WO1990015072A1 - Platelet aggregation inhibitors and related molecules - Google Patents

Abstract

A composition of matter comprising specified purified amino acid sequences. A process for making such a composition of matter. Isolated nucleic acid encoding the composition; an expression vector containing the isolated nucleic acid; a cell containing the expression vector. A method for reducing platelet aggregation in a mammal by administering the composition to the mammal in pharmaceutically effective amount. A class of mutant proteins having platelet aggregation inhibition activity. A method for reducing platelet aggregation in a mammal by administering the mutant proteins to the mammal in a pharmaceutically effective amount.

Description

PLATELET AGGREGATION INHIBITORS AND RELATED MOLECULES

This application is a continuation-in-part of U.S.

Application Serial Number 07/362,718, filed with the U.S. Patent and Trademark Office on June 1 , 1989, the entire disclosure of which is hereby expressly incorporated herein by reference thereto.

BACKGROUND OF THE INVENTION

Field of the Invention The present invention is directed to inhibitors of platelet aggregation which prevent binding to the fibrinogen receptor, the genes encoding the inhibitors, expression of the genes, and therapeutic applications of the inhibitors in diseases for which blocking intercellular adhesion mediated by the fibrinogen receptor is indicated.

More specifically, the present invention is directed to inhibitors which, by blocking the binding of fibrinogen to the platelet glycoprotein receptor Ilbllla, antagonize the final common pathway of platelet activation and act as potent antithrombotics.

Description of Background and Relevant Materials

Snake venoms have long been known to contain a variety of factors exhibiting both coagulant and anticoagulant properties on blood. A 72-residue snake venom protein named trigramin, isolated from Trimeresurus crramineus. has been shown to possess the antiplatelet activity of the venom. Trigramin (i) inhibits platelet aggregation; (ii) binds to platelets in a specific manner; (iii) does not bind to platelets from patients with Glanzmann's thrombasthenia; and (iv) inhibits fibrinogen, RGD peptides, and monoclonal antibodies to GP Ilbllla from binding to platelets, implicating trigramin as a GP Ilbllla antagonist. Huang, T.F., Holt, J.C., Lukasiewicz, H. , and Niewiarowski, S. (1987) J. Biol. Chem. 262, 16157-16163; Huang, T.F., Holt, J.C., Kirby, E.P., and Niewiarowski, S. (1989) Biochemistry 28, 661-666; Ouyang, C. and Huang, T.F. (1983) Biochem. Biophys Acta 757, 332,341.

Echistatin, a 49-residue platelet aggregation inhibitor protein from the venom of Echis carinatus. has recently been isolated, fully seguenced, and also characterized as a platelet aggregation inhibitor. Gan, Z.R. , Gould, R. J. , Jacobs, J. ., Friedman, P.A., and Polokoff, M.A., (1988) J. Biol. Chem. 263, 19827-19832; Teng, C.-M. , and Ma, Y.-H. (1988) in Hemostasis and Animal Venoms. H. Pirkle and F.G. Markland, Jr., eds., ch. 29, pp. 399-409.

Precise identification and characterization of the anticoagulant factors present in snake venom is complicated by the difficulties of obtaining sufficient raw material for study, as well as by the number and complexity of the factors in the venom. Moreover, notwithstanding that the anticoagulant factors may be extremely potent, they are usually present at very low concentrations in the venom.

These difficulties are compounded by the extreme complexity of the clotting system. While it is not necessary for purposes of the present application to detail the multiple biochemical pathways that constitute the clotting process, a large number of extrinsic and intrinsic factors, some contributing to more than one pathway, are involved and are not fully understood. The multiplicity of elements contributing to clotting, and the consequent number of possible interactions and feedback loops, makes it extremely difficult to accurately assess the effects and mechanism of an exogenously added agent. One of the results of activating the clotting pathways is the activation of platelets. This leads in turn to platelet aggregation, and to the formation of platelet plugs which adhere to the walls of blood vessels. This process is central to efficient blood clotting, as evidenced by the incidence of severe spontaneous bleeding episodes in thrombocytopenic patients and patients with inherited defects in platelet aggregation.

As previously indicated, multiple factors, including a variety of physiologic stimuli and soluble mediators, initiate activation via several pathways. These have a common final step, which entails the exposure on the platelet surface of the GP Ilbllla receptor and its subsequent binding to fibrinogen. An inhibitor that prevents this binding would antagonize platelet activation by any stimulus and would have important antithrombotic properties.

Many common human disorders are characteristically associated with a hypercoagulable state leading to intravascular thrombi and e boli. These are a major cause of medical morbidity, leading to phlebitis, infarction, and stroke, and of mortality, from stroke and pulmonary and cardiac emboli. A large percentage of such patients have no antecedent risk factors, and develop venous thrombophlebitis and subsequent pulmonary emboli without a known cause. Other patients who form venous thrombi have underlying diseases known to predispose to these syndromes.

Some of these patients may have genetic or acquired deficiencies of factors that normally prevent hyper- coagulability, such as anti-thrombin 3. Others have mechanical obstructions to venous flow, such as tumor masses, that lead to low flow states and thrombosis. Patients with malignancy have a high incidence of thrombotic phenomena, for unclear reasons. Antithrombotic therapy in this situation with currently available agents is dangerous and often ineffective.

Patients with atherosclerosis are predisposed to arterial thromboembolic phenomena for a variety of reasons.

Atherosclerotic plaques form niduses for platelet plugs and thrombi that lead to vascular narrowing and occlusion, resulting in myocardial and cerebral ischemic disease. Thrombi that break off and are released into the circulation cause infarction of different organs, especially the brain, extremities, heart and kidneys. After myocardial infarctions, clots can form in weak, poorly functioning cardiac chambers and be released into the circulation to cause emboli. All such patients with atrial fibrillation are felt to be at great risk for stroke and require antithrombotic therapy.

Patients whose blood flows over artificial surfaces, such as prosthetic synthetic cardiac valves or through extracorporeal perfusion devices, are also at risk for the development of platelet plugs, thrombi, and emboli. It is standard practice that patients with artificial cardiac valves be chronically anti-coagulated.

Thus, a large category of patients, including those with cancer, atherosclerosis, coronary artery disease, artificial heart valves, and a history of stroke, phlebitis, or pulmonary emboli, are candidates for limited or chronic anticoagulation therapy. However, this therapy is often ineffective or morbid in its own right. This is partially because the number of available therapeutic agents is limited and these, for the most part, act by reducing levels of circulating clotting factors. These agents are, therefore, not necessarily aimed at the patient's underlying hematologic problem, which often concerns an increased propensity for platelet aggregation and adhesion. They also cause the patient to be very susceptible to abnormal bleeding. Available anti-platelet agents, such as aspirin, inhibit the cyclooxygenase-induced activation of platelets only and are often inadequate for therapy.

An agent which effectively inhibited the final common pathway of platelet activation, namely fibrinogen binding, should accordingly be useful in a large group of disorders characterized by a hypercoagulable state. There are many potential sites of inhibition of platelet aggregation, including phospholipase A₂, fibrinogen, thrombin, adenylate cyclase, cyclooxygenase, thromboxane synthase or receptor, and GP Ilbllla. There has been considerable recent interest in the GP Ilbllla receptor, the most thoroughly studied member of the integrin family of cell adhesion receptors, since the interaction of fibrinogen with GP Ilbllla represents a final common pathway leading to platelet aggregation.

A class of proteins has now been identified whose members appear to function as. generalized inhibitors of binding to receptors that- bind ligands containing the RGD amino acid sequence. As this binding characterizes the adherence of cells to basement membrane proteins and is felt to be a necessary and important step in the invasion and metastasis of tumor cells, such inhibitors are also of interest as potential antineoplastic agents.

SUMMARY OF THE INVENTION The present invention is generally directed to a composition of matter comprising a purified amino acid sequence selected from the group consisting of RARGDDM^*DDY, IPRGDM*PDDR, ARGDDLDDY, and IARGDWNDDY, where M^* is methionine sulfoxide and M* is ethionine or methionine sulfoxide. Alternatively, the purified amino acid sequence may be RARGDDMDDY, in which instance the composition of matter contains essentially no composition of matter containing the carboxy-ter inal amino acid sequence PRNPHKGPAT. The composition of matter of the present invention may also, of course, contain numerous additional substances, such as adjuvants, buffers, and preservatives.

When the purified amino acid sequence is RARGDDLDDY, the carboxy-terminal sequence of the purified amino acid sequence may be CPRNPFH, CPRNPFHA, or CPRNPLHA; moreover, in this instance, the composition may contain essentially no composition of matter containing the amino acid sequence

RIARGDDLDDY. The amino acid sequence may consist essentially of pyroglutamate-CESGPCCRNCKFLKEGTICKRARGDDM*DDYCNGKTCDCPRNPHKGP;

GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPRGDM^DDRCTG

QSADCPRYH; EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCCDQCSFMKKGTICRRA-

RGDDLDDYCNGRSAGCPRNPFH; EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCC- DQCSFMKKGTICRRARGDDLDDYCNGRSAGCPRNPFHA; EAGEDCDCGSPANPCCDA-

ATCKLLPGAQCGEGLCCDQCSFMKKGTICRRARGDDLDDYCNGISAGCPRNPLHA;

SPPVCGNKILEQGEDCDCGSPANCQDRCCNAATCKLTPGSQCNYGECCDQCRFKKAGTVC-

RIARGDWNDDYCTGKSSDCPWNH; or, mixtures thereof.

The purified amino acid sequence of the present invention may be an inhibitor of platelet aggregation. More specifically, the purified amino acid sequence may inhibit the binding of fibrinogen to platelets, and may bind to the platelet GPIIb-GPIIIa complex.

The composition of matter may be of sufficient purity to yield an IC50 of no more than about 4.2 nM in an ELISA binding assay. The IC50 may be between about 1.3 and 4.2 nM, and in particular may be between about 1.4 and 3.0 nM. As an alternative measure of effect, the composition of matter of the present invention may be of sufficient purity to yield an IC50 of no more than about 610 nM in an assay for inhibition of platelet aggregation; the IC50 may be between about 80 and 610 nM, or between about 105 and 555 nM.

The composition of matter to which the present invention is directed may be derived by purifying snake venom, which may be obtained from snakes of the following genuses and species: Trimeresurus; Trimeresuruscrramineus; Aqkistrodon; Aqkistrodon rhodostoma; Bitis; Bitis arietans; Echis; and, Echis carinatus.

In another embodiment, the present invention is directed to a composition of matter including isolated nucleic acid encoding the protein component of the purified amino acid sequence described above. This nucleic acid may include a sequence encoding the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPFH. Alternatively, the nucleic acid may encode the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPFHA; or, the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPLHA.

In a further embodiment, the isolated nucleic acid may include a nucleic acid sequence encoding the amino acid sequence consisting essentially of any one of QCESGPCCRNCKFL- KEGTICKRARGDDMDDYCNGKTCDCPRNPHKGP,GKECDCSSPENPCCDAATCKLRPGA- QCGEGLCCEQCKFSRAGKICRIPRGDMPDDRCTGQSADCPRYH, EAGKDCDCGSPANPCC- DAATCKLLPGAQCGEGPCCDQCSFMKKGTICRRARGDDLDDYCNGRSAGCPRNPFH, EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCCDQCSFMKKGTICRRARGDDLDDYCN- GRSAGCPRNPFHA, EAGEDCDCGSPANPCCDAATCKLLPGAQCGEGLCCDQCSFMKKG- TICRRARGDDLDDYCNGISAGCPRNPLHA, and SPPVCGNKILEQGEDCDCGSPA- NC_3DRCC^AATCKLTPGSQCNYGECCDQC_^KKAGTVCRIARGDWNDDYCTGKSSDCPWNH. The composition of matter of the present invention may be made by the process of isolating nucleic acid encoding the protein portion of the purified amino acid sequence; ligating the nucleic acid into a suitable expression vector capable of expressing the nucleic acid in a suitable host; transforming the host with the expression vector into which the nucleic acid has been ligated; culturing the host under conditions suitable for expression of the nucleic acid, whereby the purified amino acid sequence is produced; and purifying the amino acid sequence from the host. In another embodiment, the present invention extends to an expression vector comprising nucleic acid encoding the protein component of the purified amino acid sequence as described above. The present invention also encompasses a cell containing the expression vector. This cell may be eukaryotic or prokaryotic; E. coli is an example of a suitable prokaryotic cell.

