CA2046906A1 - Soluble analogs of thrombomodulin - Google Patents

Abstract

The present invention relates to the production and use of recombinant DNA technology, to manufacture soluble thrombomodulin analogs useful, e.g., in anti-thrombotic therapy. Novel proteins, nucleic acid sequences, vectors, pharmaceuticals and methods of inhibiting thrombotic activity are disclosed herein.

Description

WO90/1~81 PCT

SOLUBLE ANALOGS OF THROMBOMODULIN

BACKGROUND OF THE INVENTION
Field of the Invention The present invention relates to the production and use of recombinant DNA technology to manufacture soluble thrombomodulin analogs useful, e.g., in antithrombotic therapy. Novel proteins, nucleic acid sequences, vectors, pharmaceuticals and methods of inhibiting thro~botic activity are disclosed herein.

Information Disclosure In the past decade, researchers and clinicians have intensified the search for improved antithrombotic compounds capable of preventing clot formation, as well as decreasing the chance of a reocclusion event after fibrinolytic therapy. Several disease states require treatment with an effective, safe anticoagulant. For example, anticoagulants are used for treatment of pulmonary embolism, to prevent deep venous thrombosis in patients undergoing prolonged bed rest or following certain surgical procedures, and for treatment of acute myocardial infarction with or without thrombolytic therapy.
Currently, anticoagulant therapies rely on heparin or warfarin compounds, each of which has various drawbacks. (Prevention of Venous Thrombosis and Pulmonary Embolism. Consensus Development Conference Statement 1986, 6(2):1-23.) Heparin in particular has several other physiological effects that impinge on hemodynamics and hemostasis. Also, these treatments necessitate very stringent laboratory monitoring, as patients are at severe risk for bleeding episodes. In these therapies, it is difficult to determine the proper dose that will provide anticoagulation yet not cause excessive bleeding.

WO90/10081 ~ ~~ PCT/US90/0095 Warfarin therapy often takes several days to become effective, and the effects require even longer to subside once the therapy is discontinued.
Thrombomodulin, a 105 kDa cell membrane protein, has been shown to have anticoagulant properties. In humans, thro~bomodulin is wldely distributed on the endothelium of blood vessels and lymphatics of all organs except the central nervous system and helps modulate the balance between the formation and the dissolution of blood clots.
Specifically, thrombomodulin controls the activlty of thrombin, a central enzyme in the coagulatlon of blood.
Thrombin's activities include conversion of fi~rinogen to fibrin and stimulation of platelet activation and-lS aggregation, ultimately resulting in blood clotformation. When thrombin is bound to thrombomodulin, however, the complex does not promote clot formation but instead acts as an anticoagulant by inhibitlng the direct procoagulant actions of thrombin, and by accelerating the activation of Protein C. Actlvated Protein C is a potent anticoagulant enzyme which disrupts the coagulation cascade. (See, for example, N. Esmon et al. (1982), J. Biol. Chem., 2S7:859-864; H.
Salem et al. (1984), J. Biol. Chem., 259:12246-51.) The DNA sequence encoding thrombomodulin has been isolated and sequenced both in its genomic form and as a cDNA molecule (R. Jackman et al. (1986), ~
Natl. Acad. Sci. USA., 83:8834-38 and (1987), 84:6425-29, both of which are incorporated herein by reference) and domains of thrombomodulin have been suggested (D.
Wen et al. (1987), Bioçhemistrx, 26:4350-57).
Functional assignments for several domains have been proposed on the basis of comparison to the DNA sequence of other known proteins, such as the LDL receptor. One study has suggested that the fifth and sixth epidermal growth factor (EGF)-like domains (within the E~F-like domain) have the capacity to bind thrombin (S. Rurosawa W090l1~81 ~ PCT/US90/00955 et al. (1988) ~. 8iol. Chem., 263:5993-5996). Another study has suggested that EGF-like domains 4,5 and 6 are sufficient to give Protein C activating activity.
(Zushi, et al., (1989) J.Biol.Chem. 264.10351.).
Unfortunately, native thrombomodulin is generally unsuitable for antithrombotic therapy. The amounts needed for commercial application makes purification from autopsy or biopsy samples impractical, since it is estimated that only about 300 lo mg throm~omodulin are present in the total endothelium of a single person.
Native thrombomodulin is membrane bound due to its inherent amino acid sequence. It is insoluble withou~ detergent treatment, generally rendering it unsuitable for systemic therapies such as for anticoagulation. Moreover, it would be problematic to produce or purify from any source, including a genetically engineered cell. The Xm of purified, detergent solubilized thrombomodulin for Protein C is only 5% that of the membrane bound protein, although the affinity is stronger when the isolated protein is incorporated into phospholipid vesicles (Galvin et al.
(1987), J. Biol. Chem., 257(5):2199-2205). It has been reported that native human thrombomodulin preparations are not very effective in blocking procoagulant activities of thrombin (Maruyama et al. (1985), J.
Clin. Invest., 75:987-991).
Soluble thrombomodulin-like molecules have been reported in human plasma and urine. These molecules have a greatly reduced ability to activate Protein C and may be degradation products of the cellular form of thrombomodulin (~. Ishii and Majerus, P., (1985), J. ClLn. Invest., 76:2178-81). These forms of thrombomodulin are present in such low amounts (approximately 0.8 mg/adult male), that they have proven difficult to characterize. More recently two groups have presented evidence that the levels of WO 90/10081 ~-? ? ~ ., Pcr/us9o/oo9s~j circulating thrombomodulin have been overestimated by a factor of 10 to 20 (Suzuki et al., (1988), J.
Biochem., 104:628-632 and references therein).
Elastase cleaved fragments of purified, native throm~omodulin have been described and partially sequenced (see Ishii, (1985), supra; Rurosawa et al., (1987), J. Biol. Chem., 262:2206-2212; Kurosawa et al J. Biol. Chem., (1988), 263:5593-5996; and Stearns et al., J. Biol. Chem., 1989, 264:3352-3356). The elastase fragment and some smaller fragments retained the ability to enhance the activation of Protein C ln vitro.
Additional references include EP 290,419 and Wo 88/05053l which discloses c~NA encoding the throm~omodulin protein. Analogs of thrombomodulin have been described in Wo 88/05053 which discloses analogs with varying numbers of EGF domains.
There exists a need for new compositions that exhibit the anticoagulant properties of thrombomodulin, are soluble in plasma, and are easily produced in large quantities. The present invention fulfills these and other needs.

SUMMARY OF THE INVENTION
This invention provides sequences of nucleic acid encoding a pepti.de selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -0 Y - asp.ser;
c) gly.ala.arg.ser. - Q: and d) ala.val.val.pro.arg.ser. - Q; where Y is selected from the sequence of amino acids provided in Table 2 ranging from a~ino acid 227 to amino acid 462, and where Q is selected from the sequence of amino acids provided in Table 2 ranging from amino acid 227 ~O90/tO08t PCT/US90/0095~
. . . , , i to amino acid 462, amino acid 350 to amino acid 462 or amino acid 227 to amino acid 497. The sequences represented by Y and Q denote the epidermal growth factor-like domains of the nati~e TM, the fourth, fifth and sixth EGF-like domains and the six EGF-like domains plus the O-linked glycosylation domain.
Also provided are polynucleotide sequences encoding the peptides described above where the sequence is selected from the group consisting of:
a) CACTG~GCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCCG~C~AGGTGGAC;
b) CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCC;
c) GGAGCCAGATCC-Z; and d) GCCGTCGTACCAAGATCC-Z;
where-X represents the nucleic acid bases numbered 879 to 1586, and where Z represents the nucleic acid bases numbered 879 to 1586, 1251 to 1586 or 879 to 1690 all numbers refer to bases depicted in Table 2 and vectors comprising such nucleotide sequences.
Further provided are substantially pure protein compositions selected from the peptides described above. Such compositions may be those wherein the peptide is a species of the peptide which exhibits about the expected molecular weight for said peptide under non-reducing chromatography. By "exhibits about the expected molecular weight" it is meant that the peptide appears as a band on a chromatographic gel within about 10% of the predicted molecular weight of the peptide based on its structure (exclusive of any glycosylation that might be present).
They may also be in a dry and salt-free form, a lyophilized powder, and/or essentially free of detergents. In another e~bodiment, the compositions may comprise one of the described peptides chemically linked to a fibrinolytic enzyme, preferably where the chemical linkage arises from a covalent bond between the enzyme and an amino acid residue of the peptide.

WO ~/1~81 PCT/US90/009~5 ... . . . . .

The fibrinolytic enzyme in ~uch a complex is preferably a streptokinase molecule bound to a plasminogen molecule or t-PA.
Pharmaceutical compositions having antithrombotic activity comprising a sterile preparation of a unit dose of thrombomodulin-like protein and having an amino acid sequence described above are disclosed, as well as methods for controlling thrombotic activity in a mammal by administering an effective amount of such compositions.
Further disclosed are compositions which comprise a biocompatible polymer having a surface which has peptides bound thereto selected from the group of peptides described above and methods for inhibiting blood clotting induced by polymers implanted into a living mammal wherein a thrombomodulin-like protein is bonded to the polymer and the poly~er is implanted into the mammal.
This invention additionally dislcoses an isolated DNA fragment containing a DNA sequence coding for multi-functional soluble human thrombomodulin (TM) analog, said analog consisting essentially of amino acid residues 350 to 462 or 390 to 462 or 22? to 462 of native TM and a targeting component. The targeting component is a sequence of amino acids capable of binding to fibrin and preferably also imparting fibrinolytic activity to the bifunctional TM analog such as the amino acid residues 4 - 530 of human tissue plasminogen activator as depicted in Table 9. These DNA
sequences are preferably operably linked to an expression control sequence and transfected into a cell. The analogs are preferably operably linked to a DNA sequence coding for a fusion protein comprising: a first amino acid sequence comprising a thro~bomodulin fragment without a stop transfer domain; and a second amino acid sequence comprising a targeting component;
wherein the bifunctional TM analog is secreted from the WO9Ot10081 PCT/US90/0095 ~ ~? ~ ~

host. Preferred sequences of TM used in these analogs are as follows: delta 1-389, delta 463-557 TM;
delta 1-226, delta 498-557 TM; delta 1-226, delta 463-557 TM; and delta 1-349, delta 463-557. Where the delta regions are missing and the numbers refer to the amino acid seouence of thrombomodulin depicted in Table

2. The bifunctional analog may have fibrinolytic activity resulting from the fusion with t-PA as described above.
The invention further provides for methods of suppressing coagulation in a patient comprising administering to the patient an amount of soluble TM
analog effective to suppress further coagulation, wherein the analog comprises TM amino acids as described above. The invention also provides for treating a patient suffering form acute myocardial infarction caused by a thrombus, said method comprising administering to the patient an amount of bifunctional soluble TM analog effective to suppress further coagulation and lyse the thrombus, wherein the bifunctional TM analog comprises TM amino acids as described above wherein said TM amino acid sequence is - fused at the C-terminus or the N-terminus to amino acids 4-530 of human t-PA.
The invention also provides for a bifunctional TM analog capable of binding thrombin with substantially the same affinity for thrombin as native thrombomodulin, said analog comprising native thrombomodulin without a functional stop transfer domain, wherein said bifunctional analog is further characterized by the ability to bind fibrin and cleave plasminogen to plasmin. In particular there is provided the analog which consists essentially of the amino acid sequence set forth in Table 9. There is also provided, TM analogs wherein said analogs consist essentially of the sisnal sequence of t-PA covalently attached to the amino terminus of the 6 EGF-like domain WO90/10081 PCT/US90/0~55 f~ ~ ., ,~
8 - ;
of thrombomodulin and the 6 EGF-like domain of thrombomodulin covalently attached to the amino terminus of t-PA.
This invention further relates to soluble thrombomodulin analogs and methods for making such analogs. More specifically these processes for making soluble thrombomodulin analogs comprising.the steps of:
culturing a host cell transformed with a DNA sequence encoding the analog; and collecting analogs secreted by the host cell; wherein the analogs lack a stop transfer sequence functional in the host cell. The host cells are preferably eukaryote cells such as yeast or mammalian cells. Of the soluble analogs it is preferred that the analog be derived from an isolated DNA fragment containing a DNA seguence coding for a soluble human thrombomodulin analog, capable of being secreted by a host cell transfected with said DNA
fragment, wherein said DNA seguence encodes a polypeptide, comprising amino acid resides 227-462 of native TM as depicted in Table 2. The preferred analogs have been describe above for the bifunctional fusion analogs and the analogs. Recombinant cells producing such analogs, the protein analogs themselves, pharmaceutical preparations and methods of using the analogs are also disclosed herein. The analogs preferably retain substantially the same or increased affinity to thrombin as native thrombomodulin.
Solubility is preferably achieved by deleting the stop transfer domain entirely or in part.

BRIEF DESCRIPTION OF THE DRAWINGS
Drawing l presents a schematic description of the role of thrombomodulin as an anti-coagulant molecule.
Drawing 2A and 2B schematically illustrate analogs disclosed herein.

Wo 90t10081 PCI/US90/0095~

9 ~ ., `, ._ ;,;
Drawing 3 illustrates two of the synthetic oligonucleotides used in Example 1 and the plasmid pUC19pcrTM7 .
Drawing 4 illustrates eukaryotic expression plasmid pTMl08 as described in Example 2.
Drawing 5 illustrates the construction of transient mammalian expression vector pTHR13 from pTHR5 as described in Example 3.
Drawing 6 illustrates the construction of baculovirus transfer vector pTMHY101 as described in Example 4.
Drawing 7 illustrates vectors used to produce fusion proteins that have both antithrombotic and fibrinolytic activity as described in Example 9.
Drawing 8 schematically illustrates a multifunctional molecule comprising a fibrinolytic exzyme (such as lys-plasminogen streptokinase complex or t-PA) covalently conjugated to a TM analog.
Drawing 9 illustrates an elution and activity profile produced during the separation of TM analog 6h-227/462 varient forms using anion exchange chromatography.
Drawing lOA shows a graph comparing the Protein C activating activity of nati~e ra~bit thrombomodulin and TM analog 6h/227-462. Drawing lOB
shows a dose response curve for TM analog 6h/227-462 in the Protein C activation assay.
Drawing llA illustrates the activity of TM
6h/227-462 in the APTT. Drawing 113 shows the activity of TM analog 6h/227-462 compared to antithrombin III
and heparin in the activated partial thromboplastin time assay.
Drawing 12 shows a graph depicting the prolongation of clotting time by TM analog 6h/227-462 measured in three assays; APTT, TCT and PT.

~0 ~/1~81 PCT/US90/00~

1 0 ~ r .
~; "
Drawing 13 shows the direct inhibition of thrombin mediated conversion of fibrinogen to fibrin by TM analog 6h/227-462.

DETAILED DESCRIPTION
In accordance with the present invention, novel compositions are provided which exhibit substantially all of the properties of native thrombomodulin except that they are made in a soluble form that is secreted ~rom the producer cell. Also provided are methods for the production of these compositions. These soluble thrombomodulin analogs can be produced economically and are easily purified and administered. A variety of therapeutic uses are anticipated, particularly with respect to anticoagulant and/or antithrombotic therapies. In order to fully appreciate the invention, the following detailed description is set forth.
"Thrombomodulin" or "TM", as used herein, refers to a protein with the biological activities of mammalian thrombomodulin typically found on endothelial cell surfaces and capable of acting as a receptor for thrombin (N. Esmon, (1987) Seminars in Thrombosis and Hemostasis, 13 (4L: 454-463 and European Patent Application No. 86904354.7, both incorporated herein by 2S reference). A DNA sequence encoding the full-length native human thrombomodulin protein has been isolated (European Patent Application No. 88870079.6, which is incorporated herein by reference). The cDNA sequence encodes a 60.3 kDa protein of 575 amino acids, which includes a signal sequence of about 18 amino acids.
Thrombomodulin within this definition includes natural allelic variations that may exist between indi~iduals.
The sequences for rabbit, bovine, mouse and human thrombomodulin exhibit a high degree of homology with one another. By analogy with other proteins, the structure of thrombomodulin can be presumptively WO ~/1~81 PCT/~IS90/00~

,.j ,, ,~, .,~
divided into domains. The term ~domain" refers to a discrete amino acid sequence that can be associated with a particular function or characteristic. The full length thro~bomodulin gene encodes a precursor peptide containing the following domains:
Approximate Amino Acid Position Domain -18--1 Signal peptide 1-226 N-terminal domain 227-462 6 EGF-like domains 463-497 O-linked Glycosylation 498-521 Stop Transfer Sequence 522-557 CytoplasmiC domain See C. S. Yost et al. (1983) Cell, 34:759-766 and D.
15 Wen et al. (1987) Biochemistry, ~26:4350-4357, both incorporated herein by reference.
In comparison to native TM, the TM analogs of the present invention have been modified to embrace selected specific regions of the native protein, that is (i) EGF-like domains 4, 5, and 6, (ii) all six cf the `six EGF-like domains EGF-like domains or (iii) the six EGF-like domains plus the O-linked glycosylation domain. Table 1 and Drawings 2A and 2B provide the specific analogs to which this invention is drawn.
The two peptides 11/6 and 11/2 refer to peptides which are similar to the elastase cleaved TM.
Elastase is an endopeptidase which cleaves at small neutral amino acids. The 11/6 refers to the number of amino acid residues present at the amino and carboxy 30 termini respectively, which are outside the 227-462 region of the 6 EGF-like domains. This peptide is analogous to the rabbit TM elastase fragment in the number of residues which are outside this region. The second TM analog is the ll/2 peptide. The 11 refers to the number of residues on the amino terminus which are beyond the 227 residue and the 2 refers to the 2 amino O 90/tO081 PCI'/US90/0095~;
r ,,. ,., :: .,:

acids residing beyond the 462 residue. The 11/2 analogue is believed to be the elastase fragment o~
human TM.
Other species contain various regions of the native thrombomodulin protein. Due to post-translational processing, these species have additional amino acid residues which originate from the signal sequence. For example, the species 4t/227-462, 4t/227-462:227-462, 4t/350-462, and 4t/227-497 all have four additional amino acids from the t-PA signal seouence.
TM analogs 6h/227-462, 6h/227:462:227-462, 6h/350-462, and 6h/227-497 all have six additional aminb acids from the hypodermin A signal sequence. These TM analogs are comprised of one or more of the EGF regions and/or the O-linked glycosylation domain. The specific regions embraced by these species are listed in the table below. Additional species are shown in Table 1 and Drawings 2A and 2B.

