AU772460B2 - Inhibitors of platelet activation and recruitment - Google Patents

Description

WO 00/23459 PCT/US99/22955

TITLE

INHIBITORS OF PLATELET ACTIVATION AND RECRUITMENT REFERENCE TO RELATED APPLICATIONS The present application is related.to U.S. Provisional Application Serial Nos. 60/104,585, filed 16 October 1998, 60/107,466, filed 06 November 1998, and 60/149,010, filed 13 August 1999.

FIELD OF THE INVENTION This invention relates to soluble CD39 compounds and compositions, the preparation thereof, and the use thereof to inhibit platelet activation and recruitment in a mammal.

BACKGROUND OF THE INVENTION CD39 is a cell-surface antigen that was originally identified as a marker for mature B cells, but is also expressed on less mature B cells, Epstein-Barr Virus-transformed B cells, activated T cells, endothelial cells and some myeloid cell lines (D6rken et al., in Leukocyte Typing IV; W. Knapp, B.

Dirken, and W.R. Gilks, Eds; Oxford University Press, New York, NY; pp. 89-90, 1989).

Monoclonal antibodies against CD39 induce B cell homotypic adhesion, an activity that may be important in the regulation of immune function (Kansas and Tedder, J. Immunol. 147:4094-4102, 1991). Molecular cloning and characterization of CD39 indicated that it is unique cell surface molecule that contains two potential transmembrane regions and a hydrophobic segment within the putative extracellular domain (Maliszewski et al., J. Immunol. 153:3574, 1994). The amino acid sequence of CD39 was reported to exhibit some homology with a guanosine diphosphatase from yeast (Maliszewski et al., supra).

In 1996, an ATP diphosphohydrolase was cloned from potato tubers (Handa and Guidotti, Biochem. Biophys. Res. Commun. 218:916, 1996). The amino acid sequences of this and several other NTPases demonstrated a high degree of similarity, particularly within several small "apyrase conserved regions" (ACR). CD39 shares these conserved regions with soluble ATP-diphosphorylase from potato tubers, other apyrases and related enzymes. It was subsequently reported that native and recombinant full-length CD39 possess E-type ATP diphosphohydrolase (ATPDase) activity (Marcus et al., J. Clin.Invest. 99:1351, 1997); Kaczmarek et al., J. Biol. Chem. 271:33116, 1996); Wang and Guidotti, J. Biol. Chem. 271:9898, 1996). ATPDases degrade nucleoside tri- and/or diphosphates, but not monophosphates (Plesner, Int. Rev. Cytol. 158:141, 1995).

Vascular endothelial cells constituitively express a cell-surface ADPase (ecto-ATP diphosphohydrolase, apyrase, EC one of at least 3 thromboregulatory systems which function in the maintenance of blood fluidity (Marcus and Safier, FASEB J. 7:516, 1983; Marcus et al., J.

Clin. Invest. 88:1690, 1991). This ecto-ADPase, which belongs to the E-type ATPDase family, rapidly metabolizes ADP in the platelet releasate, terminating further platelet recruitment and aggregation.

WO 00/23459 PCT/US99/22955 Immunoprecipitation of HUVEC detergent lysates with anti-CD39 mAb resulted in complete capture of cell-associated ADPase activity, suggesting that CD39 is the only ecto-ADPase on endothelial cells (Marcus et al., J. Clin. Invest. 99:1351, 1997). In the same study, COS cell transfectants expressing recombinant CD39 at the cell surface totally inhibited ADP-induced platelet aggregation. Thus, CD39 plays a prominent role in thromboregulation (see also, Gayle et al., J. Clin.

Invest., 101:1851, 1998).

Excessive platelet activation stimulation by an agonist) and recruitment, leading to platelet aggregation and vessel occlusion at sites of vascular injury in the coronary, carotid, and peripheral arteries, presents a major therapeutic challenge in cardiovascular medicine. Excessive platelet activation and recruitment is a contributing factor in clinical disorders including stroke, unstable angina, myocardial infarction, and restenosis following percutaneous coronary intervention including angioplasty, atherectomy, stent placement, and bypass surgery.

Glycoprotein IIb/IIIa antagonists, such as the monoclonal antibody marketed as ReoPro® (Centocor Inc.), are presently under development for the inhibition of platelet aggregation in patients undergoing percutaneous coronary intervention, and in patients with acute coronary syndromes such as unstable angina and myocardial infarction. The activation of glycoprotein Ib/IIa receptors, however, is a late event in the cascade that leads to platelet aggregation.

There is a great need to identify additional therapeutic strategies and compositions for the pharmacological neutralization of platelet reactivity (activation, recruitment, aggregation). In particular, there is a need to identify compounds and compositions which target early portions of coagulation pathways such as the ADP-dependent activation and recruitment of platelets. There is, in fact, an urgent need to identify new strategies and compositions for the treatment of stroke, which is the third leading cause of death in the United States. In the case of stroke, an advantageous therapeutic agent will reduce intravascular thrombus burden and accompanying neurological defects without increasing intracerebral hemorrhage.

SUMMARY OF THE INVENTION Soluble forms of CD39 having apyrase activity constitute a novel approach to the prevention and/or treatment of disease. The present invention provides soluble CD39 polypeptides and nucleic acids, compositions comprising a pharmaceutically acceptable carrier and a soluble CD39 polypeptide, and methods of making and using soluble CD39 polypeptides having apyrase activity. The effectiveness of soluble CD39 polypeptides have been demonstrated in vitro, ex vivo, and in vivo.

The invention is directed to soluble CD39 polypeptides selected from the group consisting of: polypeptides having an amino acid sequence as set forth in Figure 1 (SEQ ID NO:2) wherein the amino terminus is selected from the group consisting of amino acids 36-44, and the carboxy terminus is selected from the group consisting of amino acids 471-478; fragments of the polypeptides of (a) wherein said fragments have apyrase activity; variants of the polypeptides of or wherein said variants have apyrase activity; and fusion polypeptides comprising the polypeptides of WO 00/23459 PCT/US99/22955 or wherein said fusion polypeptides have apyrase activity. The invention provides compositions comprising a pharmaceutically acceptable carrier and a soluble CD39 polypeptide.

The invention is also directed to nucleic acids encoding a soluble CD39 polypeptide. The invention provides DNAs, vectors, recombinant cells, and recombinant methods for the production of soluble CD39 polypeptides.

The invention is further directed to the use of soluble CD39 polypeptides for inhibiting platelet activation and recruitment, for inhibiting angiogenesis, or for degrading nucleoside tri- and/or di- phosphates in a mammal in need of such treatment. The invention encompasses the use of a soluble CD39 polypeptide for the preparation of a medicament for inhibiting platelet activation and platelet recruitment, for inhibiting angiogenesis, or for degrading nucleoside tri- and/or di- phosphates in a mammal in need of such treatment. These and other aspects of the present invention will become evident upon reference to the following drawings, examples, and detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows the predicted amino acid sequence (SEQ ID NO:2) of human CD39. The predicted amino acid sequence contains 6 potential N-linked glycosylation sites (double underline), and 11 cysteine residues (bold face). The two predicted transmembrane regions are underlined (single underline).

Figure 2 shows the domain structure of full length CD39 and of an engineered soluble form of CD39. The locations of transmembrane regions near the amino- and carboxy-termini, the centrally located hydrophobic sequence, and a section containing the four putative apyrase conserved regions (ACR) are indicated. Cysteine residues are marked as The soluble CD39 contains a FLAG® peptide and new leader sequence and lacks the two transmembrane regions.

Figure 3 shows the immunoaffinity depletion of solCD39 from COS-1 conditioned medium (CM) following one (IX) or two (2X) rounds of adsorption. Samples were assayed for ATPase activity as described in Example 7. Data are expressed as pmoles of ATP degraded per minute.

Figure 4 shows the immunoprecipitation of solCD39 from COS-1 CM. Lane 2 shows the material that specifically bound to the antibody-coated beads. Lane 1 shows material that was preincubated with ovalbumin-coated beads to remove non-specifically bound material prior to addition of Ab-coated beads. Migration of molecular weight standards is indicated in kilodaltons (kDa).

Figure 5 shows the immunoaffinity purification and characterization of soluble CD39 (solCD39). Figure 5A shows fractions from the immunoaffinity column analyzed by SDS-PAGE, Figure 5B shows enzyme activity in the fractions, and Figure 5C shows purified solCD39 before (Lane 1) and after (Lane 2) treatment with N-glycanase.

Figure 6A shows pH optimum profiles of HUVEC membrane ecto-ADPase and recombinant solCD39 Figure 6B shows is an Eadie-Hofstee plot of rates of metabolism at different concentrations of ATP or ADP using purified solCD39 (6.5 ng).

WO 00/23459 PCT/US99/22955 Figure 7 shows inhibition of ADP-induced platelet reactivity by purified solCD39 in plateletrich plasma from a donor who had ingested aspirin. The response to increasing concentrations of ADP is shown in Fig. 7A. The effect of increasing quantities of purified solCD39 on the platelet aggregation response to 10 pM ADP is shown in Fig. 7B. Arrows indicate the addition of agonist.

Data are presented as relative light transmission vs. time (4 min duration).

Figure 8 shows a comparison of platelet reactivity as modulated by different agonists and inhibitors. The effects of CM from cells expressing solCD39 on platelet aggregation induced by 5 pM ADP (Fig. 8A) and collagen (Fig. 8B) were compared in PRP and PRP treated with 10 pM indomethacin. In Fig. 8B, 1 pg/ml collagen was used in the upper samples and 3.3 pg/ml in the lower (indomethacin-treated) samples. Fig. 8C shows the inhibition of collagen-induced aggregation by increasing quantities of solCD39 in PRP from a donor who had ingested aspirin. The arrows indicate the addition of agonist. Data are presented as relative light transmission vs time (4 min.).

Figure 9 shows the effect of FSBA-treated solCD39 on platelet reactivity. Fig. 9A shows the effects of purified solCD39, FSBA-treated solCD39, and mock-treated solCD39 (each at 4.4 pg/ml) on ASA-treated PRP after addition of 10 pM ADP. Fig. 9B shows the effects of FSBA-treated solCD39 and mock-treated solCD39 (each at 22 pg/ml) on ASA-treated PRP following addition of 3.3 pg/ml collagen. Fig. 9C shows the titration of mock-treated solCD39 (0.88-2.2 pg/ml) against FSBAtreated solCD39 (22 pg/ml). ASA-treated PRP was stimulated with 10 pM ADP. Arrows indicate addition of agonist. Data are presented as relative light transmission vs time.

Figure 10 shows pharmacokinetic analyses of solCD39 in mice. CD39 in serum was measured in the radioactive phosphate release ATPase assay or the ADPase assay Activities are expressed as pmoles nucleotide degraded per minute. The dashed line indicates the ATPase activity of 25 pg/ml of solCD39 in murine serum. Distribution (tmax 59 min (ATP); 43 min (ADP)) and clearance (t 1 /2 40 h (ATP ADP)) half-lives were determined using a biphasic curve fit.

Figure 11 shows bleeding times at 0 and 60 minutes in pigs treated with low, medium, or high doses of solCD39.

Figure 12 shows the effect of aspirin on pig platelet aggregation at baseline and day 5 after intravenous administration (Fig. 12A) and the effect of effect of high dose solCD39 on platelet aggregation at baseline and day 7 (Fig. 12B).

Figure 13 shows the inhibition of pig platelet aggregation by low, medium, and high doses of solCD39 as a function of time after bolus administration.

Figure 14 shows the concentration of CD39 in pig serum as a function of time after low, medium, or high dose administration. Distribution (t /2o 29 min) and clearance (t/23 51 h) halflives were determined using a biphasic curve fit.

Figure 15 shows the ex vivo aggregation of murine platelets. Platelets were stimulated with jtM ADP (Fig. 15A), 2.5 gg/ml collagen (Fig. 15B), or 0.1 mM sodium arachidonate (Fig. after the administration of vehicle (saline), soluble CD39 (4 mg/kg) or aspirin (5 mg/kg). Soluble WO 00/23459 PCT/US99/22955 CD39 treatment produced aggregation curves that returned to baseline following stimulation with agonists, but aspirin treatment yielded such a pattern only when arachidonate was the agonist.

Figure 16 shows reversal of the ADP-induced aggregation response in murine platelets when solCD39 is added at the peak of the aggregation response.

Figure 17 shows the inhibition of platelet (n=20, Fig. 17A) and fibrin Fig. 17B) deposition following induction of stroke in mice pretreated with 8 mg/kg soluble CD39. "Fibrin" is a positive control, "Ipsilat" is ipsilateral the ischemic hemisphere), and "Contralat" is the nonischemic hemisphere.

Figure 18 shows the comparative effects of vehicle soluble CD39 (n=67) and aspirin (n=27) on the outcome of induced stroke in mice. Fig. 18A shows cerebral blood flow, 18B shows cerebral infarct volume, 18C shows neurological score (where higher scores indicate a worse deficit (Connolly, Jr., et al., Neurosurg. 38(3):523-532 (1996)), 18D shows mortality, and 18E shows intracerebral hemorrhage. *p<0.05, *p<0.01, tp<0.001.

Figure 19 shows a covariate plot of cerebral infarct volume vs. intracerebral hemorrhage.

Vehicle (saline), aspirin (ASA, 5 mg/kg prior to stroke), soluble CD39 (4 8 mg/kg, prior to stroke), and soluble CD39 (8 mg/kg, 3 h following stroke induction in mice) are compared.

Figure 20A shows the construct used to generate CD39-/- mice by homologous recombination. The labeled restriction sites are BglII SpeI and Asp718 Figure shows a genomic Southern blot as used to identify ES clones having a disrupted CD39 allele.

Figure 21 shows the bleeding times in control aspirin-treated (5 mg/kg, solCD39-treated 8, and 20 mg/kg, n=25) and solCD39-/- mice p<0.01, p<0.001).

Figure 22 shows a comparison of stroke outcomes in control (C57BU6J x 129/J Fl) mice CD39-/- mice and CD39/- mice which were "reconstituted" with solCD39 Figure 22A shows cerebral blood flow, 22B shows cerebral infarct volume, 22C shows neurological score, 22D shows mortality, and 22E shows intracerebral hemorrhage. *p<0.05, *p<0.01, 'p<0.001.

Figure 23 is a Kaplan-Meier plot showing that solCD39 causes an improvement in survival in a stringent lung ischemia-reperfusion model.

Figure 24 shows an alignment of the N-terminal amino acid sequences of human CD39 and human CD39-L4.

DETAILED DESCRIPTION OF THE INVENTION A cDNA encoding the cell-surface molecule CD39 has been isolated, cloned and sequenced.

The nucleic acid sequence and predicted amino acid sequence of this cDNA are shown in SEQ ID NO: 1 and SEQ ID NO:2. The present invention provides methods of using soluble forms of CD39, which were constructed by removing the amino- and carboxy-terminal transmembrane domains.

Soluble CD39 retains the capacity of wildtype CD39 to metabolize ATP and ADP at physiologically relevant concentrations as well as the ability to block and reverse ADP-induced platelet activation and WO 00/23459 PCT/US99/22955 recruitment, including platelet aggregation. The use of soluble forms of CD39 is advantageous because purification of the polypeptides from recombinant host cells is facilitated, and because soluble polypeptides are generally more suitable than membrane-bound forms for clinical administration.

Because CD39 inhibits platelet activation and recruitment, and therefore platelet aggregation, the present invention provides methods and compositions for inhibiting formation of a thrombus at a site in a mammal at which platelets are inappropriately activated, methods for use in controlling platelet reactivity, thereby regulating the hemostatic and thrombotic processes, and methods of inhibiting and/or reversing platelet aggregation.

A. Hemostasis Hemostasis is defined as the arrest of bleeding from damaged blood vessels, and results from a sequence of physiologic and biochemical events. At least three interacting biological systems are involved in hemostasis: components of the blood vessels (such as the subendothelial matrix), platelets, and plasma proteins (Marcus, Disorders of Hemostasis, Ratnoff and Forbes, eds., W.B. Saunders, Philadelphia, 1996; pages 79-137; Marcus, Platelet Activation, in: Atherosclerosis and Coronary Artery Disease, vol.1, Fuster, Ross and Topol, eds., Lipincott-Raven, Philadelphia, 1996; pages 607- 637). A defect or defects in one or more of these systems can result in hemorrhagic disorder; conversely, the inappropriate activation of hemostasis culminates in development of arterial or venous thrombosis.

When a blood vessel is injured, it contracts, exposing subendothelial matrix components such as collagen, von Willebrand factor, fibronectin, thrombospondin, laminin, and microfibrils. Platelets adhere to, and are activated by, these components; collagen is an especially effective agonist for platelet activation. At least four physiologic events are initiated by platelet-collagen contact: the platelets release biologically active compounds; they express P-selectin on their cell surface (where it mediates adhesion of neutrophils, monocytes and subsets of lymphocytes); the platelet eicosanoid pathway is activated (starting with the liberation of arachidonic acid which forms prostaglandin H 2 and the platelets undergo a drastic change in shape, from smooth disks to spiny spheres.

The biologically active compounds released by platelets are numerous, and multi-functional.

Included in this group of components are serotonin, ATP, ADP, calcium, adhesive proteins (fibrinogen, fibronectin, thrombospondin, vitronectin, von Willebrand factor), growth factors (plateletderived growth factor, transforming growth factor-B, platelet factor 4) and coagulation factors (factor V, high-molecular weight kininogen, factor XI, protein S and plasminogen activator inhibitor-I (PAI- Some of these compounds play a role in the recruitment of additional platelets and/or other cells such as neutrophils and monocytes to the site of activation, whereas others are involved in feedback mechanisms to down-regulate excessive thrombus formation.

At least three separate endothelial thromboregulatory systems exist: the eicosanoids including the prostaglandins PGI 1 and PGD 2 endothelium-dependent relaxing factor (EDRF/NO); and the ectonucleotidase ATP-diphosphohydrolase (ATPDase) which has both ADPase and ATPase activities.

WO 00/23459 PCT/US99/22955 While collagen and thrombin are the prime inducers of platelet secretion, ADP is the most important agonist of platelet aggregation present in the platelet releasate. Catabolism of ADP to AMP by the ecto-ADPase blocks further recruitment of additional platelets to the site, reverses the aggregation response and blocks subsequent thrombus response.

Ecto-nucleotidase activity is demonstrable in vitro in an aggregrometry system in which EDRF/NO effects and PGI 2 production are blocked by hemoglobin and aspirin respectively (Marcus and Safier, FASEB J7:516; 1993). In this system, loss of platelet stimulatory activity in the supernatant fluid correlates with ADP catabolism. An ADPase activity has been identified in the membrane fraction of human endothelial cells; enzyme activity detected by polyacrylamide gel electrophoresis indicated both ATPase and ADPase (Marcus et al., Clin. Res. 40:226A (abstract), 1992).

B. Utility of the Claimed Invention Significant research efforts are directed to the discovery and characterization of platelet aggregation inhibitors because of the potential utility of such inhibitors in treating occlusive vascular disease. For example, WO 95/12412 discloses platelet-specific chimeric antibodies and methods of using the same in treating various thrombotic disorders. A prototype description of the efforts to develop this therapeutic agent and obtain approval for its use as a human therapeutic agent (generic name: abciximab, trade name: ReoPro®) was described by B.S. Coller in Circulation 92:2373 (1995).

CD39 is an ecto-ADPase (apyrase) located on the surface of endothelial cells. This enzyme is mainly responsible for the maintenance of blood fluidity, thus maintaining platelets in the baseline (resting) state. This is accomplished by metabolism of the major platelet agonist, adenosine diphosphate, to adenosine monophosphate, which is not an agonist. Because ADP is the most important agonist of platelet aggregation, and is present in platelet releasate, a substance which catabolizes ADP is useful in treating or preventing disease states that involve inappropriate aggregation of platelets.

Examples of the therapeutic uses of soluble CD39 and compositions thereof include the treatment of individuals who suffer from coronary artery disease or injury following myocardial infarction, atherosclerosis, arteriosclerosis, preeclampsia, embolism, platelet-associated ischemic disorders including lung ischemia, coronary ischemia, and cerebral ischemia, and for the prevention of reocclusion following thrombosis, thrombotic disorders including coronary artery thrombosis, cerebral artery thrombosis, intracardiac thrombosis, peripheral artery thrombosis, venous thrombosis, and thrombosis and coagulopathies associated with exposure to a foreign or injured tissue surface, in combination with angioplasty, carotid endarterectomy, anastomosis of vascular grafts, and chronic cardiovascular devices such as in-dwelling catheters or shunts. Other instances in which it would be useful to inhibit increased ADP release due to increased platelet stimulation would be in individuals at high risk for thrombus formation or reformation (severe arteriosclerosis), and inhibition of occlusion, reocclusion, stenosis and/or restenosis of blood vessels. Individuals who will benefit from therapies WO 00/23459 PCT/US99/22955 that involve inhibiting ADP-induced aggregation of platelets include those at risk for advanced coronary artery disease, and those that are or will be undergoing angioplasty procedures balloon angioplasty, laser angioplasty, coronary atherectomy and similar techniques). Inhibition of platelet aggregation will also be useful in individuals undergoing surgery that has a high risk of thrombus formation coronary bypass surgery, insertion of a prosthetic valve or vessel and the like), and in the prevention or treatment of deep venous thrombosis (DVT), pulmonary embolism transient ischemic attacks (TIAs) and other related conditions where arterial occlusion is the common underlying feature. In addition, the ability of CD39 to block platelet activation and recruitment is useful for preventing stroke and for treating patients experiencing stroke due to vascular occlusion. In particular, the methods, compounds, and compositions of the present invention have the ability to inhibit microvascular thrombosis, improve postischemic cerebral blood flow, and reduce cerebral infarction volumes and neurological deficit without inducing intracerebral hemorrhage, in stroke.

Soluble CD39 and compositions thereof according to the present invention can also be administered in any other therapeutic setting where it would be useful to degrade nucleoside tri- and/or diphosphates.

As an example, soluble CD39 may be used as an anti-neoplastic agent to inhibit angiogenesis and/or prevent the survival benefits that ATP provides to tumor cells, or to treat other diseases or conditions mediated by angiogenesis such as occular neovascularization.

Soluble CD39 polypeptides also have many non-therapeutic uses, since they may be used in any application where soluble ATPase and/or ADPase activity is advantageous. As an example, soluble CD39 polypeptides may be used in compositions for preserving platelets such as those described by Gepner-Puszkin Patent No. 5,378,601). As another example, soluble CD39 polypeptides may be used in pyrophosphate-based DNA sequencing methodologies such as those described by Ronaghi et al. (Science 281:336, 1998). As a further example, soluble CD39 polypeptides can be used to screen for apyrase inhibitors.

C. CD39 Polvpeptides The molecular cloning and structural characterization of CD39 is presented in Maliszewski et al. Immunol. 153:3574, 1994). CD39 contains two putative transmembrane regions, near the amino and carboxy termini, which may serve to anchor the native protein in the cell membrane. The portion of the molecule between the transmembrane regions is external to the cell. As used herein, the term "CD39 polypeptides" includes CD39, homologs of CD39, variants, fragments, and derivatives of CD39, fusion polypeptides comprising CD39, and soluble forms of CD39 polypeptides.

Soluble polypeptides are polypeptides that are capable of being secreted from the cells in which they are expressed. A secreted soluble polypeptide may be identified (and distinguished from its non-soluble membrane-bound counterparts) by separating intact cells which express the desired polypeptide from the culture medium, by centrifugation, and assaying the medium (supernatant) for the presence of the desired polypeptide. The presence of the desired polypeptide in the medium WO 00/23459 PCT/US99/22955 indicates that the polypeptide was secreted from the cells and thus is a soluble form of the polypeptide.

The use of soluble forms of CD39 is advantageous for many applications. Purification of the polypeptides from recombinant host cells is facilitated, since the soluble polypeptides are secreted from the cells. Moreover, soluble polypeptides are generally more suitable than membrane-bound forms for parenteral administration and for many enzymatic procedures.

Apyrase activity resides in the extracellular domain of CD39. Thus, for applications requiring biological activity, useful CD39 polypeptides include soluble forms of CD39 such as those having an amino terminus selected from the group consisting of amino acids 36-44 of SEQ ID NO:2, and a carboxy terminus selected from the group consisting of amino acids 471-478 of SEQ ID NO:2, and which exhibit CD39 biological activity. Soluble CD39 polypeptides also include those polypeptides which include part of either or both of the transmembrane regions, provided that the soluble CD39 polypeptide is capable of being secreted from a cell, and retains CD39 biological activity. Soluble CD39 polypeptides further include oligomers or fusion polypeptides comprising the extracellular portion of CD39, and fragments of any of these polypeptides that have biological activity.

The term "biological activity," as used herein, includes apyrase enzymatic activity as well as the ex vivo and in vivo activities of CD39. Apyrases catalyze the hydrolysis of nucleoside tri- and/or di- phosphates, but a given apyrase may display different relative specificities for either nucleoside triphosphates or nucleoside diphosphates. Biological activity of soluble forms of CD39 may be determined, for example, in an ectonucleotidase or apyrase assay ATPase or ADPase assays), or in an assay that measures inhibition of platelet aggregation. Exemplary assays are disclosed herein; those of skill in the art will appreciate that other, similar types of assays can be used to measure biological activity.

Among the soluble CD39 polypeptides provided herein are variants (also referred to as analogs) of native CD39 polypeptides that retain a biological activity of CD39. Such variants include polypeptides that are substantially homologous to native CD39, but which have an amino acid sequence different from that of a native CD39 because of one or more deletions, insertions or substitutions. Particular embodiments include, but are not limited to, CD39 polypeptides that comprise from one to ten deletions, insertions or substitutions of amino acid residues, when compared to a native CD39 sequence. The CD39-encoding DNAs of the present invention include variants that differ from a native CD39 DNA sequence because of one or more deletions, insertions or substitutions, but that encode a biologically active polypeptide. Included as variants of CD39 polypeptides are those variants that are naturally occurring, such as allelic forms and alternatively spliced forms, as well as variants that have been constructed by modifying the amino acid sequence of a CD39 polypeptide or the nucleotide sequence of a nucleic acid encoding a CD39 polypeptide.

Generally, substitutions for one or more amino acids present in the native polypeptide should be made conservatively. Examples of conservative substitutions include substitution of amino acids outside of the active domain(s), and substitution of amino acids that do not alter the secondary and/or tertiary structure of CD39. Additional examples include substituting one aliphatic residue for another, WO 00/23459 PCT/US99/22955 such as Ile, Val, Leu, or Ala for one another, or substitutions of one polar residue for another, such as between Lys and Arg; Glu and Asp; or Gin and Asn. Other such conservative substitutions, for example, substitutions of entire regions having similar hydrophobicity characteristics, are known in the art.

When a deletion or insertion strategy is adopted, the potential effect of the deletion or insertion on biological activity must be considered. Subunits of the inventive polypeptides may be constructed by deleting terminal or internal residues or sequences. Additional guidance as to the types of mutations that can be made is provided by a comparison of the sequence of CD39 to polypeptides that have similar structures, as well as by performing structural analysis of the inventive polypeptides.

The native sequence of full length CD39 is set forth in Figure 1 (SEQ ID NO:2). In some preferred embodiments the CD39 variants are at least about 70% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing; in some preferred embodiments the CD39 variants are at least about 80% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing. In some more preferred embodiments the variants of CD39 are at least about 90% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing; in some more preferred embodiments the variants of CD39 are at least about 95% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing. In some most preferred embodiments, variants of CD39 are at least about 98% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing; in some most preferred embodiments, variants of CD39 are at least about 99% identical in amino acid sequence to the amino acid sequence of native CD39 as set forth in the sequence listing. Percent identity, in the case of both polypeptides and nucleic acids, may be determined by visual inspection. Percent identity may be determined using the alignment method of Needleman and Wunsch Mol. Biol. 48:443, 1970) as revised by Smith and Waterman (Adv. Appl. Math 2:482, 1981. Preferably, percent identity is determined by using a computer program, for example, the GAP computer program version available from the Genetics Computer Group (GCG; Madison, WI, see also Devereux et al., Nucl.

Acids Res. 12:387, 1984). The preferred default parameters for the GAP program include: a unary comparison matrix (containing a value of 1 for identities and 0 for non-identities) for nucleotides, and the weighted comparison matrix of Gribskov and Burgess, Nucl. Acids Res. 14:6745, 1986, as described by Schwartz and Dayhoff, eds., Atlas of Protein Sequence and Structure, National Biomedical Research Foundation, pp. 353-358, 1979 for amino acids; a penalty of 30 (amino acids) or 50 (nucleotides) for each gap and an additional 1 (amino acids) or 3 (nucleotides) penalty for each symbol in each gap; no penalty for end gaps; and no maximum penalty for long gaps.

Other programs used by one skilled in the art of sequence comparison may also be used. For fragments of CD39, the percent identity is calculated based on that portion of CD39 that is present in the fragment.

WO 00/23459 PCT/US99/22955 The primary amino acid structure of soluble CD39 may be modified to create CD39 derivatives by forming covalent or aggregative conjugates with other chemical moieties, such as glycosyl groups, lipids, phosphate, acetyl groups and the like. Covalent derivatives of CD39 are prepared by linking particular functional groups to CD39 amino acid side chains or at the N-terminus or C-terminus of a CD39 polypeptide or the extracellular domain thereof. CD39 derivatives also include CD39 polypeptides bound to various insoluble substrates, including cyanogen bromideactivated agarose structures, or similar agarose structures, or adsorbed to polyolefin surfaces (with or without glutaraldehyde cross-linking).