In a further embodiment, the present invention is directed to a composition of matter which includes a first polypeptide having the amino acid sequence R234GD789, wherein 2, 3, 1 , 8, and 9 may be any naturally-occurring L-amino acid, the same or different, and 4 may be any naturally-occurring L-amino acid other than arginine, wherein the amino acid sequence is an inhibitor of platelet aggregation. The amino acid sequence may have an IC₅₀ (nM) of no more than about 15,000 in a platelet aggregation assay. Preferably, the IC₅₀ is no more than about 10,000, and it is most preferably no more than about 1300.

The amino acid sequence may further inhibit the binding of fibrinogen to GP Ilb-IIIa, and may have an IC₅0 of no more than about 18.1 in a Fg/GP Ilb-IIIa ELISA. Preferably, the IC₅0 is no more than about 6. Moreover, the amino acid sequence may bind to the platelet GP Ilb-IIIa complex.

Within the amino acid sequence of the first polypeptide, 4 may be selected from the group consisting of histidine, lysine, alanine, glycine, isoleucine, leucine, proline, valine, asparagine, cysteine (either reduced or linked in a disulfide bond) , glutamine, methionine, serine, and threonine. Preferably, 4 is selected from the group consisting of alanine, lysine, asparagine, glutamine, and histidine.

In a further embodiment, and still with reference to the amino acid sequence of the above-described first polypeptide, it is preferred that 2 is isoleucine or an analog thereof, 3 is proline or an analog thereof, 7 is methionine or an analog thereof, 8 is proline or an analog thereof, and 9 is aspartic acid or an analog thereof. Most preferably, 2 is isoleucine, 3 is proline, 7 is methionine, 8 is proline, and 9 is aspartic acid. The first polypeptide may include any of the amino acid sequences RAGKICRIP4GDMPDDRCTGQ; LCCEQCKFSRAGKICRIP4GDMPDDRCT- GQSADCPR; GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIP4GD- MPDDRCTGQSADCPRYH; GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAG- KICRIPAGDMPDDRCTGQSADCPRYH; GKECDCSSPENPCCDAATCKLRPGAQCGEGLCC- EQCKFSRAGKICRIPKGDMPDDRCTGQSADCPRYH; GKECDCSSPENPCCDAATCKLRPG- AQCGEGLCCEQCKFSRAGKICRIPNGDMPDDRCTGQSADCPRYH; GKECDCSSPENPCCD- AATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPQGDMPDDRCTGQSADCPRYH; GKECDC- SSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPHGDMPDDRCTGQSADCP- RYH; and combinations thereof.

In an additional embodiment, the present invention includes a composition of matter comprising a second polypeptide having the amino acid sequence 123R5D789, wherein 1, 2, 3, 7, 8, and 9 may be any naturally-occurring L-amino acid, the same or different, and 5 may be any naturally- occurring L-amino acid other than glycine. The amino acid sequence is an inhibitor of platelet aggregation, and may have an IC₅₀ (nM) of no more than about 300 in a platelet aggregation assay. Moreover, the amino acid sequence may inhibit the binding of fibrinogen to GP Ilb-IIIa, and preferably has an IC₅₀ of no more than about 3 in a Fg/GP Ilb- IIIa ELISA.

Moreover, the amino acid sequence may bind to the platelet GP Ilb-IIIa complex. Preferably, in the amino acid sequence of the second polypeptide, 5 is selected from the group consisting of alanine, isoleucine, leucine, proline, and valine. Moreover, 1 may be arginine or an analog thereof; 2 may be isoleucine or an analog thereof; 3 may be proline or an analog thereof; 7 may be methionine or an analog thereof; 8 may be proline or an analog thereof; and, 9 may be aspartic acid or an analog thereof. Preferably, 1 is arginine, 2 is isoleucine, 3 is proline, 7 is methionine, 8 is proline, and 9 is aspartic acid.

The second polypeptide may include any of the amino acid sequences RAGKICRIPR5DMPDDRCTGQ, LCCEQCKFSRAGKICRIPR5DMPDDRCT- GQSADCPR, GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPR5D- MPDDRCTGQSADCPRYH, and GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKF- SRAGKICRIPRADMPDDRCTGQSADCPRYH.

In yet another embodiment, the present invention encompasses a composition of matter comprising a third polypeptide having the amino acid sequence 12345D789, wherein 1, 2, 3, 7, 8, and 9 may be any naturally-occurring L-amino acid, the same or different; 4 may be any naturally-occurring L-amino acid other than arginine; and, 5 may be any naturally- occurring L-amino acid other than glycine. This amino acid sequence is an inhibitor of platelet aggregation, which may have an IC₅₀ (nM) of no more than about 15,000 in a platelet aggregation assay. Alternatively, and in increasing order of preference, the IC₅0 (nM) may be no more than about 10,000; no more than about 1300; no more than about 400; and, no more than about 300.

The amino acid sequence of the third polypeptide may inhibit the binding of fibrinogen to GP Ilb-IIIa. The amino acid sequence may have an IC₅₀ of no more than about 18.1 in a Fg/GP Ilb-IIIa ELISA. Preferably, the IC₅₀ is no more than about 6, and most preferably it is no more than about 3.

Moreover, the amino acid sequence of the third polypeptide may bind to the platelet GP Ilb-IIIa complex.

With regard to the amino acid sequence of the third polypeptide, 4 is preferably selected from the group consisting of histidine, lysine, alanine, glycine, isoleucine, leucine, proline, valine, asparagine, cysteine (either reduced or linked in a disulfide bond) , glutamine, methionine, serine, and threonine; and/or, 5 is selected from the group consisting of alanine, isoleucine, leucine, proline, and valine.

In particular, the third polypeptide may include the amino acid sequence GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKF- SRAGKICRIPKADMPDDRCTGQSADCPRYH.

In a further embodiment, the present invention is directed to a composition of matter comprising a polypeptide having an amino acid sequence selected from any one of the amino acid sequences for trigramin-alpha, trigramin-betal, trigramin-beta2, trigramin-gam a, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the arginine of the RGD sequence within the amino acid sequence is substituted by any naturally-occurring L-amino acid other than arginine. Alternatively, the amino acid sequence may be selected from any one of the amino acid sequences for echistatin- alphal, echistatin-alpha2, trigramin-alpha, trigramin-betal, trigramin-beta2, trigramin-ga ma, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the glycine of the RGD sequence within the amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine. In another alternative embodiment, the amino acid sequence may be selected from any one of the amino acid sequences forechistatin-alphal, echistatin-alpha2, trigramin- alpha, trigramin-betal, trigramin-beta2, trigramin-gamma, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the arginine of the RGD sequence within the amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine, and further wherein the glycine of the RGD sequence within the amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine. The present invention also contemplates a method for reducing platelet aggregation in a mammal. This method involves administering a pharmaceutically effective amount of any of the amino acid sequences and polypeptides as described above to the mammal. The amino acid sequences and polypeptides may be administered to the mammal in admixture with a pharmacologically acceptable adjuvant. This general method may also be applied to treat a mammal whose blood has an increased propensity for clotting.

Finally, the present invention is directed to compositions of matter for, variously, reducing platelet aggregation in a mammal; treating a mammal whose blood has an increased propensity for clotting; or inhibiting fibrinogen binding in a mammal. Each of these compositions contains the amino acid sequences or polypeptides of the present invention, as previously described.

BRIEF DESCRIPTION OF FIGURES Figure 1 depicts the synthetic gene coding for trigramin- gamma.

Figure 2 shows the synthetic oligonucleotide used to repair the partial trigramin-gamma clone.

Figure 3 is the sequence of DNA coding for trigramin-gamma which was inserted in pUC119.

Figure 4 depicts the expression plasmid for trigramin-gamma pTg-gamma.

Figure 5 shows the synthetic gene coding for kistrin. Figure 6 provides the results from the reverse phase HPLC purification of kistrin.

Figure 7 is the reverse phase HPLC purification of echistatin- alpha2.

Figure 8 shows the reverse phase HPLC purification of bitan isoforms. Figure 9 is the reverse phase HPLC purification of trigramin isoforms.

Figure 10 provides the amino acid sequence of native trigramin-gamma and the Lys-C and CNBr fragments used to align the sequence. Figure 11 shows the full amino acid sequences of snake venom GP Ilbllla antagonists. Figure 12 shows the results of Scatchard analysis of ²⁵I- kistrin saturation binding to unactivated and ADP-activated human washed platelets.

Figure 13 shows the results of an assay for inhibition of platelet aggregation (ex vivo) following intravenous administration of kistrin in rabbits.

Figure 14 shows the synthetic gene encoding kistrin, as used in the kistrin mutant studies described herein.

Figure 15 depicts the expression plasmid encoding the Z- kistrin fusion.

Figure 16 shows the purification of kistrin from Z domain and H64A-subtilisin.

Figure 17 provides a comparison of ELISA IC50 values of the alanine scan mutants. DETAILED DESCRIPTION

The present invention concerns the isolation of a group of platelet aggregation inhibitors, their purification and characterization, methods for their production, and their therapeutic uses. The following abbreviations are used herein: GP, for glycoprotein; CNBR, for cyanogen bromide; Lys-C, for endo- proteinase Lys-C; FAB-MS, for fast atom bombardment mass spectrometry; HPLC, for high-performance liquid chromatography; BSA, for bovine serum albumin; IPTG, for isopropyl beta-D- thiogalactopyranoside; Xgal, for 5-bromo-4-chloro-3-indoyl beta-D-galactopyranoside; PRP, for platelet rich plasma; and PPP, for platelet poor plasma. The isolation and further characterization of snake venom proteins as potential GPIIbllla antagonists to inhibit platelet aggregation as detailed herein has revealed a family of isoforms distinct from those previously described. In addition, two new proteins have been isolated: kistrin, from the Malayan pit viper Aqkistrodon rhodostoma (Calloselasma rhodostoma) . previously reported to contain a glycosylated

31,000 dalton platelet aggregation inhibitor (Huang, T.F., Wu,

Y.J. and Ouyang, C. (1987) Biochem. Biophys Acta 925, 248-257); and bitan, from the puff adder Bitis arietans.

The complete amino acid sequences for kistrin, bitan, and several new isoforms of trigramin and echistatin are presented herein. Based on the observed amino acid sequence homologies and activities, it appears that these proteins belong to a family of snake venom proteins that inhibit platelet aggregation acting via the platelet integrin receptor GP Ilbllla.