X/227-462 EGFs 1-6 X/227-462:227-462 EGFs 1-6 + EGFs 1-6 X/350-462 EG~s 4,5,& 6 X/227-497 EGFs 1-6 + O-linked Gly.
X is the number of amino acids derived from the signal sequence.
Another aspect of the present invention describes multifunctional proteins, proteins in which the TM analog is joined with sequences corresponding to other proteins that impart ~iological characteristics to the TM analog not normally associated with native thrombomodulin. These multifunctional proteins are refered to as TM
heterologs and are composed of a first functionality that is associated with native thrombomodulin, thrombin binding for example, and a second functionality that is heterologous, ie. is a functionality from a source other than native thrombomodulin. The second ~090/10081 PCT/~S90/0095 13 ~ -functionality may effect localization of the TM
heterolog so as to modify its affinity for specific tissue structures occurring ln vivo, such as cell surfaces or fibrin clots. The heterologous peptide sequence may provide an additional biological activity, such as a proteolytic activity. A preferred proteolytic activity is the enzymatic cleavage of plasminogen to plasmin. The targeting components of this invention are designated as targeting to fibrin.
It is preferred that the target also have fibrolytic activity. The multifunctional TM heterologs disclosed herein may be fusion proteins, wherein they have been provided with heterologous protein domains by N- and/or C-terminal extensions of the original TM sequences.
lS Additionally, the multifunctional TM heterolog may be a chemical conjugate to an heterologous protein through a spacer molecule. Ruger et al, (1987), Proc. Natl.
Acad. Sci. USA 84:76S9-7662, and Smith and Cassels (1988), Fibrinolysis 2:189-195 have described chemical linkages between t-PA and other molecules. These molecules have an altered affinity for cell surfaces, or enhanced affinity for fibrin.
Preferred heterologous protein domains may be encoded by nucleotide sequences isolated from tissue plasminogen acti~ator or pro-urokinase. Suitable t-PA
sequences may include, in addition to the entire protein:
Domain Amino Acids Kringle 1, Kringle 2, Protease 95-530 Kringle 2, Protease180-530 Protease 279-530 A. Chain 4-278 See L. Patthy, (1985) Cell, 41:657-663 for description of t-PA domains. For a heterologous domain isolated from human urokinase, amino acids 1-411 would be preferred (Jacobs et at. (1985) DNA, 4:139-146. Human ~'O90/1~81 PCT/US90/00955 14 ~' ~ ~
; j ! ~;
Glu-plasminogen and plasmin derivatives may also find use as heterologous domains which provide enhanced fibrinolytic activity. Amino acids 1-791 of Glu-plasminogen have such a utility. F. J. Castellino, S Methods i~ ~n~Y~oloqy 80:365-378 Academic Press (1981).
A particularly preferred embodiment of a heterologous domain of the present invention includes amino acids 4-S30 of human tissue plasminogen activator (t-PA).
In one aspect of the inYention, the TM
analogs described are secreted from the eukaryotic cells in which they are produced. As used herein, a "soluble TM analog~ is a TM analog which is soluble in an a~ueous solution and can be secreted by a cell. For pharmacological admini tration, the soluble TM analog may optionally be combined with phospholipid vesicles, detergents or other similar compounds well known to those skilled in the art of pharmacological formulation. The TM analogs of the present invention are soluble in the blood stream, making the analogs useful in various anticoagulant and other therapies.
In general, the TM analogs of the present invention are active in one or more functional assays that ~easure aspects of clotting. These can include the activation of Protein C, by acting as a co-factor of thrombin, the inhibition of the conversion of fibrinogen to fibrin by throm~in (thereby substantially inhibiting blood coagulating activity) or the inhibition of thrombin mediated platelet activation and aggregation. The latter two assays can be run on an automatic coagulation timer according to the manufacturer's specifications: Medical Laboratory Automation Inc. distributed by American Sciantific Products, McGaw Park, Illinois. (See also H. ~. Salem et al. (1984) J. 3iol. Chem., 259:12246-12251, which is incorporated herein by reference). Multifunctional proteins will also be active in assays that measure the activities not normally associated with TM. For ~`~'090/1~81 PCT/US90/009~5 15 ~;, ,..' - ;., instance, the fibrinolytic activity of fusion proteins combining TM analog sequences with, for example, t-PA
can be measured in any of several assays, such as the fibrin plate assay described by Haver~atet and Brakman, 5 (197S) Proa. in Chem. Fibrin. Thromb. 1:151-lS9.
Nucleic acid sequences encoding these proteins may be used to transform cells. These sequences, most typically DNA, may or may not contain an associated signal sequence. The resulting transformed or transfected cells can be ~ultured to readily produce these proteins in large quantities.
The nucleic acid sequences described are operable and useful in a number of host cells wnich are adapted to tissue culture. Typically, the cells are eukaryotic cells, preferably human, that can grow rapidly in standard media preparations. Prokaryotic or yeast cells may also be suitable for the use of the described invention.
This invention embraces molecular genetic manipulations that can be achieved in a variety of known ways. The r~combinant cells, plasmids, and DNA
sequences of the present invention provide a means to produce pharmaceutically useful compounds wherein the compound, secreted from recombinant cells, is a soluble derivative of thrombomodulin.
The detailed description which follows discloses alternative methods for producing soluble thrombomodulin analogs from DNA sequences encoding thrombomodulin, and is followed by specific examples of preferred methods.

Use of Thrombomodulin Analoas as an Anticoaaulant/
Antithrombotic.
The underlying pathology of thrombotic disorders is that a clot forms in response to a stimulus such as, for example, a damaged vessel wall.
This stimulus triggers the coaaulation cascade WO 90/10081 PCI'/US90/00955 generating thrombin, which has the ability to convert fibrinogen to fibrin, the matrix of the clot.
Soluble thrombomodulin analogs administered systemically will protect against thrombus formatio~
because they will inhibit the generation of thrombin, via the activated Protein C system, and/or inhibit the action of thrombin on fibrinogen without disturbing other coagulation parameters. Thus, the use of soluble thrombomodulin analogs will be both safe and effective at preventing unwanted thrombus forma~ion. The effect of thrombomodulin can be overcome by the large amounts of thrombin generated by a serious injury to vessels allowing a hemostatic plug to form.
Diseases in which thrombus formation plays a significant etiological role include myocardial infarction, disseminated intravascular coagulatio~, deep vein ~hrombosis, pulmonary embolism, septic shock, acute respiratory distress syndrome, unstable angina and other arterial and venous occlusive conditions. In all of these, as well as in other diseases in which thrombus formation is pathological, soluble thrombomodulin analogs alone or in combination with thrombolytics are useful for treatment, either to cure the disease or to prevent its progression to a more severe state. Soluble thrombomodulin analogs also provide a safe and effective anticoagulant, for e~ample, in patients receiving bioprostheses such as heart valves or patients requiring extracorporeal circulation. These compounds may replace heparin and warfarin in the treatment of, for example, pulmonary embolism or acute myocardial infarction.
In particular these compounds would find a role in the prevention of deep vein thrombosis (DVT), for instance after surgery. The formation of blood clots in the leg is itself a non-fatal condition but is very closely tied to the development of pulmonary embolism (PE). PE is difficult to diagnose and can be wo90/tO081 PCT/US90/~

17 ~ --fatal. Despite the investigation and clinical use of several prophylactic regimens, DVT and the resulting PE
remain a significant problem in many patient populations and particularly in patients undergoing orthopedic surgery. (Prevention of Venous Thrombosis and Pulmonary Embolism. Consensus Development Conference Statement 1986, 6(2):1-23.). Existing prophylactic treatments such as heparin, warfarin and dextran typically reduce the incidence of DVT in orthopedic surgery patients from more than 50% in patients at risk receiving no prophylaxis to 2S-30%
among treated patients. There are serious side effects associated with these compounds, primarily bleeding complications. Daily laboratory tests and adjustments in dosage are required to minimize bleeding episodes while retaining some efficacy.
It has been suggested that other inhibitors of thrombin such as antithrombin III may be useful to prevent DV$. Commercially available human ATIII, however, is purified from human plasma and carries with it the possibility of potentially serious contamination, such as from infectious viral particles.
In addition, while ATIII inhibits thrombin directly, it does not enhance the activation of Protein C. The ability to activate Protein C is a decided improvement in an anticoagulant therapeutic. Based on the shortcomings of existing prophylactics, an antithrombotic which is effective at preventing DVT
without predisposing the patient to bleeding or other complications could make a significant impact on patient recovery and well-being.
Angioplasty is a procedure frequently used for restoring patency in occluded arteries. Although patency may be restored, this procedure often damages 3S the endothelial lining of the artery, and blood clots begin to form as a result. Soluble thrombomodulin 18 -' analogs administered in conjunction with angioplasty will prevent this deleterious side effect.
Many acute thrombotic and embolic diseases are currently trea~ed with fibrinolytic therapy in order to remove the thrombus. The condition that has been most widely investigated is acute myocardial infarction (heart attack). Agents currently in use for treating acute myocardial infarction include streptokinase, tissue plasminogen activator and urokinase. Use of these agents can lead to serious bleeding complications. Patients who have had a thrombus removed by fibrinolytic therapy and in whom the blood flow has been restored frequently reocclude the affected vessel, i.eO, a clot reforms. Attempts have been made to prevent reocclusion by increasing the dose or duration of treatment with a thrombolytic agent, but the incidence of bleeding then in~reases.
Thus the therapeutic index for these drugs is narrow.
The use of soluble thrombomodulin analogs provides protection against reocclusion, and its specific action is local rather than systemic, i.e., where thrombin is being generated or being released from a clot. Therefore, when used in combination with a thrombolytic agent, whose dose can then be decreased, the risk of bleeding can be substantial-ly reduced.
The TM heterologs containing additional domains that impart fibrinolytic activity in combination with antithrombotic activity will provide additional and superior utilities over currently available compounds. These multifunctional proteins direct the soluble TM heterolog to the site of the fibrin clot. The fibrinolytic activity conferred upon the compound by the heterologous domain provides a superior thrombolytic agent. As the clot is lysed by ~5 the fibrinolytic action of the, for example, t-PA
domain(s), the TM domain(s) are inherently located precisely where needed to bind thrombin and inhibit any WO ~/10081 PCT/US90/0095~

1 9, ~
further growth of the clot matrix. This throm~in may be either newly generated by the coagulation pathway or released from the dissolving clot. The therapeutically effective dose of the multifunctional protein will be less than the doses of each molecule administered individually, reducing any concerns about the broader systemic action of either the TM analog or the t-PA and any conseguential undesirable side effects.
Administration of soluble thrombomodulin analogs would be by a bolus intravenous injection, by a constant intravenous infusion or by a combination of both routes. Also, soluble thrombomodulin mixed with appropriate excipients may be taken into the circulation from an intramuscular site. As used lS herein, a therapeutically effective dose is defined as that level of TM analog required to prevent for~ation of pathological clots.
Systemic treatment with thrombomodulin analogs can be monitored by determining the activated partial thromboplastin time (APTT) on serial samples of blood taken from the patient. The coagulation time observed in this assay is prolonged when a sufficient level of TM analog is achieved in the circulation.
However, this is a systemic measurement of efficacy, and perhaps a dose that is effective at the site of a clot would not be effective in prolonging the APTT.
Dosing levels and regimens can be adjusted so that an adequate concentration of thrombomodulin is maintained as measured by, for example, the APTT assay or the Protein C activation assay.
In contrast to full length thrombomodulin, the analogs of this invention should offer an improved pharmaceutical. It is anticipated that these analogs will offer superior characteristics from a manufacturing perspective, a pharmaceutical perspective or both.

W090/lOn81 PCT/~S90/0095~
s~ .~ . ,~; ,, , j General Methods Generally, the definitions of nomenclature and descriptions of general labsratory procedures used in this application can be found in T. Maniat7s et al.
Molecular Clonina. A Laboratory Manual, (1982) Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.
The manual is hereinafter referred to as Maniatis and is hereby incorporated by reference.
All enzymes were used according to the manufacturer's instructions.
Libraries are constructed in bacteriophage lambda vectors. Phage are packaged ln vitro, as described in Maniatis. Recombinant phage are analyzed by plaque hybridization as described in E3enton and Davis, (1977) Science, 196:180-182. Colony hybridization is carried out as generally described in M. Grunstein et al. (1975) Proc._Natl. Acad. Sci. USA., 72:3961-3965. Nucleotide sizes are given in either kilobases (kb) or base pairs (bp). These are estimates derived from agarose or acrylamide gel electrophoresis or from published DNA sequences.
Oligonucleotides that are not commercially available can be chemically synthesized according tc the solid phase phosphoramidi~e triester method first described by S.L. Beaucage and M.H. Caruthers, (1981) Tetrahedron Letts., 22(20~:185~-1862 using an automated synthesizer, as described in D.R. Needham-VanDevanter et al . (1984) Nucleic Acids Res., 12:6159-6168.
Purification of oligonucleotides was by either native acrylamide gel electxophoresis or by anion-exchange HPLC as described in J.D. Pearson and F.E. Regnier, (1983) J. Chrom., 255:137-149.
The seguence of the cloned genes and synthetic oligonucleotides can be verified using the chemical degradation method of A.M. Maxam et al. (1980) Methods in Enzymolooy, 65:499-560. The sequence can be confirmed after the assembly of the oligonucleotide WO90/1~81 PCT/US90/00~5 fragments into the dou~le-stranded DNA sequence using the method of Maxam and Gilbert, supra, or the chain termination method for sequencing double-stranded te~plates of R.B. Wallace et al. (1981) Gene, 16:21-26.
Southern Blot hybridization techniques were carried out according to Southern et al. (1975) J. Mol. Biol., 98:503.
This invention relates to cloning and use of genes for expression vectors in eukaryotic or prokaryotic cells. Intermediate vectors are cloned for amplification in prokaryotes such as E. coli, Bacillus or Streptomvces. ~ost preferred is E. coli because that organism is easy to culture and more fully understood than other species of prokaryotes. ~he Maniatis manual contains methodology sufficient to conduct all subse~uently described clonings in E. coli.
Strain MH-1 is preferred unless otherwise stated. All E. coli strains are grown on Luria broth (LB) with glucose, or M9 medium supplemented with glucose and acid-hydrolyzed casein amino acids. Strains with resistance to antibiotics were maintained at the drug concentrations described in Maniatis. Transformations were performed according to the method described by D.A. Morrison, (1977) J. Bact., 132:349-351 or by J.E.
Clark-Curtiss and R. Curtiss, (1983) Methods in Enzymology, 101:347-362, Eds. R. Wu et al., Academic Press, New York. Representative vectors include pBR322 and the pUC series which are available from commercial sources.

Definitions For purposes of the present invention the following terms are defined below.
The term "vector" refers to viral expression systems, autonomous self-replicating circular DNA
(plasmids), and includes both the expression and nonexpression plasmids. Where a recombinant ~0gO/10081 PCTIUS90/00955 22 ~ !~
microorganism or cell culture is described as hosting an "expression vector," this includes both extrachromosomal circular DNA and DNA that has been incorporated into the host chromosome(s). Where a vector is being maintained by a host cell, the vector may either be stably replicated by the cells during mitosis as an autonomous structure, or is incorporated within the host's genome.
The term "promoter" is a region of DNA
involved in binding the RNA polymerase to initiate transcription.
The term "operably linked" refers to a juxtaposition wherein the components are configured so as to perform their usual function. Thus, control sequences or promoters operably linked to a coding sequence are capable of effecting the expression of the coding sequence.
The term "control sequence" refers to a DNA
sequence or sequences which are capable, when properly ligated to a desired coding sequence, of affecting its expression in hosts compatible with such sequences.
Such control sequences include at least promoters in both prokaryotic and eukaryotic hosts, and optionally, transcription termination signals. Additional factors necessary or helpful in effecting expression may also be identified. As used herein, ~control sequences"
simply refers to whatever DNA sequence may be useful to result in expression in the particular host used.
The term "free of native glycosylation"
refers to material which is altered in its polysaccharide composition from that naturally produced by the human body.
A "substantially pure composition" of the peptides described herein are those which migrate largely as distinct bands on reducing gels in standard electrophoresis techniques never mind. It is preferably referring to homogeneous compositions having W090/10081 PCT/US90/00~5 ~_, ~ _ less than about 10% contamination and most preferably less than 3% contamination with peptides other than that desired.

Gene Svnthesis Pu~lication of the DNA seouence encoding human thrombomodulin facilitates the preparation of genes encoding soluble thrombomodulin analogs. The full length gene is used as a ~tarting point to construct DNA sequences encoding soluble thrombomodulin analogs. The analogs of the present invention are soluble derivatives which lack a stop transfer sequence. Further, these analogs are secreted from eukaryotic cells which have been transfected or transformed with plasmids containing genes which encode these polypeptides. Methods for making modifications, such as amino acid substitutions, deletions, or the addition of signal sequences to cloned genes are known.
The full length gene for thrombomodulin can be prepared by several methods. Ruman genomic libraries are commercially available. Oligonucleotide probes, specific to the thrombomodulin gene, can be synthesized using the published gene seouence. Methods for screening genomic libraries with oligonucleotide probes are known. The publication of the gene sequence for thrombomodulin demonstrates that there are no introns within the coding region. Thus a genomic clone provides the necessary starting material to construct an expression plasmid for thrombomodulin using known methods.
A thrombomodulin encoding DNA fragment can be retrieved by taking advantage of restriction endonuclease sites which have been identified in regions which flank or are internal to the gene.
(R.W. Jackman et al. (1987) Proc. Natl. Acad. Sci.
USA., 84:6425-6429).

W~ ~/1~81 PCT/US90/~MK~

24 .
Alternatively, the full length gene is obtained from a cDNA bank. Messenger RNA prepared from endothelial cells provides suitable starting material for the preparation of cDNA. A cDNA molecule containing the gene encoding thrombomodulin is identified as described above. Methods for making cDNA
banks are well known (See Maniatis).
Synthetic oligonucleotides can be used to construct TM genes. This is done using a series of over-lapping oligonucleotides usually 40-120 bp in length, representing both the sense and non-sense strands of the gene. These DNA fragments can be annealed, ligated and cloned.
An alternative and preferred method combines the use of synthetic oligonucleotide primers with polymerase extension on a mRNA or DNA template. This polymerase chain reaction (PCR) method amplifies the desired nucleotide sequence. RestriCtion endonuclease sites can be incorporated into the primers. U.S.
Patents 4,683,195 and 4,683,202 describe this method.
Genes amplified by the PCR reaction can be purified from agarose gels and cloned into an appropriate vector.
Genes encoding soluble thrombomodulin analogs may be constructed using the gene encoding full length thrombomodulin as a starting material. Alternatively, analogs can be constructed from synthetic oligonucleotides or by use of the polymerase chain reaction on either a DNA or RNA template to produce a pure fragment of the desired gene which can then be cloned. Combinations of these methods can be used in order to obtain the DNA sequences that are desired.
Alterations in the natural gene sequence can be introduced by the technigues of n vitro mutagenesis or by use of the polymerase chain reaction with primers that have been designed to incorporate appropriate mutations.

~ O90/10081 PCT/VS9Ot~
;~ ~ ~ . .