Fusion polypeptides of soluble CD39 within the scope of this invention include covalent or aggregative conjugates of CD39 or its fragments with other polypeptides, such as by synthesis in recombinant culture as N-terminal or C-terminal fusions. One class of fusion polypeptides are discussed below in connection with soluble CD39 oligomers. As another example, a fusion polypeptide may comprise a signal peptide (which is also variously referred to as a signal sequence, signal, leader peptide, leader sequence, or leader) at the N-terminal region or C-terminal region of a CD39 polypeptide which co-translationally or post-translationally directs transfer of the polypeptide from its site of synthesis to a site inside or outside of the cell membrane or cell wall the a-factor leader of Saccharomyces; several leader sequences are discussed in the examples that follow). It is particularly advantageous to fuse a signal peptide that promotes extracellular secretion to the Nterminus of a soluble CD39 polypeptide. In this case, the signal peptide is typically cleaved upon secretion of the soluble CD39 from the cell.

In a particularly preferred embodiment, one or more amino acids are added to the N-terminus of a soluble CD39 polypeptide in order to improve the expression levels and/or stability of the CD39 polypeptide. The one or more amino acids include an Ala residue, fragments derived from the Nterminus of another member of the CD39 family CD39L2, CD39L3, CD39L4) or from another polypeptide such as IL-2, and other peptides, either naturally-occurring or designed based upon structural predictions, capable of adopting a stable secondary structure.

In a most preferred embodiment, a soluble CD39 polypeptide is initially synthesized as a fusion polypeptide comprising: a signal peptide that promotes extracellular secretion of the soluble CD39 from the cell, the signal peptide being cleaved upon secretion, one or more amino acids added to the N-terminus of the soluble CD39 polypeptide in order to improve expression levels and/or stability, and a fragment of CD39 that possesses biological activity.

CD39 fusion polypeptides can also comprise polypeptides added to provide novel polyfunctional entities. Further, soluble CD39-containing fusion polypeptides can comprise peptides added to facilitate purification and identification of soluble CD39. Such peptides include, for example, poly-His or the antigenic identification peptides described in U.S. Patent No. 5,011,912 and in Hopp et al., Bio/Technology 6:1204, 1988. One such peptide is the FLAG® peptide, Asp-Tyr-Lys- Asp-Asp-Asp-Asp-Lys (SEQ ID NO:10), which is highly antigenic and provides an epitope reversibly WO 00/23459 PCT/US99/22955 bound by a specific monoclonal antibody, enabling rapid assay and facile purification of expressed recombinant polypeptide. A murine hybridoma designated 4E11 produces a monoclonal antibody that binds the FLAG® peptide in the presence of certain divalent metal cations, as described in U.S. Patent 5,011,912, hereby incorporated by reference. The 4E1 I hybridoma cell line has been deposited with the American Type Culture Collection under accession no. HB 9259. Monoclonal antibodies that bind the FLAG® peptide are available from Eastman Kodak Co., Scientific Imaging Systems Division, New Haven, Connecticut.

Another particularly useful class of fusion polypeptides includes those that allow localization or concentration of CD39 at a site of platelet activation and recruitment. Such fusion polypeptides comprise a moiety that specifically binds activated platelets and CD39, and can be prepared using recombinant DNA technology, or by using standard techniques for conjugation of polypeptides. For example, WO 95/12412 discloses platelet-specific chimeric antibodies and methods of using the same in treating various thrombotic disorders. These antibodies, or other platelet specific antibodies (for example, antibodies to P-selectin/CD62), are useful in forming fusion polypeptides with CD39.

Moreover, humanized or single chain antibodies can be prepared, based on such platelet specific antibodies.

Counterstructure molecules (molecules that specifically bind polypeptides expressed on the cell surface of activated platelets) and fragments thereof that bind to platelets are also useful in forming fusion polypeptides that bind specifically to activated platelets. Exemplary counterstructures include ligands for P-selectin/CD62 (see, Varki Proc Natl Acad Sci U S A 91:7390, 1994; Sammar et al., Int Inmunol 6:1027, 1994; Lenter et al., J Cell Biol 125:471, 1994).

Encompassed by the present invention are oligomers that contain CD39 polypeptides. CD39 oligomers may be in the form of covalently-linked or non-covalently-linked multimers, including dimers, trimers, or higher oligomers. Oligomers may be linked by disulfide bonds formed between cysteine residues on different CD39 polypeptides. Alternatively, oligomers may be formed by constructing fusion polypeptides of CD39 and the Fc region of an immunoglobulin molecule, such as human IgGI, to yield a CD39/Fc fusion polypeptide. The term "Fc polypeptide" as used herein includes native and mutein forms of polypeptides derived from the Fc region of an antibody.

Truncated forms of such polypeptides containing the hinge region that promotes dimerization are also included. One suitable Fc polypeptide, described in PCT application WO 93/10151 (hereby incorporated by reference), is a single chain polypeptide extending from the N-terminal hinge region to the native C-terminus of the Fc region of a human IgGI antibody. Another useful Fc polypeptide is the Fc mutein described in U.S. Patent 5,457,035 and in Baum et al., (EMBO J. 13:3992-4001, 1994).

The amino acid sequence of this mutein is identical to that of the native Fc sequence presented in WO 93/10151, except that amino acid 19 has been changed from Leu to Ala, amino acid 20 has been changed from Leu to Glu, and amino acid 22 has been changed from Gly to Ala. The mutein exhibits reduced affinity for Fc receptors. The CD39/Fc fusion polypeptides are allowed to assemble much WO 00/23459 PCT/US99/22955 like heavy chains of an antibody molecule to form divalent CD39. If fusion polypeptides are made with both heavy and light chains of an antibody, it is possible to form a CD39 oligomer with as many as four CD39 extracellular regions.

In some embodiments of the invention, oligomers comprising multiple CD39 polypeptides are joined via covalent or non-covalent interactions between peptide moieties fused to the C39polypeptides. Such peptide moieties may be peptide linkers (spacers), or peptides that have the property of promoting oligomerization. Leucine zippers and certain polypeptides derived from antibodies are among the peptides that can promote oligomerization of polypeptides.

The present invention comprises fusion polypeptides with or without spacer amino acid linking groups. For example, two soluble CD39 domains can be linked with a linker sequence, such as which is described in United States Patent 5,073,627. Other linker sequences include, for example, GlyAlaGlyGlyAlaGlySer(Gly)5Ser, (Gly4Ser)2, (GlyThrPro)3, and (Gly4Ser)3Gly4SerGly5Ser. Alternatively, CD39 can be linked to another polypeptide (non-CD39) with or without a spacer amino acid linking group. As shown in Example 9, ThrSerSer or ThrSerSerGly linkers may be used to fuse IL2 residues to soluble CD39. For the expression of soluble CD39, the inventors have made the surprising and unexpected discovery that the fusion of 12 amino acids from the N-terminus of mature human IL2 to the solCD39 coding region, results in high levels of both expression and activity in the superatants of transfected cells. Among the particularly preferred embodiments of the invention, therefore, are soluble CD39 polypeptides having an amino acid sequence SEQ ID NO:6 and nucleic acids, such as SEQ ID NO:5, that encode soluble CD39 polypeptides having an amino acid sequence SEQ ID NO:6.

The present invention further includes soluble CD39 polypeptides with or without associated native-pattern glycosylation. CD39 expressed in yeast or mammalian expression systems COS-7 cells) may be similar to or significantly different from a native CD39 polypeptide in molecular weight and glycosylation pattern, depending upon the choice of expression system. Expression of CD39 polypeptides in bacterial expression systems, such as E. coli, provides non-glycosylated molecules.

Different host cells may process polypeptides differentially, resulting in heterogeneous mixtures of polypeptides with variable N- or C-termini. Expression of soluble CD39 polypeptides in microbial expression systems, such as E. coli, generally provides a homogeneous polypeptide preparation. Polypeptides may be differentially processed by a eukaryotic cell, resulting in variable Nand C-termini, and hence yield a heterogeneous polypeptide preparation. The present invention includes polypeptides, produced by eukaryotic host cells, which have variable N-termini or C-termini.

In one embodiment of the inventive CD39 polypeptides, the amino and carboxy termini can be about five amino acids different from those disclosed herein.

The skilled artisan will also recognize that the position(s) at which a signal peptide is cleaved may differ from that predicted by computer program, and may vary according to such factors as the type of host cells employed in expressing a recombinant soluble CD39 polypeptide. A polypeptide WO 00/23459 PCT/US99/22955 preparation according to the invention may therefore include a mixture of polypeptide molecules having different N-terminal amino acids, resulting from cleavage of the signal peptide at more than one site.

D. Nucleic Acids The invention encompasses full length nucleic acid molecules encoding soluble CD39 as well as isolated fragments and oligonucleotides derived from the nucleotide sequence of SEQ ID NO: 1.

Such nucleic acid sequences may include nucleotides 178-1494 of SEQ ID NO:1 or a fragment thereof, and DNA and/or RNA sequences that hybridize to the coding region of the nucleotide sequence of SEQ ID NO:1, or its complement, under conditions of moderate stringency, and which encode polypeptides or fragments thereof of the invention.

Nucleic acid sequences encoding soluble CD39 polypeptides having altered glycosylation sites, deleted or substituted Cys residues, or modified proteolytic cleavage sites, nucleic acid sequences encoding sub-units of CD39 polypeptides or fusion polypeptides of CD39 with other peptides, allelic variants of CD39, mammalian homologs of CD39, and nucleic acid sequences encoding CD39 polypeptides derived from alternative mRNA constructs, or those that encode peptide having substituted or additional amino acids, are examples of nucleic acid sequences according to the invention.

Due to degeneracy of the genetic code, there can be considerable variation in nucleotide sequences encoding the same amino acid sequence. Included as embodiments of the invention are sequences capable of hybridizing under moderately stringent conditions prewashing solution of X SSC, 0.5% SDS, 1.0 mM EDTA (pH 8.0) and hybridization conditions of 50'C, 5 X SSC, overnight) to the DNA sequences encoding soluble CD39, and other sequences which are degenerate to those which encode soluble CD39. The skilled artisan can determine additional combinations of salt and temperature that constitute moderate hybridization stringency. Conditions of higher stringency include higher temperatures for hybridization and post-hybridization washes, and/or lower salt concentration.

In a preferred embodiment, CD39 DNAs include those that encode polypeptides that are at least about 70% or at least 80% identical in amino acid sequence to the amino acid sequence of native CD39 polypeptide as set forth in SEQ ID NO:1. In a more preferred embodiment, the encoded variants of CD39 are at least about 90% or at least about 95% identical in amino acid sequence to the native form of CD39; in a most preferred embodiment, the encoded variants of CD39 are at least about 98% or at least about 99% identical in amino acid sequence to the native form of CD39. For DNAs that encode a fragment of CD39, percent identity of the fragment is based on percent identity to the corresponding portion of full-length CD39.

Mutations can be introduced into nucleic acids by synthesizing oligonucleotides containing a mutant sequence, flanked by restriction sites enabling ligation to fragments of the native sequence.

WO 00/23459 PCT/US99/22955 Following ligation, the resulting reconstructed sequence encodes a variant having the desired amino acid insertion, substitution, or deletion.

Alternatively, oligonucleotide-directed site-specific mutagenesis procedures can be employed to provide an altered gene having particular codons altered according to the substitution, deletion, or insertion required. Exemplary methods of making the alterations set forth above are disclosed by Walder et al. (Gene 42:133, 1986); Bauer et al. (Gene 37:73, 1985); Craik (BioTechniques, January 1985, 12-19); Smith et al. (Genetic Engineering: Principles and Methods, Plenum Press, 1981); and U.S. Patent Nos. 4,518,584 and 4,737,462 disclose suitable techniques, and are incorporated by reference herein.

The well known polymerase chain reaction (PCR) procedure also may be employed to generate and amplify a DNA sequence encoding a desired polypeptide or fragment thereof.

Oligonucleotides that define the desired termini of the DNA fragment are employed as 5' and 3' primers. The oligonucleotides may additionally contain recognition sites for restriction endonucleases to facilitate insertion of the amplified DNA fragment into an expression vector. PCR techniques are described in Saiki et al., Science 239:487, 1988; Recombinant DNA Methodology, Wu et al., eds., Academic Press, Inc., San Diego, 1989, pp. 189-196; and PCR Protocols: A Guide to Methods and Applications, Innis et al., eds., Academic Press, Inc., 1990.

DNA sequences that encode CD39 polypeptides comprising various additions or substitutions of amino acid residues or sequences, or deletions of terminal or internal residues or sequences not needed for biological activity can be prepared. For example, N-glycosylation sites can be modified to preclude glycosylation while allowing expression of a homogeneous, reduced carbohydrate variant using yeast expression systems. N-glycosylation sites in eukaryotic polypeptides are characterized by an amino acid triplet Asn-X-Y, wherein X is any amino acid except Pro and Y is Ser or Thr.

Appropriate modifications to the nucleotide sequence encoding this triplet will result in substitutions, additions or deletions that prevent attachment of carbohydrate residues at the Asn side chain.

In another example, sequences encoding Cys residues can be altered to cause the Cys residues to be deleted or replaced with other amino acids, preventing formation of incorrect intramolecular disulfide bridges upon renaturation. Thus, Cys residues may be replaced with another amino acid or deleted without affecting polypeptide tertiary structure or disulfide bond formation.

Other approaches to mutagenesis involve modification of sequences encoding dibasic amino acid residues to enhance expression in yeast systems in which KEX2 protease activity is present. Other variants are prepared by modification of adjacent dibasic amino acid residues, to enhance expression in yeast systems in which KEX2 protease activity is present. EP 212,914 discloses the use of sitespecific mutagenesis to inactivate KEX2 protease processing sites in a polypeptide. KEX2 protease processing sites are inactivated by deleting, adding or substituting residues to alter Arg-Arg, Arg-Lys, and Lys-Arg pairs to eliminate the occurrence of these adjacent basic residues. Similar modification may be made to sequences encoding sites recognized and cleaved by other proteolytic enzymes. Subunits of a CD39 polypeptide may be constructed by deleting sequences encoding terminal or internal WO 00/23459 PCT/US99/22955 residues or sequences not necessary for biological activity. Sequences encoding fusion polypeptides as described below may be constructed by ligating sequences encoding additional amino acid residues to the inventive sequences without affecting biological activity.

Mutations in nucleotide sequences constructed for expression of a soluble CD39 must, of course, preserve the reading frame phase of the coding sequences and preferably will not create complementary regions that could hybridize to produce secondary mRNA structures such as loops or hairpins which would adversely affect translation of the receptor mRNA. Although a mutation site may be predetermined, it is not necessary that the nature of the mutation per se be predetermined. For example, in order to select for optimum characteristics of mutants at a given site, random mutagenesis may be conducted at the target codon and the expressed mutated polypeptides screened for the desired activity.

Not all mutations in the nucleotide sequence which encodes a CD39 polypeptide will be expressed in the final product, for example, nucleotide substitutions may be made to enhance expression, primarily to avoid secondary structure loops in the transcribed mRNA (see EPA 75,444A, incorporated herein by reference), or to provide codons that are more readily translated by the selected host, the well-known E. coli preference codons for E. coli expression.

In the genome, CD39 polypeptides are encoded by multi-exon genes. The present invention further includes alternative mRNA constructs which can be attributed to different mRNA splicing events following transcription and which hybridize with the cDNAs disclosed herein under conditions of moderate stringency. CD39 polypeptides according to the invention include allelic variations of the sequence shown in SEQ ID NO: 1, and sequences encoding CD39 polypeptides that comprise additional amino acids to those of SEQ ID NO:1.

The isolated nucleic acid sequences of this invention are sufficiently free of association with nucleic acid sequences encoding other proteinaceous material, and from other materials found in living cells, such as proteins, lipids or carbohydrates, to allow the skilled artisan to prepare vectors for the expression of soluble CD39 polypeptides.

E. Recombinant Expression Systems The present invention also provides recombinant cloning and expression vectors containing CD39 DNA, as well as host cells containing the recombinant vectors. Expression vectors comprising CD39 DNA may be used to prepare soluble CD39 polypeptides encoded by the DNA. The expression vectors carrying the recombinant CD39 DNA sequence are transferred, for example by transfection or transformation, into a substantially homogeneous culture of a suitable host microorganism or mammalian cell line. Transformed host cells are cells which have been transformed or transfected with nucleotide sequences encoding CD39 polypeptides and express CD39 polypeptides. Expressed CD39 polypeptides will be located within the host cell and/or secreted into culture supernatant fluid, depending upon the nature of the host cell and the gene construct inserted into the host cell. The WO 00/23459 PCT/US99/22955 skilled artisan will recognize that the procedure for purifying the expressed CD39 will vary according to such factors as the type of host cells employed.

Any suitable expression system may be employed. Recombinant expression vectors for expression of soluble CD39 by recombinant DNA techniques include a CD39 DNA sequence comprising a synthetic or cDNA-derived DNA fragment encoding a CD39 polypeptide, operably linked to a suitable transcriptional or translational regulatory nucleotide sequence, such as one derived from a mammalian, microbial, viral, or insect gene.

Examples of regulatory sequences include transcriptional promoters, operators, or enhancers, an mRNA ribosomal binding site, and appropriate sequences which control transcription and translation initiation and termination. Nucleotide sequences are operably linked when the regulatory sequence functionally relates to the CD39 DNA sequence. Thus, a promoter nucleotide sequence is operably linked to a CD39 DNA sequence if the promoter nucleotide sequence controls the transcription of the CD39 DNA sequence. An origin of replication that confers the ability to replicate in the desired host cells, and a selection gene by which transformants are identified, are generally incorporated into the expression vector.

In addition, a sequence encoding an appropriate signal peptide (native or heterologous) can be incorporated into expression vectors. A DNA sequence for a signal peptide (secretory leader) may be fused in frame to the CD39 sequence so that the CD39 is initially translated as a fusion polypeptide comprising the signal peptide. A signal peptide that is functional in the intended host cells promotes extracellular secretion of the CD39 polypeptide. The signal peptide is cleaved from the CD39 polypeptide upon secretion of soluble CD39 from the cell.

Regarding signal peptides that may be employed in producing soluble CD39, the native signal peptide may be replaced by a heterologous signal peptide or leader sequence, if desired. The choice of signal peptide or leader may depend on factors such as the type of host cells in which the recombinant polypeptide is to be produced. To illustrate, examples of heterologous signal peptides that are functional in mammalian host cells include the signal sequence for interleukin-7 (IL-7) described in United States Patent 4,965,195, the signal sequence for interleukin-2 receptor described in Cosman et al., Nature 312:768, 1984; the interleukin-4 receptor signal peptide described in EP 367,566; the type I interleukin-1 receptor signal peptide described in U.S. Patent 4,968,607; and the type I interleukin-I receptor signal peptide described in EP 460,846. For the expression of soluble CD39, the inventors have made the surprising and unexpected discovery that the use of a leader containing sequences derived from a human IL-2 polypeptide (SEQ ID NO:9) results in high levels of ATPase activity in the supernatants of transfected cells. Among the particularly preferred embodiments of the invention, therefore, are nucleic acids encoding soluble CD39 polypeptides having an amino acid sequence SEQ ID NO:8.

Suitable host cells for expression of CD39 polypeptides include prokaryotes, yeast or higher eukaryotic cells. Mammalian or insect cells are generally preferred for use as host cells. Appropriate cloning and expression vectors for use with bacterial, fungal, yeast, and mammalian cellular hosts are WO 00/23459 PCT/US99/22955 described, for example, in Pouwels et al. Cloning Vectors: A Laboratory Manual, Elsevier, New York, 1985. Cell-free translation systems could also be employed to produce soluble CD39 polypeptides using RNAs derived from DNA constructs disclosed herein.

Prokaryotes include gram negative or gram positive organisms, for example, E. coli or Bacilli. Suitable prokaryotic host cells for transformation include, for example, E. coli, Bacillus subtilis, Salmonella typhimurium, and various other species within the genera Pseudomonas, Streptomyces, and Staphylococcus. In a prokaryotic host cell, such as E. coli, polypeptides may include an N-terminal methionine residue to facilitate expression of the recombinant polypeptide in the prokaryotic host cell. The N-terminal Met may be cleaved from the expressed recombinant polypeptide.

Expression vectors for use in prokaryotic host cells generally comprise one or more phenotypic selectable marker genes. A phenotypic selectable marker gene is, for example, a gene encoding a protein that confers antibiotic resistance or that supplies an autotrophic requirement.

Examples of useful expression vectors for prokaryotic host cells include those derived from commercially available plasmids such as the cloning vector pBR322 (ATCC 37017). pBR322 contains genes for ampicillin and tetracycline resistance and thus provides simple means for identifying transformed cells. An appropriate promoter and a CD39 DNA sequence are inserted into the pBR322 vector. Other commercially available vectors include, for example, pKK223-3 (Pharmacia Fine Chemicals, Uppsala, Sweden) and pGEMI (Promega Biotec, Madison, WI, USA).

Promoter sequences commonly used for recombinant prokaryotic host cell expression vectors include p-lactamase (penicillinase), lactose promoter system (Chang et al., Nature 275:615, 1978; and Goeddel et al., Nature 281:544, 1979), tryptophan (trp) promoter system (Goeddel et al., Nucl. Acids Res. 8:4057, 1980; and EP-A-36776) and tac promoter (Maniatis, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, p. 412, 1982). A particularly useful prokaryotic host cell expression system employs a phage X PL promoter and a c1857ts thermolabile repressor sequence.

Plasmid vectors available from the American Type Culture Collection which incorporate derivatives of the PL promoter include plasmid pHUB2 (resident in E. coli strain JMB9, ATCC 37092) and pPLc28 (resident in E. coli RRI, ATCC 53082).

Soluble CD39 may also be expressed in yeast host cells, preferably from the Saccharomyces genus S. cerevisiae). Other genera of yeast, such as Pichia or Kluyveromyces, may also be employed. Yeast vectors will often contain an origin of replication sequence from a 2it yeast plasmid, an autonomously replicating sequence (ARS), a promoter region, sequences for polyadenylation, sequences for transcription termination, and a selectable marker gene. Suitable promoter sequences for yeast vectors include, among others, promoters for metallothionein, 3-phosphoglycerate kinase (Hitzeman et al., J. Biol. Chem. 255:2073, 1980) or other glycolytic enzymes (Hess et al., J. Adv.

Enzyme Reg. 7:149, 1968; and Holland et al., Biochem. 17:4900, 1978), such as enolase, glyceraldehyde-3-phosphate dehydrogenase, hexokinase, pyruvate decarboxylase, WO 00/23459 PCT/US99/22955 phosphofructokinase, glucose-6-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, triosephosphate isomerase, phospho-glucose isomerase, and glucokinase. Other suitable vectors and promoters for use in yeast expression are further described in Hitzeman, EPA-73,657. Another alternative is the glucose-repressible ADH2 promoter described by Russell et al. Biol. Chem.

258:2674, 1982) and Beier et al. (Nature 300:724, 1982). Shuttle vectors replicable in both yeast and E. coli may be constructed by inserting DNA sequences from pBR322 for selection and replication in E. coli (Ampr gene and origin of replication) into the above-described yeast vectors.

The yeast a-factor leader sequence may be employed to direct secretion of recombinant polypeptides. The a-factor leader sequence is often inserted between the promoter sequence and the structural gene sequence. See, Kurjan et al., Cell 30:933, 1982 and Bitter et al., Proc. Natl. Acad.

Sci. USA 81:5330, 1984. Other leader sequences suitable for facilitating secretion of recombinant polypeptides from yeast hosts are known to those of skill in the art. A leader sequence may be modified near its 3'end to contain one or more restriction sites. This will facilitate fusion of the leader sequence to the structural gene.

Yeast transformation protocols are known to those of skill in the art. One such protocol is described by Hinnen et al., Proc. Natl. Acad. Sci. USA 75:1929, 1978. The Hinnen et al. protocol selects for Trp+ transformants in a selective medium, wherein the selective medium consists of 0.67% yeast nitrogen base, 0.5% casamino acids, 2% glucose, 10 pg/ml adenine and 20 [Ig/ml uracil.

Yeast host cells transformed by vectors containing an ADH2 promoter sequence may be grown for inducing expression in a "rich" medium. An example of a rich medium is one consisting of 1% yeast extract, 2% peptone, and 1% glucose supplemented with 80 pg/ml adenine and 80 pg/ml uracil. Derepression of the ADH2 promoter occurs when glucose is exhausted from the medium.

Mammalian or insect host cell culture systems also may be employed to express recombinant CD39 polypeptides. Bacculovirus systems for production of heterologous polypeptides in insect cells are reviewed by Luckow and Summers, Bio/Technology 6:47, 1988. Established cell lines of mammalian origin may also be used. Examples of suitable mammalian host cell lines include the COS-7 line of monkey kidney cells (ATCC CRL 1651) (Gluzman et al., Cell 23:175, 1981), L cells, C127 cells, 3T3 cells (ATCC CCL 163), Chinese hamster ovary (CHO) cells, HeLa cells, and BHK (ATCC CRL 10) cell lines, and the CV1/EBNA cell line derived from the African green monkey kidney cell line CVI (ATCC CCL 70) as described by McMahan et al. (EMBO J. 10: 2821, 1991).

For the production of therapeutic polypeptides it is particularly advantageous to use a mammalian host cell line which has been adapted to grow in media that does not contain animal proteins. The use of such a cell line for the expression of soluble CD39 is described in Example 13.

Established methods for introducing DNA into mammalian cells have been described (Kaufman, Large Scale Mammalian Cell Culture, 1990, pp. 15-69). Additional protocols using commercially available reagents, such as Lipofectamine (Gibco/BRL) or Lipofectamine-Plus, can be used to transfect cells (Feigner et al., Proc. Natl. Acad. Sci. USA 84:7413-7417, 1987). In addition, WO 00/23459 PCT/US99/22955 electroporation can be used to transfect mammalian cells using conventional procedures, such as those in Sambrook et al. Molecular Cloning: A Laboratory Manual, 2 ed. Vol. 1-3, Cold Spring Harbor Laboratory Press, 1989). Selection of stable transformants can be performed using methods known in the art, such as, for example, resistance to cytotoxic drugs. Kaufman et al., Meth. in Enzymology 185:487-511, 1990, describes several selection schemes, such as dihydrofolate reductase (DHFR) resistance. A suitable host strain for DHFR selection can be CHO strain DX-B 11, which is deficient in DHFR (Urlaub and Chasin, Proc. Natl. Acad. Sci. USA 77:4216-4220, 1980). A plasmid expressing the DHFR cDNA can be introduced into strain DX-B 11, and only cells that contain the plasmid can grow in the appropriate selective media. Other examples of selectable markers that can be incorporated into an expression vector include cDNAs conferring resistance to antibiotics, such as G418 and hygromycin B. Cells harboring the vector can be selected on the basis of resistance to these compounds.

Transcriptional and translational control sequences for mammalian host cell expression vectors can be excised from viral genomes. Commonly used promoter sequences and enhancer sequences are derived from polyoma virus, adenovirus 2, simian virus 40 (SV40), and human cytomegalovirus. DNA sequences derived from the SV40 viral genome, for example, SV40 origin, early and late promoter, enhancer, splice, and polyadenylation sites can be used to provide other genetic elements for expression of a structural gene sequence in a mammalian host cell. Viral early and late promoters are particularly useful because both are easily obtained from a viral genome as a fragment, which can also contain a viral origin of replication (Fiers et al., Nature 273:113, 1978; Kaufman, Meth. in Enzymology, 1990). Smaller or larger SV40 fragments can also be used, provided the approximately 250 bp sequence extending from the Hind m site toward the Bgl I site located in the viral origin of replication site is included.

Additional control sequences shown to improve expression of heterologous genes from mammalian expression vectors include such elements as the expression augmenting sequence element (EASE) derived from CHO cells (Morris et al., Animal Cell Technology, 1997, pp. 529-534) and the tripartite leader (TPL) and VA gene RNAs from Adenovirus 2 (Gingeras et al., J. Biol. Chem.

257:13475-13491, 1982). The internal ribosome entry site (IRES) sequences of viral origin allows dicistronic mRNAs to be translated efficiently (Oh and Samow, Current Opinion in Genetics and Development 3:295-300, 1993; Ramesh et al., Nucleic Acids Research 24:2697-2700, 1996).

Expression of a heterologous cDNA as part of a dicistronic mRNA followed by the gene for a selectable marker DHFR) has been shown to improve transfectability of the host and expression of the heterologous cDNA (Kaufman, Meth. in Enzymology, 1990). Exemplary expression vectors that employ dicistronic mRNAs are pTR-DC/GFP described by Mosser et al., Biotechniques 22:150- 161, 1997, and p2A5I described by Morris et al., Animal Cell Technology, 1997, pp. 529-534.