In addition, the expression of trigramin-gamma and kistrin have been achieved in E. coli using synthetic genes. Purification

To obtain source material, lyophilized Aqkistrodon rhodostoma (Calloselasma rhodostoma) . Trimeresurus gramineus, Bitis arietans, and Echis carinatus snake venoms were purchased from either Sigma (St. Louis, Mo.) or Miami Serpenterium Laboratories (Salt Lake City, Utah) . The snake venom proteins were purified to apparent homogeneity on a G-50 Sephadex gel filtration column followed by reverse phase HPLC using a slight modification of methods previously described (Huang, T.F., Holt, J.C., Lukasiewicz, H., and Niewiarowski, S. (1987) J. Biol. Chem. 262, 16157-16163; Ouyang, C. and Huang, T.F. (1983) Biochem. Biophys Acta 757, 332,341). Specifically, an aliquot of 50 g lyophilized venom was dissolved in 1.5 ml of water and loaded onto a Sephadex G50 (fine) column (100 x 1cm²) equilibrated in either 50 mM Tris HC1, ph 7.5, or 50 mM ammonium bicarbonate, pH 8.5, containing 0.5 mM CaCl₂, and 0.02 % NaN₃> The column was run at a flow rate of 0.1 ml/min; 1.75 ml fractions were collected.

The different active proteins all eluted in the 5,000-10,000 dalton molecular weight range. Fractions were assayed using the GP Ilbllla ELISA, which is described in detail below. Active fractions were pooled and further purified by HPLC using a 5 micron 250 x 4.6 mm Vydac C18 column (218TP-54) , equilibrated in 5-10% acetonitrile, 0.1% TFA, and eluted with an acetonitrile gradient up to 30-45% at 0.5-1.0% per minute while monitoring at A₂₁. The proteins are stable in the acetonitrile/TFA buffer for several months at 4°C, and can also be lyophilized and stored at -20°C.

Three peaks of activity from the venom of A. rhodostoma were detected following reverse phase chromatography; the major protein peak represented 1.5% of the crude venom dry weight (Figure 6) . One peak of activity as observed for E^. carinatus, which was isolated as 0.4% of the crude venom (Figure 7) . There were four major, and several minor, peaks of activity observed after reverse phase HPLC purification of the venom from B. arietans; only one of the major proteins, bitan-alpha, representing 1% of the crude venom, was further characterized (Figure 8) . In contrast to observations by prior researchers, three major and three minor active protein peaks were observed by reverse phase HPLC of T. ramineus venom (Figure 9) . The three major activity peaks, designated Trigramin-betal, Trigramin-beta2, and Trigramin-gamma, were isolated as 0.18%, 0.17%, and 0.4% of the crude venom, respectively. Each protein was purified to homogeneity by re-chromatography on reverse phase HPLC and further characterized as described below. GP Ilbllla ELISA Assay

Activities were monitored using a modified fibrinogen-GP Ilbllla ELISA, which measures the inhibition of fibrinogen binding to purified human platelet GP Ilbllla receptor. The GP Ilbllla ELISA was performed by modification of the method of Nachman and Leung (Nachman, R.L. and Leung L.L.K. (1982) J. Clin. Invest.. 69, 263-269). Human fibrinogen (Kabi) was prepared by the method of Lipinska, I., Lipinski, B. , and Gurewich, V. (1974) J. Lab. Clin. Med. 84, 509-516. Platelet GP Ilbllla was prepared by the method of Fitzgerald, L.A., Leung, B. , and Phillips, D.R. (1985) Anal. Biochem. 151, 169-177, and stored frozen (-80°C) in TACTS buffer, 0.1% Triton X-100. AP3, a murine monoclonal to GP Ilbllla, was a gift from Dr. Peter J. Newman. PBST buffer refers to phosphate buffered saline containing 0.01% Tween 20; TACTS P8241S01 20 buffer contains 20mM Tris.HCl, ph 7.5, 0.02% sodium azide, 2 mM CaCl₂, 0.05% Tween 20, 150 mM NaCl.

Microtiter plates were coated with fibrinogen (10 ug/ml) and then blocked with TACTS, 0.5% BSA. The plate was washed with PBST and the sample to be determined added, followed by addition of solubilized Ilbllla receptor (40 ug/ml) in TACTS, 0.5% BSA. After incubation, the plate was washed and AP3 (1 ug/ml) added. After another wash, goat and anti-mouse IgG conjugated to horseradish peroxidase (Tago) was added. A final wash was performed, and developing reagent buffer (10 mg o-phenylenediamine dihydrochloride, 0.0212% H₂0₂, 22 mM citrate, 50 mM phosphate, pH 5.0) was incubated until color developed. The reaction was stopped with IN H₂Sθ and the absorbance at 492 nm was recorded. The specific activities (IC₅₀) of each of the snake venom proteins¹ several peptides as well as the peptide GRGDS, measured by their ability to inhibit the binding of fibrinogen to purified human platelet GP Ilbllla receptor, are presented in Table- III; the protein concentrations were normalized by amino acid composition analysis and then diluted for analysis in the ELISA. It can be seen that these proteins are potent GP Ilbllla antagonists, exhibiting activities about 100 times greater than that of the peptide GRGDS. Kistrin Comparative Binding Studies Binding studies were performed on kistrin, using both unactivated and ADP-activated washed human platelets, and purified human platelet GPIIbllla. Washed platelets were prepared from PRP incubated with 300 ng PGI₂ for 10 min, followed by apyrase (5 U/ml) for an additional 10 min. The platelet pellet was obtained by consecutive centri-fugation (1600 x g for 7 min at 25° C) and resuspension, first in 10 ml of Tyrode's buffer with 5.5 mM glucose and 2% BSA (Tyrode•s-BSA) containing 300 ng/ml PGI₂ and 1 ϋ/ml apyrase, then in 10 ml Tyrode's-BSA with 0.2 U/ml apyrase, and finally in Tyrode's-BSA. The platelets, diluted to about 50,000 platelets/microliter, were incubated for 2 hr. before use.

Washed platelets at a final concentration of 1.1-1.8 x 10⁷ platelets/ml were incubated with ¹⁵I-kistrin (11.6 microCuries/microgram; 2.5-1000 nM) and 2 mM CaCl₂ in 200 microliters Tyrode's-BSA in the absence (Figure 12, open circles) and presence (Figure 12, filled squares) of 20 microM ADP for 60 min at 25° C. Aliquots (150 microliters) were layered on 750 microliters of Tyrode's-BSA-20% sucrose in eppendorf tubes, centrifuged (14,000 x g for 6 min), the liquid was aspirated, and the cut tips containing the platelet pellet were counted on an Iso-Data series 100 gamma counter. All samples were performed in triplicate. A ¹²⁵I-fibrinogen binding control (+/~ ADP) was done to show ADP stimulation of the platelets. Data derived form the saturation binding curve (Figure 12, inset) was analyzed using Scatchard analysis. (Munson, P.J. and Rodbard, D. (1980) Anal. Biochem. 107, 220- 239.) Scatchard analysis of the binding of ¹⁵I-kistrin to unactivated and ADP-activated platelets yielded Kj values of 10.8 +/" 1-8 nM (72,000 sites), and 1.7 +/" 0.2 nM (79,000 sites) , respectively (Figure 12) . Activation of platelets with ADP results in a shift to increased binding affinity without a significant change in the number of binding sites on the platelet. In contrast, ¹²⁵I-fibrinogen does not bind to unactivated platelets; a _d value of 200-500 nM was measured for saturation or competition binding to ADP- activated washed platelets, consistent with previously reported data.

The binding of kistrin to resting and ADP-stimulated platelets was reversible even after a 60-minute incubation at 22° C. This is unlike either fibrinogen, which demonstrates time-dependent irreversible binding (Marguerie, G.A. , Edgington, T.S., and Plow, E.F. (1980), J. Biol. Chem. 255, 154-161) , or, surprisingly, trigramin-alpha, which was reported to be irreversibly bound after 5 minutes (Huang, T.F., Holt, J.C., Lukasiewicz, H. , and Niewiarowski, S. (1987) J. Biol. Chem. 262. 16157-16163).

The apparent dissociation rate constant of kistrin from activated platelets was about 2.3 times slower than that from unactivated platelets. The K for direct binding of ¹²⁵I- kistrin to the purified GPIIbllla receptor was 120 nM, substantially weaker than that observed for intact platelets. Without being bound by any particular theory, this may be due to a different conformation of the purified receptor or to the use of detergent in the assay. Nonetheless, this provides the first direct evidence that these proteins do in fact act as GPIIbllla antagonists. Platelet Aggregation Assay In addition to the GP Ilbllla ELISA assay, platelet aggregation assays for the different snake venom proteins were performed in human platelet rich plasma (PRP) as follows: 50ml whole human blood (9 parts) was drawn on 3.8% NA Citrate (1 part) from a donor who had not taken aspirin or related medications for at least two weeks. After the blood was centrifuged at 160 x g for 10 minutes at 22°C and allowed to stand for five minutes, the platelet rich plasma was decanted. Platelet poor plasma (PPP) was then decanted from the remaining blood after centrifugation at 2000 x g for 25 minutes at 22°C. The platelet count of the PRP was measured on a Baker 9000 Hematology analyzer and diluted to approximately 300,000 per microliter with PPP.

225ul PRP plus 25ul of either a dilution of the snake venom protein or a PBS control was incubated for five minutes in a Chrono-log Whole Blood Aggregometer at 25°C. An aggregating agent (collagen (1 ug/ml) , U46619 (100 ng/ml) , ADP (8 uM) was added, and the platelet aggregation recorded. Each aggregation was expressed as the percentage of the total aggregation produced when PBS was added to the PRP. The dose-dependent inhibition of platelet aggregation in ADP-stimulated platelet rich plasma was carried out on each of the snake venom proteins and peptides described herein. A very steep dose-dependence was observed for all of the proteins; the IC₅₀ values, which represent the concentration necessary to inhibit aggregation to 50% of the control aggregation, are reported in Table III. In Vitro/In Vivo Animal Studies

In order to further investigate the activity of this protein family, the in vitro and in vivo effects of kistrin in rabbits were studied. Kistrin (1.0 mg/kg) or sterile NaCl were administered to white male New Zealand White rabbits as a 5.0 ml i.v. bolus (n^5 for both groups). The rabbits were anesthetized with a mixture of 7.5 mg fluanisonet and 0.24 mg fentanyl, and blood samples were drawn onto 3.8% Na citrate (9:1) before and at 5, 10, 15, 30, 45, and 60 minutes after dosing (Figure 13) . Platelet-rich plasma was prepared by rapid centrifugation at 14,000 x g of whole blood, and aggregation responses to 10 icrogram/ml collagen were determined using optical aggregometry as described earlier. Results are expressed as percent aggregation compared to the pre-injection sample. The data illustrated are means +/- s.e.m. Platelet counts were constant measured in whole blood using a Baker 9000 Hematology Analyzer.

In vitro rabbit platelet aggregation induced by 10 icrograms/ml collagen was inhibited to 50% of control values by 1.2 +/- 0.1 microM kistrin. For in vivo evaluation, kistrin (1.0 mg/kg) was administered as a single bolus intravenous injection into rabbits, and platelet aggregation was then measured ex vivo at various time points (Fig. 13) . At five minutes, approximately 70% inhibition of collagen- induced aggregation was observed, which returned to control values by 30 minutes; platelet count remained constant throughout the experiment. This transient effect, consistent with reversible binding of kistrin to human GP Ilbllla as discussed earlier, can be accounted for by either proteolytic degradation or by rapid clearance of the active species from the circulation. Protein Composition and Sequence in addition to measuring the activities of these proteins, the proteins themselves were characterized. In order to determine the amino acid compositions, lyophilized samples were cleaved by constant boiling 6N HC1 vapor in the Millipore Picotag system for 20 hrs at 110°C. The hydrolysates were dried on a Savant speed vac concentrator and analyzed on a Beckman model 6300 amino acid analyzer.