For soluble thrombomodulin analogs constructed entirely from synthetic oligonucleotides, DNA codons may be selected to minimize secondary structure that would otherwise interfere with transcription and translation, without altering the native amino acid sequence (H. Grosjean and W. Fiers, (1982) Gene, 18:199-209). Further it may be desirable to alter the sequence of the native gene encoding thrombomodulin to incorporate unique restriction sites.
Such restriction sites find use in construction of a vector capable of expressing a gene encoding a soluble thrombomodulin analog. It may also be useful to design particular codon usage for efficiency in expression by a particular host cell.
Genes encoding soluble throm~omodulin analogs may have more than one copy of a native domain. By way of illustration and not limitation, a soluble thrombomodulin analog of the present invention may have more than one copy of the 5 EGF-liXe domains. Such a molecule may have enhanced utility in a pharmaceutical product since this region binds thrombin.
The soluble TM analogs described herein are secreted when expressed in eukaryotic cell culture.
Secretion may be obtained by the use of the native signal sequence of the thro~bomodulin gene.
Alternatively, genes encoding the soluble thrombomodulin analogs of the present invention may be ligated in proper reading frame to a signal sequence other than that corresponding to the native thrombomodulin gene. For example, the signal sequence of t-PA, (see commonly assigned co-pending USSN 074,083 filed July 16, 1987 incorporated herein by reference) or of hypodermin A or B (see commonly assigned co-pending U.S. Serial No. 148,749, filed January 27, 1989 incorporated hereby by reference) can be linked to the polypeptide. In the preferred embodime~t of the present invention, use is made of the signal sequence WO90/10081 PCT/US90/00~5 of t-PA which contains the ~econd intron of the human t-PA gene. The inclusion of the intron enhances the productlvity of the adjacent structural gene (see commonly assigned co-pending USSN #003,611 filed January 14, 1987 incorporated herein by reference).
With the analogs of this invention, that portion of the gene encoding the entire carboxyl terminal region of thrombomodulin is deleted.
Therefore, it is necessary to add a stop codon so that translation will be terminated at the desired position.
Alternatively, a stop codon can be provided by the desired expression plasmid. Additionally a polyadenylation sequence is necessary to ensure proper processing of the mRNA in eukaryotic cells encoding the soluble thrombomodulin analog. Also, it may be necessary to provide an initiation codon, if one is not present, for expression of the soluble TM analog. Such sequences may be pro~ided from the native gene or by the expression plasmid.
The TM analogs described may be encoded by genes wherein at least one N-terminal or C-terminal domain is deleted from the naturally occurring human DNA sequence encoding TM, and replaced with a DNA
sequence encoding a fibrinolytic enzyme. T-PA and pro-urokinase are fibrinolytic enzymes useful for such an analog. It may be desirable to replace both an N-terminal and a C-terminal domain of the native thrombomodulin with heterologous gene sequences. For example, in one embodiment of the present invention the gene encoding full length TM is modified so that the signal sequence and N-terminal domains, comprising amino acids -18 through 226 are deleted and replaced with DNA encoding the signal ~equence of t-PA (amino acids -32 to -l according to the numbering system in Table 6). In another embodiment, the TM analog described above may be further modified by the use of DNA encoding amino acids 4 to 530 of t-PA to replace Wo ~/1~81 PCT/US90/00955 27 - ;l the O-linked glycosylation, stop transfer, and cytoplasmic domains of native thrombomodulin (amino acids 463-557). Such a recombinant D~A molecule can be transfected into a host cell, thus providing a multifunctional protein capable of binding fibrin, activating Protein C, and converting plasminogen to plasmin.
A preferred source of the t-PA gene can be obtained by isolating the t-PA gene from an E. coli culture (strain MH-l) on deposit with American Type Culture Collection (ATCC) in Bethesda, Maryland having Accession No. 67,443. Standard cloning techniques are sufficient to obtain the t-PA plasmid and to insert heterologous domains, as desired, into genes encoding TM analogs.
The use of a heterologous domain to impart fibrinolytic activity to a TM analog may be desirable for other TM analogs described herein, and the use of the gene encoding pro-urokinase would have analogous utility in creating a multifunctional protein.
The thrombomodulin analogs of this invention are described by their amino acid sequences and by their DNA sequence, it being understood that the analogs include their biological equivalents such that 2S this invention includes minor substitutions and deletions of amino acids that have substantially little impact on the biological properties of the analogs. It should also be understood that alternative seguences could be used to express soluble thrombomodulin analogs in various host cells. Furthermore, due to the degeneracy of the genetic code, equivalent codons may be substituted to encode the same polypeptide sequence.

Cloninq Vectors Cloning vectors suitable for replication and integration in prokaryotes or eukaryotes and containing transcription and translation terminators, initiation WO90/10081 PCTJ~S90/00955 , . . .

sequences, and promoters useful for regulation of the expression of soluble thrombomodulin analogs are described herein. The vectors are comprised of expression cassettes containing at least one independent terminator sequence, sequences permitting replication of the plasmid in both eukaryotes and prokaryotes, i.e., shuttle vectors, and selection markers for both prokaryotic and eukaryotic systems.

Ex~ession of Soluble Thrombomodulin Analogs in Prokaryotic Cells In addition to the use of cloning methods in E. coli for amplification of cloned sequences it may be desirable to express soluble TM analogs in prokaryotes.
It is possible to recover a therapeutically functional protein from E. coli transformed with an expression plasmids encoding a soluble thrombomodulin analog. See Example 12.
Methods for the expression of cloned genes in bacteria are well known. To obtain high level expression of a cloned gene, such as a DNA sequence encoding a soluble thrombomodulin analog in a prokaryotic system, it is essential to construct expression vectors which contain, at the minimum, a strong promoter to direct mRNA transcription termination. Examples of regulatory regions suitable for this purpose are the promoter and operator region of the E. ~ galactosidase gene, the E. coli tryptophan biosynthetic pathway, or the leftward promoter from the phage lambda. The inclusion of selection markers in DNA vectors transformed in E. coli are useful. Examples of such markers include the genes specifying resistance to ampicillin, tetracycline, or chloramphenicol.
See Maniatis for details concerning selection markers and promoters for use in E. coli. In the described embodiment of this invention pUCl9 is used as W~90/10081 PCTtUS90/00955 a vector for the subcloning and amplification of desired gene sequences.
.

Expression of Soluble Thrombomodulin Analoas in Eukaryotic Cells It is expected that those of skill in the art are knowledgeable in the expression systems chosen for expression of the desixed TN analog and no attempt to describe in detail the various methods known for the expression of proteins in eukaryo~es will be ~ade.
The DNA sequence encoding a soluble TM analog can be ligated to various expression vectors for use in transforming host cell cultures. The vectors typically contain marker genes and gene sequences to initiate transcription and translation of the soluble lS thrombomodulin analog gene.
The vectors preferably contain a marker gene to provide a phenotypic trait for selection of transformed host cells such as dihydrofolate reductase, metallothionein, hygromycin, or neomycin phosphotransferase. The nuclear polyhedral viral protein from Autoqrapha californica is useful to screen transfected insect cell lines from S~odoDtera frugiperda and Bombvx mori to identify recombinants.
For yeast, Leu-2, Ura-3, Trp-l, and ~is-3 are known selectable markers ~Gene (1979) 8:17-24). There are numerous other markers, both known and unknown, which em~ody the above scientific principles, all of which would be useful as markers to detect those eukaryotic cells transfected with the vectors embraced by this invention.
In the examples provided below, hygromycin is included as a eukaryotic ~election marker in C~L-l cells. Of the higher eukaryotic cell systems useful for the expression of soluble TM analogs, there are numerous cell systems to select from. Illustrative examples of mammalian cell lines include RPMI 7932, ~'090/10081 ~ PCT/US90/00 VE~O and HeLa cells, Chinese hamster ovary (CH0) cell lines, WI38, BHR, COS-7, C127 or MDCK cell lines.
Cells suitable for use in this invention are commercially available from the American Type Culture Collection. Illustrative insect cell lines include Spodo~tera fru~ipe~da (fall Armyworm) and ombyx mori (silkworm).
As indicated above, the expression vector, ex. plasmid, which is used to transform the host cell, preferably contains gene sequences to initiate the transcription and sequences to control the translation o~ the soluble TM analog protein gene sequence. These sequences are referred to as expression control sequences. When the host cell is of insect or mammalian origin, illustrative expression control sequences include but are not limited to the following:
the retroviral long terminal repeat promoters ((1982) Nature, 297:479-483), SV40 promoter ((1983) Science, 222:524-527, thymidine kinase promoter (J. Banerji et 20 al. (1982) Cell, 27:299-308), or the beta-globin promoter (P.A. Luciw et al. (1983) Cell, 33:?0S-716).
The recipient vector nucleic acid containing the expression control sequences is cleaved using restriction enzymes and adjusted in size as necessary or desirable. This segment is ligated to a DNA
sequence encoding a soluble TM analog by means well known in the art.
For expression of heterologous proteins in yeast, the following promoters are useful for 30 expression: GAL1,10 ((1984) Mol. and Cell. Biol., 4:1440-48); ADH2 ((1983) J. Biol. Chem. 258:2674-2682);
and PH05 ((1982) EMB0 J., 6:675-680).
When higher animal host cells are employed, polyadenylation or transcription termination sequences need to be incorporated into the vector. An example of a polyadenylation sequence is the polyadenylation ~'090/10081 PCT/US90/009~5 t', .-~ ' ` '` ? :^
3i sequence from SV40, which may also function as a transcription terminator.
Genes incorporated into the appropriate vectors can be used to direct synthesis of proteins in either transient expression systems or in stable clones. In the former case yields are low, but the experiments are quick. In the latter case it takes more time to isolate high producing clones. Different vectors may be used for the two different types of experiments. In particular, in the case of transient expression, sequences may be included within the plasmid that allow the plasmid to replicate to a high copy number within the cell. These sequences may be derived from virus such as S~40 (e.g. C. Doyle et al.
(1985) J. Cell Biol., 100:704-714) or from chromosomal replicating sequences such as murine a~ltonomous replicating sequences (Weidle et al. (1988) Gene, 73:427-437). ~he vector for use in transient expression should also contain a strong promoter such as the S~40 early promoter (e.g., A. van Zonnenfeld et al. (1987) Proc. Natl. Acad. Sci. USA., 83:4670-4674) to control transcription of the gene of interest. While transient expression provides a rapid method for assay o~ gene products, the plasmid DNA is not incorporated into the host cell chromosome. Thus, use of transient expression vectors does not provide stable transfected cell lines. A description of a plasmid suitable for transient expression is provided by A. Aruffo & B. Seed, (1987~ Proc. Natl. Acad. Sci.
USA., 84:8573-8S77.
The disclosed embodiment makes use of C~L-1 cells. These are derived from RPMI 7932 melanoma cells, a readily available human cell line. The CHL-1 cell line has been deposited with the ATCC according to the conditions of the Budapest Treaty and has been assigned #CRL 9446, deposited June 18, 1987.

W~) 90/10081 PCI/US90/00955 ;^ r

3 2 ~~ ~
The soluble TM analogs may be expressed in yeast cells. The expression of heterologous proteins in yeast is well known. F. Sherman et al., Methods in Yeast Genetics, Cold Spring Harbor Laboratory, tl982) is a well recognized work describing the various methods which have utility for producing soluble thrombomodulin analogs in yeast.
Soluble TM analogs may alternatively be produced in the insect cell lines described above using the baculovirus system. This system has been described by V.A. Luckow and M.D. Summers (1988) Bio/Technoloqv, 6:47-55. Generally, this expression system provides for a level of expression higher than that provided by most mammalian systems. The baculovirus infects the host insect cells, replicates its genome through numerous cycles, and then produces large amounts of polyhedron crystals. The polyhedron gene can be replaced with a TM analog gene. The polyhedron promoter will then make large amounts of analog protein following infection of the culture host cell and replication of the baculovirus genome. The non-secreted gene product is harvested from the host 3-7 days post infection. Alternatively, the soluble TM
protein may be secreted fro~ the cells if appropriate signal sequences are present on the protein.
The host cells are competent or rendered competent for transfection by various means. There are several well-known methods of introducing DNA into animal cells. These include: calcium phosphate precipitation, DEAE-dextran technique, fusion of the recipient cells with bacterial protoplasts containing the DNA, treatment of the recipient cells with liposomes containing the DNA, electroporation and microinjection of the DNA directly into the cells.
See, B. Perbal, "Practical Guide to Molecular Cloning,"
2nd edition, John ~iley & Sons, New York and Wigler, et al. (1987) Cell, 16:777-785.

Wo90/1~81 PCT/US90/~X~

- 3~ ;
Culturina Cells It is preferred that the host cell is capable of rapid cell culture and able to appropriately glycosylate expressed gene products. Cells known to be suitable for dense growth in tissue culture are particularly desirable and a variety of invertebrate or vertebrate cells have been employed in the art, both normal and transformed cell lines.
The transfected cells are grown up by means well known in the art. For examples, see Biochemical Methods in Cell Culture and Virology, Kuchler, R. J., Dowden, Hutchinson and Ross, Inc. (1977). The expression products are harvested from the cell medium in those systems where the protein is secreted from the host cell or from the cell suspension after disruption of the host cell system by, e.g., mechanical or enzymatic means, which are well known in the art.

Purification of Soluble TM Analo~s The present invention provides soluble TM
analogs which are secreted by cultured recombinant eukaryotic cells. The analogs are produced in serum-free or serum supplemented media and are secreted intact. If prokaryotic cells are used, the TM analogs may be deposited intracellularly. The analogs may be 2S glycosylated or non-glycosylated. Following the growth of the recom~inant cells and concomitant secretion of TM analogs into the culture media, this "conditioned media" is harvested. The conditioned media is then clarified by centrifugation or filtration to remove cells and cell debris. The proteins contained in the clarified media are concentrated by adsorption to any suitable resin such as, for example, Q Sepharose or metal chelators, or by use of ammonium sulfate fractionation, polyethylene glycol precipitation, or by ultrafiltration. Other means known in the art may be equally suitable. Further purification of the soluble WO90/10081 PCT/US90/00~5 ~, TM analogs can be accomplished in the manner described in Galvin, J. B., et al. (1987) J. Biol. Chem., 262:2199-?205 and Salem, ~.H. et al. (1984) J. Biol.
Chem., 259:12246-12251 and in the manner described in the embodiment disclosed herein. The purification of soluble TM analogs secreted by cultured cells may require the additional use of, for example, affinity chromatography, ion exchange chromatography, sizing chromatography or other protein purification techni~aues.
Recombinant TM analogs may be produced in multiple conformational forms which are detectable under nonreducing chromatographic conditions. Removal of the higher molecular weight species and those species having a low sp~cific activity is desirable and is achieved by a variety of chromatographic techniques including anion exchange or size exclusion chromatography.
Recombinant TM analogs may be concentrated by pressure dialysis and buffer exchanged directly into volatile buffers (e.g., N-ethylmorpholine (NEM), ammonium bicarbonate, ammonium acetate, and pyridine acetate). In addition, samples can be directly freeze-dried from such volatile buffers resulting in a stable protein powder devoid of salt and detergents. In addition, freeze-dried samples of recombinant analogs can be efficiently resolubilized before use in buffers compatible with infusion (e.g., phosphate buffered saline). Other suitable buffers might include hydrochloride, hydrobromide, sulphate acetate, benzoate, malate, citrate, glycine, glutamate, and aspartate.

Che~ically Linkina the Soluble TM Analoas to Fibrinolvtic Enzvmes Blocking the active site of fibrinolytic enzymes can improve their therapeutic ratio. The p-WO 90tlO081 PCT/US90/00955 ,~ . .: .
3 5 r . . ~
anisoyl derivative of human lys-plasminogen activator complex (a complex between plasminogen and streptokinase also called APSAC or Eminase~ (Beecham Research Labs., Ltd., Middlesex, U.K.)), for example, has demonstrated clinical advantages when compared to streptokinase alone, such as a longer half-life. The acylation of the active si~e of the lys-plasminogen streptokinase complex leaves the fibrin binding site free, thus the complex retains the ability to bind to an unwanted thrombus. (Smith, RAG, Dupe, RJ, English, PD, & Geari, J., (1981) Nature 290:505-507). The active fibrinolytic complex is produced in situ after slow deacylation by hydrolysis. Tha same technique can be used to chemically link other molecules, such as a soluble TM analog, to the active site of a fibrinolytic enzyme through free amine groups on the analog. The peptides disclosed herein can be chemically linked to a fibrinolytic enzyme such as lys-plasminogen streptokinase, t-PA or the like through a covalent bond between the enzyme, and an amino acid residue within the peptide.
In one embodiment, lys-plasminogen streptokinase complex can be linked specifically and reversibly to a TM analog by using a heterobifunctional inverse substrate of plasmin (see, for example, Drawing 8). In a second embodiment, the fibrinolytic enzyme linked to the TM analog can be tissue plasminogen activator (t-PA). The acyl-enzyme is prepared according to the method described in Smith, RAG &
Cassels, (1988), Fibrinolysis, 2:198-195. Following removal of excess acylating agent by gel filtration, the acyl-enzyme is reacted with thiolated TM analog.
The acylation reaction is highly specific for the active center serine residue in either the plasminogen or the t-PA molecules. Thiolated TM is preferably produced using analoqs having only one free amine group, 11/2, 4t-227/462 or 6h-227/462, which is at the W~90/1~81 PCT/US90/0095~
,. .....

N-terminus of the molecule. Thiolated TM analog is produced by reacting the analog with 2-iminothiolane as described in Smith (1988), supra. The thiolated TM
analog is then reacted with the acylated fibrinolytic enzyme (lys-plasminogen streptokinase complex or t-PA) producing TM>SK:Plg or TM~t-PA, which can then be further purified by high performance gel permeation chromatography or other purification means as described in Smith (1988), supra.
The activity of the TM analog or the fibrinolytic enzyme can be individually assayed by methods described herein. Hydrolysis of the acyl-enzyme conjugate can be monitored on SDS gels and by release of fibrinolytic activity.
lS Chemical linkages are used to produce acyl-enzyme pro-drugs. This approach provides a reversible TM>fibrinolytic conjugate with the combined properties of both molecules, thus providing a route to thrombus specificity, via the lys-plasminogen streptokinase complex or t-PA, with a kinetic approach to the control of bleeding. The TM>fibrinolytic acyl-enzyme pro-drug acts as a relatively slowly cleared reservoir of fibrinolytic activity. Thus, the half-life of the fibrinolytic and the clot specificity of the TM analog is increased and the clearance rates and systemic effects of both are decreased.

Evaluati~n of TM Analo~s Thrombomodulin activity can be detected in a variety of assays that depend on alterations in the action of thrombin. The ability of thrombomodulin or its soluble analogs to accelerate the thrombin catalyzed activation of Protein C can be measured in the assay described by Salem et al. (above) and Galvin et al. In brief, this assay consists of two steps.
The first step is the incubation of the test thrombomodulin or TM analog with thrombin and Protein C

W~90/10081 PCT/US90/00~5 ,~ . ~ . ~ ` , , .
37 -_ ~ ^
under defined conditions. In the second step, the thrombin is inactivated with hirudin or antithrom~in III and heparin, and the activity of the activated Protein C is determined by use of a chromogenic substrate, whereby the chromophore is released by the proteolytic activity of activated Protein C. This assay is carried out with purified reagents.
Alternatively the effect of a TM analog can be measured using plasma in clotting time assays such as the activated partial thromboplastin time (APTT), thrombin clotting time (TCT) and/or prothrombin time tPT). These assays distinguish between different mechanisms of coagulation inhibition. Inhibition in any one of these assays demonstrates the presence of a lS systemic anticoagulant state. The APTT measures the effect of all of the components involved in coagulation except platelets, Factor VII and calcium ions while the TCT specifically measures the inhibition of thrombin.
The PT measures prothrombin-activity at constant lipoprotein and calcium concentrations.
These assays are used to identify soluble TM
analogs that are able to bind thrombin and to activate Protein C in both purified systems and in a plasma mllleu .
Further assays are then used to evaluate soluble TM analogs having enhanced properties by assaying them for their ability to inhibit thrombin induced fibrin formation, thrombin induced platelet activiation/aggregation and throm~in induced inactivation of protein S, as well as thrombin catalyzed activation of Factors V, VIII and XIII. One can also measure the ability of the analog to enhance the inactivation of thrombin by antithro~bin III and heparin cofactor II.
Soluble TM analogs which have utility as therapeutic agents are able to activate Protein C in physiological concentrations of calcium ion.