A useful high expression vector, pCAVNOT, has been described by Mosley et al., Cell 59:335-348, 1989. Other expression vectors for use in mammalian host cells can be constructed as disclosed by Okayama and Berg (Mol. Cell. Biol. 3:280, 1983). A useful system for stable high level WO 00/23459 PCT/US99/22955 expression of mammalian cDNAs in C127 murine mammary epithelial cells can be constructed substantially as described by Cosman et al. (Mol. Immunol. 23:935, 1986). A useful high expression vector, PMLSV N1/N4, described by Cosman et al., Nature 312:768, 1984, has been deposited as ATCC 39890. Additional useful mammalian expression vectors are described in EP-A-0367566, and in U.S. Patent Application Serial No. 07/701,415, filed May 16, 1991, incorporated by reference herein. The vectors can be derived from retroviruses. In place of the native signal sequence, a heterologous signal sequence can be added, such as the signal sequence for IL-7 described in United States Patent 4,965,195; the signal sequence for IL-2 receptor described in Cosman et al., Nature 312:768, 1984; the IL-4 signal peptide described in EP 367,566; the type I IL-1 receptor signal peptide described in U.S. Patent 4,968,607; and the type II IL-1 receptor signal peptide described in EP 460,846.

Another useful expression vector, pFLAG, can be used. FLAG® technology is centered on the fusion of a low molecular weight (lkD), hydrophilic, FLAG® marker peptide to the N-Terminus of a recombinant polypeptide expressed by the pFLAG-lTM Expression Vector (obtained from IBI Kodak).

F. Purification of soluble CD39 Polypeptides Soluble CD39 polypeptides may be prepared by culturing transformed host cells under culture conditions necessary to express CD39 polypeptides. The resulting expressed polypeptides may then be purified from culture media or cell extracts. Supernatant fluid from the cultured, transformed host cells may be concentrated using a commercially available protein concentration filter, for example, an Amicon or Millipore Pellicon ultrafiltration unit. Following the concentration step, the concentrate may be applied to a cation exchange matrix. Suitable cation exchangers include various insoluble matrices comprising sulfonic or carboxymethyl groups; sulfonic groups are preferred. The matrices can be acrylamide, agarose, dextran, cellulose or other types commonly employed in protein purification. Subsequently, an anion exchange resin is employed, for example, a matrix or substrate having pendant diethylaminoethyl (DEAE) or quaternary amino groups; quaternary amino groups are preferred. The matrices can be acrylamide, agarose, dextran, cellulose or other types commonly employed in protein purification. Additionally, a gel filtration medium may be employed to further purify CD39 polypeptides according to approximate molecular weight. Alternatively, certain of these steps may not be performed, or may be performed in the reverse order.

One or more reverse-phase high performance liquid chromatography (RP-HPLC) steps employing hydrophobic RP-HPLC media, silica gel having pendant methyl or other aliphatic groups) may be employed to further purify CD39. A substantially purified and homogeneous polypeptide having CD39 biological activity may be eluted from a polyacrylamide gel subsequent to electrophoretic separation. Some or all of the foregoing purification steps, in various combinations, can also be employed to provide a substantially purified and homogeneous recombinant polypeptide WO 00/23459 PCT/US99/22955 containing less than about 1% by mass of protein contaminants residual of production processes, or alternatively, which is greater than about 95 pure by gel electrophoresis.

Affinity chromatography may be utilized to purify soluble CD39. Affinity purification of soluble CD39 from conditioned media is described in Example 12C. Moreover, small amounts of purified CD39 may be obtained by immunoprecipitating CD39 with a monoclonal antibody, electrophoresing the immunoprecipitate on a polyacrylamide gel, excising the portion of the gel containing the CD39, and eluting the CD39 from the excised portion of the gel.

Recombinant polypeptides produced in bacterial culture are generally isolated by disruption of the host cells, centrifugation, extraction from cell pellets if an insoluble polypeptide, or from the supernatant fluid if a soluble polypeptide, followed by one or more concentration, salting-out, ion exchange, affinity purification or size exclusion chromatography steps. Finally, RP-HPLC can be employed for final purification steps. Microbial cells can be disrupted by any convenient method, including freeze-thaw cycling, sonication, mechanical disruption, or use of cell lysing agents.

Transformed yeast host cells may be employed to express CD39 as a secreted polypeptide.

This simplifies purification. Secreted recombinant polypeptide from a yeast host cell fermentation can be purified by methods analogous to those disclosed by Urdal et al. Chromatog. 296:171, 1984).

Urdal et al. describe two sequential, reversed-phase HPLC steps for purification of recombinant human IL-2 on a preparative HPLC column.

The desired degree of purity of soluble CD39 polypeptides depends on the intended use of the polypeptide. A relatively high degree of purity is desired when the polypeptide is to be administered in vivo, for example. Advantageously, soluble CD39 polypeptides are purified such that no protein bands corresponding to other (non-CD39) polypeptides are detectable upon analysis by SDSpolyacrylamide gel electrophoresis (SDS-PAGE). Following electrophoresis, the protein band may be visualized by silver staining, Coomassie blue staining, or (if the protein is radiolabeled) by autoradiography. It will be recognized by one skilled in the pertinent field that multiple bands corresponding to CD39 polypeptide may be visualized by SDS-PAGE, due to differential glycosylation, differential post-translational processing, and the like.

G. Therapeutic Compositions of CD39 Polypeptides The present invention provides compositions comprising an effective amount of a soluble CD39 polypeptide in a pharmaceutically acceptable carrier. As used herein, the terms "therapy," "therapeutic," "treat," and "treatment" generally include prophylaxis, prevention, of a disease or condition in addition to therapy or treatment for an extant disease or condition. Therapeutic compositions of soluble CD39 polypeptides may therefore need to be administered before, during, or after the presentation of symptoms. For therapeutic use, a soluble CD39 polypeptide is administered to a patient for treatment in a manner appropriate to the indication. Thus, for example, soluble CD39 pharmaceutical compositions which are administered to achieve a desired therapeutic effect can be given by bolus injection, continuous infusion, sustained release from implants or the like, or other WO 00/23459 PCT/US99/22955 suitable technique. Ideally, development of a stable form of CD39 or closely related biologically active variant would allow its use in oral form, a preferable route of administration. Since CD39 is aspirin-insensitive, these two therapeutic agents (CD39 compositions and aspirin) can be used in combination, for maximal benefit.

Typically, a soluble CD39 therapeutic agent will be administered in the form of a pharmaceutical composition comprising purified soluble CD39 in conjunction with physiologically acceptable carriers, including excipients or diluents. Such carriers will be nontoxic to patients at the dosages and concentrations employed. As described in the examples that follow, the administration of CD39 in murine and porcine models of thrombosis does not cause any observable toxic effects.

Moreover, a second dose of CD39 does not evoke any signs of immunogenicity. Ordinarily, the preparation of such compositions entails combining a soluble CD39 composition with buffers, antioxidants such as ascorbic acid, low molecular weight (less than about 10 residues) polypeptides, polypeptides, amino acids, carbohydrates including glucose, sucrose or dextrans, chelating agents such as EDTA, glutathione and other stabilizers and excipients. Neutral buffered saline or saline mixed with conspecific serum albumin are exemplary appropriate diluents.

One type of sustained release technology which may be used in administering soluble CD39 compositions is that utilizing hydrogel materials, for example, photopolymerizable hydrogels (Sawhney et al., Macromolecules 26:581; 1993). Similar hydrogels have been used to prevent postsurgical adhesion formation (Hill-West et al., Obstet. Gynecol. 83:59, 1994) and to prevent thrombosis and vessel narrowing following vascular injury (Hill-West et al., Proc. Natl. Acad. Sci.

USA 91:5967, 1994). Polypeptides can be incorporated into such hydrogels to provide sustained, localized release of active agents (West and Hubbel, Reactive Polymers 25:139, 1995; Hill-West et al., J. Surg. Res. 58:759; 1995). The sustained, localized release of CD39 when incorporated into hydrogels would be amplified by the long half life of CD39, which is demonstrated in the Examples below.

Accordingly, the soluble CD39 compositions described herein can also be incorporated into hydrogels, for application to tissues for which localized inhibition of hemostasis is desirable. For example, a hydrogel incorporating a CD39 polypeptide can be applied to tissue after surgery, to prevent or reduce post-surgical adhesion formation, or can be applied using a catheter-based delivery system following angioplasty to prevent or reduce restenosis. Those of skill in the art will be able to formulate an appropriate hydrogel by applying standard pharmacokinetic studies, for example as discussed in West and Hubbell, supra.

Effective amounts may vary, depending on the age, type and severity of the condition to be treated, body weight, desired duration of treatment, method of administration, and other parameters.

Effective dosages are determined by a physician or other qualified medical professional. Typical dosages are 0.01-100 mg/kg body weight, preferably 0.1-10 mg/kg body weight. In some embodiments a single administration is sufficient; in some embodiments the soluble CD39 polypeptide is administered on a daily basis for up to a week or as much as a month or more.

WO 00/23459 PCT/US99/22955 The biological effectiveness of soluble CD39 polypeptides is easily evaluable: at given time intervals after administration, a prolongation of the bleeding time in the setting of unchanged platelet count should be measurable if released platelet ADP has been metabolized by the CD39 composition administered. This would indicate that a therapeutic effect has likely been obtained, as said measurement correlates with clinical improvement. A therapeutic effect can also be validated by testing platelet reactivity to ADP and other platelet agonists ex vivo. Actual measurements of enzyme (apyrase) activity can also be made following administration of soluble CD39. These and other methods of measuring biological effectiveness are illustrated in the Examples below.

H. Abbreviations Used in the Specification ACR, apyrase conserved regions; AG, Affigel beads; ASA, acetylsalicylic acid; ATPDase, ATP diphosphohydolase; CHO, Chinese hamster ovary; CM, conditioned medium; DHFR, dihydrofolate reductase; FSBA, fluorosulfonylbenzoyl-adenosine; HUVEC, human umbilical vein endothelial cells; PRP, platelet-rich plasma; PTCA, percutaneous transluminal coronary angioplasty; solCD39, recombinant soluble human CD39; TBS, Tris-buffered saline

EXAMPLES

The following examples are intended to illustrate particular embodiments and not to limit the scope of the invention.

EXAMPLE 1 Assay For CD39 Expression This example describes the use of a monoclonal antibody in a FACS assay to analyze expression of CD39. The B73 mAb, a monoclonal anti-CD39, is a murine IgG1 that was derived from BALB/c mice immunized with the RPMI 1788 cell line (Rector et al., Immunology 55:481, 1985) and characterized as CD39-specific by flow cytometric analysis and immunoprecipitation/SDS-PAGE.

Monoclonal anti-CD39 is purified from ascites fluid by affinity chromatography using a protein A column, eluted with 0.05 M sodium citrate, pH 3.0, neutralized and stored a 4 0 C at a concentration of about 1 mg/ml.

WO 00/23459 PCT/US99/22955 Cells to be analyzed MP-I cells, U937 cells, U937 cells stimulated with 5 ng/ml phorbol myristate acetate (PMA), or Daudi cells) are suspended to a concentration of 106 cell in 50 pl of phosphate buffered saline (PBS) containing 100 pg/ml human IgGI, and incubated for 30 minutes.

The cells are then pelleted by centrifugation, resuspended in PBS/azide containing a first antibody (anti-CD39 or control antibody) and incubated for 30 minutes at 4°C) The cells are then washed two times in PBS/azide, resuspended, and incubated with a labeled second antibody, for example, goat anti-murine immunoglobulin conjugated to phycoerythrin, then washed again. The cells are analyzed by flow cytometry, and levels of CD39 determined.

EXAMPLE 2 Immunoselection of Cells Expressing CD39 This example describes a panning (immunoselection) technique for cells expressing CD39.

For the preparation of pan plates, purified anti-CD39 or control antibody is diluted in phosphate buffered saline containing 0.1% heat-inactivated fetal calf serum (PBS/FCS). A titration of anti-CD39 can be performed to determine the most effective concentration of anti-CD39. Pan plates are prepared by adding three ml of antibody solution or PBS/FCS alone to each plate. The plates are incubated for approximately one hour at room temperature, washed five times with PBS/FCS, and three ml of PBS/FCS containing 0.02% sodium azide are added to each plate.

The cells to be analyzed MP-1, U937, or Daudi cells) are suspended in PBS/500 pM EDTA/0.02% sodium azide (PEA) containing 5% goat serum, 5% rabbit serum and 100 pg/ml human IgG to a concentration of 2 x 106 cells/ml; 500 pl of each cell suspension is added to the prepared pan plates. The pan plates are incubated with the cell suspension for approximately two hours at room temperature, then the plates are washed gently three times with PEA containing 10% FCS (PEA/FCS), and three times with PEA. The plates are examined with a microscope, and the relative number of cells bound to each plate is determined.

EXAMPLE 3 cDNA Library Construction This example describes preparation of a cDNA library from a human B cell line referred to as MP-1, for expression cloning of human CD39.

The library construction techniques were substantially similar to that described by Ausubel et al., eds., Current Protocols In Molecular Biology, Vol. 1, 1987. Briefly, total RNA was extracted from 8M guanidine HCl-lysed MP-1 cell cultures using differential ethanol precipitation and poly (A) mRNA was isolated and enriched by oligo dT cellulose chromatography. Double-stranded cDNA was made from an RNA template substantially as described by Gubler et al., Gene 25:263, 1983. Poly(A)+ mRNA fragments were converted to RNA-cDNA hybrids using reverse transcriptase primed with random hexanucleotides. The RNA-cDNA hybrids were then converted into double- WO 00/23459 PCT/US99/22955 stranded cDNA fragments using RNAase H in combination with DNA polymerase I. The resulting double-stranded cDNA was blunt-ended with T4 DNA polymerase.

Unkinased unphosphorylated) BglII adaptors were ligated to 5' ends of the above bluntended cDNA duplexes, using the adaptor cloning method described in Haymerle et al., Nucleic Acids Res. 14:8615, 1986. Under the described conditions, only the 24-mer oligonucleotide (top strand) will covalently bond to the cDNA during the ligation reaction. The non-covalently bound adaptors (including the complementary 20-mer oligonucleotide described above and any unligated adaptors) were removed by gel filtration chromatography at 65°C, leaving 24 nucleotide non-selfcomplementary overhangs on the cDNA termini.

The adaptored cDNA was inserted into adaptored pDC303, a mammalian expression vector that also replicates in E. coli. pDC303 was assembled from pDC201 (a derivative of pMLSV, previously described by Cosman et al., Nature 312: 768, 1984), SV40 and cytomegalovirus DNA and comprises, in sequence with the direction of transcription from the origin of replication, the following components: SV40 sequences from coordinates 5171-270 containing the origin of replication, enhancer sequences and early and late promoters; cytomegalovirus promoter and enhancer regions (nucleotides 671-63 from the sequence published by Boechart et al. (Cell 41:521, 1985); (3) adenovirus-2 from coordinates 5779-6079 containing the first exon of the tripartite leader (TPL), segment 7101-7172 and 9634-9693 containing the second exon and part of the third exon of the TPL and a multiple cloning site (MCS) containing sites for XhoI, KpnI, Smal and BglI; SV40 segments from coordinates 4127-4100 and 2770-2533 containing the polyadenylation and termination signals for early transcription; adenovirus-2 sequences from coordinates 10532-11156 of the virusassociated RNA genes VAI and VAII of pDC201; and pBR322 sequences from coordinates 4363- 2486 and 1094-375 containing the ampicillin resistance gene and origin of replication.

The MP-1 cDNA library in pDC303 was introduced into E. coli strain DHIOB by electroporation. Recombinants were plated to provide approximately 5,000 colonies per plate. These recombinants were pooled to give a bulk stock of approximately 500,000 recombinants for screening.

DNA was prepared from transformed bacteria and isolated by cesium chloride centrifugation.

EXAMPLE 4 Molecular Cloning of Human CD39 cDNA This example describes the isolation of a DNA molecule encoding CD39 from the expression cloning library described in Example 3.

A. Round I: Transfection and Immunoselection The isolated plasmid DNA was transfected into a sub-confluent layer of COS-7 cells using DEAE-dextran and a chloroquine treatment substantially according to the procedures described in McMahan et al., EMBO J. 10:2821; 1991.

COS-7 cells were maintained in transfection and growth medium (Dulbecco's modified Eagles' medium containing 10% fetal calf serum, 50 U/ml penicillin, 50 pg/ml streptomycin, 2 WO 00/23459 PCT/US99/22955 mM L-glutamine and 50 pg/ml gentamicin) and were plated to a density of approximately 1.5 x 106 cells in 10 ml transfection and growth medium in 10 cm dishes. Medium was removed from adherent cells growing in a layer to approximately 70% confluency, and replaced with 10 ml complete medium containing 66.5 pM chloroquine. About 500 pl of a DNA solution (5 pg DNA, 0.5 mg/ml DEAEdextran in transfection and growth medium containing 66.5 pM chloroquine) was added to the cells and the mixture was incubated at 37 0 C in 10 C02 for about five hours.

Following incubation, media was removed and the cells were shocked by addition of 5 ml transfection and growth medium containing 10% DMSO (dimethylsulfoxide) for 2.5 20 minutes.

Shocking was followed by replacement of the solution with 10 ml fresh transfection and growth medium. Twelve plates of cells were grown in culture for two to three days to permit transient expression of the inserted DNA sequences. The cells were trypsinized after about 24 hours of growth in order to remove them from the plates. After an additional one to two days, cells expressing CD39 were selected by panning, essentially as described in Example 2. The cells were incubated in the mAb 73 pan plates for two hours at room temperature, after which unbound cells were removed by gently rinsing three times with PEA/FCS, then three times with PEA.

The cells that were not removed by rinsing were expressing CD39; cells expressing CD39 were lysed by the addition of 700 pl lysing buffer containing sodium dodecyl sulfate (SDS) and incubation for 20 minutes at room temperature. Lysates were transferred from each dish to individual microfuge tubes containing 100 pl of 5 M NaCI. The tubes were capped, mixed thoroughly by inverting about 20 times, and stored at 4°C overnight. After overnight incubation at 4°C, high molecular weight DNA (debris) was removed by centrifugation, and 2 pg of glycogen was added to each supernatant. The supernatants were then extracted twice with phenol/chloroform and once with chloroform/isoamyl alcohol. DNA was ethanol precipitated, washed with 80% ethanol, and vacuum dried. The purified DNA was then electroporated into E. coli, which were then plated out on ampicillin plates. A large-scale transformation was carried out in this manner, yielding a total of approximately 48,000 colonies (sub-library DNA was prepared from the colonies using CsCI; frozen stocks of the colonies were prepared at the same time.

B. Round II: Electroporation and Immunoselection The DNA from sub-library 1 was electroporated into COS cells (10 x 10 cm plates).

Transfected COS cells were incubated, harvested and panned substantially as described for Round I above. DNA was isolated and a sub-library (sub-library 2) of approximately 50,000 independent colonies was prepared substantially as described above.

C. Round III: Electroporation and FACS Selection The DNA from sub-library 2 was electroporated into COS cells (10 x 10 cm plates).

Transfected COS cells were incubated and harvested substantially as described for Round I above.

The harvested cells were analyzed by FACS substantially as described in Example 1 above. A small subpopulation of cells expressing CD39 was observed, and was sorted out from the larger mixture of WO 00/23459 PCT/US99/22955 cells; DNA was isolated from the sorted cells. A sub-library (sub-library 3) of approximately 5,000 independent colonies was prepared. The DNA was pooled into 10 pools of approximately 500 colonies each; isolated DNA and frozen stocks of bacteria were prepared for each pool.

D. Round IV: Transfection and Immunoselection The DNA from sub-library 3 was transfected into COS cells using DEAE-dextran and chloroquine treatment, and incubated, substantially as described for Round I above, except that the cells were incubated on fibronectin-treated, chambered slides (10 slides, 1 for each pool, and 4 control slides) instead of 10 cm plates. After two days of growth, the cells were harvested as described, and analyzed by FACS substantially as described in Example 1 above, as well as by a slide dipping technique. In the slide dipping technique, the slides were incubated with 12 5 I-labeled mAb 73 and fixed with glutaraldehyde. The results were determined by autoradiography using light microscopy to detect cell containing silver granules.

Two pools containing approximately 500 individual clones each were identified as potentially positive for production of CD39. The pools were titered and plated to provide plates containing an average of approximately 150 colonies each. A replicate nitrocellulose filter was prepared from each plate; each plate was then scraped to provide smaller pools of plasmid DNA.

E. Round V: Transfection and Immunoselection COS-7 cells were transfected with the DNA from the smaller pools by DEAE-dextran, according to the same procedure described above. The transfected cells were screened by slide dipping and FACS as described previously. Two of the smaller pools contained clones that were positive for CD39 as indicated by the presence of an expressed gene product that bound mAb 73.

A total of 156 colonies was picked from the replicate filter corresponding to one of the positive smaller pools, and inoculated into culture medium for overnight growth. After overnight growth, the cultures were arranged in a matrix format of 12 rows and 13 columns. Subpools of culture medium were prepared by pooling medium from each row and each column for a total of 24 subpools.

The subpools were used to prepare DNA for a final round of transfection and screening. An intersection of a positive row and a positive column indicated a potential positive colony One potential positive colony clone) was identified.

A streak plate was prepared from the positive clone (clone and minipreps of DNA were made from nine individual colonies from the streak plate. The DNA was digested with Bgl II and analyzed by SDS-PAGE. Nine of nine individual colonies from clone 1 contained identical inserts of 1.8 2.0 Kb. A single isolate that contained the 1.8 2.0 Kb insert was picked and inoculated into ml culture medium for overnight growth. DNA was prepared and sequenced by dideoxynucleotide sequencing. The nucleotide and deduced amino acid sequence of clone 1 is given in SEQ ID NO: 1. A cloning vector containing human CD39 sequence, designated pCD39 was deposited with the American Type Culture Collection, Rockville, MD (ATCC) on September 29, 1992, under the Budapest Treaty, and assigned accession number 69077. A murine homolog of CD39 was isolated by WO 00/23459 PCT/US99/22955 cross-species hybridization; the amino acid sequence of the murine homolog is described in Maliszewski et al., J. Immunol. 153:3574, 1994.

EXAMPLE Preparation of CD39 mAbs This example describes the preparation of additional monoclonal antibodies against CD39, including antibodies against the region that contains apyrase activity. Preparations of purified CD39 fragments exhibiting ADPase activity, for example, or transfected cells expressing such CD39 polypeptides, are employed as immunogens to generate monoclonal antibodies against CD39 using conventional techniques, such as those disclosed in U.S. Patent 4,411,993. DNA encoding CD39 fragments can also be used as an immunogen, for example, as reviewed by Pardoll and Beckerleg in Immunity 3:165, 1995. Such antibodies are useful for interfering with CD39-induced platelet aggregation, as components of diagnostic or research assays for CD39 or CD39 activity, and in affinity purification of CD39.

To immunize rodents, CD39 immunogen is emulsified in an adjuvant (such as complete or incomplete Freund's adjuvant, alum, or another adjuvant, such as Ribi adjuvant R700 (Ribi, Hamilton, MT), and injected in amounts ranging from 10-100 pg subcutaneously into a selected rodent, for example, BALB/c mice or Lewis rats. DNA may be given intradermally (Raz et al., Proc. Natl. Acad.

Sci. USA 91:9519, 1994) or intramuscularly (Wang et al., Proc. Natl. Acad. Sci. USA 90:4156, 1993); saline has been found to be a suitable diluent for DNA-based antigens. Ten days to three weeks days later, the immunized animals are boosted with additional immunogen and periodically boosted thereafter on a weekly, biweekly or every third week immunization schedule.

Serum samples are periodically taken by retro-orbital bleeding or tail-tip excision for testing by dot-blot assay (antibody sandwich), ELISA (enzyme-linked immunosorbent assay), immunoprecipitation, or other suitable assays, including FACS analysis. Following detection of an appropriate antibody titer, positive animals are given an intravenous injection of antigen in saline.

Three to four days later, the animals are sacrificed, splenocytes harvested, and fused to a murine myeloma cell line NSI or preferably Ag 8.653 [ATCC CRL 1580]). Hybridoma cell lines generated by this procedure are plated in multiple microtiter plates in a selective medium (for example, one containing hypoxanthine, aminopterin, and thymidine, or HAT) to inhibit proliferation of nonfused cells, myeloma-myeloma hybrids, and splenocyte-splenocyte hybrids.

Hybridoma clones thus generated can be screened by ELISA for reactivity with CD39, for example, by adaptations of the techniques disclosed by Engvall et al., Immunochem. 8:871, 1971 and in U.S. Patent 4,703,004. A preferred screening technique is the antibody capture technique described by Beckman et al., J. Immunol. 144:4212, 1990. Positive clones are then injected into the peritoneal cavities of syngeneic rodents to produce ascites containing high concentrations mg/ml) of anti- CD39 monoclonal antibody. The resulting monoclonal antibody can be purified by ammonium sulfate precipitation followed by gel exclusion chromatography. Alternatively, affinity chromatography based WO 00/23459 PCT/US99/22955 upon binding of antibody to protein A or protein G can also be used, as can affinity chromatography based upon binding to CD39 polypeptide. An alternative strategy is to employ full-length CD39 immunogen, selecting for antibodies that bind CD39, and winnowing out those that bind to previously defined epitopes, for example by screening with a fragment of CD39 that represents a previously defined epitope.

Monoclonal antibodies are also prepared by immunizing CD39 knockout mice, such as those described in Example 19D, with CD39 immunogen. Since the entire CD39 sequence is seen as "foreign" in the knockout mice, this strategy can lead to the generation of antibodies recognizing epitopes that are shared across species lines, including antibodies that antagonize or agonize CD39 bioactivity.

EXAMPLE 6 Physiological Activity of CD39 This example demonstrates that CD39 is the endothelial cell ecto-ADPase responsible for inhibition of platelet function. Human umbilical vein endothelial cells (HUVEC) constitutively inhibit platelet responsiveness to prothrombotic stimuli by catabolism of exogenous platelet-derived ADP.

The endothelial ecto-ADPase has been identified as CD39 (Marcus et al., J. Clin. Invest. 99:1351, 1997). Anti-CD39 antibodies immunoprecipitated ADPase activity from a preparation derived from endothelial membranes, and COS cells transfected with an expression vector comprising CD39 acquired ecto-ADPase activity whereas COS cells transfected with a control vector did not. Ecto- ADPase activity was measured in a manner similar to that described in Marcus et al., J. Clin.

Investigation 88:1690, 1991, by conversion of 1 4 C-ADP to AMP by transfectant monolayers as well as membrane preparations, and was greater than or comparable to activity of intact HUVEC monolayers and solubilized membranes HUVEC mRNA was analyzed by RT-PCR using primer pairs derived from the sequence of the human CD39 lymphoid cell activation antigen with emphasis on its N-terminal portion. Lymphoid CD39 cDNA was used for direct comparison of PCR product sizes. The data demonstrated identity between HUVEC and lymphoid CD39 in the 4 fragments spanning the portion analyzed (approximately 1250 of the 1850 bp of lymphoid CD39). Northern blot analyses revealed that the mRNA for CD39 in HUVEC was expressed in the same band pattern as in MP-I cells, from which CD39 was originally cloned.

Confocal microscopy and fluorescence activated cell sorting, using mAb73, were used to determine if HUVEC cells expressed CD39. The FACS protocol was substantially as described in Example 1. For confocal microscopy, cells (human umbilical vein endothelial or transfected COS-1 cells) grown on coverslip glass were washed with PBS and fixed with 3% paraformaldehyde for minutes at room temperature. Auto fluorescence was quenched by treatment with 50mM NH 4 CI for minutes. Cells were then incubated in PBS containing 5% NGS (normal goat serum) plus 0.1% triton X-100 to block non-specific binding and to permeabilize cells. Cells were then incubated with WO 00/23459 PCT/US99/22955 anti-CD39 antibody at 5 pg/ml in PBS containing 5% NGS triton X-100 for 1 hour at room temperature. Following three washes with PBS containing 5% NGS triton X-100 cells were incubated with goat anti-mouse labeled with Texas Red (Molecular Probes) at 5 pg/ml in addition to mM YOYO (Molecular Probes) for nucleic acid counter stain, for 1 hour at room temperature.

Cells were washed 3 times with PBS containing 5% NGS triton X-100 and mounted in 100 mg/ml DABCO (1,4 diaxabicyclo octane) (Sigma) in 50% glycerol. Cells were then viewed with Multiprobe 2001 laser scanning confocal microscope (Molecular Dynamics). One image was collected of CD39 staining (Texas Red) and a second image was collected of cell nuclei (YOYO).

Both the confocal microscopy and FACS experiments demonstrated that HUVEC express CD39. The patterns of expression were similar to those seen in cells transfected with full-length human CD39.

The physiological activity of CD39 was illustrated by the ability of CD39-transfected COS cells to inhibit and completely reverse platelet aggregation by 10 pM ADP. CD39-transfected COS cells, as well as MP-I cells and HUVEC, metabolized this quantity of ADP to AMP within three to four minutes and, when added to platelet rich plasma (substantially as described in Marcus et al., supra), they rapidly reversed platelet aggregation. This activity occurred within the time frame of platelet adhesion to injured subendothelium, a process leading to immediate ADP release, recruitment of additional platelets and formation of a hemostatic plug or thrombus. This time course paralleled platelet inhibition by CD39-expressing cells, and was commensurate with their ADPase activities.

The activity of ADPase/CD39 was independent of formation of other known thromboregulators, nitric acid or prostacyclin. These results demonstrate the importance of ADPase/CD39 as a physiological, constitutively expressed endothelial cell thromboregulator.