The amino acid sequences of kistrin, bitan, and the isoforms of trigramin and echistatin were determined by Edman degradation of the reduced and carboxymethylated proteins and peptides derived by Lys-C proteolytic digestion. (For echistatin-alpha 2, the protein had to be treated with pyroglutamate aminopeptidase before any sequence could be obtained.)

The snake venom proteins (0.5 nmol) were loaded directly onto a model 470A Applied Biosystems gas phase sequencer equipped with a 120A PTH amino acid analyzer. In the case of echistatin-alpha 2, 1 nmol of protein was pretreated with 0.1 ml 50 mM sodium phosphate, pH 7.3, 1 mM EDTA, 10 mM beta-mercaptoethanol, containing 0.5 mg pyroglutamate aminopeptidase (Boehringer) , for 3 hours at 50°C. PTH amino acids were integrated with a Nelson Analytical model 3000 data system; data analysis was carried out on a Vax 11/785 Digital Equipment System (Henzel, W.J., Rodriguez, H. , and Watanabe, C. (1987) J. Chromatogr. 404, 41-52).

For reduction and carboxymethylation, approximately 10 ug of each of the different purified snake venom proteins were dried in vacuo in a Savant speed vac concentrator followed by solubilization in 100 ul of 6 M guanidine HCl, 10 mM Tris pH 8.3, 1 mM EDTA and 2 mM DTT for 1 hr at 37°C. After the addition of iodoacetic acid (6 mM) for 1 hr at 37°C in the dark, the carboxymethylated sample was then isolated by HPLC (VYDAC-C18) using a 0.1% TFA/acetonitrile gradient from 5-10 to 30-35% acetonitrile at 0.5% per minute. The reduced and carboxymethylatedproteins eluted generally about 5-10 minutes later than the native proteins.

After carboxymethylation, alkylation, and purification by HPLC, about 10 ug of each lyophilized snake venom protein was solubilized in 100 ul of 10 mM Tris pH 8.5, 1 mM EDTA, and 10% acetonitrile. Endoproteinase Lys-C was added in a 1:10 (wt:wt) ratio and digested for 6 hours at 37°C. The digest was then chromatographed on a VYDAC C18 column with a 5-10 to 35-70% gradient of acetonitrile in 0.1% TFA at 0.5-1.0% per min. Peaks detected at A₂κ were then analyzed by FAB-MS and sequenced by Edman degradation. For cyanogen bromide (CNBr) digests, 1 nmol of reduced carboxymethylated, HPLC purified, lyophilized snake venom protein was solubilized in 100 ul fof a solution of 1 mg CNBr in 1 ml of 80% formic acid in the dark for 20 hrs at 25°C. After a 1:2 dilution with water, the reaction was separated by HPLC (VYDAC C18) with a 5-10 to 30-65% acetonitrile gradient at 0.5-1.0% per minute. The peptides detected at A₂₁ were analyzed by FAB-MS and sequenced by Edman degradation.

In general, the order of the first two Lys-C peptides were established by NH₂ sequencing of the intact protein. Incomplete digests with Lys-C and digests with CNBr were subsequently performed to provide the overlaps necessary to complete the amino acid sequence analysis. An example of the sequencing analysis is shown in Figure 10 for trigramin-gamma. Amino acid composition analysis of the different proteins agreed within experimental error with that predicted from the protein sequence (Table I) .

The complete sequences for all of the proteins are shown in Figure 11; the sequences of echistatin-alphal and trigramin-alpha, components which have been previously isolated, are provided for comparison purposes. Residues which are conserved in greater than 50% of the sequences are boxed; the conserved cysteine residues are shaded and the RGD region common to each of the inhibitors is in a heavy box. The <Q at the amino terminus of echistatin-alpha2 refers to a pyroglutamyl residue. The methionine at position 28 in echistatin-alpha2 is oxidized to the sulfoxide; the methionine at position 52 of kistrin can be readily oxidized to the sulfoxide by H₂0₂ to produce kistrinox-

FAB-mass spectrometry (FAB-MS) of the isolated peptide fragments and the native molecule was used to confirm the sequence analysis and to verify that the entire sequence had been obtained? this data is shown in Table II.

In order to carry out FAB-MS, protein and peptide fractions (ca. 50 pmol each) were vacuum evaporated to dryness and redissolved in 2 to 3 microliters of 70 percent formic acid. The entire fraction was dried onto the probe tip and resuspended in 1.4 microliters of either thioglycerol or m-nitrobenzyl _jalcohol:70% formic acid (50:50). Fast atom bombardment mass spectra were obtained on a JEOL HX 110/110 tandem mass spectrometer operated in the two sector mode. An excellent correlation was found between the observed and calculated molecular masses of both the peptide fragments and the native protein. In addition, the mass spectral data for the native protein was consistent with all of the cysteines present in their oxidized form as disulfides. Evidence for any alkylation by IAA of the native protein could only be observed after reduction with DTT. Evidence for one of the minor peaks in the kistrin purification was obtained that showed an M + 16 peak for the mass spectrum for both the native protein and the Lys-C4 peptide fragment, which contains one methionine. Oxidation of native kistrin with H₂0₂ yielded a protein with an identical retention time on reverse phase HPLC and mass spectral data to the minor peak. This protein, designated kistrin_0Xf is most likely the Met-52 sulfoxide.

The assignment of the echistatin-alpha 2 sequence containing Met-28 sulfoxide was based on a) the mass spectral data; b) the amino acid sequencing data; c) the pyroglutamyl blocked amino terminus; and d) the lack of cleavage by CNBr, which is indicative of oxidation of the methionine (Means, G.E. and Feeney, F.E., Chemical Modification of Proteins Holden-Day, San Francisco, 1971 pp. 164, 204-25). This sequence differs from echistatin-alpha 1 in that it lacks the Ala-Thr residues at the carboxy terminus; contains an amino pyroglutamyl group, likely derived from cyclization of glutamine, conserved in all of the other proteins at this position; and contains an oxidized methionine. It is difficult to determine the precise in vivo native protein sequence due to the potential for chemical oxidation or hydrolysis during both the venom extraction and protein isolation. Gene Synthesis, Cloning, and Expression Having thus measured the proteins' activities and determined their amino acid compositions and sequences, the use of recombinant techniques to produce them was investigated, starting with trigramin-gamma. Based on the foregoing amino acid composition and sequence analyses, a gene for trigramin-gamma was assembled from 10 synthetic DNA oligonucleotides ranging from 28 to 55 base pairs long and sharing a 27-28 base pair overlap with neighboring oligonucleotides.

The resulting DNA fragment (Figure 1) from Pst I to Sal I was cloned into E. coli using standard methodologies (Maniatis, T. , Fritsch, E.F., Sambrook, J. (1982) Molecular Cloning (Cold Spring Harbor, N.Y.)). The individual DNA oligonucleotides were phosphorylated, annealed together by gradual cooling from 85°C to 4°C, and then ligated into Ml3 mplδ digested with Pst I and Sal I (Yanisch-Perron, C. , Viera, J., and Messing J. (1985) Gene, 33, 103-119).

The ligation mixture was transformed into E. coli JM101 and plated on LB in 2YT top agar containing 40 microM IPTG and 200 micrograms/ml XGAL. Clear plaques were picked and correct insert size verified by miniscreening. The DNA of the clones containing the correct insert size was sequenced (Sanger, F., S. Nicklen, S., and A.R. Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467) . A deletion was repaired in one of the clones by cassette mutagenesis replacing the Xba I to Sal I fragment with a Xba I to Eco RI fragment shown in Figure 2.

In order to cut the clone with Xba I, the entire Pst I to Sal I fragment was subcloned into pUC119 (Viera, J. and Messing, J. (1987) Methods in Enzymology. 153, 3-11) carrying the AMP" marker, to allow DNA preparation from a dam^" strain, GM48 (Backman, K. (1980) Gene. 11, 169-171) . The clone in M13 mplδ was cut out with Pst I and Sal I and ligated into pUC119 also digested with Pst I, Sal I, and calf alkaline phosphatase. The ligation was transformed into GM48 and plated on LB containing 50 micrograms/ml carbenicillin. The correctly constructed clones in GM48 were verified by miniscreening the plasmid DNA. New oligonucleotides (Figure 2) were phosphorylated and ligated into the pUC119 clone digested with Xba I, Eco RI, and calf alkaline phosphatase. The ligation mixture was transformed into JM101 and plated on LB containing 50 micrograms/ml carbenicillin. The correct construction was determined by miniscreening the DNA.

The corrected trigramin-gamma clone, confirmed by dideoxy sequence analysis (Figure 3) , was cut out of the pUC119 vector with Pst I and Ava I and cloned into pB0475 that had been digested with Nsi I, Ava I, and calf alkaline phosphatase to form plasmid pTg-gamma (Figure 4) . The plasmid pB0475 (Wells, J.A. and Cunningham, B.C., U.S.S.N. 07/264,611, filed October 28, 1988) contains the alkaline phosphatase promoter and StII signal sequence (Chang, C.N., Rey, M. , Bochner, B. , Heyneker H., and Gray, G. (1987) Gene 55, 189-196), the origin of replication for the fl phage and pBR322 from base pairs 1205 through 4361 containing the origin of replication and the beta lactamase gene.

A 1% inoculum from a saturated culture of E. coli 16C9 tonA phoA E15 (argF-lac)169 deoC2, a derivative of W3110 (Bennett, W. , Bochner, B., and Chang, C.N. , U.S.S.N. 07/224,520, filed July 26, 1988) cells grown overnight at 37°C in LB media containing 50 micrograms/ml carbenicillin carrying the pTg-gamma plasmid, was added to 20 ml of AP5 media (0.15% glucose, 0.22% casein amino acids, 0.03% yeast extract, 1.6 mM MgS0₄, 20 mM NH₄CI, 1 mM KC1, 70 mM NaCl, 120 mM triethanolamine.HCl, pH 7.4) containing 50 micrograms/ml carbenicillin and grown at 37° for 24 hrs to an OD₅₅₀ of 2.0. The cells were harvested by centrifugation and frozen at -20°C for 1 hour, then subjected to osmotic shock in 1 ml of 10 mM Tris pH 7.5 and mixed at 4°C for 1 hour. The cell debris was removed by centrifugation; trigramin-gamma was purified from the pooled periplasmic supernatant of 16 separate shake flasks by the reverse phase HPLC methods described above.

Aliquots were assayed for activity by the GP Ilbllla ELISA. Recombinant trigramin-gamma was purified by the methods outlined above. The purified material was identical to the natural protein based on a) identical retention time on reverse phase HPLC; b) identical amino terminus (EAGED) ; c) identical molecular weight as measured by FAB-MS; and d) identical specific activity as measured by the GP Ilbllla ELISA. Having thus succeeded in the production of trigramin- gamma by recombinant techniques, the application of these techniques was extended to kistrin. The gene for kistrin was assembled from 8 synthetic DNA oligonucleotides, 38 to 60 base pairs long, sharing an 18 base pair overlap with neighboring oligonucleotides resulting in the DNA fragment shown in Figure 5 from Pst I to Eco RI. The DNA oligonucleotides were phosphorylated, annealed together from 85°C to 4°C, and then ligated into M13 mpl8 cut with Pst I and Eco RI. The ligation mixture was transformed into E. coli JM101 and plated on LB with 2YT top agar containing 40 microM IPTG and 200 micrograms/ml XGAL. Clear plaques were picked and correct insert size verified by miniscreening. The DNA of the clones containing the correct insert size was verified by DNA sequencing.