W~90/10081 PCT/US90/0095 38 ~ l Formulation and Use of ThrolmPomodulin Analoas Soluble thrombomodulin analogs described herein may be prepared in a lyophilized or liquid formulation. The material is to be provided in a concentration suitable for pharmaceutical use as elther an injectable or intravenous preparation.
These compounds can be administered alone or as mixtures with other physiologically acceptable active materials, such as one-chain t-PA, or inactive materials, or with suitable carriers such as, for example, water or normal saline. These compounds can be administered parenterally, for example, by injection. Injection can be subcutaneous, intravenous or intramuscular. These compounds are administered in pharmaceutically effective amounts and often as pharmaceutically acceptable salts, such as acid addition salts. Such salts can include, e.g., hydrochloride, hydrobromide, phosphate, sulphate, acetate, benzoate, malate, citrate, glycine, glutamate, and aspartate, among others. The analogs described herein may display enhanced in vivo activity by incorporation into micelles. Methods for incorporation into ionic detergent micelles or phospholipid micelles are known.
An antithrombotic agent can be prepared using the soluble TM analogs described herein and can consist of a completely purified thrombomodulin analog alone or in combination with a thrombolytic agent as described a~ove. Compounds of the present invention which are shown to have the above recited physiological effects can find use in numerous therapeutic applications such as, e.g., the inhibition of blood clot formation.
Thus, these compounds can find use as therapeutic agents in the treatment of various circulatory disorders, such as, for example, coronary or pulmonary embolism, strokes, as well as the prevention of reocclusion following thrombolytic therapy, and these W~90t10081 PCT/US90/0095~
.. .. ! . . ~
39 i - ; .
compounds have utility in the cessation of further enlargement of a clot during an infarction incident.
Further, the compounds disclosed can be useful for treatment of systemic coagulation disorders such as DIC, which is often associated with septicemia, certain cancers and toxemia of pregnancy.
These compounds can be administered to mammals for veterinary use, such as with domestic animals, and for clinical use in humans in a manner similar to other therapeutic agents, that is, in a physiologically acceptable carrier. In general, the administration dosage will range from about O.OOOl to lO0 mg/kg, and more usually O.OOl to O.l mg/kg of the host body weight. These dosages can be administered by constant infusion over an extended period of time, until a desired circulating level has been attained, or preferably as a bolus injection.

coatin~ of Biomaterials with TM Analo~s The use of altered prosthetic endovascular or cardiovascular devices anywhere in the circulation system results in the formation of thrombus, a blood-derived mass as a pathological consequence of activating hemostatic mechanisms under variable flow conditions. Typically, thrombogenesis in association with prosthetic endovascular or cardiovascular devices includes the following sequence:
(a) exposure of the surface to circulating blood;
(b) platelet adherence, aggregation and release of platelet components;
(c) thrombin generation and fibrin formation;
(d) thrombin dissolution which requires plasmin generation and fibrinolysis. In general, when blood contacts an artificial surface, the surface will rapidly acquire a layer of absorbed plasma proteins which will mediate subsequent thrombotic events. This series of events also follows when blood is circulated Wo9O/1oo8~ PCT/US90/0 through an extracorporeal device, such as a heart/lung machine.
It has been desirable to introduce various coatings onto the polymeric surfaces of such blood-contacting devices to promote thromboresistance.Thrombomodulin represents a new class of molecule suitable for creating a thromboresistant surface. It is especially suitable as such a surface since it has no known inhibitors and will be available to function in this capacity for extended periods of time.
The TM analogs described herein are particularly advantageous for this purpose as they are closely related to the protein fragment which is derived when full length TM is digested with porcine pancreatic elastase. The long-term stability of immobilized proteins is of paramount importance. Thus, the smaller, proteolytically resistant TM analog is more advantageous than the full length molecule which can be proteolysed by enzymes in the blood, resulting in the potential loss of active component from the biomaterial surface. The TM analogs will be particularly preferable over the use of the full length molecule, inter alia, during periods of physiological stress, e.g., inflammation, where potent white cell proteases, including leukocyte elastase, have access to the biomaterial surface.
The TM analogs may be used to coat polymers used in a wide variety of biological applications including, but not limited to, arteriovenous shunts, intravascular shunts (eg., umbilical, angiographic), vascular grafts, heart valves, artificial joints, pacemakers, left ventricle assist devices, and the like.
The TM analogs are bonded to a biocompatible polymer. Biocompatible polymers may be any suitable polymeric biomaterial or combination thereof known and used in the art for biological application such as ~O90/10081 PCT/US90/00955 4l polyurethanes, silicone elastomers, hydrogels (e.g., poly(hydroxyethyl methacrylate), polyesters, polyethers, polyvinyl alcohol, and the like.
The TM analog may be bonded to coat the polymer material following activation of the biopolymer. Activation ~ethods are known in the art and may utilize amino, carboxyl, hydroxyl or sulfhydryl functions cn the compound to be coated. Activation may be achieved through a variety of known ~ono- and/or bi-functional reagents, including, but not limited to, glutaraldehyde, carbodiimide activated COOH, isocyanate, cyanuric acid, or hydrosuccinimide esters.
Spacer arms, known in the art, may optionally be used.
The single amine function at the N-terminus of some of the TM analogs disclosed herein, such as ll/2, 4t/227-462 and 6h/227-462 provide a hishly selective means to covalently couple this protein in a highly oriented array to activated biomaterial surfaces. Once the biocompatible polymer has been coated, it may be implanted in a ma~mal as necessary according to the teaching in the art for the procedure at hand or used in any device that contacts blood where the blood must remain anticoagulated.
The following examples are offered by way of illustration and not by way of limitation.

EXAMPLES
In all discussions of sequences the convention in the art is that the terms 5' and 3' are used to indicate direction on the coding strand of a gene and its surrounding sequences. Unless otherwise stated, all single stranded DNA illustrated in this document represents the coding strand of the gene.

WO90/10081 PCT/US90/00~5 Example 1. Derivation of Oliqonucleotides Encodina Thrombomodulin Anal QgS

a. Isolation of genes encoding amino acids 227-462.
Human DNA was used to isolate a gene encoding the 6 EGF-like domains of thrombomodulin corresponding to amino acids 227-462. This bNA was isolated from fetal liver according to the method of Blin, N and D.
W. Stafford, (1976) ~ucleic Acids Res. 3:2303. The DNA
was then used as a template in a polymerase chain reaction with synthetically derived primers selected to embrace the desired regions tsee Drawings 2 and 3, and Tables 1 and 3) .
The following steps provide a means to obtain 15 a DNA insert encoding amino acids (aa) 227-462 and uses primers #1033 and #1034 (see Drawing 3). It is understood that by modifying the procedures set forth below by using alternative primers, other TM analogs may be obtained.
The sequence of the #1033 and #1034 primers correspond to the 5' and 3' ends of the desired domain;
but they have been modified so that they contain a BamHI site. A termination codon (TGA) was introduced following base 1586. The polymerase chain reaction was run under the conditions described by Saiki, et al., (1988) Science, 320:1350~13S4 except that the initial temperature of annealing was 37'C. After 10 cycles, the annealing temperature was raised to 45-C for the remaining 30 cycles. An aliquot of the reaction products was separated on a 5% polyacrylamide gel and visualized by ethidium bromide staining. A band of the predicted size (700 bp) could clearly be seen. To confirm the identity of the insert, one can optionally sequence this band or hybridize it to an internal probe.

j ~ ~ ,~

b. Isolation of genes encoding other regions of TM.
The polymerase chain reaction as herein described was used in the same manner to isolate additional fragments of thrombomodulin corresponding to the regions listed in Table 1 and shown schematically in Drawing 2A. In particular, these regions embrace one or more of the EGF-like domains and the O-linked glycosylation dormain. The sequences of the primers selected to produce the desired regions are shown in Table 3.

c. Cloning plasmids cotaining the soluble TM
analog genes.
i. pUCl9pcrTM7 The remainder of the polymerase chain reaction mixture described in part (a.) above was restricted with BamHI, separated on a 5% polyacrylamide gel, and the 700 bp band was excised and eluted. It was ligated to pUCl9 that had been restricted with BamHI and the new plasmid was transformed into E. coli strain DH5~. White colonies that grew on a medium containing ampicillin and 5-bromo-4-chloro-3-indolyl-~-D-galactoside were picked onto a grid and hybridized by the Grunstein~Hogness technique with a synthetically derived gene corresponding to aa 283-352 of TM that had been cut out of a cloning plasmid (pTM2.l) with EcoRI
and HindIII before labelling with 32p by random priming (Boehringer Mannheim).
After exposing the filters to X-ray film the one colony that hybridized to the pTM2.l probe (pUCl9pcrTM7) was grown up. DNA was extracted and was analyzed by restriction with either BamHI or ~glII.
The restriction digests were separated on an agarose gel and the ethidium bromide stained bands confirmed that there was a 700 bp insert and that it had the correct restriction ~ap for the 6 EGF-like domains o~

W390/1~81 PCT/~IS90/~95~
~. .: ,~ .,j thrombomodulin. This gel was also transferred to nitrocellulose and analyzed by hybridization with the 32p labelled insert from pTM2.l. The 700 bp insert released after digestion with BamHI hybridized to this probe as did the 486 bp BglII fragment proving that indeed this clone contained the 6 EGF-like domains of thrombomodulin. Verification that no mutations had been introduced into pUCl9pcrTM7 during the poly~erase chain reaction and cloning procedures was obtained by sequencing the insert. (see Drawing 4) ii. Cloning plasmids containing other ~M
analog genes Other cloning plasmids, such as pTM309 and pTM323 were constructed using methods simil~r to those described in (i.) Plasmid pTM309 contains amino acids 350-462 of native TM (EGF-like domains 4,5,& 6) and pTM323 contains amino acids 227-497 (EGF-like domains 1-6 + O-linked glycosylation domain).
Additional plasmids were constructed that contain other TM analog gene sequences. (See Drawing 2A and Table l).

Example 2. Production of Human TM Analoas in Mammalian Cells.

a. Mammalian expression vectors for TM analogs ~his example provides a mammalian expression vector comprising the analog genes of Example l.
Drawing 4 provides an overview of the construction of one of these vectors. ~he genes are operably linked to the signal sequence of human tissue plasminogen activator. The expression plasmid, pPAl24, contains a promoter contained within the three copies of the long terminal repeats derived from Harvey Sarcoma virus for the expression of cloned genes. This plasmid was derived from pPAl19, and pSC672, both described in 35 detail in co-pending USSN #074,083, filed July 16, ~090~1008~ PCT/US90/00955 ,. . ,, , ,~
,~, ~
1987, incorporated herein by reference. A BglII - BclI
fragment containing the SV40 polyadenylation region was isolated from pSC672. This fragment was cloned into pPAll9 which had been digested with BglII and ~clI. In the resulting plasmid, pPA124, both the BglII and BclI
sites remained intact. Plasmid pPA124 contains the t-PA signal sequence adjacent to an appropriate restriction site and this signal sequence also contains the second intron of the human t-PA gene.
The gene encoding the soluble TM analog was removed from pUC19pcrTM7 by treatment with Ba~HI and ligated to pPA124 that had been treated with BglII.
Transformants were screened for the presence of the insert in the correct orientation, that is in which the t-PA signal sequence was linked to the 5' end of the throm~omodulin insert encoding an open reading frame.
This plasmid, pTM101, was then digested with ClaI and ligated to a ClaI fragment containing the dhfr gene under the control of the SV40 promoter. The ClaI
fragment is described in W088/02411 at page 26.
Transformants were screened for the presence of this dhfr cassette and then the orientation relative to the plasmid was determined by restriction mapping (pTM103).
Plasmid pT~103, containing the dhfr sequence in the divergent direction to the thrombomodulin sequence, was treated with BclI and a DNA fragment encoding a gene providing hygromycin resistance on a BamHI fragment was ligated into the plasmid. Clones were selected, after transformation into E. coli strain DH5~, by their ability to grow on plates containing both ampicillin and hygromycin B. The orientation of the hygromycin B gene relative to the plasmid was determined by restriction mapping. One plasmid, pTM108, in which the hygromycin B gene lies in the opposite orientation to the TM gene, was grown up in culture. This plasmid has the sequences encoding the TM analog under the control of the triple LTR promoter, WO90/1~81 PCT/US90/00955 with both a gene that confers hygromyc~n-~ resistance and ona that encodes dhfr present on the plas~id. The thrombomodulin sequence was linked to the tissue plasminogen activator signal sequence, ensuring its secretion. (See Table 6) b. Transfection, selection and amplification of stable mammalian clones.
For the transfection, l0 ~g of pTMl08 was mixed with Lipofectin reagent (Bethesda Research Laboratories) and added to a monolayer of 105 CHL-l host cells in 6-well plates. Forty-eight hours after transfection, a known number of cells were plated onto selective media. Resistance to hygromycin B was used as the selection marker. CHL-l cells transfected with the bacterial hygromycin B gene can survive growth in 0.3 mg/ml hygromycin B.
The transfection or selection frequency was 2/103 and was determined as the number of colonies arising after selection, divided by the total number of cells plated. The culture supernatant was shown to contain l.5 U/ml TM activity after 24 hours in contact with the cells.
A population of cells resistant to the first selection conditions were then subjected to a second 25 round of selective pressure. Either l00nM or 500nM
methotrexate (MTX) was added to the growth medium to select for transfectants that expressed the dhfr gene.
Only clones which had amplified the dhfr gene would be able to grow in this high level of MTX. In the process of gene amplification, other plasmid sequences will be co-amplified with the dhfr gene and thus lead to increased gene expression of the non-selectable gene as well. Resistant clones were apparent after 5 to 6 weeks. Individual clones resistant to these levels of MTX were isolated and assayed. A culture after selection in l00nM MTX was shown to produce 4.9-14.7 U

47 ~' ^~ ;
per ml of protein C activating activity (see below). A
pooled population was plated into a ten-fold greater concentration of MTX (l~M or 5~M). Clones were again recovered from this selection step and assayed. At each step clones were shown to produce and secrete TM
analog into the culture medium.

Example 3. Transient Expression of TM Analoas in Mammalian Cells a. Construction of Transient Mammalian Expression Vectors.
Two vectors suitable for use in transient expression experiments were constructed from commercially available plasmids.
i. pPAl33 Plasmid pPAl33 is a modification of plasmid pLl (Okayma, H. and Berg P., (1983) Mol. Cell Biol.
3:280-289) purchased from Pharmacia LRB Biochemical (Piscataway, NJ). This plasmid, PLl, carries the 5V40 origin of replication the early region transcription ZO promoter and mRNA splicing sequences. It was used to construct a plasmid containing a polylinker sequence that adds a unique series of restriction sites, the t-PA signal sequence under the control o~ the SV40 promoter and the SV40 origin or replication. The mRNA
splicing sites were removed. Construction of this plasmid, pPAl29, is described in detail in co-pending, commonly assigned patent application USSN 345,372, incorporated herein by reference. To create pPAl33, pPAl29 was digested with BglII and another polylinker sequence was inserted immediately downstream from the t-PA signal seguence. This polylinker creates a plasmid with a unique series of restriction sites (BglII, NotI, SnaBI and SplI).

~090/1~81 PCT/US90/~

ii. pTHR5 Plasmid pTHR5 is a modification of plasmid pCDM8 purchased from Invitrogen ~San Diego, CA) and carries the cytomegalovirus immediate early promoter.
Plasmid pPAl33 was digested with XhoI and NotI to release the t-PA signal sequence which was isolated and cloned into pCDM8 that had been digested with XhoI and NotI so that the t-PA signal sequence is under the control of the cytomegalovirus promoter. (See Drawing 5).

b. Transient Expression Plasmids Containing TM
Analog Genes.
The cloning plasmids containing TM analog gene sequences described in Example l were digested with BamHI and/or BglII and/or NotI. ~he TM gene sequences were isolated and ligated into either pPAl33 or pTHR5 immediately adjacent to the t-PA signal sequence to insure that the transiently expressed peptides would be secreted into the cell culture medium. A model plasmid, pTHR13, is shown in Drawing 5. Table 2 lists other TM analog gene containing plasmids and their parent plasmids. Plasmid pTHRl3 was made by digesting cloning plasmid pTM301 with BamHI and BglII- and ligating the TM analog gene fragment into the BglII site of pTHR5. pTHR13 contains the gene sequence coding for the 6 EGF-like domains of thrombomodulin operably linked to the t-PA signal sequence.
These plasmids were transfected into SV40 transformed African Green Monkey Xidney Cells (COS-l) by the Lipofectin technique. In controls, COS-l cells were either transfected with the parental plasmids or mock transfected with phosphate buffered saline (PBS).
Protein C activation assays were done using conditioned cell culture medium to determine the presence of TM
~5 analog. Table 7 depicts a list of transient expression plasmids.

WOg0/10081 PCT/US90/00~
.J

Example 4. Production of human thrombomodulin analoas in insect cells $nfected wit~ a baculovirus ex~ression vector.
Autographa çalifor~ia nuclear polyhedrosis virus (AcNPV) was used as an expression vector for human thrombomodulin (TM). By using specifically constructed plasmids, the protein coding sequence for selected portions of human thrombomodulin were linked to the signal sequence of hypodermin A. The hybrid hypodermin A-TM analog genes were inserted downstream of the polyhedrin transcription signal, and the TM-polyhedrin hybrid genes were transferred by ho~ologous recombination to an infectious AcNPV expression vector.
In this virus the wild type AcNPV polyhedrin gene has been replaced with the TM analog gene in the recombinant virus. S~odo~tera frugiRerda (sf9) cells were then infected for human TM analog production.

a. Construction of AcNPV Transfer Vectors i. Vectors with the Hypodermin A signal sequence: pHYl and pSC716.
Two oligomers, COD#1198 and COD#ll99 were synthesized, see Table 6. These oligomers contain the Hypodermin A signal sequence, a translation initiation codon, a BglII cloning site, a BamHI 5' overhand and a Kpnl 3' overhang. COD#1198 and COD#ll99 were annealed and cloned into pSC654, a pUCl9 derivative, creating pHYl. See Drawing 6.
Plasmid pHYl was restricted with BamHI and EcoRI, releasing the hypodermin A signal sequence.
This sequence was then ligated to pSC714 to create the vector pSC716. Plasmid pSC714 is a derivative of pVL1393, obtained from Summers, et al. The only difference between the two is that in pSC714, the BglII
si~e has ~een destroyed.

~VO90/10081 PCT/US90/00955 .. ,~ , , , ~ ,~
;
ii. Construction of a vector in which the hypodermin A signal sequence is linked to a gene encoding TM amino acids 227-462.
The BamHI fragment from pUCl9pcrTM7 was cloned into the BglII site of pHYl and the orientation was chosen such tha~ the hypodermin A signal seouence was adjacent to amino acid 227. This plasmid is ~HYlOl.
iii. Construction of the AcNPV transfer vector: pTMHYlOl.
Plasmid pHYlOl was treated with BamHI/EcoRI
which releases the Hypodermin A signal sequence linked to the TM analog coding seouence. Shuttle vector pVLl393 contains a partially deleted AcNPV polyhedrin gene and unique BamHI and EcoRI cloning sites. The BamHI/EcoRI fragment from pHYlOl was inserted downstream of the polyhedrin promoter, thus creating a plasmid, pTMHYlOl, in which the hybrid gene was under the control of the polyhedrin promoter. This plasmid is shown in Drawing 6.
iv. Construction of other AcNPV transfer vectors Transfer plasmids containing other TM analog gene sequences were constructed using a strategy similar to that outlined above. Fragments from the cloning plasmids described in Example l above were cloned into pSC716 in frame so that the TM analog gene sequence was fused to the hypodermin A signal sequence.
See Table 8 for a list of transfer vectors and the gene sequences contained therein.

b. Construction of AcNPV Transfer Vectors pTMHYl02 and pTMHYl03.
A new signal sequence is synthesi2ed that contains a BalI cloning site, a Ba~HI S' overhang and a KpnI 3' overhang. The synthetic fragments are annealed W~90/10081 PCT/US90/0095i s~ ~, , .

and cloned into pSC654 as described above creating pHY2.
New genes consisting of human analog ll/6 or human analog ll/2 are isolated by using hu~an DNA in the polymerase chain reaction as described in Exampl~
l. The primers introduce a BalI site at the 5' end and a termination codon and an EcoRI site at the 3' end (see Table 3). After treatment with BalI and EcoRI, the genes are cloned into pHY2, which has also been treated with BalI and EcoRI. The EcoRI site is a unique site in the original pSC654 vector used to make pHY2. The new plasmids pHYl02 and pHYl03 contain the signal sequence of hypodermin A fused to genes that encode analog ll/6 and analog ll/2, respectively.
Plasmids pHYlO2 and pHYl03 are treated with BamHI and EcoRI and the fragment released is cloned into pVLl393 at its unique BamHI and EcoRI sites creating plasmids pTMXYl02 and pTMHYl03 in which the hybrid genes are under the control of the polyhedrin promoter.

c. Transferring Genes into the AcNPV Genome.
Plasmids pTMHYlOl, pTHR22, etc~ containing a gene encoding the Hypodèrmin A signal sequence and a TM
analog coding sequence, were transferred to the AcNPV
genome by in vivo recombination. A calcium phosphate precipitation technique modified for insect cells was used according to Summers and Smith. Briefly, the individual transfer vectors and AcNPV DNA were cotransfected into sf9 cells basically as follows: a T25 flask was seeded with 2xlO6 cells. The cells were allowed to attach for one Xour at room temperature.
Then one microgram of transfer ~ector and l ~g AcNPV
DNA coprecipitated in calcium phosphate were incubated with SF9 cells for 4 hrs. TMN-FH media supplemented 3S with 10% FBS was then replenished. A stock virus was created.