EXAMPLE 7 Phosphate Release Assay for ATPase Activity This example describes an ATPase assay that may be used to track enzyme activity. Samples (approximately 100 pl of either concentrated CM or purified polypeptide) are combined with 20 pl of assay buffer (200 mM HEPES, 1.2 M NaCI, 50 mM KCI, 15 mM CaCl 2 15 mM MgCl2 and 3 mM ATP) and sterile water is added to a final volume of 200 pl. Radiolabeled ATP (0.8 pCi 7 32

P]

ATP; Amersham, Arlington Heights, IL) is added and the mixture incubated for 20 minutes at 37 0

C.

Stop mix (0.5 ml of 20% activated charcoal/1 M HC1) is added and the reaction is placed on ice for minutes. After centrifugation (14K rpm for 10 minutes), the supernatant is assayed for free 3P using a scintillation counter. Data are expressed as raw counts or net counts, or as picomoles of ATP degraded per minute.

EXAMPLE 8 Binding Assay This example describes an assay to asses the binding of CD39 polypeptides to CD39 antibodies by biospecific interaction analysis (BIA) using a biosensor, an instrument that combines a WO 00/23459 PCT/US99/22955 biological recognition mechanism with a sensing device or transducer. An exemplary biosensor is BIAcoreTM, from Pharmacia Biosensor AB (Uppsala, Sweden; see Fagerstam Techniques in Protein Chemistry II, ed. J.J. Villafranca, Acad. Press, NY, 1991). BIAcoreTM uses the optical phenomenon surface plasmon resonance (Kretschmann and Raether, Z. Naturforschung, Teil. A 23:2135, 1968) to monitor the interaction of two biological molecules. Molecule pairs having affinity constants in the range 105 to 1010 M 1 and association rate constants in the range of 103 to 106 Ms-1, are suitable for characterization with BIAcoreTM.

The biosensor chips are coated with CD39 antibody mAb73). The different constructs of CD39 to be assessed are then added at increasing concentrations; the chip is regenerated between the different constructs, for example, by the addition of sodium hydroxide. The resultant data can analyzed to qualitatively or quantitatively asses production of CD39 polypeptides. Affinity of the CD39 polypeptides for the CD39 antibodies can also be determined. In a similar manner, other monoclonal antibodies or polypeptides that specifically bind CD39 can be immobilized on a biosensor chip to asses the binding of various CD39 polypeptides.

EXAMPLE 9 Transient Expression of Soluble CD39 Polypeptides This example describes the preparation of constructs for the transient expression of soluble CD39 polypeptides.

A. Reagents Used The B73 mAb, a murine IgGI recognizing human CD39, was kindly provided by Dr. Guy Delespesse Montreal, Quebec, Canada). The M2 mAb recognizing the FLAG® peptide (DYKDDDDK, SEQ ID NO:10), a murine IgG1, was prepared at Immunex Corp. Affigel 10 (Bio- Rad, Hercules, CA) and CNBr-activated Sepharose 4B (Pharmacia Biotech, Piscataway, NJ) immunoaffinity columns were prepared according to manufacturers' instructions. Typically, coupling efficiencies in the range of 3-5 mg mAb per ml of affinity gel slurry were obtained.

B. Construction of a Soluble CD39 (solCD39) Expression Plasmid To generate a soluble molecule having the properties of CD39 the N-terminal and C-terminal portions of CD39, including the two transmembrane regions (see Fig. were removed. To allow transport of soluble CD39 into the medium, a leader sequence providing for secretion was added at the amino terminus of the polypeptide.

Constructs of soluble human CD39 (solCD39) were made in the mammalian expression vector pDC206 (Kozlosky et al. Oncogene. 10:299, 1995), utilizing human IL2 (huIL2), human growth hormone (huGH) and murine 1L7 (mulL7) leaders).

WO 00/23459 PCT/US99/22955 The DNA sequences between the putative transmembrane regions of full-length CD39, including nucleotides 178-1494 of SEQ ID NO:1, were amplified using PCR and the C-terminal transmembrane coding region was replaced with a stop codon. The PCR product was fused to a synthetic DNA fragment encoding an 8 amino acid peptide tag (FLAG®) and ligated with a muIL7 leader (muIL7L) into the plasmid pDC206 vector via Spel and Bgl2 restriction sites. This construct encoded N-terminally FLAG-tagged solCD39.

Alternate leaders were introduced by ligating the Spel/Bgl2 FLAG-solCD39 fragment into two different pDC206 plasmids, with leaders derived from: human growth hormone (huGHL), and a human proinsulin/IL2 fusion polypeptide (huIL2L, Cullen, DNA. 7:645, 1988). The coding region of the latter construct, which is shown in SEQ ID NOs:25 and 26, includes sequences encoding the huIL2 leader (hulL2L, nucleotides 1-72, amino acids 1-24 in SEQ ID NO:25), the first 12 amino acids of mature human IL2 (nucleotides 73-108, amino acids 25-36 in SEQ ID NO:25), a four amino acid linker (nucleotides 109-120, amino acids 37-40 in SEQ ID NO:25), the FLAG tag (nucleotides 121-144, amino acids 41-48 of SEQ ID NO:25), and sol CD39 (nucleotides 145-1461, amino acids 49- 487 of SEQ ID The constructs comprising the muIL7 leader, the human growth hormone leader, and the human proinsulin/IL2 leader were designated pIL7LFlagSolCD39, pGHLFlagSolCD39, and pIL2LFlagSolCD39 respectively.

Each construct was used to transiently transfect subconfluent layers of COS-1 cells using DEAE dextran followed by chloroquine as described by Cosman et al., Nature 312:768, 1984. As a negative control, a CD40 ligand construct (pIl2LCD401ig, Spriggs et al., J. Exp. Med. 176:1543, 1992) was also transfected into COS-1 cells.

C. Preparation of Conditioned Medium from solCD39 Transfectants The transfected COS-1 cells were incubated (37 0 C, 5% C0 2 in 0.5% FCS-supplemented DMEM-F12 medium in 10 cm 2 Petri dishes or 175 cm 2 tissue culture flasks. After 5 days, conditioned medium (CM) from these cultures was collected, and cells and debris were removed by centrifugation.

The CM was concentrated 4-10 fold using a pressurized, stirred cell fitted with a YM-10 (10 kD cutoff) membrane (Amicon Corp., Danvers, MA).

D. ATPase Activity in Conditioned Medium from solCD39 Transfectants ATPase activity in the CM from solCD39 transfectants (100 gL of 10-fold concentrated supernatant) was assayed essentially as described in Example 7, except that the 10X assay buffer contained 30 mM cold ATP. The results are shown in TABLE 1.

The transfections were repeated and the CM (10, 20 and 30 gL, unconcentrated) was assayed essentially as described in Example 7. The results are shown in TABLE 2. Because the pIL2LFlagSolCD39 showed higher ATPase activity in COS-1 supernatants than pIL7LFlagSolCD39 and pGHLFlagSolCD39, this construct was selected for further investigation. ATPase levels in CM WO 00/23459 PCT/US99/22955 from COS-I cells transfected with plL2LFlagSolCD39 increased with time in culture over at least 4 days post-transfection.

TABLE 1 ATPase Activity in Concentrated CM from solCD39 Transfectants Sample CPM Release x 10 (raw counts) pl2LFlagSolCD39 99.96 pIL7LFlagSolCD39 39.47 pGHLFlagSolCD39 21.14 pIL2LCD401ig 10.53 media only 7.89 TABLE 2 ATPase Activity in CM from solCD39 Transfectants Sample Vol (gL) CPM Release x 10 3 (net counts) pIL2LFlagSolCD39 10 24.71 43.92 56.93 pIL7LFlagSolCD39 10 5.01 9.75 14.23 pGHLFlagSolCD39 10 5.51 7.22 9.95 E. Immunoaffinity Depletion of solCD39 from COS-1 CM To confirm that recombinant solCD39 accounted for the ATPase activity observed in the CM, CM from COS-I transfectants was incubated with immunoaffinity beads prior to enzyme assay.

CM was collected from COS-I cells transfected with pIL2LFlagSolCD39, which had been cultured for 5 days in DMEM/FI2 supplemented with 5% FCS. A 100 pl aliquot of drained Affigel beads (AG) conjugated with either chicken ovalbumin, antiFLAG mAb, or anti-CD39 mAb was added per ml of CM. CM was subjected to one or two cycles of binding with one of the following: ovalbumin-conjugated AG, M2 mAb-conjugated AG, or B73 mAb-conjugated AG. Each cycle involved continuous gentle agitation of the slurry for 14 h at 4°C followed by centrifugation to recover supernatants for a subsequent binding cycle or for ATPDase activity measurements.

As shown in Fig. 3, immunoprecipitation with anti-CD39 mAb-conjugated beads resulted in removal of over 80% of ATPase activity from CM. Over 95% ATPase activity was removed with a second antibody adsorption step. Immunoprecipitation (2 cycles) with anti-FLAG mAb-coated beads also resulted in substantial depletion of enzyme activity. Two rounds of preincubation with a control (ovalbumin-conjugated beads) did not remove significant ATPase activity from the supernatants.

WO 00/23459 PCT/US99/22955 F. Immunoprecipitation of Recombinant solCD39 To characterize recombinant solCD39 polypeptide expression, COS-1 cells were transfected with mammalian expression vectors encoding cell surface CD39 (pHuCD39, Marcus et al., J. Clin.

Invest. 99:1351, 1997), tagged soluble CD39 (pIL2LFlagSolCD39), or soluble CD40 ligand (plL-2L- CD40L) and grown in 5% FCS-supplemented DMEM/F12 medium in 10 cm 2 dishes. Two days after transfection, the medium was replaced with Cys/Met-free medium and cells were incubated for 1 h at 37°C. The culture medium was replaced with fresh Cys/Met-free medium supplemented with 5 pl of (Amersham, Arlington Heights, IL) in order to label newly synthesized polypeptides, and cells were cultured for 5 h at 37°C. CM from the metabolically radiolabeled cells was collected, purified of cells and debris by centrifugation and sterile filtration, and stored at 4 0 C until further use.

For radioimmunoprecipitation, 500 pl of 3 5 S-labeled CM was added to 250 pl of 3% BSA in Tris-buffered saline (TBS), pH 7.7, followed by addition of 50 pl of a 80% slurry of mAb-coated AG beads. In some cases, 35S-labeled CM were incubated with ovalbumin-coated AG beads to remove nonspecifically binding materials prior to addition of Ab-coated beads. After incubation for 14 h at 4°C, beads were removed by centrifugation and washed three times in cold TBS.

For SDS-PAGE analysis, 35 pl of 4-fold concentrated reducing sample buffer (250 mM Tris/HCl, pH 6.8, 8 SDS, 40 glycerol, 20% 2-mercaptoethanol, 0.05% bromophenol blue dye) was added to each AG pellet, boiled for 5 min, and loaded onto a 8-16% Novex (San Diego, CA) polyacrylamide gel. Gels were electrophoresed at 25 mA, prepared for autoradiography by soaking in Enhance (NEN Life Science Products, Boston, MA) for 1 h and in for 20 min, followed by vacuum drying at 80 0 C. Gels were exposed to Kodak (Rochester, NY) Xomat AR film for 2 h, then developed.

As shown in Fig. 4, IL-2L-FlagsolCD39 transfectants secreted a radiolabeled protein of -66 kD that was recognized by anti-CD39. This protein was not detected in anti-CD39 immunoprecipitated CM from COS-1 cells transfected with a vector encoding full-length CD39 (including N-terminal and C-terminal hydrophobic regions), or with a vector encoding a secreted protein, CD40 ligand. Anti-FLAG mAb immunoprecipitated a similar-sized band from CM of the plL-2L FlagsolCD39 transfectant, consistent with the presence of the FLAG® peptide in recombinant solCD39. Preclearing radiolabeled culture superatants with anti-ovalbumin-coated beads failed to remove the 66 kD band, indicating that binding to anti-CD39 and anti-FLAG was specific. Beads coated with an irrelevant, isotype matched control antibody failed to immunoprecipitate the 66 kD band from solCD39 containing CM.

G. Preparation of Additional solCD39 Fusion Constructs Restriction enzymes were used to prepare a DNA fragment comprising nucleotides 1 through 1488 of SEQ ID NO:1, the coding region of which would be expected to encode a fragment of CD39 lacking the second (most C-terminal) transmembrane domain. Appropriate linker oligonucleotides WO 00/23459 PCT/US99/22955 were prepared (SEQ ID NOs:14 and 15), and used in a three-way ligation of the CD39 DNA, the linker oligonucleotides, and an expression vector comprising regulatory elements allowing expression of a resulting fusion polypeptide in mammalian cells along with DNA encoding a mutated human immunoglobulin Fc (SEQ ID NOs:16 and 17) region that exhibits reduced affinity for Fc receptors (nucleotides 42 through 740 of SEQ ID NO: 16). This construct was referred to as CD39A2Fc, and when transfected into cells resulted in the expression of a fusion polypeptide comprising amino acids 1 through 474 of SEQ ID NO:1 and amino acids 1 though 232 of SEQ ID NO:17, which could be detected on the surface of transfected cells using either anti-CD39 or anti-human IgG.

A PCR technique was employed to prepare a fragment of DNA from the CD39A2Fc construct that also lacked the first, most amino-proximal transmembrane region, but included CD39 DNA from nucleotides 181 through 1488 (amino acids 39 through 474) of SEQ ID NO:1 and the Fc mutein DNA from CD39A2Fc (using the linkers shown in SEQ ID NOs:18 and 19).

The resulting DNA was then ligated into an expression vector that included DNA encoding the murine Interleukin-7 leader sequence (SEQ ID NO:20) ligated immediately proximal to the CD39encoding sequence. This construct was designated CD39Al,A2Fc.

DNA encoding the FLAG® peptide (SEQ ID NO:10) and a codon corresponding to nucleotides 178, 179 and 180 of SEQ ID NO:I was inserted into the CD39Al,A2Fc construct, in between the leader sequence and the CD39-encoding sequence, to provide a detectable tag for the amino terminus of the fusion polypeptide (using the linkers shown in SEQ ID Nos:21 and 22). The tagged construct was referred to as FlagCD39Al,A2Fc.

Another construct was prepared that removed the Fc mutein sequences, and added codons corresponding to nucleotides 1489 through 1494 of SEQ ID NO:1 immediately downstream of the CD39 sequences (using the linkers shown in SEQ ID Nos:23 and 24). This construct was designated FlagCD39Al,A2.

Each of the constructs was transfected into mammalian cells and protein levels were assayed on the surface, in the interior, or in the supernatant fluid of transfected cells using antibodies to FLAG®, CD39, or human IgG.

EXAMPLE Expression and Activity of pIL2LSolCD39 To facilitate the establishment of a stably producing transfectant in CHO cells, an untagged version of soluble human CD39 (solCD39) was constructed. A 523 bp Spel/Ndel fragment containing the FLAG® tag and the first 163 amino acids (aa) of pIL2LFlagSolCD39 was removed, and replaced with a similar fragment from a C-terminally FLAG®-tagged solCD39. Thus the entire solCD39 coding region was reconstituted, sans FLAG®, while retaining the HuIL2 leader and mature IL2 residues. This construct was designated pIL2LSolCD39. The coding region of pIL2LSolCD39, WO 00/23459 PCT/US99/22955 which is shown in SEQ ID NO:7, includes sequences encoding the huIL2 leader (huIL2L, nucleotides 1-72, amino acids 1-24 in SEQ ID NO:7), the first 12 amino acids of mature human IL2 (nucleotides 73-108, amino acids 25-36 in SEQ ID NO:7), a three amino acid linker (nucleotides 109-117, amino acids 37-39 in SEQ ID NO:7), and sol CD39 (nucleotides 118-1434, amino acids 40-478 of SEQ ID NO:7).

To determine whether activity was affected by removal of the N-terminal FLAG® tag, COS-I cells were transfected with plL2LFlagSolCD39 and pIL2LSolCD39 and supernatants (sups) were harvested after 5 days. Samples of 10, 20 and 30 tL of lx sups were assayed for ATPase activity as described in Example 7. As shown in TABLE 3, activity was not affected by removal of the Nterminal FLAG® tag.

TABLE 3 ATPase Activity in CM from pIL2LFlagSolCD39 and pIL2LSolCD39 Transfectants Sample Vol (glL) CPM x 103 (net counts) pIL2LFlagSolCD39 10 23.6 40.2 54.4 pL2LSolCD39 10 20.1 36.0 51.0 EXAMPLE 11 Preparation of Additional solCD39 Fusion Constructs A. Preparation and Characterization of Triml and Trim2 Variants To characterize the effect of the 12 mature human IL2 (huIL2) residues on the expression of solCD39, the huDL2 residues were removed during the construction of nucleic acid sequences encoding two additional variants of pIL2LSolCD39: pIL2LTriml ("Triml") and pIL2LTrim2 ("Trim2").

The pIL2LTriml variant was constructed by purifying a Hind3/Bgl2 restriction fragment from pIL2LSolCD39 which contained the entire solCD39 coding region except for the first four amino acids. This fragment was ligated with a synthetic oligo cassette (containing the huIL2 leader and the first amino acid of mature huIL2) into Smal/Bgl2 digested pDC206. The huIL2 leader was thus reintroduced and joined to solCD39 with an intervening alanine residue.

The plL2LTrim2 variant was constructed in a similar fashion using a Spel/Bgl2 fragment from pIL2LSolCD39 and a synthetic oligo cassette containing the huIL2 leader and the linker-encoded sequences (with the first codon altered to alanine). Thus, the hulL2 leader was restored with an intervening Ala-Ser-Ser linker preceding solCD39.

The N-terminal portions of the pIL2LSolCD39, pIL2LTriml and pIL2LTrim2 polypeptides are compared below, with the predicted cleavage points indicated as WO 00/23459 PCT/US99/22955 pIL2LSolCD39 (SEQ ID NO:11) MALWIDRMQLLSCIALSLALVTNS*APTSSSTKKTQLts sT QNK...

pIL2LTriml (SEQ ID NO:12) MALWIDRMQLLSCIALSLALVTNS A T*QNK...

pIL2LTrim2 (SEQ ID NO:13) MALWIDRMQLLSCIALSLALVTNS as*sT QNK...

The polypeptide encoded by the Triml construct has the sequence SEQ ID NO:27. Residues 26-464 are a soluble portion of CD39 and the cleavage of the leader sequence is between Ser24 and The expression of Triml and Trim2 constructs was analyzed in COS-] cells cultured in 10 cm plates. After 5 days of incubation, Ix supematants were examined via ELISA (using anti-CD39) and the phosphate-release assay described in Example 7. As shown in TABLE 4, the specific activities (based on concentrations determined by ELISA) of Trim] and Trim2 were equivalent to pIL2LFlagSolCD39. Expression levels, however, appeared to be reduced 3-4 fold.

TABLE 4 SolCD39 Expression and Activity in CM from pIL2LSolCD39 an Trim Transfectants Sample [CD39] pg/mL Activity (pmol ATP/min/pg) x 103 pIL2LSolCD39 0.75 5.67 pIL2LTriml 0.21 8.38 pIL2LTrim2 0.21 6.67 COS-1 cells were also transfected with pIL2LSolCD39, pIL2LTriml and pIL2LTrim2 and cultured in T175 flasks (30 mL). 5-day/lx sups were then analyzed via ELISA. As shown in TABLE 5, the ELISA results again indicated lower expression levels for the Triml and Trim2 variants.

To further characterize the effect of the human IL2 (huIL2) residues on the expression of solCD39, the pIL2LSolCD39, pIL2LTriml, and pIL2LTrim2 products were purified using anti-CD39 coated sepharose. The N-terminal amino acid sequence was determined for each of the purified polypeptides. For solCD39 the N-terminus was APTSSSTKKT. (residues 25-34 of SEQ ID NO:11). For Triml the N-terminus was ATQNKALPEN. .(residues 25-34 of SEQ ID NO:27). The Trim2 polypeptides had heterogeneous N-termini.

TABLE SolCD39 Expression in CM from pIL2LSolCD39 and Trim Transfectants Sample [CD39] gg/mL pIL2LSolCD39 0.796 pIL2LTriml 0.143 pIL2LTrim2 0.113 WO 00/23459 PCT/US99/22955 B. Preparation and Characterization of Trim3 and Trim4 Variants Nucleic acids encoding additional solCD39 variants, designated pIL2LTrim3 ("Trim3") and pIL2LTrim4 ("Trim4"), are also constructed using a synthetic oligo cassette strategy. The N-terminal portions of the solCD39, Trim3 and Trim4 polypeptides are compared below. The predicted cleavage points are indicated as pIL2LSolCD39 MALWIDRMQLLSCIALSLALVTNS*APTSSST KKTQLtssTQNK (SEQ ID NO:11) pIL2LTrim3 MALWIDRMQLLSCIALSLALVTNS*A ST KKTQLtssTQNK...

(SEQ ID NO:28) pIL2LTrim4 MALWIDRMQLLSCIALSLALVTNS ST*KKTQLtssTQNK (SEQ ID NO:29) The polypeptide encoded by the Trim3 construct has the sequence SEQ ID NO:28. Residues 36-474 are a soluble portion of CD39 and the predicted cleavage of the leader sequence is between Ser24 and Ala25. The polypeptide encoded by the Trim4 construct has the sequence SEQ ID NO:29.

Residues 35-473 are a soluble portion of CD39 and the predicted cleavage of the leader sequence is between Thr26 and Lys27.

The pIL2LTrim3, and pIL2LTrim4 polypeptides are expressed and purified using anti-CD39 coated sepharose. The N-terminal amino acid sequence and specific activity are determined for each of the polypeptides.

C. Preparation and Characterization of solCD39-L4 Fusion Polypeptides The CD39 gene family is reported to contain at least four human members: CD39, CD39L2, CD39L3, and CD39L4 (Chadwick and Frischauf, Genomics 50:357, 1998). CD39-L4 is reported to be a secreted apyrase (Mulero et al., J. Biol. Chem. 274(29):20064, 1999). Additional solCD39 variants are constructed by fusing N-terminal sequences from CD39L2, CD39L3, or CD39L4 to a soluble portion of CD39. The N-terminal amino acid sequences of human CD39 and human CD39-L4 are aligned in Fig. 24.

For one construct, CD39-L4-1, a nucleic acid encoding CD39-L4 amino acid residues 1-37 (Metl to Ser37 of SEQ ID NO:31) is fused to a nucleic acid encoding CD39 residues 38-476 (Thr38 to Thr476 of SEQ ID NO:2). The polypeptide encoded by the CD39-L4-1 construct has the sequence SEQ ID NO:3. Residues 1-37 are from CD39-L4, residues 38-476 are a soluble portion of CD39, and the predicted site of cleavage of the leader sequence is between Ala20 and Val21.

For another construct, CD39-L4-2, a nucleic acid encoding CD39-L4 amino acid residues 1- 48 (Metl to Leu48 of SEQ ID NO:31) is fused to a nucleic acid encoding CD39 residues 49-476 (Tyr49 to Thr476 of SEQ ID NO:2). Another construct, CD39-L4-3, is identical to CD39-L4-2 except that the Cys residue at position 39 (Cys39) is replaced by another amino acid, preferably Ser. The polypeptides encoded by the CD39-L4-2 and CD39-L4-3 constructs have the sequence SEQ ID NO:4.

WO 00/23459 PCT/US99/22955 Residues 1-48 are from CD39-L4, residues 49-476 are a soluble portion of CD39, and the predicted site of cleavage of the leader sequence is between Ala20 and Val21.

Additional constructs are constructed by fusing a portion of the CD39-L4 N-terminal coding region to the CD39 N-terminal coding region. After expression in recombinant cells, the N-terminal sequence, enzymatic activity, and platelet inhibitory activity is determined for each of the polypeptide products.

D. Preparation and Characterization of IgkappaLsolCD39 Nucleic acids encoding an Igkappa leader sequence fused to amino acids from IL-2 and to solCD39 are also constructed. One such construct encodes a polypeptide having an Igkappa leader and four amino acids from IL-2 fused to the N-terminus of the CD39 soluble portion (set forth as residues Thr38 to Thr476 of SEQ ID NO:2). The N-terminus of the encoded polypeptide is therefore: (amino acids 1-32 of SEQ ID where amino acids Metl-Gly20 are the Igkappa leader, Ala21-Ser24 are from IL-2, and Thr25-Glu32 is the beginning of solCD39 sequences. The predicted cleavage site is indicated as The polypeptide encoded by the IgkappaLsolCD39 construct has the sequence SEQ ID NO:30. Residues 25-463 are a soluble portion of CD39 and the predicted cleavage of the leader sequence is between and Ala21. After expression in recombinant cells, the N-terminal sequence, enzymatic activity, and platelet inhibitory activity is determined for each polypeptide product.

EXAMPLE 12 Development of a Stably Transfected Cell Line Secreting solCD39 A CHO cell line expressing solCD39 was generated to improve recombinant solCD39 polypeptide production.

A. Preparation of Constructs and Cell Lines for the Stable Expression of Soluble CD39 Polypeptides The solCD39 cDNA insert, containing the recombinant solCD39 sequence and the IL-2 leader but not the FLAG® sequence, was excised from the pIL2LSolCD39 plasmid by XmaI/BglII digestion, then inserted into 2A5Ib, an expression vector containing the DHFR gene and optimized for stable CHO cell lines (Morris et al., In Animal Cell Technology: From Vaccines to Genetic Medicine, M.J.T. Carrondo, B. Griffiths, and J.L.P. Moreira, editors, Kluwer Academic Publishers, Boston. 529- 534, 1997).

The solCD39-2A51b plasmid was transfected into CHO cells using Lipofectamine (GIBCO BRL; Gaithersburg, MD) according to manufacturer's recommendations. The CHO cell line used in these studies, DX B-I 1, had been adapted to serum-free suspension culture conditions. Transfected cells were grown in modified DMEM-F12 medium, supplemented with peptone, glutamine, glucose, transferrin, lipids, and IGF-I (insulin-like growth factor 1; used solely when cultures were induced for protein expression). After 3 days growth, the cells were transferred to selective medium lacking WO 00/23459 PCT/US99/22955 hypoxanthine and thymidine. Stock cultures were grown at 37°C in suspension, and passaged every 2- 3 days. Induction cultures were grown at 31 C in suspension, with IGF-I and sodium butyrate (1-2 mM). Cell density at start of induction cultures was 1.5-2x106 cells/ml. The average induction period was 7 days, at which time CM was collected for further analyses.

B. TLC Assays for ADPase and ATPase Activities in CM Containing solCD39 Following growth in selective medium, CM from CHO cell cultures was analyzed for ATPase and ADPase activities. ADPase assays (Marcus et al., J. Clin. Invest. 88:1690, 1991) were primarily used in determining enzyme kinetics and pharmacokinetics. Test samples were incubated with 50 pM 4 C] ADP (NEN Life Science Products) in assay buffer (15 mM Tris, 134 mM NaCl, 5 mM glucose, pH 7.4, containing 10 pM Ap5A (P',P 5 -di[adenosine-5']pentaphosphate, I mM ouabain, 10 pM dipyridamole, and 3 mM CaCl2) in a total volume of 50 pl (5 min, 37 0 Reactions were stopped by placement on ice and addition of 10 pl "stop solution" (160 mM disodium EDTA, pH 7.0, 17 mM ADP, 0.15 M NaCI) to block further metabolism of ADP. Nucleotides, nucleosides, and bases were separated by TLC using isobutanol/1-pentanol/ethylene glycol monoethyl ether/NH 4 0H/H 2 0 (90: 180: 90: 120). Radioactivity was quantified by radio TLC scanning (RTLC multiscanner; Packard, Meriden, CT). Results were calculated as averages of duplicate to quadruplicate measurements after subtraction of buffer blanks (consistently of total radioactivity). Data were expressed as percentage of ADP metabolized or as pmol ADP metabolized per minute per pl CM. A unit of activity is the quantity of enzyme which will degrade 1 pmol of ADP in 1 min at 37 0 C. Identical assays were performed using ATP as a substrate in order to examine the kinetics of the ATPase activity of CD39.

As shown in TABLE 6, the stably transfected CHO cells secreted 20-fold higher levels of both enzyme activities compared to CM from transfected COS-I cells.

TABLE 6 Comparison of ADPase and ATPase Activities in CM Containing solCD39 Cell Type ADPase (pmol/min/pl) ATPase (pmol/min/pl) CD39 (CHO) 1403 970 CD39 (COS-1) 70 44 C. Affinity Purification of solCD39 from Stably Transfected CHO Cells Thirty ml of 10-fold concentrated CM from solCD39-secreting CHO cells was added to 3 ml of B73 mAb-coated Sepharose 4B gel slurry and mixed overnight at 4 0 C. The affinity matrix was pelleted by centrifugation, washed 3 times with PBS, and added to a plastic column. Specificallybound protein was eluted by the addition of 0.1 M triethylamine, pH 11.5. Fractions (1.2 ml each) were collected in tubes containing 120 pl of neutralizing solution (1 M sodium phosphate, monobasic; pH 4.3) and analyzed for protein content by SDS-PAGE, followed by Coomassie Blue staining.

WO 00/23459 PCT/US99/22955 Biological activity was determined using an ATPase assay as described in Example 7, so that peak fractions could be pooled, buffer exchanged into PBS, and concentrated 5-fold. N-linked sugars were removed from purified protein using a kit from Oxford Glycosystems (Rosedale, NY). The recombinant solCD39 was analyzed by SDS-PAGE.

A prominent band of -66 kD was present in early eluted fractions, with a peak of Coomassie Blue staining around fraction 5 (Fig. 5A). Over 90% of the protein present was found as this major band. Overloading the polyacrylamide gel did reveal some smaller molecular weight contaminants, however, these appear to be related to the antibodies present on the column and not to the protein loaded on the column.