A clone with the correct sequence was cut from the M13 mpl8 vector with Pst I and Sty I and cloned into pB0475 that had been digested with Nsi I and Sty I and calf alkaline phosphatase. (The plasmid pB0475 is discussed above.) Correct constructions were verified by DNA miniscreening. The pKistrin construction was then transformed into E. coli 16C9 for expression experiments. A 1% inoculum from a saturated culture of E. coli 16C9 tonA phoA E15 (argF-lac)169 deoC2, grown overnight at 37°C in LB media containing 50 micrograms/ml carbenicillin carrying the pKistrin plasmid, was added to 20 ml of AP5 media containing 50 micrograms/ml carbenicillin and grown at 37°C for 24 hrs to an OD₅₅₀ of 2.0. The cells were harvested and osmotically shocked as above. The cell debris was removed by centrifugation and the supernatant, containing the proteins from the periplasmic space, was assayed for activity (Ilbllla ELISA) and kistrin (kistrin radioimmunoassay) . The expression level of kistrin was estimated to be between 0.5 to 5.0 microgram/ l of periplasmic supernatant. Kistrin was purified from the pooled periplasmic supernatant of separate shake flasks by the reverse phase HPLC methods described above.

In view of the successful production of trigramin-gamma and kistrin using recombinant methodologies, it is reasonable to expect that echistatin-alpha2, trigramin-betal, trigramin- beta2, and bitan-alpha may be similarly obtained.

Thus, the present invention has brought to light a family of snake venom platelet aggregation inhibitors, consisting not only of newly-discovered proteins such as bitan and kistrin, but also of previously-undetected and unappreciated isoforms such as trigramin-betal, -beta2, and -gamma, and echistatin- alpha2.

It is evident from Figure 11 and Table II that there is strong sequence homology, as well as essentially identical GPIIbllla antagonist activity, among this family. A Dayhoff data base search for proteins similar to the different snake venom proteins failed to turn up any with substantial homology. The RGD recognition sequence common to many adhesive proteins is always found, as well as the highly conserved cysteines, which are all oxidized and likely to have similar, though not identical, disulfide linkages. The snake venom proteins are about 20 to 100-fold more potent in the inhibition of fibrinogen-GPIIbllla binding than are the peptides GRGDS or GRGDV, indicating that either other determinants on the protein are important for binding, or that the RGD is conformationally restricted in the protein in a manner that enhances binding. It is interesting to note that the methionine sulfoxide in kistrin_ox and echistatin-alpha 2, both near the RGD region, has little effect on activity when compared to the others. Of additional interest is the issue of specificity to other RGD binding integrins, such as the vitronectin alpha_vbeta₃ or fibronectin alphas betai receptors; a recent report has implicated trigramin-alpha as possibly blocking the vitronectin or fibronectin receptors on C32 melanoma cells or fibroblasts, respectively (Knudsen, K.M. , Tuszynski, G.P. Huang, T.F., and Niewiarowski, S. (1988) Exp. Cell Res. 179, 42-49) .

MUTANT KISTRIN STUDIES As previously discussed, all of the proteins in the snake venom family of platelet aggregation inhibitors to which the present invention relates contain an RGD sequence. This sequence is thought to be critical for binding to the receptor (Plow, E. F. and Ginsburg, M. H. (1989) Progress in Hemostasis and Thrombosis. Coller, B. S., ed. , (W. B. Saunders Company, Philadelphia) vol 9, pp. 117-156). The structural-functional relationship of some variants of the RGD sequence has been investigated in small peptides by measuring the inhibition of fibrinogen binding to ADP stimulated platelets (Plow, E.F., Pierschbacher, M. D. , Ruoslahti, E., Marguerie, G.A. , and Ginsburg, M. H. (1985) Proc. Natl. Acad. Sci. USA, 82:8057-8061). The results of this study demonstrated that single conservative substitutions at the arginine, glycine, or aspartic acid positions markedly reduced inhibitory activity, strongly suggesting that the RGD sequence is essential for highly active compounds.

Peptide synthesis has been used to a limited extent in an attempt to assess the importance of certain residues of echistatin in platelet aggregation. (Garsky, V. M. , Lumma, P. K. , Freidinger, R. M. , Pitzenberger, S. M. , Randall, W. C. , Veber, D. F. , Gould, R. J., Freidman, P. A., (189) Proc. Natl. Acad. Sci. USA. 86, 4022-4026) . In this study, when R24 was replaced with ornithine or alanine, 3- and 18-fold decreases, respectively, were observed in inhibition of platelet aggregation. The effect of these same substitutions for arginine in the tetrapeptides RGDF or RGDD appear to be much more pronounced.

The difficulty of assessing the role played by the RGD sequence in platelet aggregation inhibition is compounded by the fact that each of these venom proteins binds with considerably higher affinity (>100 fold) to GP Ilb-IIIa than do shorter peptides such as GRGDS. This may be the result of additional binding interactions provided through residues outside the RGD sequence. Alternatively, conformational restraints placed upon the RGD sequence by other portions of the molecule may hold it in a strict position favorable for binding, such that it represents the only major binding epitope. Thus, in the Garsky et al. study noted above, a possible involvement of the carboxy terminus of echistatin in preventing platelet aggregation was noted, but this effect was observed by deleting this portion of the protein. A deletion of this scope may have an effect on the overall structure of the protein, and may not be representative of the binding interactions with GP Ilb-IIIa. However, the lack of an amino terminal portion, as represented in echistatin, suggests that at least this region is not essential in the other venom proteins for their observed high affinity binding to GP Ilb- IIIa.

Several regions are highly conserved among the snake venom inhibitors (Figure 11) . It is possible that these conserved residues interact with GP Ilb-IIIa to enhance binding. Alternatively, the RGD epitope may simply be presented in a specific conformation that enables tight binding; the fact that certain amino acid replacements among the different venom proteins can be made without significantly affecting binding affinities supports this idea.

The ability to make substitutions in the RGD sequence would be useful for several reasons. It is well known that there are many integrin receptors that interact with RGD peptides. These may not interact with the mutants described herein, and thus it may be possible to 'target' a particular type of receptor through tailoring of a suitable RGD mutant. In addition, the ability to alter the RGD sequence without substantial loss of activity offers the potential to manipulate the antigenicity of this class of platelet aggregation inhibitors. In order to investigate these possibilities, it was necessary to design an amino acid substitution protocol which would permit comparisons of the activities of venom proteins having amino acid substitutions at various locations. Alanine was chosen as the substituting amino acid, and kistrin was used to examine the effects of alanine substitution along the protein chain. Because of the small size and neutral character of alanine, alanine replacements in a protein sequence have been used to test for important interactions between molecules (Cunningham B. C. , and Wells, J. A. (1989) Science 244:1081-1085). Alanine replacements have a minimal affect on a protein's structure, but can serve to eliminate stabilizing interactions of kistrin with GP Ilb-IIIa. Site- directed mutagenesis can be used to replace amino acids with alanine. In addition, other mutations were made to probe the relative importance of the RGD region in this protein.

In general terms, the following study was commenced by constructing a double-stranded DNA plasmid which could both express kistrin as a secreted fusion to the Z domain of protein A, and be used to make a single-stranded DNA template to facilitate site-directed mutagenesis. Site-directed mutagenesis was performed by annealing an oligonucleotide encoding the desired change to the single-stranded DNA template, recreating a double-stranded plasmid in which one strand (containing the oligonucleotide) encoded the desired mutation, and transforming the plasmid into a strain of repair-deficient E. coli. This strain replicates both strands, producing a population of mutant and wild-type plasmids.

The mutant plasmid was isolated and transformed into another strain of E. coli for expression of the mutant Z- kistrin fusion. ^• The fusion was isolated from the cell culture, purified, and enzymatically cleaved to provide the mutant kistrin, which was then characterized. The characterization included DNA sequence, amino acid composition, mass (using FAB-MS) , and activity as measured by the Fg/GP Ilb-IIIa ELISA and platelet aggregation assays.

The materials and methods employed for this study were generally those previously described and discussed herein. Specifically, all enzymes for DNA manipulations were from Bethesda Research Laboratories, except E. coli DNA polymerase 1 large fragment (Klenow) from Boehringer Mannheim and T4 DNA Ligase from New England Biolabs. Native kistrin, fibrinogen, GP Ilb-IIIa, and reagents for the Fg/GP Ilb-IIIa ELISA and platelet aggregation assays were as described previously. Bacterial strains used included E. coli strain JM101 (Δlac-pro. supE thi[F'traD36 proAB+ lacl^Q ZΔM15]) (Messing, J. (1979) Technical Bulletin, NIH Publication No. 79-99, 2, 43-48), 16C9 (an E. coli strain derivative of W3110 (tonA; ATCC27325)) (Bennett, W. , Bochner, B. , and Chang, C.N., U.S.S.N. 07/224,520, filed July 26, 1988), and BMH 71-18 mutL (Kramer, B. , Kramer, W. , and Fritz, H.. -J. (1984) Cell 38:879-887) . Oligonucleotides were synthesized using hydrogen phosphate chemistry (Froehler, B.C., Ng, P. G. , Matteucci, M. D. (1986) Nucleic Acids Res. 14:5399-5407), and purified by poly-acrylamide gel electrophoresis.

All DNA manipulations were performed essentially as previously described (Sambrook, J., Fritsch, E. F. , and Maniatis, T. (1989) Molecular Cloning 2nd Edition, Cold Spring Harbor, N.Y.), unless otherwise indicated. The standard one- letter abbreviation for amino acids is used to identify each kistrin mutant. For example, changing arginine 49 in kistrin to lysine is represented as R49K. Double mutants are named as strings; for example if, in addition to R49K, glycine 50 was changed to alanine, the mutant would be called R49K/G50A. Construction of Synthetic Kistrin Gene

A synthetic gene for kistrin was assembled from 10 synthetic oligonucleotides, ranging from 46 to 65 base pairs long and sharing a 15 base pair overlap with neighboring oligonucleotides. The individual DNA oligonucleotides were phosphorylated, then gradually annealed together from 85°C to 4"C. The resulting Pstl-EcoRI DNA fragment (Figure 14) was ligated into M13 mplδ (Messing, J., Gronenborn, B. , Muller- Hill, B., and HofSchneider, P.H. (1977) Proc. Natl. Acad. Sci. USA, 74:3652-3646., Yanisch-Perron, C. , Viera, J. , and Messing, J. (1985) Gene, 33:103-119) and transformed into E. coli. Clones with the correct insert size were analyzed by dideoxy sequence analysis (Sanger, F. , S. Nicklen, S., and A.R. Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA, 74:5463-5467) until a clone containing the proper sequence was obtained. Construction of Secretion-Expression Plasmid

The secretion-expression plasmid pZkis was constructed by ligating three fragments. The first was a 4.87-kilobase Nsil (filled in using Klenow) to Styl fragment from pB0475 (Wells, J.A. and Cunningham, B.C., U.S.S.N. 07/264,611, filed October 28, 1988). This contains the alkaline phosphatase promoter, Stll signal sequence (Chang, N.C., Rey, M. , Bochner, B. , Heynecker, H. , and Gray, G. (1987) Gene 55:189-196) and pBR322 (Bolivar, F. , Rodriguez, R.L., Greene, P.J., Betlach, M.C., Heynecker, H.L. , and Boyer, H.W. (1977) Gene 2:95-113) from base pairs 1369 through 4361. This fragment also contains the plasmid origin of replication and the β lactamase gene as well as the fl origin of replication from M13 mplδ (Rsal(filled in) to Ahall(filled in)) (Messing et al. 1977) which had been inserted into the pvull site of pBR322. The second fragment was a O.lδ-kilobase Fspl (blunt) to Narl coding for the Z domain of protein A in pZAP (Carter et al. 19δ9 and Nilsson, B. , Moks, T. , Jansson, B., Abrahmsen, L. , Elmblad, A., Holmgren, E., Henrichson, C, Jones, T.A. , and Uhlen, M. (19δ7) Protein Engineering 1:107-113). The third fragment, taken from the M13 mplδ vector described previously, was a 0.23-kilobase Narl to Styl containing the synthetic gene for kistrin with a linker sequence (Figure 14) suitable for cleavage by H64A-subtilisin (Carter et al. 19δ9) (Figure 15) . The Nsil (filled in the Klenow) to Fspl ligation, which put the Z domain reading frame out of phase, was repaired by deleting one base using site directed mutagenesis. The resulting plasmid encodingwild-type kistrin was used to make a single-stranded DNA template, by virtue of the fl origin of replication, in order to perform site-directed mutagenesis. The mutagenized plasmid was then used to express the kistrin mutants. Site-directed mutagenesis was performed according to the following procedure. Construction of Mutant Kistrin Gene

Site directed mutagenesis was performed as described previously (Wells, J.A., Cunningham, B. C. , Graycar, T. P. and Estell, D. A. (1966) Philos. Trans R. Soc. London A 317, 415;^" Kramer et al. (184); Carter, P. Bedoulle, H. and Winter, G. (1985) Nucleic Acids Res. 13:4431-4443). In general, oligonucleotides were designed which encode the desired amino acid change in addition to a silent change which adds or destroys a unique neighboring restriction site. This allowed for convenient selection of the mutant plasmid by using a restriction enzyme targeted for the new unique site.