W~90/1~81 PCT/US90/00~
, The genes contained on plasmids pTMHY102 and pTMHY103 are transferred into the AcNPV genome in the same manner as described above.

d. Detection and Purification of Recombinant Viruses.
Plaque assays were performed to detect recombinant viruses. The transfection stocks were diluted to 10-4, 10-5, 10-6 on day 7 post transfection.
A control vector in which ~-galactosidase is under the control of the polyhedrin promoter was also included in the experiment. The results showed that the control vector assayed with X-gal gave an average of 2%
recombinant blue plaques. There were approximately 2%
occlusion negative recombinant TM plaques observed.
Occlusion negative plaques from the cells transfected with the TM transfer vectors were pic~ed and replated at lo~l, lo~2 and 10-3 dilution of plaque assay. on day 7, the plates showed 100% pure occlusion negative recombinant plaques. A single colony for each of the desired TM analogs, (6h/227-462, 6h/227-462:227-462, 6h/350-462, 6h/227-497, etc.) was selected for further work.
.

e. Production of Soluble Thromoomodulin Analog.
i. Production of TM activity by the cotransfection stocks.
2x~06 sf9 cells were infected with 1 ml of the transfection stock as described above. on day 3 the cells were pelleted and resuspended in serum free Excell 400 medium (JR Scientific) for another three days. Supernatant was collected and, when assayed, showed the presence of TM activity.

~'O90/tO~81 PCT/US90/009~
"

ii. Growth of viral stocks.
T25 flasks were seeded at a density of 1.SX1O6 sf9 cells per 5 ml. An isolated recombinant plaque was picked (see above) and used to infect the sf9 cells in TMN-FH media supplemented with 10S FBS for viral stocks. On day 3, supernatants from the cultures were collected and were shown to contain TM a~tivity.
iii. Production of TM activity by recombinant AcNPV 6h-227/462 Five 30 ml haker fla~ks were 6eeded with 2x107 cells and infected with 2 ml of the above-prepared recombinant viral stoc~. A spinner flask was also infected with one of the virus stocks prepared above.
On day 4 the supernatan~ was collected for assay. The results showed that insect cells infected with recombinants AcNPV 6h-2~7/462 (and the others) were secreting thrombomodulin activity.
iv. Production of TM activity by other recombinants The methods described above for cotransfection and production of viral stocks were used to produce infected insect cells which secrete other TM analogs including 6h/(227-462)2, 6h/350-462, 6h/227-497, etc.

Example 5. TM ~nalo~: t-PA Chimaeric Proteins 2S The t-PA gene is used for the assembly of genes encoding chimaeric fusions of a fibrinolytic enzyme with thrombomodulin polypeptide ~equences. The t-PA gene used for this purpose was isolated from plasmid pPA003, which is on deposit with the ATCC, 30 accession No. 67293. Plasmid pPA003 is described in detail in co-pending, commonly assigned, patent application USSN 003,611 ~nd is incorporated herein by reference. Tables 7 and 8 depict a list of mammalian expression plasmids and baculovirus transfer plasmids containing thro~bomodulin analog genes fused with the t-PA gene.

~090/10081 PCT/US90/00955 a. TM Heterologs Synthesized in Mammalian Cells.
i. TM Analogs fused to the N-terminus of t-PA
The 3' ends of the TM nnalog genes described in the previous examples all have termination codons which must ~e renoved in order to create chimaeric proteins containing thrombomodulin gene sequences fused to the N-terminus of t-PA. ~rimer~ for use in a ~CR
were synthesized so that when incorporated into the DNA
sequences of the TM analogs the termination codons at the 3' ends were deleted and a BglII ite was introduced. The 5' ends of the ~M genes remained unchanged. These primers were used in a ~eries of PCRs that synthesized variant fragments of the TM analcg genes described in Example l. These new TM analog genes were ligated into either of the two transient expression vectors described in Example 3, pPAl33 or pTHR5, so that the TM analog gene was fused to the t-PA
signal sequence and followed by a BglII site. These plasmids were treated with BglII and a gene encoding amino acids 4 to 530 of mature t-PA was inserted. This fragment was generated by PCR using the primers shown in table 3. The plasmids that result contain a gene that encodes a fusion protein consisting of a signal sequence from t-PA followed by the TM analog which is fused to mature t-PA under control of either the SV40 promoter or the cytomegalovirus promoter. See Drawing 7 .
ii~ TM Analogs fused to the C-terminus of ~-PA
The termination codon ~t the 31 end of the t-PA gene was deleted using the ~me ~ethod described above. Two primers were ~yn~ha ized for the PCR that $ntroduced a new BglII site immediately 3' to the proline at amino acid 530, which also deletes the termination codon, and cre~tes a BamHI site at amino acid 4 in place of EglII site found in the native t-PA
.

WO90/1~81 PCT/US90/00 5S ~ ~
f, gene sequence. The t-PA gene used A5 a template for the PCR was isolated from plasmid pPA509. This plasmid is on deposit with the ATCC, accession #67443, and is described in detail in co-pending, commonly assigned, patent application USSN 074,083, incorporated herein by reference. The new t-PA gene frag~ent was inserted into the BglII ~ite of p~nx5 so that the t-PA gene is fused to the t-PA ~ignal sequence, creating plasmid pTHRl6. See Drawing 7.
Thrombomodulin ~nAlog gene ~eguences, digested with ~glI~ and/or Bam HI were inserted into pTHR16 at the BglII site creating plasmids that encode soluble secreted fusion proteins whose N-terminus is the natural N-terminus of t-PA fused at the natural C-terminus of t-PA to the N-terminus of the TM analog gene. Schematic examples of these fusion peptides are shown in Drawing 2B. The DNA and amino acid sequence of t-PA fused with the 6 EGF-like domains of TM is shown in Table 9.

b. TM Heterologs Synthesized in Insect Cells TM heterolog gene sequences for expression in insect cells were constructed by methods similar to those described above. For example, a plasmid for producing one TM heterolog in which the 6 EGFs were operably linked to the N-terminus of t-PA was constructed by digesting plasmid pTHR6 with NheI and NotI and inserting the gene fragment into pTHRlO after digestion with the ~ame e~zymes. pTHR6 is a transient expression vector containing the 6 EGF5 fused to t-PA
for expression in mam~alian cells. pTHRlO is a ba~ulovirus transfer vector containing gene sequences for the 6 EGF-like domains. The resulting plasmid, pTHR25, contains gene ~equences for the 6 EGF-like domains and t-PA in a ~aculovirus transfer vector.
In another example where the 6 EGF-like domains were fused to the C-terminus of t-PA, a SmaI-U090/10081 PCT/US90/~

s6 ~ :
~;. . , . .. - ,i NotI fragment removed from pTHRl7 and containing the 6 EGF domains was ligated into the SmaI-NotI sites of pTXR12.

Example 6. ~ifieation of TM AnalPs 6h~227-462 TM analog 6h/227-462 was purified from conditioned media from cells infected with AcNPV-6h/227-462 by removal of cell de~ris, followed by five chromatography ~teps: 1) Q Sepharose, 2) thrombin affinity, 3) gel filtration, 4) anion exchange, and 5) a secor.d gel filtration step. Substantially pure 6h/227-462 (>95%) is obtained after the second chromatography step (throm~in affinity), hswever, when analyzed on SDS-PAGE gels, multiple active forms of 15 analog 6h 227-462 were o~served. These variant forms were separated from each other using anion exchange with Mono Q resin. The gel filtration steps effect an exchange of buffers. All chromatography steps were performed at 4- C.

a. Materials Some of the chromatographic resins were purchased from commerical sources. Q Sepharose and Sephadex G25 were purchased from Sigma (St. Louis, M0), and Mono Q 5/STM from Pharmacia L~B ~Piscataway, NJ).
DFP-thrombin agarose was prepared approximately as follows: 360 mg of bovine thrombin in 100 ml of 20 ~M Na phosphate, pH 7.5 was added to approximately lO0 ml of a 50% Affigel lO resin ~lurry and mixed overnight at 4-C. The Affigel lO was prepared for use as described by the manuacturer and equilibrated with the load buffer. Residual active esters were blocked by the addition of lO0 ml o~ O.lM
glycine (pH 5.6) for one hour at 4-C. The gel was then equilibrated with 30 mM Tris-HCl, 2M NaCl, pH 7~5, and 20 ~l of DFP was added to give a final concentration of about lmM DFP. After 16 hrs of mixing at 4-C an W`~90/10081 PCT/US90/~5 57 , j additional 6 ~1 of DFP was added and mixing continued for 4 additional hours. The resin was then washed with 20 mM Tris-HCl, 2 M NaCl p~ 7.5 and stored at 4-C.
Thrombin activity was mea~ured using the Ka~i S-2238 substrate and indicated that >86% of the thrombin was removed from the solution, and presumably coupled to the resin, giving a final concentration of about 6 mg of thrombin per ml of resin. The enzymatic activity of the DFP treated resin ~as ~1% of the starting activity.

b. Production of pure 6h/227-462.
Conditioned media was harvested and clarified by cross-flow filtration on a Microgon 0.3 ft2 hollow fiber filter unit at about pH 6.0, then diluted by 50%
with ultra-pure water. The p~ was adjusted from about 6.0 to about pH 5.2 with glacial acetic acid. The adjusted media was then loaded onto a colu~n of Q
- Sepharose resin at a flow rate of about 110-150 ml/min.
The co~umn had previously been equilibrated with about

4 column volumes of wash buffer 1 (117 mM Na acetate, 0.02% NaN3 pH 5.0). After loading, the column was washed with wash buffer 1 followed by wash buffer 2 (25mM Na acetate, O.lMNaCl, p~5.0) then the TM analog was eluted with wash buffer 2 containing 0.3 M NaCl, pH
5Ø This column matrix can be reused by stripping with wash buffer 2 + 2.0 M NaCl.
Column fractions containing activity as measured in the Protein C activation ssay (see Example 11) were pooled and diluted 30% with of 0.3 M NaCl, 20 mM Tris-HCl, 0.5 mM CaCl2, 0.02% NaN3, p~ 7.5. The pH
of the diluate was measured and adjusted to about 7.5 with NaOH. The ionic strength of the pool was about the ionic strength of a solution of 0.3 M NaCl. This adjusted pool was loaded overnight by gravity onto a thrombin agarose column pre-equil$brated with the same buffer used to dilute the Q Sepharo~e pool. The column W~ ~/10081 PCT/US90/00955 58 ,~ ~
t` _ .,' J
was washed with diluent buffer, and the TM analog was removed from the ~atrix with 1.5 M GuHCl, 2.0 M NaCl, 20 mM Tris HCl, 1 ~M Na EDTA, 0.02% NaN3, pH 7.5.

c. Evaluation of Purity SDS Polyacrylamite Gel Electrophoresis was performed by the method of ~aemmli using 3.3~
acrylamide in the stacking and 12.5S acrylamide in the running gel. Nonreduced zamples were diluted in Laemmli sample solubilization ~uffer (50 mM Tris-HCl, pH 6.8, 25% glycerol, 2% SDS, and .01% bromphenol blue) and loaded directly onto the gel. Reduced samples were incubated for 7 minutes at lOO-C in sample solubilization buffer containing 10 ~M dithiothreitol.
Following incu~ation, iodoacetamide was added to a final concentration of 50 rM, and the sample incubated an additional 5 minutes at room temperature, bef ore loading on the gel. Pharmacia LMW Calibration Rit protein standards were used for MW markers, and the gels were silver stained.
Under reduced-alkylated conditions, a single broad band was seen at about 53 kDa, while under non-reducing conditions, a doublet was seen at about 31 and 36 kDa, as well as low levels of a higher MW ~aterial (having an apparent molecular weight of about 52 kDa).
Both bands in the doublet were active as shown by elution of the bands from an SDS gel, followed by a Protein C activation assay. The 52 kDa material seen in non-reduced gels, which was estimated at 10% of the total protein by gel scanning, also appeared to be derived from active thrombomodulin as shown by the activity of the eluted high bands in the Protein C
activation assay. Thus all material visualized on the non-reduced gel had TM activity. The lower MW bands seen in non-reduced gels were the predominant form or species of the isolated protein. The preferred species of peptide is one having a higher TM activity being W~90/1~81 PCT/~S90/00955 5g ~
1: , ,. ' , J
free of other less active thro~bomodulin-like peptides.
This preferred ~pecies corresponds to the slower migrating of the two lower MW bands. Such peptide species were purified below to provide under non-S reducing chromatography a single ~pecies having theexpected ~pparent molecular weight predicted for the particular peptide.
The TM analog 6h/227-462 ~ene has two potential N-linked glycosylation ~ites ~t ~mino acids 364 and 391 and ~wo potential O-linked glycosylation sites at amino acids 319 and 393. The thrombin affinity purified product was shown to be glycosylated by treatment with N-glycanase which resulted in an apparent decrease in MW of about 2000 daltons.

d. Amino Acid Composition and Seguencing.
i. Sample Preparation.
Since the expected product has 36 Cys residues in 18 disulfide bridges, ~amples were reduced and alkylated prior to N-terminal analysis. Guanidine in a Tris buffer was added to 1.5 ml of product (about 3435 pmoles, 90 ~g) to give a final concentration of 4M
guanidine, 0.1 M Tris-HCl, 1 mM EDTA, pH 8.5 in a volume of 3.0 ml. The resultant solution was reduced with a 43 fold molar excess of DTT over disulfide (2.67 ~moles, 0.89 ~M final concentration) for 3 hr at 37-C, then alkylated for 60 minutes at 37-C by the addition of a 2.6 fold molar excess of iodoacetic ~cid over DTT
(6.94 ~moles, 2.3 ~M final concentration). The sample was purified using a Waters Delta PakC-4 reverse phase H~LC with a linear gradient of water + 0.1% TFA to 75%
CH3CN + 0.1% TFA. Only a single protein peak was recovered following the reduction and alkylation protocol. Aliquots were dried down in a Savant evaporator for 6equencing and amino acid analysis.
ii. Amino Acid Analysis.

i The amino acid composition was determined as described above in the determination of specific activity starting at the hydrolysis step. The ami~o acid composition found for the protein aqrees well with what was expected (see Table 4). The exceptions are tryptophane and cysteine, whic~ are not accurately measured by the methods used here.
iii. Sequencing.
Amino acid sequence analysis was performed on an Applied Biosystems Protein/P~ptide 6equencer (Model 4?7 connected to a sooA data module or a Model #470A
connected to a chart recorder). The phenylthiohydantoin (PTH) amino acids were identified on-line with a 120A
Applied Biosystems PTH Analyzer by reverse phase HPLC
using a Brownlee PTM-C18 cartridge (2.1 x 222 mm).
The reduced-alkylated sample purified as described above gave the following sequence:
Ala-Val-Val- Pro-Arg(5)-ser-Cys-Ser-Val-Glu(10)-Asn-Gly-Gly-Cys/Glu-Glu(15)-His-Ala-Cys/Glu-Asn-Ala(20)-lle-Pro-Gly-Ala-Pro(25)-Arg-Cys/Glu-Gln-xxx-Pro(30)-Ala-Gly-Ala- m -Leu(35)-Gln-Ala-Asp-Glu~Gly.
sequence found gives an excellent fit with that expected for the cloned material. In several positions (i.e., 14,18,27), it was not unexpectedly impossible to distinguish between an assignment of Cys or Glu, while in position 29, identified as xxx, a clear signal for Cys was not evident. Also, due to the sequence Ala-Ala (33-34), the second Ala (yyy) could not be unequivocally assigned. Otherwise, all residues were as predicted from the DNA ~equence.

e. Separation of variant TM analog species.
i. ~uffer exchange The substantially pure, aetive TM analog 6h/227-462 was applied to a Sephadex G25 column and recovered in 0.2S N-ethylmorpholine acetate (NEM) pH
7Ø This step removes GuHCl and NaCl.

'.~O90/10081 PCT/US90/00955 61 t ~ V
ii. Anion Exchange Chromatography As noted above, more than one form of TM
analog is detected in throm~in-affinity purified material run on an SDS-PAGE gel under non-reducing conditions. A method was developed to resolve these variants. TM analog 6h/227-462 collected from the Sephadex G25 column was applied to a Mono Q column (Pharmacia, 10 mioron particles, quarternary amine) pre-equilibrated with 0.2% N-ethylmorpholine (NEM).
pH7Ø After washing with this buffer the various forms were separated using a gradient of 0 to 0.4 M
NaCl. The elution and activity profiles are shown in Drawing 9. Samples of each fraction were evaluated on an SDS-PAGE gel under non-reducing conditions. Three distinct bands with slightly different mobilities could be seen on the stained gel. Fractions containing peptides with liXe mobilities were pooled (A = fxns 32-35, B = fxns 40-44, C = fxns 70-71) a~d then assayed for total protein content and for activity in the Protein C activation assay. The specific aotivities are listed in the table below. No inactive peptides were detectable in any of the fractions.
Test MaterialSDecifio Activity rU/ma~
Mono Q Load166,000 + 12,000 2S Sxns 32 - 35416,000 + 19,000 fxns 40 - 44262,000 + 4,000 fxns 70 - 7167,600 + 5,000 Example 10. ~ ça~ion of TM Analo~ 4t~227-462 Conditioned cell culture ~edia was collected and the ionic strength adjusted to be equivalent to about the ionic ~trength of 0.3 M NaCl using 1 M NaCl.
The adjusted ~edia was loaded directly onto a thrombin affinity column that had previously been washed with 0.3 M NaCl, 20 mM Tris HCl, 0.5 mM CaCl2 and 0.02% NaN3, pH7.5 (wash buffer). After loading, the column was washed with wash buffer, then the TM analog was removed WO90/1~81 PCT/US90/009 from the column matrix with 1.5 M GuHCl, 2.0 M ~aCl, 20 mM Tris, 1 mM Na EDTA and 0.02~ NaN3. A peak containing TM activity was collected And applied to a Sephadex G25 column and washed throu~h with 0.2% NEM to effect a S buffer exchanqe (removal of NaCl and GuHCl). The active fractions were pooled and applied to a colu~n of Mono Q 5/5 (Pharmacia, LXB) in 0.2% NEM, pH 7Ø The TM analog was selectively eluted with a gradient of 0 to 0.6 M NaCl in 0.2% NEM pH7Ø Fractions containing TM activity were pooled and ~amples run on SDS-PAGE
gels under ~oth reducing and non-reducing conditions.
~he TM analog appeared to be about 90-95~ pure. The remaining high molecular weight contaminatinq material was removed by molecular exclusion c~romatography. The activity containing fractions from the Mono Q column were lyophilized to dryness and resuspended in (380 ul) of 1.0~ NEM, pH 7.O then applied to a Superose 12TU
(Pharmacia, Piscataway, N~) column at 0.5 ml/min. The active material Gollected from this chromatography step was estimated to be 99% pure TM analog by silver stained SDS-PAGE gel analysis.