ATPDase activity of the affinity column fractions correlated with the intensity of protein bands on SDS-PAGE (Fig. 5A, 5B). ATPDase activity was barely detectable in the anti-CD39 column flowthrough, indicating that affinity purification is an effective means of isolating biologically active recombinant solCD39. Treatment of the purified protein with N-glycanase for 18 hours to remove N-linked oligosaccharides caused the broad band of protein at 66 kD to resolve into a much tighter band of protein at approximately 52 kD, the predicted size for solCD39 (Fig. The total protein yield from 1 L of CHO-solCD39 CM was -2 mg. Production of solCD39 was scaled up to 10 liter bioreactors. The resultant conditioned medium contained approximately 100 pg/ml of solCD39 according to ELISA analysis. Thus, each 10 L bioreactor run would expected to produce 500-1000 mg of recombinant polypeptide. CHO cell lines expressing additional solCD39 constructs are similarly prepared and characterized.

EXAMPLE 13 Expression of solCD39 in Veggie-CHO and CS-9 Cells In this example, soluble CD39 is expressed in CHO cells that have been adapted to grow in suspension in media that does not contain animal proteins (see Rasmussen et al., Cytotechnology 28:31, 1998), or in the presence of IGF-1 in the clonal cell line CS-9.

The dihydrofolate reductase-deficient Chinese hamster ovary cell line, DXB 1 I-CHO is commonly used as a host cell for the production of recombinant polypeptides. DXB 1 -CHO was adapted to grow in suspension. A serum-free host named Veggie-CHO was then generated by adapting DXB 11 -CHO cells to growth in serum-free media in the absence of exogenous growth factors such as Transferrin and Insulin-like growth factor (IGF-1). The latter adaptation was achieved by a gradual reduction of serum supplementation in the media and the replacement of serum with low levels of growth factors, IGF-1 and transferrin, in an enriched cell growth media. The suspension adapted serum-free adapted cells were then weaned off these growth factors. The resulting Veggie- CHO cells maintain vigorous growth and high viability as well as a DHFR-deficient phenotype in media that is serum-free and also free of animal-derived proteins. CS-9 cells were also derived from DXB I1-CHO cells. The suspension adapted serum-free adapted cells were adapted to grow in the WO 00/23459 PCT/US99/22955 absence of transferrin, then individual clones were isolated. The CS-9 clone was chosen for its stable recombinant protein expression.

Veggie-CHO cells and CS-9 cells are used as a host cell line for the stable, high level expression soluble CD39 polypeptides using methods similar to those described in Example 12.

EXAMPLE 14 Biochemical Properties of Affinity-Purified solCD39 Purified solCD39 material was subjected to N-terminal amino acid sequencing and mass spectroscopy. Quantitative amino acid analysis of peak fractions from the affinity column yielded a ratio of amino acid residues consistent with calculated values for human CD39. The Nterminus of the pIL2LSolCD39 product had the following sequence: APTSSSTK K T Q LtssT (residues 25-41 of SEQ ID NO: 1).

The first 12 residues represent the mature hulL2 residues; residues 13-15 (tss, lower case) are linker-encoded residues; residues 16,17, etc. (T Q are solCD39.

Using the TLC assay system described in Example 12B, the ADPase activity of the membrane-bound HUVEC ecto-ADPase was determined at different pHs in buffers containing 100 mM bis-trispropane (Sigma, St. Louis, MO). This was compared to the ADPase activity of purified solCD39 at these pHs. Kinetic constants for CD39 metabolism of ATP and ADP were determined by measuring the initial rates of reaction as analyzed in the TLC system. ADP or ATP at 2.5-150 pM were incubated separately with 2x l0 9 M solCD39.

As shown in Fig. 6A, the pH optima for the ecto-ADPase on the surface of HUVEC and for affinity-purified recombinant solCD39 ADPase activities were between pH 8 and 8.5. This indicated that recombinant solCD39 would be maximally active under the same physiological conditions as native CD39/ecto-ADPase.

Initial rates of ATP and ADP metabolism by recombinant solCD39 were determined as shown in Fig. 6B, and kinetic constants were derived. The Km and Vma for ADP were 5.9 pM and 72 pmol/min, respectively; for ATP a Km of 2.1 pM and Vm. of 26 pmol/min were determined. The assays were performed with 2x 10 9 M solCD39, yielding kca, of 720 min"' (ADP) and 260 min"' (ATP).

Thus, the specificity constant, kcat/Km (1.2x 10 min-' was identical for ADP and ATP. The specific activity for purified recombinant solCD39 was 11 U/mg for ADP and 4 U/mg for ATP.

EXAMPLE Platelet Inhibitory Properties of solCD39 This example shows that recombinant affinity purified solCD39 is effective as an inhibitor of platelet activation and recruitment.

After obtaining informed consent from volunteers, blood was collected via plastic tubing using acid citrate-dextrose (38 mM citric acid; 75 mM sodium citrate; 135 mM glucose) as WO 00/23459 PCT/US99/22955 anticoagulant. Where indicated, volunteers had ingested 650 mg acetylsalicylic acid (ASA) 18 h prior to blood donation. Platelet-rich plasma (PRP) was prepared with an initial whole blood centrifugation (200g, 15 min, 25°C), and a second centrifugation of the PRP (90 g, 10 min) to eliminate residual erythrocytes and leukocytes. The stock suspension of PRP was maintained at room temperature under 5% CO2-air.

A. Platelet Aggregation Studies PRP containing 1.22x 108 platelets was pre-incubated (3 min, 37 0 C) in an aggregometer cuvette (Lumiaggregometer; Chrono-Log, Havertown, PA) alone or in combination with test samples containing solCD39. Total volumes were adjusted to 300 pl with TSG buffer (Marcus et al., J. Clin.

Invest. 88:1690, 1991; Marcus et al., J. Clin. Invest. 99:1351, 1997). After the 3 min preincubation, platelet agonists (ADP or collagen) were added at the concentrations indicated, and the aggregation response recorded for 4-5 min. Where indicated, 10 pM indomethacin (Sigma, St. Louis, MO) was added to PRP to inhibit endogenous cyclooxygenase activity.

As shown in Fig. 7, the addition of 10 pM ADP to PRP alone resulted in a full, irreversible aggregation response; partially reversible aggregation occurred at lower ADP concentrations.

However, in the presence of only 3.3 pg/ml solCD39, platelet aggregation induced by 10 pM ADP was abruptly terminated and the curve rapidly returned to baseline. Importantly, the extent of aggregation was reduced to levels below those observed with 1 pM ADP. Higher concentrations of solCD39 had an even more profound inhibitory effect, virtually eliminating the initial burst of aggregation elicited by 10 pM ADP.

Platelet responsiveness to 5 pM ADP was examined in PRP treated with and without the cyclooxygenase inhibitor indomethacin (10 pM), in the presence of CM containing solCD39 from COS-I and CHO cells. As shown in Fig. 8A, indomethacin treatment resulted in partial reversal of ADP-induced platelet aggregation in the absence of solCD39. In contrast, CM containing solCD39 were capable of completely abrogating platelet responses to ADP, whether PRP was indomethacintreated or not.

Inhibition of platelet reactivity by CD39 was not limited to blocking the agonistic effects of ADP, as shown in Fig. 8B and 8C. Collagen, which is another critical platelet agonist, was used at 1 pg/ml to induce platelet aggregation. The presence of solCD39 markedly reduced the response to collagen compared to control (Fig. 8B, upper curves). A similar inhibitory effect of solCD39 was observed in PRP treated with indomethacin (Fig. 8B, lower curves), when collagen was used at 3.3 pg/ml. As shown in Fig. 8C, the effect of solCD39 on collagen-induced aggregation was dose dependent.

B. Inactivation of Enzymatic Activity of solCD39 and the Effect on Inhibition of Platelet Activation To demonstrate that the ability of solCD39 to inhibit platelet activation and recruitment was due to the enzymatic activity of solCD39 and not to some other property, the solCD39 was reacted WO 00/23459 PCT/US99/22955 with FSBA (Fluorosulfonylbenzoyl-adenosine), an ATP analog that inhibits collagen-induced platelet activation (Colman et al., Blood 68:565, 1986) and binds irreversibly with ATPDases found on several cell types (Sevigny et al., Biochem. Biophys. Acta 1334:73, 1997; Sevigny et al., Biochem. 312:351, 1995).

SolCD39 (2 nmol) was combined with 2 ml labeling buffer (100 mM Hepes, pH 7.4, 200 mM NaCI, 4% dimethylformamide [vol/vol]), 400 pl 5 mM FSBA (Sigma Chemical Co.) dissolved in ethanol, and 1.52 ml water. A mock-treated sample was also prepared in which the FSBA solution was replaced with water. After incubating at 37* for 90 min., the samples were centrifuged in a filter unit (Amicon Corp.) for 1 hour at 5,500 rpm and buffer exchanged into PBS to remove unreacted material. The effect of FSBA-treated solCD39 on platelet reactivity is shown in Figure 9.

Induction of platelet activation by ADP (Fig. 9A) or collagen (Fig. 9B) was significantly inhibited by either purified solCD39 or mock FSBA-treated solCD39. In contrast, incubation with FSBA-treated solCD39 did not have a significant effect on platelet activation. A comparative titration of mock-treated solCD39 verses FSBA-treated solCD39 (Fig. 9C) indicated that 22.0 gg/ml of FSBAtreated sol CD39 gave a similar aggregation profile as 0.88 pg/ml of mock-treated solCD39. This indicated that 96% of the aggregation inhibitory activity of solCD39 was lost after FSBA derivitization. Analyses of residual ADPase activity of FSBA- treated solCD39 by the radio-TLC assay system demonstrated that approx. 94% of the enzymatic activity was blocked, while the phosphate release assay indicated that a similar percentage of the ATPase activity was lost as well.

C. Mutagenesis Studies To identify amino acids involved in the biological activity of solCD39, site directed mutagenesis was used to alter selected amino acid residues in CD39. Mutants were assayed for enzymatic (ATPase and ADPase) and platelet inhibitory (dose-dependent inhibition of platelet aggregation) activities. For one series of mutants, residues within the conserved apyrase regions were replaced with alanine.

Platelet inhibitory activity correlated generally with enzymatic activity. The E174A mutant (residues are numbered as in Figure 1) was completely devoid of enzymatic activity and had no effect on platelet responsiveness; the S218A mutant retained less than 10% of ADPase activity and approx.

of platelet inhibitory activity. Glutamatel74 and Serine218 therefore appear to be important for both the enzymatic and platelet inhibitory activities of CD39.

Additional mutant forms of CD39 are expressed and assayed for enzymatic and platelet inhibitory activity in order to identify mutants with increased or decreased activity as well as mutants that preferentially catalyze the ATPase or ADPase reaction.

WO 00/23459 PCT/US99/22955 EXAMPLE 16 Persistence of solCD39 Following In Vivo Administration in Mice Balb/c mice (6-8 weeks of age; maintained under specific pathogen-free conditions; Jackson Laboratory, Bar Harbor, ME) were intravenously injected with 50 pg recombinant solCD39 in 100 pi sterile saline NaCI). No overt external difficulties were noted in the animals following injection. At various times after injection 10, 30 min, 1, 2, 4, 8, 24 pairs of mice were bled by cardiac puncture and euthanized. Serum was prepared from each blood sample and frozen until assay.

The presence of biologically active solCD39 in serum samples was measured in ATPase and ADPase assays. The data were fit using Deltagraph (Deltapoint, Monterey, CA). The best fit was derived using double exponential decay. Where indicated, specificity of enzyme activity was determined by incubating serum samples with anti-CD39 mAb-coated beads to remove CD39 prior to testing for ATPase activity.

As shown in Fig. 10, the data obtained best fit a biphasic exponential curve. The amount of ATPase activity from 25 pg/ml of solCD39 placed in murine serum is presented for comparison. The t 1 /a (distribution phase) was calculated to be 59 min in the ATPase assay and 43 min in the ADPase assay. Approximately 55-65% of apyrase activity was cleared from the circulation during this phase.

The elimination phase had a tP/ 2 1 of approximately 40 h in both assays. Preclearing the 10 min, 2h, and 24 h time point samples with anti-CD39 mAb-coated beads completely eliminated serum ATPase/ADPase activities. These data also demonstrate that the assays specifically detect recombinant human solCD39.

EXAMPLE 17 Pilot Dose Ranging Study in Yorkshire-Hampshire Pigs SolCD39 was administered to Yorkshire-Hampshire pigs, which have been developed as a porcine model of thrombosis. Following intravenous injection, CD39 persisted in the circulation and was capable of inhibiting platelet aggregation and recruitment for as much as a week following injection. This is in marked contrast to many other therapeutic agents used for platelet inhibition, wherein the duration of inhibition is very short.

Ten pigs were randomly assigned to receive solCD39 in low (72 pg/kg), medium (221 pg/kg), or high (670 pg/kg) doses. Aspirin was administered orally on a daily basis. Placebo controls consisted of aspirin. Saline controls and solCD39 were administered as a single bolus. Time points were measured following this administration. Blood samples were obtained via an external jugular vein catheter. Bleeding times were measured in pigs receiving placebo controls and in those receiving solCD39 at baseline and at 60 minutes. ADP-induced platelet aggregation was measured at specific time intervals following administration. The concentration of CD39 in serum as a function of time was measured using an ELISA assay.

Administration of solCD39 was well tolerated. It did not induce anemia or thrombocytopenia and, importantly, a second dose of solCD39 could be administered without observable ill effects, such WO 00/23459 PCT/US99/22955 as hypotension, thrombocytopenia, or hemorrhage. Clot retraction was normal following all experiments, indicating that platelet function was essentially normal.

A. Effect of solCD39 on Bleeding Time Bleeding time is an absolute measure of platelet function. As shown in Figure 11, solCD39 induced a prolonged bleeding time. This indicated that a therapeutic effect had been obtained via a mild interference with platelet function. These mild increases in bleeding time were similar to those obtained by aspirin administration. This indicates that the hemorrhagic defect was mild.

B. Effects of Aspirin and solCD39 on Platelet Aggregation Figure 12A shows the effect of aspirin on platelet aggregation at baseline and at day 5, and Figure 12B shows the effect of high dose solCD39 on platelet aggregation at baseline and at day 7.

Peak heights from the platelet aggregation curves for each of the three solCD39 doses are plotted in Fig. 13. The platelet aggregation data are also compared by plotting relative areas from the platelet aggregation curves for each of the three solCD39 doses. A dose of 670 pg/kg inhibited greater than of ADP induced platelet aggregation. The inhibitory effect was long-lived, with 30% inhibition (after high dose solCD39) at two weeks. These experiments show that solCD39 has potent and long lasting anti-platelet effects, and that these effects are superior to those obtained using aspirin.

C. Persistence of solCD39 in Serum The persistence of solCD39 in porcine serum, as determined by ELISA, is shown in Fig. 14.

Distribution and clearance half-lives were determined using a biphasic curve fit. The t, 2 a (distribution phase) was calculated to be 29 minutes. The elimination phase had a tt 1 3 of approximately 51 hours.

SolCD39 biological activity (ADPase activity) also exhibited a long elimination half-life, approaching 5-7 days, and could still be detected over two weeks after administration. During this time there were no changes in hematologic parameters and no evidence of hemorrhage despite tripling of the bleeding time.

D. Percutaneous Transluminal Coronary Angioplasty (PTCA) Study Porcine platelets and fibrinogen were labeled with "'Indium and 1 25 Iodine respectively for infusion into pigs. Twelve pigs were sedated and anesthetized, and randomly assigned to receive intravenous solCD39 (670 pg/kg) plus heparin and ASA or intravenous saline placebo plus heparin and ASA. Oral ASA was given to all pigs for at least three day prior to the coronary angioplasty procedure, and heparin (100 U/kg) was given at the time of the angioplasty. One to three days prior to the angioplasty an external jugular line was inserted to administer the labeled platelets and fibrin, CD39 or saline, and to facilitate blood draws. Labeled platelets and fibrinogen were given approximately 18 hours prior to balloon injury. Coronary arteries were injured using an oversize WO 00/23459 PCT/US99/22955 balloon. A coronary guide catheter was first advanced into the ascending aorta. An oversized balloon was then advanced into a coronary vessel and inflated at 6 to 8 atmospheres for a total of thirty seconds. The balloon was then deflated and withdrawn. The average ratio of balloon size to vessel size was 1.32 for the placebo group and 1.29 for the CD39 group.

Platelet aggregation and bleeding time were measured 30 minutes after administration. The pigs were killed 24 hours after balloon injury and solCD39 administration, and the labeled platelet 'Indium) and fibrin 2 5 lodine) deposition per cm 2 was measured in the injured coronary artery segments. The results are summarized in TABLE 7. CD39 administration was well tolerated without bleeding or hemodynamic complications. Moreover, no bleeding was noted during PTCA or after sheath removal and there was no significant difference in hematocrit or platelet counts between the groups.

These results show that the administration of solCD39 results in a significant inhibition of platelet aggregation and prolongation of bleeding time, as well as a trend toward inhibition of platelet and fibrin deposition, after balloon injury in animals. The results also suggest that CD 39 has a minimal risk of inducing bleeding.

TABLE 7 Effects of solCD39 After Balloon Injury Platelet Fibrin Deposition Inhibition of Treatment Deposition Ratio Ratio Bleeding Platelet Time Aggregation Placebo 1.78 0.4 0.71 0.14 3.03 0.2 1 solCD39 1.25 0.19 0.62 0.10 7.00 0.81 80 2 p=value 0.2 0.5 0.009 0.001 After the radioactivity decayed, toluidine-blue stained injured coronary artery segments were examined histologically, in order to further characterize the extent of thrombus formation. A blinded observer qualitatively evaluated the degree of histologic injury in the coronary segments by assessing, on a scale of 1-4 with 4 being the most severe injury, the severity of medial and internal elastic lamina tear, medial separation, and hemorrhage. A composite injury score was obtained by totaling the three individual scores. The medial injury scores for the placebo and CD39 groups were 2.5 and 2.2 respectively; medial separation scores for the placebo and CD39 groups were 2.0 and 1.6 respectively; the degree of hemorrhage for the placebo and CD39 groups were 2.3 and 2.5 respectively. The composite injury scores for the placebo and CD39 groups were 6.6 and 6.2 respectively. These in vivo results correlate well with results, reported herein, obtained in vitro and ex vivo.

EXAMPLE 18 Soluble CD39 Provides Additive Inhibition of Platelet Aggregation Over Aspirin and Abciximab An ex vivo study was performed in order to evaluate the additive inhibition of platelet aggregation when soluble CD39 is added to platelet rich plasma from patients receiving: placebo, WO 00/23459 PCT/US99/22955 aspirin, clopidogrel, ticlopidine, or abciximab. Each group consisted of three to six patients. The clopidogrel, ticlopidine, and abciximab groups also received aspirin. Baseline platelet aggregation was measured for each group, in response to the platelet agonists ADP, collagen, or the Thrombin Receptor Activating Peptide TRAPI.

6 SolCD39 (10 tg/ml or 100 pg/ml) was then added and the additional inhibition of platelet activation (over baseline, in response to the platelet agonists) was measured in each of the five groups. The result are shown in TABLE 8.

TABLE 8 Additive Inhibition of Platelet Aggregation by Soluble CD39 Group Placebo Aspirin Clopidogrel Ticlopidine Abciximab Baseline ADP 84 4 69 5 58 6 76 3 0 0 Collagen 85 1 62 8 57 9 71 17 0 0 TRAP 94 ±2 66 ±6 51 2 26 5 46 6 SolCD39 gig/ml ADP 0±0 4±2 5±2 10+2 0±0 100% 2 97% 92% 86% 100% Collagen 75 2 21 ±4 31 ±9 48 18 0 0 11% 68% 48% 37% 100% TRAP 70 3 38 ±7 35 1 19 ±6 22 26% 45% 31% 33% 52% SolCD39 100 tg/ml ADP 0+0 1±0 0±0 2±2 0±0 100% 99% 100% 97% 100% Collagen 57 5 16 ±4 21 6 37 15 0 0 33% 75% 64% 53% 100% TRAP 65 4 26 5 23 3 18 ±7 22 9 63% 55% 36% 52% Platelet aggregation, arbitrary units Percent inhibition relative to same agonist in the absence of CD39 Soluble CD39 at a concentration of 10 pg/ml synergistically inhibited ADP, collagen, and TRAP mediated platelet aggregation in patients on aspirin (p<0.001), and this effect was independent of clopidogrel and ticlopidine. Abciximab alone abolished platelet aggregation due to ADP and collagen, but CD39 provided synergistic inhibition of platelet aggregation induced by TRAP (p<0.007). Soluble CD39 at 100 g.g/ml provided increased inhibition of platelet aggregation to all agonists. These results were also seen in vitro. Collagen and TRAP induce platelet aggregation via mechanisms in addition to ADP release and recruitment, so the ability of CD39 to inhibit collagen and TRAP-mediated platelet aggregation suggests additional versatility of CD39 as an antithrombotic agent.

WO 00/23459 PCT/US99/22955 EXAMPLE 19 Soluble CD39 Inhibits Thrombosis and Limits Ischemic Cerebral Injury in Wild Type and Reconstituted CD39 Null Mice The above examples suggested that soluble CD39 would inhibit ADP-mediated amplification of platelet recruitment in distal microvessels, thereby reducing thrombosis after stroke. The following experiments illustrate the use of CD39 in a microvascular thrombosis (murine ischemic stroke) model.

Soluble CD39 inhibited microvascular thrombosis and conferred cerebroprotection in stroke. A notable feature of the solCD39 treatment was the low incidence of intracerebral hemorrhage relative to that reported for other antithrombotic agents.

A. Materials and Methods C57BL/6J mice (6-8 wk) were obtained from Jackson Laboratories (Bar Harbor, ME).

Untreated mice, and mice treated with 4 mg/kg solCD39, with 5 mg/kg aspirin or phosphate buffered saline, were anesthetized and heparinized (10 prior to blood collection via cardiac puncture. gL of 3.8% trisodium citrate was added to each mL of blood. Samples from 6-8 mice were pooled and platelet-rich plasma (PRP) was prepared by centrifugation. The PRP contained 400-700 x 10 3 platelets per IL. All experiments were completed within 2 hours of blood collection. PRP (200 gtL) was preincubated, for 3 min. at 37 0 C, with 100 |gL Tris-buffered saline buffer (15 mM NaCI, 5 mM glucose, pH 7.4) in an aggregometer cuvette (Lumiaggregometer; Chrono-Log, Havertown, PA), and the platelet agonists ADP, collagen, or sodium arachidonate were added at the final concentrations indicated. Aggregation responses were recorded for 2-4 min, and expressed as area under the curve (height times width at 1/2 height).

The effects of soluble CD39 were tested in a previously validated murine model of stroke injury (Choudhri, et al., J. Clin. Invest. 102:1301-1310 (1998); Connolly, Jr., et al., J. Clin.

Invest. 97:209-216 (1996); and Connolly, Jr., et al., Neurosurg. 38(3):523-532 (1996)).

Anesthetized mice were maintained at 37 2 0 C during and for 90 min following surgery. A midline neck incision was made and the right carotid artery exposed. Middle cerebral artery occlusion was accomplished by advancing a 13-mm heat-blunt tipped 6-0 nylon suture via an arteriotomy in the external carotid stump. The external carotid artery was cauterized to secure hemostasis, and arterial flow re-established. Carotid artery occlusion never exceed 3 min. The occluding suture was removed after 45 min and cautery was again locally applied to prevent bleeding at the arteriotomy site. Surgical staples were used for wound closure.

Doppler measurement of cerebral cortical blood flow, neurological score (Huang, et al., Science 265:183-1885 (1994)), calculation of infarct volume, measurement of cerebral thrombosis using 1 1In-labeled platelets (Choudhri, et al., J. Clin. Invest. 102:1301-1310 (1998) and Naka, et al., Circ. Res. 76:900-906 (1995)), detection of intracerebral fibrin (Choudhri, et al., J.

Clin. Invest. 102:1301-1310 (1998)), and measurement of intracerebral hemorrhage (Choudhri, et WO 00/23459 PCT/US99/22955 al., J. Clin. Invest. 102:1301-1310 (1998) and Choudhri, et al., Stroke 28:2296-2302 (1997)) were measured as previously described. The results are described below.

B. Soluble CD39 Abrogates the Ex Vivo Aggregation of Murine Platelets Platelet-rich plasma was obtained from mice 1 hour after injection of saline (vehicle), soluble CD39, or aspirin. Ex vivo platelet aggregation was studied to ascertain the relative potency of solCD39 as compared to aspirin (which can improve the outcome following a transient ischemic attack). Platelets from control and aspirin-treated mice strongly aggregated following stimulation with ADP (Fig. 15A) or collagen (Fig. Soluble CD39 abrogated platelet aggregation in the presence of ADP, and attenuated aggregation in the presence of collagen and arachidonate. In contrast, aspirin treatment only blocked platelet reactivity to arachidonate (Fig. 15C). The platelets from mice pretreated with solCD39 showed an initial aggregation in the presence of arachidonate, but rapidly disaggregated and returned to the resting state before a full response occurred (Fig. C. Soluble CD39 Is Effective Even When Added at the Peak of the Aggregation Response ADP (5 pM) was added to mouse platelets in vitro to induce an aggregation response.

Soluble CD39 (2.5 pg/ml or 1.25 pg/ml) was added at the peak of the aggregation response. The solCD39 immediately reversed the aggregation response, as shown in Fig. 16. This result demonstrates that SolCD39 is able to reverse an aggregation response, rapidly returning platelets to a resting state, even when added at the peak of the response. This result likely reflects the fact that at the peak of the aggregation response ADP is prominent in the releasate from the aggregating platelets.

Soluble CD39 metabolizes this ADP to the biologically inactive compound AMP almost instantaneously, accounting for the rapid descent of the aggregation curve in Fig. 16, right side.

D. Soluble CD39 Reduces the Sequelae of Stroke Intravenously injected soluble CD39 showed therapeutic utility in stroke. Soluble CD39 inhibited platelet accumulation in the ipsilateral cerebral hemisphere following induction of stroke, as shown in Fig. 17A. Similarly, solCD39 decreased the level of fibrin accumulation in the ipsilateral hemisphere (vs. contralateral) as measured by Western blot analysis using a fibrin specific antibody (Fig. 17B).

The ability of solCD39 to reduce thrombosis, as measured by decreased platelet and fibrin deposition, was accompanied by improved postischemic cerebral perfusion 24 hours after stroke induction, as shown in Fig. 18A. In contrast, when aspirin was administered at a clinically relevant dose (that inhibited the ex vivo response of platelets to arachidonate) no improvement was seen in postischemic cerebral blood flow (Fig. 18A).

Preoperatively administered solCD39 conferred a dose-dependent diminution of cerebral infarct volume, as measured by digital histological analysis (Fig. I 8B). Aspirin, in contrast, showed a WO 00/23459 PCT/US99/22955 tendency to decrease cerebral infarct volume, although this effect was not statistically significant. The administration of soluble CD39 either prior to, or up to 3 h following, stroke reduced both neurological deficit (Fig. 18C) and mortality (Fig. 18D).

The effects of soluble CD39 and aspirin on the development of intracerebral hemorrhage following stroke are shown in Fig. 18E. Aspirin increased intracerebral hemorrhage (as measured spectrophotometrically) significantly, but there was no significant increase in intracerebral hemorrhage at any dose of soluble CD39 tested. At these doses, soluble CD39 inhibited both platelet and fibrin accumulation and promoted an increase in postischemic blood flow, as shown in Figs. 17A, 17B, and 18A. Figure 19 shows a covariate plot of cerebral infarct volume vs. intracerebral hemorrhage for each treatment, and indicates that aspirin is less capable of reducing infarct volume and preventing intracerebral hemorrhage than soluble CD39. In summary, at the doses tested in the mouse stroke model, solCD39 conferred protection without inducing the bleeding problems that often accompany anti-thrombotic therapy regimens.

E. CD39 Null Mice Can be Reconstituted with Soluble CD39 CD39-/- mice were generated using a gene targeting vector in which exons 4-6, encoding apyrase conserved regions 2-4 (Handa, M. Guidotti, Biochem. Biophys. Res. Commun. 218:916- 923 (1996); Wang, T.F. Guidotti, J. Biol. Chem. 271:9898-9901 (1996); Maliszewski, et al., J. Immunol. 153:3574-3583 (1994); and Schoenborn, et al., Cytogen Cell Gen. 81(3-4):287- 280 (1998)), were replaced with a PGKneo cassette, as shown in Fig. 20A. The gene targeting vector, in which a 4.1 kb Spel-BglI fragment containing exons 4-6 was replaced with a PGKneo cassette, was introduced into 129-derived ES cells. Cells were selected in G418 and gancyclovir. Nine ES clones with a disrupted CD39 allele, as identified by genomic Southern blot analyses of Bglll digested DNA as shown in Fig. 20B, were injected into blastocysts and the resulting chimeras crossed to C57BU6 to produce CD39 heterozygotes. CD39-/- mice were generated at the expected Mendelian frequency from CD39 intercrosses. The CD39-/- mice used in the experiments described below represent random C57BU6 x 129 hybrids.

Homozygous CD39-/- mice were overtly normal, and did not display an obvious phenotype in the unperturbed state. Hematological profiles, including erythrocyte parameters, platelet counts, leukocyte counts and differentials, and coagulation screening, were normal. As shown below, the CD39-null mice did not exhibit a prothrombotic phenotype unless challenged by experimental stroke.