The oligonucleotide was designed to extend 9 to 12 bases beyond these mismatch changes in both the 3' and 5' directions to facilitate annealing to the single stranded DNA template. The oligonucleotide was phosphorylated, annealed to the template, and then klenow and deoxynucleotide triphosphates (dNTPs) were added to fill in the template. Ligase was added to complete the double stranded plasmid; one strand is the wild-type template, and the other is the mutated copy.

The ligation mixture was transformed into repair deficient E. coli strain BMH 71-18 mutL cells, which then produce plasmid from each of the two strands of DNA, creating a population of wild-type and mutant plasmids. A miniscreen of these cells provided a mixture of both plasmids, which were then restriction selected or restriction purified to enrich for the mutant population. (In restriction selection, if a unique restriction site has been deleted, cutting with the corresponding restriction enzyme will destroy the wild-type plasmid prior to transformation of the mixture into E. coli. In restriction purification, if a unique restriction site has been added, cutting with the corresponding restriction enzyme will linearize the mutant plasmid so that it can be purified by gel electrophoresis, religated, and transformed into E. coli.)

Template was made from clones by the addition of the helper phage K07 and analyzed using by dideoxy sequence analysis (Sanger et al. 1977, Vieira, J. and Messing, J. (1987) Methods in Enzymology 153, 3-11) for the sequence encoding the desired change. Having obtained a plasmid encoding mutant kistrin as a secreted fusion protein, mutant kistrin was then obtained by transferring the plasmid to E. coli. strain 16C9 for expression. Kistrin is expressed as a fusion to the Z domain of protein A, which is secreted into the periplasmic space of E. coli by virtue of the Stll signal sequence. The linkage between the Z domain and kistrin contains the sequence Ala-Pro-Gly-Phe-Ala-His-Tyr-Gly-Lys, where Ala-Pro is the end of the Z domain, Gly-Phe-Ala-His-Tyr is the recognition site for H64A-subtilisin which cleaves the 44 carboxy terminal side of Tyr, and Gly-Lys is the beginning of kistrin (Figure 14) (Carter et al. 1989). Expression and Purification of Mutant Kistrin

Overnight cultures grown in 2YT media (Zoller, M. J. and Smith, M. (1987) Methods in Enzymology 154:329-350) containing 50 ug/ml carbenicillin were used to inoculate (1% inoculum) a 1 L flask containing 250 ml of 2YT, 50ug/ml carbenicillin. The cells were grown for 16 hrs and then harvested by centrifugation and the cell pellets frozen at -20°C for 1 hr. The frozen pellets were resuspended in 12.5 ml of 10 mM Tris pH 7.5 and shaken at 4^βC for 1 hr. The cell debris was removed by centrifugation and the supernatant transferred to a tube containing 0.25 ml of IgG Sepharose (Pharmacia) equilibrated in 50 mM Tris pH 7.5, 150 mM NaCl, and 0.05% Tween 20. The tube was shaken at room temperature for 20 min; the IgG Sepharose was allowed to settle and the supernatant was discarded.

Thus, the IgG Sepharose was washed twice with 10 ml of 50 mM Tris pH 7.5, 150 mM NaCl, and 0.05% Tween 20, and then washed with 10 ml of 0.5 mM NH₄OAC before being transferred to a column. The 2-kistrin fusion was eluted with 1.5 ml of 1 N HOAc and the eluent dried by lyophilization. Dried Z- kistrin fusion was solubilized in 50 μl of 100 mM Tris pH 8.6, 0.05% Tween 20 and digested with H64A-subtilisin (Carter, P., Nilsson, B. , Burnier, J.P., Burdick, D. and Wells, J.A. (1989) Proteins 6:240-248) at a ratio of 10 to 1 (w/w) for 2 days at 37°C. Cleaved kistrin is then purified from the Z domain and H64A-subtilisin by reverse phase HPLC using a 10 to 40% gradient of 0.01% TFA/acetonitrile at 1%/min (Dennis et al. 1990) .

Cleavage with H64A-subtilisin was very specific for the linker region of the fusion protein, and HPLC was an efficient method employed as the last purification step. All the disulfides of kistrin appeared to form spontaneously as determined by FAB-MS and specific activity of r-kistrin; no refolding steps were necessary. Although the specific arrangement of the disulfide bonds in kistrin has not yet been elucidated, it is known that reduction of the disulfides leads to much less active compounds.

All of the kistrin mutants were constructed in pZkis, expressed, and purified as described above. The Z domain fusion allowed for a quick and simple purification step using IgG Sepharose which yielded >95% pure kistrin fusion. After cleavage with H64A-subtilisin, the kistrin mutants could be purified away from other contaminants by HPLC with only minor changes in retention time of the mutants (± 2% acetonitrile) (Figure 16) . The yield from a 250 ml culture grown in 2YT was generally about 40 to 80 μg of mutant, although for a few mutants (D15A, L21A, Q26A, T58A, P65A, H68A) the yield was lower (about 5-10 μg) , and one mutant, F38A, was never observed. Because kistrin has been found to be quite stable in 20% acetonitrile, 0.01% TFA, the mutants were stored in the HPLC eluent until needed. 0 1

46 Chemical Characterization of Mutant Kistrin

Amino acid analysis was used to quantitate each mutant; substitutions in the amino acid composition could readily be detected. Generally, in addition to DNA sequence analysis, mutants were verified by Fast Atom Bombardment-Mass Spectrometry (FAB-MS) for the correct mass; selected mutants from the RGD region are shown in Table IV. FAB-MS which monitored for the correct molecular mass proved to be a rapid and sensitive method for mutant verification. Fast Atom Bombardment-Mass Spectrometry. Protein fractions (ca. 200 pmol each) were vacuum evaporated to dryness and redissolved in 2 to 3 microliters of 70 percent formic acid. The entire fraction was dried on to the probe tip and resuspended in 1.4 μL of either thioglycerol or m-nitrobenzyl alcohol:70% formic acid (50:50) . Fast atom bombardment mass spectra were obtained on a JEOL HXllO/110 tandem mass spectrometer operated in the two sector mode. Amino Acid Analysis. Lyophilized samples were cleaved by constant boiling 6 N HC1 vapor in the Millipore Picotag system for 20 hrs at 110 "C. The hydrolysates were dried on a Savant speed vac concentrator and analyzed on a Beckman model 6300 amino acid analyzer. Characterization of Mutant Kistrin Activity

The kistrin mutants obtained as described above were characterized as to their inhibition of fibrinogen binding to GP Ilb-IIIa, and as to inhibition of platelet aggregation. The conditions for these assays have been previously described herein.

Alanine replacements were made for every amino acid throughout kistrin, except for positions where a glycine or alanine already existed. Cysteine residues were also omitted from the scan since disruption of the disulfides could lead to larger changes than simple intermolecular interactions between kistrin and GP Ilb-IIIa. Results. The IC₅₀ values of each purified alanine mutant in the Fg/GP Ilb-IIIa ELISA are plotted in Figure 17. Changes in the RGD region of kistrin have a significant impact on the inhibition of immobilized fibrinogen binding to GP Ilb-IIIa, while alanine replacements outside of this region seem to have little or no detectable effect. The Fg/GP Ilb-IIIa ELISA IC₅₀ for R49A increased about 5-fold over wild-type kistrin to about 15 nM, while the IC₅₀ for D51A increased over 5000 fold to 17 μM, indicative that D51 is critical for high potency. The mutant G50A showed no change in its ability to inhibit immobilized fibrinogen binding to GP Ilb-IIIa (IC₅₀ = 2.5 nM) . Additional mutants were made in the RGD region of kistrin and are shown in Table V. The effect of these mutations is more apparent in the platelet aggregation assay. In this assay the D51A and D51E mutants were at least 100-fold less active. Other substitutions at D51 lead to much less potent compounds. Mutant R49A was 63-fold less potent; mutant R49K and the double mutant R49K/G50A appear to be nearly as potent as wild-type kistrin in both the Fg/GP Ilb-IIIa ELISA and platelet aggregation assays. Other mutations at R49 also give active, though less potent, compounds.

Replacement of G50 or R49 led to analogs of kistrin that were far more potent than one would have predicted based on the available peptide data. Thus, the G50A mutant was only 2-fold less potent in the platelet aggregation assay than wild-type, whereas the potency of GRADSP is about 23-fold less potent than GRGDSP as measured by a fibrinogen binding assay (Plow et al. (1985)) . In fact, multiple mutations can be made in the RGD region of the protein without significant reduction in activity, as demonstrated by the R49K/G50A analog. Summary

It appears that many substitutions for the amino acid sequence of native kistrin are well tolerated, and lead to compounds of potential use as antithrombotic agents. Based on these studies, it is reasonable to expect that similar substitutions may be made within the RGD sequences of other members of this family of snake venom inhibitors, including but not limited to those whose sequences are provided in Figure 11, which may be expected to substantially retain their inhibition activity. It is also reasonable to expect that, at least, not only the amino acid substitutions actually carried out and described herein, but also substitutions of analogs of those amino acids, would yield comparable results. in defining analogs, the recognized categories of conservative amino acid analogs are aromatic residues (F, H, Y, W) ; charged basic residues (H, K, R) ; charged acidic residues (D, E) ; aliphatic neutral nonpolar residues (A, G, I, L, P, V) ; and aliphatic neutral polar residues (C, M, N, Q, S, T) . Moreover, the term amino acid, as used herein, refers to naturally-occurring L-amino acids normally found in proteins, unless otherwise specifically indicated. The commonly-used one- and three-letter abbreviations for amino acids are used herein (Lehninger, A.L., Biochemistry. 2d ed. , pp. 71-92, (1975), Worth Publishers, N.Y.).

In view of the efficacy of these proteins as inhibitors of fibrinogen binding to the GP Ilbllla receptor, and the feasibility as demonstrated herein of using recombinant techniques to produce them, the present invention may have application in the treatment of a large group of disorders associated with, or characterized by, a hypercoagulable state. Representative of such disorders are genetic or acquired deficiencies of factors which normally prevent hypercoagulability; mechanical obstructions to venous flow, such as tumor masses, prosthetic synthetic cardiac valves, and extracorporeal perfusion devices; atherosclerosis; and coronary artery disease.