Example 7. Assavs for Thrombomodulin ~naloas.
a. Materials Rabbit thrombomodulin, hirudin and human Protein C were supplied by American Diagnostica. Human thrombin is available from a variety of noncommercial and commercial sources. Bovine thrombin was purchased from Miles Labs, Dallas, Texas. D-valyl-L-leucyl-L-arginine-p-nitroanilide (S-2266) and D-Phe-Pip-Arg-p-nitroanilide (S-2238) were purchased from Kabi Diagnostica.
Bovine serum albumin (fraction V), citrated human plasma, and APTT reagent were purchased from Sigma Chemicals. Microtiter plates were supplied by 35 Corning (#25861-96). All other reagents were of the highest grade available.

~090/10081 PCTIUS90!009~5 63 ~ -b. Methods and ~esults.
i. Protein C Activation Assay (Chromogenic) This assay was performed by mixing 20 ~l each of the following proteins in a microtiter plate:
thrombomodulin sample (unknown or standard), thrombin (3 nM), and Protein C (1.5 ~). The assay diluent for each protein was 20 ~M Tris-HCl, O.l M NaCl, 2.5 ~M
CaCl2, 5 mg/ml BSA, pH 7.4. The wells were incubated for 2 hours at 37-C, after which Protein C activation was terminated by the addition of 20 ~l of ~irudin (0.16 unit~l, 370 nM) in assay diluent and incubation for an additional lO minutes.
The amount of activated Protein C formed was detected by adding lO0 ~l of l.0 mM S-2266 (in assay diluent), and continuing to incubate the plate at 37 C.
The a~sorbance at 405nm in each well was read every lO
seconds for 30 minutes, using a ~olecular Devices plate reader. The absorbance data was stored, and the change in absorbance per second (slope) in each well was calculated. The change in absorbance per second is proportional to pmole/ml of activated Protein C.
This ratio was determined empirically using varying concentrations of totally activated Protein C. Samples containing lO0~ activated Protein C were generated by mixing Protein C at 0 to l.5 ~M with 60 nM rabbit TM
and 30 nM thrombin, incubating for 0 to 4 hours, adding hirudin and measuring S2266 activity as above.
Conditions under which 100% of the Protein C was activated were defined as those in which the S2266 activity (A405/sec~ reached a plateau.
A unit of activity is defined ~s l pmole of activated Protein C generated per ml/min under the reagent conditions defined above. The response of both purified rabbit TM and conditioned medium containing analog 6h~227-462 in this assay is shown in ~rawing lOA. In Drawing lOB the enhancement of protein C
activation by purified TM analog is depicted.

WO90/10081 PCT/US90/0~5 64 , ~-~ '.''J
In some cases the activity values reported were calculated using rabbit thrombomodulin as a standard. By using amino acid analysis to deduce protein mass, it has been determined that l nmole of TM
analog 6h-227/462 has activity eguivalent to 1 nmole of rabbit thrombomodulin.
Other TM analogs showed Protein C activation enhancing activity including 4t/227-462, 4t/227-462:227-462, PA:227-462(mammalian cell6), 6h/227-462:PA, 6h/227-462:227-462, PA:227-462(insect cells), 6h/350-462, and 6h/227-497.

ii. Determination of Specific Activity of TM
Analog 6h/227-462.
TM analog samples were dialyzed exhaustively lS against 0.2% ethylmorpholine acetate, pH 7.5, and assayed for activity and amino acid composition (no activity was lost due to the dialysis). Samples were prepared for analysis by dialysing samples into 0.2% N-ethylmorpholine-acetate, 0.02% NP-40, pM 7.5 using a 6000 Spectropor membrane and dried down in hydrolysis tubes. Hydrolysis was performed in evacuated tubes in a vapor phase at llO C for 22 hours using 6 N constant boiling HCl containing 1% phenol. The amino acid analysis was perfor~ed after samples were dried down and derivatized with phenylisothiocyanate (PITC). They were analyzed by reversed phase HPLC on a 4~m alkyl binding ~Cl8) silica column with detection at 254 nm.
The chromatogram was developed by a gradient from Eluant A (131 mM NaOAc, pH 6.4, 0.047S TEA, 1 pm EDTA) to Eluant B (60% C~3CN in water). Pierce Protein hydrolysate is used as a standard ~nd norleucine is incorporated as an internal control. Nmoles protein from the amino acid composition were converted to ng using a molecular weight value of 26,200 for the TM
analog 6h-227/462 peptide. Samples had an activity in WO 90/10081 PCI/llS90/00955 r 65 :. .; -. . , the chromogenic as~ay of 0.25 units per ng of protein, or 250,000 units/mg.

iii. Inhibition of the Activated Partial Throm~oplastin Time (APTT).
The for~ation of a clot from citrated plasma is triggered by the ~ddition o~ ~rain cephalin in ellagic acid (nAPTT reagentn), ~nd calcium ion. The time required ~or the clot to form i~ reproducible and increases proportionally with the addition of thrombomodulin. Reagents for the APTT are incubated at 37-C. before mixiny, except for tha citrated plasma, which is kept at 4-C.
The reaction was carried out as follows: lO0 ~1 of Sigma Citrated Plasma was added to a plastic 15 cuvette (Sarstedt #67.742), incubated at 37-C for 1 min; 100 ~l of Sigma APTT reagent was added and the mixture incubated for 2 min at 37-C; 100 ~l of test sample (or control buffer) and 100 ~l 25 mM CaCl2 were added and the cuvette was immediately placed in a Hewlett-Packard 8451A spectrophotometer equipped with a circulating water bath to keap the cuvette at 37 C
during reading. The absorbance due to light scattering at 320 n~ was measured every 0.5 seconds, fro~ 15 to 120 seconds, timed from the addition of CaCl2. A plot of absorbance vs. time yields a sigmoidal curve, with the clotting time defined as the time at which the slope is the ~teepest, corresponding to the infle~tion point of the curve.
A dilution curve was generated for pure analog 6h/227-462 in the AP~T assay (6ee Drawing llA) (~g amounts shown were in a 100 ~l ~ample volume). T~e APTT was prolonged in a dose dependent fashion.
Purified 6h/227-462 was compared to heparin and antithro~bin III in the APT$ ~ssay. As seen in Drawing llB, TM analog 6h/227-462 ~howed a greater ability to prolong the APTT at lower amounts than W090/1~81 PCT/US90/~X~
66 1' ~ ` ?, either heparin or AT III. The guantities of heparin and AT II used in this experiment are not effective doses i~ v~vo suggestin~ that a TM analog will be more effective than either.

iv. Inhibition of thrombin clotting time (TCT) and prothrombin reaction (PT).
Both t~e PT and TCT are determined using the Hewlett-Packard 8452 A diode-arr~y spectrophotometer used 20r the APTT. For the PT reaction, 90 ul of either TM analog 6h~227-462 or PBS was added to 20 ul thromboplastin and 90 ul 25 ~M CaCl2 in a cuvette. The mixture was incubated for 1 minute at 37 C, then 100 ul of citrated plasma was added. After loading the cuvette into the spectrophotometer,the absorbance due to light scattering at 320 nm was measured every 0.5 seconds, from 15 to 120 seconds, timed from the addition of the plasma. A plot of absorbance vs. time yields a eigmoidal curve, with the clotting time defined as the time at which the slope is the steepest, corresponding to the inflection point of the curve.
The TCT is evaluated in the same manner. ~he initial reaction mixture contains 100 ul citrated plasma, 2S ul of 100 mM CaCl2 and 162.S ul of either PBS
or TM analog. After 1 minute, 12.5 ul of thombin is added. The clotting time is measured as described above.
Drawing 12 shows that the TM analog 6h/227-462 prolongs both the PT and the TCT.

v. Direct ~nticoagulant activity - Inhibition of thrombin catalyzed conversion of fibrinogen to fibrin.
Thrombin and varying amounts of TM analog 6h/227-462 were incubated for 2 ~inutes at 37 C in microtitre plate wells. The total initial reaction 3S volume was 50 ul P8S +7.5 mM CaCl2 and 90 nM thrombin.

WO 90/10081 PCI/lJS90/009S5 67 ~ - r -,;
A~ter initial incubation, 100 ul of 3.75 mg/~l human fibrinogen was added per well, and the thrombin induced formation of fibrin was followed by measuring the change in absorbance at 405 nm in a Molecular Devices Vmax sp~ctrophotometer (Molecul~r Devices, Menlo Park, CA). The end-point of the ~88ay was the time at which SOS o~ the final absorbance was r~ached. ~esidual t~rombin activi~y wa~ dotermin~d by ref~rence to a thrombin ~tandard c~rve, which linearly relates the r~ciprocal o~ t~e thrombin conc~ntration to the clotting time. As can be seen in Drawing 13, Tff analog 6h/227-462 inhibited the ability of thrombin to convert fibrinogen to fibrin in a dose dependant manner .

vi. Inhibition of platelet activation and aggregation.
The effects of TM analog 6h/227-462 on thrombin activation of plat~lets was tested by the methods of ~smon, et al, (1983) J. Biol. C~em.
258:12238-12242. When evaluated using this assay, TM
20 analog 6h/227-462 inhibited ~he thrombin mediated activation and aggregation of platelets.
vii. Fibrinolysis - Fibrin plate assay.
Th~ fibrinolytic activity of TM heterologs, fusions of TM analogs and t-PA, was evaluated using zonal clearing on plasminogen-enriched fibrin plates as described by Haverkatet and Brak~an, (1975) Proa. in Chem. Fibrin. ThrQmb. 1:151-159. TM het~rologs PA:227-462(mammalian cells), PA:350-462(mammalian cells), 6h/227-462:PA, and PA:227-462(insoct cell5) all ~xhi~ited ~ibrinolytic activity u~ing this ~ssay.

viii. Inhibition of Factor XIII activation.
TM analog 6h/227-462 inhibited tbe activation of Factor XIII by thro~bin as measured by the method of Polgar, et al., (I987) T~rom~. Haemostas. 58:140.

ix. Additional measures of TM antlthrom~otic activity.
1) TM analog's inhibition of activation of Factor V by thrombin is measured by the method described by Esmon et al., ~. Bio. Chem., (1982), 257:7~44-7947.
2) Inhibition of the TM analog thrombin complex by antithrombin III and heparin cofactor II is measured as described by Jakubow6ki et al., 1986.
3) TM nalog~s inhibition of the inactivation of protein S by thrombin is measured by the method described by Thompson ~ Salem, J. Clin. Invest., (1986), 78(1):13-17.
ix. Additional measures of TM antithro~botic activity.

c. Activity of TM ~nalog In vi~o The in vivo activity of 6h/227-462 was tested in a mouse model of disseminated intravascular coagulation (Xumada et al (1987), El~od 71(3):728-733).
Three ~roups of six mice each were anaesthetised and injected with 37.5 units of thrombin via the tail vein.
In mice receiving no other treatment this caused pulmonary embolism rapidly leading to death. ~wo groups were pretreated by injection with IM analog 25 6h/227-462 (3 min. prior~ at lx or 2x the concentration of thrombin.
Groups treated with 6h/227-462 were protected from lethal effects of thrombin. The survivi~g mice in the control group (thrombin alone) were comato~e for 30 minutes whereas mice in the pretreatment groups appeared nor~al immediately on rec~very from the ~naesthetic.

~090/tO081 PCT/~'S90/0095 69 t , ~ ~
. i, , ~, - ., Example 8. Formulation and Use.

a. Concentration.
Methods were developed to concentrate the TM
analog 6h/227-462 in a variety of ~olutions under pressure without significant losses in activity.
A number of aliquots of purified sample were concentrated with an Amicon 8050 ctirred-cell concentrator equipped with YM-lO mem~rane operated at 40 psi nitr~gen gas ~t 20-C.
i. A 56 ml sample containing ~M analog (774 U/ml) in 20 mM Tris-HCl, 1.5 M Guanidine-~Cl, 2 M NaCl, l mM EDTA, with O.O9S NP-40, pH 7.5 was buffer exchanged 2 times with 15 volumes of O.2% N-ethylmorpholine-acetate (17.4 mM), pH 7.5 containing -15 0.02% azide. The final concentration in 2.3 ml was estimated at (2~,108 U/ml) and the recovery of Protein C activating activity was 99%.
ii. A solution of the TM analog (544 U/ml) in 20 mM Tris-~Cl, l.O M NaCl. O.l mM EDTA, 0.02% NP-40, pH 7.S was concentrated 5-fold without buffer exchange to give a 95% recovery of the original activity.
iii. A lO ml solution of the TM fragment (544 U/ml) in 20 mM Tris-HCl, l.O M NaCl, O.1 mM EDTA, 25 0.02% NP-40, pH 7.5 was buffer exchanged 3 times with 10 volumes of the N-ethylmorpholine buffer to a final concentration of 1,509 U/ml (2.0 ml) with a 60%
recovery of the original activity.
iv. A sample containing 1,960 U/ml TM in 2M
30 NaCl, 1.5 M guanidine-HCl, 1 ~M EDTA, 20 ~M Tris-HCl, p~ 7.5, containing no NP-40, wa~ concentrated lO-fold using an Amicon Centricon lO ~icroconcentrator device.
Recovery of activity was >90%.
In addition, the TM analog can be concentrated by drying and resolubilization as described below.

WO90/10081 PCTI~S90/009 ~ ?

b. Dialysis.
The stabil~ty of t~e TM analog during dialysis was demonstrated.
Dialysis was performed using 12,000-14,000 molecular weight cutoff dialysis tubing at 4-C.
Purified TM analog (1,921 U/ml) in 2 ~ NaCl, 1.5 M
guanidine-HCl, 1 ~M EDTA, 20 mM Tris-HCl, pH 7.5, was extensively dialyzed against 0.2% N-ethyl~orpholine (NEM) (pH 8.8). No loss in actiYity was detected although the protein was diluted by a factor of 1.2 (final concentration was 1,661 U/ml).
A solution containing 450 U/~l of TM analog in 20 mM Tris-HCl, pH 7.5 containing 2 M NaCl, 1.5 M
guanidine, 1.0 ~M ED~A + 0.02% NP-40 was dialyzed against the N-ethylmorpholine acetate bl~ffer at pH 7.5 with no NP-40. Recovery of activity was 85%.

c. Stability during freezing and thawing.
The stability of the TM analog after freezing and thawing under various conditions was evaluated.
A solution of ~M analog in 20 mM Tris-HCl, 1.O M NaCl, 0.1 mM EDTA, 0.02% NP-40, pH 7.5 was used as the starting material. Each of four aliquots were concentrated 5-fold and three were also buffer exchanged 3 x with 10 volumes of the solvents listed below. Each were then put through 5 freeze-thaw cycles, freezing in a dry ice: HeOH bath and thawing at room temperature.
1) 5-fold concentration of initial solution concentration 2) 5-fold concentration after buffer exchange into 0.2 M Tris-HCl, 0.02% azide, p~ 7.5 3) 5-fold concentration after buffer exchange into 0.2 M Tris-HCl, 0.2 octylglucoside, 0.02% azide, pH 7.5 WO 90/10081 PCr/US90/0095!;

7 1 i ~ ; j 4) 5-fold concentration after buffer exchange into 0.2% N-ethylmorpholineacetate (17.4 mM), 0.02 azide, pH 7 ~ 5 ~
All samples showed >95% recovery of acti~ity following the ~reeze-thaw cycles.

d. Drying and Resolubilization.
Four equal aliquots of TM analog in 0.S ml solutions of 0.02% NEM (pH 8.8) were prepared in Eppendorf tubes, and to two of these ~liquots NP-40 was added to a final concentration of 0.02%. The aliquots were taken to dryness in a Savant microconcentrator.
Either 50 ~1 of P9S (phosphate buffered saline) or of 0.2% NEM (p~ 8.8) were used to resuspend each sample overnight at 4 C.
The Protein C activity assay was carried out on all samples. The results are pro~ided in Table 5.
In another experiment, a sample to purified TM analog 6h/227-462 was buffer exchanged on Sephades G25 into PBS then diluted 5-fold with pure water. one ml samples, containing O.22 mg of TM analog protein, were lyophilized, and one was then resuspended in 0.2 ml water, a final concentration of 1.1 mg/ml. Greater that 95~ of the activity was recovered. The remaining lyophilized protein was stored for several week at -70 degrees C with no loss of activity upon resuspension.
These experiments demonstrate that:
a) The ~M analog can be efficiently dialysed from the thrombin eluent buffer into the volatile buffer N-ethylmorpholine at pH 8.8 in the absence of detergent, b) TM analog samples in the NEM buffer can be freeze-dried and stored in the absence of salts, detergents or other sta~ilizers, c) samples of the TM ~nalog can be resolubilized in either NEM or PBS with or without NP-40 resulting in little or no loss of biological activity. Ot~er salts that are of use include hydro~romide, sulfate, acetate, citrate, malate, WO9O/1~81 PCT/US90/0~5 S?.~ r ~

borate, lactate, glycine, glutamate and aspartate, among others, and d) samples can ~e effectively concentrated by this procedure, to at least l.lmg/ml.

Example 9 Expression Qf TM An~lQa Activitv in E. coli.
A prokaryotic expression vector was contructed for the expres~ion of ~ TM ~nalog in E~
çQli. The parent plasmid was pRITS, purchased from Amersham, which contain~ a gene coding for Protein A
under the control of the Protein A promoter and having the Protein A signal 6equence. This plasmid was digested with BamHI ~nd a BamHI fragment from plasmid pUCl9pcrTM7 was inserted. The ~amHI fragment carries the gene coding for the 6 EGF-liXe domains of native thrombomodulin. The resulting plasmid, pTHR8 was transformed into E coli. A culture was grown up and a partial purification of the TM analog protein was done as described by Amers~am using an IgG Sepharose column.
This protein was active in the Protein C activation assay.