Under those conditions, the defect was abolished and normal blood fluidity was restored by administration of soluble CD39. Bleeding times of CD39-/- mice were normal, indicating that normal blood flow in an unperturbed animal is not dependent upon endogenous expression of CD39. As is seen in normal mice, CD39-/- animals exhibited markedly increased bleeding times following the administration of aspirin or following administration of increasing doses of solCD39 as shown in Fig.

21. CD39-/- mice subjected to focal cerebral ischemia exhibited diminished blood flow following reperfusion as compared to genetically matched controls (Fig. 22A), indicating that endogenous CD39 WO 00/23459 PCT/US99/22955 contributes to maintenance of hemostasis during episodes of vascular injury. When solCD39 (8 mg/kg) was administered to the CD39-/- mice, these mice were "reconstituted" as shown by a postischemic blood flow similar to untreated controls. CD39-/- mice demonstrated increased cerebral infarction volume as compared to genotype-matched controls following induced stroke (Fig. 22B).

CD39-/- mice "reconstituted" with solCD39 had markedly diminished infarct volume, indicating a protective effect of solCD39. Other parameters (neurological deficit scores, overall mortality, and intracerebral hemorrhage) did not differ between groups (Fig. 22 C, D, E).

These results demonstrate that CD39 inhibits microvascular thrombosis and confers cerebroprotection without inducing intracerebral hemorrhage in a murine model of stroke. Soluble CD39 decreased platelet deposition, fibrin deposition, and cerebral infarction volume. Soluble CD39 reduced infarction volume and restored postischemic blood flow even when administered three hours following stroke induction. This result is important because the average patient experiencing a stroke appears in the emergency room approximately three hours after the initial event occurs. The ability to treat patients with solCD39 after three hours provides an important advantage over many other agents designed to inhibit platelet reactivity.

EXAMPLE Soluble CD39 Improves Survival in a Mouse Ischemia Model C57BU6 mice were anesthetized and ventilated, and their thoraces were opened to surgically expose both pulmonary hila. Either physiological saline or soluble CD39 (8 mg/kg) was administered intravenously, after which the left pulmonary hilum was cross-clamped for one hour. The cross-clamp was removed for three hours of reperfusion, and then a cross-clamp was applied to the right hilum for a thirty minute observation period. This latter maneuver effectively removed the normal lung from circulation, so that the mouse must survive on the function of the post-ischemic left lung. The results are shown, in the form of a Kaplan-Meier survival plot, in Figure 23. All of the mice given saline died prior to the thirty minute time point whereas all of the sol39-treated mice survived for thirty minutes.

The long lasting effects of soluble CD39 are also shown to be clinically useful in the reduction of complications of atherosclerosis, such as myocardial infarction, stroke, and peripheral vascular occlusion. Patients suffering from these conditions demonstrate an abundance of activated platelets in their circulation, and such activated platelets have a lowered threshold for ADP stimulation. Soluble CD39 metabolically deletes ADP from the fluid phase of activated platelets and reverses their prothrombotic characteristics.

The relevant disclosures of publications cited herein are specifically incorporated by reference. The examples presented above are not intended to be exhaustive or to limit the scope of the invention. The skilled artisan will understand that variations and modifications and variations are possible in light of the above teachings, and such modifications and variations are intended to be within the scope of the invention.

Throughout the description and claims of this specification, the word "comprise" and variations of the word, such as "comprising" and "comprises", is not intended to exclude other additives, components, integers or steps.

The discussion of documents, acts, materials, devices, articles and the like is included in this specification solely for the purpose of providing a context for the present invention. It is not suggested or represented that any or all of these matters formed part of the prior art base or were common general knowledge in the field relevant to the present invention as it existed in Australia before the priority date of each claim of this application.

*oooo *5* ooo IC W:\ilonaSharon\SJJspeciksp64 53a EDITORIAL NOTE APPLICATION NUMBER 64115/99 The following Sequence Listing pages 1 to 34 are part of the description. The claims pages follow on pages 54 to 57.

WO 00/23459 PCT/US99/22955 SEQUENCE LISTING <110> Maliszewski, Charles R.

Gayle III, Richard B.

Price, Virginia L.

Gimpel, Steven D.

Immunex Corporation <120> Inhibitors of Platelet Activation and Recruitment <130> 2879-WO <140> <141> <150> US 60/104,585 <151> 1998-10-16 <150> US 60/107,466 <151> 1998-11-06 <150> US 60/149,010 <151> 1999-08-13 <160> 31 <170> PatentIn Ver. <210> <211> <212> <213> <220> <221> <222> 1 1599

DNA

Homo sapiens

CDS

(67)..(1596) <400> 1 ccacaccaag cagcggctgg gggggggaaa gacgaggaaa gaggaggaaa acaaaagctg ctactt atg gaa gat aca aag gag tct aac gtg aag aca ttt tgc tcc Met Glu Asp Thr Lys Glu Ser Asn Val Lys Thr Phe Cys Ser 1 5 aag aat atc cta gcc atc ctt ggc ttc tcc tct atc ata gct gtg ata Lys Asn Ile Leu Ala Ile Leu Gly Phe Ser Ser Ile Ile Ala Val Ile 20 25 get ttg ctt gct gtg ggg ttg acc cag aac aaa gca ttg cca gaa aac Ala Leu Leu Ala Val Gly Leu Thr Gin Asn Lys Ala Leu Pro Glu Asn 40 gtt aag tat ggg att gtg ctg gat gcg ggt tct tct cac aca agt tta Val Lys Tyr Gly Ile Val Leu Asp Ala Gly Ser Ser His Thr Ser Leu 55 tac ate tat aag tgg cca gca gaa aag gag aat gac aca ggc gtg gtg 108 156 204 252 300 WO 00/23459 Tyr Ile Tyr PCT/US99/22955 Lys Trp Pro Ala Glu 70 Lys Giu Asn Asp Thr Gly Vai Vai atc tca aaa ttt Ile Ser Lys Phe cat caa His Gin gta gaa gaa tgc Val Giu Giu Cys agg Arg 85 gtt aaa ggt cct Val Lys Giy Pro gga Giy gtt Vai cag aaa gta aat Gin Lys Val Asn gaa Glu 100 ata ggc att tac Ile Gly Ile Tyr ctg Leu 105 act gat tgc atg Thr Asp Cys Met gaa Glu 110 348 396 444 aga gct agg gaa Arg Ala Arg Giu gtg Val1 115 att cca agg tcc Ile Pro Arg Ser cag Gin 120 cac caa gag aca His Gin Glu Thr ccc gtt Pro Val 125 tac ctg gga Tyr Leu Gly gag ttg gca Giu Leu Ala 145 gcc Ala 130 acg gca ggc atg Thr Ala Gly Met cgg Arg 135 ttg ctc agg atg Leu Leu Arg Met gaa agt gaa Giu Ser Glu 140 ctc agc aac Leu Ser Asn 492 540 gac agg gtt ctg Asp Arg Val Leu gat Asp 150 gtg gtg gag agg Val Val Giu Arg agc Ser 155 tac ccc Tyr Pro 160 ttt gac ttc cag Phe Asp Phe Gin gg t Gly 165 gcc agg atc att Ala Arg Ile Ile act Thr 170 ggc caa gag gaa Giy Gin Giu Giu gg t Gly 175 gcc tat ggc tgg Ala Tyr Gly Trp att Ile 180 act atc aac tat Thr Ile Asn Tyr ctg Leu 185 ctg ggc aaa ttc Leu Giy Lys Phe agt Ser 190 588 636 684 cag aaa aca agg Gin Lys Thr Arg tgg Trp 195 ttc agc ata gtc Phe Ser Ile Val cca Pro 200 tat gaa acc aat Tyr Giu Thr Asn aat cag Asn Gin 205 gaa acc ttt Giu Thr Phe ttt gta ccc Phe Val Pro 225 gga Gly 210 gct ttg gac ctt Ala Leu Asp Leu ggg Gly 215 gga gcc tct aca Giy Ala Ser Thr caa gtc act Gin Val Thr 220 gct. ctg caa Ala Leu Gin 732 780 caa aac cag act Gin Asn Gin Thr atc Ile 230 gag tcc cca gat Giu Ser Pro Asp aat Asn 235 ttt cgc Phe Arg 240 ctc tat ggc aag Leu Tyr Giy Lys gac Asp 245 tac aat gtc tac Tyr Asn Val Tyr aca Thr 250 cat agc ttc ttg His Ser Phe Leu tgc Cys 255 tat ggg aag gat Tyr Giy Lys Asp cag Gin 260 gca ctc tgg cag Ala Leu Trp Gin aaa Lys 265 ctg gcc aag gac Leu Ala Lys Asp att.

Ile 270 828 876 924 cag gtt gca agt Gin Val Ala Ser aat Asn 275 gaa att. ctc agg Glu Ile Leu Arg gac Asp 280 cca tgc ttt cat Pro Cys Phe His cct gga Pro Gly 285 tat aag aag gta gtg aac gta agt gac ctt tac aag acc Tyr Lys Lys Val Val Asn Val Ser Asp Leu Tyr Lys Thr ccc tgc acc Pro Cys Thr 300 972 290 295 WO 00/23459 aag aga ttt Lys Arg Phe 305 PCTIUS99/22955 gag atg act ctt Glu Met Thr Leu cca Pro 310 tte cag cag ttt Phe Gin Gin Phe gaa Glu 315 ate eag ggt Ile Gin Gly 1020 att gga Ile Gly 320 aac tat caa caa Asn Tyr Gin Gin tgc Cys 325 cat caa age at~e His Gin Ser Ile etg Leu 330 gag etc tte aae Glu Leu Phe Asn ace Thr 335 eca Pro agt tae tgc cct Ser Tyr Cys Pro cca etc eag ggg Pro Leu Gin Gly 355 tac Tyr 340 tee eag tgt gec Ser Gin Cys Ala tte Phe 345 aat ggg att tte Asn Gly Ile Phe 1068 1116 1164 gat ttt ggg gca Asp Phe Gly Ala ttt Phe 360 tea get ttt tae Ser Ala Phe Tyr ttt gtg Phe Val 365 atg aag ttt Met Lys Phe act gag atg Thr Giu Met 385 tta Leu 370 aac ttg aca tea Asn Leu Thr Ser gag Glu 375 aaa gte tet cag Lys Val Ser Gin gaa aag gtg Glu Lys Val 380 gag ata aaa Glu Ile Lys 1212 1260 atg aaa aag tte Met Lys Lys Phe tgt Cys 390 get cag ect tgg Ala Gin Pro Trp gag Glu 395 aca tct Thr Ser 400 tac get gga gta Tyr Ala Gly Val aag Lys 405 gag aag tac ctg Glu Lys Tyr Leu agt Ser 410 gaa tac tgc ttt Glu Tyr Cys Phe tct Ser 415 ggt ace tac att Gly Thr Tyr Ile etc Leu 420 tee etc ett etg Ser Leu Leu Leu eaa Gin 425 ggc tat cat tte Gly Tyr His Phe aca Thr 430 1308 1356 1404 get gat tee tgg Ala Asp Ser Trp gag Glu 435 eac ate cat tte His Ile His Phe att Ile 440 gge aag ate eag Gly Lys Ile Gin gge age Gly Ser 445 gac gee ggc Asp Ala Gly eca get gag Pro Aia Glu 465 tgg Trp 450 act ttg gge tac Thr Leu Gly Tyr atg Met 455 etg aac ctg ace Leu Asn Leu Thr aac atg ate Asn Met Ile 460 ace tat gte Thr Tyr Val 1452 1500 caa eca ttg tee Gin Pro Leu Ser aca Thr 470 cct etc tee cac Pro Leu Ser His tee Ser 475 tte etc Phe Leu 480 gge ttg Gly Leu 495 tag atg gtt eta tte Met Val Leu Phe ett ate ttt cac Leu Ile Phe His 500 tee Ser 485 etg gte ett tte Leu Val Leu Phe aca Thr 490 gtg gee ate ata Val Ala Ile Ile 1548 1596 aag cet tea tat Lys Pro Ser Tyr tte Phe 505 tgg aaa gat atg Trp Lys Asp Met gta Val 510 1599 <210> 2 <211> 510 WO 00/23459 <212> PRT <213> Homo sapiens <400> 2 PCTIIJS99/22955 Met Glu Asp Thr 1 Ile Leu Tyr Tyr Val1 Lys Arg Gly Al a 145 Phe Tyr Thr Phe Pro 225 Leu Gly Leu Al a Gly Lys Glu Val Glu Ala 130 Asp Asp Gly Arg Gly 210 Gin Tyr Lys Ala Ile Val Gly Ile Val Trp Pro Giu Cys Asn Glu 100 Val Ile 115 Thr Ala Arg Val Phe Gin Trp, Ile 180 Trp Phe 195 Ala Leu Asn Gln Gly Lys Asp Gin 260 Lys 5 Leu Leu Leu Ala Arg Ile Pro Gly Leu Gly 165 Thr Ser Asp Thr Asp 245 Ala Giu Gly Thr Asp Giu 70 Val1 Gly Arg Met Asp 150 Ala Ile Ile Leu Ile 230 Tyr Leu Ser Phe Gln Ala 55 Lys Lys Ile Ser Arg 135 Val Arg Asn Val1 Gly 215 Glu Asn Trp Asn Ser Asn 40 Gly Giu Gly Tyr Gin 120 Leu Val1 Ile Tyr Pro 200 Giy Ser Val Gin Asp 280 Val1 Ser 25 Lys Ser Asn Pro Leu 105 His Leu Giu Ile Leu 185 Tyr Al a Pro Tyr Lys 265 Lys 10 Ile Ala Ser Asp Gly 90 Thr Gin Arg Arg Thr 170 Leu Giu S er Asp Thr 250 Leu Thr Ile Leu His Thr 75 Ile Asp Glu Met Ser 155 Gly Gly Thr Thr Asn 235 His Ala Phe Ala Pro Thr Gly Ser Cys Thr Giu 140 Leu Gin Lys Asn Gin 220 Ala Ser Lys Cys Val Giu Ser Val1 Lys Met Pro 125 Ser Ser Glu Phe Asn 205 Val Leu Phe Asp Pro 285 Ser Lys Ile Ala Asn Val Leu Tyr Vai His Phe Val Giu Arg 110 Vai Tyr Giu Glu Asn Tyr Giu Gly 175 Ser Gin 190 Gin Giu Thr Phe Gin Phe Leu Cys 255 Ile Gin 270 Asn Leu Lys Ile Gin Gin Ala Leu Leu Pro 160 Ala Lys Thr Vai Arg 240 Tyr Val Ala Ser Asn Giu Ile Leu Arg 275 Pro Cys Phe His Giy Tyr Lys WO 00/23459 Lys Val Val 290 Phe Glu Met 305 Asn Tyr Gin Tyr Cys Pro Leu Gin Gly 355 Phe Leu Asn 370 Met Met Lys 385 Tyr Ala Gly Thr Tyr Ile Ser Trp Glu 435 Gly Trp Thr 450 Glu Gin Pro 465 Met Val Leu Leu Ile Phe Asn Thr Gin Tyr 340 Asp Leu Lys Val Leu 420 His Leu Leu Phe Val Leu Cys 325 Ser Phe Thr Phe Lys 405 Ser Ile Gly Ser Ser 485 Ser Pro 310 His Gin Gly Ser Cys 390 Glu Leu His Tyr Thr 470 Leu Asp 295 Phe Gin Cys Ala Glu 375 Ala Lys Leu Phe Met 455 Pro Val Leu Tyr Gin Gin Ser Ile Ala Phe 345 Phe Ser 360 Lys Val Gin Pro Tyr Leu Leu Gin 425 Ile Gly 440 Leu Asn Leu Ser Leu Phe Lys Thr Phe Glu 315 Leu Glu 330 Asn Gly Ala Phe Ser Gin Trp Glu 395 Ser Glu 410 Gly Tyr Lys Ile Leu Thr His Ser 475 Thr Val 490 Pro Cys 300 Ile Gin Leu Phe Ile Phe Tyr Phe 365 Glu Lys 380 Glu Ile Tyr Cys His Phe Gin Gly 445 Asn Met 460 Thr Tyr Ala Ile Thr Gly Asn Leu 350 Val Val Lys Phe Thr 430 Ser Ile Val Ile PCT/US99/22955 Lys Arg Ile Gly 320 Thr Ser 335 Pro Pro Met Lys Thr Glu Thr Ser 400 Ser Gly 415 Ala Asp Asp Ala Pro Ala Phe Leu 480 Gly Leu 495 His Lys Pro Ser Tyr Phe Trp Lys Asp Met Val 500 505 510 <210> 3 <211> 476 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <400> 3 Met Ala Thr Ser Trp Gly Thr Val Phe Phe Met Leu Val Val Ser Cys 1 5 10 Val Cys Ser Ala Val Ser His Arg Asn Gin Gin Thr Trp Phe Glu Gly 25 WO 00/23459 WO 0023459PCTIUS99/22955 Ile Tyr Tyr Val Lys Arg Gly Ala 145 Phe Tyr Thr Phe Pro 225 Leu Gly Ala Lys Phe 305 Phe Gly Lys Giu Val Glu Ala 130 Asp Asp Gly Arg Gly 210 Gin Tyr Lys Ser Val 290 Giu Leu Ile Trp Giu Asn Vai 115 Thr Arg Phe Trp Trp, 195 Ala Asn Gly Asp Asn 275 Val Met Ser Ser Thr Gin Asn Lys Ala Leu Pro Giu Asn Vai Lys 40 Val Pro Cys Giu 100 Ile Al a Val1 Gin Ile 180 Phe Leu Gin Lys Gin 260 Giu Asn Thr Leu Al a Arg Ile Pro Gly Leu Gly 165 Thr Ser Asp Thr Asp 245 Ala Ile Val Leu Cys 325 Asp Giu 70 Val Gly Arg Met Asp 150 Ala Ile Ile Leu Ile 230 Tyr Leu Leu Ser Pro 310 Ala 55 Lys Lys Ile Ser Arg 135 Val1 Arg Asn Val Gly 215 Giu Asn Trp Arg Asp 295 Phe Gly Giu Gly Tyr Gin 120 Leu Val1 Ile Tyr Pro 200 Gly Ser Val Gin Asp 280 Leu Gin Ser Asn Pro Leu 105 His Leu Giu Ile Leu 185 Tyr Al a Pro Tyr Lys 265 Pro Tyr Gin Ser Asp Gly 90 Thr Gin Arg Arg Thr 170 Leu Giu Ser Asp Thr 250 Leu Cys Lys Phe His Thr 75 Ile Asp Giu Met Ser 155 Gly Gly Thr Thr Asn 235 His Ala Phe Thr Giu 315 Thr Gly Ser Cys Thr G lu 140 Leu Gin Lys Asn Gin 220 Ala Ser LJys His Pro 300 Ile Ser Val Lys Met Pro 125 S er Ser Giu Phe Asn 205 Val Leu Phe Asp Pro 285 Cys Gin Leu Vai Phe Glu 110 Vai Giu Asn Giu Ser 190 Gin Thr Gin Leu Ile 270 Giy Thr Gly Tyr His Val1 Arg Tyr Glu Tyr Gly 175 Gin Giu Phe Phe Cys 255 Gin Tyr Lys Ile Ile Gin Gin Ala Leu Leu Pro 160 Ala Lys Thr Val Arg 240 Tyr Val Lys Arg Gly 320 Asn Tyr Gin Gin His Gin Ser Ile Leu 330 Giu Leu Phe Asn Thr Ser 335 WO 00/23459 Tyr Cys Pro Tyr Ser Gin Cys Ala Phe Asn Gly Ile Ph 340 345 Leu Gin Gly Asp Phe Gly Ala Phe Ser Ala Phe Tyr Ph 355 360 36 Phe Leu Asn Leu Thr Ser Glu Lys Val Ser Gin Glu Lj 370 375 380 Met Met Lys Lys Phe Cys Ala Gin Pro Trp Glu Glu I] 385 390 395 Tyr Ala Gly Val Lys Glu Lys Tyr Leu Ser Glu Tyr Cq 405 410 Thr Tyr Ile Leu Ser Leu Leu Leu Gin Gly Tyr His PI 420 425 Ser Trp Glu His Ile His Phe Ile Gly Lys Ile Gin G 435 440 4 Gly Trp Thr Leu Gly Tyr Met Leu Asn Leu Thr Asn M 450 455 460 Glu Gin Pro Leu Ser Thr Pro Leu Ser His Ser Thr 465 470 475 <210> 4 <211> 476 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <220> <221> VARIANT <222> (39) <223> Any amino acid, preferably Cys or Ser <400> 4 Met Ala Thr Ser Trp Gly Thr Val Phe Phe Met Leu V 1 5 10 Val Cys Ser Ala Val Ser His Arg Asn Gin Gin Thr T 25 Ile Phe Leu Ser Ser Met Xaa Pro Ile Asn Val Ser A 40 Tyr Gly Ile Val Leu Asp Ala Gly Ser Ser His Thr S 55 Tyr Lys Trp Pro Ala Glu Lys Glu Asn Asp Thr Gly V 70 75 ie ie Is le ys he Ly et Leu 350 Val Val Lys Phe Thr 430 Ser Ile Val Phe Ser Leu Val PCT/US99/22955 Pro Pro Met Lys Thr Glu Thr Ser 400 Ser Gly 415 Ala Asp Asp Ala Pro Ala al rp la er al Ser Glu Thr Tyr His Cys Gly Leu Ile Gin WO 00/23459 Val Giu Glu Lys Val Asn Arg Giu Val 115 Gly Ala Thr 130 Ala Asp Arg 145 Phe Asp Phe Tyr Gly Trp Thr Arg Trp 195 Phe Gly Ala 210 Pro Gln Asn 225 Leu Tyr Gly Giy Lys Asp Ala Ser Asn 275 Lys Val Val 290 Phe Giu Met 305 Asn Tyr Gin Tyr Cys Pro Leu Gin Gly 355 Cys Glu 100 Ile Al a Val1 Gin Ile 180 Phe Leu Gin Lys Gln 260 Glu Asn Thr Gin Tyr 340 Asp Arg Ile Pro Gly Leu Gly 165 Thr Ser Asp Thr Asp 245 Al a Ile Val Leu Cys 325 Ser Phe Val Gly Arg Met Asp 150 Al a Ile Ile Leu Ile 230 Tyr Leu Leu Ser Pro 310 His Gin Gly Lys Ile Ser Arg 135 Val1 Arg Asn Val Gly 215 Giu Asn Trp Arg Asp 295 Phe Gin Cys Ala Giy Tyr Gin 120 Leu Val Ile Tyr Pro 200 Gly Ser Val Gin Asp 280 Leu Gin Ser Ala Phe 360 Pro Leu 105 His Leu Giu Ile Leu 185 Tyr Al a Pro Tyr Lys 265 Pro Tyr Gin Ile Phe 345 Ser Gly 90 Thr Gin Arg Arg Thr 170 Leu Giu Ser Asp Thr 250 Leu Cys Lys Phe Leu 330 Asn Ala Ile Asp Glu Met Ser 155 Gly Gly Thr Thr Asn 235 His Ala Phe Thr Glu 315 Giu Gly Phe Ser

CYS

T'hr G lu 140 Leu Gin Lys Asn Gin 220 Al a Ser Lys His Pro 300 Ile Leu Ile Tyr Lys Met Pro 125 Ser Ser Giu Phe Asn 205 Val Leu Phe Asp Pro 285 Cys Gin Phe Phe Phe 365 Phe Glu 110 Vai Giu Asn Glu S er 190 Gin Thr Gin Leu Ile 270 Gly Thr Gly Asn Leu 350 Val PCTIUS99/22955 Val Gin Arg Ala Tyr Leu Glu Leu Tyr Pro 160 Gly Ala 175 Gin Lys Giu Thr Phe Val Phe Arg 240 Cys Tyr 255 Gin Val Tyr Lys Lys Arg Ile Gly 320 Thr Ser 335 Pro Pro Met Lys Phe Leu Asn Leu Thr Ser Giu Lys Val Ser Gin Giu Lys Val Thr Giu 370 375 380 WO 00/23459 Met Met Lys Lys Phe Cys Ala Gin Pro Trp Glu Glu I 385 390 395 Tyr Ala Gly Val Lys Glu Lys Tyr Leu Ser Glu Tyr C 405 410 Thr Tyr Ile Leu Ser Leu Leu Leu Gin Gly Tyr His P 420 425 Ser Trp Glu His Ile His Phe Ile Gly Lys Ile Gin G 435 440 4' Gly Trp Thr Leu Gly Tyr Met Leu Asn Leu Thr Asn M( 450 455 460 Glu Gin Pro Leu Ser Thr Pro Leu Ser His Ser Thr 465 470 475 <210> <211> 1365 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <220> <221> CDS <222> (1)..(1362) <400> gca cct act tea agt tct aca aag aaa aca cag cta a Ala Pro Thr Ser Ser Ser Thr Lys Lys Thr Gin Leu T 1 5 10 cag aac aaa gca ttg cca gaa aac gtt aag tat ggg a Gin Asn Lys Ala Leu Pro Glu Asn Val Lys Tyr Gly I 25 gcg ggt tct tct cac aca agt tta tac ate tat aag t Ala Gly Ser Ser His Thr Ser Leu Tyr Ile Tyr Lys T 40 aag gag aat gac aca ggc gtg gtg cat caa gta gaa g Lys Glu Asn Asp Thr Gly Val Val His Gin Val Glu G 55 aaa ggt cct gga ate tea aaa ttt gtt cag aaa gta a Lys Gly Pro Gly Ile Ser Lys Phe Val Gin Lys Val A 70 75 att tac ctg act gat tgc atg gaa aga gct agg gaa g Ile Tyr Leu Thr Asp Cys Met Glu Arg Ala Arg Glu V 90 tec cag cac caa gag aca ccc gtt tac ctg gga gcc a 9 Le ys ie ly et Lys Phe Thr 430 Ser Ile agt Ser gtg Val cca Pro tgc Cys gaa Glu att Ile gca PCT/US99/22955 Thr Ser 400 Ser Gly 415 Ala Asp Asp Ala Pro Ala ct hr tt le gg rp aa lu at .sn tg 'al icg tea Ser ctg Leu gca Ala agg Arg ata Ile cca Pro ggc acc Thr gat Asp gaa Glu gtt Val ggc Gly agg Arg atg 48 96 144 192 240 288 336 WO 00/23459 WO 0023459PCT/US99/22955 Ser cgg Arg gtg Val agg Arg 145 aac Asn gtc Val ggg Gly gag Glu aat Asn 225 tgg Trp agg Arg gac Asp ttc Phe caa Gin 305 Gin His ttg ctc Leu Leu 115 gtg gag Val Glu 130 atc att Ile Ile tat ctg Tyr Leu cca tat Pro Tyr gga gcc Gly Ala 195 tcc cca Ser Pro 210 gtc tac Val Tyr cag aaa Gin Lys gac cca Asp Pro ctt tac Leu Tyr 275 cag cag Gin Gin 290 agc atc Ser Ile Gin Giu Thr Pro Val Tyr Leu Gly Ala Thr 100 105 agg Arg agg Arg act Thr ctg Leu gaa Glu 180 tct Ser gat Asp aca Thr ctg Leu tgc Cys 260 aag Lys ttt Phe c tg Leu atg Met agc Ser ggc Gly ggc Gly 165 ac c Thr aca Thr aa t Asn cat His gc c Ala 245 ttt Phe ac c Thr gaa Giu gag Glu gaa Giu ctc Leu caa Gin 150 aaa Lys aat Asn caa Gin gc t Aia agc Ser 230 aag Lys cat His ccc Pro atc Ile ctc Leu 310 agt Ser agc Ser 135 gag Giu ttc Phe aat Asn gtc Val ctg Leu 215 ttc Phe gac Asp cct Pro tgc Cys cag Gin 295 ttc Phe gaa Giu 120 aac Asn gaa Giu agt Ser cag Gin act Thr 200 caa Gin ttg Leu att Ile gga Gly acc Thr 280 gg t Gly aac Asn gag Giu tac Tyr gg t Giy cag Gin gaa Giu 185 ttt Phe ttt Phe tgC Cys cag Gin tat Tyr 265 aag Lys att.

Ile acc Thr t tg Leu cc Pro gcC Ala aaa Lys 170 acc Thr gta Val cgc Arg tat Tyr gtt Vai 250 aag Lys aga Arg gga Gly agt Ser cca Pro 330 gca Ala ttt Phe tat Tyr 155 aca Thr ttt Phe cc Pro ctc Leu ggg Gly 235 gca Ala aag Lys ttt Phe aac Asn t ac Tyr 315 ctc Leu gac Asp gac Asp 140 ggc Gly agg Arg gga Gly caa Gin tat Tyr 220 aag Lys agt Ser gta Val gag Giu tat Tyr 300 tgc Cys cag Gin agg Arg 125 ttc Phe tgg Trp tgg Trp gct Ala aa c Asn 205 ggc Giy gat Asp aat Asn gtg Vali atg Met 285 caa Gin cct Pro ggg Gly Aia Giy 110 gtt ctg Val Leu cag ggt Gin Giy att act Ile Thr ttc agc Phe Ser 175 ttg gac Leu Asp 190 cag act Gin Thr aag gac Lys Asp cag gca Gin Ala gaa att Giu Ile 255 aac gta Asn Val 270 act ctt Thr Leu caa tgc Gin Cys tac tcc Tyr Ser gat ttt Asp Phe 335 Met ga t Asp gcc Aia atc Ile 160 ata Ile ctt Leu atc Ile t ac Tyr c tc Leu 240 ctc Leu agt Ser cca Pro cat His c ag Gin 320 ggg Gly 384 432 480 528 576 624 672 720 768 816 864 912 960 1008 tgt gcc ttc aat ggg att ttc ttg cca.