The present invention has of necessity been discussed herein by reference to certain specific methods and materials. It is to be understood that the discussion of these specific methods and materials in no way constitutes any limitation on the scope of the present invention, which extends to any and all alternative materials and methods suitable for accomplishing the objectives of the present invention. TABLE I

AMINO ACID COMPOSITION OF SNAKE VENOM PROTEIN GP πbma ANTAGONISTS

isoiπ Tgβ] Tgβ? Tgγ PiTaπg EchP_.

AMINO ACID

Asx i.i- ( 11.4(11) 11.4(11) 11.6(11) 14.4 (14) 8.1 (8)

Thr 1.8 (2) 1.8 (2) 1.8(2) 1.9 (2) 3.7 (4) 2.1 (2)

Ser 3-6 (4) 3.1 (3) 2.8 (3) 3.0 (3) 4-2(5) 1.1 (1)

Glx 7.3(7) 3.7 (4) 3.9 (4) 5.0 (5) 7.6(7) 3-6 (3)

Gly 7.0(7) 9.3 (9) 9.1 (9) 9.2(9) 8.0 (8) 4-9 (5)

Ala 5.2(5) 71(7) 7.9 (8) 8.4(8) 5.2 (5) 1-4 (1)

Cys 12.3 (12) 11.9(12) 12,0 (12) 11.4(12) 13.9 (14) 7.4 (8)

Val 0(0) 0(0) 0(0) 0(0) 2-1 (2) 0(0)

Met 0.8 (1) 0.9 (1) 0.9 (1) 1.0 (1) 0(0) 0.8 (1)

Be 1.8(2) 0.8 (1) 0.8 (1) 1.9 (2) 2.0 (2) 1.0 (1)

Leu 2.1 (2) 3.1 (3) 3.1 (3) 5.0 (5) 2.0 (2) 1-2 (1)

Tyr 1.1 (1) 0.5 (1) 0.5 (1) 0.8 (1) 1.5 (2) 0.9 (1)

Phe 1.0 (1) 1.9 (2) 1.9(2) 0.9 (1) 1.0 (1) 0.9 (1)

His 1.0 (1) 1.1 (1) 1.0(1) 0.9 (1) 1.4(1) 1.0 (1)

Lys 4.2(4) 4.1 (4) 4.1 (4) 2.9 (3) 5.3 (5) 5.1 (5)

Arg 5.7(6) 5.0(5) 5.1 (5) 4.1 (4) 3.5(4) 3.8 (4)

Pro 5.8(6) 5.9(6) 6.0(6) 4.9 (5) 5.0 (5) 3.8 (4)

T N/A f 0) N/A f 0) N/A (0) N/A (0) N/A (2) N/A (0)

Total 67.8(68) 71.6(72) 72.3 (73) 73.0(73) 82.83(83) 47.1(47)

(! residues per mol from composition analysis of acid hydrolysis.

P residues per mol found by sequence analysis are indicated in parentheses.

O) the value reported assumes 2 tryptophans are present. TABLE π

FAB-MASS SPECTROMETRY OF SNAKE VENOM PROTEIN

GP Ilb-IIIa ANTAGONISTS

Observed Calculated Molecular Mass ^a Molecular Mass ^b

(a u) (amu)

Kistrin 7318±2 7318.30

Kistrin_ox 7334±2 7334.30

Bitan-α 8987±4 8989.98

Trigramin-βl 7551±2 7550.49

Trigramin-β2 762312 7621.57

Trigramin-γ 7563±2 7561.51 r-Trigramin-γ 7560±2 7561.51

Echistatin-o2 5243+2 5242.95^c

[a] Protein fractions (ca.50 pmol each) were lyophilized and redissolved in 2 to 3 μl of 70 % formic acid. The entire fraction was dryed onto the probe tip and resuspended in 1.4 μl of either thioglycerol or m-nitrobenzyl alcohol:70% formic acid (50:50). FAB-MS data were obtained on a JEOL HXl 10/110 tandem mass spectrometer operated in the two sector mode. The data reported are corrected for the ionized M+l (H⁺) or M+23 (Na⁺) peak that is observed.

[b] Molecular mass data was calculated from the sequence of the native proteins, assuming all cysteines form disulfide bonds.

[c] Calculated mass assumes an amino terminal pyroglutamate and the methionine oxidized to the sulfoxide. TABLE m GP Ilb-IIIa Antagonist Activity Summary

Fg GP Ilb-ma Human Platelet

Solid Phase ELISA Aggregation Assay

Compound IC₅0, nM (n) IC50, nM (n)

Snake Venomg Kistrin 2.7 ± 1.4 (5) 128 + 35 (4)

135 + 15 (2) (U46619)

105 ±25 (2) (Collagen)

Kistrin o 2.4 + 0.7 (3) 138 ±40 (2) Kistrin red/cm >1500 (1) n.d. Bitan-α 1.8 ± 0.4 (3) 108 ± 2 (2)

Trigramin-βl 3.0 ± 1.2 (4) 300 ± 80 (3)

Trigraπώι-β2 2.3 ± 0.1 (2) 170 (1)

Trigramiπ-γ 2.2 ± 0.2 (2) 240 (1) r-Trigramin-γ 1.7 (1) n.d.

Echistatin-O-2 2.7 ± 0.5 (3) 555 ±55 (2)

Peptide

GRGDS 205 ±70 (3) 225,000 ± 70,000 (3)

S .J3STITUTE SHEET TABLE IV

FAB-MS of Kistrin Mutants in the RGD Region

COMPOUND MASS foa.. MASS fobs.)

R46A 7234 7236.8

I47A 7277 7278.3

P48A 7293 7292.8

R49K 7291 7291.2

R49A 7234 7233.5

G50A 7333 7333.9

D51 E 7333 7335.1

D51A 7275 7275.6

M52F 7335 7334.5

P53A 7293 7292.8

D54A 7275 7274.3

D55A 7275 7273.7

R56A 7234 7233.1

SUBSTITUTE SHEET 54

TABLE V

Claims

WHAT WE CLAIM IS:

1. A composition of matter comprising a purified amino acid sequence selected from the group consisting of RARGDDM^*DDY, IPRGDM^#PDDR, RARGDDLDDY, and IARGDWNDDY, where M^* is methionine sulfoxide and M* is methionine or methionine sulfoxide.

2. A composition of matter comprising the purified amino acid sequence RARGDDMDDY, containing essentially no composition of matter comprising the carboxy-terminal amino acid sequence PRNPHKGPAT.

3. The composition of matter as defined by claim 1 wherein said purified amino acid sequence is RARGDDLDDY, and further wherein the carboxy-terminal sequence of said composition of matter is selected from the group consisting of CPRNPFH, CPRNPFHA, and CPRNPLHA.

4. The composition of matter as defined by claim 3, containing essentially no composition of matter comprising the amino acid sequence RIARGDDLDDY.

5. The composition of matter as defined by claim 1 wherein said purified amino acid sequence is RARGDDM^*DDY, containing essentially no composition of matter comprising the amino acid sequence RARGDDMDDY.

6. The composition of matter as defined by claim 1, wherein said purified amino acid sequence consists essentially of a sequence selected from the group consisting of pyroglutamate-CESGPCCR CKFL EGTICKRARGDDM*DDYCNGKTCDCPRNPHKGP, GKECDCSSPENPCC DAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPRGDM^#PDDRCTG- QSADCPRYH^EAG DCDCGSPANPCCDAATCKLLPGAQCGEGPCCDQCSFMKKGTICRR- ARGDDLDDYCNGRSAGCPRNPFH,EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCC- DQCSFMiπGTICRRARGDDLDDYCNGRSAGCPRNPFHA,EAGEDCDCGSPANPCCDAAT- CKIiPGAQCGEGLCCDQCSFMKKGTICRRARGDDLDDYCNGISAGCPRNPLHA, SPPVCG- KILEO^EDCDCGSPANCQDRCCNAATCKLTPGSQCNYGECCDQCRFKKAGTVCRIARGDW- NDDYCTGKSSDCPWNH, and mixtures thereof, where M* is methionine or methionine sulfoxide.

7. The composition of matter as defined by claim 1, wherein said purified amino acid sequence is an inhibitor of platelet aggregation.

8. The composition of matter as defined by claim 7, wherein said purified amino acid sequence inhibits the binding of fibrinogen to platelets.

9. The composition of matter as defined by claim 8, wherein said purified amino acid sequence binds to the platelet GPIIb-GPIIIa complex.

10. The composition of matter as defined by claim 7, of sufficient purity to yield an IC50 of no more than about 4.2 nM in an ELISA binding assay.

11. The composition of matter as defined by claim 10, of sufficient purity to yield an IC50 of between about 1.3 and 4.2 nM in an ELISA binding assay.

12. The composition of matter as defined by claim 11, of sufficient purity to yield an IC50 of between about 1.4 and 3.0 nM in an ELISA binding assay.

13. The composition of matter as defined by claim 7, of sufficient purity to yield an IC50 of no more than about 610 nM in an assay for inhibition of platelet aggregation.

14. The composition of matter as defined by claim 13, of sufficient purity to yield an IC50 of between about 80 and 610 nM in an assay for inhibition of platelet aggregation.

15. The composition of matter as defined by claim 14, of sufficient purity to yield an IC50 of between about 105 and 555 nM in an assay for inhibition of platelet aggregation.

16. The composition of matter as defined by claim 7, made by purifying snake venom.

17. The composition of matter as defined by claim 16, wherein said venom is derived from snakes of the genus Trimeresurus.

18. The composition of matter as defined by claim 17, wherein said venom is derived from snakes of the species Trimeresurus qramineus.

19. The composition of matter as defined by claim 16, wherein said venom is derived from snakes of the genus Aqkistrodon.

20. The composition of matter as defined by claim 19, wherein said venom is derived from snakes of the species

Aqkistrodon rhodostoma.

21. The composition of matter as defined by claim 16, wherein said venom is derived from snakes of the genus Bitis.

22. The composition of matter as defined by claim 21, wherein said venom is derived from snakes of the species Bitis arietans.

23. The composition of matter as defined by claim 16, wherein said venom is derived from snakes of the genus Echis.

24. The composition of matter as defined by claim 23, wherein said venom is derived from snakes of the species Echis carinatus♦

25. A composition of matter comprising isolated nucleic acid encoding the protein component of the purified amino acid sequence as defined by claim 1.

26. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPFH.

27. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPFHA.

28. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence RARGDDLDDY and the carboxy-terminal amino acid sequence CPRNPLHA.

29. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of QCESGPCCRNCKFLKEGTICKRARGDDMDDYCNGKTCDC- PR PHKGP.

30. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKF- SRAGKICRIPRGDMPDDRCTGQSADCPRYH.

31. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCCDQCSFMKK- GTICRRARGDDLDDYCNGRSAGCPRNPFH.

32. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of EAGKDCDCGSPANPCCDAATCKLLPGAQCGEGPCCDQC- SFMKKGTICRRARGDDLDDYCNGRSAGCPRNPFHA.

33. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of EAGEDCDCGSPANPCCDAATCKLLPGAQCGEGLCCDQC- SFMKKGTICRRARGDDLDDYCNGISAGCPRNPLHA.

34. The composition of matter as defined by claim 25, wherein said isolated nucleic acid comprises a nucleic acid sequence encoding the amino acid sequence consisting essentially of SPPVCGNKILEQGEDCDCGSPANCQDRCCNAATCKLTP- GSQCNYGECCDQCRFKKAGTVCRIARGDWNDDYCTGKSSDCPWNH.

