, V. (1974) J. Lab. Clin. Med. 84,509-516

34. Laemmli, U. K. (1970) Nature 2 2 7 , 680-685 35. Morrissey, J. H. (1981) Anal. Biochem. 117 , 307-310 36. Niewiarowski, S., Kornecki, E., Hershock, D., Tuszynski, G. P.,

Pidard, D., Bennett, J. S., Soria, C., Soria, J., Dunn, D., Kieffer, N., and Nurden, A. T. (1985) J. Lab. Clin. Med. 106,651-660

37. Mustard, J. F., Perry, D. W., Ardlie, N. G., and Packham, M. A. (1972) Br. J . Haematol. 2 2 , 193-204

38. Kornecki, E., Tuszynski, G. P., and Niewiarowski, S. (1983) J. Bid. Chem. 285,9349-9356

39. Born, G. V. R., and Cross, M. T. (1963) J. Physiol. (Lond.) 168,

40. Ardlie, N. G., and Han, P. (1974) Br. J . Haematol. 2 6 , 331-356 41. Scatchard, G. (1949) Ann. N . Y. Acad. Sci. 5 1 , 660-672 42. Segal, I. H. (1976) Biochemical Calculations, 2nd Ed., p. 50, John

43. Brown, J. H., and Bowles, M. E. (1966) Toxicon 3, 205-212 44. Marguerie, G. A., Edgington, T. S., and Plow, E. F. (1980) J. Biol.

45. Coller, B. S. (1986) J. Cell Biol. 103 , 451-456 46. Ruggeri, Z. M., Houghten, R. A., Russel, S. R., and Zimmerman,

T. S. (1986) Proc. Natl. Acad. Sci. U. S. A . 8 3 , 5708-5712 47. Coller, B. S., Folts, J . D., Scudder, L. E., and Smith, S. R. (1986)

48. Hanson, S., Harker, L., Ruggeri, Z., and Zimmerman, T. (1986)

178-195

Wiley and Sons, New York

Chem. 255 , 154-161

Blood 6 8 , 783-786

Circulation 7 4 , Suppl. 11, 143 (abstr.)

Trigramin Interaction with Glycoprotein IIb-IIIa Complex 16163

I Glu 270 7 ASP 230

ReriduerlMole I

14 12 B 2

12 I 3

1

86

8820