Cys Aia Phe Asn Gly Ie Phe Leu Pro 325 WO 00/23459 PCT/US99/22955 gca ttt tca get ttt tac ttt gtg atg aag ttt tta aac ttg aca tca 1056 Ala Phe Ser Ala Phe Tyr Phe Val Met Lys Phe Leu Asn Leu Thr Ser 340 345 350 gag aaa gtc tct cag gaa aag gtg act gag atg atg aaa aag ttc tgt 1104 Glu Lys Val Ser Gin Glu Lys Val Thr Glu Met Met Lys Lys Phe Cys 355 360 365 gct cag cct tgg gag gag ata aaa aca tct tac get gga gta aag gag 1152 Ala Gin Pro Trp Glu Glu Ile Lys Thr Ser Tyr Ala Gly Val Lys Glu 370 375 380 aag tac ctg agt gaa tac tgc ttt tct ggt acc tac att ctc tec ctc 1200 Lys Tyr Leu Ser Glu Tyr Cys Phe Ser Gly Thr Tyr Ile Leu Ser Leu 385 390 395 400 ctt ctg caa ggc tat cat ttc aca get gat tcc tgg gag cac ate cat 1248 Leu Leu Gin Gly Tyr His Phe Thr Ala Asp Ser Trp Glu His Ile His 405 410 415 ttc att ggc aag ate cag ggc agc gac gcc ggc tgg act ttg ggc tac 1296 Phe Ile Gly Lys Ile Gin Gly Ser Asp Ala Gly Trp Thr Leu Gly Tyr 420 425 430 atg ctg aac ctg acc aac atg ate cca get gag caa cca ttg tcc aca 1344 Met Leu Asn Leu Thr Asn Met Ile Pro Ala Glu Gin Pro Leu Ser Thr 435 440 445 cct ctc tcc cac tcc acc taa 1365 Pro Leu Ser His Ser Thr 450 <210> 6 <211> 454 <212> PRT <213> Artificial Sequence <400> 6 Ala Pro Thr Ser Ser Ser Thr Lys Lys Thr Gin Leu Thr Ser Ser Thr 1 5 10 Gin Asn Lys Ala Leu Pro Glu Asn Val Lys Tyr Gly Ile Val Leu Asp 25 Ala Gly Ser Ser His Thr Ser Leu Tyr Ile Tyr Lys Trp Pro Ala Glu 40 Lys Glu Asn Asp Thr Gly Val Val His Gin Val Glu Glu Cys Arg Val 55 Lys Gly Pro Gly Ile Ser Lys Phe Val Gin Lys Val Asn Glu Ile Gly 70 75 Ile Tyr Leu Thr Asp Cys Met Glu Arg Ala Arg Glu Val Ile Pro Arg 90 WO 00/23459 Ser Gin His Arg Leu Leu 115 Val Val Giu 130 Arg Ile Ile 145 Asn Tyr Leu Val Pro Tyr Gly Gly Ala 195 Giu Ser Pro 210 Asn Val Tyr 225 Trp Gin Lys Arg Asp Pro Asp Leu Tyr 275 Phe Gin Gin 290 Gin Ser Ile 305 Cys Ala Phe Ala Phe Ser Giu Lys Val 355 Ala Gin Pro 370 Gin 100 Arg Arg Thr Leu Giu 180 Ser Asp Thr Leu Cys 260 Lys Phe Leu Asn Ala 340 Ser Trp Giu Met Ser Gly Gly 165 Thr Thr Asn His Aia 245 Phe Thr Giu Giu Gly 325 Phe Gin Glu Thr Glu Leu Gin 150 Lys Asn Gin Ala Ser 230 Lys His Pro Ile Leu 310 Ile Tyr Giu Glu Pro Ser Ser 135 Giu Phe Asn Val Leu 215 Phe Asp Pro Cys Gin 295 Phe Phe Phe Lys Ile 375 Val Giu 120 Asn Giu Ser Gin Thr 200 Gin Leu Ile Gly Thr 280 Gly Asn Leu Val Val1 360 Lys Tyr 105 Glu Tyr Gly Gin Giu 185 Phe Phe Cys Gin Tyr 265 Lys Ile Thr Pro Met 345 Thr Thr Leu Leu Pro Al a Lys 170 Thr Val Arg Tyr Val 250 Lys Arg Gly Ser Pro 330 Lys Giu Ser Gly Ala Phe Tyr 155 Thr Phe Pro Leu Gly 235 Al a Lys Phe Asn Tyr 315 Leu Phe Met Tyr Ala Asp Asp 140 Gly Arg Gly Gin Tyr 220 Lys S er Val Giu Tyr 300 Cys Gin Leu Met Ala 380 Thr Arg 125 Phe Trp, Trp Ala Asn 205 Gly Asp Asn Val Met 285 Gin Pro Gly Asn Lys 365 Gly Ala 110 Val 3 in Ile Phe Leu 190 Gin Lys Gin Giu Asn 270 Thr Gin Tyr Asp Leu 350 Lys Val1 PCTLJS99/22955 Gly Met Leu Asp Gly Ala Thr Ile 160 Ser Ile 175 Asp Leu Thr Ile Asp Tyr Ala Leu 240 Ile Leu 255 Val Ser Leu Pro Cys His Ser Gin 320 Phe Gly 335 Thr Ser Phe Cys Lys Giu Lys Tyr Leu Ser Giu Tyr Cys Phe Ser Gly Thr Tyr Ile Leu Ser Leu 385 390 395 400 WO 00/23459 PCTIUS99/22955 Leu Leu. Gin Gly Tyr His Phe Thr Ala Asp Ser Trp Giu His Ile His 405 410 415 Phe Ile Gly Lys Ile Gin Giy Ser Asp Ala Giy Trp Thr Leu Gly Tyr 420 425 430 Met Leu Asn Leu Thr Asn Met Ile Pro Ala Giu Gin Pro Leu Ser Thr 435 440 445 Pro Leu Ser His Ser Thr 450 <210> 7 <211> 1437 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <220> <221> CDS <222> (1434) <400> 7 atg Met 1 agt Ser aaa Lys gtt Val1 tac Tyr cat His gtt Val gcc Ala ctt Leu aca Thr aag Lys atc Ile caa Gin cag Gin ctg Leu.

gca Ala cag Gin tat Tyr tat Tyr gta Val aaa Lys tgg Trp ctt Leu cta Leu ggg Gly aag Lys gaa Glu.

gta Val1 100 atc Ile 5 gtc Vali act Thr att Ile tgg Trp gaa Glu.

aat Asn gac agg atg caa ctc ctg tct tgc att gca cta.

Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu.

10 aca Thr ag t Ser gtg Val cca Pro 70 tgc Cys gaa Glu aac Asn tca Ser ctg Leu 55 gc a Al a agg Arg ata Ile agt Ser acc Thr 40 gat Asp gaa.

Glu.

gt t Val1 ggc Giy gca Al a 25 cag Gin gcg Ala aag Lys aaa Lys att Ile 105 cct Pro aac Asn ggt Gly gag Glu ggt Gly 90 tac Tyr cag Gin act Thr aaa Lys tct S er aat Asn 75 cct Pro c tg Leu tca Ser gca Ala tct Ser gac Asp gga Gly act Thr agt Ser t tg Leu.

c ac His aca.

Thr atc Ile gat Asp tct S er cca Pro aca Thr ggc Gly tca Ser tgc Cys 110 aca Thr gaa- Giu.

ag t Ser gtg Val1 aaa Lys atg Met aag Lys aac Asn tta Leu.

gtg Vai ttt Phe gaa.

Glu.

96 144 192 240 288 336 384 aga gct agg gaa gtg att cca agg tcc Arg Aia Arg Giu. Val Ie Pro Arg Ser 115 120 cac caa gag aca ccc gtt His Gin Glu. Thr Pro Val 125 WO 00/23459 WO 0023459PCT/US99/22955 tac ctg Tyr Leu 130 gga gcc acg gca Gly Ala Thr Ala ggc Gly 135 atg cgg ttg etc Met Arg Leu Leu agg Arg 140 atg gaa agt gaa Met Giu Ser Giu gag Giu 145 ttg gca gac agg Leu Ala Asp Arg gtt Val1 150 ctg gat gtg gtg Leu Asp Val Val gag Giu 155 agg agc ctc agc Arg Ser Leu Ser aac Asn 160 432 480 528 tac ccc ttt gao Tyr Pro Phe Asp ttc Phe 165 cag ggt gcc agg Gin Gly Ala Arg atc le 170 att act ggc caa Ile Thr Gly Gin gag gaa Giu Giu 175 ggt gcc tat Giy Ala Tyr cag aaa aca Gin Lys Thr 195 ggc Gly 180 tgg att act atc Trp Ile Thr Ie aac Asn 185 tat otg ctg ggc Tyr Leu Leu Gly aaa ttc agt Lys Phe Ser 190 aat aat cag Asn Asn Gin agg tgg ttc agc Arg Trp Phe Ser ata Ile 200 gte cca tat gaa Val Pro Tyr Giu acc Thr 205 gaa acc Giu Thr 210 ttt gga got ttg Phe Gly Ala Leu gac Asp 215 ott ggg gga gcc Leu Giy Giy Ala tct Ser 220 aca caa gto act Thr Gin Vai Thr ttt Phe 225 gta coo caa aao Val Pro Gin Asn eag Gin 230 act ate gag too Thr Ilie Giu Ser eca Pro 235 gat aat got etg Asp Asn Ala Leu caa Gin 240 672 720 768 ttt ego ctc tat Phe Arg Leu Tyr ggc Gly 245 aag gao tao aat Lys Asp Tyr Asn gtc Val 250 tao aca oat ago Tyr Thr His Ser ttc ttg Phe Leu 255 tgc tat ggg Cys Tyr Gly cag gtt gca Gin Val Ala 275 aag Lys 260 gat oag gca etc Asp Gin Ala Leu tgg Trp 265 eag aaa otg gc Gin Lys Leu Ala aag gao att Lys Asp Ile 270 oat cot gga His Pro Gly 816 864 agt aat gaa att Ser Asn Giu Ilie oto Leu 280 agg gac eea tgo Arg Asp Pro Cys ttt Phe 285 tat aag Tyr Lys 290 aag gta gtg aac Lys Val Val Asn gta Vai 295 agt gao ott tao Ser Asp Leu Tyr aag Lys 300 aoe ccc tgo ace Thr Pro Cys Thr aag Lys 305 aga ttt gag atg Arg Phe Glu Met aoct Thr 310 ctt oca ttc cag Leu Pro Phe Gin cag Gin 315 ttt gaa ato eag Phe Giu Ile Gin ggt Gly 320 912 960 1008 att gga aao tat Ile Gly Asn Tyr caa Gin 325 caa tge oat oaa Gin Cys His Gin ago Ser 330 ato otg gag etc Ile Leu Giu Leu tto aac Phe Asn 335 ace agt tac Thr Ser Tyr tgo Cys 340 cot tao toe eag Pro Tyr Ser Gin tgt Cys 345 gee tte aat ggg Ala Phe Asn Gly att ttc ttg Ile Phe Leu 350 1056 ooa eca ete cag ggg gat ttt ggg gca ttt tea get ttt tao ttt gtg 1104 WO 00/23459 Pro Pro Leu. Gin Gly Asp Phe 355 atg aag ttt tta aac ttg aca Met Lys Phe Leu Asn Leu Thr 370 375 act gag atg atg aaa aag ttc Thr Glu Met Met Lys Lys Phe 385 390 aca tct tac gct gga gta aag Thr Ser Tyr Ala Gly Val Lys 405 tct ggt acc tac att ctc tcc Ser Gly Thr Tyr Ile Leu Ser 420 gct gat tcc tgg gag cac atc Ala Asp Ser Trp Giu His Ie 435 gac gcc ggc tgg act ttg ggc Asp Ala Gly Trp Thr Leu Gly 450 455 cca gct gag caa cca ttg tcc Pro Ala Giu Gin Pro Leu Ser 465 470 <210> 8 <211> 478 <212> PRT <213> Artificial Sequence <400> 8 Met Ala Leu Trp Ile Asp Arg 1 5 Ser Leu. Ala Leu Val Thr Asn Lys Thr Gin Leu Thr Ser Ser Val Lys Tyr Gly Ile Val Leu 55 Tyr Ile Tyr Lys Trp Pro Ala 70 His Gin Val Glu Glu Cys Arg Val Gin Lys Val Asn Glu Ile 100 Gly 360 tca Ser tgt Cys gag Giu ctc Leu cat His 440 tac Tyr Ala gag Giu gc t Ala aag Lys ctt Leu 425 ttc Phe atg Met Phe aaa Lys cag Gin tac Tyr 410 ctg Leu att Ile ctg Leu Ser gtc Val cct Pro 395 c tg Leu c aa Gin ggc Gly aac Asn Ala tct Ser 380 tgg Trp agt Ser ggc Gly aag Lys ctg Leu 460 Phe 365 cag Gin gag Giu gaa Giu tat Tyr atc Ile 445 acc Thr Tyr gaa Giu gag Giu tac Tyr cat His 430 cag Gin aac Asn PCTIUS99/22955 Phe Vai aag gtg 11! Lys Vai ata aaa 12i Ile Lys 400 tgc ttt 12 Cys Phe 415 ttc aca 12 Phe Thr ggc agc i3 Giy Ser atg atc 13 Met Ile 52 00 48 96 44 92 aca cct ctc tcc cac tcc acc taa Thr Pro Leu Ser His Ser Thr 475 1437 Met Ser Thr 40 Asp Giu Vai Gly Gin Al a 25 Gin Aila Lys Lys Ile 105 Leu 10 Pro Asn Gly Giu Gly 90 Tyr Leu.

Thr Lys Ser Asn 75 Pro Leu Ser Ser Ala Ser Asp Gly Thr Cys Ser Leu His Thr Ile Asp Ile Ser Pro Thr Gly Ser Cys 110 Ala Thr Giu Ser Vai Lys Met Leu.

Lys Asn Leu Val Phe Glu WO 00/23459 WO 0023459PCT/US99/22955 Arg Tyr Giu 145 Tyr Gly Gin Giu Phe 225 Phe Cys Gin Tyr Lys 305 Ile Thr Pro Met Thr 385 Thr Ala Leu 130 Leu Pro Al a Lys Thr 210 Val Arg Tyr Val1 Lys 290 Arg Gly Ser Pro Lys 370 Giu Ser Arg 115 Gly Al a Phe Tyr Thr 195 Phe Pro Leu Gly Ala 275 Lys Phe Asn Tyr Leu 355 Phe Met Tyr Giu Al a Asp Asp Gly 180 Arg Gly Gin Tyr Lys 260 Ser Val1 Giu Tyr Cys 340 Gin Leu Met Al a Val Thr Arg Phe 165 Trp Trp Ala Asn Gly 245 Asp Asn Val Met Gin 325 Pro Gly Asn Lys Gly 405 Ile Ala Val 150 Gin Ile Phe Leu Gin 230 Lys Gin Giu Asn Thr 310 Gin Tyr Asp Leu Lys 390 Val1 Pro Gly 135 Leu Gly Thr Ser Asp 215 Thr Asp Ala Ile Val1 295 Leu Cys Ser Phe Thr 375 Phe Lys Arg 120 Met Asp Ala Ile Ile 200 Leu Ile Tyr Leu Leu 280 Ser Pro His Gin Gly 360 Ser Cys Giu Ser Arg Val Arg Asn 185 Val Gly Giu Asn Trp 265 Arg Asp Phe Gin Cys 345 Ala Glu Ala Lys Gin Leu Val Ile 170 Tyr Pro Gly Ser Val 250 Gin Asp Leu Gin Ser 330 Ala Phe Lys Gin Tyr 410 His Leu Giu 155 Ile Leu Tyr Ala Pro 235 Tyr Lys Pro Tyr Gin 315 Ile Phe Ser Val Pro 395 Leu Gin Arg 140 Arg Thr Leu Giu S er 220 Asp Thr Leu Cys Lys 300 Phe Leu Asn Ala Ser 380 Trp S er Giu 125 Met Ser Gly Gly Thr 205 Thr Asn His Ala Phe 285 Thr Giu Giu Gly Phe 365 Gin Giu Giu Thr Glu Leu Gin Lys 190 Asn Gin Ala Ser Lys 270 His Pro Ile Leu Ile 350 Tyr Giu Giu Tyr Pro Ser Ser Glu 175 Phe Asn Val Leu Phe 255 Asp Pro Cys Gin Phe 335 Phe Phe Lys Ile Cys 415 Val Giu Asn 160 Giu Ser Gin Thr Gin 240 Leu Ile Gly Thr Gly 320 Asn Leu Val Val Lys 400 Phe WO 00/23459 PCT/US99/22955 Ser Gly Thr Tyr Ile Leu Ser Leu Leu Leu Gin Gly Tyr His Phe Thr 420 425 430 Ala Asp Ser Trp Glu His Ile His Phe Ile Gly Lys Ile Gin Gly Ser 435 440 445 Asp Ala Gly Trp Thr Leu Gly Tyr Met Leu Asn Leu Thr Asn Met Ile 450 455 460 Pro Ala Glu Gin Pro Leu Ser Thr Pro Leu Ser His Ser Thr 465 470 475 <210> 9 <211> 24 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic signal sequence <400> 9 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 Ser Leu Ala Leu Val Thr Asn Ser <210> <211> 8 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic peptide <400> Asp Tyr Lys Asp Asp Asp Asp Lys 1 <210> 11 <211> 43 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <400> 11 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 WO 00/23459 PCT/US99/22955 Ser Leu Ala Leu Val Thr Asn Ser Ala Pro Thr Ser Ser Ser Thr Lys 25 Lys Thr Gin Leu Thr Ser Ser Thr Gin Asn Lys <210> 12 <211> 29 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <400> 12 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 Ser Leu Ala Leu Val Thr Asn Ser Ala Thr Gin Asn Lys <210> 13 <211> 31 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <400> 13 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 Ser Leu Ala Leu Val Thr Asn Ser Ala Ser Ser Thr Gin Asn Lys 25 <210> 14 <211> 87 <212> DNA <213.> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 14 ccggctggac tttgggctac atgctgaacc tgaccaacat gatcccagct gagcaaccat tgtccacacc tctctcccac gagcccc 87 <210> <211> 87 WO 00/23459 PCTIUS99/22955 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> gatcggggct cgtgggagag aggtgtggac aatggttgct cagctgggat catgttggtc aggttcagca tgtagcccaa agtccag 87 <210> 16 <211> 740 <212> DNA <213> Homo sapiens <220> <221> CDS <222> (42)..(737) <400> 16 cggtaccgct agcgtcgaca ggcctaggat atcgatacgt a gag ccc aga tct tgt 56 Glu Pro Arg Ser Cys gac Asp gcg Ala atc Ile gaa Glu cat His cgg Arg aag Lys aaa Lys ccg Pro tcc Ser gac Asp aa t Asn gtg Val gac Asp act Thr tca Ser cgg Arg cc t Pro gcc Ala gtc Val tac Tyr cac His gtc Val ac c Thr gag Glu aag Lys agc Ser aag Lys 105 aca Thr ttc Phe cc t Pro gtc Val aca Thr gtc Val tgc Cys tgc Cys ctc Leu gag Glu aag Lys aag Lys 75 ctc Leu aag Liys cca Pro ttc Phe gtC Val ttc Phe 60 ccg Pro acc Thr gtc Val ccg Pro ccc Pro aca Thr 45 aac Asn C99 Arg gtc Val tcc Ser tgc Cys cca Pro 30 tgc Cys tgg Trp, gag Glu ctg Leu aac Asri 110 cca Pro 15 aaa Lys gtg Val tac Tyr gag Glu cac His 95 aaa Lys cag Gin gca Ala cc Pro gtg Val gtg Val cag Gin 80 c ag Gin gcc Al a cc Pro cct Pro aag Lys gtg Val gac Asp t ac Tyr gac Asp ctc Leu cga Arg gaa Giu gac Asp gac Asp ggC Gly aac Asn tgg Trp cca Pro gaa Glu 130 gcc Ala acc Thr gtg Val gtg Val agc Ser o tg Leu gcc Ala 115 cca Pro gag Glu ctc Leu agc Ser gag Glu acg Thr aa t Asn 100 ccc Pro cag Gin ggc Gly atg Met cac His gtg Val tac Tyr ggc Gly atg Met gtg Val 104 152 200 248 296 344 392 440 cag aaa acc atc tcc aaa gcc aaa ggg Gin Lys Thr Ile Ser Lys Ala Lys Gly 120 125 WO 00/23459 PCT/US99/22955 tac ace ctg ccc cca tec egg gat gag ctg acc aag aac cag gtc agc 488 Tyr Thr Leu Pro Pro Ser Arg Asp Glu Leu Thr Lys Asn Gin Val Ser 135 140 145 ctg ace tgc ctg gtc aaa ggc ttc tat ccc agg cac ate gcc gtg gag 536 Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Arg His Ile Ala Val Glu 150 155 160 165 tgg gag agc aat ggg cag ccg gag aac aac tac aag ace acg cct ccc 584 Trp Glu Ser Asn Gly Gin Pro Glu Asn Asn Tyr Lys Thr Thr Pro Pro 170 175 180 gtg ctg gac tec gac ggc tec ttc ttc etc tac age aag etc acc gtg 632 Val Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Lys Leu Thr Val 185 190 195 gac aag age agg tgg cag cag ggg aac gtc ttc tea tgc tec gtg atg 680 Asp Lys Ser Arg Trp Gin Gin Gly Asn Val Phe Ser Cys Ser Val Met 200 205 210 cat gag gct ctg cac aac cac tac acg cag aag age etc tec ctg tct 728 His Glu Ala Leu His Asn His Tyr Thr Gin Lys Ser Leu Ser Leu Ser 215 220 225 ccg ggt aaa tga 740 Pro Gly Lys 230 <210> 17 <211> 232 <212> PRT <213> Homo sapiens <400> 17 Glu Pro Arg Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro Ala 1 5 10 Pro Glu Ala Glu Gly Ala Pro Ser Val Phe Leu Phe Pro Pro Lys Pro 25 Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val 40 Val Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr Val 55 Asp Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gin 70 75 Tyr Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu His Gin 90 Asp Trp Leu Asn Gly Lys Asp Tyr Lys Cys Lys Val Ser Asn Lys Ala 100 105 110 WO 00/23459 Leu Pro Ala Pro Met Gin Lys Thr Ile Ser Lys Ala Lys Gly 115 120 125 Arg Glu Pro Gin Val Tyr Thr Leu Pro Pro Ser Arg Asp Glu 130 135 140 Lys Asn Gin Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr 145 150 155 His Ile Ala Val Glu Trp Glu Ser Asn Gly Gin Pro Glu Asn 165 170 Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe 180 185 190 Ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gin Gin Gly Asn 195 200 205 Ser Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr 210 215 220 Ser Leu Ser Leu Ser Pro Gly Lys 225 230 <210> 18 <211> 18 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 18 ctttccatcc tgagcaac <210> 19 <211> 36 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 19 aaaaaactag tcagaacaaa gctttgccag aaaacg <210> <211> 24 <212> PRT <213> Mus sp.

PCT/US99/22955 Gin Pro Leu Thr Pro Arg 160 Asn Tyr 175 Leu Tyr Val Phe Gin Lys WO 00/23459 PCT/US99/22955 <400> Met Phe His Val Ser Phe Arg Tyr Ile Phe Gly Ile Pro Pro Leu Ile 1 5 10 Leu Val Leu Leu Pro Val Thr Ser <210> 21 <211> 46 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 21 ctagttctgg agactacaaa gatgacgatg acaaaaccca gaacaa 46 <210> 22 <211> 46 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 22 agctttgttc tgggttttgt catcgtcatc tttgtagtct ccagaa 46 <210> 23 <211> 89 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide <400> 23 ccggctggac tttgggctac atgctgaacc tgaccaacat gatcccagct gagcaaccat tgtccacacc tctctcccac tccacctaa 89 <210> 24 <211> 89 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Synthetic oligonucleotide WO 00/23459 PCTIUS99/22955 <400> 24 ggccttaggt ggagtgggag agaggtgtgg acaatggttg ctcagctggg atcatgttgg tcaggttcag catgtagccc aaagtccag <210> <211> 1464 <212> DNA <213> Artificial Sequence <220> <221> CDS <222> (1)..(1461) <220> <223> Description of Artificial construct of human CD39 Sequence: Fusion <400> atg gcc Met Ala 1 ctg tgg atc Leu Trp Ile 5 gac agg atg caa Asp Arg Met Gin ctg tct tgc att Leu Ser Cys lie gca cta Ala Leu agt ctt gca Ser Leu Ala aaa aca cag Lys Thr Gin ctt Leu gtc aca aac agt Val Thr Asn Ser gca Ala 25 cct act tca agt Pro Thr Ser Ser tct aca aag Ser Thr Lys gat gac aaa Asp Asp Lys cta act agt tca Leu Thr Ser Ser gga Gly gac tac aaa gat Asp Tyr Lys Asp acc cag Thr Gin aac aaa gca ttg Asn Lys Ala Leu cca Pro 55 gaa aac gtt aag Glu Asn Val Lys tat Tyr ggg att gtg ctg Gly Ile Val Leu gat Asp gcg ggt tct tct Ala Gly Ser Ser cac His 70 aca agt tta tac Thr Ser Leu Tyr atc Ile 75 tat aag tgg cca Tyr Lys Trp Pro gca Ala 192 240 288 gaa aag gag aat Glu Lys Glu Asn gac Asp aca ggc gtg gtg Thr Gly Val Val cat His 90 caa gta gaa gaa Gin Val Giu Glu tgc agg Cys Arg gtt aaa ggt Val Lys Gly ggc att tac Gly Ile Tyr 115 cct Pro 100 gga atc tca aaa Gly Ile Ser Lys ttt Phe 105 gtt cag aaa gta Val Gin Lys Val aat gaa ata Asn Glu Ile 110 gtg att cca Val Ile Pro 336 384 ctg act gat tgc Leu Thr Asp Cys atg Met 120 gaa aga gct agg Glu Arg Ala Arg gaa Glu 125 agg tcc Arg Ser 130 cag cac caa gag Gin His Gin Glu aca Thr 135 ccc gtt tac ctg Pro Vai Tyr Leu gga Gly 140 gcc acg gca ggc Ala Thr Ala Gly 432 atg cgg ttg ctc agg atg gaa agt gaa gag ttg gca gac agg gtt ctg WO 00/23459 WO 0023459PCT/US99/22955 Met 145 Arg Leu Leu Arg Met 150 Giu Ser Glu Giu Leu 155 Ala Asp Arg Val ttt gac ttc cag Phe Asp Phe Gin 175 Leu 160 ggt Gly gat gtg gtg gag Asp Val Val Glu agg Arg 165 ago ctc agc aac Ser Leu Ser Asn tac coo Tyr Pro 170 528 gcc agg atc Ala Arg Ile ate aac tat Ile Asn Tyr 195 att Ile 180 act ggc caa gag Thr Gly Gin Giu gaa Giu 185 ggt gce tat ggc Giy Ala Tyr Gly tgg att act Trp Ile Thr 190 tgg tte ago Trp Phe Ser 576 624 ctg ctg ggc aaa Leu Leu Gly Lys ttc Phe 200 agt eag aaa aca Ser Gin Lys Thr agg Arg 205 ata gte Ile Val 210 cca tat gaa ace Pro Tyr Giu Thr aat Asn 215 aat cag gaa acc Asn Gin Giu Thr ttt Phe 220 gga got ttg gao Gly Ala Leu Asp ott Leu 225 ggg gga goc tot Gly Gly Ala Ser aca.

Thr 230 eaa gto act ttt Gin Val Thr Phe gta Val 235 ccc caa aac eag Pro Gin Asn Gin act Thr 240 672 720 768 ate gag toe eca Ile Giu Ser Pro gat Asp 245 aat get etg eaa Asn Aia Leu Gin ttt Phe 250 ego etc tat ggc Arg Leu Tyr Gly aag gao Lys Asp 255 tac aat gte Tyr Asn Vai etc tgg cag Leu Trp Gin 275 tao Tyr 260 aea cat ago tte Thr His Ser Phe ttg Leu 265 tge tat ggg aag Cys Tyr Gly Lys gat eag gca Asp Gin Aia 270 aat gaa att Asn Giu Ile 816 864 aaa etg gee aag Lys Leu Aia Lys gao Asp 280 att eag gtt gea Ile Gin Val Ala agt Ser 285 etc agg Leu Arg 290 gao eca. tge ttt Asp Pro Cys Phe oat His 295 cet gga tat aag Pro Gly Tyr Lys aag Lys 300 gta. gtg aac gta Val Val Asn Val agt Ser 305 gac ett tao aag Asp Leu Tyr Lys ace Thr 310 coo tgo ace aag Pro Cys Thr Lys aga Arg 315 ttt gag atg aot Phe Giu Met Thr ott Leu 320 912 960 1008 eca ttc eag cag Pro Phe Gin Gin tt t Phe 325 gaa ate eag ggt Giu Ile Gin Gly att Ile 330 gga aac tat eaa Gly Asn Tyr Gin caa tgo Gin Cys 335 cat eaa. age His Gin Ser eag tgt gee Gin Cys Ala 355 ate Ile 340 etg gag etc tto Leu Giu Leu Phe aae Asn 345 aec agt tao tge Thr Ser Tyr Cys eet tao tee Pro Tyr Ser 350 ggg gat ttt Gly Asp Phe 1056 1104 tte aat ggg att Phe Asn Gly Ile tte Phe 360 ttg eca eca eto Leu Pro Pro Leu cag Gin 365 ggg gca ttt tea get ttt Gly Ala 370 Phe Ser Ala Phe tac Tyr 375 ttt gtg atg aag Phe Val Met Lys t tt Phe 380 tta aac ttg aca.