WO 90/1~81 PCT/US90/00 73 ;~

Table 1 Name Pri~ers (COD # ~ amino acids

5 ' 3 ' pTM301 1034 1411 aa 227--462 pTM302 1034 1408 aa 227--344 pTM312 1034 1433 aa 227--344 pTM303 1034 1209 aa 227--462 pT~304 1034 1412 aa 227--421 pT~314 1034 1435 aa 227--421 pTM305 1034 1410 ~a 227--386 pTM315 1034 1434 aa 227--386 p~M306 1292 1411 aa 427--462 pTM307 1293 1410 aa 351--386 pTM317 1293 1434. aa 351--386 lS pT~308 1293 1412 aa 351--421 pTM318 1293 1435 aa 351--421 pTM309 1293 1411 aa 351--462 pTM310 1294 1412 aa 390--421 pTM320 1294 143S aa 390--421 pTM311 1294 1411 aa 390--462 pTM322 1034 1481 aa 227--498 pTM323 1293 1481 aa 351--498 WO 90/tO081 PCI/US90/00955 WO 90/10081 PCI`/US90/009~;5 '? ' ~` - .`;
'` .: , O ~ OC~ O C~ ~ O ~ ~ ~ ~
~ .~ ~-~ ~ ~ c n ~ ~ ~
O ~ _ ~ ~ ~_ 2' C) O ~ ~ ~ G~ ~L
C O 5_ ~ :~ C n r ~ .~ G~ ~
~ ~ 1~.1. ~ 1-- S~' ~ (D 1~ ~ 1-- ~ C~ ~
c n ~~ ~ ~ :~ ~ o D 1 0 GOl ~ 88 8 i~ 3 3 ~ ~ i~ ~- 8 8, . ~ ~ ~ ~ ~ ~
C ~ ~ ~ ;~ _ ~ E~ ~ - ~
~ ~ ~ ~ ~ ~ _ ~ ~ ~ ~
~ ~ ~ ~ ~' ~ ~ _ ~ ~ ~ ~ ~
~ ~ æ~ ~ ~ ~ Q~ E
O Q O D ~g ~ L ~ C~ n D Q c~
~ Q t-- Q O 1-- Q
Q O ~: Q O P~ O C~ P) n G~ ~ Q
o c~ D' o Q ~ O Q D' ;~ ~ O W ~
c) n ~ n 1-- G~ n ~ D Q 'c C~ n 1--o Q ~ Q o U~
n G~ C O Q Dl ~ ~ C) ~ O
o D ~ O t-t ~ 8 ~ o 8 ~?
c~ n ~n ~ Q o O n n O Q U~ Q o ~n ~ ~ cn ~ G~ Q r ;E ~ Y~ o ~ ;~ o ~ c~
Q ~ n) I E~ :~ ~ ~ Q o ~D Q o ~-- c~

Q o 2 0Q D 1 D D ~ D 9 t~ Q Q ~n o ~ 8 ~ o ~ ~
0 ~ D~ ~ 0 D :~ D 0 C7 n n c~ ~ ~ D ~ Q 9 tn ;E~
æ
8 ~ c~ g ~ 0 ~ 0 o c~ C c~ Q ~s ~ O ~ C~ c~ c~
c~ n ~ o o ~ ~ o 0 ~ G~ n 5 O Q :~
1~ 3~ Y~ n 1-- c~ n ~ n 1--Q D L O ~ O ~ ~ C5 n ~
O G~ D ~Q O ~ O Q D 3:~ ~3 ~ n cs ~
Q 91- ~ O O ~ Q 01~ ~ 5- C~ n ~-~ ~ ~ 5 ~ 8 ~ 8 8 ~ ~ 8 ~

Q O I C.~ O C~ O G~ O ~ G~ n C~ y ~ g ~

WO 90/10081 PCI`/US90/0~
. .

) O
8 ~ ~ ~ ~ 3, P ~ G~ ' 3 3 ~ æ ' ' ;s ~ i ~ ~ ~ ~ ~ ~ ~ ' ~ ~ ~0 ~ _ ~ ~ ~ tn ~ ~ ~7 ~, ~
_ t~ ~r ~ C I G~ n o 3 ~
8~ 1 ~ ~
C
D
~ ~
~ ~ ~ ~ ~ ;~ ~ ~s . :~ o . ;:~ Q O S~, ~33 ~ ~ ~ I ~ ~ ~ , D
a 5~
c~ ~ o o ~ Q In c, ~ D c.~
~ ~ ~ 3 ~c~ 0 ~ ~ ~
~ 8 o ~ i ~ P. n ~ Q I o ~ ~ ~
C~; ~ ~ ,~ I o G~ n ~ ~ n G~ ~ In c~ ~ I c~ ~o WO 90/10081 PCI/US90/009~;

/ 7 :, e ~ I ~ I æ ~

~ ~ 1 ~ g ,, ~

n c~ u~ n rr~ o I ~ ~ c WO gO/10081 PCI/US90/00955 F` ' ~

e ~ tn,U~ ~ I
~ l e z:~ . ~ I

~ r ,.
7 9 .~

Table 3. Synthetic Oligonucleotide Primers COD #1292 aa 427 CysGluAsnGlyGlyPhe 5'ATCGGATCCTGCG~AA~CGGCGGCTcc pri~er/coding seqence BamHI

C~D ~1293 aa 350 CysPbeArgAlaAsnCys 5'GTGGGATCCTGCTTCAGAGCCAACTGC primer/coding seguence BamHI

COD t 1294 aa 390 CysAsnGlnThrAlaCys 5'CAGGGATCCTGCACCCAGACTGCCTGT prLmer/coding ~equence BamHI

COD ~1408 aa 339 LeuValAspGlyGluCys S'(CTGGTGGACGGCGAGTGT) coding sequence GACCACCTGCCGCTCACACACCGCCGGCGCCT primer sequence NotI

COD ~1409 aa 456 ArgHisIleGlyThrAspCys 5'(CGCCACATTGGCACCGA GT) coding sequence GCGGTGTAACCGTGGCTGACATCTCGCCGGCGTAG primer ~eguence NotI

COD ~1410 aa 381 HisGluProHisArgCys 5'(CACGAGCCGCACGGACGT) coding seguence GTGCTCGGCGTGTCCACGGTCTCGCCGGCGTT primer sequence NotI

WO90/10081 ~ PCT/US90/0~5 .. :
Table 3 ~ontd.
COD ~1411 aa 456 ArgHisIleGlyThrAspCysSTOP
5'(CGCCACATTGGCACCGACTGTTGA) coding ~equence GCGGTGTAACCGTGGCTGACAACTCGCCGGCGT primer ~equence NotI

COD ~1412 aa 416 AspAspGlyPheIleCys 5'(GACGACGGTTTCATCTGC) co~ing ~equence CTGCTGCCAAAAGGATACGCGCGGCCGGCTG pri~er sequence NotI

COD ~1433 aa 339 LeuValAspGlyGluCysSTOP
5'(CTGGTGGACGGCGAGTGTTGA) coding sequence GACCACCTGCCGCTCACAATCCGCCGGCGCCT primer sequence Not~

COD ~1434 aa 381 HisGluProHisArgCysSTOP
5'(CACGAGCCGCACGGACGTTGA) coding sequence GTGCTCGGCGTGTCCACGATCCGCCGGCGTT primer 6equence NotI

COD ~1435 aa 416 AspAspGlyPheIleCysSTOP
5'(GACGACGGTTTCATCTGCTGA) coding sequence CTGCTGCCAAAGGATACGATCCGCCGGCGGCTG pri~er ~equence NotI

COD ~1480 aa 462 CysAspSerGlyCysValAspSTOP
5'(TGTGACTCCGGCAAGGTGGACTGA) cDding ~equence ACACTGAGGCCGTTCCACCTGACTCTTAAGCT pri~er ~equence EcoRI

WO90/10081 PCT/~S90/00955 81 f` ~
. .
Table 3 contd.
COD tl479 aa 459 &lyThrAspCysAspSerSTOP
5'(GGCACCGACTGTGACTCCTGA) coding ~equence CCGTGGCTGACACTGAGGACTCTTAAGCAG
EcoRI

COD ~1478 aa 216 HisTrpAlaArgGluAlaPro 5'CCATGGCCACTGGGCC~GCGAGGCGCCG pri~er/coding ~equence ~alI

COD tl481 aa 490 ProAlaValGlyLeuValHisSerSTOP
5'(CCGGCCGTGGGGCTCGTGCATTCGTGA) coding sequence GGCCGGCACCCCGAGCACGTAAGCACTCGCCGGCGGTA primer seq.
NotI
Oligomer primers for 6 EGF-like domains COD~ 1289 - 3' aa 456 aa 462 ArgHisIleGlyT~rAspCys ~CGCCACATTGGCACCGACTGTAGATCTGGC) coding ~equence GCGGTGTAACCGTGGCTG~CATCTAGACCG priuer ~equence BglII
Oligomer primers for t-PA
COD ~1302 - 3' aa 524 aa 530 IleArgAspAsnMetArgPro (ATTCGTGACAACATGCGTCCGAGATCTGGA) coding ~equence TAAGCA GTTGTACGCAGGCTCTAGACCT primer sequence BglII
ATTCGTGACAACATGCGTCCGTGATCTGGA nati~e ~equence ~OD ~1305 - 5' aa 2 aa 11 AlaArgSerTyrGlnValIleCysArgAsp GCGGGATCCTACCAAGTGATCTGCAGAGAT pri~er/coding sequence BamHI
GCCAGATCTTACCAAGTGATCTGCAGAGAT native ~equence WO 90tlO081 PCI/US90/0095 82 s' ;~; ? '! '~ ~
.
Table 4. Amino Acid Composition Results Amino Acid ~?~ected ~ound ASX 30 28. 6 GLX 30 28 . 7 SER 13 12. 4 GLY 21 19.2 HIS 8 8.0 T~IR 10 9.9 AI~ 20 19 . 2 PR0 22 21. 5 VAL 11 9 . 5 MET 2 2 . 0 ILE 8 7.7 I~EU lO 11. 0 PHE 8 7 . 8 LYS 0 0.1 TYR 7 6 . 8 WOgO/1~81 PCT/US90/~ ~5 83 _ - . , ~ . .

Table 5. Activity of Thrombomodulin Analogs Following Drying.

Sam~le U/~l_ua/ml % Recoverv Dialyzed starting material 1661 8.88 PBS 601ub. 17277 92.4 104 PBS (NP-40) ~olub. 16635 88.9 lO0 NEM solub. 13568 72.5 82 NEM (~NP-40) solub. 1793995.9 108 84 ~~ .
Iable 6 Si~ Sff3us~es t.-pA S~gnal S~uence -32 ~a MetAspAlaMetLysArgGlyLeuCysCysValLeuLeuLeuCysGlyAlaValPhe ATGGATGCAATGAAGAGAGGGCTCTGCTGTGTGCTGCTGCTGTGTGGAGCAGTCTTC
TACCTACGTTACTTCTCTCCCGAGACGACACACGACGACGACACACCTCGTCAGAAG
-13 aa ValSerProSerGlu GluIleHisAlaArgPheArgArgGl~AlaAr~
GTTTCGCCCAGCCAGIINT~ON AIGAAATCCATGCCCGATTCAGAAGAGGAGCCAGA
CAAAGCGGGTCGGTC CTTTAGGTACGGGCTAAGTCTTCTCCTCGGTCT
+4 Ser TCC
AGG

Hypoder~ln A S1cn~1 Sequ~nce - p~Yl MetLeuLysPheValIleLeuLeuCysSerIleAlaTyrVal COD #1198 GATCATGCTCAAGTTTGTTATTTTATTGTGCAGTATTGCCTATGTT
BamHI TACGAGTTCA~ACAATAAAATAACACGTCATAACGGATACAA
PheGlyAlaValValProAr~SerProArg TTCGGTGCCGTCGTACCAAGATCTCCCCGG
AAGCCACGGCAGCATGGTTCTAGAGGGGCCCATGG COD ~1199 BglII KpnI

~ypodermin A S~gnal Sequence - p~Y2 MetLeuLysPheVallleLeuLeuCysSerIleAlaTyrVal COD #1482 GATCATGCTCAAGTTTGTTATTTTATTGTGCAGTATTGCCTATGTT
BamHI TACGAGTTCAAACAATAAAATAACACGTCATAACGGATACAA
PheGly ~pnI
TTCGGCCAGGTAC
AAGCCGGTC
BalI COD #1483 WO90/tO081 PCT~S90/00955 , Table 7. Transient Expression Plasmids Vector Pa~çnt TX Çene T~ Analo~
pTHRl pPA133 aa 227-462(-) 4t~227-462a pTHR7 pTHR5 aa 227-462(-) 4t/227-462b pTHR13 pTHR5 aa 227-462(+) 4t/227-462~+) pTHRl9 pTHR5 aa 350-462 4t/350-462 pTHR20 pTHR5 aa 227-462(-) + 4t~227-462:227-462 227-462(+) pTHR21 pT~R5 aa 227-497 4t/227-497 pTHR30 pTHRS aa 227-421 4t/227-421 pTHR31 pTHR5 aa 350-421 4t/350-421 pTHR32 pTHR5 aa 390-462 4t/390-462 pTHR33 pTHR5 aa 350-386 4t/350-386 pTHR34 pTHR5 aa 390-421 4t/390-421 pTHR35 pTHR5 aa 427-462 4t/427-462 pTHR43 pTHR5 aa 227-343 4t/227-343 pTHR44 pTHR5 aa 227-386 4t/227-386 TM Heteroloa pTHR4 pPA133 aa 227-462(-) 227-462:PA a pTHR6 pTHR5 aa 227-462(-) 227-462:PA b pTHR17 pTHR5 aa 227-462(+) PA:227-462 pTHR18 pTHR5 aa 350-462 PA:350-462 +/- indicates presence or absence of a stop codon at the 3' end of the TM gene WO90/1~81 PCT/~S90/00~5 .. ,~ ,~
Table 8. ~
Vector p~ent TM Gene ~M Analoa pTHR10 pSC716 aa 227-462(+) 6h/227-462 pTHRll pSC716 aa 227-462(-) + 227-462(+) 6h/(227-462)2 pTHR22 pSC716 aa 350-462 6h/350-462 pTER23 pSC716 aa 227-497 6h/227-497 pTHR45 pSC716 aa 350-421 6h/3S0-421 pTHR46 pSC716 aa 350-386 6h/350-386 pTXR47 pSC716 aa 390-421 6h/390-421 pTHR48 pSC716 aa 427-462 6h/427-462 pTMHY102 pHY2 aa 216-468 pTM~Y103 p~Y2 aa 216-464 TM Heterolo~
pTHR24 pSC716 aa 227-462(-) 6h/227-462:PA
pTHR25 pSC716 aa 227-462(+) PA:227-462 +/- indicates presence or absence of a stop codon at the 3' end of the TM gene wo 90/10081 PCI/US90/0095 Tab eFUSED TO THE C-TEPIMINUS OF
MATIJRE TISSUE PLASMINOGEN ACTIVATOR

cleavage sites lor M~ture t-PA
t-PA aa ~ 4 GlyAlaArgSerTyrGlnValIleCys GGAGCCAGATCCTACCAAGTGATCTGC
CCSCGGTCTAGGATGGTTCACTAGACG
B~l lU8am Hl ArgAspGluLysThrGlnMetIleTyrGlnGlnHisGlnSerTrpLeuArgProValLeu AGAGATGA~AAAACGCAGATGATATACCAGC~ACATCAGTCATGGCTGCGCCCTGTGCTC
TCTCTACTTTTTTGCGTCTACTATATGGTCGTTGTAGTCAGTACCGACGCGGGACACGAG

Ar~SerAsnArgValGluTyrCysTrpCysAsnSerGlyArgAlaGlnCysHisSerVal AGAAGCAACCGGGTGGAATATTGCTGGTGCAACAGTGGCAGGGCACAGTGCCACTCAGTG
TCTTCGTTGGCCCACCTTATAACGACCACGTTGTCACCGTCCCGTGTCACGGTGAGTCAC
~0 ProValLysSerCysSerGluPr~rgCysPheAsnGlyGlyThrCysGlnGlnAlaLeu CCTGTCAA~AGTTGCAGCGAGCCAAGGTGTTTCAACGGGGGCACCTGCCAGCAGGCCCTG
GGACAGTTTTCAACGTCGCTCGGTTCCACAAAGTTGCCCCCGTGGAC~GTCGTCCGGGAC

TyrPheSerAspPheValCysGlnCysProGluGly~heAlaGlyLysCysCysGluIle TACTTCTCAGATTTCGTGTGCCAGTGCCCCGAAGGATTTGCTGGGAAGTGCTGTGA~ATA
ATGAAGAGTCTAAAGCACACGGTCACGGGGCTTCCTAAACGACCCTTCACGACACTTTAT

AspThrArgAlaThrCysTyrGluAspGlnGlyIleSerTyrArgGlyThrTrpSerT~r GATACCAGGGCCACGTGCTACGAGGACCAGGGCATCAGCTACAGGGGCACGTGGAGCACA
CTATGGTCCCGGTGCACGATGCTCCTGGTCCCGTAGTCGATGTCCCCGTGCACCTCGSGT

WO ~/1~81 PCT/US90/~K~

r.. , . : ,' ,,i Table 9 C~nt~.
AlaGluserGlyAlaGlucysThr~snTrpAsnserserAlaLcuAlaGlnLysproTyr GcGGAGAGTGGcGccGAGTGcAccAAcTGGA~QGcAGcGc:;TTGGt:CCAGAAGCCCTAC
CGccscTcAccGcGGcTcAcGTGGTTGAccTTGTcGTcGcGcAAccGGGTcTscGG~ATG

SerGlyArgArgproAspAlaIleArgLeuGlyLeuGlyAsnHlsAsnTyrcysArgAsrl AGCGGGCGGAGGCCAG~CGCCATCAGGCTGGGCCTGGGGAACCACA~CTACTGCAGAAAC
TcGcccGccTccGG$cTGcGGTAGTccGAcccGGAccccT$GGTGTTGATGAcGTcTTTG
1 !;0 ProAspArgAspserLysproTrpcysTyrvalpheLysAl~GlyLysTyrserserGlu CcAGATcGAGAcTc~AAGcccTGGTGcTAcGTcTTTAAGGcGGGGAAGTAcAGcTcAGAG
GGTCTAGCTCTGAGTTTCGGGACCACGATGCAGAAATTCCGCCCCTTCATGTCGAGTCTC

PheCysSerThrProAlaCysSerGluGlyAsnSerAspCysTyrPheGlyAsnGlySer TTC~GCAGCACCCCTGCCTGCTCTGAGGGAAACAGTGACTGCTACTTTGGGAATGGGTCA
AAGACGTCGTGGGGACGGACGAGACTCCCTTTGTCACTGACGATGAAACCCTTACCCAGT

AlaTyrArgGlyThrGlnSerLeuT~rGluSerGlyAlaSerCysLeuProTrpAsnSer GCCTACCGTGGCACGCAGAGCCTCACCGAGTCGGGTGCCTCCTGCCTCCCGTGGAATTCC
CGGATGGCACCGTGCGTCTCGGAGTGGCTCAGCCCACGGAGGACGGAGGGCACCTTAAGG

Met~leLeuIleGlyLysValTyrThrAlaGlnAsnProSerAlaGlnAlaLeuGlyLeu ATGATCCTGATAGGCAAGGTTTACACAGCACAGAACCCCAGTGCCCAGGCACTGGGCCTG
TACTAGGACTATCCGTTCCAAATGTGTCGTGTCTTGGGGTCACGGGTCCGTGACCCGGAC

GlyLysHisAsnTyrCysArgAsnProAspGlyAspAlaLysProTrpCys~isValLeu GGCAAACATAATTACTGCCGGAATCCTGATGGGGATGCCAAGCCCTGGTGCCACGTGCTG
CCGTTTGTATTAATGACGGCCTTAGGACTACCCCTACGGTTCGGGAC QCGGTGCACGAC

LysAsnArgArgLeuThrTrpGluTyrCysAspV~lProSerCysSerThrCysGlyLeu AAGAACCGCAGGCTGACGTGGGAGTACTGTGATGTGCCCTCCTGCTCCACCTGCGGCCTG
TTCTTGGCGTCCGACTGCACCCTCATGACACTACACGGGAGGACGAGGTGGACGCCGGAC

WO90/1~81 PCT/US90/~ ~5 270 Table 9 Ccnbd.
ArgGlnTyrserGlnproGlnpheArgIleLysGlyGlyLeupheAlaAspIleAlaser AGAcAGTAcAGccp~GccTcAGTTTcGcATcAAAGGAGGGcTc5TcGccG~cATcGccTcc TcTGTcATGTcGGTcGGAGTcA~AGcGTAGTTTccTcccGAGAAGcGGcTGTAGcGGAGG