35. The composition of matter as defined by claim 1, wherein said composition is made by the process of: a) isolating nucleic acid encoding the protein portion of said purified amino acid sequence; b) ligating said nucleic acid into a suitable expression vector capable of expressing said nucleic acid in a suitable host; c) transforming said host with said expression vector into which said nucleic acid has been ligated; d) culturing said host under conditions suitable for expression of said nucleic acid, whereby said purified amino acid sequence is produced; and e) purifying said amino acid sequence from said host.

36. An expression vector comprising the isolated nucleic acid as defined by claim 25.

37. A cell comprising the expression vector as defined by claim 36.

38. The cell as defined by claim 37, wherein said cell is eukaryotic.

39. The cell as defined by claim 37, wherein said cell is prokaryotic.

40. The cell as defined by claim 39, wherein said cell is E. coli.

41. A method for reducing platelet aggregation in a mammal, comprising administering a pharmaceutically effective amount of the composition of matter as defined by claim 7 to said mammal.

42. The method as defined by claim 41, further comprising administering said composition of matter to said mammal in admixture with a pharmacologically acceptable adjuvant.

43. A method for treating a mammal whose blood has an increased propensity for clotting, comprising administering a pharmaceutically effective amount of the composition of matter as defined by claim 7 to said mammal.

44. The method as defined by claim 43, further comprising administering said composition of matter to said mammal in admixture with a pharmacologically acceptable adjuvant.

45. A composition of matter for reducing platelet aggregation in a mammal, comprising the composition of matter as defined by claim 7.

46. A composition of matter for treating a mammal whose blood has an increased propensity for clotting, comprising the composition of matter as defined by claim 7.

47. A composition of matter for inhibiting fibrinogen binding in a mammal, comprising the composition of matter as defined by claim 7.

48. A composition of matter comprising a polypeptide having the amino acid sequence R234GD789, wherein 2, 3, 7, 8, and 9 may be any naturally-occurring L-amino acid, the same or different, and 4 may be any naturally-occurring L-amino acid other than arginine, further wherein said amino acid sequence is an inhibitor of platelet aggregation.

49. The composition of matter as defined by claim 48, wherein said amino acid sequence has an IC₅₀ (nM) of no more than about 15,000 in a platelet aggregation assay.

50. The composition of matter as defined by claim 49, wherein the IC₅₀ (nM) is no more than about 10,000.

51. The composition of matter as defined by claim 50, wherein the IC₅₀ (nM) is no more than about 1300.

52. The composition of matter as defined by claim 48, wherein said amino acid sequence inhibits the binding of fibrinogen to GP Ilb-IIIa.

53. The composition of matter as defined by claim 52, wherein said amino acid sequence has an IC50 of no more than about 18.1 in a Fg/GP Ϊlb-Illa ELISA.

54. The composition of matter as defined by claim 53, wherein the IC₅₀ is no more than about 6.

55. The composition of matter as defined by claim 52, wherein said amino acid sequence binds to the platelet GP Ilb- Ilia complex.

56. The composition of matter as defined by claim 48, wherein 4 is selected from the group consisting of histidine, lysine, alanine, glycine, isoleucine, leucine, proline, valine, asparagine, cysteine (either reduced or linked in a disulfide bond) , glutamine, methionine, serine, and threonine.

57. The composition of matter as defined by claim 56, wherein 4 is selected from the group consisting of alanine, lysine, asparagine, glutamine, and histidine.

58. The composition of matter as defined by claim 48, wherein 2 is isoleucine or an analog thereof, 3 is proline or an analog thereof, 7 is methionine or an analog thereof, 8 is proline or an analog thereof, and 9 is aspartic acid or an analog thereof.

59. The composition of matter as defined by claim 58, wherein 2 is isoleucine, 3 is proline, 7 is methionine, 8 is proline, and 9 is aspartic acid.

60. The composition of matter as defined by claim 59, wherein said polypeptide includes the amino acid sequence RAGKICRIP4GDMPDDRCTGQ.

61. The composition of matter as defined by claim 60, wherein said polypeptide includes the amino acid sequence

LCCEQCKFSRAGKICRIP4GDMPDDRCTGQSADCPR.

62. The composition of matter as defined by claim 61, wherein said polypeptide includes the amino acid sequence GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIP4GDMPDDRCT- GQSADCPRYH.

63. The composition of matter as defined by claim 62, wherein said polypeptide includes the amino acid sequence selected from the group consisting of GKECDCSSPENPCCDAATCK- LRPGAQCGEGLCCEQCKFSRAGKICRIPAGDMPDDRCTGQSADCPRYH,GKECDCSSPE- NPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPKGDMPDDRCTGQSACPRYH, GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPNGDMPDDRCT- GQSADCPRYH,GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIP- QGDMPDRCTGQSADCPRYH,GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFS- RAGKICRIPHGDPDDRCTGQSADCPRYH, and combinations thereof.

64. A composition of matter comprising a polypeptide having the amino acid sequence 123R5D789, wherein 1, 2, 3, 7, 8, and 9 may be any naturally-occurring L-amino acid, the same or different, and 5 may be any naturally-occurring L-amino acid other than glycine, further wherein said amino acid sequence is an inhibitor of platelet aggregation.

65. The composition of matter as defined by claim 64, wherein said amino acid sequence has an IC₅0 (nM) of no more than about 300 in a platelet aggregation assay.

66. The composition of matter as defined by claim 64, wherein said amino acid sequence inhibits the binding of fibrinogen to GP Ilb-IIIa.

67. The composition of matter as defined by claim 66, wherein said amino acid sequence has an IC₅₀ of no more than about 3 in a Fg/GP Ilb-IIIa ELISA.

68. The composition of matter as defined by claim 66, wherein said amino acid sequence binds to the platelet GP Ilb- Ilia complex.

69. The composition of matter as defined by claim 64, wherein 5 is selected from the group consisting of alanine, isoleucine, leucine, proline, and valine.

70. The composition of matter as defined by claim 64, wherein 1 is arginine or an analog thereof, 2 is isoleucine or an analog thereof, 3 is proline or an analog thereof, 7 is methionine or an analog thereof, 8 is proline or an analog thereof, and 9 is aspartic acid or an analog thereof.

71. The composition of matter as defined by claim 70, wherein 1 is arginine, 2 is isoleucine, 3 is proline, 7 is methionine, 8 is proline, and 9 is aspartic acid.

72. The composition of matter as defined by claim 64, wherein said polypeptide includes the amino acid sequence RAGKICRIPR5DMPDDRCTGQ.

73. The composition of matter as defined by claim 72, wherein said polypeptide includes the amino acid sequence

LCCEQCKFSRAGKICRIPR5DMPDDRCTGQSADCPR.

74. The composition of matter as defined by claim 73, wherein said polypeptide includes the amino acid sequence GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPR5DMPDDRCTG- QSADCPRYH.

75. The composition of matter as defined by claim 74, wherein said polypeptide includes the amino acid sequence GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPRADMPDDRCTG- QSADCPRYH.

76. A composition of matter comprising a polypeptide having the amino acid sequence 12345D789, wherein: a) 1, 2, 3, 7, 8, and 9 may be any naturally- occurring L-amino acid, the same or different; b) 4 may be any naturally-occurring L-amino acid other than arginine; and, c) 5 may be any naturally-occurring L-amino acid other than glycine, further wherein said amino acid sequence is an inhibitor of platelet aggregation.

77. The composition of matter as defined by claim 76, wherein said amino acid sequence has an IC₅₀ (nM) of no more than about 15,000 in a platelet aggregation assay.

78. The composition of matter as defined by claim 77, wherein the ICso (nM) is no more than about 10,000.

79. The composition of matter as defined by claim 78, wherein the IC₅₀;<_nM) is no more than about 1300.

80. The .composition of matter as defined by claim 79, wherein the IC₅₀ (nM) is no more than about 400.

81. The composition of matter as defined by claim 80, wherein the IC₅₀ (nM) is no more than about 300.

82. The composition of matter as defined by claim 76, wherein said amino acid sequence inhibits the binding of fibrinogen to GP Ilb-IIIa.

83. The composition of matter as defined by claim 82, wherein said amino acid sequence has an IC₅₀ of no more than about 18.1 in a Fg/GP Ilb-IIIa ELISA.

84. The composition of matter as defined by claim 83, wherein the IC₅₀ is no more than about 6.

85. The composition of matter as defined by claim 84, wherein the IC₅₀ is no more than about 3.

86. The composition of matter as defined by claim 82, wherein said amino acid sequence binds to the platelet GP Ilb- IIIa complex.

87. The composition of matter as defined by claim 76, wherein 4 is selected from the group consisting of histidine, lysine, alanine, glycine, isoleucine, leucine, proline, valine, asparagine, cysteine (either reduced or linked in a disulfide bond) , glutamine, methionine, serine, and threonine.

88. The composition of matter as defined by claim 76, wherein 5 is selected from the group consisting of alanine, isoleucine, leucine, proline, and valine.

89. The composition of matter as defined by claim 88, wherein 4 is selected from the group consisting of histidine, lysine, alanine, glycine, isoleucine, leucine, proline, valine, asparagine, cysteine (either reduced or linked in a disulfide bond) , glutamine, methionine, serine, and threonine.

90. The composition of matter as defined by claim 89, wherein said polypeptide includes the amino acid sequence

GKECDCSSPENPCCDAATCKLRPGAQCGEGLCCEQCKFSRAGKICRIPKADMPDDRCTG- QSADCPRYH.

91. A composition of matter comprising a polypeptide having an amino acid sequence selected from any one of the amino acid sequences for trigramin-alpha, trigramin-betal, trigramin-beta2, trigramin-gamma, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the arginine of the RGD sequence within said amino acid sequence is substituted by any naturally-occurring L-amino acid other than arginine.

92. A composition of matter comprising a polypeptide having an amino acid sequence selected from any one of the amino acid sequences for echistatin-alphal, echistatin-alpha2, trigramin-alpha, trigramin-betal, trigramin-beta2, trigramin- gamma, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the glycine of the RGD sequence within said amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine.

93. A composition of matter comprising a polypeptide having an amino acid sequence selected from any one of the amino acid sequences forechistatin-alphal, echistatin-alpha2, trigramin-alpha, trigramin-betal, trigramin-beta2, trigramin- gamma, and bitan-alpha, as shown in Figure 11, or combinations thereof, wherein the arginine of the RGD sequence within said amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine, and further wherein the glycine of the RGD sequence within said amino acid sequence is substituted by any naturally-occurring L-amino acid other than glycine.

94. A method for inhibiting platelet aggregation in a mammal, comprising administering a pharmaceutically effective amount of the composition of matter as defined by any one of claims 48, 60, 71, 79, 80, and 81 to said mammal.

95. The method as defined by claim 94, further comprising administering said composition of matter to said mammal in admixture with a pharmacologically acceptable adjuvant.

96. A method for treating a mammal whose blood has an increased propensity for clotting, comprising administering a pharmaceutically effective amount of the composition of matter as defined by any one of claims 48, 64, 76, 91, 92, and 93 to said mammal.

97. The method as defined by claim 96, further comprising administering said composition of matter to said mammal in admixture with a pharmacologically acceptable adjuvant.

98. A composition of matter for inhibiting platelet aggregation in a mammal, comprising the composition of matter as defined by any one of claims 48, 64, 76, 91, 92, and 93.

99. A composition of matter for treating a mammal whose blood has an increased propensity for clotting, comprising the composition of matter as defined by any one of claims 48, 64, 76, 91, 92, and 93.

100. A composition of matter for inhibiting binding of fibrinogen to platelets in a mammal, comprising the composition of matter as defined by any one of claims 48, 64, 76, 91, 92, and 93.