Leu Asn Leu Thr 1152 WO 00/23459 PCT/US99/22955 tca gag aaa gtc tct cag gaa aag gtg act gag atg atg aaa aag ttc 1200 Ser Glu Lys Val Ser Gin Glu Lys Val Thr Glu Met Met Lys Lys Phe 385 390 395 400 tgt get cag cct tgg gag gag ata aaa aca tet tac get gga gta aag 1248 Cys Ala Gin Pro Trp Glu Glu Ile Lys Thr Ser Tyr Ala Gly Val Lys 405 410 415 gag aag tac ctg agt gaa tac tgc ttt tct ggt acc tac att etc tec 1296 Glu Lys Tyr Leu Ser Glu Tyr Cys Phe Ser Gly Thr Tyr Ile Leu Ser 420 425 430 ctc ctt ctg caa ggc tat cat ttc aca gct gat tec tgg gag cac ate 1344 Leu Leu Leu Gin Gly Tyr His Phe Thr Ala Asp Ser Trp Glu His Ile 435 440 445 cat ttc att ggc aag ate cag ggc agc gac gcc ggc tgg act ttg ggc 1392 His Phe Ile Gly Lys Ile Gin Gly Ser Asp Ala Gly Trp Thr Leu Gly 450 455 460 tac atg ctg aac ctg ace aac atg ate cca get gag caa cca ttg tec 1440 Tyr Met Leu Asn Leu Thr Asn Met Ile Pro Ala Glu Gin Pro Leu Ser 465 470 475 480 aca cct etc tec cac tec ace taa 1464 Thr Pro Leu Ser His Ser Thr 485 <210> 26 <211> 487 <212> PRT <213> Artificial Sequence <400> 26 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 Ser Leu Ala Leu Val Thr Asn Ser Ala Pro Thr Ser Ser Ser Thr Lys 25 Lys Thr Gin Leu Thr Ser Ser Gly Asp Tyr Lys Asp Asp Asp Asp Lys 40 Thr Gin Asn Lys Ala Leu Pro Glu Asn Val Lys Tyr Gly Ile Val Leu 55 Asp Ala Gly Ser Ser His Thr Ser Leu Tyr Ile Tyr Lys Trp Pro Ala 70 75 Glu Lys Glu Asn Asp Thr Gly Val Val His Gin Val Glu Glu Cys Arg 90 Val Lys Gly Pro Gly Ile Ser Lys Phe Val Gin Lys Val Asn Glu Ile 100 105 110 WO 00/23459 PCTJUS99/22955 Leu Thr Asp Cys Met Giu Arg Ala Arg Giu Val Ile Pro Gly Arg Met 145 Asp Ala Ile Ile Leu 225 Ile Tyr Leu Leu Ser 305 Pro His Gin Gly Ser 385 Ile Ser 130 Arg Val1 Arg Asn Vai 210 Gly Giu Asn Trp Arg 290 Asp Phe Gin Cys Al a 370 Glu Tyr 115 Gin Leu Val Ile Tyr 195 Pro Giy Ser Val Gin 275 Asp Leu Gin Ser Ala 355 Phe Lys His Leu Giu Ile 180 Leu Tyr Ala Pro Tyr 260 Lys Pro Tyr Gin Ile 340 Phe Ser Val1 Gin Arg Arg 165 Thr Leu Giu Ser Asp 245 Thr Leu Cys Lys Phe 325 Leu Asn Aia Ser Glu Met 150 Ser Gly Gly Thr Thr 230 Asn His Ala Phe Thr 310 Giu Giu Gly Phe Gin 390 Thr 135 Giu Leu Gin Lys Asn 215 Gin Al a Ser Lys His 295 Pro Ile Leu Ile Tyr 375 Glu 120 Pro Ser Ser Giu Phe 200 Asn Val1 Leu Phe Asp 280 Pro Cys Gin Phe Phe 360 Phe Lys Val Giu Asn Giu 185 Ser Gin Thr Gin Leu 265 Ile Giy Thr Gly Asn 345 Leu Val Val Tyr Leu Giu Leu 155 Tyr Pro 170 Gly Ala Gin Lys Giu Thr Phe Val 235 Phe Arg 250 Cys Tyr Gin Val Tyr Lys Lys Arg 315 Ile Gly 330 Thr Ser Pro Pro Met Lys Thr Giu 395 G ly 140 Ala Phe Tyr Thr Phe 220 Pro Leu Gly Ala Lys 300 Phe Asn Tyr Leu Phe 380 Met Ala Asp Asp Gly Arg 205 Gly Gin Tyr Lys Ser 285 Val1 Giu Tyr Cys Gin 365 Leu Met Thr Arg Phe Trp 190 Trp Ala Asn Gly Asp 270 Asn Val Met Gin Pro 350 Giy Asn Lys Ala Val Gin 175 Ile Phe Leu Gin Lys 255 Gin Giu Asn Thr Gin 335 Tyr Asp Leu Lys Gly Leu 160 Gly Thr Ser Asp Thr 240 Asp Ala Ile Val Leu 320 Cys Ser Phe Thr Phe 400 Cys Ala Gin Pro Trp Giu Giu Ile Lys Thr Ser Tyr Ala Gly Val Lys 405 410 415 WO 00/23459 Glu Lys Tyr Leu Leu Leu.

435 His Phe Ile 450 Tyr Met Leu 465 Thr Pro Leu Leu 420 Gin Gly Asn Ser Ser Gly Lys Leu His 485 Giu Tyr Ile Thr 470 Ser Tyr Cys Phe Ser Gly Thr Tyr Ile 425 430 His Phe Thr Ala Asp Ser Trp, Giu 440 445 Gin Gly Ser Asp Ala Gly Trp Thr 455 460 Asn Met Ile Pro Ala Glu Gin Pro 475 Thr PCT/US99/22955 Leu Ser His Ile Leu Gly Leu Ser 480 27 <211> 464 <212> PRT <213> Artificiai Sequence <220> <223> Description of Artificiai construct of human CD39 Sequence: Fusion <400> 27 Met Ala 1 Ser Leu Giu Asn Ser Leu Val Val Lys Phe Met. Giu Pro Val Ser Giu 130 Ser Asn 145 Giu Glu Leu Al a Val1 Tyr His Val1 Arg Tyr 115 Glu Tyr Gly Trp Leu Lys Ile Gin Gin Al a 100 Leu Leu Pro Al a Ile 5 Val Tyr Tyr Val1 Lys Arg Gly Ala Phe Tyr 165 Asp Thr Gly Lys Giu 70 Val Glu Ala Asp Asp 150 Gly Arg Asn Ile Trp 55 Giu Asn Val1 Thr Arg 135 Phe Trp Met Ser Val1 40 Pro Cys Glu Ie Al a 120 Val Gin Ie Gin Ala 25 Leu Ala Arg Ile Pro 105 Gly Leu Gly Thr Leu 10 Thr Asp Giu Val Gly 90 Arg Met Asp Ala Ile 170 Leu Gin Ala Lys Lys 75 Ile Ser Arg Val Arg 155 Asn Ser Asn Gly Giu Gly Tyr Gin Leu Val 140 Ile Tyr Cys Lys Ser Asn Pro Leu His Leu 125 Glu Ile Leu Ile Ala Ser Asp Gly Thr Gin 110 Arg Arg Thr Leu Ala Leu Leu Pro His Thr Thr Gly Ile Ser Asp Cys Giu Thr Met Giu Ser Leu Gly Gin 160 Gly Lys 175 WO 00/23459 WO 0023459PCTIUS99/22955 Phe Asn Val Leu 225 Phe Asp Pro Cys Gin 305 Phe Phe Phe Lys Ile 385 Cys Phe Giy Met Ser Gin Thr 210 Gin Leu Ile Giy Thr 290 Gly Asn Leu Val Val1 370 Lys Phe Thr Ser Ile 450 Gin Giu 195 Phe Phe Cys Gin Tyr 275 Lys Ile Thr Pro Met 355 Thr Thr Ser Al a Asp 435 Pro Lys 180 Thr Val Arg Tyr Val1 260 Lys Arg Gly Ser Pro 340 Lys Glu Ser Giy Asp 420 Ala Ala Thr Phe Pro Leu Gly 245 Al a Lys Phe Asn Tyr 325 Leu Phe Met Tyr Thr 405 Ser Gly Glu Arg Gly Gin Tyr 230 Lys Ser Val Giu Tyr 310 Cys Gin Leu Met Ala 390 Tyr Trp Trp Gin Trp Ala Asn 215 Gly Asp Asn Val Met 295 Gin Pro Gly Asn Lys 375 Gly Ile Glu Thr Pro 455 Phe Leu 200 Gin Lys Gin Glu Asn 280 Thr Gin Tyr Asp Leu 360 Lys Val Leu His Leu 440 Leu Ser 185 Asp Thr Asp Ala Ile 265 Val Leu Cys Ser Phe 345 Thr Phe Lys Ser Ile 425 Gly Ser Ile Leu le Tyr Leu 250 Leu Ser Pro His Gin 330 Gly Ser Cys Giu Leu 410 His Tyr Thr Val Pro Gly Gly Glu Ser 220 Asn Val 235 Trp Gin Arg Asp Asp Leu Phe Gin 300 Gin Ser 315 Cys Ala Ala Phe Giu Lys Ala Gin 380 Lys Tyr 395 Leu Leu Phe Ile Met Leu Pro Leu 460 Tyr Aia 205 Pro Tyr Lys Pro Tyr 285 Gin Ile Phe Ser Val 365 Pro Leu Gin Gly Asn 445 Ser Giu 190 Ser Asp Thr Leu Cys 270 Lys Phe Leu Asn Ala 350 Ser Trp Ser Gly Lys 430 Leu His Thr Thr Asn His Ala 255 Phe Thr Giu Glu Giy 335 Phe Gin Glu Giu Tyr 415 Ile Thr Ser Asn Gin Ala Ser 240 Lys His Pro Ile Leu 320 Ile Tyr Giu Giu Tyr 400 His Gin Asn Thr WO 00/23459 PCT/US99/22955 <210> 28 <211> 474 <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial construct of human CD39 Sequence: Fusion <400> 28 Met Ala Leu 1 Ser Leu Ala Thr Ser Ser Ile Val Leu Trp Pro Ala Glu Cys Arg Asn Glu Ile Val Ile Pro 115 Thr Ala Gly 130 Arg Val Leu 145 Phe Gin Gly Trp Ile Thr Trp Phe Ser 195 Ala Leu Asp 210 Asn Gin Thr 225 Trp Leu Thr Asp Glu Val Gly 100 Arg Met Asp Ala Ile 180 Ile Leu Ile Ile 5 Val Gin Ala Lys Lys Ile Ser Arg Val Arg 165 Asn Val Gly Glu Asp Thr Asn Gly Glu 70 Gly Tyr Gin Leu Val 150 Ile Tyr Pro Gly Ser 230 Arg Asn Lys Ser 55 Asn Pro Leu His Leu 135 Glu Ile Leu Tyr Ala 215 Pro Met Ser Ala 40 Ser Asp Gly Thr Gin 120 Arg Arg Thr Leu Glu 200 Ser Asp Gin Ala 25 Leu His Thr Ile Asp 105 Glu Met Ser Gly Gly 185 Thr Thr Asn Leu 10 Ser Pro Thr Gly Ser 90 Cys Thr Glu Leu Gin 170 Lys Asn Gin Ala Ser Leu Thr Glu Ser Val 75 Lys Met Pro Ser Ser 155 Glu Phe Asn Val Leu 235 Phe Ser Cys Lys Lys Asn Val Leu Tyr Val His Phe Val Glu Arg Val Tyr 125 Glu Glu 140 Asn Tyr Glu Gly Ser Gin Gin Glu 205 Thr Phe 220 Gin Phe Leu Cys Ile Thr Lys Ile Gin Gin Ala 110 Leu Leu Pro Ala Lys 190 Thr Val Arg Tyr Ala Gin Tyr Tyr Val Lys Arg Gly Ala Phe Tyr 175 Thr Phe Pro Leu Gly 255 Leu Leu Gly Lys Glu Val Glu Ala Asp Asp 160 Gly Arg Gly Gin Tyr 240 Lys Gly Lys Asp Tyr Asn Val Tyr Thr His 245 WO 00/23459 PCT/US99/22955 Asp Gin A Asn Glu I: 2' Val Asn Vc 290 Met Thr L 305 Gin Gin C Pro Tyr S( Gly Asp P1 3! Asn Leu T] 370 Lys Lys P] 385 Gly Val L, Ile Leu S Glu His I: 4 Thr Leu G 450 Pro Leu S 465 <210> 29 <211> 473 la Leu 260 le Leu 75 al Ser eu Pro ys His er Gin 340 he Gly 55 hr Ser he Cys ys Glu er Leu 420 le His 35 ly Tyr er Thr Trp Gin Lys Leu Ala Lys Asp Ile Gin Val Ala Ser 265 270 Arg Asp Phe Gin 325 Cys Ala Glu Ala Lys 405 Leu Phe Met Pro Asp Leu Gin 310 Ser Ala Phe Lys Gin 390 Tyr Leu Ile Leu Leu 470 Pro Tyr 295 Gin Ile Phe Ser Val 375 Pro Leu Gin Gly Asn 455 Ser Cys 280 Lys Phe Leu Asn Ala 360 Ser Trp Ser Gly Lys 440 Leu His Phe Thr Glu Glu Gly 345 Phe Gin Glu Glu Tyr 425 Ile Thr Ser His Pro Ile Leu 330 Ile Tyr Glu Glu Tyr 410 His Gin Asn Thr Pro Cys Gin 315 Phe Phe Phe Lys Ile 395 Cys Phe Gly Met Gly Thr 300 Gly Asn Leu Val Val 380 Lys Phe Thr Ser Ile 460 Tyr Lys 285 Lys Arg Ile Gly Thr Ser Pro Pro 350 Met Lys 365 Thr Glu Thr Ser Ser Gly Ala Asp 430 Asp Ala 445 Pro Ala Lys Val Phe Glu Asn Tyr 320 Tyr Cys 335 Leu Gin Phe Leu Met Met Tyr Ala 400 Thr Tyr 415 Ser Trp Gly Trp Glu Gin <212> PRT <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Fusion construct of human CD39 <400> 29 Met Ala Leu Trp Ile Asp Arg Met Gin Leu Leu Ser Cys Ile Ala Leu 1 5 10 WO 00/23459 Ser Leu Al Ser Ser Th 3 Val Leu As Pro Ala Gi Cys Arg Va.

Giu Ile Gi Ile Pro Ar 11 Ala Gly Me 130 Val Leu As 145 Gin Gly Al Ile Thr Ii Phe Ser I 19 Leu Asp Le 210 Gin Thr Il 225 Lys Asp T) Gin Aia LE Giu Ile LE Asn Vai SE 290 PCT/US99/22955 Leu Thr a Leu r Gin p Ala u Lys 1 Lys y Ile 100 g Ser 5 t Arg p Val *a Arg *e Asn 180 .e Val u Gly .e Glu ~r Asn ~u Trp 260 ~u Arg ~r Asp Vai Thr Asn Ser Ser Thr Lys Lys Thr Gin 25 Asn Gly Giu Gly Tyr Gin Leu Val1 Ile 165 Tyr Pro Gly Ser Val1 245 Gin Asp Leu Lys Ser Asn 70 Pro Leu His Leu Glu 150 Ile Leu Tyr Al a Pro 230 Tyr Lys Pro Tyr Al a Ser 55 Asp Gly Thr Gin Arg 135 Arg Thr Leu Glu Ser 215 Asp Thr Leu Cys Lys 295 Leu 40 His Thr Ile Asp Giu i2 0 Met Ser Gly Gly Thr 200 Thr Asn His Ala Phe 280 Thr Pro Thr Gly Ser Cys 105 Thr Giu Leu Gin Lys 185 Asn Gin Al a Ser Lys 265 His Pro Giu Ser Val Lys 90 Met Pro Ser Ser Giu 170 Phe Asn Val Leu Phe 250 Asp Pro Cys Asn' Leu Val 75 Phe Glu.

Val Glu Asn 155 Glu Ser Gin Thr Gin 235 Leu Ile Gly Thr Gly 315 Val1 Tyr His Val Arg Tyr Glu 140 Tyr Gly Gin Glu Phe 220 Phe Cys Gin Tyr Lys 300 Lys Ile Gin Gin Ala Leu 125 Leu Pro Ala Lys Thr 205 Val Arg Tyr Val Lys 285 Arg Tyr Tyr Val Lys Arg 110 Gly Ala Phe Tyr Thr 190 Phe Pro Leu Gly Al a 270 Lys Phe Gly Lys Giu Val1 Glu Ala Asp Asp Gly 175 Arg Gly Gin Tyr Lys 255 Ser Val Glu Ile Trp Giu Asn Val Thr Arg Phe 160 Trp, Trp, Ala Asn Gly 240 Asp Asn Val Met Gin 320 Thr Leu Pro Phe Gin Gin Phe Giu Ile Gin Ile Gly Asn Tyr 305 310 WO 00/23459 Gin Cys His Tyr Ser Gin Asp Phe Gly 355 Leu Thr Ser 370 Lys Phe Cys 385 Val Lys Glu Leu Ser Leu His Ile His 435 Leu Gly Tyr 450 Leu Ser Thr 465 <210> <211> 463 <212> PRT Gin Cys 340 Aia Giu Al a Lys Leu 420 Phe Met Pro Ser 325 Ala Phe Lys Gin Tyr 405 Leu Ile Leu Leu Ile Phe Ser Val Pro 390 Leu Gin Gly Asn Ser 470 Leu Asn Ala Ser 375 Trp Ser Gly Lys Leu 455 His Giu Gly Phe 360 Gin Giu Giu Tyr Ile 440 Thr Ser Leu Ile 345 Tyr Giu Giu Tyr His 425 Gin Asn Thr Phe Asn 330 Phe Leu Phe Val Lys Val Ile Lys .395 Cys Phe 410 Phe Thr Gly Ser Met Ile Thr Pro Met Thr 380 Thr Ser Ala Asp Pro 460 Ser Pro Lys 365 Glu Ser Giy Asp Ala 445 Ala Tyr Leu 350 Phe Met Tyr Thr Ser 430 Gly Giu PCTIUJS99/22955 Cys Pro 335 Gin Giy Leu Asn Met Lys Ala Giy 400 Tyr Ile 415 Trp Giu Trp Thr Gin Pro <213> Artificial Sequence <220> <223> Description of Artificial construct of human CD39 <400> Met Giu Thr Asp Thr Leu Leu Leu 1 5 Gly-Ser Thr Gly Ala Pro Thr Ser Asn Val Lys Tyr Gly Ile Val Leu 40 Leu Tyr Ile Tyr Lys Trp, Pro Ala 55 Val His Gin Val Glu Giu Cys Arg 70 Phe Val Gin Lys Val Asn Giu Ile Sequence: Fusion Trp, Thr 25 Asp Glu Val Gly Val 10 Gin Ala Lys Lys Ile 90 Leu Asn Gly Giu Gly 75 Tyr Leu Lys S er Asn Pro Leu Leu Ala Ser Asp Gly Thr Trp, Leu His Thr Ile Asp Val Pro Thr Gly S er Cys Pro Giu Ser Val Lys Met WO 00/23459 WO 0023459PCT1US99/22955 Glu Val Giu Asn 145 Giu Ser Gin Thr Gin 225 Leu Ile Gly Thr Gly 305 Asn Leu Val Val1 Arg Tyr Glu 130 Tyr Gly Gin Glu Phe 210 Phe Cys Gin Tyr Lys 290 Ile Thr Pro Met Thr Ala Leu 115 Leu Pro Al a Lys Thr 195 Val Arg Tyr Val Lys 275 Arg Gly Ser Pro Lys 355 Giu Arg 100 Gly Ala Phe Tyr Thr 180 Phe Pro Leu Gly Ala 260 Lys Phe Asn Tyr Leu 340 Phe Met Glu Ala Asp Asp Gly 165 Arg Gly Gin Tyr Lys 245 Ser Val1 Giu Tyr Cys 325 Gin Leu Met Val Thr Arg Phe 150 Trp Trp Ala Asn Gly 230 Asp Asn Val Met Gin 310 Pro Gly Asn Lys Ile Ala Val1 135 Gin Ile Phe Leu Gin 215 Lys Gin Glu Asn Thr 295 Gin Tyr Asp Leu Lys 375 Pro Gly 120 Leu Gly Thr Ser Asp 200 Thr Asp Ala Ile Val 280 Leu Cys Ser Phe Thr 360 Phe Arg 105 Met Asp Ala Ile Ile 185 Leu Ile Tyr Leu Leu 265 Ser Pro His Gin Gly 345 Ser Cys Ser Arg Val1 Arg Asn 170 Val Giy Giu Asn Trp 250 Arg Asp Phe Gin Cys 330 Al a Giu Al a Gin Leu Val Ile 155 Tyr Pro Gly Ser Val 235 Gin Asp Leu Gin Ser 315 Al a Phe L~ys Gin Tyr 395 His Leu Giu 140 Ile Leu Tyr Ala Pro 220 Tyr Lys Pro Tyr Gin 300 Ile Phe S er Val Pro 380 Gin Arg 125 Arg Thr Leu Glu Ser 205 Asp Thr Leu Cys Lys 285 Phe Leu Asn Ala Ser 365 Trp Giu 110 Met Ser Gly Gly Thr 190 Thr Asn His Ala Phe 270 Thr Giu Giu Gly Phe 350 Gin Glu Thr Glu Leu Gin Lys 175 Asn Gin Ala Ser Lys 255 His Pro Ile Leu Ile 335 Tyr Giu Glu Pro Ser Ser Giu 160 Phe Asn Val Leu Phe 240 Asp Pro Cys Gin Phe 320 Phe Phe Lys Ile 370 Lys Thr Ser Tyr Ala Gly Val Lys Giu Lys Leu Ser Glu Tyr Cys 400 385 390 WO 00/23459 PCT/US99/22955 Phe Ser Gly Thr Tyr Ile Leu Ser Leu Leu Leu Gin Gly Tyr His Phe 405 410 415 Thr Ala Asp Ser Trp Glu His Ile His Phe Ile Gly Lys Ile Gin Gly 420 425 430 Ser Asp Ala Gly Trp Thr Leu Gly Tyr Met Leu Asn Leu Thr Asn Met 435 440 445 Ile Pro Ala Glu Gin Pro Leu Ser Thr Pro Leu Ser His Ser Thr 450 455 460 <210> 31 <211> 58 <212> PRT <213> Homo sapiens <400> 31 Met Ala Thr Ser Trp Gly Thr Val Phe Phe Met Leu Val Val Ser Cys 1 5 10 Val Cys Ser Ala Val Ser His Arg Asn Gin Gin Thr Trp Phe Glu Gly 25 Ile Phe Leu Ser Ser Met Cys Pro Ile Asn Val Ser Ala Ser Thr Leu 40 Tyr Gly Ile Met Phe Asp Ala Gly Ser Thr

Claims (13)

1. A polypeptide having the structure X-Y wherein X is selected from the group consisting of an Ala residue and heterologous peptides capable of adopting a stable secondary structure and Y is a soluble CD39 polypeptide selected from the group consisting of: polypeptides having an amino acid sequence as set forth in Figure 1 (SEQ ID NO:2) wherein the amino terminus is selected from the group consisting of amino acids 36 to 44, and the carboxy terminus is selected from the group consisting of amino acids 471 to 478; fragments of the polypeptides of wherein said fragments have apyrase activity; variants of the polypeptides of or wherein said variants have apyrase activity.

2. The polypeptide of claim 1 wherein Y is a soluble CD39 polypeptide selected from the group consisting of: polypeptides having a sequence consisting of amino acids 38-476 or 39-476 of SEQ ID NO:2; variant polypeptides that are at least 70% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity; variant polypeptides that are at least 80% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity; variant polypeptides that are at least 90% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity; *25 variant polypeptides that are at least 95% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity; variant polypeptides that are at least 98% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity; and variant polypeptides that are at least 99% identical in amino acid sequence to amino acids 36 to 478 of SEQ ID NO:2 or to a fragment thereof, wherein said variant polypeptides have apyrase activity.

3. The polypeptide of claim 1 wherein X comprises a peptide fragment from the amino terminal portion of mature IL-2, CD39-L2, CD39-L3 or CD39-L4. *o oo*

4. A polypeptide having the structure A-B-C wherein A is 1-20 amino acids from the amino terminal portion of mature IL-2, B is a linker of0-15 amino acids, and C is the soluble CD39 polypeptide Y of claim 1. A soluble CD39 polypeptide selected from the group consisting of: SEQ ID NO: 6, amino acids 25-464 of SEQ ID NO:27, amino acids 25-474 of SEQ ID NO:28, amino acids 27-473 of SEQ ID NO:29, amino acids 21-476 of SEQ ID NO:3, amino acids 21- 476 of SEQ ID NO:4, or amino acids 21-463 of SEQ ID NO:30; and fusion polypeptides comprising the polypeptides of wherein said fusion polypeptides have apyrase activity.

6. The soluble CD39 polypeptide of claim 5 having an amino acid sequence selected from the group consisting of SEQ ID NO: 6, amino acids 25-464 of SEQ ID NO:27, amino acids 25-474 of SEQ ID NO:28, amino acids 27-473 of SEQ ID NO:29, amino acids 21-476 of SEQ ID NO:3, amino acids 21-476 of SEQ ID NO:4, and amino acids 21-463 of SEQ ID

7. The soluble CD39 polypeptide of claim 6 having the sequence of amino acids 21-463 of SEQ ID

8. An isolated nucleic acid encoding a polypeptide of one of claims 1-7.

9. The nucleic acid of claim 8 wherein said nucleic acid is DNA.

10. The DNA of claim 9 having a sequence selected from the group consisting of: SEQ ID NO:5; and DNA sequences which, due to degeneracy of the genetic code, encode the polypeptide encoded by SEQ ID

11. The DNA of claim 9 wherein said DNA further encodes a leader peptide operably linked to the N-terminus of the polypeptide, wherein the leader peptide facilitates the extracellular secretion of the polypeptide.

12. The DNA of claim 11 wherein the leader peptide comprises all or part of a leader from IL-2, proinsulin, human growth hormone (huGH), 1L7, or Igkappa.

13. The DNA of claim 12 wherein the leader peptide comprises the sequence SEQ ID NO:9.

2879-WO PCT/US 99/22955 g y u i -1 2000 14. The DNA of claim 11 having a sequence selected from the group consisting of SEQ ID NO:7; and DNA sequences which, due to degeneracy of the genetic code, encode the polypeptide encoded by SEQ ID NO:7. A vector comprising the nucleic acid of claim 8. 16. The vector of claim 15 wherein said vector is an expression vector. 17. The vector of claim 16 wherein said vector is a eukaryotic expression vector. 18. A eukaryotic expression vector comprising the sequence SEQ ID 19. A eukaryotic exnression vector comprising the sequence SEQ ID NO:7. A recombinant cell comprising the nucleic acid of claim 8. 21. The cell of claim 20 wherein said cell is a prokaryotic cell. 22. The cell or claim 20 wherein said cell is a eukaryotic cell. 23. The cell of claim 22 wherein said cell is a COS cell or a CHO cell. 24. The cell of claim 23 wherein said cell is a CHO cell that has been adapted to grow in suspension and in the absence of serum. A recombinant CHO cell comprising a nucleic acid sequence SEQ ID NO:5 or SEQ ID NO:7. 26. A process for preparing a soluble CD39 polypeptide comprising culturing a recombinant cell according to claim 20 under conditions that permit expression of the CD39 polypeptide and recovering the CD39 polypeptide from the culture. 27. The process of claim 26 wherein the recombinant cell is a eukaryotic cell. 28. The process of claim 26 wherein the recombinant cell is a CHO cell that has been adapted to grow in suspension and in the absence of serum. 56 29. A polypeptide produced according to the process of claim 26. A polypeptide produced according to the process of claim 28. 31. A composition comprising a pharmaceutically acceptable carrier and a polypeptide according to one of claims 1-7. 32. A composition comprising a pharmaceutically acceptable carrier and a polypeptide according to claim 29. 33. A composition comprising a pharmaceutically acceptable carrier and a polypeptide according to claim 34. A method of inhibiting angiogenesis in a mammal in need of such treatment comprising administering a therapeutic amount of a biologically active soluble CD39 polypeptide. A polypeptide according to claim 1, substantially as hereinbefore described with reference to any one of the Examples. DATED: 23 February 2004 PHILLIPS ORMONDE FITZPATRICK Attorneys for: IMMUNEX CORPORATION *aa. •g *o*o* go•• oo*