HisProTrpGlnAlaAlA Ile~heAlaLys~,isArgArgSerProGlyGluArgPhe~eu CAccccTGGcAGGcTGccATcTTTGccAAGcAcAGGAGGTcGcccGGAGAGcGGTTccTG
GTGGGGAccGTccGAcGGTAGAp~AcGGTTcGTGTccTccAGcGGGccTcTcGccAAGGAc CysGlyGlyIle~euIleSerSerCysTrpIleLeuSerAlaAla~sCysPheGlnGlu TGcGGGGGcATAcTcATcAGcTccTGcTGGATTcTcTcTGccGcccAcTGc5TccAGGAG
ACGCCCCCGTATGAGTAGTCGAGGACGACCTAAGAGAGACGGCGGGTGACGAAGGTCCTC

ArgPheProProHisHisLeuThrValIleLeuGlyArgThrTyrArgValValProGly AGGTTTCCGCCCCACCACCTGACGGTGATCTTGGGCAGAACATACCGGGTGGTCCCTGGC
TccA~AGGcGGGGTGGTGGAcTGccAcTAGAAcccGTcTTGTATGGcccAccAGGGAccG

GluGluGluGlnLysPheGluValGluLysTyrIleValHisLysGluPheAspAspAsp GAGGAGGAGCAGAAATTTGAAGTCGA~AAATACATTGTCCATAAGGAATTCGATGATGAC
CTCCTCCTCGTCTTTAAACTTCAGCTTTTTATGTAACAGGTATTCCTTAAGCTACTACTG

ThrTyrAspAsnAspIleAlaLeuLeuGlnLeuLysSerAspSerSerArgCysAlaGln ACTTACGACAATGACATTGCGCTGCTGCAGCTGA~ATCGGATTCGTCCCGCTGTGCCCAG

GluSerSerValValArgThrValCysLeuProProAlaAspLeuGlnLeuProAspTrp GAGAGCAGCGTGGTCCGCACTGTGTGCCTTCCCCCGGCGGACCTGCAGCTGCCGGACTGG
CTCTCGTCGCACCAGGCGTGACACACGGAAGGGGGCCGCCTGGACGTCGACGGCCTGACC

ThrGluCysGluLeuSerGlyTyrGlyLys~lsGluAlaLeuSerProPheTyrSerGlu ACGGAGTGTGAGCTCTCCGGCTACGGCAAGCATGAGGCCTTGTCTCCSTTCTATTCGGAG
TGccTcAcAcTcGAGAGGccGATGccGTTcGTAcTccGGAAcAGAGGAAAGATAAGccTc 430 Table 9 Conbd.
ArgLeuLysGluAlaHisValArgLeu~yrProSerSerArgCysT~rSerGlnHisLeu CGGCTGAAGGAGGCTCATGTCAGACTGTACCCATCCAGCCGCTGCACATCACAACATTTA
GCCGACTTCCTCCGAGTACAGTCTGACATGGGTAGGTCGGCGACGTGTAGTGTTGTAAAT
4~0 LeuAsnArgThrValThrAspAsnMetleuCysAlaGlyAspThrArgSerGlyGlyPro CTTAACAGAACAGTCACCGACAACATG~TGTGTGC~GGAGACACTCGGAGCGGCGGGCCC
G~ATTGTCTTGTCAGTGGCTGTTGTACGACACACGACCTCTGTGAGCCTCGCCGCCCGGG

GlnAlaAsnLeuHlsAspAlaCysGlnGlyAspSerGlyGlyProLeuValCysLeuAsn CAGGCAAACTTGCACGACGCCTGCCAGGGCGATTCGGGAGGCCCCCTGGTGTGTCTGAAC
GTCCGTTTGAACGTGCTGCGGACGGTCCCGCTAAGCCCTCCGGGGGACCACACAGACTTG

AspGlyArgMetThrLeuValGlyIleIleSerTrpGlyLeuGlyCysGlyGlnLysAsp CTACCGGCGTACTGAAACCACCCGTAGTAGTCGACCCCGGACCCGACACCTGTCTTCCTA

ValProGlyValTyrThrLysValT~rAsnTyrLeuAspTrpIleArgAspAsnMetArg GTCCCGGGTGTGTACACAAAGGTTACCAACTACCTAGACTGGATTCGTGACAACATGCGA
CAGGGCCCACACATGTGTTTCCAATGGTTGATGGATCTGACCTAAGCACTGTTGTACGCT
t-PA T M
aa s~o aa 227 ProARGSERCysSerValGluAsnGlyGlyCysGluHisAlaCysAsnAlaIleProGly CCGAGATCCTGCAGCGTGGAGAACGGCGGCTGCGAGCACGCGTGCAATGCGATCGCTGGG
GGCTCTAGGACGTCGCACCTCTTGCCGCCGAC:GCTCGTGCGCACGTTACGCTAGGGACCC
Bgl IllBam Hl AlaProArgCysGlnCysProAlaGly~laAlaLeuGlnAlaAspGlyArgSerCysThr GCTCCCCGCTGCCAGTGCCCAGCCGGCGCCGCCCTGCAGGCAGACGGGCGCTCCTGCACC
CGAGGGGCGACGGTCACGGGTCGGCCGCGGCGGGACGTCCGTCTGCCCGCGAGGACGTGG

AlaSerAlaThrGlnSerCysAsnA~pLeuCysGluHlsPheCy~ValProAsnProAsp GCATCCGCGACGCAGTCCTGCAACGACCTCTGCGAGCACTTCTGCGTTCCCAACCCCGAC
CGTAGGCGCTGCGTCAGGACGTTGCTGGAGACGCTCGTGAAGACGCAAGGGTTGGGGCTG

Wo9O/t~81 PCTtUS90/0~5 91 1, i . -.:. ~. , ;;
284 Table 9 Conb~.
GlnProGlySerTyrSerCy~MetCysGluThrGlyTyrArgLeuAlaAlaAspGln~is CAGCCGGGCTCCTACTCGTGCATGTGCGAGACCGGCTACCGGCTGGCGGCCGACCAACAC
GTCGGCCCGAGGATGAGCACGTACACGCTCTGGCCGATGGCCGACCGCCGGCTGGTTGTG
~04 ArgCyQGluAspValAspAspCyQIleLeuGluproSerProCysProGlnArgCysVal CGGTGCGAGGACGTGGATGACTGCATACTGGAGCCCAGTCCGTGTCCGCAGCGCTGTGTC
GCCACGCTCCTGCACCTACTGACGTATGACCTCGGGTCAGGCACAGGCGTCGCGACACAG

AsnThrGlnGlyGlyPheGluCysHl~CysTyrProAQnTyrAspLeuY~lAspGlyGlu AACACACAGGGTGGCTTCGAGTGCCACTGCTACCCTAACTACGACCTGGTGGACGGCGAG
TTGTGTGTCCCACCGAAGCTCACGGTGACGATGGGATTGATGCTGGACCACCTGCCGCTC

CysValGluProValAspProCy3PheArgAlaAsnCysGluTyrGlnCysGlnProLeu TGTGTGGAGCCCGTGGACCCGTGCTTCAGAGCCAACTGCGAGTACCAGTGCCAGCCCCTG
ACACACCTCGGGCACCTGGGCACGAAGTCTCGGTTGACGCTCATGGTCACGGTCGGGGAC

AsnGln~hrSerTyrLeuCysValCy~AlaGluGlyPheAlaProIleProHisGluPro AACCAAACTAGCTACCTCTGCGTCTGCGCCGAGGGCTTCGCGCCCATTCCCCACGAGCCG
TTGGTTTGATCGATGGAGACGCAGACGCGGCTCCCGAAGCGCGGGTAAGGGGTGCTCGGC

HisArgCysGlnMetPheCysAsnGlnThrAlaCysProAlaAspCysAspProAsnT~r CACAGGTGCCAGATGTTTTGCAACCAGACTGCCTGTCCAGCCGACTGCGACCCCAACACC
GTGTCCACGGTCTACAAAACGTTGGTCTGACGGACAGGTCGGCTGACGCTGGGGTTGTGG

GlnAlaSerCysGluCysProGluGlyTyrIleLeuAspAspGlyPheIleCysThrAsp CAGGCTAGCTGTGAGTGCCCTGAAGGCTACATCCTGGACGACGGTTTCATCTGCACGGAC
GTCCGATCGACACTCACGGGACTTCCGATGTAGGACCTGCTGCCAAAGTAGACGTGCCTG

IleAspGluCysGluAsnGlyGlyPheCysSerGlyV~lCysXisAsnLeuProGlyThr ATCGACGAGTGCGAAAACGGCGGCTTCTGCTCCGGGGTG~GCCACAACC5CCCCGGTACC
TAGCTGCTCACGCTTTTGCCGCCGA~GACGAGGCCCCACACGGTGTTGGAGGGGCCATGG

92 ç~
Table 9 Ccntd. TM-aa 4 62 PheGlucysIlecysGlyproAspserAla~euAlaArgHisIleGlyThr GcccGccAcATTGGcAccGAcTGTTTcGAGTGcATcTGcGGGcccGAcTcG
CGGGcGGTGTAAccGTGGcTGAcAAAGcTcAcGTAGAcGcccGGGcTGAGc AspCysQP
GCCCTTTGAGGATCT
CGGGAAACTCCTAGA
Bam HU~I 11

Claims (61)

WHAT IS CLAIMED IS:

1. A sequence of nucleic acid bases encoding a peptide selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

2. A polynucleotide sequence encoding the peptides of claim 1 where the sequence is selected from the group consisting of:
a) CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCCGGCAAGGTGGAC;
b) CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCC;
c) GGAGCCAGATCC-Z; and d) GCCGTCGTACCAAGATCC-Z;
where X represents the nucleic acid bases numbered 879 to 1586, and Z represents the nucleic acid bases selected from the group numbered 879 to 1586, 1251 to 1586 or 879 to 1690 wherein all numbers refer to bases as provided in Table 2.

3. A polynucleotide sequence according to claim 2 having a sequence of CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCCGGCAAGGTGGAC.

4. A polynucleotide sequence according to claim 2 having a sequence of CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCC.

5. A polynucleotide sequence according to claim 2 having a sequence of GGAGCCAGATCC-Z.

6. A polynucleotide sequence according to claim 2 having a sequence of GCCGTCGTACCAAGATCC-Z.

7. A recombinant vector comprising a sequence of nucleic acids encoding a peptide selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

8. A vector of claim 7 where the sequence of nucleic acid is selected from the group consisting of:
a) CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCCGGCAAGGTGGAC;
b) CACTGGGCCAGGGAGGCGCCGGGCGCTTGGGAC-X-GACTCC;
c) GGAGCCAGATCC-Z; and d) GCCGTCGTACCAAGATCC-Z;
where X represents the nucleic acid bases numbered 879 to 1586, and Z represents th nucleic acid bases selected from the group numbered 879 to 1586, 1251 to 1586 or 879 to 1690 wherein all numbers refer to bases as provided in Table 2.

9. A substantially pure protein composition selected from a peptide of the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - zsp.ser:
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

10. A peptide of claim 9 having an amino acid sequence of:
his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. - Y -asp.ser.gly.lys.val.asp.

11. A peptide of claim 9 having an amino acid sequence of his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. - Y -asp.ser.

12. A peptide of claim 9 having an amino acid sequence of gly.ala.arg.ser. - Q.

13. A peptide of claim 9 having an amino acid sequence of ala.val.val.pro.arg.ser - Q.

14. A composition comprising a peptide of claim 9 wherein said peptide is a species of said peptide which exhibits about the expected molecular weight for said peptide under non-reducing chromatography.

15. A composition of claim 9, wherein the peptide is chemically linked to a fibrinolytic enzyme.

16. A composition of claim 15, wherein said chemical linkage arises from a covalent bond between the enzyme nnd an amino acid residue within the peptide.

17. A composition of claim 15, wherein said enzyme is a streptokinase molecule bound to a plasminogen molecule and wherein the peptide is bound to the active site of the plasminogen molecule.

18. A composition of claim 15 wherein the fibrinolytic enzyme is t-PA.

19. A composition of claim 9 wherein the composition is in a dry and salt-free form.

20. A pharmaceutical composition having antithrombotic activity comprising a sterile preparation of a unit dose of thrombomodulin-like protein free of contaminants of human origin and having an amino acid sequence selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462: and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acid as provided ia Table 2.

21. A pharmaceutical composition of claim 20 having an amino acid sequence of his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. - Y -asp.ser.gly.lys.val.asp.

22. A pharmaceutical composition of claim 20 having an amino acid sequence of his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. - Y -asp.ser.

23. A pharmaceutical composition of claim 20 having an amino acid sequence of gly.ala.arg.ser. - Q.

24. A pharmaceutical composition of claim 20 having an amino acid sequence of ala.val.val.pro.arg.ser. - Q.

25. A method of controlling thrombotic activity in a mammal by administering an effective amount of a sterile aqueous solution of a thrombomodulin-like protein, free of contaminants of human origin and having an amino acid sequence selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids 6elected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

26. A composition comprising a biocompatible polymer haviny a surface wherein the surface has bound thereto peptides selected from the group consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

27. A method for inhibiting blood clotting induced by polymers implanted into a living mammal, said method comprises the bonding of a thrombomodulin-like protein to the polymer and implanting the polymer into the mammal, said protein is selected from the group of peptides consisting of:
a) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser.gly.lys.val.asp;
b) his.trp.ala.arg.glu.ala.pro.gly.ala.trp.asp. -Y - asp.ser;
c) gly.ala.arg.ser. - Q; and d) ala.val.val.pro.arg.ser. - Q;
wherein Y is the sequence of amino acids ranging from amino acid 227 to amino acid 462; and Q is the sequence of amino acids selected from the group consisting of amino acid 227 to amino acid 462, amino acid 350 to amino acid 462 and amino acid 227 to amino acid 497 wherein all numbers refer to amino acids as provided in Table 2.

28. An isolated DNA fragment containing a DNA sequence coding for multi-functional soluble human thrombomodulin (TM) analog, said analog consisting essentially of amino acid residues 350 - 462 of native TM and a targeting component.

29. An isolated DNA fragment encoding a bifunctional TM analog according to Claim 28, consisting essentially of amino acid residues:
, ;
, ;
, ; and , .

30. A DNA fragment according to Claim 28 wherein said targeting component comprises a polypeptide sequence which imparts fibrinolytic activity to the bifunctional TM analog.

31. An isolated DNA fragment according to Claim 28, wherein said targeting component comprises amino acid residues 4 - 530 of human tissue plasminogen activator as depicted in Table 9.

32. A DNA sequence according to Claim 28 operably linked to an expression control sequence.

33. A cell transfected with 2 DNA sequence according to Claim 32.

34. A method of producing a bifunctional soluble TM analog comprising culturing a host transformed with at least one recombinant DNA segment having an expression control sequence operably linked to a DNA sequence coding for a fusion protein comprising:
a first amino acid sequence comprising a thrombomodulin fragment without a stop transfer domain;
and a second amino acid sequence comprising a targeting component;
wherein the bifunctional TM analog is secreted from the host.

35. A method according to Claim 34 wherein said first amino acid sequence is selected from the group consisting of:
, ;
, ;
, ; and , .

36. A method according to Claim 35, wherein the bifunctional analog has fibrinolytic activity.

37. A method according to Claim 36, where said targeting component consists essentially of amino acids 4 -530 of human t-PA.

38. A method of suppressing coagulation in a patient comprising administering to the patient an amount of soluble TM analog effective to suppress coagulation, wherein the analog comprises TM amino acids selected from the group consisting of:
, ;
, ;

, ; and , .

39. A method of treating a patient suffering form acute myocardial infarction caused by a thrombus, said method comprising administering to the patient an amount of bifunctional soluble TM analog effective to suppress further coagulation and lyse the thrombus, wherein the bifunctional TM analog comprises TM amino acids selected from the group consisting of:
, ;
, ;
, ; and , .
wherein said TM amino acid sequence is fused at the C-terminus or the N-terminus to amino acids 4-530 of human t-PA.

40. A method according to Claim 39, wherein the TM analog retains substantially the same or greater affinity for thrombin as native thrombomdulin.

41. A bifunctional TM analog capable of binding thrombin with substantially the ame affinity or greater as native thrombomodulin, said analog comprising native thrombomodulin without a functional stop transfer domain, wherein said bifunctional analog is characterized by the ability to bind fibrin and cleave plasminogen to plasmin.

42. A TM analog according to Claim 41, wherein said analog consists essentially of the amino acid sequence set forth in Drawing 9.

43. A TM analog according to Claim 41, wherein said analog consists essentially of the signal sequence of t-PA covalently attached to the amino terminus of the 6 EGF-like domain of thrombomodulin.

44. A TM analog according to Claim 41, wherein said analog consists essentially of the 6 EGF-like domain of thrombomodulin covalently attached to the amino terminus of .
t-PA.

45. A process for making soluble thrombomodulin analogs, said method comprising the steps of:
culturing a host cell transformed with a DNA
sequence encoding the analog; and collecting analogs secreted by the host cell;
wherein the analogs lack a stop transfer sequence functional in the host cell.

46. A method according to Claim 45, wherein the host cell is a eukaryotic cell.

47. An isolated DNA fragment containing a DNA sequence coding for a soluble human thrombomodulin analog, capable of being secreted by a host cell transfected with said DNA fragment, wherein said DNA
sequence encodes a polypeptide, comprising amino acid resides 390-462 of native TM.

48. An isolated DNA fragment according to Claim 47 wherein said polypeptide is selected from the group consisting of:
, ;
, :
, ; and , .

49. A DNA sequence according to Claim 47 operably linked to an expression control sequence.

50. A cell transfected with a DNA sequence according to Claim 49.

51. A method of producing a soluble thrombomodulin (TM) analog comprising culturing a host transformed with at least one recombinant DNA segment, the segment comprising an expression control sequence operably linked to a DNA sequence coding for TM analog without a functional stop transfer domain, wherein the TM analog is secreted from the host and selected from the group consisting of:
, ;
, ;
, ; and , .

52. A method according to Claim 51, wherein the TM analog DNA sequence comprises a signal sequence.

53. A method according to Claim 51, wherein the host is a eukaryotic cell line.

54. A method of suppressing coagulation in a patient comprising administering to the patient an amount of soluble thrombomodulin (TN) analog effective to suppress coagulation, wherein the TM analog is selected from the group consisting of:
, ;
, ;
, ; and , .

55. A method according to Claim 54 wherein the TM analog retains substantially the same or increased affinity to thrombin as native thrombomodulin.

56. A soluble thrombomodulin analog capable of binding thrombin with substantially the same affinity as native thrombomodulin, said analog comprising native thrombomodulin without a functional stop transfer domain.

57. An analog according to Claim 56, wherein said analog lacks any amino acid from the stop transfer domain of native thrombomodulin.

58. A method of activating anticoagulant activities of thrombin comprising complexing the thrombin with a soluble thrombomodulin analog of Claim 56.

59. A method of treating a patient requiring anticoagulant therapy, said treatment comprising administering a therapeutically effective dose of the soluble TM analog of Claim 56.

60. A method of mitigating the risk of reocclusion in a patient recovering from myocardial infarction comprising administering a therapeutically effective dose of the soluble TM analog of Claim 56.

61. A method of prophylactically treating a patient at risk from suffering deep vein thrombus said method comprising administering to the patient a therapeutically effective dose of the soluble TM analog of Claim 56.